Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Use and effects of chemical agents on Australian personnel in Vietnam - Royal Commission (Hon. Mr Justice P. Evatt) - Final report, dated 31 July 1985 - Report - Volume 4 - Cancer

Download PDF Download PDF

The Parliament of the Commonwealth of Australia


Commissioner: The Hon. Mr Justice P. Evatt, DSC , LLB

Final Report—July 1985

Volume 4: Cancer

Presented 22 August 1985 Ordered to be printed 19 September 1985

Parliamentary Paper No. 291/1985



Commissioner: The Hon. Mr Justice Phillip Evatt DSC, LLB.

A Judge of the Federal Court of Australia


July 1985


Australian Government Publishing Service Canberra 1985

© Commonwealth of Australia 1985

ISBN 0 644 04339 3 ISBN 0 644 04343 1 Set of Volumes Report Volume Four

Printed by Canberra Publishing and Printing Co., Fyshwick, A.C.T.


Commissioner: The Hon. Mr Justice Phillip Evatt DSC

G.P.O. Box 4842 Sydney, N.S.W. 2001 Telephone: (02) 239 6222

Your Excellency,

Secretary: Mr B.D. Meade

31 July 1985

In accordance with Letters Patent issued to me on 13 May 1983, 27 June 1984, 3 August 1984 and 23 April 1985, I have the honour to present to you the Final Report of my inquiry.

I believe that the Report complies with those Letters Patent and that my task is therefore completed.

Yours sincerely


His Excellency the Right Honourable Sir Ninian Stephen, A.K., G.C.M.G., G.C.V.O., K.B.E. Governor-General and Commander-in-chief Government House CANBERRA A.C.T. 2600









(1) The Present Era 15

(2) Summary 18

(3) Research Patterns in Chemical Carcinogenesis 1945 to Date 19

(4) Current Views on Criteria for Evidence of Carcinogenicity of Chemicals 20


(a) AVHS Mortality Report 25

Personnel 26

Study Design 27

Cancer Analysis 28

Conclusion 32

D.l (b) Evidence from Australian Vietnam Veterans 33 Conclusion 36

D.l (c) Incidence of Lymphoma Known to DVA 36

D.l (d) Incidence of Soft Tissue Sarcoma Known to DVA 40

Conclusion 45

D . 2 United States Human Data 46

(a) Ranch Hand 1 (1983) 48

Conclusion 51

(b) Ranch Hand II (1984) 51

Conclusion 61

(c) Veterans' Administration(U.S.A.) 62

(d) Greenwald et al 63

Conclusion 64

(e) Kogan et al 65

Conclusion 69

Summary re U.S. Personnel 70

E. OCCUPATIONAL EVIDENCE - APPLICATORS E.1 Victoria - Departmental Study (1985) 71

Conclusion 74


E .2 Occupational - Applicators in New Zealand - Smith et al (1982, 1983) 75

Conclusion 81

E .3 Finland - Riihimaki et al (1982, 1983) 81

Conclusion 88

E . 4 Sweden - Hogs-tedt and Wester lund (1980) 88

Conclusion 90

E . 5 Axelson and Hardell (i) Axelson 90

(ii) Hardell (a) North Sweden Study 92

(b) South Sweden Study 93

(c) Hardell's Malignant Lymphoma Study 94 (iii) Axelson's Re-Analysis of his 1973 Study & Up-Date Study 95

(a) Critical Examination of Axelson Studies 100

(i) Epidemiology and its Purpose 101 (ii) Lack of Expertise 105

(iii) Inaccuracy of the Data 106

(iv) Weakness of the Conclusions 120 (v) Differences between Axelson and Hardell 124

Conclusion 126

(vi) Critical Examination of Hardell Studies 126 (a) The North Sweden Study 128

(b) The South Sweden Study 156

(c) The Malignant Lymphoma Study 160

(d) The Nasopharyngeal Cancer Study 166

(e) The Colon Cancer Study 169

(f) Effect of Interview 169

(g) The Axelson Technique 178

(h) Conclusion 180


F.l Nitro W.Va. USA 185

F.l (a) Zack and Suskind 185

(b) Zack and Gaffey 187

(c) Moses et al 190

(d) Suskind and Hertzberg 191

Conclusion 194

F .2 Dow, Midland II. USA 194

F .2 (a) Cook et al 195

Conclusion 196

F . 2 (b) Ott et al 196

Conclusion 198

Summary 198


F .3 Soft Tissue Sarcoma Issue F .3 (a) Honchar and Halperin 199

(b) Cook and Johnson, Moses and Selikoff and Johnson et al 203

(c) Fingerhut et al 204

Conclusion 208

F .4 BASF Ludwigshafen, Germany 208

F .4 (a) Thiess and Frentzel-Beyme 210

Conclusion 214

F.S Seveso, Italy 214

Summary 219

F . 6 Table of Comparative Exposures 219

G. ANIMAL AND IN VITRO DATA ANALYSES RE 2,4-D. 2,4.5-T and TCDD G.l Introduction 222

G .2 Mutagenicity (i) 2,4-D 223

Conclusion 234

(ii) 2,4.5-T 234

Conclusion 242

(iii) TCDD 242

Conclusion 248

G . 3 Carcinogenicity 249

(i) 2,4-D 250

Conclusion 256

(ii) 2,4,5-T 256

Conclusion 260

(iii) TCDD 260

Conclusions 271


H.l The Herbicides 273

Diguat 273

Paraguat 275

Picloram 278

Dalapon 289

Diuron 290

Monuron 292

Bromacil (Hyvar-X) 297

Cacodylic Acid 299

Distillate-Creosote 303

Borate-Chlorate 305

H .2 The Insecticides DDT 307

Malathion 323

Dieldr in 335

Chlordane 350

Lindane 362

Diazinon 369


Pyrethrins 371

Diethyl-m-Toluamide (DEET) 372

Dimethyl-Phthalate (DMP) 372

Di-n-Butyl Phthalate (DBP) 373

Η. 3 Drugs Dapsone 375

Chloroquine, Primaquine & Paludrin 383

Η. 4 Solvents Dimethyl Sulphoxide (DMSO) 385

Distillate 386

Kerosene 387











Introduction Standard of Proof Ascertainment of Claims Exposure



Toxicology and Safe Doses Health Effects General


VII Health Effects, Reproductive Outcomes and Birth Anomalies


VIII Health Effects, Cancer


IX Health Effects, Mental



Mortality Class Action Status of W A A Interim Report and S.47


XIV Benefits and Treatment


XV Conclusions and Recommendations Epilogue




Cancer A disorder of uncontrolled cell growth, also called malignant neoplasm.

Carcinogen An agent which causes an increase in cancers, or malignant neoplasms.

Carcinogenesis The multi-stage process that converts normal cells to neoplastic (cancerous) cells and forms cancers from these cells by clonal proliferation.

Cell Transformation An alteration in cellular behaviour and phenotype.

Chromosomal Aberration A structural or numerical alteration to a chromosome.

Clastogen An agent that produces structural chromosomal aberrations.

Co-carcinogen An agent operating in conjunction with a carcinogen to augment malignant neoplasm induction.

Complete Carcinogen An agent capable of producing cancers without the assistance of any other exogenous agent.

DNA - Deoxyribonucleic Acid The chemical form in which genetic information is stored in the cell. Each normal human cell contains 23 pairs of chromosomes each of which contains a long DNA molecule as a "blueprint" for human development.

Gene Amplification A process resulting in the presence of multiple copies of a gene that is normally present as a single copy.

Genotoxic A general expression to cover toxic effects to genetic material in germinal or somatic cells.


Initiation That process whereby a normal cell or cell undergoes the change which begins the process of malignant neoplasm production.

In vitro That l s , in a petri dish or a test tube.

Latency period The period elapsing between exposure to a carcinogen and the growth of the tumour to the point where it becomes clinically evident.

Mutation Any detectable heritable change in the genetic material not caused by normal segregation or genetic recombination. The term includes changes in DNA

composition, sequence arrangement, structure and position that result in permanent alteration of the gene including chromosomal alterations.

Numerical Chromosome Aberrations The loss or gain of one or more chromosomes or

chromosome pairs.

Oncogene A cellular oncogene is a gene which, if altered either in structure or activity, may contribute to malignant transformation. A viral oncogene is an oncogene derived from a cell but carried by a virus, which may

contribute to malignant transformation in mammalian cells.

Point Mutation A mutation where no change in chromosome structure is detected by microscopy: more recently, a mutation involving a single gene or a single base within the


Progression Sequential development of atypical cells.

Promotion That process of inter-action with an initiated cell whereby the process of carcinogenesis is advanced or hurried by an agent which cannot by itself initiate malignant neoplasm production.

Sister Chromatid Exchange An equal symmetrical exchange between the chromatids of a single chromosome.


Structural Chromosomal Aberrations Breaks or re-arrangements within or between


Throughout this Chapter repeated reference is made to VVAA1s initial submission, to its submission related specifically to toxicology and to its final submission. Each submission has been exhibited and is numbered 1040,

1878 and 1879 respectively.





A repeated claim made by those purporting to represent

members of VVAA for some years leading up to the

commencement of this Commission was that Agent Orange had

caused and was causing (inter alia) cancer among

Australian Vietnam veterans. That this is so is reflected

in the Letters Patent, particularly paragraphs (h) and (i)

which read:

AND without restricting the scope of your inquiry. We direct you to inquire particularly into the following matters, namely:-(h) the toxic properties, resulting from use

separately or in combination, in relation to humans of each of the chemical agents used, with particular emphasis on, but not limited to, any direct or indirect carcinogenic, mutagenic, teratogenic or neurotoxic properties and the extent and duration of exposure necessary to affect humans. ( i )

(i) evidence relating to the effects of any exposure to chemical agents on the mental and physical health and well-being of

Australian personnel and any effects on the mental and physical health and well-being of their spouses.


Cancer is an emotive topic. Many Australians will die as a

result of it, perhaps as many as one in four. It is clear

that it is not a disease of the old only, although as life

expectancy increases (as a result of the control of other

causes of death) the raw numbers of those dying as a

result of cancer increase although age-adjusted incidence

does not when tobacco-caused cancers are omitted.1

Using 31 December 1981 as a convenient cut-off point, the

1972 Australian statistics were such that it would then

have been expected that from amongst the total Vietnam

veteran cohort which had returned to Australia between

1965 and 1972, some thirty veterans would die from cancer

before the end of 1981. These thirty deaths, of course,

would have been expected from that group of Australians

without the operation of any additional cause to those

governing the then statistical rate. Accordingly, if

Agent Orange, or indeed, any other chemical agent used in

Vietnam, caused cancers amongst Vietnam veterans leading

to their deaths, then increases in the number of expected

deaths would be demonstrated. This aspect is discussed

later herein.

VI11 -2


The claims made by VVAA in relation to cancer, as in other

areas, lack definition.

In its initial submission lodged with the Commission in

November 1983 W A A made the following claims:

(i) 2,4-D is carcinogenic;

( ii) 2,4,5-T is carcinogenic;

(iii) TCDD is carcinogenic (at p 54);

(iv) W A A does not have the resources to carry out

morbidity studies. However, lists are being

prepared by each State from questionnaires (at p


(v) The lists show the veterans' complaints ... the

lists are prepared from questionnaires supplied

to the W A A (at p 61);

(vi) W A A 1s submission is that cancers are unduly

prevalent in causes of deaths of Vietnam


veterans. In particular, rare forms of cancer

are manifest (at pp 62 and 63).

It is clear then that as long ago as November 1983 W A A

had material in hand relating to members' health. From

"Debrief" (the official publication of WAA) of August

1983 it is known that the questionnaire referred to in

W A A 1 s submission went out not later than August 1983.

That questionnaire specifically referred to cancer.


In mid 1984 mortality lists were supplied by W A A to

the Commission. These are dealt with in detail in Chapter

X of this Report. Extracts from those lists became Exhibit

1746. No claim has been made that this exhibit or the

data from which the exhibit was compiled, support W A A ' s

claim that cancer generally or any rare form thereof was

"unduly prevalent in the causes of death".

By letter dated 12 January 1984 Monsanto Australia Limited

sought particulars of the matters set out in W A A 1s

November 1983 submission. Amongst the questions asked

wer e: 2 7

27. What are the number of deaths which it is alleged have occurred in the Vietnam

veterans group, and what is it alleged is the statistical expectation of death? If


practicable we should be grateful if you would furnish us with the names of those who have died.

29. How many deaths is it alleged have been

caused by cancer amongst the Vietnam

veterans group, and what are the "rare forms of cancer" referred to and what have been their frequency as a cause of death within the group? If practicable we would be

grateful if you would furnish us with

relevant names, dates and causes of death.

VVAA gave a considered reply. In a letter dated 11 May

1984 it answered:

27. A total number is still being collected. The DVA may be able to help. Names are


29. See 27. The two rare types of cancer are soft tissue sarcoma and lymphoma.3.

The total number of deaths from cancer was never provided

by W A A to this Commission and no information has been

forthcoming as to the prevalence of the "rare forms of


In May 1984 VVAA presented to the Commission a written

submission on the topic of toxicology. The general claims

as to the carcinogenic characteristics of TCDD and 2,4,5-T

are repeated in that submission (at p 64) and mention is

made of lung cancers, skin cancers (at p 48), gastro

intestinal cancers (at p 52), soft tissue sarcoma (at p

57), and malignant lymphoma (at p 58).


No allegation is made of a specific link between 2,4,5-T

or TCDD and any particular type of cancer. although in

what appears to be a general literature review, the work

of H a r d e n , Sandstrom, and Eriksson is discussed (at pp

57-60) . The submission also refers to a letter in the

Lancet dealing with a commentary by Coggon and Acheson (at 4 P 61).

There are also passing references to WHO's documents and

to EPA's reports leading to the claim - "In addition to

other effects TCDD and 2,4,5-T together are carcinogen,

teratagen (sic) and mutagen" (at p 64).

In its final submission lodged on 22 February 1985 VVAA

makes the following statement:

The submission is that amongst the chemicals sprayed in Vietnam are a number who (sic) because they are mutagens, carcinogens, promoters or immune suppressants have the potential to cause

or promote cancer.(at p 111)

It is appreciated that medical science does not fully

understand the nature of the disease. Nonetheless, the

generality of WAA's submissions has left the Commission

with the task of necessarily elucidating possible

mechanisms, identifying possible carcinogens, collecting

V I 1 1-6

and analysing the data in relation to the many chemicals

mentioned„ albeit in passing, and applying appropriate

carcinogenic principles to the data collected.

Accordingly, it seems appropriate to examine-

A. The theory of carcinogenesis,

B. The history of chemical carcinogenesis.

C. Principles and criteria for assessing the

carcinogenicity of particular chemicals.

D. The evidence that exists in relation to the

chemicals to which Australian personnel and

American servicemen in Vietnam might possibly

have been exposed.

E. & F. The evidence that exists in relation to the

health effects of individuals exposed (E)

occupationally, or (F) industrially as a result

of accidents involving relevant chemicals.

G. Animal and in vitro data analysis in respect of

relevant chemicals.


It will be remembered that a finding has been made by the

Commission that on the balance of probabilities most

Australian personnel were not exposed to Agent Orange or

indeed to any other herbicide at all or in any event to

any extent sufficient to have caused effects of the kind

alleged by WAA. 5

But, because there is a doubt about the relevant level of

exposure of Australian personnel to herbicides, some

assumptions are made implicitly in this chapter of the

Report about exposure. It should not be thought that the

Commission, in making such assumptions, is making any

finding of a relevant degree of exposure to such



Medical science does not understand the mechanisms of

action of a single causal factor of cancer and does not

fully understand the nature of 6 cancer. In fact the

mechanisms of production of cancer remain the subject of a 7 number of theories.

The history of cancer research is dotted with discarded

theories. For the two decades from 1920 it was believed

V111 - 8

that any association between cancer and chemicals was

limited and was based solely on repeated applications of a

particular chemical. Research proceeded on the so-called

initiation and promotion theory with scientists such as

Berenblum (1941) and Shubik (1947), through their croton

oil experiments, leading the way along the path of this

two-stage theory.8

However. Shubik has now moved away from such theory to a 9

theory of step progression. Experimental evidence

suggests that carcinogens, like other toxic substances, do

exhibit a dose-response relationship i.e. have a threshold

which must be reached before any effect occurs. In his

statement prepared for this Commission Dr Shubik said it

is "apparent that tolerance levels for carcinogens are

likely to exist".10

On the other hand, "single molecule" and "no threshold"

theories, although unable to be rejected out of hand on a

"theoretical model" basis, lack experimental or other

evidentiary support.

Currently the most widely (but not universally) accepted

theory of carcinogenesis is that the formation of a cancer

is dependent upon a mutation of cells in the body in which


the cancer occurs. (Brusick, Dost, Shubik).11 Thus,

the fact that a particular substance causes mutations in a

body may be some indication of the carcinogenic

characteristics of the substance; but not all mutagens are

carcinogens and a substance which is carcinogenic in one

species may not be carcinogenic in another. There are

substances which may be mutagenic in lower life forms but

which have no genotoxic effect in higher forms of life

just as there are substances which have been so shown in

one strain of a particular animal but have not been shown

to be carcinogenic in another strain of the same animal.

Further, there are other substances which are carcinogenic 12

in one type of animal but not in humans.(Dost, Brusick)



The knowledge that chemical compounds or mixtures of them

might cause cancer originates from clinical studies of

occupationally exposed groups. The classic instance was

the discovery by Sir Percival Pott in the 18th Century

that chimney sweeps developed cancer of the scrotal skin

after prolonged exposure to soot.


This observation stimulated experimental studies in the

late 19th and early 20th century. Researchers tried to

produce cancer in laboratory animals by painting them with

soot and coal tar.

13 No one succeeded until Yamagiwa et al persisted for months in applying coal tar to the skin of a rabbit's

ear. They were lucky in their choice of both tissue and

animal, because it is now known that dose, area of

application and species can be crucial in such


During the period between these experiments and World War

II there were two major areas of investigation.

The first was of occupational or iatrogenic cancers which

were studied both in human populations and in the


One example was the identifying of 2-napthylamine as the

probable cause of human bladder cancers amongst workers in

the Aniline dye industry. This was achieved by a seven

year study, involving large doses to dogs.

Another more confusing example was the identifying of

benzopyrene (B(a)P) as a principal carcinogen in coal


tar. B(a)P is a product of organic combustion and is

ubiquitous in the human environment. Low levels are found

in polluted air, in barbecued meat and in tobacco smoke.

The significance of these low levels in cancer induction

is unknown.

However, the levels at several parts per million would be

easily detected by 1980s analytical methods, and perhaps

perceived as high. For a lay person levels can be

confusing - a contamination level of 20 ppm may be said to

be dangerous: 20 ppt sounds much the same but is in fact

1 thousandth of the strength.

B(a)P is to be found in the polluted atmosphere of modern

urban areas. But atmospheric pollution can only be a

small factor in the current lung cancer epidemic:

cigarettes are unquestionably a major factor.

Other compounds which cause cancer in animals are present

in cigarette smoke. Thus even when a human carcinogen is

clearly epidemiologically identified, the toxicologist has

a daunting task. The mechanisms and precise factors

remain a mystery.

V I 11-12

The second major area of investigation was triggered by

the discovery in 1920 that the food additive "butter

yellow" (p-dimethylaminoazobenzine [dab]) caused liver

cancers in rats. This led to the routine testing of

"synthetic" additives.

Another key event was the discovery of the carcinogenic

properties of 2-acetylaminofluorene, a proposed

pesticide. It became a valuable research tool and its

discovery provoked routine testing of proposed pesticides

at the instance of US regulatory agencies.

On an entirely different front researchers sought to

discover and understand the mechanisms of cancer

induction. Some sought to f ind a relationship between

chemical structure and biological activity. Others studied

the metabolism of established carcinogens in animal

systems. Some compounds needed to change or metabolise

before they became active and slow and painstaking

research is continuing.

A third approach involved investigation of biological and

pathological factors that affect the process of cancer 14

induction. For example, Deelman found that by cutting

the skin he could augment the numbers of cancers caused by


coal tar application. This led to the view that a

multiphased process was involved and many studies

exploring the "initiation" and "promotion" theory ensued.

Some revealed factors that inhibited cancer induction and

the modern chemotherapeutic drug search began.

A further interesting discovery was that some chemicals

had a dual effect, inducing cancers at the site of

application and also augmenting the incidence of the

cancers found also in the control animals. This effect is

puzzling and also poses a major problem in the assessment

of modern cancer bio-assays.

Another discovery of importance was that cellular changes„

perhaps equatable with cancer, would be produced in tissue

cultures ("in vitro") by the application of suspected


Differences in susceptibility between species also

attracted interest and experiment. 1 5


Oppenheimer1s accidental discovery that inert plastics

inserted under the skin of rodents caused cancer seemed to

require revision of theories regarding chemical

interaction as essential for cancer production.


B .(1) The Present Era

For the purposes of this discussion the present era will

date from the end of World War II. although in truth major

changes started taking place about 1955.

The first area of concern about the effect of chemicals.

and particularly new chemicals. as potential causes of

cancer centred around the food additive and pesticide


Initially chemicals were tested in a bio-assay that was

designated as a 2 year chronic toxicity test. This test

was undertaken using the rat as the principal species

although some mice were used. It was assumed that this

test would uncover a variety of possible chronic effects

and no special emphasis was placed on this test as one for


Well known toxicologists, Barnes and Denz from the U.K.,

believed such tests to be of little value and felt that a

subacute test of three months. if conducted properly,

would reveal the majority of toxic effects. Indeed, their

views expressed in a review article in 1953 predicted

modern attitudes.16


In the course of some of the earlier chronic toxicity

tests undertaken by the U.S. Food and Drug Administration

many subsequent findings were predicted. DDT, for

example, was found to be productive of liver tumours in

mice; this was not considered to be of practical

importance at that time since the tumours were all benign.

Only malignancies were thought to be of relevance.

Amidst an atmosphere of some hysteria the US Congress in

the 1960s passed an amendment requiring the banning of any

food additive found to give rise to cancer in man or in

animal via an appropriate route of administration. This

"Delaney Amendment" removed the freedom of the regulator

to make a value judgment.

This amendment was based on the theory that no "no-effect"

level for chemical carcinogens could be determined: any

dose could be visualised as productive of irreversible

cellular change and potentially cancer. Such theory was

based, inter alia, on the German work of Druckrey et al 17 and has some validity.

Nonetheless Dr Shubik is of the view that no-effect levels

for carcinogens do exist. The Commission agrees; at least

V I 1 1-16

in the sense that, at low dose levels, increases in risk

will be so low that they will be unmeasurable and


There are many who believe that the Delaney Amendment

represents an oversimplification of the problem and that

it led to the banning of some products of value (such as

DDT and saccharin). It does not apply to therapeutic drugs

with anomalous results. For example, the early high dose

oral contraceptives, then known to cause tumours in rats

were permitted to be marketed causing tumours in women, 19 albeit at low incidence.

A value judgment by the regulator had been possible in

this latter case but was unavailable in the former. The

risks of obesity by the use of sugar as a sweetener are

greater than the risk of cancer induction by the use of , . 20


The bio-assay of chemicals for possible carcinogenic

activity has proceeded. Food additives, pesticides and

drugs have largely been tested by manufacturers in order

to obtain the necessary approval to market their

products. Requirements are applied in a similar manner

on a world wide basis and studies are carefully reviewed


by committees of WHO and FAO such as the Joint Expert

Committee on Food Additives (JECFA) and the corresponding

committee on pesticide residues. Joint Expert Committee on

Pesticide Residues (JMPR). Both these committees review

procedures regularly and set approved levels for use of

these chemicals.

The National Toxicology Program (NTP) of the U.S. has

introduced another dimension. In this program selected

chemicals are, in essence, "screened" for cancer induction

in standard tests in mice and rats. The tests performed

may bear no relationship to human use conditions of the


B .(2) Summary

Certain specific product categories are now routinely

tested for carcinogenicity according to planned

protocols. Other more generalized long-term animal tests

are carried out, notably in the U.S., in large scale

"screening" programs. Many of the results obtained have

been exceptionally difficult to interpret in practical


V I 11-18

B. (3) Research Patterns in Chemical Carcinogenesis:

1945 to Date

The first part of this era was concerned with following up

leads obtained in the three preceeding decades. Many new

classes of chemical carcinogens were discovered. Some of

these were more readily amenable to biochemical and

molecular biological study and thus more in-depth

investigation of the interaction of these compounds with

cellular components has become possible.

Also. newer techniques of cell biologists involving the

use of tissue cultures and of bacterial systems have

played an increasingly important role in such studies.

Since the early part of this century Bovari's theory that

cancer may be the result of a somatic mutation has been

considered possible. In the past decade a series of

techniques have been introduced for detecting mutations in

various bacterial and other systems. To correlate in

vitro mutagenesis with chemical carcinogenesis in vivo has

been found to be unreliable.


B. (4) Current Views on Criteria for Evidence of

Carcinogenicity of Chemicals

Criteria for assessing evidence for the carcinogencity of

chemicals were reviewed by a panel of the National Cancer

Advisory Board of the U.S. in 1977. In its introductory

statement this group stated:

The criteria that are described are guidelines and not rigid. universal criteria. The

complexity of the problem dictates that the evaluation of potential human hazards of a given agent must be individualized in terms of the chemical and metabolic aspects of that agent, its

intended use(s). the data available at the time the decision must be made, and other factors pertinent to the case under consideration. Each case must be considered on its own and the

criteria appropriate for one agent may not necessarily apply to another....

The Overall Assessment Process

Chemical carcinogenesis is a rapidly moving field and great quantities of data have been

accumulated during the past decade. Even though an individual experiment may yield only

suggestive information, this information may be of considerable importance when considered with other data.

Clearly, when the primary source of data comes from epidemiological studies in man, it may be possible to evaluate a chemical and institute scientifically-based preventive measures. However, even in the instances where data are available from humans, the data must be

supplemented with information from other sources before a conclusion can be reached.

For example. toxicological evaluation of carcinogenicity has classically relied upon long

VI 11-20

term in vivo studies as the primary source of data. Such studies have been performed in a routine manner, and evaluations have followed predetermined formulas. This rote method is

rapidly giving way to evaluations that take into account findings from in vitro tests, metabolism studies, and biometric analyses as well as any

other available information. One of these methods alone cannot produce a reliable estimate of a chemical's risk to man, but taken together they provide an estimate with a high level of


Carcinogens act via different mechanisms. which result in their having different magnitudes of risk to man. Even though there is no basis for the exact extrapolation of risk from experimental animal to man, current advances, if exploited to

the fullest, can provide a basis for

distinguishing the degrees of risk from different carcinogens. The scientific criteria should be reviewed often, and scientific advances should be fully adopted.

The scientific characterization of human risks from carcinogens involves the evaluation and integration of data from many disciplines. It requires scientific impartiality to review all appropriate data, both negative and positive,

including statistical estimations of low-dose response. Quantitative characterization of human risk requires scientific experience and

judgment. Because of the strengths and

weaknesses of the data to be evaluated in the assessment of human risk and the complexity of the problem, case-by-case analysis is most appropriate.21

In spite of the inability to derive a generic

classification of carcinogens, chemical carcinogens can,

in principle, be divided into two types. One type gives

non-threshold dose responses, is stochastic in mechanism,

and has some probability of producing carcinogenic effects


at any dose. The second type gives threshold dose

responses and, theoretically, has a no-effect level. A

few chemicals can be placed, provisionally. in one

category or the other: but for the bulk of chemical

carcinogens, no one can discern in which compartment they

fall. By dealing with chemicals case-by-case and by 22

studying mechanisms, better results can be achieved.



In order to assess the carcinogenic hazard of a particular

chemical the following factors are considered:

(a) Epidemiological evidence;

(b) Experimental animal evidence;

(c) Indirect biological evidence;

(d) Indirect chemical evidence;

(e) Conjecture.

The process was explained by Dr Bernard Stewart. Dr

Stewart is the Principal Research Fellow in the Childrens'

V I 11-22

Leukaemia and Research Unit, Prince of Wales Childrens'

Hospital, Randwick, Sydney. He is a member of a number of

consultative committees convened by the International

Agency for Research on Cancer, (IARC) an agency

established under the auspices of the World Health

Organisation (WHO). This agency which is supported by 16

National Governments undertakes research programs in

cancer epidemiology and chemical carcinogenesis. Dr

Stewart's membership of such committees as are referred to 2 3 in his curriculum vitae, establishes him as an expert

of international renown in the chemical carcinogenic field.

Dr Stewart said:

In essence there are only two areas of evidence - direct evidence - which we have that carcinogenic activity is taking place. Either one has

evidence of a change in the incidence of cancer in a human population when contrasted with the population - some group within the human

population when contrasted with the population as a whole - this is the study of epidemiology - or one is able to produce malignant tumours in experimental animals as a consequence of exposing

those animals to the chemical in question. Now other evidence may infer carcinogenic activity but in rigorous terms these are the only two areas in which the word carcinogen can be used in any definitive sense.

Inferences of carcinogenic activity may be drawn from certain biological testing procedures. I would imagine the Inquiry has previously heard of the Ames Test which is a method of demonstrating

that a chemical is capable of inter-acting with genetic material - in this case in bacteria - and


it is one of a series of so-called short term tests which infer carcinogenic activity but by no means prove it. There may be an inference of carcinogenic activity or carcinogenic risk simply because a carcinogen is detected in the new context; that is to say, one might, for example, discover asbestos fibres in the water supply and one knows that asbestos is carcinogenic and therefore an inference can be made of a

carcinogenic risk but again, of course, it is an inference.

And finally for completeness, certain

"inferences" of carcinogenicity are wholly based on conjecture.24

Dr Stewart also said:

The best known short term test is the Ames Test or the microsomal mediated mutation of salmonella ... As I have said once, and I will stress again, this is an extremely useful test procedure. It

tells - it indicates that a chemical is capable of being metabolised by mammalian enzymes to produce a reactive intermediate capable of interacting with DNA. It infers (the possibility of) carcinogenic activity but it certainly does not prove it.25.

By way of emphasising that epidemiological evidence and

animal evidence are the direct and most persuasive

evidence, reference is made to the evidence of Dr Shubik.

He said: I

I have now come to the unequivocal conclusion that it is cheaper and just as effective, to toss a coin about mutagenicity or not, rather than to go to the trouble of putting these compounds into bacterial systems so that as far as mutagenesis and carcinogenesis are concerned. there is a correlation with a certain number of carcinogens and certain chemical groupings. But I think the


overall assessment must be that it is not a very- good test and I really would not want to say that because something was mutagenic, or not, that it was necessarily going to be carcinogenic.26

Accordingly, the Commission will consider these five

factors (where possible) in evaluating the carcinogenic

hazard of relevant chemicals.


1. Australian Human Data

(a) AVHS Mortality Report

Again the Commission is fortunate in having available to

it direct data on cancer outcomes amongst Australian

Vietnam veterans. These data are contained in the AVHS

Mortality Report which is dealt with in considerable

detail in Chapter X.

Insofar as it relates to cancer the Commission reiterates

the following:

On 7 September 1984 the Director of the Commonwealth

Institute of Health submitted the Australian Veterans 27

Health Studies Mortality Report to the Minister for

Veterans' Affairs.



The group of medical. scientific and statistical experts

which prepared the Mortality Report included:


Dr Fett who was the Principal Investigator graduated in

1978 Bachelor of Medicine, Bachelor of Surgery and

Bachelor of Medical Science, with first class honours in

each degree.

He obtained the Masters Degree in Public Health at Harvard

University, majoring in epidemiology.

As Principal Investigator, Dr Fett came into contact with

the Commission on a regular basis and his intelligence and

integrity are acknowledged.

He was assisted by:




LINA FORCIER BSc(Hons), Postgrad Dip Genet


This group had the benefit of the services of a consultant

epidemiologist. Dr John D Mathews, BSc. MD, BS, PhD,

FRACP. At the time of preparation of the Mortality Report

Dr Mathews was the Principal Research Fellow at the

National Health and Medical Research Council.

Subsequently he was appointed Director of the Menzies

School of Health Research in Darwin and the Robert Menzies

Professor of the University of Sydney.28 Dr Mathews' 2 9 curriculum vitae reveals him to be highly qualified in

public health and epidemiology. He was chosen to be this

Commission's Senior Consultant.

Study Design

The Mortality Report analyses the results of a

retrospective cohort study of mortality among Australian

National Servicemen of the Vietnam conflict era. The

decision to undertake the study was made in 1981.30 Its

object was to assess the possible effects of Vietnam

service on subsequent mortality of Australian personnel by

determining whether death rates amongst Australian Vietnam

veterans were higher than amongst comparable non-veterans

for all causes of deaths combined as well as for various

other specific causes of death considered individually.

The population formally studied included all former


Australian National Servicemen from the Vietnam conflict

era other than:

those who had enlisted after February 1971;

those who had served in the army for less than 12


those who had died during army service within two

years of enlistment; and

those who had died from the direct effect of combat

injuries received in Vietnam.

This population was followed until 1 January 1982 in order

to determine whether each member was alive or dead (vital

status). The follow-up was extremely successful, the rate

of trace exceeding 94% for both veterans and

non-veterans.31 Analyses of the causes of death among

19,205 Vietnam veterans and 25,677 non-veterans were then

undertaken (i.e. a total of 44,882 National Servicemen).

Cancer Analysis

The analysis of deaths from cancer made in the Mortality

Report reveals no statistically significant difference in

VI11 -28

the rates between veterans and non-veterans. The relative

mortality rate is 0.99 (95% confidence intervals: 0.6,

1.6). In view of the claims made by WAA, the further

finding that there is no statistically significant

difference in the death rates for the veterans and the

non-veterans from soft tissue sarcoma, non Hodgkin1s 3 2 3 3

lymphoma or Hodgkins Disease are of particular


The number of deaths observed amongst veterans from all

neoplasms (I CD 140-239) was 31 compared with an expected

number of 32.9 (Ratio 0.94; rounded down in the Mortality

Report to 0.9). As the principal author of the Report, Dr

M.J. Fett, pointed out in his evidence, this represents an

observed rate of 1.6 per 1000 persons compared with 1.7

34 per 1000 expected. Amongst non-veterans the number of deaths observed from all neoplasms was 41 compared with 43

expected (Ratio 0.95; rounded up in the Mortality Report

to 1.0). Thus, the ratios of observed to expected cancer

deaths amongst veterans and non-veterans are essentially

,, 35

the same.

Deaths from cancer were anlaysed into groupings as follows:

VI11 - 2 9

Cause of Death ICD 8 Codes Description

171 Malignant neoplasm

of connective and other soft tissue

200, 202 Non-Hodgkin1s


201 Hodgkin's disease

204-207 Leukaemia and


203, 208-209 Other neoplasms of

lymphatic and haematopoietic tissue

140-170, Other malignant

172-199 neoplasms

210-229 Benign neoplasms

230-239 Neoplasms of

unspecified nature

and It was found that the death rates amongst both

veterans and non-veterans in each of the eight groupings

are similar to the death rates in the Australian male


Two (2) soft tissue sarcomas were found amongst the

veterans, 3 cases of non-Hodgkin's lymphoma and none of

Hodgkin's Disease. The 5 deaths amongst veterans from

these three causes are to be compared with the 7 deaths

observed amongst non-veterans from the same three 37 causes. As indicated above, none of the figures for

V I 11-30

the observed incidence of any of these types of cancer was

statistically significant.

In all there were 47 deaths in the category "other

malignant neoplasms" (ICD Codes 140-170 and 172-199). Of

these, 23 occurred amongst veterans (23.5 expected; ratio

0.98). Upon examining these 23 cases there was no

evidence of any excess of neoplasms of any particular type

including gastro-intestinal, lung and skin cancers.38

In his evidence about the Mortality Report Dr Mathews


The results of this study should be reassuring in the sense that the post Vietnam death rates of National Servicemen were not increased relative to those expected for the Australian male



As has been said in relation to the mortality study by other witnesses, and can be seen from the data presented there, there is no evidence to suggest that the patterns of mortality are related to chemical exposure■(Emphasis added)40

Dr Mathews' views are inferentially supported by the

contents of Table 3.15 of the Mortality Report which

demons tra te quite clearly that death rates did not vary

significantly with calendar year of first emplanement to


Vietnam either for all veterans or for veterans within

their corps groupings.41

Indeed there was no evidence of any interaction between

year of departure for or duration of Vietnam service for

those most likely to be exposed (ie infantry) on the one 4 2

hand and death rates on the other. As Dr Mathews

agreed, "It did not matter when you enlisted, it did not

make a difference to the mortality rate deduced from this „ . ..4 3 s tudy."

One of the purposes of Table 3.16 is to segregate the

Vietnam conflict into phases which correspond with the

differing intensities of herbicide spraying in those areas

in which Australian troops were centred. This analysis

enabled Dr Mathews to state that, "one found no difference

across the periods of low to medium to high to no




The Mortality Report provides no support for the claims

made by VVAA as to undue prevalence of cancers generally,

or soft tissue sarcomas and malignant lymphomas

specifically as causes of death amongst Australian Vietnam


veterans. In fact the Mortality Report strongly refutes

such claims, and is powerful evidence that there is no

present association between Vietnam service and cancer.

D . 1. (b) Evidence From Australian Vietnam Veterans

Amongst the witnesses called by VVAA on health effects

were two who suffered from cancer. They are dealt with in

the section of this Report called Health Effects - 4 B General. But it is appropriate to deal with them also

under this topic heading. (As was done in the Health

Effects section, these two veterans are not identified

other than by a number in order to keep their anonymity.

Each of the two is numbered in accordance with the number

used in the Health Effects section.)

The first veteran was No. 4. He was in Vietnam from June

1965 until June 46 1966 . He operated ou t of Bien Hoa

with occasiona1 visits to other parts of Vietnam,

including Vung Tau on two occasions.

In 1983 he was diagnosed as suffering from a Grawitz

4 8

tumour of the kidney.

VI11 -3 3

This cancer was accepted by DVA as pensionable because an

increased incidence of such tumours has been associated

with heavy cigarette smoking and the veteran's consumption

of cigarettes had increased markedly during and as a 4 9

result of his Vietnam service.

The veteran gave no evidence of exposure to herbicides in

general nor to Agent Orange in particular and he did not

assert that he had even seen any effects of spraying in

country he passed through.50

The Commission knows of no medical or scientific

literature which associates the incidence of Grawitz

tumour with exposure to Agent Orange, other herbicides, or

chemicals used for vector and other insect control in

South Vietnam.

The second was Veteran No 3. He arrived in Saigon in

December 1967 and remained in Vietnam until August 51 1968. Except for a period of 6-7 weeks when he was at

Nui Da t he spent his time in Vietnam stationed at Vung


During his tour of Vietnam he worked substantially as a

motor mechanic repairing vehicles and equipment. That


equipment: included a truck mounted with spraying


Whilst he was in Vietnam the herbicides used for perimeter

spraying did not include Agent Orange or its


constituents. Nor does he claim any exposure


consequent upon the Holt/Lugg trials which occurred in

Nui Dat during the period of his service.

His only claimed association with spraying was that he

observed men with portable foggers in the evenings at Nui

Dat and there was also an occasion when he observed Number

1 Armed Regiment spraying with big water tankers as part

of the ground preparation for a workshop which was to be

built. The substance sprayed on that occasion was said by

him to be like that' used every evening for fogging for


mosquitoes. The substance used for spraying for

mosquitoes was of course malathion.

In 1980 he was found to have a spot on his right testicle

and although the biopsy of the sample after removal was

said to be negative for cancer he was later found to be

suffering from a seminoma which is a form of testicular . 5 6

cancer. This cancer was removed in July 1980.


The Commission has not been able to find any evidence in

the literature that any form of testicular cancer is

associated with exposure to chemicals. No scientific

evidence has been tendered by VVAA to suggest such an



A study of the incidence of testicular cancer was

tendered by Counsel Assisting the Commission and that

study shows an incidence rate higher amongst professional,

administrative and clerical workers than amongst manual

workers who might be expected to be exposed either in

outdoor labouring duties or in factory work.


The evidence in neither case supports a claim that

exposure to chemicals in South Vietnam was a cause either

directly or indirectly of their respective cancers and

indeed neither has a cancer of a type referred to in

VVAA1s claim.

D.l.(c) Incidence of Lymphoma Known to DVA

Ms Robyn King of the Department of Veterans' Affairs (DVA)

produced evidence from departmental records concerning

death from lymphoma amongst Vietnam veterans up to 22

V111 - 3 6

November 1984 together with evidence of the incidence of

lymphoma amongst those veterans.58 These exhibits

detailed the kinds of neoplasms including their I CD code

classifications as well as the names. dates of service and

dates of first diagnosis in respect of each of the 47

Vietnam veterans involved.

The Commission observes that the spraying of Agent Orange

commenced in January 1965 and was suspended on 15 April 59 1970. No defoliation operations employing fixed-wing

aircraft were conducted in Phuoc Tuy Province after 30

June 1968 and all defoliation in that Province ceased on . 60

or prior to 1 January 1969.

A total of 47 veterans were diagnosed as having lymphoma.

Two (2) of these were diagnosed before they went to

Vietnam. Of the balance. 21 have a requisite minimum

latency period of 10 years. Cancers caused by exposure to

chemical agents take from 5 to 25 years to develop, with

10 years plus being generally accepted as an appropriate

minimal latency period.61 Indeed, of the 21, 5 have only

the bare minimum. (It should be borne in mind that

latency periods are based on assumptions. Radiation-caused

leukemia is known to have shorter latency periods and

latencies for lymphoma might possibly be shorter than the

VI11 -37

10 years nominated. 5-10 might be appropriate. As

appears below this distinction became unimportant.)

Of the 21 cases with the requisite minimum latency period,

6 were in Vietnam at a time outside the period during

which Agent Orange was sprayed and one of them (Case 44)

may have been in Vietnam outside that period.

This leaves 15 (or perhaps 14) Vietnam veterans who

suffered from lymphoma and who served in Vietnam during

the period when there was a possibility of being sprayed

with Agent Orange.

Although there has been no specific evidence that any of

these were in fact sprayed with Agent Orange or other

herbicides, the Commission assumes for the present

analysis necessary exposure.

In the result, 45 Vietnam veterans have suffered from

malignant lymphoma after commencement of Vietnam service.

To investigate any increase in incidence one must compare

that number with the number to be expected in the like

population of Australian males of like age who have not

been to Vietnam. The expected number of malignant

V I 11-38

lymphoma can be estimated by summing, over all veterans,

the product of their period at risk of lymphoma after

commencement of their Vietnam service and the lymphoma

incidence rates from Australian cancer registries.

Because lymphoma incidence rates vary with age, the age of

the Vietnam veterans has to be taken into account in the


The statistician associated with the AVHS studies. Dr

Michael Adena, performed the necessary calculations and

was called to give evidence. He calculated the expected

number to be between 56 and 58.

An incidence of 45 as against 56 to 58 expected, provides

in itself strong evidence against any causal association

between the experience of Australian personnel in Vietnam

and the incidence of malignant lymphoma amongst them.

If allowance is made for a latency period of not less than

10 years from the date of commencement of service in

Vietnam and if those Australian personnel who went to

Vietnam after 1970 are excluded, then the number of

malignant lymphomas expected in the cohort would be 28.77

and the number of malignant lymphomas observed is 17.

Thus, whether allowance is made for a latency period or

VI11-3 9

not, the number of malignant lymphomas observed amongst

Vietnam veterans is lower than the expected number.

The Commission therefore concludes that Vietnam service

did not give rise to an increased incidence of malignant

lymphoma and there is no support for the proposition that

exposure to Agent Orange, herbicides of other types or

indeed any of the chemical agents used in Vietnam caused

that disease amongst Vietnam veterans. The Commission

also observes that a comparison of DVA material with the

most thoroughly collected AVHS material reveals that DVA

would have recovered by their researches at least 90% of

those suffering from malignancies. Even if one adds a

full 10% to the numbers recovered, the incidence does not

approach that statistically expected.

D.l.(d) Incidence of Soft Tissue Sarcoma Known to DVA

Four cases of soft tissue sarcoma amongst Vietnam veterans

are known to DVA. One is still living and three are

dead .




1 1972

2 1974

3 ■ 1977

4 N/A

These four DVA file numbers are recorded in a confidential


Another veteran originally had his cause of death coded as

a soft tissue sarcoma. In the course of the AVHS

Mortality Study a review of the available data showed this

to be an incorrect coding and that it should have been ICD

199; that is, not soft tissue sarcoma. This case is

referred to in the Mortality Report.62

The Commission sets out in the following table some

further details.



1 2/9/69 May 1971 1 yr 8 mths

2 17/2/69 November 1972 2 yrs 9 mths

3 17/11/69 December 1972 3 yrs 1 mth

4 13/1/70 September 1983 13 yrs 8 mths


Thus it can be observed that one only of the four cases

has a latency period of or in excess of 10 years. Any

association between the occurrence of the soft tissue

sarcomas in cases 1, 2 and 3 and service in Vietnam or

exposure to herbicides or other chemical agents is

therefore highly improbable.

Case 4 was a member of the 110 Signal Squadron and was

based at Nui Da t in the Phuoc Tuy Province. All

defoliation ceased in that Province some six months prior

to the arrival of Case 4 in Vietnam63 and Agent Orange


was wholly suspended on 15 April 1970. Case 4 was

thus in Vietnam at a time when no spraying of Agent Orange

took place in the Phuoc Tuy Province.

In addition, examination of the four cases of soft tissue

sarcoma reveals that they are all of different types. If

some specific agent was causing cancers one might expect

them to be of similar pathology.

Broadly speaking, the incidence rate of soft tissue

sarcoma in the Australian population is 2 per 100,000 per


The Surveillance Epidemiology and End Result Program 65 (SEER) has produced a table showing the estimated


incidence of soft tissue sarcoma for various age

groupings. That table is as follows:

Table lc

Aqe Group No. of Males Estimates Cases of STS

(Yrs) (1000) No. Rate per 100.000

30-34 980 20 2.0

35-39 907 14 1.5

40-44 740 17 2.3

45-49 590 22 3.7

50-54 552 33 6.0

TOTAL 3769 106 2.8

Dr Adena adopted a similar procedure with these figures to

that adopted by him in respect of malignant lymphomas.

If a period from the commencement of the calendar year of

first service of each veteran in Vietnam until the end of

1984 is taken and if mortality amongst Vietnam veterans is

assumed to be similar to that amongst other Australian

males a total of 683,532 person years of risk results.

If a soft tissue sarcoma incidence rate of 2 per 100,000

head of population per annum is adopted, the expected

number of soft tissue sarcoma cases amongst the Vietnam

veterans cohort would be 13.7.


However, if allowance is made for a latency period of at

least 10 years from the commencement of service in Vietnam

of each veteran and if those Australian personnel who

first went to Vietnam in or after 1971 (that is, after

cessation of spraying) are excluded. the number of person

years of risk is 272,737: the number of soft tissue

sarcomas then to be expected amongst the Vietnam veterans

would be 5.45. Thus the comparison between the number of

soft tissue sarcomas expected and those observed is either

13:4 or 5:1 depending on whether latency is taken into

account or not.

The Commission is aware that the 1 healthy worker1 effect

is known to play a less significant role with regard to

cancer than other diseases.66 Accordingly, it may be

argued that the shortfall between the number of soft

tissue sarcomas observed and the number expected is more

than one would anticipate. The result in a sense is "too


Such a shortfall could be the result of under-reporting,

misclassification of disease, chance and/or a genuinely

lower incidence rate amongst Vietnam veterans.


Soft tissue sarcoma is a serious disorder, commonly

fatal. Pursuant to regulation 65A of the Regulations made

under the Repatriation Act 1920 veterans are entitled to

full and free hospital, medical and surgical treatment for

cancer whether or not it is related to war service.

Local Medical Officers are well aware of Regulation 65A.

In addition there has been a great deal of public

controversy about cancer and its possible relationship

with exposure to Agent Orange in Vietnam.

It seems unlikely then that DVA would not become aware of

a cancer case amongst Vietnam veterans. The testing

procedures comparing DVA information with AVHS information

indicates a 90% recovery rate.67 The Commission is

inclined to assign the relatively small number of soft

tissue sarcomas amongst Vietnam veterans to chance

differences in diagnostic practice rather than to any

protective effect.


The evidence before the Commission concerning the

incidence of soft tissue sarcoma not only does not support

WAA's claim that exposure to chemical agents in Vietnam


has given rise to an increase in deaths from or in the

incidence of that form of cancer: it strongly suggests

that there is no such causal effect.

Thus, considering human data in relation to Australian

personnel, there is no evidence to support the claim that

Australian Vietnam veterans are dying or suffering from

cancer or any particular form of cancer in higher numbers

or at an earlier age than expected for such a group. As

has been said earlier, this data is the best data

available for answering such a guestion.

D.2 United States Human Data

Of course Australian personnel were not alone in Vietnam.

A responsible search for human data involves analysis of

available studies of American servicemen. Such analyses

enable the Commission to test the consistency of the

Australian data. The most important studies are the Ranch

Hand studies.

Operation Ranch Hand was the code name given to the

tactical military project for the aerial spraying by

fixed-wing aircraft of herbicides in Vietnam between 1962

and 1971.68 The spray flights were known as trail dust


missions. A total of 1,269 US personnel served in the

Ranch Hand project between 1962 and 1971.69 Of this

number, 22 were killed in action leaving 1,247 Ranch Hand

personnel as possible subjects for study.

. 70

As the Commission has already concluded Ranch Hand

personnel were the group which was most heavily exposed to

Agent Orange in the Vietnam conflict. The evidence of

Colonel Hubbs and Mr Dudenhoeffer, members of the Ranch 71

Hand team, has been referred to earlier in this Report.

There are two Ranch Hand studies. The first, conducted by

Colonel Lathrop as Principal Investigator and called Ranch

Hand I was a matched cohort study of mortality as at 31

December 1982. It involves 1,247 Ranch Hand personnel and

6,171 matched controls.

The second study, of which Colonel Lathrop was again the

Principal Investigator, was called Ranch Hand II. This

was a baseline Morbidity Study making a comparison between

Ranch Hand personnel living at the commencement of the

study and a comparison group matched by age, race and



D .2 (a) Ranch Hand I (1983)72

In the Ranch Hand I Study an analysis of Ranch Hand

personnel and matched controls was undertaken. Ranch Hand

personnel were accepted as having been exposed to Agent

Orange. The exposed study population was defined as those

individuals who were formally assigned to the USAF

organisations responsible for the aerial dissemination of

herbicides and insecticides in Vietnam from 1962 to

1971. A total of 1269 Ranch Hand personnel were

identified and a population of 24,971 individuals was

assembled for comparison purposes. These had not been

occupationally exposed to herbicides or insecticides in . 74

Vietnam or otherwise. After matching had taken place

there remained 1241 Ranch Hand personnel who formed the

Ranch Hand cohort. They were matched for age. race, sex

and occupational category from the comparison population 7 5 in the proportion 1:5.

Analysis of mortality from cancer revealed four (4) deaths

amongst Ranch Hand personnel compared with 39 amongst the

comparison group. This did not indicate any elevation of

relative risk for the Ranch Hand personnel, the relative

risk for cancer mortality being a reassuring 0.503. with a

95% confidence interval of 0-1.024.76

V I 11-48

An analysis of malignant neoplasms according to site was

carried out. No malignant neoplasms of the bone,

connective tissues, skin or breast (ICD 170 to 175), i.e.

no soft tissue sarcomas (ICD 171) were found and there

were no cases of malignant neoplasms of lymphatic or

hematopoietic tissue (ICD 200-208), i.e. no malignant

lymphomas (ICD 200, 202), and no cancers of the digestive 77 organs.

An analysis of the neoplasms according to their morphology

revealed that there were no deaths from malignant lymphoma

or Hodgkin's Disease amongst Ranch Hand personnel, whereas

there were deaths from both malignant lymphoma and

Hodgkin's Disease amongst the comparison group.78 (NB:

None would be "expected" in such a small cohort: the nil

is non probative).

In addition to comparing the Ranch Hand personnel's

mortality experience with that of the Air Force comparison

group, their mortality experience was compared with that

of three other comparison groups, namely, all United

States white males. non-disabled retired military officers

and enlisted personnel (obtained from the Department of

Defense mortality information) and the West Point class of

1956. No death rate comparisons based on specific causes

VI11-4 9

were carried out in relation to the US males or retired

officer groups. However, an examination of the deaths from

cancer amongst Ranch Hand officers compared with the West

Point officer group was carried out. This showed that 283

Ranch Hand Officers were at risk compared with 458 West

Point Officers and that whereas there were no deaths from

cancer amongst the Ranch Hand Officers there were six (6) 79

amongst the West Point Officers. A calculation of the

relative risk derived from the relevant table in the

Study80 could not be made because of the absence of

deaths in the Ranch Hand Officers in the analysis of

interest, but, as the study points out, "this finding is

consistent with the apparent but not significantly

decreased Ranch Hand cancer mortality noted in the Ranch

Hand versus comparison group analysis".81

The conclusions stated in the Ranch Hand (Mortality) Study


The mortality analyses described in this report have not revealed any adverse death experienced in the herbicide/dioxin exposed cohort . . . there

is no indication that operation Ranch Hand personnel have experienced any increased mortality or any unusual patterns of death in time or by cause. They are not dying in

increased numbers, at earlier ages or by

unexpected causes, (emphasis added)82

V I 11-50

Not only are Ranch Hand personnel "not dying in increased

numbers, at earlier ages, or by unexpected causes"83 but

this most highly exposed group compares more than

favourably with both their West Point equivalents and the

comparison group in relation to death from cancer.


The Ranch Hand I study does not support the claim made by

W A A of an undue prevalence of deaths from cancer or from

specific types of cancer amongst Vietnam veterans. It

provides suggestive evidence to the contrary which alone

would be unconvincing because of the small numbers


D.2 (b) Ranch Hand II (1984)84

This study examines the health outcomes of 1241 Ranch Hand

personnel - 1208 living Ranch Hand personnel were matched

with 1238 comparisons for age. sex and occupational

category.88 The members of the comparison group were

drawn from Air Force personnel who had been assigned to a

variety of cargo mission organisations throughout South

East Asia during the period 1962 to 1971 and had not been

occupationally exposed to herbicides or insecticides.86

VI11 -51

The Ranch Hand and comparison groups were surveyed by

questionnaire and physically examined. Contact with the

subjects was made initially in November 1981 and continued

through until November 1982, with the predominance of

mailing of questionnaires taking place in the first half

of 1982.87 The physical examinations took place over a

period which extended until December 1982.88 Thus. the

two groups were effectively followed for a minimum latency

period in excess of 10 years and, as far as the physical

examinations were concerned, for a minimum of

approximately 11 years.

Analyses were conducted in relation to a number of health

effects. They included the incidence of all neoplasms,

neoplasms according to site (site specific analysis),

neoplasms according to type (morphology) and the

relationship between the incidence of cancer as a whole

and of various types of cancer each related to an Index of

Exposure which was developed for the Ranch Hand personnel.

Fourteen (14) malignant systemic neoplasms were found

amongst Ranch Hand personnel. An identical number was

found amongst the comparison group.89

In the site specific analysis, no cancers of connective 90

tissue were found amongst the Ranch Hand group nor was


any indication of site specificity or "target site" found


amongst the Ranch Hand personnel.

In addition the analysis of the morphology of the systemic

malignant neoplasms revealed:

(i) a nil incidence of non-Hodgkin1s lymphoma amongst 92 the Ranch Hand personnel; and

(ii) a nil incidence of Hodgkin's Disease amongst the 93 Ranch Hand personnel;

Analysis of the incidence of skin cancer showed that they 94

all occurred amongst non-black individuals.

Thirty-five (35) malignant skin cancers were found amongst

Ranch Hand personnel and 27 amongst the comparison


group. Seventy-five percent (75%) of all verified

neoplasms amongst Ranch Hand personnel were in fact

cancers of the skin. For the comparison group the figure

was 73%. Of the skin cancers reported amongst Ranch Hand

personnel, 44% were verified as malignant.96 The

similarity of these percentages in the exposed and

unexposed groups rather points to some common causal

factor and hence away from Agent Orange.


Analysis of the malignant skin cancers by cell type

revealed that 31 of the 35 occurring in Ranch Hand

personnel were of the basal cell type, ie non-melanomic.

Non-melanomic cancers accounted for 91% of the Ranch Hand

skin cancers and for 93% of those occurring in the

comparison group. The similarity of these percentages for

both the exposed and unexposed groups again points to a

common causal factor and away from Agent Orange.

The study reports that:

These findings are consistent with reported data that nonmelanoma cancer of the skin is the most common malignant neoplasm in the white population of the United States.97

Analysis of the incidence of malignant skin cancers by

reference to site of occurrence98 showed that 26 (81%)

of the Ranch Handers and 20 (80%) of the comparison group 9 9

had their skin cancers on the face, head and neck.

This is consistent with the effects of ultra-violet

radiation (sunlight), which has been implicated in the

medical literature as the dominant factor in the

development of non-melanomic skin cancer.100 The number

of systemic cancers amongst Ranch Hand personnel and the

comparison group was identical and does not support the

claim by W A A of an increase in cancers amongst a


population exposed to Agent Orange or other phenoxy

herbicides. The difference in the incidence of skin

cancer between the Ranch Hand personnel and the comparison

group was reported as being "of borderline significance"

statistically101 and further work is presently (June

1985) being undertaken to ascertain the relationship

between the incidence of skin cancer in both groups and

the geographic areas of residence of the group members.

As ultra violet radiation is known to cause skin cancer

the data favours it as the common causal factor and points

away from Agent Orange or other chemical exposure.

An index of herbicide exposure was compiled separating the

Ranch Hand personnel into those who had low, medium and

high exposures. Analysis of the incidence of cancer

related to the herbicide exposure index showed no

dose-response relationship between herbicide exposure and

the occurrence of either skin or systemic cancers in the

Ranch Hand group. Indeed a suggestive negative was noted

amongst the enlisted flying group.102

The conclusion derived from the study includes the


There are no significant group differences for malignant or benign systemic tumors.


Overall there is no consistent data to show that the Ranch Handers are developing uncommon systemic cancers, or cancer in unusual sites, or

at a younger age.

Both systemic and skin cancers in the Ranch Hand group do not correlate consistently with the herbicide exposure index.103

This study has not identified statistical group differences for illnesses commonly attributed to dioxin exposure.104

A written statement prepared by Dr Marvin Schneiderman, a

statistician called by WAA, claimed that:

There was an increase among the study cases in Ranch Hand II in the observed number of a few types of tumors, including possibly the testicular tumours.105

Dr Schneiderman sought to support this claim by referring

to a paper. This paper became Exhibit 1671. In that paper

he compared the Ranch Hand II results with information

culled from the Armed Forces Institute of Pathology

(AFIP). Whilst Table 4 of this paper conceded that any

elevations in the relative risk for Ranch Hand personnel

with respect to certain selected types of cancers were

statistically significant only in relation to those of the

skin (as to which see the discussion above) and of the lip

and oral cavity. Dr Schneiderman extended the claim in his

oral testimony before the Commission so as to claim that

"there is the excess of respiratory cancers"106 amongst


the Ranch Hand personnel. It should be noted, however,

that he admitted that the information on which he based

both this statement and the statements contained in Table

4 to Exhibit 1671 came only from Tables Appendix XX, X-7


and X-5 in the Ranch Hand II study.

In making his oral statement about respiratory cancer Dr

Schneiderman has not compared like with like. The AFIP

figures for respiratory system cancer do not include as

many cancer sites as those covered by Table X-5 in the

Ranch Hand II study. Table X-5 in fact indicates that the

3 cancers observed in the Ranch Hand group, and used by Dr

Schneiderman as the basis for his statement, include not

only respiratory cancers but also " intra-thoracic"

cancers. In this context, it should also be noted that in

the Ranch Hand I study, there were no deaths from cancers

of the bronchus and lung amongst the Ranch Hand group

personnel but there was one amongst the comparison

group.108 Reference to Table X-6 in Ranch Hand II Study

shows that there were in fact only two cancers of the

bronchus and lung amongst the Ranch Hand personnel.

However Dr Schneiderman admitted in the course of

cross-examination that he had ignored the contents of this

Table 1 OQ . Not only did Dr Schneiderman fail to take

into account the information in Table X-6 but he also


admitted that the information in Table X-6 in fact

indicates that in respect of mortality and morbidity for a

whole range of malignant neoplasms of the bronchus and

lung "the Ranch Handers fared better than the

comparisons".1· 1 ·0 It should also be noted that in making

his comparison Dr Schneiderman counted only one cancer

from the comparison group rather than the two indicated in

Table X-5.111

As a result of his use of incomplete and inaccurate data

the calculations on which he based Table 4 of Exhibit 1671

are not relevant and his assertion that the relative risk

for "respiratory cancer" amongst Ranch Hand personnel is

2.44 (admittedly not statistically significant) is not

correct. When the correct information was used and a

comparison made between all malignant neoplasms of the

bronchus and lung in Ranch Hand personnel and in the

complete comparison group (ie 1238) the relative risk

computed by Dr Schneiderman was not 2.44 as appearing in


Table 4 in his paper. but only 1.582, a figure which

was not suggested by Dr Schneiderman to be, and was not in

fact, statistically significant.

Dr Schneiderman also claimed in his Table 4 that the

relative risk of bladder cancer in Ranch Hand personnel

when contrasted with the comparison group was 1.75.


Although he admitted that this was not statistically

significant Dr Schneiderman asserted that there was "an

excess" of bladder cancers amongst the Ranch Hand . 113 personnel.

Once again in making his assertion. Dr Schneiderman relied

solely upon the material in Appendix XX of Ranch Hand II


Study. He ignored the contents of Table X-6 in the

Ranch Hand II Study. This contained the following data:


(Extracted from Table X-6) (Ranch Hand II Study)

ICD-10 Morbidity

Codes Nomenclature Ranch Hand Comparison

M 801-804 Epithelial neoplasms Bladder 0 1

M 812-813 Transitional cell Papillomas and Carcinomas Bladder 2 1

i.e. 2 Ranch Hand versus 2 comparison group.

When asked in cross-examination to compare the full

cohorts of Ranch Hand personnel (1208) and comparison

group (1238) Dr Schneiderman computed the relative risk of

bladder cancer amongst the Ranch Hand personnel to be only 115 1.054 and no claim was made that this figure is

statistically significant - indeed, its very proximity to

unity indicates that it is not.


In his written statement Dr Schneiderman also submitted

that Ranch Hand II may "possibly" indicate an increased

risk of testicular cancer amongst the Ranch Hand

personnel.116 However, in cross-examination, 117 he

had to concede that only 2 Ranch Hand personnel had

testicular cancer compared with 1.1 expected, that this

did not have any statistical significance and was not a

surprising result amongst the number which made up the

Ranch Hand population.118

In making the claim in Table 4 to Exhibit 1671 that the

relative risk for "lip and oral" cancers was 1.75 for

Ranch Hand personnel , he once again looked only at the

contents on Appendix XX of the Ranch Hand II Study. He

thus failed to take into account the notes to Table X-5

(the Morbidity-Site Specific Table). These stated that

all Ranch Hand personnel and members of the comparison

group who had malignant neoplasms of the lip, oral cavity

and pharynx (ICD Code 9th edition. Codes 140-149) reported 1 2 0

a history of cigarette and cigar

, . 119

smoking. This

observation is even more noteworthy in view of the

assertion in Dr Schneiderman 1 s own 1984 paper as to " the 120

absence of cigarette smoking data".


None of Dr Schneiderman1 s attempts to detract from or

weaken the conclusions stated in Ranch Hand II are to the

point and they do not in any way detract from those

conclusions which were based upon exhaustive statistical

analyses of an immense amount of data expertly gathered

and carefully analysed over a period of years at a cost of

some $US11 million.

Before leaving Dr Schneiderman it should be pointed out

that he did not in any way seek to deny the conclusions

stated in the Ranch Hand II Study itself, nor did he

attack or seek to undermine the conclusions in the Ranch

Hand I Study. In relation to the Ranch Hand II Study, he

merely carried out a series of calculations based on a

number of assumptions. These calculations gave him

answers which were mathematically accurate, but since he

ignored or selectively used the data in the Study, his

calculations amount to no more than a facile, but

meaningless, exercise demonstrating that his approach was

more appropriate to the advocate than to the scientist

with complete detachment.


The Ranch Hand II Study does not support VVAA1s contention

that service in Vietnam has given rise to an increased

V II1-61

incidence of cancers or of rare forms of cancers. Indeed„

the study points to the absence of any such effect

together with an absence of any association between

herbicide exposure and cancer incidence in U.S. Vietnam


D.2(c) Veterans' Administration (U.S.)

The U.S. Veterans' Administration maintains computerised

patient treatment and discharge records. A search of

these found 234 cases of soft tissue sarcoma (ICD 171)

amongst Vietnam era veterans. Forty one percent (41%) of

these Vietnam era veterans had served in Vietnam and it

was found that 86, ie 37% of the cases of soft tissue

sarcoma, occurred amongst the Vietnam veterans. Dr

Mathews, the Senior Epidemiologist Adviser to the

Commission, said of such a relationship that:

The most important obvious inference to be made from this study is that the proportion of tumours occurring in those who went to Vietnam at around 40% is exactly the same as the proportion of those . . . who were in the armed services in the Vietnam era. Now, that would suggest there was no effect of Vietnam service.121

Although the findings from this study are yet to be

reported and full details of the method of analysis are

yet to be ascertained, the figures certainly do not reveal

VI11 -62

any undue prevalence in the incidence of soft tissue

sarcoma amongst Vietnam veterans or, indeed, any elevation

at all in such incidence.


D . 2(d) Greenwald et al

In this study the New York State Cancer Registry was used

to identify all the living and deceased men with soft

tissue sarcomas (ICD 171) diagnosed in New York State

(excluding New York City) between 1962 and 1980 and who

were between the ages 18 to 29 at any time between 1962

and 1971. The 285 males so identified were then matched

with controls. Amongst the 285 subjects there were 151

living and 134 who had died. Live cases were matched with

live controls and deceased cases were matched with

deceased controls. Each case and control (in the case of

deceased persons the next-of-kin) was interviewed by

telephone and data were collected about military service,

occupation and other matters which might touch upon

herbicide exposure. It was found that 11 of the 285 cases

of soft tissue sarcoma (3.8%) had served in Vietnam

compared with 18 of the 285 living controls (6.3%) and 9

of the 133 dead controls (6.7%) . Thus "there was no

association found with service in Vietnam when soft tissue

sarcoma cases were classified into sub-groups according to

V I 1 1-63

Most relevantly for the type of soft tissue sarcoma.

the present Inquiry, the study reports that:

No association was found between soft tissue sarcoma and any of the study variables including Agent Orange ... 124

The negative result of this study, which does take into

account exposure to Agent Orange, has been criticised by

Dr Schneiderman on the basis that the latency period

involved is too short. However, the mean interval between

commencement of Vietnam service and diagnosis of soft

tissue sarcoma was only fractionally short of 10 years,

namely 9 years and 7 months and the maximum interval was 12 5 13 years and 8 months. In addition, it should be

remembered that the study was undertaken in response to

claims made in the United States in relation to soft

tissue sarcoma which were similar to those made by VVAA in

the present Inquiry.



The study by Greenwald et a 1 does not support the claims

by VVAA that exposure to Agent Orange in Vietnam has led

to an undue prevalence of cancers generally and of soft

tissue sarcomas and malignant lymphomas in particular

amongst Vietnam veterans. In fact it provides contrary



D. 2(e) Kogan et al 126

This study which was completed in January 1985. looks at

the patterns of death amongst Vietnam veterans who had

served in Vietnam for a period of at least 6 months in the

period between 1 July, 1958 and 1 April 1973, veterans who

had not served in Vietnam and non-veteran males from the

State of Massachusetts, U.S.A. Although the study

concluded that there was a statistically significant

elevation in the number of deaths due to connective tissue

cancer amongst the Vietnam veterans when compared with

both non-Vietnam veterans and non-veteran males this

conclusion is based on only 9 deaths and on limited

information from only death certificates as to the cause 1 77 of death. In addition:

(i) no information was available as to the actual

duration of service in Vietnam of any Vietnam

veteran or as to the actual dates when such

service took place. All that was known was that

each Vietnam veteran in the study had served at

least 6 months in Vietnam at some time in the

period between 1 July 1958 and 1 April 1973;128

VI11 - 6 5

(ii) no Information was available as to exposure (if

any) of the study subjects whilst they were in 129 Vietnam;

(iii) no information was available as to exposures (if

any) of any of the subjects before they went to

Vietnam or following their return or, if they had

been so exposed. to what and when;

(iv) no information was available as to the date of

onset of the cancer from which each subject later


(v) the conclusion was based on only 9 cases of soft _ . 130

tissue sarcoma;

(vi) there was a spread of types amongst the soft

tissue sarcomas found, there being 5 different

„ * 131 types found.

It is material that the authors themselves note the above

considerations and comment:

The present study was not based on either

adequate numbers of deaths or adequate exposure information to help resolve (the) important issue.


i.e. whether any association exists between soft tissue 132

sarcoma and exposure to phenoxy herbicides.

. 13 3

This is a comment with which Dr Mathews agreed.

Details of the deaths from soft tissue sarcoma amongst the

United States Vietnam veterans who had soft tissue

sarcomas are as follows:


Extracted from Exhibit 1872 Table 7.

Case Year


Year Discharged

Year i Death

1 1969 1971 1975

2 1967 1970 1976

3 1965 1967 1976

4 1967 1972 1977

5 1964 1967 1977

6 1970 1971 1978

7 1970 1971 1982

8 1971 1974 1982

9 1961 1966 1983

Even if it be assumed that each of the Vietnam veterans in

the above table went to Vietnam immediately upon induction

into the Army and was exposed to Agent Orange immediately


on his arrival there (i.e. he had no basic training) and

that the onset of the cancer in each subject was

coincident with his death, there may be as few as 3 cases

of the 9 cases (Cases 5, 7 and 9) in which the latency

period would be greater than 10 years and the worst

situation on such assumptions would be 6 cases with such a

13 4

latency period (Cases 3, 4, 5, 7, 8 and 9).

It is clear that Case 8 could not have been exposed to

Agent Orange in Vietnam as his total period of army

service was outside the period when Agent Orange was

sprayed in Vietnam, namely, between 1965 and 15 April 135 1970. He must therefore be excluded.

This leaves 5 cases (Cases 3, 4, 5, 7 and 9) with a

latency period of 10 or more years. These cases have a

mean latency of 11 years 2 months and a maximum period of

13 years - both of which periods are close to the latency

periods in the study undertaken by Greenwald et al.

If, somewhat more realistically, it be assumed that each

of the Vietnam Veterans referred to in Table III above was

inducted into the army on 1 January in the year of his

induction. that he then underwent a period of basic

training before being sent to Vietnam (say 6 months).


that he was exposed to Agent Orange on the day of his

arrival in Vietnam and that his soft tissue sarcoma was

diagnosed in the middle of the year in which his death

occurred, then an entirely different picture emerges in

relation to latency periods. Only 3 cases (Cases 5, 7 and

9) would have latency periods of or in excess of 10



Cases 1, 2, 3, 6, 7 and 9 may have been in Vietnam at a

time when Agent Orange was not being or had not been 13 7 sprayed and, as already pointed out. Case 8 was

certainly there after spraying had ceased.


Notwithstanding the methodological and numerical problems,

the fact is that the study found a statistically

significant excess of soft tissue sarcomas.

Latency period criticisms are based on uncertain


The apparent excess may be due to chance, misdiagnosis or



Unlike the study by Greenwald et al, the study by Kogan et

al does not have regard to exposure to Agent Orange or to

any other chemical and so it says nothing about exposure

to phenoxy herbicides. The fact is that it adds nothing

to the available store of knowledge on the issue of

whether there is an association between exposure to

phenoxy herbicides and soft tissue sarcoma.

In the final analysis the Commission finds the data

inconsistent with other work and accordingly suspect. It

cannot. however. be ignored.

Summary re U.S. Personnel

Amongst United States personnel there is no evidence to

indicate adverse outcomes in deaths from cancer or in the

incidence of cancer which are statistically significant or

which would support the claim that United States Vietnam

Veterans have died or are dying or suffering from systemic

cancers, soft tissue sarcomas, malignant lymphomas or from

any other particular form of systemic cancer in higher

numbers or at earlier ages than expected for such a group.



1 Victoria - Departmental Study - (1985)

Affidavit evidence was given by Mr W C Parsons, Chairman

of the Vermin and Noxious Weeds Destruction Board in the

Department of Conservation, Forest and Lands in the State

of Victoria concerning the use in that Department of both

2.4- D and 2.4,5-T.138 This evidence indicated that both

2.4- D and 2,4,5-T had been used in the Department since

about 1950 and that experiments had been conducted as to

the effectiveness of a mixture of equal parts of 2,4-D and

2,4,5-T for the control of blackberry.

The growth of blackberry in Victoria is quite extensive

and its control takes approximately 20% of the effort

expended in controlling noxious weeds in that State. For

example, in the financial year 1982/1983, approximately

150,000 man hours were spent by employees of the

Department on the treatment of blackberry, much of the 139

work involving the use of 2,4,5-T.

Mr Parsons estimated that from 630 employees in the

Department, more than half at some time during the year

would be involved in the treatment of blackberry. This is


undertaken over 5 months during summer and early autumn.

A substantial number of employees had been with the

Department and involved in the spraying for many years, . or 140

some for up to 25 or 30 years.

Mr Parsons provided statistics for each of the years from

1959 to 1983 inclusive. These showed that a total of

1,062,000 kg of active ingredient of 2,4-D and 2,4.5-T

(375,200 kg of 2,4.5-T; 686,800 kg of 2.4,-D) were used by

the Department. This total does not include the 2,4-D

used in 1964/65 as the figures for this period were not


This long-term and extensive use of 2,4-D and 2.4,5-T. is

not believed to have given rise to health problems amongst

the employees engaged in the work, except for some

transient irritations. These occurred in few employees

and arose mainly from accidental spillages onto the skin

and into the eyes. However, in view of current

controversy about the use of 2,4,5-T a study of causes of

death and of selected health end points was undertaken.

In this study 2,181 men, who, during the years 1951 to

1970 had been employed by the Department spraying 2,4-D

and 2,4,5-T and who had been so employed for a period of


at least of 12 months, were followed through to 1984. Of

this total. 2,023 (93%) were traced. Of these, 1331 were

living and 692 were deceased.142

Some indication of the intensity of exposure of these

employees can be obtained from the evidence given by Mr J 143 W Bramford. He had been employed by the Department . 14 4

since August 1976 and. quite apart from regular

immersions in the tank of 2,4,5-T to adjust the agitator

(including one instance when he was locked in a tank for a 14 5

period by one of his more playful workmates), it is

clear that he was heavily exposed to spray. The users of

both knapsacks and vehicle mounted equipment were

frequently made very wet with the 2,4,5-T and 2,4-D.

Analysis of the causes of death amongst these men showed

that 169 (24%) of those who were deceased had died from

cancer. Of these, only one had died from soft tissue

sarcoma compared with expected deaths of 0.7. Five (5)

had died from bladder cancer compared with 5.7 expected, 9

from stomach cancer compared with 17.7 expected and 3 from

malignant lymphoma compared with 3.3 expected. In no

category of cancer was any statistically significant

excess observed. Of these figures, Dr Mathews said

"..... there appears to be no dramatic increase in any of

V I 1 1-73

the tumours that have previously been of concern in 146

relation to herbicide exposure."

When a comparison is made between the extent of exposure

of the Victorian sprayers and the sprayers involved in the

Swedish studies of Hardell and Axelson, the impact of this

statement becomes apparent. Dr Mathews confirmed that the

periods of exposure of the Victorian sprayers would have

been much longer in each year because of the longer


spraying season.

The data in this study and the conclusions on that data

expressed by Dr Mathews, are consistent with the results 14 8

of the studies in New Zealand by Smith et al (1984),

in Finland by Riihimaki et al ( 1984) 149 and in Sweden by 150 Hogstedt et al (1980).


This Victorian study provides strong evidence that

exposure to phenoxy herbicides does not increase deaths

from cancer in general or from rare forms of cancers such

as soft tissue sarcoma or malignant lymphoma.


E . 2 Occupational - Applicators in New Zealand 151

Smith et al (1982. 1983)

This case control study of males who suffered from soft

tissue sarcomas which had been reported by public

hospitals to the New Zealand Cancer Registry between 1976

and 1978 was undertaken by Dr Allan H Smith. He was then

a member of the Epidemiology Unit at the Wellington School

of Clinical Medicine in New Zealand, and is currently

Professor in the Department of Biomedical and


Environmental Health Sciences, Berkeley, U.S.A.

The purpose of the study was to determine whether any

association existed in New Zealand between the use of

phenoxy herbicides and the incidence of soft tissue


sarcoma. This study was designed to overcome

criticisms that had been validly made in scientific

circles of the design of Professor H a r d e n 1s


studies. Each case was matched with a control chosen

at random from other cancer patients in the Cancer


The use of cancer patients as controls was to overcome

both interviewer and recall bias. According to the

evidence given by Professor Smith it should not have any 155 corresponding disadvantage.


Originally there were 112 cases and a like number of

controls. However for various and appropriate reasons the

number of cases was reduced to 80 and the number of

controls to 92. Cases and controls were then classified

according to the extent of exposures to phenoxy acid


used in New



and 2,4-D are those most commonly

An initial report of the Study was published in 1982

before the question of exposure had been finally

assessed. This report had classified cases and controls 156 by reference to occupation.

After final assessment of exposure had been carried out

analysis revealed relative risks of 1.3 for those exposed

for at least 5 days subject to a latency period of 10

years, of 1.6 for those exposed for a minimum of 1 day

subject to a latency period of 5 years. Neither of these

figures was statistically significant. Professor Smith

described the analysis of those exposed for at least 5

days and subject to a latency period of 10 years as:

The more relevant exposures ... in that we have ignored any exposures in the 10 years prior to registration, in which period we know any

exposure is probably irrelevant . . . So, it is


only exposures over 10 years prior that are probably relevant.!·57

Professor Smith pointed out that professional applicators

in New Zealand spray for 6 or more months each year and

many have been spraying for a number of years.158 This

is to be contrasted with the Swedish situation:

Where spraying is confined to one month of the year when it is possible to do so ... So one

would therefore expect a high risk amongst professional sprayers in New Zealand. So that, even though there are not numerically very many of them, we would still have expected to find some if there was a real association between spraying 2,4,5-T and soft tissue sarcoma.-1 -59

When the relative risks were calculated after stratifying

for year of birth and whether or not the patient or a

relative was interviewed. rates increased from 1.3 to 1.4

(10 year latency period) and from 1.6 to 1.7 (5 year

latency period) . None of these relative risks were

statistically significant.

The size of the sample was such that the study was very

likely to detect an odds ratio around 3 and would

certainly have detected an odds ratio in the order of 5 or

6 as suggested by the Harden studies. (emphasis



It is very significant that no soft tissue sarcoma cases 161

were found amongst those who had worked as sprayers,

the most heavily exposed group.

In his oral testimony. Professor Smith drew attention to

the decrease in relative risk when regard was had to the

longer latency period. He said:

We also removed any exposure that was less than 5 days in total duration and the reason for that is that those exposures were also probably

irrelevant in that the notion that you could get cancer from less than five days' exposure is most implausible and certainly for those greater than five days should be greater. So by focusing on the relevant period in time and focusing more on relevant durations, if there was a trend and if that 1.6 had increased to, say, 2.2 in the fourth line, or 3, one would produce that as evidence for a real association going on. Instead of that, not only did we not find that it stayed the same, but it dropped ... 162

In his written statement 163 it is pointed out that:

(a) In addition to an unknown number of

unregistered applicators there were then 500 workers registered in New Zealand engaged in full time ground spraying and in total there were some 2,000 workers professionally

involved in phenoxy acetic acid herbicide spraying either from the air or on the

ground whose exposure was very much greater than that of the patients in the study;

(b) Not one case of soft tissue sarcoma was found in such a person.

(c) In fact none of the soft tissue sarcoma

cases had ever worked full time in spraying phenoxy acetic acid herbicide;

VI11 -78

(d) The documentation of the exposures to

2,4,5-T and 2,4-D of those involved in the study was at least as good as that in the

Harden studies.

Professor Smith's conclusion from his study is that:

The results concerning phenoxy acetic acid exposure from this case control study do not generally support the hypothesis that exposure to phenoxy acetic acid herbicides cause soft tissue sarcoma in New Zealand.164

and he deposed that his results are "inconsistent with the

Hardell study results".165

Professor Smith in his oral evidence given at San

Francisco in September 1984 confirmed that his analysis of

the occupations of the soft tissue sarcoma cases when

compared with controls do not show any significant


Professor Smith indicated that he had carried out an

update of his earlier papers (Exhibits 1239 and 1244) and

had prepared a paper in that regard which was then in

press. The paper was made available to the Commission and

became Exhibit 1540, being then restricted to Counsel

only. The Commission is aware that the paper has now been

published in the Journal of the National Cancer Institute,

U.S. November 1984, Vol 73, Pt 5, pp 1111-1117.


In Exhibit 1540 he further analysed the data involved in

his study (as updated) consequent upon two further

interviews of the 82 cases and 92 controls completed after

publication of Exhibit 1226.

Table 4 of Exhibit 1540 sets out the relative risks (odds

ratios) for various occupations and activities involving

exposure to phenoxy herbicide. A comparison of findings

(Table 9) by Harden in his Northern and Southern Sweden

studies with the studies in New Zealand confirms that:

They are ... in conflict, as one cannot defend, at least on statistical grounds, the proposition that the true odds ratio might lie somewhere in between those found in the two countries.

The New Zealand study has sufficient power to test the

hypothesis. In addition, the study was large enough to

show that, even if there were a real causal link, it is

statistically highly improbable that the true relative

risk would be greater than 2.9.167 The New Zealand

study also examined exposure to chlorophenols. The

findings "were also largely negative."168

Professor Smith presented as a highly qualified, careful,

detached, scientist in whose work this Commission has

VI1 1-80

confidence. The circumstances of exposure. its annual

duration, the number of years and the number of persons

exposed, together with an absence of methodological

blemishes in the study make the negative findings in this

study very persuasive in relation to the question of

association between soft tissue sarcoma and phenoxy

herbicide exposure. Further, comparison of the result with

those of the following two studies in Finland and Sweden

is also significant.


The New Zealand study provides no support for claims by

W A A that exposure to phenoxy herbicides is associated

with an increase in the incidence of or death from cancer

generally or from either soft tissue sarcoma or malignant


E .3 Finland - Riihimaki et al (1982, 1983)

This study involves an examination of the rate and causes

of death amongst a prospective cohort of 1926 men employed

in Finland for a period of at least 2 weeks in the years

between 1955 and 1971 in the spraying of 2,4-D and/or

2,4,5-T. The cohort was assembled in 1972 and was then

V I 11-81

followed until 1980 when the first report169 was


These men used portable manually operated or motor driven

knapsack rigs, except in the case of the Highway Authority

which used mainly moving vehicles. Throughout the 1950s

and 1960s the methods of application and the working

habits of the workers did not change appreciably and no

precautions were taken against exposure. The commercial

herbicide preparations most extensively used by the

applicators was a 2:1 mixture of emulsified esters of

2,4-D and 2,4,5-T dissolved in water, although at times

pure 2,4,5-T preparations were used .

Analysis of some old herbicide preparations used in

Finland in the years 1962 to 1967 suggest that the TCDD

content in the 2,4,5-T was in in the range 0.1 to 0.9


During prospective follow-up to 1980 the cohort

accumulated 16,694 person years of observation. In this

period 105 men died from natural causes against 155

expected. Deaths from cancer totalled 2_6 in the period 171

compared with an expected 36.5. When the period was

divided into two segments (1972-1976 and 1977-1980) 9

VI1 1-82

deaths from cancer were observed in the first segment

(18.1 expected) and 17 deaths were observed in the second

segment (18.4 expected).

Analysis by types of cancer „ with both ten and fifteen

year latency periods, did not reveal any significant

excesses in any area and in fact demonstrated that the 172

cancers were spread amongst a number of sites. The

analyses showed that with a latency period of not less

than 10 years, 20 deaths from cancer occurred compared

with 24.3 cancer deaths expected, giving a relative risk

of 0.82.

Three-quarters of the study population had 2 to 7 working

weeks of exposure whilst one-quarter had 8 weeks or

17 3

more. In this context it should be noted those

exposed for 8 weeks or more during 5 years had no higher

mortality from cancer than those who were less

. 174


When a latency period of 15 years is used in the

calculations, the number of cancers observed relative to

the number expected falls markedly in both the more and

the less heavily exposed sub-groups. In addition, the

incidence of cancer either falls or remains in


substantially the same proportions whether a nil latency

period or a 10 year latency period is used in the



(Extracted from Exhibit 852 Table 5)

Exposure Cause No


10 year Latency

15 year Latency

6 wks or

more during )A11 ) Natural 34 29 15

5 years ) A11 Cancer 8 8 1

8 wks or

more during ) A11 ) Natural 23 21 13

5 years ) A11 Cancer 5 5 1

A comparison of observed and expected deaths for cancer is

also revealing.



(Extracted from Exhibit 852 Table 5)

Exposure No

Latency 0 E

10 year Latency 0 E

15 year Latency 0 E

6 wks or )

more during ) 5 years 8 15.4 8 11.3 1 5

Ratio 0: E 0.51 0.70 0.17

8 wks or )

more during ) 5 years )

5 12.0 5 9.5 1

Ratio 0: E 0.41 0.52 0.19

(0 - Observed; E - Expected)

Thus the ratio of observed deaths from cancer falls in

each category (nil, 10 years, 15 years) with the increase

in exposure. This is consistent with and, indeed, mirrors

the findings of Professor A H Smith in New Zealand. Such

a trend not only does not suggest any association with

exposure to 2,4-D or 2,4,5-T, but rather is to the

_ 175


The contents of Table 4 of Exhibit 852 are also very

revealing. They show that in employment where there is

exposure, whether with the State Railways, Highway

Authority, Forestry Authority or Power Company, there is

no increase in the number of deaths from cancer amongst

spray applicators compared with that expected.



(Extracted from Exhibit 852 Table 4)

Employer Person


Observed Deaths

Expected Deaths

State Railways 2209 10 10.6

Highway Authority 2302 3 5.4

Forestry Authority 3343 6 6.9

Power Company 591 1 1.4

The finding in relation to Forestry Authority employees is

consistent with the results obtained in respect of Swedish

Foresters by Hogstedt and Westerlund (when a 10 year

latency period was allowed for).

Tumours as a cause of death amongst the exposed

applicators were analysed according to site and type. No

cases of soft tissue sarcoma or malignant lymphoma were

found and the number of "cases of the most common types of

tumour. namely lung cancer and stomach cancer closely

corresponded to the expected figures".176

In the abstract of the first report published in 1982 the

authors state:

No increase of cancer mortality was detected„ and the distribution of cancer types was


unremarkable. No cases of death from lymphomas or soft tissue sarcomas were found.177

and they conclude that:

The present results do not support the previously shown association between mixed herbicide exposures and increased overall cancer risks as far as specific exposures to 2,4-D and 2,4,5-T

are concerned. Neither did our observations support the finding that a markedly increased risk of lymphoma is associated with exposure to

chlorinated phenoxy acids.

A further follow-up of this cohort was undertaken. The

results were published by Riihimaki et al in 1983.178

This shows that again there were no cases of soft tissue

sarcoma or lymphoma. Analysis of the data (again with a

10 year latency period included in the calculation)

demonstrates that:

There were no significant differences between the observed and expected figures in relation to cancer incidence.'79

In addition the authors conclude:

With regard to mortality and morbidity of all cancers the rates were 70-82% of the national figures.180

Statistically. the studies of Riihimaki et al are

inconsistent with the results of the Swedish cohort

studies by Hardell and Axelson (i.e. Axelson's second


analysis of the data). The EPA Review Draft comments

about the studies of Riihimaki:

The smallest detectable risk ... was only 3.1 ... The study therefore appears powerful enough to detect relative risks even smaller than those seen in the Swedish and West German studies.181


These Finnish studies do not support the claims by W A A

concerning increased incidence of cancer generally, of

lymphoma or of soft tissue sarcoma. They are negative and

inconsistent with the conclusions of Hardell and Axelson.

E .4 Sweden - Hogstedt and Westerlund (1980)182

This study examines the incidence and causes of death

amongst men employed by a Swedish forestry company between

1954 and 1967 whose use of phenoxy herbicides commenced in

1953 , increased in 1954 and continued beyond 1967. To be

included in the study an employee had to have used phenoxy

herbicides for more than 4 days within the study period.

The study analyses were based on 158 men who were exposed

and 244 who were not. The 158 exposed men included a

sub-group of 16 foremen.


No deaths from soft tissue sarcoma or malignant lymphoma

were observed.

Analysis of the incidence and causes of death amongst the

exposed men showed that of a total of 29 deaths,183 8 184

occurred from cancer compared with 9.8 expected.

When a 10 year latency period was incorporated into the

analyses, the number of deaths from cancer amongst the

exposed group was 6 compared with 6.6 expected.188

Mortality amongst the exposed workers and unexposed

workers taken as a whole was similar and in both cases

less than in the population as a whole, although mortality

amongst the foreman sub-group was higher than expected

from the population rates. The authors draw attention to

the exposure of this sub-group to DDT and to the WHO Study

of 1979 which associates exposure to DDT with increases in

the freguency of lung and liver tumours.186 There was

no evidence in the exposed workers of any increase in

mortality with longer duration of exposure to herbicides.

The tumours in the exposed group were not clustered to any

site187 and a comparison of the exposed and the

unexposed groups reveals "a cause of death pattern that

largely agrees with national statistics expected


V I 11-89


This Swedish study does not support the claim by W A A that

the incidence of lymphomas, of soft tissue sarcomas or of

cancers generally is increased as a consequence of

exposure to Agent Orange or other phenoxy herbicides. The

findings are to the contrary and consistent with Smith's

New Zealand studies.

E .5 Axelson and Hardell

(i) Axelson

In 1973 Professor O. Axelson undertook a study of cancer

mortality amongst 348 railroad workers exposed to 2.4-D,

2,4,5-T. amitrole, diuron, monuron and/or a number of

other herbicides during the period from 1951 to

18 9

1972. As far as exposure to 2,4,5-T is concerned the

Commission notes that 1963 was the first year that a

herbicide containing 2,4,5-T was used by the Swedish

190 Railways. Professor Axelson1 s study is dealt with in detail commencing at p 95. Briefly, he reported that:

There seems to be a possible association between excess tumour mortality and exposure to amitrole


and its combinations, especially if a latent period is considered . . . on the contrary in the cohort exposed to phenoxy acids and combinations a fairly good agreement is found between the expected and the observed deaths independent of

the latent period (emphasis added).191

With latency periods of nil, 3 years and 5 years he found

that tumour incidence amongst phenoxy acids sprayers were

2 (2.97 expected), 2 (2.22 expected) and 2 (1.68 expected)

i.e. the incidence remained constant and was not different

from the number of tumours expected. No soft tissue

sarcomas or malignant lymphomas were found by him in any

of the members of the study cohort.

In 1977 Dr L Harden discovered and reported on 7 cases of 192 soft tissue sarcoma and Professor Axelson re-analysed 19 3 his own earlier data. He then concluded that:

If exposure to phenoxy acids alone was

assumed . . certain indications suggest

that an excess tumour mortality may also be referable to phenoxy acids.195


On the basis of these investigations it is not possible to select any special herbicide as carcinogenic, but the suspicion against phenoxy acids as a group of preparations may have become

somewhat increased, (emphasis added)196

V I11-91

This second conclusion was based on an assumption and was

tentatively expressed.

E . 5 (ii) Harden

(a) North Sweden Study

In 1977 Dr Hardell reported having found 7 patients with

soft tissue sarcoma and some exposure to phenoxy acids and

other herbicides in the 10 to 20 years preceding


diagnosis He then obtained funding to carry out a

study in North Sweden in respect of soft tissue sarcoma

and its possible relationship with exposure to

herbicides.198 This and his subsequent studies are

analysed in detail later in this section at p 96. Briefly

he reports:

There is a definitely increased risk of malignant mesenchymal soft tissue tumors in subjects handling phenoxy acids or chlorophenols .... It is. however, impossible to assess the effects of

the individual chemical substances separately, (emphasis added).199

This conclusion cannot properly be regarded as support for

the claims by W A A in relation to Agent Orange and soft

tissue sarcoma.


E .5 (i i) (b) South Sweden Study

As a consequence of the results obtained in his North

Sweden Study, Dr Harden commenced a study in the five

southern-most counties of Sweden in 1979 . This study

involved 38 deceased and 72 living cases (110) of soft

tissue sarcoma diagnosed and reported to the Cancer

Registry in the years 1974 to 1978. These cases were then

matched for age and place of residence to controls.

Following analysis of the data gathered (as to which see

comments later) Dr Hardell reports that the risk ratio for

soft tissue sarcoma:

After exposure to phenoxy acids or. chlorophenoIs, was found to be 5.1 in the matched material ... and after dissolving of matching 4.7. (emphasis added)200

His conclusion:

This study indicates that exposure to phenoxy acids may constitute a risk factor in the

development of (soft tissue sarcomas.)201

It should be noted that once again there was not, nor on

the data available, could there be, any singling out of

phenoxy acids.

V I 1 1-93

E .5 (ii)(c) Harden's Malignant Lymphoma Study

In January 1979, Dr Hardell reported having found 11

people with malignant , . 202 u

lymphoma who had been the

subject of "possible exposure to phenoxy acids or

chlorophenols . (emphasis ,, ^ 203


As a result of this finding he undertook a matched case

control study on patients in the age group 25 to 85 who

had been admitted at the Department of Oncology in Umea

during the years 1974 to 1978 suffering from either

non-Hodgkin's lymphoma or Hodgkin's Disease.

A total of 60 cases of Hodgkin's Disease and 105 cases of 204

non-Hodgkin's lymphoma were included in the study.

These cases were then matched by sex. age and place of

residence with controls. The study reports a relative risk

for exposure to phenoxy acids and chlorophenols of 4.8,

and that "no significant dose-response relationship could

205 be demonstrated". It should also be observed that 5 of the cases involved in this study had been exposed to

only multi-chlorinated phenoxy acids (MCPA) (nil of the

controls) and 7 cases were exposed only to 2,4-D - (2 of

the controls).


Both MCPA and 2,4-D are admitted by Dr Harden to be

"Phenoxy acids not contaminated by phenoxy chlorinated

dibenzo dioxins (PCDD) or phenoxy chlorinated

dibenzofurans (PCDF)"206 (i.e. they contain no TCDD).

Dr Hardell1s conclusion was:

The present investigation . . . suggests . . . that exposure to organic solvents,

chlorophenols and/or phenoxy acids

constitutes a risk factor for the incidence of malignant lymphoma. (emphasis added)207

E . 5 (iii) Axelson's Re-Analysis of his 1973 Study &

Up-Date Study

As stated earlier in E .5.(i). Professor O. Axelson in 1972

undertook a study (first analysis) of cancer mortality

amongst railway workers following rumours of excess lung

cancer mortality amongst those exposed to 2,4-D and

2.4.5-T.208 In an endeavour to determine if the facts

justified the rumour he examined mortality amongst a

cohort of 348 railway workers. Inclusion in the cohort

depended upon a worker having a given duration of exposure

to herbicides during the period from 1951 to 1971.

On analysis of the data Professor Axelson found a total of

18 deaths from all causes compared with 20.54 expected and


a total of 6 deaths from tumours compared with 4.88

expected when a zero latency period was allowed for. For

tumours, therefore, the relative risk was close to unity

(namely 1.2) and not statistically significant.

As far as deaths from lung cancer were concerned, 0.83 had

been expected and 2 were found. Neither of these occurred

in the sub-cohort said to be exposed to phenoxy acids and

both of them (cases 256 and 257) occurred in people who 209

had been smokers for a long time, perhaps decades.

In this first analysis Professor Axelson concluded that:

There seems to be a possible association between excess tumour mortality and exposure to amitrole and its combinations, especially if a latent period is considered. On the contrary in the cohort exposed to phenoxy acids and combinations a fairly good agreement is found between the expected and observed deaths independent of the latent period.(emphasis added)210

Thus his first analysis did not demonstrate an association

between phenoxy herbicides and an increase in cancer 211 .

incidence and in fact Professor Axelson admitted that

his conclusion was:

As close as one can get to a negative

epidemiologically.212 1 1

VI11- 9 6

After becoming aware of the observations by Dr Hardell,

i.e. "the first clinical report about soft tissue sarcoma 213

and phenoxy herbicide exposures" Professor Axelson

re-analysed his data.

Two (2) cancer cases which in his first analysis had been

classified as having been exposed to amitrole were now

stated to be cases in which "it is possible that phenoxy

acid exposure alone may have occurred" (emphasis 2i4 added). When these cases were classified in this way

(i.e. such an assumption was made) an excess of mortality

from tumours was obtained amongst individuals with phenoxy

• ^ 21B acid exposure.

Thus to produce an excess of tumours in the relevant

subcohort a change in classification had to be effected,

i.e. the finding of an excess in that cohort depends on

only two (2) cases - the 2 who were re-classified.

It is no doubt because of considerations such as these

that the re-analysis concedes that:

(i) The material is unsatisfactorily small;216


(ii) The exposure pattern is complex; and

VI11 - 97

(iii) It is not possible to clearly select any special

herbicide as carcinogenic.218

The conclusions in the re-analysis are quite nebulous in

relation to the subcohort said to have been exposed to

phenoxy acids:

Certain indications suggest that an excess tumour mortality may also be referable to phenoxy acids.2· * - *

On the basis of these investigations it is not possible to clearly select any special pesticide as carcinogenic but the suspicion against the phenoxy acids as a group of preparations may have become somewhat increased.(emphasis added)22®

Professor Axelson later followed the cohort through to

October 1978 and. using the same exposure data as before,

he undertook another analysis (update). This time,

however. he introduced a latency period of 10 years and

reached a conclusion which was different from the

conclusion in his first analysis. There are two (2)

reports of his update in evidence. 221 They are not

identical. In both versions however the following

statement is made:

Tumour incidence was also updated but provided little additional information. The incidence data have therefore been omitted from this presentation.222


This omission is interesting since in his first analysis

Professor Axelson states that:

The tumour incidence data, however, may be more relevant than the mortality data . . .223 ;

It is reasonable to conclude that nothing abnormal was

found in relation to tumour incidence in the update - an

indicator which Professor Axelson had regarded as "more

relevant" than mortality.

In his update Professor Axelson concluded that:



No specific type of tumour is predominating.

(ii) The aspects of causal relationship are rather 225 unclear.

(iii) Those exposed in 1962 or later did not show a

clear excess mortality.226

(iv) This finding might be interpreted in different

ways, eg, the variety of herbicides during the

early period could be of importance and/or the

work conditions may have been more primitive and

the herbicide handling more careless, resulting 227

in a higher degree of exposure.


These findings do not single out phenoxy acids. Indeed

the comment in (iv) above together with the discussion

concerning Table I to Exhibit 763 (at p 115 et seq) and

the absence of any other data as to the herbicides used by

the Swedish railways suggests that Professor Axelson1s

work in fact exculpates 2,4,5-T. At the very least, there

are serious problems in his work arising from confounding


It is interesting to note that nowhere in the initial

submission of W A A or in its submission on Toxicology or

in its final submission is any reference made to Professor

Axelson1s first analysis. his re-analysis, his update or

to any of his conclusions and no reliance whatsoever has

been placed by W A A on his work or on his evidence,

although he was called as a witness before the Commission.

E.5 (iii)(a) Critical Examination of Axelson Studies

However, because Professor Axelson seems to suggest that

his first analysis, re-analysis and update in some way

support Dr Hardell's findings, they should be examined.

In carrying out this examination it is desirable to look



(i i) The lack of expertise of those engaged in

Axelson's study;

(iii) The inaccuracy or uncertainty of the data

involving problems in the selection of the

cohort. in the determination of exposure and

arising from wrong diagnoses;

(iv) The weakness of the conclusions; and

(v) The inconsistency between the results obtained by

Professor Axelson and those obtained by Dr


E.5 (iii)(a)(i) Epidemiology and its Purpose

Epidemiology is a discipline which can be used to predict

the risk of individuals in a human population developing a

particular disorder or to demonstrate an apparent

association or lack of association between an observed

health phenomenon and a given substance or

substances.228 Epidemiology involves the mathematical

testing of a possible association between an observed

(i) Epidemiology and its purposes;


health phenomenon and a postulated source of that



However, the demonstration of a mathematical association

between the health phenomenon and the postulated source

does not of itself establish causal relationship. Indeed

it has been said that an epidemiological result cannot be

proof positive of causal relationship. it merely 230

demonstrates statistical association. But as stated

earlier at Chapter 11-16 in this Report, it can provide 231 strong evidence of causality.

A statistical association may reflect a true association

or only an indirect or artefactual association. An

indirect association between a factor and a disease may be

produced by the presence of another factor, either known

or unknown, which is common to both the disease and the

factor which has been demonstrated to be statistically

associated with the disease (confounding).

An artefactual association can result from use of a

methodology which involves some form of bias. The bias

may be involved in the selection of the individuals used

in the study (selection bias) or it may come about because

of the manner in which, or person by whom, the information


Bothfor the study is elicited ( observational bias) ,

forms of bias can give rise to invalid data and can

corrupt the conclusion.

According to Professor Axelson observational bias may take

different forms. It may be bias of the subject arising,

for example, out of memory problems or it may be bias in

the investigator himself.


So that observer bias includes investigator bias and subject bias? -- Yes. 233

People who are suffering from a particular disease are

more apt to think about their condition and to try to find

an explanation in exposure to various situations or agents 234

which they believe may have caused their disease.

This can affect the accuracy of the data obtained.

Likewise the accuracy of data obtained may be affected by

the subject having knowledge of the result which the

investigator is seeking to achieve. just as it may be

affected by the data collector being made aware of the

object of the study.


In assessing the result of an epidemiological study the

strength of the association, the presence of a

dose/response relationship, confirmation or replication by

others and the specific nature of the response are all

relevant factors, as is Knowledge that the data on which

23 5

the result is based are accurate. So if the data on

which a study is based are subject to doubt, then the

conclusions of the study must be treated as doubtful at


Proper epidemiological practice requires that the events

which generate an hypothesis should be kept separate from 237

the events used to test that hypothesis. Professor

Axelson himself, after some quibbling about the matter.

agreed that:

To my understanding there is a tendency for some statisticians to take the view that one should not use individuals or such observations that have started - yes, started the study - I would not say generated the hypothesis because that is generated from facts outside as well.238

However the analysis presently under consideration

breached this basic principle, as is clear from the



What generated this study of course was an assertion of three lung cancers was it not? -- Yes.

One of whom was subsequently excluded... -- Yes

That left you with two and they were included as cases in your study were they not? --

Yes. 239

E.5 (iii) (a) (ii) Lack of Expertise

A questionnaire was sent out in conjunction with the

gathering of data. At the time the questionnaire was

designed and sent out and at the time the data were

gathered. Professor Axelson had no university

qualifications or professional training in epidemiology or

statistics,240 and he was not a mathematician.241

Indeed he admitted that at the time he undertook his work

he had not had any training whatsoever in epidemiology.


other than self-training - by reading books. It is

true that later in 1972 he did attend 14 days of lectures

in Finland but that, as he conceded, "did not have any


effect because it came later".

This lack of training and of experience may well explain

the absence of a written protocol for the study 244 and

the absence of any formal development of a plan for the

study detailing its design, its purpose and how it would

24 5

be carried out.

V I 11-105

E.5 (iii ) (a) (iii) Inaccuracy of the Data

Where exposure to a given substance is the variable in a

study the need for the data about exposure to be as

accurate as possible is fundamental, since inaccuracy in

those data can affect the correctness of the

, . 246


The data collection for the first analysis was undertaken

at a time when there was considerable controversy in

Sweden about the use of phenoxy herbicides, particularly

2,4,5-T and a great deal of adverse publicity both in the

newspapers and on television. In addition, there were

rumours about an association between cancer and use of 247

such herbicides amongst railway workers. Thus the

climate in the community generally and in the study

population in particular raises a serious question of

information bias.

People might have said I was exposed to Phenoxy acids, even though the employment record suggested that it was Amitrole?-- Yes, that was a possibility as we said before . . .248

This problem, was heightened amongst railway workers by 24 9

the distribution of questionnaires, the first

question in which was:


Have you experienced any health problems which you relate to herbicides?250

The questionnaires were sent first to the Chief Physician

of the Railways, from him to the various districts in the


railway network and then from the district to the men.

Whilst Professor Axelson did not know who it was that 252

actually filled out the questionnaires it is clear

that they went to the bulk of those included in the

253 cohort and to others as well, all of whom had asked for health check-ups. This appears from Professor 254 Axelson1s evidence before the Royal Commission and

from his 1980 EPA deposition where he admitted that the

workers had asked for health check-ups at the relevant


They all asked for health check-ups at the time so we had to, in a way, provide some camouflage for what we were actually looking at and what we

were interested in.255

The questionnaires were distributed at the very time when

lists of the workers said to have been exposed to

herbicides in the period 1951 to 1972 were being compiled

by, or had been requested from the railways.256 An 257

examination of a copy of the questionnaire reveals

that it sought no information concerning the extent of


exposure to any particular herbicide or chemical and that

the instructions which formed part of it did not

differentiate between different herbicides. They were all

grouped together as "weed and brush killers". In view of

the form of the instructions which appeared at the head of

the questionnaire, and the first question in it, it is not

surprising that Professor Axelson had to concede that it

was "fairly clear what the investigation was all

about"258 and that "We did not camouflage that at


Despite the fact that the distribution of the

questionnaires at such a sensitive time was likely to have

exacerbated the problem of recall bias, the information

which they contained in relation to exposure was not 260 .

used. Instead the cohort was compiled from exposure

information derived from lists supplied by the Swedish


It took a number of attempts to compile the study cohort.

In the first attempt information as to exposure was sought

from the Swedish railways. The information was provided

in lists which were the result of "a joint assessment of

exposure from the railway company and from the trade union

representatives".261 That first or basic set of lists

VI11- 108

was of little use "because we got in material which was

fairly bad from some districts"262 and "so, therefore we

repeated the whole procedure once more".263

To this end instructions were given for the compilation of

another set of lists:

... At least in this respect, because we had a bad experience there from the first attempt, so we had to be more detailed in the second one before we gave up.264

MR O'KEEFE: Doctor, let me be clear about something: You got a first lot of lists from the railways which were rejected because of differences in the various data provided from different districts, correct? -- Yes.

That is -- incomplete and inconsistent --Yes

THE COMMISSIONER: That is list one.

MR O'KEEFE: That is list one? -- Yes. List two then came and in respect of that you made some mechanical exclusions? --Yes.

That is. you excluded everybody who had only exposure up to an including 45 days? --Yes.

That was a purely mechanical process? --Yes.

Then you submitted a manuscript which became the subject of publicity? -- Yes.

THE COMMISSIONER: Excuse me, where did that manuscript go to, I did not make a note of it? Where did you send that manuscript? -- To the National Board of Occupational

Safety and Health.


MR O'KEEFE: That was the subject of media publicity? -- Yes.

And that led to a number of persons coming forward claiming that they should have been in the study? -- Yes.

THE COMMISSIONER: Excuse me while I just interrupt. Those people coming forward to the Railways?

MR O'KEEFE: To the Railways? - Yes.

That led to a third lot of lists; correct? -- Yes.265

In compiling the second set of lists there were instances

in which the union insisted that the names of particular

workers should be added and, after discussion, additions

were made. The lists were then signed as "mutally agreed"

between the railways and the unions;266 however they

probably included workers whose names were put on despite

the fact that they did not appear in the railways' records

as persons who had been exposed for 45 days or more.

Do I understand then that some of the persons included in the list as a result of disputation by the unions were persons who, in the railway records, did not appear as persons who had been exposed for 45 days or more? -- Yes, well, were probably such

individuals which were not appearing in the files as exposed ... 267

Professor Axelson admitted that he did not know how many

of the workers fell into this category.268


In addition Professor Axelson conceded that the people at

the Swedish railways who gathered the material had no

training in data collection269 and, after i some

hesitation, admitted tha t he may have agreed at an

Environmental Protection Agency (EPA) hearing in

Washington in 1980 that the way in which exposure was

determined was for the people compiling the lists to look

at the salaries that were paid, deduce from them what work

tasks were undertaken by the various workers and to imply


from that what exposure there could have been.


A perusal of the EPA transcript shows that he did so

agree. It was also admitted by Professor Axelson that the

process involved in the collection of exposure data was "a 272 sort of guessing" and tha t the extent to which any

particular person may have been exposed was also


"guesstimated". In relation to exposure the

following answer is also significant:

And would you agree with me that in the railway records there was no indication in relation to any individual of any particular

preparation or preparations of a chemical nature which were in fact involved in his job? -- I believe not.274

The people who compiled the lists relating to exposure had

problems with their task:


There were, obviously, people who were not possible exactly to classify in terms of exposure and for those they had to come up with some sort of estimation and when I say they, it refers to representatives of the railway company and of the

trade unions.275

No checking of exposure was ever undertaken by Professor

Axelson or his assistants.276 Indeed the data which

were provided in the second and third sets of lists were

accepted with, "no further investigation in relation to

the cohort". ... "not even a spot check in relation to one 277 or some random number". and

. . . except for excluding the persons who were exactly 45 days of alleged exposure and some accidental inclusions of those who were under 45 days. (Professor Axelson) took no part in the selection of the cohort (emphasis added).. 170

In these circumstances it can readily be understood why

Professor Axelson took no responsibility for the accuracy

of the exposure data:

You did not yourself take any steps to verify the accuracy of the data supplied in respect of exposure did you? -- I did not take any

steps, no.

The resources available for this study were small, were they not? -- Yes.

... you were given data and you accepted it and used it? -- Yes.279

V I I 1-112

Professor Axelson1s concern about the accuracy of the data

is manifest by his decision "to modify the cohort

selection basis after the data had been gathered"280 by

increasing it (perhaps somewhat naively) from a minimum of

45 days exposure to a minimum of 46 days of exposure.

The information relating to phenoxy herbicide exposure

used in the first analysis, the re-analysis and the update

was obtained from the lists supplied by the Swedish

railways.281 These lists set out details of potential

exposure under three headings namely. Group A, Group O and

Group F. Whilst Professor Axelson said that 'A' in the

Group A heading stood for amitrole,282 an examination of

Exhibit 1370 shows that a number of specific herbicides

are referred to by name under that heading.

The heading to Group O was "other herbicides".283 the

heading to Group F was said by Professor Axelson to stand 284 for "phenoxy acids". However this turned out not to

be so as emerged when the Professor was asked:

Group F was all brush killers? -- I think it is important at this stage to define brush killer. We have not the corresponding word in Swedish, you see.

Was not the word at the top of that column a word which you advised Mr Warren, in 1980. meant all brush killers? -- Yes, that would

probably be the best translation.285

V I 11-113

by the Swedish railways in the years against which entries

are made, is hard to understand. Either Table 1 indicates

how much herbicide was used in years dealt with in the

Table or it does not. If it does so indicate, then it is

destructive of the conclusions set out in the update for

the reasons indicated in this part of the Report.

If it does not indicate how much herbicide was used by the

Swedish railways in the years referred to, then there are

two further possibilities, namely Professor Axelson either

believed that it did so indicate or he did not so

believe. It is inconceivable that he did not so believe.

However, if that is the situation, then the inclusion of

the Table in his study would amount to a gross departure

from the standards of scientific honesty which one is

entitled to expect. If he did believe the contents of

Table 1 to be accurate, even though they are not accurate,

such a situation underlines the loose, untrained and

inexperienced approach which was involved in the study and

emphasises the totally unsatisfactory nature of the

exposure data.

Because of the significance of Professor Axelson1s change

of stance in relation to this matter the Commission asked:


THE COMMISSIONER: I am rather shattered to find that Table there, which I have studied in some depth, to find that it has got nothing to do with the price of fish - if I may use

a colloquialism of this country. I had

assumed the men had been exposed, or there wgs potential exposure, to that amount of tnat chemical and I find I am wrong in

making such an assumption? -- Yes.It does not tell explicitly the exposure of the men . . 295

However, a little later Professor Axelson said that the

Table " reflects. indirectly at least, how much of 296 herbicides that was used".

This Table shows that no herbicide containing 2,4,5-T or

any substance that could potentially contain 2,3,7,8 -297

TCDD was used by the Swedish Railways prior to 1963.

When Professor Axelson undertook the updating of his data

he ascertained that two cases (268 and 334 ) which had

previously been counted as cancers298 had not had a

diagnosis of cancer confirmed by the National Central


Bureau of Statistics. These cases were therefore

excluded from the update300 on the grounds of


Can I help you this way? If you might check me - there are 23 cases in Table IV? -- Yes.

If you go to Table I which shows observed and expected number of deaths from tumours you have 17? -- Yes.


If you go back to Table IV you have case 138 is not dead? -- Yes.

So he is one exclusion from the 23? -- Yes.

Case 388 is not dead. So he is an exclusion? -- Yes.

Case 135 I missed - he is not dead either; right -- Yes.

And case 098 is not dead; right? -- No.

So there are four that are not dead; correct? -- Yes.

And 4 from 23 equals 19? -- Yes.

So you have two to account for? -- Additional, yes.

And it would seem quite proper to infer that you had excluded 268 and 334 on the grounds of wrong diagnosis? -- Yes.301

Error in diagnosis can be quite important in a study in


which the numbers involved are small, as is the case


with the work of Professor Axelson.

The "other exposures" of the workers involved in the two 304

analyses and the update were unknown. Indeed, the

update expressly stated that:

It should also be made clear that the recognized exposures have been mixed with other unknown exposures: unfortunately there was no possibility of sub-dividing the material in this respect.

(emphasis added)^0^ 5


In relation to Professor Axelson1s work it must be

observed that:

It was carried out by untrained and inexperienced


There was no proper protocol;

It was based on a cohort, the selection of which

was completely outside the control and

supervision of those conducting the study;

No responsibility for the accuracy of the

exposure data was accepted;

Those conducting the work used data which they

believed may have been inaccurate.

The data on which the work was based were


Thus, in relation to the information on which Professor

Axelson's study was based the Commission concludes:

(a) It was obtained by untrained and inexperienced



(b) There was no proper protocol to guide them in

their data collection;

(c) Exposure data were obtained in a climate of

considerable controversy concerning the very

subject of the study;

(d) The questionnaires distributed made the purpose

of the information apparent;

(e) In a number of cases the data relating to

exposure were little more than a guess;

(f) Persons who were counted as having been exposed

to phenoxy acids may not have had any such


(g) The number and extent of confounding factors was


(h) There was inaccuracy in diagnosis data.

E .5 (iii)(a)(iv) Weakness of the Conclusions

As has been noted already, the conclusions in relation to

the effect of phenoxy herbicide exposure reached by


Professor Axelson in each of the three studies are

different; the first negative, the second raising a fairly

nebulous possibility based upon an assumption as to

exposure, the third dependent on small numbers and

involving a lack of specificity and accuracy in exposure


These considerations alone would be adequate to justify

the conclusion that Professor Axelson1s work does not make

any worthwhile contribution to the resolution of the

question whether or not there is an association between

exposure to phenoxy herbicides and the incidence of

cancer. Such a conclusion is supported by Dr Armstrong

who analysed Professor Axelson's work.306

There are. however, additional factors within the Axelson

material which reinforce such a conclusion.

In his first analysis Professor Axelson used a maximum

latency period of 5 years. In his later analysis the

latency period was changed to 10 years to bring it "more 307

in line with standard epidemiological approaches".

Ten years was a period which "seemed to be a reasonable

latency or induction latency period".308

V I 11-121

After lengthy cross-examination concerning the Tables

in Exhibits 763 and .11, Professor Axelson conceded that if

he had adopted the same latency period in his first

analysis as he did in the update, the total number of

tumours observed in the cohort said to have been exposed

to phenoxy acids and combinations would be reduced from 310

five (5) to three (3) namely cases 173, 328 and 3 88.

They are the only cases which had both the potential for

exposure to phenoxy acids and combinations as well as the

necessary minimum latency period.

This reduced number of three (3) is to be compared with 311 2.83 expected.

And the relative risk then is 1.06? ... Yes, insignificant in any case both as it stands and as it would be now. (emphasis



This conclusion would flow through to the re-analysis.

When a minimum latency period of 10 years is adopted and

the data set out in Table 1 to Exhibit 763 are accepted as

correct, the number of tumours observed in the so-called

"phenoxy acid" sub-cohort in the update is reduced to 314

three (3) compared with 3.14 expected. This gives a 3 1 5 315 relative risk of 0.95. However the figure of 3.14



for the number of tumours expected would need some

reduction316 but, according to Professor Axelson, the

extent of the reduction could not be calculated from the 317 data provided. He was unable to agree that the

adjustment would probably result in a figure of between

2.22 and 2.44 (Relative Risk of 1.35 and 1.22).318 It

is the Commission's view that figures of this kind can

fairly be derived from the data presented and that they

give an indication of the number of tumours to be expected.

As can be seen from Table VII there is only one instance

(case 328 - cancer of the prostate) in which the first 319 exposure occurred after 1963.


Derived from Exhibit 11 Table 4 and Transcript 3829-3830.

Case First year of exposure

No assigned by Axelson

127 1957

172 1958

173 1958

268 1960

317 1958

328 1967

388 1961

113 1960


In the original analysis, the re-analysis and the update,

the number on which the relative risks for the subcohort

said to be exposed to phenoxy acids and combinations were

calculated are small and the existence of an excess

depends entirely on the classification as to exposure

which is assigned to given cases.

The exposure classifications have been changed in the two

analyses as far as some cases are concerned and, in any

event, as has been demonstrated above, the exposure data

are inadequate. unreliable and non-specific.

E .5 (iii)(a)(v) Differences between Axelson and Harden

The following are differences between Axelson and

Hardell. It should be remembered that they are not

inconsistencies since Axelson1s results were obtained by a

small cohort study whilst Hardell1s were by case control

studies. The differences merely show that the studies do

not support each other.

(a) Hardell claims that there is an increased risk of soft

tissue sarcoma and malignant lymphoma both of the

Hodgkins and non-Hodgkins types from exposure to

phenoxy herbicides. In Axelson1s work there were no


cases of soft tissue sarcoma or of non-Hodgkins

lymphoma found and only one case of Hodgkins Disease.

(b) It is claimed that Hardell1s work indicates

specificity in the type of tumour produced by phenoxy

herbicide exposure namely, soft tissue sarcoma and

malignant lymphoma. However in his update Axelson

concludes that, "No specific type of tumour can be

considered to predominate."320

(c) Hardell asserts relative risks in excess of 4 for the

association of soft tissue sarcoma and phenoxy

herbicide exposure. In his first analysis Axelson's

finding in relation to the "phenoxy acids and

combination" subcohort was as close to a negative as 3 21 one can get. . In his re-analysis, even when the

exposure classification was changed for two (2) cases

so as to produce an excess of tumours observed over

those expected in the subcohort exposed to phenoxy

acids and combinations, the relative risk was still

only 1.6 and not statistically significant.

In his update the relative risk when properly

determined (ie with exclusion of wrong diagnoses,

adoption of a 10 year latency period and using the

VI11-12 5

data detailed in Exhibit 763 Table 1) lies close to

unity and is certainly not statistically significant.


The Commission concludes that because of the number of

faults, together with the extent of uncertainty or error

in the data. Professor Axelson1s doubtful positive

conclusions should be given little or no weight.

E .5.(vi) Critical Examination of Hardell Studies

In sections E .5(ii)(a), (b) and (c) above, three studies

of Dr Lennart Hardell concerning the effects of herbicide

exposure on various populations in Sweden were discussed.

This section of the Report contains a critical examination

of those and two other Hardell studies:

(a) The North Sweden Study concerning the incidence of 3 22 soft tissue sarcoma;

(b) The South Sweden Study also concerning the incidence 323 of soft tissue sarcoma; 3 2 4


(c) The Malignant Lymphoma Study;


(d) The Nasopharyngeal Cancer Study;

(e) The Colon Cancer Study;326 which had as its

essential purpose the verification of the conclusions

in the studies referred to in paragraphs (a), (b) and 327 (c) above.

Before the North Sweden study began Dr Harden had

published his preliminary clinical observations which

provided the impetus for the study.328 He did the same 329

before undertaking the lymphoma study.

Since Dr Hardell's studies are the only epidemiological

evidence which support an association between exposure of

humans to phenoxy acids and an increased incidence of soft

tissue sarcoma and malignant lymphoma. the Commission will

examine them closely. It has noted the comments of Chief

Judge Weinstein in the Agent Orange Product Liability


Litigation that the s tudies "were widely recognised as 3 30 flawed" and the remarks of Mr Justice Nunn in the

„ „ 331

Nova Scotia case where he said that the Hardell

studies were "unacceptable as proof of the results

claimed" and found 1 332 "that they cannot be taken at face i „ 332 value".


E.5(vi)(a) The North Sweden Study

In August 1977 Dr Harden reported clinical observation of

seven (7) initial and two (2) later instances of soft

tissue sarcoma found amongst people said to have been

exposed to phenoxy herbicides during the 10 to 20 years

preceding diagnosis. These were patients at the Umea 3 3 3

Hospital Cancer Centre between 1970 and 1976. He


referred to these cases as "a series".

3 3 5

In this first report Dr Hardell states that between

1970 and 1976, 87 patients with malignant mesenchymal

tumours (soft tissue sarcomas - STS) had attended the Umea

Oncological Centre. Fifty-five of these (63%) were men

and of these 43 were still working. Nine (9) of the men

who were still working were forestry workers. four (4)

worked in agriculture and forestry and six (6) at saw

mills or in the pulp industry - a total of 19 or 42.2%.

This mix of occupations is not surprising in view of the

geographic situation of Umea in Sweden's north and the

extent to which forestry, activities associated with

forestry and, to a lesser extent, agricultural pursuits

are undertaken in the region. Hardell wrote.

Exposure to phenoxy acids is probably relatively common in the three forest districts which made up the Umea clinic's catchment area.336


This leads to doubt about the conclusions that there was

an excess amongst forestry workers.

In the report of his clinical observations in relation to

the seven (7) initial cases Dr Hardell states that "the

exposure was quite massive in all cases" (emphasis added) 3 37 and that the exposure of five (5) of them was

"direct"338 In his North Sweden study these seven (7)

initial cases are described as "a series of patients with

soft tissue sarcomas and massive exposure to phenoxy 3 3 9 acids"(emphasis added). An examination of the

histories of these patients disclosed that these

statements are inaccurate. His use of the word "massive"

is revealing; so also perhaps, is the editing of the case

histories from the hospital records.340

Case 1 had sprayed with 2,4,5-T for barely one week in

1963 and 1964, two weeks in 1965, one month in 1966, one

month in 1967 and 2 weeks in 1968 . He was diagnosed as

suffering from soft tissue sarcoma in 1976. An

appropriate and conservative latency period of ten years

suggests that the relevant exposure for Case 1 is less

than three weeks of herbicide spraying.

Case 2 sprayed for 3 weeks in each of three or four

summers during the 1950s and 1960. He was diagnosed as

V I 1 1-129

suffering from soft tissue sarcoma in 1976. Thus his

exposure was a maximum of twelve weeks spread over twenty


Case 3 was diagnosed as suffering from soft tissue sarcoma

in 1976. His only association with herbicides was that

"thousand litres of phenoxy acids" (plural; emphasis

added) were stored for about eleven months each year in

the workshop in which he was employed. Some of these had

been "stored partly in open containers" so that he could

smell the material and i t is stated that he "had also 341

handled the preparation". (emphasis added)

In what circumstances or for what period the containers

were left open (e.g. whether this was only when they were

being handled), the extent and manner of handling of "the

preparation" (singular) by the patient, what "the

preparation" handled was or what the "phenoxy acids"

(plural) were is not revealed.

Similarly whether or not the man was wearing gloves or

other protective clothing when he handled the

. . 342 .

herbicide is not stated in the report and remains



The description does not indicate a great extent of 3 4 3

exposure. Dr Harden saw this patient and appears to


have taken a part in his extended treatment. He must

therefore have known the extent of the exposure, if his

approach had been rigorous. He was however unable to help 345 the Commission.

He conceded however that:


He (ie Case 3) never sprayed.

and that he classified him as unexposed for the purposes 347 of his North Sweden Study.

Case 4 was exposed to unspecified phenoxy acids for about

twenty days per annum in the years between 1961 and 1966.

As the date of his diagnosis is not revealed the precise

extent of his relevant exposure cannot be calculated. If

one takes the date of observation, ie 1976 as the date of

diagnosis, the exposure is probably less than 20 days,

considerably less than the New Zealand sprayers studied by . , 348

Smith et al.

Case 5 was exposed to herbicides for a total of six weeks

in 1945 and 1946. The nature of the herbicides is not


revealed but having regard to the date on which 2,4-D and

2,4.5-T were introduced (almost certainly after 1950) into 3 4 9 ,

Sweden, it is most improbable that they were phenoxy

acids. However. Case 5 did spray phenoxy acids for two

weeks in each of the years 1960 to 1968. He was diagnosed

as suffering from soft tissue sarcoma in 1974. Adopting

an appropriate minimum latency period of ten years, his

exposure was a maximum of eight weeks spread over four

years - again considerably less than the New Zealand

sprayers studied by Smith et al.

Case 6 was diagnosed in 1972 as having soft tissue

sarcoma. The only evidence that he had any exposure was

that "over 4-5 summers in the fifties . . . he cut and

removed grass along the verges sprayed by the Highway

Department" and that "he had worked and 1 ived in forests 3 50 sprayed with phenoxy acids".

In Dr Hardell' s report of this case neither the

relationship between the time of cutting of the grass and

the time of herbicide spraying nor the actual herbicide

used is revealed. However, it is very unlikely that

2,4,5-T would have been used for grass killing since it is

guite unsuitable for that purpose. It is also improbable

that the grass would have been cut before the herbicide 1 1

VI11 -13 2

had time to take effect and kill the grass. To spray

herbicide on grass and then cut the grass immediately

would be pointless. Dr Hardell concedes as much when he

says, "there were some days in between" the two

_ 351 events.

During the cross-examination he seemed to shift his ground


. was the bushes which were sprayed and not the grass actually.

So he cut grass in the area where bushes had been sprayed some time before?---- Yes".

(emphasis added)352

As to living and working in a forest:

Did you ever define before writing this article where in relation to his place of residence the spraying in the forest had taken

place?--- No.

So whether it was a kilometre or two kilometres or right next door to his house was not a

matter that you had data on?--- Correct.

And whether or not he worked within those sort of distances of a place which had been sprayed was not a matter that you had data

on?--- Correct.

And whether or not the spraying had taken place a day or a week or months before he worked in a particular place was not a matter on which you had data?--- Correct.

And whether or not he was ever actually in contact with the substances that had been sprayed in the forest was not a matter on which you had data?--- Correct.353



Whether the forest was sprayed once a year or once every five years and where it was

sprayed in relation to him and when he was in the forest in relation to the spraying was not a matter that you could deduce from the record?--- Yes".354

3 55

This man had never sprayed himself.

Case 7 was only diagnosed as having soft tissue sarcoma in

1970, but his symptoms extended back to mid-1969. He "had

worked in forests up to 1970"356 where "he carried out

thinning in phenoxy acid sprayed areas from 1956 to the

3 57

beginning of the 1960s." (emphasis added) Would he

have been thinning immediately after spraying had taken

place? The report says nothing about the actual temporal

association and Dr Hardell was unable to assist the

Commission. 358 He also worked "in conjunction with

spraying a few

, _ 359

years later. What this means is

unknown as is the nature of the spray or sprays used.

This sparse data360 was described as:

... is a fair summary of the total information you were able to get about exposure of that man?--- Yes . 361

Thus it is doubtful that there was any real exposure of

Cases 3, 5, 6 and 7 but if there was any, it was minimal.


What does one conclude then about Dr H a r d e n 1 s

description, where in his study he states:

What is remarkable ... is that a rare type of

tumour is involved, that the exposure was quite massive in all cases, that the latency period corresponds .... (emphasis added).362

When confronted Dr Hardell sought to escape from his


... what I meant with massive by that time was direct exposure to phenoxy herbicides as compared to more indirect exposure.

Doctor, direct exposure to phenoxy herbicides is what you meant by massive. was it?--- Yes. (emphasis added)363

He then defined "direct" in the following way:

By direct exposure, by that time it was those who had somehow received this stuff on

themselves, either their clothes or the skin, (emphasis added)364

This was a most unconvincing evasion.

The description of the exposure of the cases in the series

as "massive" and "quite massive" is not only inaccurate,

it is emotive. It is not the language of a detached


VI11-13 5

As to Case 6, Dr Hardell relented:

What caused you to have that assumption for case No. 6?--- We discussed it already. I_

wouldn't sav to be massively exposed any more.

... you shouldn't have said it in 1977, should you?--- No. (emphasis added)365

After significant assistance he finally conceded that such

a description was inaccurate for Case 7 as well as for

„ , 366 Case 6.

The Commission finds that the description of the exposure

of the initial seven (7) cases as massive is inaccurate.

The background of Hardell's work is also important. In or

about April 1977 a ban on the use of 2,4,5-T was imposed

in Sweden367 for reasons which the Commission judges to

be political rather than scientific,368 At that time

phenoxy herbicides were the subject of intensive debate

and there was even some alarm about 2,4,5-T.369 The

controversy became so heated (even in the cold of the

north of Sweden) that it led to accusations that an

eminent academic from Umea, Professor Rappe, was a liar

because he defended the use of Hormoslyr, which contained


Dr H a r d e n 's clinical observations received widespread 371 publicity. He gave a number of newspaper and radio

interviews and appeared on television.372 He was

increasingly identified in local and other media as

involved in the scientific work which was seeking to link

specific types of cancer with the use of phenoxy


herbicide. He was even described as the "battering 374

ram" against phenoxy herbicides.

This publicity is relevant to possible information bias in

his studies. Because of the extent of the controversy in

Sweden information bias was always likely to be a

problem. It was accentuated by Dr H a r d e n 1 s

identification with the cause. As Professor K. Sune 3 7 5 Larsson hinted, Dr Hardell1 s publicity of his preliminary clinical observation may have influenced the

findings at least of his soft tissue sarcoma studies.

The Commission observes that when Professor Larsson gave

evidence (in Sydney in April 1984 and in Zurich in October

1984 ) it was not put to him in cross-examination that his

evidence on this matter was incorrect. The Commission

also observed that in San Francisco in September, 1984, Dr

Hardell did not deny Professor Larsson1s statements.376


It is clear that he had probably known about Professor

Larsson1s remarks as he had insisted that a condition of

his attendance before the Commission at San Francisco was

that he be permitted to read the transcript of Professor

Larsson's evidence before being called.

Dr Hardell expressly recognised that in a climate where

information bias was likely "it would be most undesirable

for a researcher who was going to undertake a study to do

anything that would exacerbate or make worse the situation 377 that existed at that time".

However that is precisely what he did. Not only did he

give newspaper interviews, appear on radio and television

and pose for photographs but he also took part in public

meetings and generally ensured that his work was brought

- -v. * 378 to the fore.

Dr Hardell's studies have been criticised on several

bases. These include: ( i )

(i) The inclusion of the data which generated the

hypotheses which the studies were designed to test;

V I 11-138

(ii) Information bias as a result of:

(a) selective recall by the cases because of the

prevailing publicity and other factors including

a preliminary phone call from Dr Hardell;

(b) a difference in the completeness of the

histories of exposure of cases when compared

with those of controls;

(c) interviewer bias arising out of the fact that

the interviewers knew the purpose of the various


(iii) inadequate, unsatisfactory and inaccurate exposure


(iv) methodological problems arising from:

(a) the inadequacy of the instructions for the

telephone interviewer;

(b) the form of the questionnaires which required

only "yes" or "no" answers and gave no

opportunity for a "don't know" reply or for 37 9 explanations;


(c) the linking by the results of 12 different

histological types of sarcoma with exposure - an

improbable situation.

(vi) the presence of confounding variables for which

adjustment could not, on the available data, be made.

In the Commission's view each of these criticisms is

validly applied to Dr Hardell1s North Sweden Study and

most of them also apply to his later studies.

Dr Hardell1s North Sweden Study was a case-control study

of 52 male patients in the age bracket 26-80 who were

treated at the Umea Oncological Centre between 1970 and


Each case was matched for sex, age and town of residence

with 4 controls. Dr Hardell then contacted all subjects

by telephone. This was followed by a letter and printed

questionnaire.380 The letter was signed by Dr Hardell

and was on the letterhead of the Umea Oncological

Centre.381 The questionnaire was headed "Umea Regional

Hospital Oncological Centre".382


North Sweden is a fairly sparsely populated area and

although Umea is a relative small place383 it is a

centre for the north. Its hospital is a specialist 384

hospital dealing with cancer and it is the only such

hospital in the north of Sweden.385 All soft tissue

sarcomas would be referred to it.386 and its cancer

section would be well-known to the people of the region.

Although the principles of epidemiology require that

hypothesis generating data and hypothesis testing data be

kept separate Dr Hardell included the hypothesis

generating cases in his North Sweden Study.387

There were 13 soft tissue sarcoma cases included in the

Study as "exposed". At least 6 of these were drawn from

the 7 initial cases,388 The evidence suggests that as

many as eight from the original 9 patients may have been


Doctor would you agree with me that as many as eight of the nine patients referred to in your clinical observations, Exhibit 1515, could have been included as exposed cases in your soft tissue sarcoma study?--- Yes. As well as the rest of the 52 cases .389

Of the initial 7 cases which were clearly included in the

study only 1, namely Case 3, was treated as unexposed.


The first general criticism is clearly justified in

relation to the North Sweden Study.

The extent of publicity which preceded Dr Hardell1s North

Sweden Study and his prominence in the debate which was

then raging about phenoxy herbicides have already been

discussed. In addition every one of the living cases "had

already been treated in the Department of Oncology at the

3 90

University of Umea" and both the letter accompanying

the questionnaire and the questionnaire itself were linked

with the Oncological Centre at Umea.

The questionnaire:

(a) specifically referred to Hormoslyr (2,4.5-T) in

questions 10, 14 and 28;

(b) included 7 questions (out of a total of 31) which were

either concerned with the spraying of herbicide or

nominated a herbicide;

(c) contained 20 questions which expressly referred to

exposure to some form of chemical or chemicals;

(d) included as its remaining questions, 8 which were

concerned only with occupation or place of work. 1


which dealt with medical details, 1 which asked about

cigarettes and 1 at the end of the questionnaire,

namely question V, which gave the subject an

opportunity to add any "additional information".

A cursory reading of the questionnaire demonstrates that

its major concern was with exposure to chemicals, the

emphasis being on herbicides. particularly Hormosylr

(2,4,5-T). In the context of the then current controversy

about 2,4,5-T the subject matter of the inquiry would have

been clear to all who received the letter and


Having regard to all these considerations it is perhaps

not surprising that Dr Hardell was finally constrained to


I admit that there could be a recall bias which probably is in our study".(emphasis added)·^1

Clearly, information bias taints the study and the second

general criticism [(ii) (a) supra] is made out.

VI11-14 3

As well, exposure to phenoxy herbicides had been raised

with each of the patients referred to in the clinical 392 observations and the diagnosis of cancer and its

association with the patient's history of exposure 393

discussed with the patient. Indeed, in taking the

history, the connection between the patient's cancer and . 3 9 4

his exposure to phenoxy herbicides was "probed". At

least one of those initial cases "spontaneously" told (Dr

Hardell) about his work as a forestry man and then he

constantly talked about spraying within that

3 9 5

occupation. No such probing or volunteering of

association can be found in relation to the controls!

Thus the general criticism in paragraph (ii)(b) supra is

made out.

The questionnaires were not numbered when they went out.

Each revealed the name of the patient. Numbers were

assigned to the cases only after the interviews had taken 3 9 6 place. The telephone interviewer for the study (Miss

Damber) had no written instructions in respect of the


interview. Dr Hardell had no direct knowledge of her

mode of questioning398 and it appears that she was not 3 9 9

experienced in such studies. The criticisms in

paragraphs (ii)(c) and (iv)(a) (supra) seem also to be



A perusal of the questionnaire400 and the accompanying 401 letter shows that the criticisms in paragraphs

(iv)(b) and (c) above are also valid.

The exposure data of the cases are both unsatisfactory and

inadequate. The statement about exposure contained in the

study is curious:

Exposure to phenoxy acids or chlorophenols was admitted by 36.5% of the patient group....

(emphasis added)402

but nowhere in the questionnaire do the words "phenoxy

herbicides" occur.403

For the purposes of the study, exposure was said by Dr

Harden to include "working in a freshly sprayed moist

area.404 However, whether a person had worked in such

an area or had "wet contact" was not dealt with by the 405 questionnaire. The only question which touched upon

this subject being question 14 which asked whether a

person had "worked in treated area":

And that doesn't tell you if you answer the question whether the area had been treated five minutes before you worked there, five days or five months before you worked there does it?--- No.


This is even more significant when only "yes" or "no"

answers were sought in the questionnaire406 and

In order to get that sort of data one would have to go to the interview?--- Yes.407

It is also clear from the use made of the initial 7 cases

referred to in the clinical observations408 that no

uniform approach to exposure was maintained (cf supra

Cases 6 and 7 in the clinical observations). In addition

Dr Harden said that questionnaires were sent to the

employers of all those cases and controls who had been 409

working in forestry, (ie 50 persons in total) and "we

could only get answers from (twenty) persons since the 410

employers did not have them keep the records". As a

result the information was regarded by Dr Hardell as "weak

data and nothing I can set up on a scientific basis and , 411

argue about".

Some of the data on which the North Sweden Study is based 412

are included in a paper which was published in 1979.

This permits the additional data relating to exposure and

the substances involved to be analysed in a limited way.

The numbers assigned to the cases in that paper cannot

however correlate with those in Dr Hardell1s clinical


observations. Dr Hardell himself was not able to do

it. All that is known is that at least 6 of the initial 7

cases and possibly as many as 8 of the initial 9 cases

observed by Dr Hardell were included as "exposed" in the

North Sweden Study.

Table III to Exhibit 1281A contains details relating to

the 13 soft tissue sarcoma cases who are said to have been

exposed to phenoxyacetic acids. The Commission examined

six of these cases, namely Cases 160, 180, 390, 490, 510

and 520.

The exposure of Case 390 is recorded in the following

manner: "2,4-D (?)".

In cross-examination it emerged that Case 390 had only 2

days of exposure at some time between 9 and 19 years prior

to diagnosis. This occurred:

While he was mist blowing?--- Yes.

With an uncertain chemical but perhaps 2,4-D?--- Yes. (emphasis added)414

_ 415

Neither MCPA nor 2,4-D contains 2,3,7,8-TCDD.



Case 160 was stated in the Study to have had 7 days

exposure 19 years before diagnosis. However„ it emerged

that the only information about the exposure of this case

was that it had occurred while he was working as a

supervisor when he had either sprayed the material himself 416

or he had got it on himself or- on his clothes.

It will be recalled that Dr Alastair Hay said of this

study that if only 2 cases had been wrongly included in

Table III of Exhibits 1281A the findings would not be


statistically significant, i.e. the result depends on

a claimed excess of only 2 soft tissue sarcomas in the

exposed category. Cases 390 and 160 could well be

classified as "unexposed".

The Commission also considered latency periods:

(i) Case 180 was said to have had 7 days of exposure to

MCPA alone at some time between 13 and 15 years


before the study.

(ii) Case 490 had a latency of between 5 and 10

V I 11-14 8

(iii) Case 510 had an exposure of 5 months "somewhere

between 3 and 9 years before the date of


diagnosis". The maximum latency period for

this case was thus 9 years, a period said to be


"improbably short".

(iv) Case 520 had a latency period which was between 8 422 and 12 years.

Thus in relation to the 13 "exposed" cases included in the

study the following emerges:

(a) at least 6 and perhaps as many as 8 formed part of

the hypothesis generating data;

(b) 3 had either doubtful or minimal exposure;

(c) 1 had a clearly inadequate latency and another 1 or

even 2 had a latency period which was barely


Dr Harden did not dispute this:

Dr Hardell, looking at Table III on page 714 of Exhibit 1281A, is it fair to say that of (the) 13 cases that are included, as many as eight of them could be cases that were

VI11-14 9

your clinical referred to observations?--- Yes in 423

Dealing with Case 490 and 510, it would be right to say that the latency period revealed is significantly less than that usually attributed to soft tissue sarcoma?--- Yes.

And the same is probably true of Case 520?--- Yes.424

In the result 10 or even 11 of the 13 cases could properly

be excluded on grounds of wrong classification or as

irrelevant to the question of causation.

Even looking at a best case from Dr Hardell' s point of

view, exclusions would not be fewer than:

- 6, because they formed part of the hypothesis

generating data; plus

- 1, because of no evidence of exposure to phenoxy

herbicides; plus

- 1, with a latency period which was clearly


Only 5 cases then remain, a number which does not produce

a statistically significant excess.


If proper principles of epidemiology are applied and the

initial 7 cases excluded. the relative risk would fall "to

something just below 3" (as was admitted by Dr


Hardell) . This fall in the relative risk is brought

about by eliminating only 1 of the errors in the study.

The reduced figure is still influenced by the effects of

subject recall bias. interviewer bias and the presence of

confounding factors for which no adjustment has been

made. Dr Hardell did not have material from which this

figure could be checked in San Francisco. He offered to

forward the data. What he in fact did was re-calculate

from data still not seen by the Commission. In a letter

addressed to the Commission in Sydney and dated 7 November

1984 he said:

. . . There was some confused discussion about the exposure to phenoxy acids. chlorophenols and organic solvents among the cases of malignant lymphoma. It was argued that the risk could not be attributed to phenoxy acids. I have looked back into my papers and separated those cases and controls only exposed to phenoxy acids. Although it is for scientific reasons not appropriate to regard those individuals with joint exposure to these substances as unexposed I have done so in my calculations: 17 cases and 11 controls were only exposed to phenoxy acids. This gives a

relative risk (crude) of 3.3 with 95% confidence limits of 1.6-7.0. still significant findings. The risk would be higher if those individuals

with joint exposure were excluded from the group of unexposed individuals - as a more appropriate method. Anyhow I have made the calculation as it was asked for during the cross-examination and I

do still find an association.426


The criticisms remain valid and the data remain unchecked

by the Commission.

The Commission also regards the heading of Table IV of

Exhibit 1281A (ie the Table headed "Exposure to

Phenoxyacetic Acids Only") as misleading. Exposure to

phenoxyacetic acids alone cannot be determined although Dr

Hardell asserted that he had analysed "exposure to phenoxy 427 acids alone" and obtained a relative risk of 5.3 in

relation to such exposure. It is apparent both from the

study itself and from his cross-examination that no such

analysis was done. The study states that:

It is ... impossible to assess the effects of the individual chemical substances separately, since practically all patients may have been exposed to chlorinated dioxins as well, including

tetrachlorodibenzodioxin (TCDD) . . . and also to other compounds. (emphasis added)428

In Tables to Exhibit 1516 in which the results of the

so-called analysis of "exposure to phenoxy acetic acids 429

alone" are set out. (emphasis added) persons who had 430

also been exposed to chlorophenols are included. In

this context it should be remembered that Dr Hardell

claims that exposure to chlorophenols is associated with a 431

relative risk for soft tissue sarcoma of 6.6 - a

confounding variable of importance.


Another consideration relevant to the accuracy or adequacy

of the exposure data is the attempt to verify exposure

with employers. This gave rise to results which were:

Uncertain and difficult to evaluate. Records of individual working manuals had not been kept, and the answers were mainly based upon reminiscence. Replies from the employers were obtained for 20/50 persons involved.43^

The exposure data for the 13 cases in Table IV of Exhibit

1281A depend very much on the outcome of the interviews as

is clear from the following:

And you rely quite heavily on that interview don't you, for exposure?--- Yes of course.

And there were no written instructions for that exposure . . . ?--- No.

.... for that interview, were there?--- No.

And had the lady that did it, Ms Damber, ever done any interviewing for epidemiological studies before this one?--- No.^33

An examination of Tables VIII and IX set out later in this

section reveals that the relative risk of the cases when

compared with controls more than doubled after interview.

There is also a marked disparity in the effect of

interview on the percentage of controls rejected when

compared with cases - nearly 4:1. Thus not only is there


validity in the criticisms that the theoretical

possibility of interviewer bias exists, but Dr Hardell's

own later analyses provide material which suggests that

it, in fact, operated.

All cases were given the suffix O. Controls were given

suffixes of 1, 2, 3 or 4. Thus, those classifying the

subjects in the study "knew just by looking whether a


person was a case or a control". This provides

further opportunity for information bias.

A preliminary classification of the subjects was carried

out by Dr Hardell and his assistant and then discussed . 435

with others, including Professor Axelson. The

classifying process was carried out with knowledge of the

numbers assigned to the subjects who were to be

classified, so "there was no question as to classification

being blind".436

For some deceased subjects the fact that their clothes

smelt was to be sufficient to classify them as


exposed. This information was obtained from "the

4 3 8

next-of-kin who washed the clothes".

The study itself admits the presence of a number of

confounding factors:


*Diesel oil - about which there was insufficient 4 3 9

information for a proper evaluation.

*Other pesticides - about which no information was

. . .. 440


It is relevant to the problem of confounding to note that

Dr Harden conceded that none of the 13 exposed cases of

soft tissue sarcoma in the study were exposed to

phenoxyacet ic acids alone. One (1) had high grade

exposure to chlorophenols and Dr Hardell was unable to say

how many had been exposed to organic solvents , but he

conceded that a number had

. 441

been and that no

calculations whatsoever could be made about the effect of

organic solvents because there were just no data on 442 them.

In the conclusions to his study Dr Hardell refers to TCDD

as a substance to which "nearly all exposed persons were 4 4 3

also exposed." (emphasis added) However two of the

cases clearly had no such exposure (Cases 180 and 390) and

they represent in excess of 15% of the total number of

exposed cases.

VI11-15 5

The Commission has noted the limited conclusion to which

Dr Hardell came as a result of his North Sweden study. It

regarded him as a strong witness, but influenced by his

anxiety to preserve the integrity of his study. The North

Sweden study is open to criticism and the Commission finds

3 30

that is rightly regarded "as flawed" and


"unacceptable as proof of the results claimed".

E.5(vi)(b) The South Sweden Study

Following his North Sweden Study Dr Hardell undertook an

investigation into "the potential effects of MCPA, 2,4-D

and the analogous phenoxy propionic acids" all of which

were widely used in agriculture in the southern areas of 444 Sweden. The study was commenced in February 1979

and, as in the case of the North Sweden study, the data

gathering took place in the spring, with the telephone

interviews in this particular study being conducted in


May/June 1978 - a time when the seasonal debate was 4 4 6 more intense. The study involved 110 patients

diagnosed between 1974 and 1978 as suffering from 13 447

different types of soft tissue sarcoma. These were

matched with 219 controls.


The gathering of exposure data involved contacting people

in advance by telephone, then sending them a nine page

guestionnaire and later supplementing the information

derived from the responses by means of selective telephone

interviews of respondents who had worked in agriculture,

forestry or horticulture during the relevant period. In

some cases the interviewer also contacted employers,

neighbours and others "to elucidate possible exposure" at


least amongst those who were deceased.

Exposure to phenoxy acids of less than one day was not

considered nor was exposure within 5 years before the year

of diagnosis, ie a 5 year latency period was adopted.

The Commission notes that the Swedish railway workers in

the study were dealt with differently from those in

Professor Axelson1s study. As Dr Hardell states in his


The embankments of Swedish State Railways have been sprayed mainly with amitrole, but pesticides including phenoxy acids have also been used. It was not possible to determine completely all the

substances to which individuals had been exposed, despite contact with the people themselves, their workmates, or their employers. Therefore, railway workers who reported exposure to

pesticides were regarded as unexposed to phenoxy acids. (emphasis added)


Eighty-five (85) of the 110 cases (i.e. 77.3%) were not

exposed to any phenoxy acids or chlorophenols. Of the

remaining 25, 7 are said to have been exposed to phenoxy

acids other than 2,4,5-T and 11 to chlorophenols alone.

This left only 7 cases who were exposed to 2,4,5-T as well

as to other phenoxy acids and known and unknown chemical

substances, included amongst which were organic solvents.

Based upon these 7 cases and 1 control Dr Hardell

calculated the point estimate of the relative risk of


exposure to 2,4,5-T and other phenoxy acids as 17.0.

This is an extraordinary figure and so out of step as to

lack credibility.

4 51

Relative risks dealt with in the South Sweden study

were reported as follows:

1. Exposure to phenoxy acids or chlorophenols - 5.1

(Table 2);

2. Exposure to all phenoxyacetic acids - 6.8 (Table


3. Exposure to all phenoxyacetic acids exluding

2,4,5-T - 4.2 (Table 3);


4. Exposure to chlorophenols alone - 3.3 (Table 3

and page 31).

An examination of a different print of the South Sweden

Study. Exhibit 1517, is helpful. It contains material

which is not included in Exhibit 1281C or Exhibit 716. In

Exhibit 1517 Dr Hardell admits that "when exposure data

are collected via questionnaires and interviews a certain

possibility exists that the cases will have a greater

interest in the questions than the healthy controls do"

(page 17). This concession is not included in the later

versions of the study. Also for reasons which are

unexplained. the results reported in the two studies

differ. Thus in Table 6 of Exhibit 1281C the relative

risk amongst agriculture/forestry workers exposed to

phenoxy acids is calculated at 6.4 based upon 13 cases and

5 controls whereas a relative risk of 5.7 is reported in

Table 7 of Exhibit 1517 (page 18).

The conclusion of the study would seem to reflect concern

about the accuracy of the results because despite the high

relative risks reported the conclusion expressed in the

study is very guarded:


This investigation has indicated that exposure to phenoxy acids and chlorophenols might constitute a risk factor in the development of soft tissue 4 c j sarcomas.

Dr Harden is careful at least about final conclusions.

Save only for non-inclusion of hypothesis generating data,

the same criticisms are available of this study as those

of the North Sweden Study. The media publicity which the

North Sweden Study attracted is likely to have accentuated

the problem of information bias in the second study.

The Commission is of the view that this South Sweden study 3 30

has properly been "widely regarded as flawed",

treated as "unacceptable as proof of the result

3 31

claimed" and such that it "cannot be taken at face

, „ 332 value".

E .5.(vi)(c) The Malignant Lymphoma Study

In January 1979 Dr Hardell published preliminary clinical

findings relating to 17 male patients who were suffering


from malignant lymphoma. As a result he undertook

his third study, the Malignant Lymphoma Study. In this

study Dr Hardell deals with 169 patients aged from 25 to 4 5 4

85 who were diagnosed in the period from 1974 to 1978


as suffering from either Hodgkin1s Disease or

non-Hodgkin's lymphoma. The cases included those 17 . 4 5 5

patients referred to in the preliminary findings.

Cases were matched for sex, age, place of residence and

year of death with a total of 338 controls. Information

about exposure to various chemicals including organic

solvents, chlorophenols and phenoxy acetic acids was

obtained by questionnaire.

However there were severe problems about identifying the

substances to which any individual had been exposed.

Indeed Dr Harden conceded in cross-examination:

In this paper concerned with lymphoma, as you told me before the adjournment, you do not present any separation of figures which deal with the number of cases or controls exposed

to phenoxy acetic acid alone?--- No.

And it is right to say, is it not. that in the end Table 3 tells us about numbers who are exposed to phenoxy acetic acid, low-grade chlorophenols, some high-grade

chlorophenols, and would you agree as well organic solvents?--- Yes.4^6

On Dr Hardell's analysis of the data the relative risk for

exposure to phenoxy acids or chlorophenols was 6.0, to 457

phenoxy acetic acids, 4.8 and to chlorophenols. 4.6.


A fivefold increased risk of malignant lymphoma was found

for those exposed to phenoxy herbicides and

4 58

chlorophenols. The increased risk for those with

high grade exposure to chlorophenols was eightfold and

4 5 9

threefold for those with low grade exposure. An

increased risk was also found for those exposed to organic

i _ 460 solvents.

The study did not reveal any difference between the risk

of Hodgkins Disease and non-Hodgkins lymphoma despite the

fact that Hodgkin's Disease seems to occur in two peaks, 461

one affecting the young and one affecting the old.

There is also thought to be some viral association in the

aetiology of Hodgkin's Disease as far as the younger group 4 62 is concerned and the aetiologies of Hodgkin's Disease

and non-Hodgkin's lymphoma appear to be different.

4 6 3

Exhibit 1518 does not differentiate between the two.

It groups them together and. in addition, includes


lymphomas which are unclassifiable.

At Page 10 of the text in Exhibit 1518 it is stated that

"exposures to phenoxy acid was analysed separately

excluding all persons who had had high grade exposure to 465 chlorophenols". However, this is not correct, as is

clear from the following:


When one goes to the analysis about phenoxy acid cases, one must keep in mind firstly that a number of them may have been exposed to low grade chlorophenol exposure?--- Yes.

And five of them were exposed to high grade chlorophenol exposure?--- Yes.^66

On examining Table 3 to Exhibit 1518 (page 5) it is clear

that the heading to that Table (namely, "Exposure to

Phenoxy Acids in Cases and Controls After Those Exposed to

Chlorophenols were Excluded and Matching Dissolved") is

4 67

also quite inaccurate and misleading. The exclusions

ought, at very least, to be restricted to high grade

exposure468 and even then it is not fair "because there

are five cases of high grade exposure included" amongst 469 those in the Table.

Thus the Table in question in Exhibit 1518 is not looking

at those cases which were exposed only to chlorophenols.

It includes cases who were exposed to phenoxy acids, 14

who were exposed to a low level of chlorophenols and 5

instances of exposure to a high level of



The frontispiece of Exhibit 1518 shows that it was

translated for the EPA and "revised in accordance with

comments by Dr s Hardell and Axelson". This is, in the

Commission's view significant, because Dr Hardell had

sought to excuse these errors by saying:


... There must be some misinterpretation ... and I guess that must be the Swedish paper too.471


Do you wish to say again that the heading to Table 3 in this paper Exhibit 1518, must be wrong because of the translation?--- It could be so, but it is scientifically no

problem because this paper hadn't been published.4/2

Ultimately, however. Dr Harden admitted that he had

revised the translation of the Exhibit "including the 4 7 Ί heading on Table 3".

Dr Hardell was also asked about Table 5 to Exhibit 1518.

That table is headed, "Exposure to organic solvents in

cases and controls after matching was dissolved and those

exposed to phenoxy acids (F) or with high grade exposure

to chlorophenols (K) were excluded, assuming that there


was no joint exposure". Exposure to organic solvents

was characterized as low grade (I) and high grade (II) and

in Table 5 there are 23 cases who had both phenoxy acid

exposure and high grade chlorophenol exposure. Dr Hardell

was asked:

Can you tell me how many of those were exposed to chlorophenols with a high grade

exposure?--- No.475

V I 11-164

He was also asked:

Doctor can you tell me from the paper how many people were exposed to phenoxy acids

only?--- In respect to chlorophenols?

No, just phenoxy acids and nothing else?--- No. 476

Although Dr Hardell "set out in this paper to examine the

relationship between phenoxy acid exposure and lymphoma"

he never published a figure in respect to phenoxy acid 477 exposure alone. However, he claimed that he had


carried out this analysis but did not submit it for



Thus the study is misleading in the form of its Tables.

There are also multiple confounding factors which have not

(and cannot) be adjusted for. The study is also subject

to the same criticisms as the North Sweden Study,

including the fact that hypothesis generating data were

used in it.

No one has been able to replicate the findings in Dr

Hardell1s studies relating to soft tissue sarcoma and

malignant lymphoma and the results of his three studies

are contrary to the findings by Professor Allan Smith in

the New Zealand studies and to the results of the

industrial studies which are dealt with later in this

Chapter and which are tabulated in Table X.


This absence of replication, the absence of specific

outcome (i.e. 12 types of soft tissue sarcoma,

non-Hodgkin's malignant lymphoma and Hodgkin's Disease),

admitted information bias, the presence of significant

confounding factors, the unreliability of the exposure

data and the other factors detailed above all indicate

that the statistical associations asserted by Dr Hardell

are suspect. The Commission cannot, on the balance of

probability, accept them as supporting an inference of

causal connection between soft tissue sarcoma, malignant

lymphoma and exposure to phenoxy herbicides.

E .5.(vi)(d) The Nasopharyngeal Cancer Study

In 1981-82 in conjunction, inter alia, with Professor

Axelson, Dr Hardell undertook a case-control study of

nasal and nasopharyngeal cancer and their relation to

phenoxy acid or cholorophenol exposure. This Study480

dealt with 71 patients (being 44 cases of nasal cancer and

27 of nasopharyngeal cancer) who were aged between 25 and

85 years, who had been reported to the Swedish Cancer

Registry in the years 1970-1979 and who were resident in

the three most northern counties of Sweden at the time of

diagnosis. These cases were matched with 541 controls. A

questionnaire which was identical with that used in


previous studies was sent to the cases, or the next-of-kin

of deceased cases. It is claimed by Dr Hardell that the

study in question "follows the methodological design" of

the previous studies undertaken by him. The controls were

those who had been used in earlier studies in North

Sweden, but no further contact was made with them for the

purposes of this study. On analysis of the data "no

significant association was found" between the cancers of 481

interest and phenoxy herbicides.

Dr Hardell claimed that this study could be used as a

cross-check on previous studies. However, despite his

claim that this and previous studies used the same

procedures there are a number of differences between the

Nasopharyngeal Cancer study and Dr Hardell' s earlier

studies. It is noted that in the Nasopharyngeal Cancer


(i) No contact was made by telephone with the cases

before they received the questionnaire;

(ii) there was no publicity linking phenoxy herbicides

and nasal cancer at the time the study was

undertaken and the debate about phenoxy herbicides

had probably abated with the banning of 2,4,5-T in

Sweden in 1977;


(iii) no hypothesis generating cases were included amongst

the cases examined in the study;

(iv) the interviewer who made telephone contact with the

cases was "blind", since the questionnaires for this

study were mixed with identical questionnaires from 482 the Colon Cancer Study.

The existence of a number of differences in the procedures

adopted for this Study when compared with earlier studies

means that this study cannot be used as a verification of

the earlier studies. The fact that a negative result was

obtained may suggest that the criticisms advanced in

respect of Dr Hardell1s first three (3) studies are


It is also of interest to note that, contrary to what had

been done by Professor Axelson in his study of Swedish

Railway Workers, employees of the Swedish Railways who

reported exposure to pesticides:

... were considered unexposed to phenoxy acids, since the type of preparation could not be

established with certainty.483


E.5.(vi)(e) The Colon Cancer Study

4 8 4

The Colon Cancer study involved 157 male colon cancer

patients who had been extracted from the Swedish Cancer

Registry and were resident in northern Sweden. Each was

sent a questionnaire and interviewed. Again the same

control group as used in the earlier study was used for

this study, but no further contact was made with the

controls. The results show no difference in the incidence

of colon cancer amongst those exposed to phenoxy acids and

it is argued by Dr Hardell that this verifies that there

was no information bias in his earlier studies.

There are, however, differences in the methodology for the

Colon Cancer study when compared with the earlier

studies. For example the cases were not telephoned by Dr

48 5

Hardell before they received a questionnaire, the

interviewer did not have knowledge either of the object of

the study or of the classification of patients, and there

had been no publicity linking colon cancer with phenoxy


E.5.(vi)(f) Effect of Interview

In cross-examination Dr Hardell was shown Table 1 of

Exhibit 1519, a document which had been produced by him in


Washington in the 1980 EPA proceedings as a true analysis

of his previous studies. In those proceedings it had been

marked as Exhibit 957.

This Table shows that before interviews were conducted:

- in the North Sweden Study, 28.8% of the 52 cases

of soft tissue sarcoma (i.e. 15 cases) had

originally been classified as exposed and 13.6%

of the 206 controls (i.e. 28 controls) had been

so classified;487

- in the South Sweden Study, 16.4% of the 110 cases

of soft tissue sarcoma (i.e. 18 cases) had been

classified as exposed and 4.6% of the 209

controls (i.e. 10 controls) had been so


- in the Malignant Lymphoma Study, 24.9% of the 169

cases of soft tissue sarcoma (i.e. 42 cases) had

been classified as exposed489 and 11% of the

335 controls (i.e. 37 controls) had been so



After interview 13 cases and 14 controls classified as

exposed remained in the North Sweden study. In the South

Sweden study 14 cases and a mere five (5) controls

remained after interview. In the Malignant Lymphoma study

41 cases remained after interview but only 24

_ . 491



These results are set out in Table VIII below.



North Sweden South Sweden Malignant STS Study STS Study Lymphoma Study

Controls - Before Interview 28 10 37

Controls - After Interview 14 5 24

Reduction 50% 50% 35%

Cases - Before Interview 15 18 42

Cases - After Interview 13 14 41

Reduction 13% 22% 2.1%

Comparative) Effect of ) Interview ) Controls ) 3.8 : 1

Vs. )

Cases )

2.27 : 1 16.66 : 1

The differential effect of the interview on cases and

controls is at least suggestive of interviewer bias.

A comparison of Table 1 of Exhibit 1519 in the present

Commission (formerly Exhibit 957 before the U.S. EPA) and


Exhibit 12 8IB shows that the percentage of exposed cases

in the Malignant Lymphoma study in Exhibit 1519 is 24.9%

However, in Table 1 of Exhibit 1281B it is shown as

When this difference was pointed out to Dr Harden he said

that this change had taken place after he had been

cross-examined in the 493 EPA proceedings and that the

change was effected because:

I must have overlooked some exposed (sic) ones which I had not noticed before.494

Dr Hardell then had to admit that by changing that figure

4 9 5

he had built up the relative risk from 2.7 to 3.

That such a mistake should have escaped him and the four

people who had gone through the data concerning the

4 9 6

Malignant Lymphoma study is curious. When confronted

with this he changed this answer and said that it did not

escape everybody:

Because we are discussing different criteria for exposure. This first part says all phenoxy acid exposure and that means exposure that has been notified somehow in the questionnaire or by

interviews, but during the evolution of the exposure it has turned out that these people are not exposed.497


When pressed on his answer he admitted that this

explanation was wrong since Table 1 of Exhibit 1519 was

dealing with information derived from questionnaires

alone, unadjusted by any value judgments made in the

assessment process and unaffected by the results of the

interview ie not taking into account the effect of . 498


When the data derived from the questionnaires alone are

used, a point estimate for the relative risk revealed by

the study is 2.6 compared with 5.3 when the effects of the 499

interviews are taken into account.

If in respect of Dr H a r d e n 1 s first three studies a

comparison is made between the point estimates obtained

using data from the questionnaires alone and the point

estimates obtained when the effects of the interviews on

such data are taken into account. the relative risks are:

2.6 for the North Sweden study - compared with 5.3 as


4.1 for the South Sweden study - compared with 6.8 as


2.8 for the Malignant Lymphoma study - compared with

4.8 as published.

V I 1 1-174

If the data are changed in the way Dr Harden changed them

between the publication of Exhibit 1519 (1980) and the

publication of Exhibit 1281B (1981) the figure becomes 3

compared with 4.1 for the Malignant Lymphoma study.500

These results are set out in Table IX below:



(Extracted from Exhibits 1516. 1517. 1518. 1519 and SFO Transcript pp 273-275




RR*- Before Interview 2.6 4.1 2.7

RR - After Interview 5.3 6.8 4.8

Increase in ) RR following) interview )

103.8% 65.8% 77.8%

* Relative Risk

It is to be noted that the figures in Table IX illustrate

the possible effect of just one of the defects for which

the Hardell studies have been criticised „ namely

interviewer bias.

When Dr Hardell was cross-examined on Table 1 of the Colon

VI 11-175

Cancer study, the heading on which is "Exposure

Frequencies to Phenoxy Acids", he gave the following



If you look at the heading in Table 1, is it right to say that the cases you are looking at under the heading Lymphoma are cases that were exposed only to phenoxy

acids?--- No. 502


..... We have already stated some of these were exposed to chlorophenol and organic solvents, but they are included.503

In this context it is relevant to return to the 13 exposed

cases of soft tissue sarcoma included in the North Sweden

study. Dr Hardell agreed that none of these cases were

exposed only to phenoxy acids. One of them in fact had

high grade chlorophenol exposure and there was an unknown 504

number who had low grade chlorophenol exposure.

Whilst Dr Hardell was unable to say how many were exposed 505

to organic solvents, he agreed that some of them were.

Although Dr Hardell claimed that organic solvents had been

excluded from the analysis of the North Sweden study,506

when he was asked to clarify this he referred to a passage

at page 715 of Exhibit 1281A. However, this passage does


no t make any reference to organic solvents.507 When

pressed about this Dr Hardell said they were included

"with the emulsifiers which contain organic

, 508

solvents". In the end, however. he conceded that

organic solvents were not dealt with in the paper:

It is not dealt with, is it, they are not dealt with?--- Not in the text, but back in the


Back in the files but the paper is silent about it?--- Yes . 509

As to organic solvents, note the following:

Put diesel oil on one side. Doctor, and tell me how I can, from the data included in the

paper (Exhibit 1518), make any calculation whatsoever about organic solvents leaving diesel oil aside?--- You can't.

No, of course you can't. There is just no data in the paper about that, are there?--- That is correct.510

Dr Hardell was then asked about information bias:

And in the case of cancer I put to you that the existence of controversy and of publicity was one factor that could operate in that regard.---- Yes .

And you may have also interviewer bias?--- Yes.

As a result of which there is a differential treatment by the interviewer between cases and controls .--- Yes .

V I 11-177

Which may result in a difference or apparently different risk ratio when the input of the interviewer is taken into account on a

question of exposure?--- Yes.

And you may, as well, you may in addition, I'm sorry, have a problem of classification that could influence a result?--- Classification of what?

Of persons as to whether or not they were exposed to a particular substance?--- Yes.511 ...

E.5(vi)(g) The Axelson Technique

Both Dr Hardell and Professor Axelson sought to rely upon

a mathematical calculation undertaken by Professor Axelson

to demonstrate that the procedures adopted by Dr Hardell

in his early studies did not give rise to information

Dr Hardell was asked about the Axelson technique at pages

285 et seq of the Transcript of evidence taken in San

Francisco in September 1984 and from his cross-examination


on this topic it emerged that the Axelson technique

depends upon a number of assumptions: 1

(1) both cases and controls will have like recall on the

subject of exposure and such recall will be




(ii) the classification of workers into categories of

agriculture and forestry (AF) on the one hand and

other occupations (O) on the other hand is accurate;

(iii) there will be an equality of incidence in the soft

tissue sarcoma amongst the unexposed members of the

AF group and the O unexposed members of the



(iv) there is no other factor which operates in one group


and does not operate in the other.

It would seem that the technique assumes the result sought

to be proved. Certainly, as Professor Allan Smith has

pointed out in this respect:

There seems to be little basis for an assumption that there would be no differential recall as between cases and controls.517

In addition Professor Smith points out that it is possible

on a theoretical (i.e. mathematical) basis:

That diminution (due to recall bias) in the proper numbers of unexposed cases employed in agriculture or forestry could be compensated for and masked by an increased incidence of the disease among workers in agriculture and forestry

V I 11-179

not exposed to phenoxy herbicides but exposed to other environmental factors prevalent in their industry, (emphasis added)518

Professor Smith also expressed the view that there was

"little value to be found in the application of the

Axelson technique to the South Sweden study" and that such

technique was not a reliable indicator of the absence of

recall bias in that study.518

The Commission accepts Professor Smith on this point.

E .5(vi)(h) Conclusion

As to the Hardell studies the Commission is of the view

that not only are they "widely recognised as flawed".

"unacceptable as proof of the results claimed" and such

" that they cannot be taken at face value" but that the 519

analysis above clearly demonstrates the flaws. They

are contrary to a large body of soundly based opinion and

the results of a number of well conducted studies. The

results of Hardell's studies have not been replicated.

Accordingly the Commission does not accept the Hardell

studies as proving, on the balance of probabilities, any

causal association between Soft Tissue Sarcoma and

Lymphoma and exposure to 2,4-D. 2,4,5-T, and TCDD.



(a) General

The manufacture of 2,4,5-T on a large scale commercial

basis did not commence in the United States, United

Kingdom or Europe until the late 1940s. Since that time

there have been a number of industrial accidents or like

occupational incidents which have resulted in employees

being exposed to phenoxy herbicides, or to the various

chemicals used for, or produced in the course of,

manufacture of 2,4,5-T and its precursors. The principal

industrial incidents are set out in Table X.



YEm T m 1955/77

m w n i f f l i /

ICOOKN Mieairro Nitro,W.\£u U.S.A.

CHEMKHS Η4ϋΝΓ 2,4,b-T 2,4,5-HP HID ttiD tetharriL



(i) Bplceam

(ilX tn p a tirra L


(6) 122

(b) 884





( m s s SB DMH

9 t o 1

35 Yes 1**

1948/4y rcarirheaiy Vfcstfalai Gsnnany

K rita c h lo n irra l HID 2,4,5/KP Ctxxpaticnal 17 1951 NIL to* to*


1952 Mld-Miine Q1MWY 2T475=TP HP

HID teCH M rcchlaracEtic a d d Methanol

CcnpatixTal 60 1960 NEL to* to*


In M g d o fm a m f

2,4,3-lltP HID HB; t o l l

B p lasim 75 1979 7 to to

1933/71 1956 1966

H tneacuim ci /,3 ,3 -Ί α / GtHrhle HID

& B dnt HB

cte Claire H$$tE

ii j u x ira ric m i iuu a i r a x

i i ) Qplcsicn 17

( ill) D p k s k r 21

- tO DtCRTOHN -- * *

1954/bb ijoenrunger Ingleheirn ^r?75=T 2,4,5-OCP

TTTD CDcxpaticral 31 1976 Nil

NaCH; 1CB Mrcjmlarracetic a d d

195b Ltiancnd- Alkalai tomric, N.J. U.SJL

z,4-u 2,4,5-T HID torrxhlcaxaaetic a d d toCH; Seetic a d d

Ccopaticral 29 1971 Nil Nd Nd

1963 Hiilipe-

Dphar Arebezdan Η Ι1Λ 1

2,4,b-Tu/ ItetradTlarfcBTaaie HID ad iim T rid ilrm L m ate

toCH; totham l

B p lcsim 106 1977 Nil Nd Nd

lybl/71 1964 LEW- Midland

Michigan, U. U.S.A.

2.4.5- T 2.4.5- HP HID Sodiun MTDchkxo-

(1) C capaticral

(ii) Oxupatim al







Nd Nd

Yes Nd

teCH __

13> ala± nl

1965/68 Sxilara αΗΗΗΌνΜΟΑ 2,4,5 Ji HB; HID

Ife>achlcrcbsT33ie tony vrknam charicals S herhiddes

Cdapaticnal 78 1978 2 Nd Nd

1954 U.S.S.K. 2,4,5-T OxCLpaticftal 128 1970 181

1968 H a lite (tine

Chankals Uiit) B risk e r Eedv^iire


2,4,5-ί ο ^ 1CB ICED CrtixdichlarcbaiaaTe NaCH

Bploeicn 79 1978 Nil to to

13/2/'/i diem ice-tete Linz, ALEIRIA 2,4,5-T HID Cccupatimal 50A00 - - * *

1973/75 (apprak) >tna=oito t o r n *


j^tit3chlanxhgHte iCP t e e ard Ccta


CticLpetitral 41 1978 Nil to* to*

Fbctnctes * ard ** appear cn the follcwing page.


Either the information which is available or the

time of follow-up involved is such that firm

negative conclusions cannot be drawn from these

incidents. Nowhere in the relevant literature,

however, is there any suggestion of unusual

cancer rates amongst the employees involved in

any of these incidents.

The one death from cancer in the classifications

1 lymphatic and hematopoietic tissues' (ICD

200-209) observed in this study is to be compared

with 3.15 deaths expected in these

classifications. Thus the Standardized Mortality

Ratio (SMR). i.e. 0.32, is negative. (Exhibit

1916 Table 4). The coding of the cause of death

in the case in question was I CD 201, Hodgkin1 s

Disease. (Exhibit 1916 Table 11). The employee

involved was one who had not been exposed to

2,4,5-T or chemicals involved in its manufacture,

(Exhibit 1916 Table 9). This instance of

Hodgkin's Disease cannot, therefore, be causally

associated with exposure to phenoxy herbicides.

In each of the instances referred to in Table X above

either acute high level exposure, chronic high level

exposure or exposure over a substantial period was


involved. In each instance the level of exposure was

sufficient to give rise to an outbreak of chloracne,

frequently severe. It is significant that cancers were

observed in only four of the studies.

The one case of malignant lymphoma which was found in an

exposed person amongst all those involved in the 14 major

industrial incidents which have occurred since 1949 is not

statistically significant in the study in which it appears.

The two cases of lung cancer which were observed in the

Spolana study occurred two and three years respectively

after exposure and both of the men involved had a history 520 of smoking. The extreme shortness of the periods

between relevant exposure and manifestation of these two

neoplasms together with the existence of a known cause for

the lung cancers observed, namely smoking, argue strongly

against any causal connection between these two lung

cancers and exposure to phenoxy herbicides.

However, it is still necessary to examine in detail the

results of those studies which reveal some incidence of

cancer amongst exposed employees or the public. These

involve, in respect of employees, Monsanto1s Nitro plant,

Dow's Midland plant and BASF's plant at Ludwigshafen.


In addition, the accident in July 1976 in a manufacturing

plant at Seveso in Italy which led to some 36,000 people

possibly being exposed to TCDD is examined. Here. of

course, the latency period following that accident is

presently (1985) less than 10 years.

F. 1 Nitro W.Va. U.S.A.

Four studies of the coincidence and causes of death of

employees at Monsanto1s Nitro W.Va. factory are considered.


F.l (a) Zack and Suskind

Dr Suskind et al undertook a study of the incidence and

causes of death amongst a cohort consisting of 121 male

employees at Monsanto1s Nitro W. Va. factory. These

workmen had all developed chloracne following a severe

explosion at the factory on 8 March, 1949 in the course of

manufacture of 2,4,5-trichlorophenol. For the purposes of

the analyses each member of the cohort was taken to have

entered the study on the date of the accident and the

vital status of each member was ascertained as at 31

December, 1978. There was thus a follow up period of 29



Nine (9) deaths from malignant neoplasms were observed in

the cohort compared with 9.04 deaths expected (SMR 1.00).

Examination of the cancer types involved in these nine


deaths reveals;

(i) No cases of soft tissue sarcoma (ICD 171);

(ii) No cases of non-Hodgkin's lymphoma (ICD 299,202);

(iii) One (1) case of Hodgkin's Disease (ICD 201);

(iv) No cases of neoplasms of the stomach, liver or

other digestive organs (ICD 150 to 159);

(v) Five (5) cases of lung cancer (ICD 162 to 163)

(2.85 expected). Four of these cases had a

history of smoking.

The conclusion from the study analyses is that there is no

indication of excess of deaths due to malignant neoplasm


in the exposed employees. Although the cohort is

relatively small the extent of follow-up was complete and

the period of follow-up longer than in any previous

study. The study thus represents "the best opportunity so

far to study the long-term effects of TODD on



V I 11-18 6

The Commission treats its negative conclusion in relation

to cancer incidence, especially in relation to soft tissue

sarcoma, as persuasive. Dr Suskind, a dignified and

scholarly researcher whose work is of the highest quality,

was good enough to share his raw data with the Commission

and its advisers in Cincinnati, Ohio. He freely gave two

days of his time so as to permit close analysis.

As soft tissue sarcoma and lymphoma are quite rare the

negative is persuasive only: the expected numbers would

have been small in any event. The study is consistent

with the other negatives.


F.l.(b) Zack and Gaffey

Another examination of the incidence and causes of death

amongst the employees in the Nitro plant of Monsanto was

undertaken by Zack and Gaffey.

The mortality experience amongst 884 white males employed

on the hourly payroll for 1 or more years between 1

January 1955 and 31 December 1977 was examined. All were

successfully traced until 31 December 1977, when their

vital status was ascertained. 721 (82%) were found to be

living and 163 (18%) were verified as having died.


Opportunity for exposure to 2,4,5-T as determined from

work history records was ascertained in respect of all but

1 of the deceased. Of the remaining 162 deceased workers

it was found that 58 had been exposed to 2,4,5-T and 104

. , _ 526

had not.

Thirty-five (35) deaths from malignant neoplasms were

found amongst the members of the group compared with 30.92

expected (SMR 1.13). Deaths from relevant neoplasms

amongst both the exposed and the unexposed workers were as


V I11-188


(Extracted from Exhibit 1916 Table 4)


Causes 0 E SMR Significance

Stomach (ICD 151) 1 1.63 0.61 No

Liver (ICD 155-156) 0 0.60 - No

All other digestive organs 3 6.42 0.47 No

Lung (ICD 162. 163 ) 14 9.91(sic) 1.41 No

* Lymphatic & Hematopoietic tissue (ICD 200-209) 1 3.15 0.32 No

** Bladder (ICD 188) 9 0.91 9.89 Yes


Ο : Observed E : Expected SMR: Standardized Mortality Ratio

Of the total of 14 lung cancers observed 6 occurred in the exposed group (3.57 expected) and 8 in the

unexposed group (6.42 expected).527

* The single soft tissue sarcoma (ICD 171.9) occurred in the unexposed group528 (expected about .1).

** The incidence of bladder cancer had previously been examined in connection with the exposure of some employees in the plant to PAB. a substance which had been used in the rubber processing section of the plant before 1955. A study had identified PAB as a

carcinogen associated with bladder cancer.529

Analysis of deaths by reference to exposure showed the

proportion of cancer deaths amongst workers exposed to

2,4,5-T to be lower than in the unexposed group, i.e. an


SMR of 82 for the exposed group and an SMR of 122 for the

unexposed group.

The authors conclude that the SMR for bladder cancer is

the only one which is significant amongst those relating

to malignant neoplasms; otherwise the Mortality Ratios for

malignant neoplasms amongst those exposed to 2,4,5-T

indicate no unusual patterns of mortality.

The authors compare the results in relation to the 884

workers in this study with those of the 121 workers the

subject of Suskind's mortality study arising out of the

1949 incident. Leaving aside PAB effects, they comment:

The results of the two studies are similar in that neither shows an excess in deaths from any site among the malignant neoplasms.530


F.l.(c) Moses et al

In 1979 Moses et al carried out a cross-sectional survey

of 226 workers who had been involved in the manufacture of


2,4,5-T at Nitro at some time between 1948 and 1969.

One hundred and seventeen (52%) of the subjects were found

to have chloracne, the bulk of them having acquired it as

a result of the explosion in March 1949 or its aftermath.

V I 1 1-190

Lifetime occupational histories were obtained and detailed

physical and dermatological examinations carried out. In

order to assess the degree of past contamination in the

plant, a sample of 2,4,5-T from an earlier year of

manufacture (1969) at the plant was obtained and


examined. It was found to contain 6ppm of TCDD (cf a

mean of 1.98 ppm for Agent Orange was used in

_ . 534


Despite the extensive physical and dermatological

examinations which were carried out as part of the study

no cancers, either systemic or of the skin, are reported

and a conclusion of the study is:

It is unlikely that permanent. severe and debilitating toxicological sequelae are inevitable after .exposure to TCDD sufficient to produce chloracne.

It should be noted that only 55% of those invited took

part in the study but of these only 6 failed to complete

full examination.

. 536

F.l.(d) Suskind and Hertzberq

In 1979, Dr Suskind carried out a further study of workers

in Monsanto1s Nitro plant to determine the long term


health effects of work-place exposure to the processes

involved in manufacturing 2,4,5-T.

The population studied consisted of two cohorts: 204

workers who had clearly been exposed to 2,4,5-T and its

contaminants and 163 who had not. The extent of exposure

of the members of the exposed cohort ranged from one month

to 20 years and commonly this exposure was intermittent 537 over a period of years.

The mean age of the exposed group was 56.7 + 0.7 years.

The members of the unexposed group were younger, the mean

age being 46.2 + 1 year. Thus it was to be expected that

the incidence of cancer might be somewhat higher amongst

the older (exposed) group.538

Members of the two groups were interviewed and physically

examined and after analysis of the data so derived, the

study reports that:

An excess of cancer of all sites was not observed in the mortality analysis of the group exposed to the run-away reaction materials nor was an increased prevalence of cancer of all sites observed in this study population. Cancer of the

liver was not observed in this study population, (emphasis added).


Three bladder cancers were observed. All of these

occurred in workers who had been exposed to PAB. In

addition all of the cancers found were in those "aged 50

years or older and the percentage in the exposed group was

not significantly different from that in the not-exposed „ 540 group".

Analysis of the incidence of skin cancer revealed that it

was not associated with exposure to 2,4,5-T and that there

was no statistically significant association (RR 1.5) 541

between the incidence of skin cancer and chloracne.

Indeed, as is noted from Table 4 to the study "there was

no association of findings of basal cell epithelioma . . . 542 with exposure".

This negative finding is important in light of the

findings in Ranch Hand 11 concerning the incidence of

basal cell epithelioma amongst Ranch Hand personnel, 543

namely, that skirt cancers found in the Ranch Hand

personnel did not correlate with the Ranch Hand herbicide

exposure index.

The results of the study by Suskind and Hertzberg are

weakly negative in relation to cancer generally and in

relation to soft tissue sarcomas and malignant lymphomas


considered as a sub-group. In addition the study provides

important negative evidence in respect of the incidence of

basal cell skin cancer.


The several Nitro studies are negative in relation to soft

tissue sarcoma and malignant lymphoma and for cancer

generally, except in relation to PAB related bladder

cancer. They certainly provide no support for the claim

by W A A that exposure to Agent Orange results in an undue

prevalence in deaths from cancer generally or from soft

tissue sarcoma or malignant lymphoma in particular.

F .2 Dow, Midland II. U.S.A.

Workers at the plant of the Dow Chemical Company at

Midland, Michigan were the subject of two mortality

studies. One was concerned with the mortality experience 544

of 61 employees (Cook et al). The other was

concerned with the mortality experience of 204 men (Ott et


This study examines the mortality experience of a cohort

of 61 Dow employees who were exposed to TCDD in 1964 as a

result of which 49 of them developed chloracne. The vital

status of these 61 men was followed to 31 December


1978 when it was found that of 4 deaths in the group

3 were due to malignant neoplasms (1.6 expected). One of

these was a fibrosarcoma, a type of soft tissue sarcoma.

It should be observed that only approximately 0.035 soft

tissue sarcomas would be expected.

The information in Table XII has been extracted from the 548 report of the study.

F. 2 (a) Cook et al546

Case Exposure Service Chloracne Cancer Type 1st


Year of Death Age

1 High 35 yr s (Query) Adenocarcinoma 1950 1s 1972 62

2 Low 33 yrs (Query) Fibrosarcoma 1964 1975* 53

3 Low 31 yes Yes Glioma 1949 1976 56

* Diagnosed in 1973549

In respect of it the authors comment as follows:

In the cohort studied here only one of the cancer deaths occurred in an individual with documented chloracne ... These findings, and the lack of a dose-response based on exposures ... suggest that

the cancer mortality experienced by this cohort was a chance occurrence.” 0


That this "was a chance occurrence" is supported by the

fact that the period between first exposure of the worker

involved and the date of his death is only 9 years i.e.

less than the generally accepted (although somewhat

speculative) minimum latency period for such a type of



The elevation in neoplasms in the study of the exposed Dow

workers by Cook et al is negative in relation to cancer

and does not provide support for the claims by W A A in

relation to the incidence of deaths from cancer generally,

or from soft tissue sarcoma and malignant lymphoma

specifically in Australian personnel exposed to Agent

Orange in Vietnam: this is in part also because of the

probable magnitude of dose.

F .2 (b) Ott et al551

This study examines the mortality experience of 204

persons exposed to 2,4,5-T during the course of its

manufacture. The durations of exposure of those included

in the study ranged from less than one year to a maximum

of approximately 10 years extending over a period from


1950 to 1971. The vital status of each of the

workers was followed through to 31 December„ 1976. Of 1Λ

deaths found to have occurred (20.3 expected) only one (.1)

was due to malignant neoplasm (3.6 expected) and no soft

5 5 3

tissue sarcomas or malignant lymphomas were found.

The single death from malignant neoplasm was from

respiratory cancer in a 63 year-old retired worker who had

a history of having smoked up to two packets of cigarettes

a day.

The number of malignant neoplasms amongst those workers

exposed to 2,4,5-T for less than one year was nil (2.3

expected) and only 1 (1.3 expected) amongst those who had 554

more than one year and up to 10 years of exposure. A

detailed examination of exposures indicated that many of

the workers included in the study had been "exposed to a

variety of other chemical compounds during their


The concentrations of 2,4,5-T (or its sodium salt) in the

atmosphere of the workplace were measured. They were 3

found to range up to 6.21 mg/m (6.21 ppm) and, in parts

of the plant where the finishing of the product was

carried out, dust levels were often sufficiently high to

be noticeably irritating, giving rise to "nasal



irritation, sneezing, bitter taste". These data

indicate that the levels of exposure were fairly high and

the periods during which the employees worked in the

departments in question were more than transient. Indeed,

in many cases, such periods were quite protracted.

The authors comment that:

No adverse mortality effects have been observed in association with the work environment.557



The results of this study are negative. They do not

support WAA's claims that the incidence of death from

cancer generally or from soft tissue sarcoma or malignant

lymphoma specifically is increased amongst persons exposed

to Agent Orange. The study is small and its results should

ba treated accordingly, but. nonetheless, it is a valuable

addition to the available data.

F .2 Summary

None of the studies of mortality and morbidity undertaken

at Dow's Midland plant amongst a large number of workers

acutely and severely exposed to 2,4,5-T and TCDD and/or


occupationally exposed to 2,4,5-T and the products of its

manufacture (including TCDD) over a long period provide

support for the claim by W A A that exposure to 2,4,5-T or

TCDD or Agent Orange gives rise to an increase in deaths

from or in the incidence of cancer generally or soft

tissue sarcoma and malignant lymphoma specifically.

F . 3 Soft Tissue Sarcoma Issue

F .3 (a) Honchar and Halperin


In a letter to the Lancet in 1981 Honchar and

Halperin recoded a neoplasm referred to in the study by 5 5 9 Zack and Suskind from ICD 173.9 to ICD 171 and having

done so counted it as a soft tissue sarcoma. They added

this to the soft tissue sarcoma referred to in the study

undertaken by Cook et al560 at the Dow Chemical

Company's plant and then added to that the soft tissue

sarcoma referred to in the study by Zack and Gaffey.561

Whilst conceding that in none of the studies was any

statistically significant excess in the incidence of soft

tissue sarcoma observed, they commented that:

When combined . . . the three cases from the four cohorts suggest a common pattern. This finding supports the concept of combining small cohorts

VI11-19 9

of workers with a common occupational exposure or analyses.562

The last sentence of this comment by Honchar and Halperin

is curious since it asserts that the result (end) is

itself, in some unstated way, a verification of the method

(means) i.e. the end justifies the means. The adding

together of one aspect of different studies can be sound

from an epidemiological point of view but is fraught with

danger: to recode and ignore the negative studies is at

best unsound.

In evidence given before the Commission in St Louis,

Missouri, Dr W Gaffey, an epidemiologist, said:

It is not possible to base a valid probability statement on a combination of one aspect of four different studies in the way in which Honchar and Halperin have done. 563

He gave three reasons for this:

(i) In each study a different criterion of exposure

was adopted (not a completely valid criticism).

(ii) The "other exposures" of the workers are

different from one study to another.


(iii) Other· populations of exposed workers should have 564

been included in the analysis but were not;

failure so to do introduces a form of selection

The logic of adding together one aspect of these four

different studies was also questioned by Dr

Armstrong, 566 „

He saw the process1 of combination as

raising "a logical issue" and an argument that because of

the differences in exposure "that combination is


Honchar and Halperin1s concept of combining one element of

different studies involves the statistical vice of

calculating a probability after the event.568 Dr Gaffey

said that this was an additional reason "why the kind of

conclusion implied by Honchar and Halperin is


inappropriate" and in this context he drew attention

to the statement by Professor Stallones that "the

statistical probability of something happening after it 570 has already happened is one".

However, it must be observed that the process of

combination would be valid if:


(i) There was a logical connection between the types

of exposures (this is in fact true);

(ii) A11 exposures of these types were included,

regardless of the results. (In practice


F. 3. (b) Cook and Johnson,571 Moses & Selikoff572 and 573 Johnson et al.

Valid or invalid, the approach of Honchar and Halperin led

Dr Cook of Dow to write to the Lancet in 1981 advising of

an additional case of soft tissue sarcoma which had been

found in the Dow workforce.

In addition, Moses and Selikoff advised of another case

among the Nitro population and finally Johnson et al

advised of two cases: they were said to be a father and

son who had been diagnosed as suffering from malignant

schwannoma. Both worked at Monsanto's plant at Sauget

Illinois which manufactured chlorinated phenols and other

organic materials.

In his report Dr Johnson asserted that the father "had

prolonged exposure before the tumour developed" but that

" the son did not have a long latency period between the

V I 1 1-202

date of his first exposure at Monsanto and the diagnosis

of his disease." However, at a previous job (4 years

earlier)" the son had some exposure to chlorophenol."

It thus appeared that there were 7 soft tissue sarcomas

included in an unknown total number of deaths amongst

workers employed in chemical plants.


F . 3. (c) Finqerhut et al

Notwithstanding the difficulties of adding together one

aspect of different studies and drawing inferences from

the aggregation, concern was expressed at the apparently

large number of soft tissue sarcomas which had occurred.

As a consequence a detailed examination of the diagnosis

and the exposure of each of the seven persons involved was

undertaken by Dr Fingerhut et al on behalf of the U.S.

National Institute of Occupational Safety and Health


The result of this detailed examination was that of the

seven (7) cases of soft tissue sarcoma reported, three (3)

had not experienced exposure to 2,4,5-T or trichlorophenol 57 5 at all (Cases 5. 6 and 7). The tissue slides of all


seven (7) persons were examined. Of the four (4) cases

who had experienced some exposure to 2,4,5-T op

trichlorophenol„ two (2) of them were found not to be soft

tissue sarcomas (Cases 2 and 3). In both of these cases


the neoplasms were carcinomas. Thus instead of there

being seven (7) cases of soft tissue sarcoma amongst

exposed workers as reported in the various studies and in

the case reports above there were in truth only two (2)

instances of soft tissue sarcoma amongst workers who had

in fact been exposed to 2,4,5-T or trichlorophenol,

namely. Cases 1 and 4.

A comparison of the position as initially reported and as

determined in the light of the detailed re-examination for

NIOSH is set out in Table XIII which is extracted from

Exhibits 1241 and 1247.

V I11-204

V I I I - 2 0 5

TPBtE x m

(Extracted frotTEgiiblts 1241 ard 1247)


BQHIDQf EPdCFOT ckgdkl tifittatiC . FDiSHUr C H O m E


EXPC6LFE* ιιι<^ιΐιΐΝ


SHIY/ FDQFHJT KbK-ki' c n st £Î“ΠΕ


X· rack and- Sutidrri 1980

Mxsamo NLtro roli«-puiiL Fibrous


SIS Y6S 1946** 1.9 yrs 'ILF


1978 29


2. rack arfl- Gaffe/ 1963

M itsnto Nitre iPdero- carcincnH

Iipcearoora iBcbm

CSitdrntB Yes 1950 2.0 yrs 2, 4, b - T


2, 4, b - τ 09581“ 1972 (71971)


3. cock ec

al 1980 u v lih ro -


LcdliliTiB (LemBt- csi£) 1964 j.a yrs Μ&1πθ3Έβ"1ϋέ - TCP

1CP 1973 9

4. Lbck et

al 1981 Lfcw tblicTHnt Fibrous Ristiocytaia

SiiD Yes 1951 19 yrs ιοί· iperattt* Flarrt irech. (TCPEeptl)

ΊΟ ly/9 28

5. M dses

al 1981 KttSSnLO Nitzo tyjoid

neurogenic &rcora SiS Isb NA minu3TaixE;


N3E 1980 ΝΔ

6. uciiticn



s r

lilifo - sarmra

SIS No m Nk ΙιχχΑχχίοι


btne 1980 Nk


m 931

Mzusanto Sfeuaet Τ1Ί yyjcad Iipcearoora

S'iS “Hd blk n r a m c n

v o t e r

FEnS 1980 Nt\

"* ‘id 2 ,4 ,5-Tf 'ILUj c r 'i d

** Ltite cr hire. LBte cf first emosure vrkrrwi. *** Wbrted fc r I I

In view of the background circumstances which led to a

re-evaluation being undertaken by Fingerhut et al it is

important to know that the NIOSH group responsible for

that re-evaluation included both Honchar and Halperin.

The accuracy of the exposure data used by Fingerhut et al,

so far as it relates to those persons employed by

Monsanto, was confirmed by the evidence given before the

Commission by Dr Gaffey in St Louis, Missouri in the 577

course of its overseas hearings.

Because of the claim by Moses et al as to the exposures of

Case 5, Dr Gaffey produced work records and gave detailed

evidence of the work history of Case 5 from the

commencement of his employment with Monsanto in 1943 until

his retirement on 1 August 1975. This evidence clearly

demonstrated that during that time Case 5 was a truck

driver, haulier and maintenance worker at Monsanto1s Nitro

plant for the period from 1945 until 8 April 1947. Because

2,4,5-T was not manufactured at that plant until October

1948 (and then only on a pilot basis, with commercial

production not beginning until 1949) he could not then

have been exposed.578 On 8 April 1947 he became a

storeroom clerk and remained in that position until 29 May

1969. During this time his place of work as a clerk


was well removed from the place of manufacture of


2,4,5-T. u There is no suggestion that Case 5 was

exposed to 2,4,5-T or TCDD in the position of trainee

instrument mechanic which he took up at the end of May

1969. Manufacture of 2,4,5-T was discontinued at Nitro in

early August, 1969.581 This evidence if accepted rebuts

the suggestion by Moses et al582 that this case "had had

potential exposure as a truck driver, hauler (sic),

maintenance worker" (emphasis added). The Commission

accepts Dr Gaffey as a witness of truth and a careful

reliable researcher, particularly as the evidence given by

Dr Gaffey in this regard was not challenged in

cross-examination by Counsel for W A A who were present at

the St Louis hearings.

No evidence was led from Dr Schneiderman, when he was

later called on behalf of WAA, to dispute:

(i) what Dr Gaffey had said either in relation to his

criticism of Honchar and Halperin on

epidemiological grounds;

(ii) Dr Gaffey1s evidence in relation to the analysis of

Case 5's work history;

V I I 1-207

(iii) the support by Dr Gaffey for the conclusions of

Fingerhut et al, that there were only two (2) cases

of soft tissue sarcoma amongst the industrial

workers examined in the studies in Table X and

in the case reports who had in fact been exposed to

2,4,5-T or trichlorophenol.

F . 3 . Conclusion

Only two (2) cases of soft tissue sarcoma were found

amongst those industrial workers examined in the studies

in Table X and referred to in the case reports who had in

fact been exposed to 2,4,5-T or tr ichlorophenol and hence

to TCDD.

F .4. BASF Ludwiqshafen, Germany

In November 1953 an explosion occurred in the course of

manufacture of 2,4,5-trichlorophenol at the BASF plant at

Ludwigshafen, Germany as a result of which 74 workers were

exposed to the products of the explosion including TCDD.

As appears from the study published in 1982 by Thiess et 583 al which is discussed in the following paragraphs, 5 8 4 584 the bulk of the men soon developed chloracne. Five


years after the event an employee who was carrying out

some welding on part of the plant which had been involved

in the explosion lifted his face mask in order to wipe

sweat from his face and four days later he also developed 585 chloracne. Six months after this he was hospitalised

for enlargement of his liver and pancreas and in the

course of being examined for this a tumour was found in

his left upper abdomen. Within a further three months

i.e. nine months after his exposure, he died. At post

mortem, in addition to the tumour. he was found to have

pancreatitis, a duodenal ulcer with perforation of the

stomach and extending into the pancreas, liver abscesses

and chloracne and, at an inquest arising out of his death,

the Coroner found an association between his death and the

explosion. In the light of present-day knowledge the

extreme shortness of the period which elapsed between his

exposure and the finding of a developed tumour would

indicate an absence of causal association between the two

- as was quite fairly conceded by Dr Alistair Hay.586

Indeed in his evidence he stated:

It is not my view that the man died as a result of this exposure.587


F . 4 . (a) Thiess and Frentzel-Beyme

In 1976 a study was undertaken of the 74 workers who had

been contaminated with TCDD as a result of the explosion

(Thiess and Goldman;588 Thiess and


Frentzel-Beyme ). All 74 were traced. In a report of

the study by Thiess and Goldman made in September


1976, two gastric carcinomas in workers aged 64 and

66 are noted.


A further study of the 74 workers was undertaken by

Thiess and Frentzel-Beyme in 1978. In it the members of

the exposed group were compared with an internal control

group in the factory matched for age and date of entry

into BASF's plant, but otherwise selected at random. The

exposed group was followed through until January 1978

(i.e. 24 years of follow up).

This study found that the incidence of malignant neoplasms

as a whole amongst the exposed workers did not exceed that

expected. However four (4) of the six (6) malignant

neoplasms found were in the gastro-intestinal tract and

these include two cases of cancer in excess of the number


expected in the age group 65 to 6 9 years. The study

findings were statistically significant in only that

V I 11-210

sub-group, namely the age group 65 - 69 years. The

authors' comment on this is important:

We are not sure if this is a chance event due to special circumstances at BASF.593

A further follow up of the cohort was made until

594 1979. Once again all members of the cohort were traced. As a result of this it was found that the total

number of deaths from neoplasms amongst the members of the

cohort was seven (7) compared with 4.22 expected for the

geographic area of Ludwigshafen. No cases of soft tissue

sarcoma (ICD 171) were found nor were any cases of

non-Hodgkin's lymphoma or cases of Hodgkin's Disease (ICD

200-202) found. However, 3 bronchial carcinomas (ICD 162)

were found compared with 1.36 expected for the

Ludwigshafen area and 3 stomach cancers were found

compared with 0.61 expected for that area. Two carcinomas

in living persons (one bronchial and one prostatic) were


also found.

The authors report that:

Because of the small size of the cohort and the small absolute number of deaths from any

particular cause, the results of this study do not permit any definite conclusions concerning the carcinogenic effect of dioxin exposure.596

V I 11-211

Some further comments should be made in respect of this


(i) Although the control group is said to have been

matched by age and date of entry into the factory

there is no indication that the group was matched

by sex. This is important since stomach cancer

mortality rates for females are lower than those

for males and lung cancer mortality rates for

females are only a fraction of the rates for

males (although with the increase in female

smoking beginning during and after World War II

that situation is rapidly changing).


(ii) Table 1 in the latest study shows that at

ages under 45 (where mortality is low) the total

number of years of exposure of the control cohort

is 13.4% greater than that of the dioxin exposed

cohort, whereas for ages over 45 (where mortality

is high) the exposure of the dioxin cohort is

6.6% greater than that of the control cohort.

Using Australian 1971 mortality rates to obtain a

rough estimate of expected deaths, this

difference in the ages of the two groups would


mean that an extra 1.6 deaths would be expected

in the older group on account of age alone. The

data show 13 deaths in the dioxin exposed cohort

and 10 in the control cohort. This difference of

3 could occur merely by chance.

However, even assuming it is not a chance

finding, if the difference in age of the two

groups is allowed for (with appropriate mortality

rates back to 1953) it is likely that the "true"

difference in incidence between the two groups

would be about one death - a result which would

not be significant statistically.

(iii) The fact that the three (3) deaths from stomach

cancer do constitute a chance finding can be

demonstrated. There were no deaths from cancer

of the stomach from 1953 to 1970. There were

three (3) deaths from this cause between 1970 and

1975 but none between 1975 and 1980. Thus a

calculation of observed deaths versus expected

death in the cohort would produce very different

results depending on when it was carried out.598


It is also of importance to recall that no excess of

stomach cancers emerges in any of the other industrial 599

studies referred to above, or in Ranch Hand I or in

Ranch Hand II600 or in the Australian Mortality

Report.601 This absence of any excess of stomach

cancers in other studies supports the suggestion by Thiess

and Frentzel-Beyme that the excess at BASF "may be due to

. „ 602



The BASF studies do not support the claims of W A A in

relation to soft tissue sarcoma and malignant lymphoma,

nor in relation to cancer generally.

F .5. Seveso-Italy

On 10 July 1976 as a result of a rupture in a pressure

reactor during the course of 2.4,5-trichlorophenol

manufacture at the ICMESA factory near Seveso, Italy, a

dense white cloud of gases, vapours, liquid droplets and

solid particulates was released. The cloud drifted down

wind and settled over nearby fields and houses affecting

an area in excess of 1,600 hectares which had a population

of more than 36,000 people.

V I 1 1-214

A lengthy report of studies conducted following the

accident entitled "Anatomy of a TCDD Spill: The Seveso

Accident" by G. Reggiani was exhibited.603

The area affected was divided into three zones.604

Zone Area Population

A 85 ha 736

B 220 ha 5,000

R 1200 ha 31,000

The TCDD contamination levels in the soil were carefully

measured in these zones in December 1976. They varied

2 2

between 5,477 ug/m in Zone A (average 235.5 ug/m ) to 2 less than 5 ug/m in Zone R. The maximum TCDD level in

Zone B (in which people have lived continuously since the 2

accident) was then 43.8 ug/m with an average for Zone B

, _ . 2 605

of 3 ug/m .

It would appear that these levels reflect a reduction in

surface concentration as a result of the process of

photodegradation deposed to by Dr D Crosby in his oral

evidence given before the Commission.606

It is therefore reasonable to postulate a general level of

TCDD concentration which was initally much higher in all

V I 11-215

zones including Zone A in which people continued to live

for the 16 days immediately after the accident.607

Chemical analysis of vegetation showed concentrations of

TCDD in the vegetation in the period between 16 July to 23

2 608

July 1976 as up to 15,480 ug/m .

Table XIV gives a comparison of possible TCDD

concentration in Vietnam, immediately after Agent Orange

spraying and the TCDD concentrations at Seveso at

different times. Thus by comparison with possible levels

of TCDD contamination in Vietnam, those in Seveso were

high and the duration of possible exposure was much more

protracted - every day and all day, rather than transient.

V I 11-216




Direct Zone Vegetation Soil

spraying in the air

(initial) July December

1976 1976

0.58 A 1300-15,840 to 20,000 to 547.9


B - to 50 to 43.8



* All values are expressed in ug/m

In the few days immediately following the accident 447

persons were diagnosed as suffering from acute dermatitis

caused by the caustic chemicals in the cloud (sodium

trichlorophenate and glycolophenate)609 and in the

period thereafter a total of 187 persons (159 children and

28 adults) developed chloracne.610 The population of

Zone A was evacuated. The populations of Zones B and R

continued to live in their respective zones.

Since 1976 the health of the people in the Seveso area,

and more particularly the health of those who suffered

some injury as a consequence of the accident, has been



To date no increase in deaths from any particular cause

has been observed amongst the inhabitants. Cancer is one

of the particular causes of death examined.

A cancer registry has been set up in the Seveso area but

no increase in the deaths from that cause has been

detected, although it should be observed that the period

which has elapsed to date is just short of 9 years. In

view of this, any conclusions in relation to cancer which

can be drawn from the experience at Seveso are limited.

One conclusion that can be drawn from the situation

however is that even if TCDD were to be ultimately shown

to be carcinogenic, the latency period for cancers argued

to arise from such exposure is greater than 9 years. This

accords with conventional modern thinking.

The people who are living in Zone B have experienced

exposure to TCDD for more than 9 years now. The

concentration of TCDD has no doubt fallen over the years

but it was, at least for some months, much more that any

Australian personnel would have experienced from direct

spraying in Vietnam.


E. & F.l. Summary

The numerous studies which have been carried out from time

to time over a period of more than 35 years in relation to

a large number of industrial workers and others who have

been acutely and, at times, severely exposed to 2.4,5-T

and/or TCDD and occupationally exposed to 2,4,5-T and its

contaminants do not support the claims by W A A that

exposure to Agent Orange is associated with an increased

incidence of death from cancers generally or from soft

tissue sarcomas or malignant lymphomas in particular.

Indeed, they provide a strong and consistent body of

evidence suggesting that there is no cancer producing

effect from exposure to Agent Orange in South Vietnam.

E. & F.2. Table of Comparative Exposures

As has been seen from the various examinations of the

health and mortality experiences of applicators and

workers in industries in which exposure to 2,4,5-T and

TCDD has occurred, there has been no increase in the

incidence of cancer generally or in the incidence of soft

tissue sarcoma or malignant lymphoma, in particular.


The increased risk asserted by Axelson is said to be a

risk arising from either amitrole (his first

conclusion)612 or from a combination of herbicides with

a "suspicion against the phenoxy acids" (his second

conclusion),613 (emphasis added). The findings of

Hardell do not, and on the data could not, relate

specifically to 2,4,-D. 2,4.5-T or TCDD.

Against this background it is worthwhile to compare the

levels of TCDD contamination to which industrial workers

have been exposed and a "worst case" situation for

Australian personnel who served in Vietnam.


LEVELS OF TCDD CONTAMINATION Extracted from Exhibits 1262 and 147.




SPOLANA Czechoslovakia Occupational 2,420,000 ug/m2 (2,240 mg/2)

PHILIPS DUPHAR Amsterdam Industrial accident

10,000 ug/m2 (100 mg/m2)

COALITE Manchester

Industrial accident

40,000 ug/m2 (40 mg/m2)


Industrial accident

14,000 ug/m2 (14 mg/m2)


These levels of contamination are to be compared with the

"worst case" levels of possible contamination calculated

for Australian personnel in jungle areas in Vietnam.

These are examined in detail in this Report in Exposure 614 Models. They are only a fraction of the levels of

contamination which occurred in the industrial incidents

referred to in Table X in which the TCDD levels are

known. Thus the level of TCDD contamination measured at

the Coalite plant is more than 10,000 times greater than

the possible level of TCDD contamination in the lower

jungle compartment. whilst at Spolana the level of TCDD

contamination is nearly 7,000,000 times greater than that

in the lower jungle compartment in Vietnam immediately

after a single spraying (0.348 ug/m2) . 6 ^ Even if

spraying over an open unforested (Savannah) area in

Vietnam is assumed (2.030 ug/m2) . the comparative

depositions are 1:20,000 (Coalite) and 1:200,000 (Spolana).



2.4.5-T AND TCDD


Laboratory test systems have been developed for the short

term testing of the geno toxic properties of chemicals.

The most widely used is the Ames test. This test employs

a number of strains of salmonella typhimurium for the

purposes of assays in which the number of mutations in the

bacterial medium, which is susceptible to the effects of

mutagenic chemicals, is measured. In other tests that are

conducted, different end-points are measured. Some of

these may have considerable relevance to mutagenicity in

humans; others may not.617

False positive results as well as false negative results

can arise from these „ . 618

tests. Because of these various

considerations it is necessary to look at more than one

set of results from a given type of test and also at

results from more than one type of test in order to form

an opinion as to the mutagenic potential of a particular


An examination of various tests for mutagenicity and

carcinogenicity of 2,4-D, 2,4,5-T and TCDD and the


opinions of various witnesses called before the Commission

together with the conclusions of the Commission in this

regard are set out.


(i) 2,4-D

2,4-D has been extensively evaluated for

potential using a wide range of test systems,

summarised in Table XVI.


These are


σ ο ;


(Extracted from exhibits 1268, 130, 180 and 1351) SUMMARY OF MUTAGENICITY OF 2,4-D



. Microbial 1. S .typhimurium not stated neg Anderson et al (1972)

2. S .cerevisiae 1000 ppm weak Siebert & Lamperle

pos (1974)

3. S .cerevisiae 100 ppm pos Zetterberg (1974)

4. S .typhimurium not stated neg Styles (1973)

5. B.subtilis 100 ppm neg Shirasu et al (1976)

6. E.coli 100 ppm neg Shirasu (1976)

7. E.coli pos Leifer et al (1981)

8. S .typhimurium 100 ppm neg Shirasu (1976)

9. S.cerevisiae 600 ppm posa Zetterberg et al (1977)

10. S typhimurium 500 ppm neg Zetterberg (1978)

11. E.coli neg Nagy et al (1975)

Insects 1. Drosophila 1629 ppm neg Vogel & Chandler (1974)

2. Drosophila 1000 ppm weak Magnusson et al

pos (1977)

Cell Cultures 1. Embryonic kidney 10 ppm pos Bongso & Basrur (1973)

2. Peripheral 1000 ppm neg Bongso & Basrur (1973)


3. Human fibroblasts 2.2 ppm pos Ahmed et al (1977a)

4. Hamster lung cells 0.2 ppm pos Ahmed et al (1977b)

5. Human lymphocytes 0.02 ppm pos Pilinskeyh (1974 )

Micronucleus Test 1. Mice 100 mg/kg neg Jenssen & Renberg


Dominant Lethal 1. Mice 125 mg/kg neg Epstein et al (1972)

2. Mice 40 mg/kg negb Lamb et al (1981b)

Host Medicated 1. Mice-S.cerevisiae 240 mg/kg neg Zetterberg et al (1977)

2. Mice-S.typhimurium 240 mg/kg neg Zetterberg et al (1977)

3. Mice-S.typhimurium not stated neg Buselmaier et al (1973)

Thymidine Uptake 1. (mouse) 200 mg/kg pos Seiler (1979)

Animal Cytogenetics 1. Rat bone marrow not stated neg Styles (1973)

2. Mouse bone marrow 40 mg/kg neg Lamb et al (1981b)


3. Mouse bone marrow 300 mg/kg pos Pilinskaya (1974)

Effective only at pH 4.5 With 40 mg 2,4,5-T and 2.4 ug TCDD daily for 8 weeks.


Dr Dost pointed out that there is "a lack of positive

response in most tests: and that 2,4-D "is without

significance as an environmental mutagenic hazard",620

notwithstanding the positives in C 3 and 5 above.

This view is shared by the WHO in the I ARC Monograph

1982621 in which the evidence for mutagenicity from the

short term tests is evaluated as 1 1 inadequate" . 622 _..


category reflects the politics of internal tensions within

agencies like IARC.

It is a category in which there were generally negative or

only up to two positive test results accompanied by two or

more negative test results.^23 Many of such data would

tend to express a negative conclusion but this category

accommodated both regulators and more rigorous scientists.

A like result is reached by Seiler in his review of the

genetic toxicology of phenoxy acids other than

6 24

2,4,5-T and by Dr Brusick when he stated his

conclusion that "very little evidence exists to suggest

that 2,4-D .... or related salts and esters are

unequivocal mutagens ", 625

V 111-225

Dr Brusick also said that, "Essentially all bacterial

assays which detect mutagenic chemicals were negative for

2,4,-D";626 and that no significant chromosome effects

were observed in tests on drosophila (fruit fly).627

In relation to the in vitro mammalian assays undertaken.

Dr Brusick noted that there are "No reliable data

available to demonstrate that ...(2,4-D) is capable of

inducing base change mutation in cultured mammalian

cells"628 and that except for a study by Pilinskaya (the 629

credibility of which was open to grave doubt) 2,4-D

had not been shown to induce chromosomal aberrations in

bovine peripheral blood cells.

In vivo studies are generally considered to be more

relevant for hazard assessment than results obtained in

vitro.630 In addition factors such as mode of

administration, metabolism and body distribution are

important in assessing risks for the toxic phenomena

associated with genetic toxicity.631

In vivo studies which should provide the more useful

results for genetic risks analysis have produced negative

results in a range of tests which included micro-nucleus,

bone marrow SCE and dominant lethal tests.632 Again the

V I 11-226

exception was a study by Pilinskaya633 about which Dr

Brusick commented:

The .... study is of great concern regarding its credibility.634


The data of Pilinskaya may not be very useful for extrapolation because of the unusually low-dose level used in the in vitro study.635

Dr Brusick expanded upon this in his cross-examination

when he said:

A positive study at a low dose would be suspect if there were a number of studies conducted at higher dose levels that failed to show an


Dr Brusick in his "Addendum to Reviews Submitted"637

classified n-butyl 2,4-D negative as a mutagen.

This conclusion is in line with that of Dr Dost referred

to above and is also supported by the summary of tests

conducted by Jones et al638 as well as by the collection 639

of studies in the IARC Monograph of 1977.

Dr Shearer. who was called on behalf of VVAA, gave

evidence contary to the above views of Professor Dost and

Dr Brusick.

V I 11-227

In her written statement. Dr Shearer said:

2.4- D has been shown to cause point mutations in animal cells, chromosome damage in human

lymphocytes as well as mouse bone marrow cells. 2.4- D has been shown to inhibit synthesis of DNA in testicular cells in mice.640

A close examination of the references on which she relies

for these statements, namely. Exhibits 1597-1602, reveals

the following:

(i) Exhibits 1597 and 1600 are studies by Ahmed et

al. These have defects which were conceded by Dr


(a) In the first study, the dose used approached the

EPA tolerance limits; there were no positive

controls used; and the effect which Ahmed

reported was accompanied by the curious result of

2,4-D being more toxic at lower doses and less

toxic at higher dose.641

(b) The second study reported an induction of

unscheduled DNA synthesis (UDS) in human

fibroblast cultures.

V I11-228

Of these Ahmed studies Dr Brusick commented that:

This data is currently suspect because . . . the results using the same techniques cannot be reproduced in another laboratory . . . and because 2,4-D as well as other pesticides reported positive by Ahmed have been tested for UDS in different assays considered to be more reliable and were found not to be active . . . Consequently Ahmed's data must be viewed with great caution.

I have no confidence in the results.642

and that the work must:

... be viewed with suspicion since no

confirmation of this work has been reported.643

(i i) Exhibit 1598 a report of the work of Pilinskaya.

This has already been dealt with, as have the

adverse comments in relation to it by Dr Brusick,

earlier in this section.

(iii) Exhibit 1599 Korte and Jalal. A "preliminary

finding" based on only four (4) test subjects.

An analysis of Table I to the report of the study

does not substantiate any adverse genetic effect

at low doses which have statistical

significance. Dr Shearer quoted it in support of

such a proposition, which it does not sustain.

V I 11-22 9

In relation to the high dose results the authors

stated that they " indicate the need for further

investigation"„644 i.e. the results were

equivocal. However, Dr Shearer was not prepared

to agree with the authors1 conclusion in this 645 regard. She based her statement on her own

interpretation of the results rather than the

authors1, a fact which she concealed from the

Commission in her written statement.

Exhibit 1601 Basrur et al. In the course of her

cross-examination Dr Shearer conceded646 that

this study was not authority for the proposition

for which she quoted it, namely "2,4-D has been

shown to cause ... stimulation of cell division 647

in non-dividing muscle cells".

Exhibit 1602 Seiler et al. Whilst Dr Shearer

quoted this study to support her statement she

failed to draw attention to the "pitfall"

referred to by the author of the study - a

pitfall into which she fell. Seiler pointed out

that his test results may have been caused in one

of two ways, one only of which was inhibition of

testicular DNA and that further testing was


necessary to resolve the ambiguity in the test

results.648 In addition Dr Shearer conceded in

cross-examination that this test does not


actually show (chromosome) "breakage".

However, even if the results were to be regarded

as unambiguous, it would still be open to

question whether such a finding has any

significance in relation to mutagenesis.650

Thus examination of the material on which Dr Shearer

relied in order to dispute the views expressed both by Dr

Dost and Dr Brusick shows her only support to be data

which are either ambiguous, scientifically unreliable or

irrelevant or an interpretation of studies which is

contrary to the interpretation of the primary


In this context it should be remembered that Dr Shearer is

a vocal and emotionally committed opponent of the use of

2,4-D.651 She engaged in intemperate attacks on Dr

Dost,652 Dr Brusick653 and Carmel1i654 implying

dishonesty to them. She took an advocate's role

inappropriate to a scientific consultant and in general

her evidence was both unsatisfactory and unhelpful.

V I 11-231

The VVAA's final submission refers to 2,4-D in a five line

reference (at p 41) and by reference to page 5120 of the

transcript seeks to suggest that Dr Shearer said that

2.4-D was carcinogenic.

A perusal of transcript page 5120 reveals that no such

assertion is made. The subject being discussed at

6 5 5

transcript page 5120 is a paper by Burton et al.

This was used by Dr Shearer to support a statement by her

that 2.4-D was rapidly absorbed through inhalation.656

The experiment referred to in Burton's paper involved

anaesthetizing male rats. inserting a tracheal cannula and

then injecting a solution of herbicide into the rats'

lungs by means of a syringe inserted through the cannula.

At the end of periods from 0.5 to 120 minutes the rats'

lungs were excised and examined. The results were

reported as showing that 2,4-D was absorbed at essentially

the same rate as that reported for other lipid-insoluble

compounds of similar molecular size.

This paper provides no support for the proposition that

2,4-D is either mutagenic or carcinogenic or that it

causes birth defects as asserted in VVAA's final


V I 11-2 32

In addition, it should be pointed out that nowhere in the

references set out in that section of WAA's final

submission does Dr Shearer say that absorption of 2,4-D

into the lung causes transmissible cellular injury,

changes in genetic material or indeed any of the things

ascribed to her in the third paragraph on page 41 of

WAA's final submission.

Dr Poliak

In the course of his evidence on behalf of WAA, Dr Poliak

dealt with 2,4-D and examined the studies which have been

performed in relation to possible chromosomal and

mutagenic effects of 2,4-D. He was not able to point to

any study in either animals or humans which demonstrated

6 57

any such damage. When he detailed the studies

performed to detect possible mutagenic effects of 2,4-D he

did not refer to any which are not included in Table XVI

set out earlier.

Indeed there are a number of studies referred to in Table

XVI to which he did not advert.

Having regard to the review he had undertaken it is

significant to note that whilst in the discussion section


of his written statement658 he stated that 2,4-D has 6 5 9

"many undesirable and deleterious effects" and that

it has "a number of toxic and hazardous effects",660 he

did not ascribe a mutagenic effect to 2,4-D.


2,4-D is not mutagenic.

G.2.(ii) 2,4,5-T

The mutagenicity and chromosomal toxicity of 2,4,5-T have

been tested in a large number of in vitro investigations.

These are detailed in Table XVII.

V I 11-234






s. cerevisiae + rev. mutation* Zetterberg et al. 1977

s . cerevisiae - gene conversion Siebert & Lamperle, 1974 B. subtilis - recombination Shirasu et al. 1976

E. coll - rev. mutation Shirasu et al. 1976

E. coll - for. mutation Fahrig, 1974

S. typhimurium - rev. mutation Carere et al. 1976

S . typhimurium - rev. mutation Styles, 1973

s. typhimurium - rev. mutation Buselmaier et al, 1973

s . typhimurium - rev. mutation Shirasu et al. 1976

S. typhimurium (8 mutants )

rev. mutation Anderson et al . 1972

S. coelicolor - for. mutation Carere et al. 1976

s . marcescens - rev. mutation Buselmaier et al. 1973

s . marcescens - ref. mutation Fahrig, 1974

(2 strains)

Only at pH 4.5 or lower






1800 ppm

1000 ppm 15 days

250 ppm

1000 ppm (+1000 ppm 2,

250 ppm

1 ppm (2.4,5-T Bee)

200 ppm

50 ppm



Neg.recessive lethal

Slight post, recessive

Neg. recessive lethal

Slight pos. recessive* 4-D)

Neg nondisjunction. chromosome loss, induced exchanges

Pos. aberrations in** oocytes & nurse cells

Neg. in unstable gene in white locus

Various oocytic aberrations


Vogel & Chandler, 1974

Majumdar & Golia, 1974

Majumdar & Golia, 1974

Magnusson et al, 1977

Magnusson et al, 1977

Darving & Sunner, 1971

Rasmuson & Svahlin, 1978

Darving, 1975

Only by combining test data. Viewed separately the results were negative.

According to Magnusson et al, this defect was not expressed as genetic disorder in offspring.

V I 11-236






Gerbil bone marrow 70,100 mg/kg daily x 5 days

Pos. gaps, breaks Neg. exchange*

Ma jumdar & Hall. 1973

Gerbil bone marrow 30 mg/kg x 5 days

No effect Majumdar & Hall. 1973

Mouse bone 100 mg/kg Neg.

micronuclei Jenssen 1976 and Renberg,

Mouse bone marrow 2.5 mg/kg Pos. gaps,


Davring 1977 and Hultgren.

Mouse bone marrow 40 mg/kg/day x 8 weeks

No effect** Lamb et al, 1981

Dominant 40 mg/kg/day No effect** Lamb et al. 1981

Cultured human leucocytes

10-4 107 M

Dose related chromatid breaks

Fujita et al 1975

* The absence of exchange type aberrations is important as it indicates that the positive effect observed could be due to toxicity.

** With 40 mg 2,4-D, 2.4 ug TCDD/kg/day.

Dr Dost gave evidence that a review of the above results

shows that there was only 1 microbial test which was

positive out of 21 different tests conducted. namely, that

of Zetter berg et a 1, 1977. In order to achieve this

result the "pH had to be lowered to at least 4.5 in order

V I11-237

to produce an effect, indicating that the anionic form of

the herbicide does not penetrate bacterial cells".662

Dr Brusick agreed with Dr Dost in relation to 2,4,5-T

stating that essentially all bacterial assays were

negative.663 in dealing with the study by Zetterberg et

al (above) he pointed out that the dose levels required to

induce mutations were 35 ug/ml which is inconsistent with

other data reported in yeast strains like that used in the

Zetterberg study in question. He expressed the view that

these results "raise several questions since one would

expect similar responses against this array of yeast tests

if the results truly reflected genetic toxicity".664 He

took the view that the " Inconsistent findings may be the

product of problems in data interpretation or ... due to

pH dependent changes in compound transport".665

He also expressed the view that there was "very little

evidence ... to suggest that 2.4.5-T or related salts and

esters are unequivocal mutagens".666

This is supported by the results obtained by Jones et

al667 in their study in which in vitro and in vivo

mutagenicity tests for 79 chemicals, including 2,4,5-T

were carried out.


In the 1977 IARC Monograph on 2,4,5-T668 the tests

referred to (with the exception of those by Ma jumdar et

al) were all negative.

The tests conducted by Ma jumdar et al on drosophila (fruit

fly) and in the bone marrow cells of the Mongolian gerbil

showed increased sex linked recessive lethals and an

increase in chromatid gaps, breaks and fragments

respectively. These results could not be replicated by


In addition, in relation to the test involving the gerbils

Dr Brusick commented that the lack of any exchange-type

aberrations in that test "substantially lowers my concern

for mutagenic risk since the observed effect could be

attributed to toxicity".669

In the 1982 IARC Monograph670 the animal and human

evidence concerning 2,4,5-T is reviewed, the conclusion

being that the evidence of mutagenicity is "inadequate" in

the sense earlier described in this Report.

V I 11-239

WAA's Case

As against this, no evidence called by VVAA showed 2,4,5-T

to be mutagenic. Witnesses called by W A A included Dr s

Silbergeld, Shearer and Poliak.

Dr Silbergeld dealt with 2,4,5-T. Her written

statement671 does not assert that it was either

mutagenic or carcinogenic.

Indeed, inferentially, she asserts the contrary by

postulating as the mechanism for possible male mediated

birth defects a dysfunction in hormonal regulation rather

that a mutagenic change.672

Dr Shearer merely mentioned 2,4,5-T in passing673 and

the litany of adverse effects to which she referred (which

did not include mutagenic damage) were not ascribed to

2,4,5-T but to TCDD.674

Dr Poliak dealt with 2, 4.5-T at length. There were no

studies considered by him in microbial. insect or

mammalian systems which are not included in Table XVII

above. He was not able to point to any study in such

organisms or systems which established that 2,4,5-T was



He did, however, introduce into this section of his 675 ,

statement reference to a study by Yoder et al

(1973).676 Dr Poliak described this study as "one of

the few investigations involving man which approaches the

rigors of a controlled laboratory study".677 However,

when he was cross-examined about this study678 he

conceded that:

(i) the persons involved in the study had been

exposed to 17 insecticides and 14 herbicides;679

(ii) 2,4,5-T was the herbicide to which they had been

least freguently exposed;680

(iii) it was "not possible from the results of this

study to draw any separate conclusions about . . .


(iv) it was not a laboratory study at all. This fact

was emphasised by his answer to a question by the


THE COMMISSIONER: Nowhere approaching it? --Nowhere.682


This isHe also referred to a study by Yef imenko. 683

written in Russian and is impossible to evaluate.

However, the English abstract does not assert that 2, 4, 5-T

is mutagenic.


2, 4, 5-T is not mutagenic.

G. 2. ( iii) TCDD

A large number of tests have been carried out to determine

if TCDD is mutagenic. A summary of these is set out in

Table XVIII.





TA 98

TA 100

TA 1530 TA 1531





PLATE INC.** 0 0 0 NTT







1532 1534 1535 1537 1538 1950



0 NT 0 0 NT NT

0 NT 0 0 NT NT

0 NT 0 0 0 0

0 NT 0 0 0 0






1975 1978 G46

NT NT NT Hussain et al

1972* *

NT NT 0 Seiler 1973*

NT NT NT McCann 1978

NT NT NT McCann 1978

0 0 0 Gilbert et al


0 0 0 Gilbert et al


NT NT NT Geiger and Neal


NT NT NT Geiger and

Neal, 1981

0 NT NT Zeiger 1983

NT = Not tested; QR = CXiestionable response; + = Positive response; 0 = Negative response * Not able to be replicated by McCann et al, Gilbert et al, Geiger et al (2 attempts). ** Done under both aerobic and anaerobic conditions

In the I ARC Monograph. 1982684 WHO examined the evidence

for mutagenic activity in short term tests and found that

it was "inadequate" to classify TCDD as a mutagen.

Dr Brusick pointed out TCDD was " judged by an expert

review panel as essentially without mutagenic activity

when tested in a variety of bacterial and mammalian model

„ „ 685


Dr Dost was of the opinion that at most TCDD could be

considered as no more than a weak mutagen in animals686

and that evidence associated with human exposure was



The ERA in its Exposure Draft688 stated that

"mutagenicity assays in micro organisms have been used to

assess the genotoxic effects of 2,3,7,8-TCDD; however the

results of most of these assays indicated little potential

for mutagenic effects".689 In addition having reviewed

available animal data and the human evidence the

conclusion was expressed that:

Overall, the data indicate little potential for the interaction of 2,3,7,8-TCDD with nucleic acids or the ability of 2,3,7,8-TCDD to produce chromosomal aberrations.690

VI 11-244

Dr Shearer referred to TCDD but did not assert that it was

mutagenic. In the long list of adverse effects which she

ascribed to TCDD691 she did not include mutagenicity.

In addition, the authority she relied upon for her long

list of adverse effects must be examined. It is a paper

by Esposito and others. 692

In it the authors point out that TCDD has been "most

extensively tested" but that "information implicating

2,3,7,8-TCDD as a mutagen is scarce and conflicting" and

they detail the results of the studies by Hussain and

_ .. 693


Dr Silbergeld did not claim that TCDD was mutagenic.

Dr Hay694 stated that "the evidence implicating TCDD is

conflicting" and that "a majority of investigations find

6 9 5

that TCDD is not a mutagen in the Ames test".

The source of the conflict referred to by Dr Hay consisted

of the study by Hussain et al696 together with a study

by himself, Ashby and others which utilized baby hamster

kidney cells (BHK cells) and was said to produce a

positive result.697

VI11-24 5

The Commission notes that although Dr Hay seemed to claim

that he had taken part in these tests and claimed credit

for them both in his book "The Chemical Scythe, Lessons of


2,4,5-T and Dioxin"


statement, it

and in

is clear

his written

from his

cross-examination 700 that he did not himself undertake

the tests. The work was undertaken by Drs Ashby, Styles


and Elliott and Dr Hay's function was to write up the . _ 702 results.

In cross-examination Dr Hay admitted that the test on

which he sought to rely was "not a test for


He was also cross-examined about the study of Hussain et

al704 and when shown a graphic representation of

Hussain's results, said:

I would be disturbed to find that line was flat along the bottom - I would expect it to

MR O'KEEFE: What would you say about that? -- I think I would be disturbed by those

findings. I would want to repeat that


You would think that that experiment was one that was of doubtful validity, would you? --Yes, I would.

Do you realize that that graph is, in fact, a graph from the Hussain 1972 test that you

referred to this morning ? -- No I do not.706


The graph in question became an exhibit. and is

indeed a graph from Hussain.


Dr Brusick was also asked about the baby hamster kidney

cells (BHK) test undertaken by Ashby et al and reported by

Dr Hay. Of it Dr Brusick said:

These are not tests to determine mutagenic activity, so positive effects or negative effects in these tests do not - are not relevant to the classification of a chemical as a mutagen or non-mutagen.708

He pointed out that the results of the test referred to by

Dr Hay had been published in 1977 and that they were

referred to as the BHK 21 test. He said:

This test has been subjected to a great deal of criticism so far as its reliability to detect chemicals which might be carcinogens ... In fact the test . . . has been found to be quite

ineffective. compared to other transformation systems;709


The use of this test and the reliance upon the data from this test. in my opinion. is not

justified in making a very strong case for carcinogenic potential, and I might add that virtually no laboratory now uses this test,

including Dr Styles in his laboratory;710


VI11 -24 7

Dr Styles himself prepared a review of the method in which he drew attention to the limitations, the methodological limitations of this test.711

Dr Brusick was then asked about the Hussain study and its 712 results. He said that he was familiar with the


particular study and that "other individuals using

the same strain and derivatives of that strain which are

supposed to be more sensitive than strain TA-15-3 2 . . .


have failed to replicate these results". As a

consequence he was of the opinion that the results

obtained by Hussain "were either unique to that laboratory

or to that given time that the study was conducted or that


they were possibly in error". (emphasis added).

Also, the inability of others to replicate these studies

"would tend to make one question these results",716

(emphasis added).

He concluded by saying that the level of concern that

phenoxy herbicides might be human mutagens or carcinogens

would be minimal at best and that such a "conclusion

appears to be entirely consistent with animal and human


information currently available".


TCDD is probably not mutagenic.


G. 3 . Carcinogenicity

As can be seen from the foregoing examination, 2.4-D and

2,4,5-T are clearly not mutagenic; TCDD is probably not.

These conclusions are important in the evaluation of the

carcinogenicity potential of Agent Orange as it contains

each of these substances - TCDD being present only in

minute quantities - a mean of 1.98 ppm.718

As already stated, the theory which is currently most

widely held concerning the cause of cancer is that its

initiation arises out of and involves some mutagenic

cellular change. In the light of the negative conclusions

in relation to the mutagenicity of 2,4-D, 2,4,5-T and TCDD

it is, even on a theoretical basis, improbable that Agent

Orange is carcinogenic.

As Dr Brusick stated:

Chemicals without clear evidence for mutagenic activity, especially in vivo, would not in my opinion be good candidates as human


This improbability is consistent with the weight of the

epidemiological evidence arising out of exposures to some

or all of the constituent elements of Agent Orange.


For completeness, however, it is useful to look at the

studies and expert opinions concerning the carcinogenic

potential of each of these substances.

G.3.(i) 2,4-D

The extensive short term studies undertaken in relation to

the genetic toxicity of 2,4-D strongly suggest that it is

not carcinogenic. The two most useful studies were 720

undertaken by Innes et al (1969) and Hansen et al



These studies involved male and female animals, dosages at

varying levels to 1250 mg/kg and periods of dosage up to

two years. The results of both were negative.

The conclusions of Hansen et al were that:

No significant effect on growth rate, survival rate, organ weights or hematologic values was noted;

and that:

No target organ tumors were observed; the

individual tumor types were randomly and widely distributed and of the type normally found in aging Os borne-Mend el rats. 7“


Pathological examination of the tumours revealed that they

were not "target organ" types. but rather that they were

randomly distributed amongst types normally found in

ageing Osborne-Mendel rats.

In addition, the number of animals which survived the two

year study did not differ from group to group, whereas

with a test substance which is in fact carcinogenic. one 723

would expect fewer survivors in the high dose groups

All these considerations support the conclusion of the

authors -

That a carcinogenic effective 2,4-D has not been shown.(emphasis added) '^4


In another study (Arknipov and Koslova, 1974) two (2)

instances of tumour were reported in a group of 165

treated animals (male and female rats) after 23 months of

2,4-D administration and one (1) in an untreated animal

after 27 months. The precise amount of 2,4-D administered

was not specified but was said to be 10% of the LD50

(unspecified) .

In a second test involving 300 female mice no tumours

developed during the animals' lifetime. Dr Dost has

pointed out that there are difficulties in the study with


methodological details,

tumours and the purity-consequence there is "

, 727 work".

the protocol for identification of

of the chemicals used.726 As a

doubt as to the meaning of the

The Scientific Advisory Panel of the EPA evaluated 2,4-D

and "found no basis for considering 2,4-D a carcinogenic

hazard, either to applicators or as an environmental

intoxicant.1,728 Dr Brusick examined available studies

and concluded that:

The evidence for carcinogenic potential is not very convincing and, in my opinion, would

indicate an almost negligible level of concern that phenoxy herbicides act in vivo as initiating agents at levels of exposure associated with commercial application, (emphasis added)72?

Dr Shubik also reviewed the data. He said:

It would be my conclusion that 2,4-D should be classified as non-carcinogenic in animal systems, and, therefore, unlikely to be in humans.730

It should also be noted that in respect of 2,4-D:

(i) WHO has fixed an acceptable daily intake for man

as 0.3 mg/kg of body weight;

V I 1 1-252

(ii) The US Occupational Safety and Health

Administration standards allow employee exposure 3 of 10 mg/m in the working atmosphere for each

8 hour shift of a 40 hour week;

(iii) The Federal Republic of Germany has a like

731 standard;

(iv) There is no question of 2,4-D being contaminated

with 2,3.7,8-TCDD.

Against this substantial body of evidence was the view

expressed by Dr Shearer. She claimed that the

experimental design of the studies undertaken by Hansen et

7 3 2 7 3 3

a 1 and Innes et al was inadequate and that the

Hansen study showed "some indications of


carcinogenicity". However:

(i) This is not what the investigators said; and

(ii) In arriving at her conclusion she depended upon a

7 3 5 "re-analysis of the Hansen data" by

Dr Reube r7 36 which she said "demonstrate

statistically significant increases in malignant 7 37 tumors when analysed by organ systems".


When cross-examined about what was actually done by

Reuber, Dr Shearer admitted that no independent work had

been undertaken by him but he is said to have looked at

some of the slides from the Hansen studies as well as

reports from other sources, including the Russian study of

Arkhipov et al.738


Although she relied on Reuber's paper. Dr Shearer was

unaware that it had been evaluated by Dr Squires in a

review undertaken by Dr Squires for the ERA and was also

unaware that Dr Squires had disagreed with Reuber's


Dr Shubik who commented on the Reuber paper rejected his

conclusion. He said:

It should be noted that this study by Hansen et al has been re-recorded in a most unusual and unorthodox manner by Reuber. This paper is presented as if it were original data; it is, however, merely a re-evaluation of the previous study and concludes that carcinogenic effects were, indeed, seen in several different organ sites. These conclusions are not supported in any wav by the data. (emphas is a d d e d ) /4i

Dr Shearer also made an assertion that 2,4-D had been

"conclusively established to be carcinogenetic in man"

74 2

(emphasis added). Her statement in this regard is


wholly inconsistent with an affidavit sworn by her in

Southern Oregon Citizens v. Toxic Sprays Incorporated on 743 744

13 June 1982. In cross-examination she admitted

that her affidavit contained the following statement:

Although there is no scientific consensus about conclusive evidence of carcinogenicity of 2,4-D, there is certainly a possibility, especially in regard to dioxin contaminants which are just beginning to be quantified and studied.

2,4-D does not contain 2,3,7,8-TCDD - dioxin: the

Commission knows of no responsible assertion that it does.

To seek to overcome the predicament in which she found

herself, she claimed that although she had signed this

affidavit, it was not sworn by her but was merely a

document, the last page of which she signed at midnight 745 one night. In this way she sought to disclaim

responsibility for the inconsistency in her two statements.

For an "expert" witness to concede that she signed a

"last" page without reading what went before (because it

was not yet in existence! ) and unaware, so it was said,

of the implications of the oath, beggars belief.

Frankly, the lady appears to be inexpert and,

accordingly, her evidence is unacceptable.


C o n c l u s i o n

2,4-D is not carcinogenic.

G.3.(ii) 2,4.5-T

Only a limited number of studies have been undertaken to


test the carcinogenic potential of 2,4,5-T.

Studies by limes et al revealed that none of the phenoxy

herbicides, including 2,4,5-T caused tumour formation at


the maximum tolerable daily dose.

Similarly, a two year study undertaken by Dow (Kociba et

74 8

al) revealed no increase in the incidence in the

number of tumours in rats when they were given doses

between 3 mg/kg/day and 20 mg/kg/day over a period of two

years. Whilst the highest dose induced significant

general toxicity, there was no increase in tumour


74 9

A. study of Muranyi-Kovacs (1976) reported a small

increase in the incidence of tumours in one group of

animals which had a longer life span than others in the

study. This study is difficult to interpret. Dr Shubik


7 50

pointed out that it was inadequately reported. When

2.4.5- T was administered to one strain of mice (XVII/G) at

80 ppm in their diet it was not carcinogenic although the

2.4.5- T contained 0.05 ppm of dioxins, but was reported as

7 51 positive when administered to another strain (C3HF).

In respect of this study he commented:

This contention is virtually impossible to check since the tumours are classified into incidental tumours and non-incidenta1 tumours for purposes of evaluation. It is then concluded that the

incidence of non-incidental tumours is

significantly increased whereas the incidence of incidental tumours is not. There is no precise listing in this paper of which tumours were placed in which category and this statistical

manipulation of the data cannot be checked. My personal conclusion from reviewing the data is that this is an entirely negative study. In this instance, perhaps, it is safest to conclude that

this study does not add anything of consequence one way or the other.(emphasis added)752

in his oral testimony he stated that the mice used in the 7 5 3

study were of a very special strain and that if the

study "were analysed in the usual manner, one would find 754 no difference whatsoever".

He also commented that he had looked at "other data

provided by these people and it very often introduces the 7 5 5 same level of confusion". Dr Shubik's examination


led him to the conclusion that it is most unlikely that

7 5 6

further studies would find 2.4,5-T to be carcinogenic.


2.4.5- T was reviewed by the I ARC in 1977 and was not

determined to be carcinogenic. Another review was carried

out by the I ARC in 1982.758 Again having examined both

animal and human data the evidence for carcinogenicity of

2.4.5- T and its esters to humans and to animals was

classified as "inadequate" (as defined).

Dr Brusick also examined the available material. He

concluded that:

The evidence for carcinogenic potential (of 2,4,5-T) is not very convincing and, in my opinion, would indicate an almost negligible level of concern that phenoxy herbicides act in vivo as initiating agents at levels of exposure associated with commercial application. Negative human epidemiology studies in which the only exposure was to phenoxy herbicides support this conclusion. Low level contamination with TCDD should not contradict this conclusion

...(emphasis added).759

and that phenoxy herbicides-

... were essentially without activity in vivo even at levels several hundred times the worst case human exposure during herbicide

application. The small amount of genetic and reproductive data obtained in human studies is uniformly negative. Consequently, the level of concern that phenoxy herbicides might be human mutagens, carcinogens or teratogens would be


minimal at best. This conclusion appears to be generally consistent with animal and human information currently available.760

Dr B Stewart in his evidence (Exhibit 1349) pointed out


Based on the animal data alone, 2.4,5-T would have to be judged as not carcinogenic for


Dr Shearer expressed no view about 2,4,5-T and Dr

Silbergeld made no assertion that 2,4,5-T was carcinogenic

- her only reference to it being that animal data can be

used as a predictor of its potential in that regard.762

Dr Hay made reference to 2,4,5-T.763 However, in his

treatment of it he did not claim that it is carcinogenic

in animals nor assert it to be carcinogenic in humans,

although he does advert to one of Hardell's studies.

The Commission notes the conclusion of the UK Advisory

Committee on Pesticides:

3.1 What we have had to consider in this Review is whether there is any sound medical or

scientific evidence that humans or other living creatures, or our environment, would come to any harm if cleared 2,4,5-T herbicides continue to be used in this country for the recommended purposes

and in the recommended way. We have found none. . . .764


C o n c l u s i o n

2,4.5-T is not carcinogenic in humans.

G.3. (iii) TCDD

Only one thorough evaluation of the carcinogenic potential

of TCDD in animals has been undertaken. That was a study

by Kociba et _ 765

al, which extended over two years and

involved four groups of 50 male and 50 female rats at

doses between 0 and 0.1 ug/kg/day of TCDD.

This study showed that at the highest dose significant

non-carcinogenic pathology occurred. At 0.001 ug/kg/day

no treatment-related effects were identified. At 0.01

ug/kg/day of TCDD some increase in hepatocellular nodules

and focal hypoplasia in the lungs were found whilst at the

highest dose excess hepatocellular carcinomas were found

in the female rats but not in the males.

Dr Dost comments that the interpretation of the study is

difficult because of the extensive non-tumour related

injury (increased mortality, morphological changes in

various parts of the body etc. )766 at the highest dose.

He points out that the general rule in carcinogenetic

V I I 1-260

studies is that "the high dose must be tolerated with

essentially no non-tumour related pathology other than

weight loss that may not exceed 10%".767 The authors of

the study reported:

(i) no increase in neoplasms occurred

receiving sufficient TCDD during the

study period to induce slight

manifestations of toxicity;768 and

(ii) doses sufficient to induce severe toxicity

increased the incidence of some types of

neoplasms in rats, whilst reducing the incidence

, „ 769

of other types.

770 Contrasted with the Kociba study is that of Kouri who found that large single doses (100 ug/kg) produced no

tumours although the dose administered was sufficient to

kill almost half the treated animals.

Dr Dost's conclusion from the data is:

The various studies of TCDD carcinogenic

potential seemed to lead to a conclusion that it is not an effective carcinogen.7 7 0 771·

in rats

two year

or no

V I I I - 2 6 1

He was of the view that:

It may be safely assumed that the carcinogenic potential of the chemical is not primary and that it is subject to the same dose-response kinetics as are the various enzyme inductive effects,

including the probable presence of a threshold effect.772

Dr Shubik considered the Kociba study and expressed the

view that the medium level of administration gave

77 3

"essentially a border-line negative result" and that

there was a clear negative result at the lowest level.

Dr Shubik also considered the US National Toxicology


Program (NTP) assay, the comments by the Peer Review

Panel, its criticism of the interpretations of the NTP 775

study and the conclusion by the Panel that the liver

and thyroid tumours found in the study could well be

attributed to hepato-toxicity induced by the high level of

dose and he came to the conclusion that:

An objective analysis of the NTP study must dismiss it as flawed in design and

interpretation"! (emphasis added)776

apparently accepting the validity of the criticisms by the

Peer Review Panel. The NTP undertook a further study but

once again the result was criticized by the Peer Review „ 777


VIII - 2 6 2

Dr Shubik was of the opinion that the results of that

further study should be interpreted with caution. He was

also of the view that, "an assessment of human risk cannot

be made" based upon this study. He further said:

I do not believe that the data in this study

provides an adequate foundation for concluding that TCDD is a carcinogen in this model


Consequently, the most he was prepared to say that TCDD 779

was in the very suspect compartment.

The I ARC reviewed the toxicity of TCDD in 1978 and again

in 1982 . In 1982 it concluded that the evidence for

carcinogenicity to animals was "sufficient „780 . „ " but that

evidence for carcinogenicity to humans was


Based upon these and other data, Dr Stewart expressed the

opinion that TCDD demonstrated activity as a promoting

agent in animals but he said:

Whilst there is no possibility of stating (on the basis of experimental evidence alone) that cancer would be caused by service in Vietnam if an

increased incidence were revealed, exposure to TCDD would be the most fruitful area for

speculation regarding cause.(emphasis added)78^

V I 1 1-263

Beside these careful analyses is to be placed the evidence

of Dr Silbergeld. In her written statement she claimed


Dioxin is a complete carcinogen ... In addition, I accept as a matter of reasonable scientific certainty the work of Pitot and others that dioxin is also one of the most powerful promoters which has ever been studied in mammalian systems.

(emphasis added)783

Dr Silbergeld1s admitted role as advocate rather than


detached scientist, her evasions and lack of

frankness, her broad assertions unsupported by authority

and her highly-coloured exaggerations have been noted by

the Commission.

However, a particular example of evasion and of lack of

frankness relates to her reference to Pitot. In the

course of expanding on the above reference, she said:

What Dr Pitot has shown - and others now as well - is that in addition to dioxin's ability to act as a complete carcinogen - that is to cause

cancer all by itself in laboratory animals - it has a very great power in two stage assay for cancer activity ...(emphasis added).785

It is clear from an examination of the paper by Pitot et

al786 that the author makes no such assertion. When

this was put to Dr Silbergeld she stated that it was

Nebert who said TCDD was a complete carcinogen.787

V I 11-2 64

When this assertion was pursued with her and when she was

asked to identify the paper in which Dr Nebert referred to

TCDD as "complete carcinogen", she nominated Exhibit

1685.788 That Exhibit makes no such statement. When

this was pointed out to her she said:

Well, I would rely on the work by Kociba.789

When the Commission then sought to ascertain the

reference, instead of Kociba1s work, she referred to a

paper by Renate Kimbrough. However. she finally had to

admit, in answer to a question from the Commission, that

in none of the papers on which she had sought to rely, did

any of the authors describe TCDD as "a complete

carcinogen". When thus caught out, she said:

That sentence appeared in the ERA 2 Volume

Review. 790

7 91

When that document was shown to her, she had to admit 792

that the phrase did not occur in it.

Dr Silbergeld1 s assertion in Exhibit 1676 that TCDD is a

complete carcinogen is to be compared with her description

of it in the paper which she co-authored with Dr Mattison

and Ms Nightingale, in which she describes TCDD as only:

VIII - 2 6 5

A tumour promoter in animals previously exposed to an initiator.793

Such a description coincides with her characterisation of

TCDD in the report which she made for the benefit of the 794 US District Court at the request of Special Master 795 Feinberg and as directed by Chief Judge Weinstein.

This document was submitted by her on or about 29 May,

1984 and was entered in the records of the Federal

79 6

District Court.

In the course of argument on Motions for Summary Dismissal

of that action (heard on 18 September 1984), Chief Judge

Weinstein (at page 41) referred to the document as "the

study bought and paid for by the Court".

In that document she did not, at any stage, describe

dioxin or TCDD as a complete carcinogen. She described it 797

as "a promoter or late stage carcinogen" and in an

addendum bearing date 13 June, 1984, she again described

TCDD as "a promoter or late stage carcinogen".798

In that addendum she relied upon the work of Pitot et al

as authority for TCDD being a promoter not a complete

carcinogen, as she at first asserted before this


V I 1 1-266

The Commission, on the evidence before it, is of the view

that the true opinion of Dr Silbergeld is that TCDD is not

a complete carcinogen, that it acts merely as a promoter 799

and then only at "high doses", and at a late stage.

But no conclusion should or will be based on that opinion


One further study. Van Miller et al (1977),800 should be

dealt with in relation with TCDD. That study looked at

the effect of doses of TCDD between 0.001 ppb (9.0003

ug/kg/week) to 1,000 ppb (500 ug/kg/week) fed to

Sprague-Dawley rats in their diet for a period of 78

weeks. At week 65 of treatment the animals then surviving

were subjected to laparotomy and biopsies were taken from

any gross tumours observed. At the end of the experiment

all animals were sacrificed and examined. Table XIX deals

with the study findings.



(Extracted from Exhibits 232 and 1271)

Tumours in Rats - Van Miller Study

Dose Benign Tumours Malignant Tumours

Total Number of Tumours

Number of Rats with Tumours

All Malign %

0 0 0 0 0/10 0/10

1 ppt 0 0 0 0/10 0/10

5 ppt 1 5 6 5/10 5/10 (50%)

50 ppt 2 1 3 3/10 1/10 (10%)

500 ppt 2 2 4 4/10 2/10 (20%)

1 PPb 0 4 5 4/10 4/10 (40%)

5 PPb 8 2 10 7/10 2/10 (20%)

From Table XIX it can be seen that the number of malignant

tumours decreased as the dose of TCDD increased from 5 ppt

through 50 ppt, 500 ppt and 1 ppb up to 5 ppb.

In addition, in the higher dose levels significant

non-tumour toxicity was observed. It should also be noted

that no tumours of any kind were found in the control

animals. According to the ERA this "is unusual for


Sprague-Dawley rats".

A wide variety of neoplasms was observed in the study

animals, a finding which the authors point out "is not

V I 11-268

consistent with many of the known carcinogens".802 The

conclusion reached in the study was that:

While the relative small number of animals as well as the diversity of results in this study does not allow the conclusion that TCDD is a carcinogen, the increased incidence of neoplastic

alterations in rats fed levels of TCDD as low as 5 ppt is of obvious concern.(emphasis added)803

Thus it can be said of the study that:

(i) No target organ or particular cancer type was


(ii) Curiously, no tumours were found in the control

animals; (the result is "too good")

(iii) There was a fall in the incidence in tumours,

both benign and malignant, as the dose increased;


(iv) The conclusion of the authors was far from a

positive one.

Such reasoning may have led the EPA to comment:

These factors may tend to lessen the reliance which can be placed on the positive results of this study.804


and to regard the study as "only suggestive evidence".805

A like response was evoked from Dr Shubik who said that

although the study was positive at high dosages it had

aspects which made its meaning unclear806 and he did not

regard it as support for the proposition that TCDD was an

extremely potent producer of malignancies.807

His views in this regard confirm the opinion expressed by

Dr Dost that "some aspects of the study have raised

questions which have made its meaning unclear".808

Dr Silbergeld made no reference to the Van Miller study in

her statement,809 in her list of references,810 or in

her oral testimony before the Commission. Nor did she

refer to or rely upon it in the paper she co-authored with

Dr Mattison,811 although she did refer to the works of

Pitot et al and Kociba et al in that paper. It is a

reasonable inference, therefore, that Dr Silbergeld did

not regard the study by Van Miller et al as an assistance

in her attempt to establish the carcinogenicity of TCDD in


The absence of reliance upon the study by Dr Silbergeld

and those called on behalf of WAA, combined with the

V I 1 1-270

comments made by the EPA, Dr Dost and Dr Shubik in

relation to the study and the negative conclusion by its

authors support the submission that it should not be

regarded as establishing that TCDD is carcinogenic either

in animals or in humans


(i) The weight of evidence is very strong against

TCDD being a complete carcinogen;

(ii) The most that can be said is that at high doses

TCDD may be a promoter in some animals;

(iii) The available evidence establishes that if TCDD

is a promoter in some animals it is subject to a

dose-response relationship, and is a promoter at

most and only when administered at high doses

over an extended period;

(iv) The evidence does not establish that TCDD is a

carcinogen in humans;

(v) The evidence does not support a conclusion that

TCDD has or would cause cancer at doses likely or

even possible in Vietnam.

V I 11-271


It will be recalled that the original claims of VVAA in

relation to cancer listed 2,4-D, 2,4.5-T and TCDD as the

three chemicals that were its prime concern. However, in

subsequent submissions W A A has claimed that many of the

chemical agents used in Vietnam have caused health

problems to Australian personnel leading to cancer.

Accordingly, reference is now made in this regard to each

of the chemical agents referred to not only in W A A 1 s

submissions before the Commission but also those in

several lists of chemicals appearing in different issues

of "Debrief", the official publication of WAA.

V I 1 1-272



It is not mentioned in W A A 1s original submission.

In its submission specifically related to toxicology, (at

p 87) it appears amongst a list of five over the following

"these chemicals appear to have been used in small

quantities in South Vietnam".

The chemical compound is not mentioned at all in the final

submission of WAA. Neither Dr Shearer, Dr Silbergeld nor

Dr Hay mentioned it.

It follows that no case is made by W A A that diquat is


Nevertheless, in accordance with its policy, the

Commission instructed Counsel Assisting to carry out

appropriate research in relation to diquat.

A WHO Joint Meeting in 1970 reviewed the carcinogenicity

data on diquat. It concluded

VI 1 1-273

Groups of rats (25 males and 25 females per

group) were fed diquat ... for two years at

dosages of 0.125. .250, 500 or 1000 parts per million in the diet. After 56 days the 1000

parts per million level was discontinued because of lack of growth and mortality in males and females. ... (at the end of two years) no

adverse effects were reported in ... tumour

incidence and gross and microscopic


In another experiment (seven groups of rats, 25 males and

25 females per group) were fed diquat dichloride for two

years. The dose levels were 0, 10, 50, 100, 250, 500 or

1000 parts per million. (At the end of two years) growth,

food consumption, survival tumour formation ... gross and

microscopic examination ... revealed no effects differing

from the controls. WHO concluded that diquat was not


The same organisation, in collaboration with the Food and

Agriculture Organisation, (FAO), reviewed the toxicity in

1979 and in connection with carcinogenicity concluded

No increased incidence of tumours has been observed in two years of dietary feeding

experiments in rats.8!3

In 1981 the Commonwealth Department of Health reviewed the

carcinogenicity of diquat. The Department then stated:


Despite the theoretical suspicion that diquat and paraquat could be carcinogenic, there is, as yet, no evidence from the medical literature that surviving victims of paraquat or diquat

poisoning have developed lung or any other cancer. Results of chronic animal studies support this.814

Neither Dr Stewart nor Dr Shubik, expert consultants to

the Commission, were asked whether they thought diquat was


Dr Ian Munro in his statement to the Commission815

stated simply "diquat is neither carcinogenic nor

mutagenic". He was not cross-examined on this statement.

The Commission concludes that diquat is not carcinogenic

and that Australian personnel have not suffered nor are

they likely to suffer from cancer as a result of exposure

to it in Vietnam.


In its original submission to the Commission W A A refers

to paraquat only once and in the following sentence "this

chemical is highly toxic in an acute lethal sense. It has

a wide range of pulmonary effects."

V I 11-27 5

In its submission relating specifically to toxicology it

is mentioned only at page 87 in the same sentence as


It is not mentioned in W A A 1s final submission. Neither

Drs Shearer„ Silbergeld, Barsotti nor Hay mentioned

paraquat in their evidence.

Thus, no case has been made by W A A that paraquat is a


WHO in 1970, reviewed the toxicity of paraquat. In

connection with its carcinogenicity the meeting reported

Four groups of rats (30 males and 30 females, with sixty of each sex for controls), were fed diets containing paraquat at levels of 0, 50, 125

and 250 parts per million for two years. No

adverse effects were seen at any level tested on growth, survival, behaviour, tumour incidents ... microscopic examination of tissues and organs revealed no adverse effects . 816

The WHO meeting of 1972 reported no increase in tumour

incidence in experiments.

In 1975 WHO after a joint meeting with FAO concluded that

there was no increase in tumour incidence amongst rats

given paraquat.817


In 1981 the Commonwealth Department of Health reviewed the

carcinogenicity of paraquat and said:

Despite the theoretical suspicion that paraquat could be carcinogenic there is. as yet. no

evidence from the medical literature that surviving victims of paraquat poisoning have developed lung or other cancer. Results of chronic animal studies support this.818

Imperial Chemical Industries PLC and ICI Australia Limited

made a submission to the Royal Commission.819

As to carcinogenicity of paraquat that submission said:

Paraquat also has been subjected to extensive toxicological evaluation in experimental animals ... experimental evidence and epidemiological studies together with long experience in use demonstrate that paraquat poses no carcinogenic,

teratogenic or reproductive hazard to man.82®

The Commission has reviewed the literature referred to in

that submission and adopts the paragraph above as

correctly evaluating the literature.

The toxicity of paraquat was again reviewed by WHO in

1984. They said:

Fletcher et al (1972a) performed a

carcinogenicity study on mice at dietary level of


25, 50 and 75 milligrams per kilogram per day for eighty weeks...... but no sign of tumour growth or atypism. Nor was any increased tumour

incidence reported in the rats in a two year study with the maximum dietary level of 250 milligrams per kilogram in the diet...... 821

Thus paraquat has not been found to be teratogenic or

carcinogenic in long term studies on rats and mice. In

vitro mutagenicity studies have been inconclusive although

generally suggestive of weak potential activity, while in

vivo studies were negative. The Commission concludes that

paraquat is not carcinogenic and that no Vietnam veteran

has or will contract cancer as a result of exposure to it

in Vietnam.


W A A 1s initial submission to the Commission does not

mention picloram.

In its submission on toxicology under the heading picloram

the following appears:

Carcinogenicity of Picloram" Reuber in Journal of Toxicology and Environmental Health 7:207-22 1981 reports that picloram is highly carcinogenic in rats and mice. It is submitted that it is

extremely likely that it is also carcinogenic in man.(at p 84.)


In its final submission W A A says "Dr Shearer822 reports

on page 1 that picloram has been shown to cause cancer in

rats and mice". Neither Dr s Hay, Shearer or Silbergeld

mentioned picloram in their oral evidence.

The National Research Council, Canada, undertook a full

evaluation of the possible environmental risks associated

with the use of picloram in 1974. On the carcinogenicity

of picloram in animals, the Council's report said:

There are no reports on carcinogenicity and mutagenicity studies, although a two year study in rats (Wolf. 1965; McCollister and Lang, 1969) indicates no increase in tumour incidence in the

test groups.823

In the Review of Literature (1981) Vol 1 the following

data on the carcinogenicity of picloram was referred to

and evaluated. That is:

Mice Two groups of B6C3F1 hybrid mice, containing 50 of each sex per group, were fed picloram (90% pure) in the diet in time weighted average doses

of 5,062 and 2,431 ppm for 80 weeks (NCI,

1978)(Exhibit 1616). Mice were observed an additional 33 weeks, were killed, and all organs were examined for gross and histological

changes. The incidence of lesions in the treated mice were compared to matched and pooled control groups.

At the end of 17 weeks, one low-dose and five high-dose females had general body tremors. During the first year of feeding, the condition

V I 11-27 9

of the mice was reported to be comparable with controls. During the second year of feeding, clinical signs of toxicity became increasingly evident among treated mice. Slight

hyperactivity, the presence of rough hair coats, and abdominal distention were noted in treated animals. There was no significant excess

mortality among treated mice as compared to controls. At the termination of the study, no significant difference between treated animals and controls was apparent in the incidence of

neoplastic and nonneoplastic lesions.

Rats Two groups of Osborne-Mendel rats, 50 of each sex per group, were fed time weighted average

concentrations of 14,875 and 7,437 ppm picloram in the diet for 80 weeks (NCI, 1978) (Exhibit 1616). The rats were observed for an additional 10 weeks, after which they were killed and

examined for gross tissue and histological changes. The incidence of lesions was compared to matched and pooled control groups.

During the second 6 months of feeding, treated rats were observed to have a moderate incidence of diarrhea, hematuria, and rough hair coats. Clinical signs of toxicity became more apparent

in the test animals during the second year of feeding and included dermatitis, tachypnea, dark urine, diarrhea, and vaginal bleeding. Upon examination of tissues, feci of cellular

alteration and neoplastic nodules of the liver were seen in treated rats. The increased

incidence of neoplastic nodules in the liver was significant only for females receiving the higher dose of picloram. Statistical analysis showed a significant dose-related trend among female rats towards increasing liver nodule incidence with increasing dose. These nodules were interpreted to be benign tumours by the examining

pathologist. Hepatocellular carcinoma was observed in one low-dose male and one high-dose female; no hepatocellular carcinoma was observed among matched controls. This finding was

reported to be statistically insignificant. A high incidence of follicular and C-cell

hyperplasia and neoplasia of the thyroid gland was observed among low-dose male and high-dose female rats, but this finding was not significant.

V I 11-280

Summary and Evaluation In the only well-controlled study found on picloram, female rats fed picloram at doses high enough to induce clinical signs of toxicity developed a significant excess of neoplastic nodules in the liver (NCI, 1978) (Exhibit

1616). No other carcinogenic lesions were found in treated rats or mice. The investigating pathologist concluded that picloram is not carcinogenic in these strains of mice and rats, but could induce benign liver tumours in female Osborne-Mendel rats.

However, after calculating the upper and lower confidence limits of the relative risk interval, it was found that the upper limit of relative risk was greater than 1. The result of this

statistical test was interpreted as indicating that the bioassay design could have failed to reveal a positive carcinogenic effect. The statistical analysis of relative risk suggests that more studies be done before any definite conclusions are made regarding the possible carcinogenicity of picloram.824

Reuber in "Carcinogenicity of Picloram" (1981)825

reviewed "every known study of the carcinogenicity of

picloram" and stated that the results and conclusions

reached by him were based on his examination of the

histological sections of the National Cancer Institite

(NCI) picloram rat study and the NCI picloram mouse study.

In connection with the NCI picloram rat study (1978)

Reuber said:

Neoplasms at all sites, as well as malignant neoplasms, were increased in both low-and high-dose picloram-treated male and female rats. The malignant neoplasms were both carcinomas and



There were similar increases in endocrine neoplasms in male and female rats given

picloram. Carcinomas were present more often in treated rats. Neoplasms of the adrenal, mainly carcinomas, were increased in both male and female rats on the low and high doses of

picloram. There were increased benign and malignant neoplasms and carcinomas of the pituitary in both female and male rats given both doses. Neoplasms of the thyroid were observed in female rats ingesting both doses and males

ingesting the low dose of picloram.

Increased incidences of neoplasms, particularly malignant, of the reproductive organs were found in female rats receiving picloram.

Neoplasms of the liver were increased in male and female rats receiving picloram.

Chronic renal disease was increased in high-dose male rats. Atrophy of the testes in

pic loram-treated male rats was severe and dose-related.826

In connection with the NCI picloram mouse study (1978)

Reuber said:

Male mice ingesting the low dose of picloram developed a higher incidence of neoplasms of the liver and of atrophy of the testes than did male mice receiving the high dose. This indicates

that the high dose was toxic and that less

picloram was ingested by the mice given the high dose than by the mice given the low dose.

Mice of both sexes ingesting the high dose of picloram developed high incidences of neoplasms of the spleen. It was previously noted that mice ingesting low doses of dieldrin tended to develop lymphomas, whereas those given higher doses

developed neoplasms of the liver (Walker et al, 1972) .

Summary Picloram is carcinogenic for the spleen of B6C3F m^le and female mice. Male mice also

developed atrophy of the testes.827


In February 1982 the Poisons Schedule (Standing) Committee

of the National Health and Medical Research Council in

Australia received the Reuber paper for consideration. The

Committee also noted:

Extracts from "Pesticide and Toxic Chemical News" of April 5. 1981, which reported that Dr Reuber had received a "very stern warning" from his supervisor for releasing the results of the study

before they could be scrutinized by the National Cancer Institute. Dr Reuber's competence had also been questioned.

However, the Committee also noted that the statistical analyses had been done by the

Frederick Cancer Institute and that the paper had been reported in a reputable scientific journal and had therefore undergone peer review. The study could therefore not be discounted on the basis of the conflict between Dr Reuber and the organization with which he was employed.

The Secretary of the Committee was requested to write to Dow Chemicals, seek their comments on the paper and, if possible, their rebuttal of the results".828

Then, at the 55th Meeting of the PACSC on the 6 and 7 May


The Committee noted comments from Dow Chemical (Australia) Limited on the paper "Carcinogenicity of Picloram" by M.D. Reuber. Comments on the paper had been requested by the Committee after

it had noted Dr Reuber's paper and the

controversy at the February 1982 Meeting.

The comments indicated that neither Dow, the National Cancer Institute nor the US

Environmental Protection Agency had considered the toxicology of picloram to indicate possible carcinogenicity.


The toxicology reviewed by PACSC also did not indicate a carcinogenic risk.

However„ Dow had decided to repeat the subacute and chronic studies for consideration by the NCI and the Secretary was reguested to obtain the results of these studies when available.

No further action was considered necessary.829

In the Review of Literature (1984) Vol 3, Reuber's review

of the picloram carcinogenicity studies was again

commented upon. The authors said:

Reuber (1981: Exhibit 1617) reexamined the histological sections from the NCI studies on rats and mice described in JRB. Review of Literature (1981, Vol 1: Exhibit 894). Treated animals were compared to a pooled control group consisting of 10 matched controls and 30

untreated animals from studies on other

chemicals. In both male and female rats, Reuber found an increase in the incidence of total benign and malignant neoplasms, endocrine gland neoplasms, thyroid gland neoplasms, pituitary gland carcinomas, adrenal gland carcinomas, and benign and malignant neoplasms of the liver. Malignant neoplasms of the reproductive tract and neoplasms of the uterus were significantly

increased in the low-dose group of female rats. Male rats showed treatment-related atrophy of the testes and chronic renal disease. Treated mice showed an increased incidence of neoplasms of the

spleen in high-dose males and females and of testicular atrophy in low-dose males.

A number of the tissues examined by Reuber (1981) were tissues in which tumour diagnosis is

sometimes controversial. The author did not report the histopathological criteria he used to classify lesions. Therefore, it is not possible

to evaluate his diagnoses. Comparing three histopathological reviews of the liver slides from the pic lor am-treated rats revealed that the


results obtained by Kasza (1980), reading the slides blind, were not significantly different from those obtained by NCI (1978). However, the

results obtained by Reuber (1981) differed significantly from both of the other diagnoses. The number of tissues examined by Reuber also differed from the number examined by NCI. In some cases Reuber examined more tissues than NCI, probably tissues that were autolyzed or from animals that had been lost, as these tissues are not usually examined in NCI bioassays. In other cases, he examined fewer tissues than NCI, as he excluded tissues from some animals that died before termination of the study. In view of

these questionable features of Reuber's review, it does not provide sufficient evidence to change NCI's conclusion (supported by Kasza's

independent review) that picloram induced only benign tumours in high-dose female rats. Dow Chemical company (1983b) reported that they are performing a carcinogenicity bioassay, started in

1982, that will provide additional evidence as to whether picloram is a carcinogen.830

(In summary) chronic exposure provided only limited evidence that picloram was a liver carcinogen in one sex of one species, the

problems with the study could have affected the outcome. Therefore, it is necessary to wait for the results of the Dow Chemical Company cancer study before determining if picloram is a


Surprisingly, Dow Chemical (Australia) Limited, in its

submission to the Commission832 does not deal at all

with the carcinogenicity of picloram.

Dr Philippe Shubik, in his statement to the Commission

dealt with the carcinogenicity of picloram in the

following terms:


This compound was assayed in a standard U.S. National Cancer Institute bioassay using Osoborne-Mendel rats and B6C3F1 mice (1978: Exhibit 1616). The rats were given 744 or 372 mg/kg/day and the mice fed either 640 or 320 mg/kg/day. Only the female rats were reported to have had an increased incidence of benign liver

tumors. M. Reuber (1981: Exhibit 1617) has reported that on re-reviewing this study that there was an increased incidence of various malignant tumors. The recent review of this

literature by Clements Associates, Inc (Review of Literature (1984) Vol 3: Exhibit 894) takes issue with the techniques used by Reuber in selecting his material. It is not possible for me to take a stand in this instance without

additional information.

In view of the other characteristics of picloram - namely that it appears not to be metabolized it seems unlikely that it will prove to be

carcinogenic in the study that is now apparently underway. Judgment of the potential hazard from this compound should await the completion of the new study (by Dow), (emphasis added)0133

Dr Bernard Stewart in his statement to the Commission

indicated that there was no international agency for

research on cancer evaluation of the carcinogenicity of

picloram but added:

A bioassay of technical grade picloram for carcinogenicity was conducted by the National Cancer Institute (USA). Picloram was fed to rats and mice. No tumours were found in male or

female mice or male rats at incidences that could be significantly associated with treatment. In female rats, however, the incidence of neoplastic nodules of the liver, and benign tumours was associated with treatment with picloram.834


and he said that at worst, the results of the above NCI 835 (1978) study constituted "limited" evidence of

. . . 836


Dc Stewart also dealt with the matter in his oral evidence

Dr Ian Munro dealt with the carcinogenicity of picloram in

his oral evidence during questioning by Counsel

Assisting. The following exchange occurred:

.... The National Cancer Institute of the United States has tested picloram for carcinogenicity in rats and mice and negative results were found in mice but a very slight increase was found in

hepatic tumours in rats given a maximum tolerated dose, that is, the maximum tolerated dose is a near toxic dose of picloram. You will recall I mentioned that there were some indications that

picloram did induce some changes in liver

pathology when administered to rats and that these changes, 1 think are predisposed to the development of these hepatic benign tumours. These are not malignant tumours, these are benign tumours of the liver which occur there

spontaneously anyway ---At the highest level of toxicity and keeping them alive?--- Yes, yes, yes, but I think what is

important is that the picloram has been tested extensively in those short term tests for carcinogenicity looking for genetic end points as I pointed out previously. These results have been persistently negative. So my impression is

that the positive, so-called positive effects. in terms of benign hepatic tumours in rats is probably due to accompanying hepatic toxicity and that there is again no evidence to indicate that this is a carcinogen per se. (emphasis added)837


In the Commission's view Reuber is "out on his own" in

believing picloram to be carcinogenic. In any event the

study upon which he relies is a study of dosages orders or

magnitude higher than any possible Vietnam service related

dose. As well, dosages were applied over a period which

amounts to almost the whole of the life of the rat.

By comparison, soldiers in Vietnam would perhaps have

received very small doses irregularly over a period

equivalent to, at most, a sixtieth of their normal life


In the Commission's view, picloram is not relevantly a

carcinogen. Accordingly, the Commission concludes that no

Vietnam veteran from Australia has or is likely to suffer

from cancer as a result of exposure to picloram in Vietnam.


The possibility that picloram is carcinogenic in animals

cannot be excluded. However, the data are such as to

provide no basis of inference of a carcinogenic hazard to


VII 1^288


W A A ' s original submission does not mention the compound.

In its submission related specifically to toxicology the

compound is mentioned on page 85 as a heading over the

following sentence "dalapon is a carcinogen and primary


In its final submission dalapon receives no mention.

No evidence was called to support WAA's allegation that

dalapon is a carcinogen. No literature reference is given.

Dalapon is an exempt compound for the purpose of the

Poisons Schedule in Australia which is a reflection of its

low toxicity. The research of those Assisting indicates

that dalapon has never been referred to by any of the

investigatory bodies and there is just nothing to suggest

that it is carcinogenic. For example, WHO has published

no data on this compound.

The Commission finds that dalapon is not a carcinogen and

that it has not or will not cause cancer amongst Vietnam

veterans as a result of exposure to that chemical agent in




In its primary submission to the Commission this compound

is not mentioned by WAA.

In its submission related specifically to toxicology it

appears only in a list over the words "these chemicals

appear to have been used in small quantities in South


In its final submission W A A includes diuron in a list

over the following words:

Diuron and bromacil have mutagenic activity in multiple bacterial test systems. Monuron is oncogenic and diuron is suspect. Dr Shearer. Exhibit 1615 page 4 and report cited. See also Exhibit 1654a 1655.(at p 33.)

Dr Shearer gave oral evidence in addition to her written

statement. Her training was as a nurse. Her evidence

was based on literature search only. She has no personal

experience of any of the relevant chemicals but claimed

expertise in genetic toxicology. Any experience in

pharmacology was restricted to that gained during her

nursing training.838

V I 11-290

Her expertise by comparison with the toxicologists truly

expert in the field is negligible. Her suspicion of

diuron is based only upon the US EPA reference of it for

further testing. This reference was based upon its

chemical similarity to diurene.

The Commonwealth Department of Health considered the

toxicity of diuron.839 It concluded "no mutagenic,

carcinogenic or teratogenic effects of diuron in man or

animals are known to have been reported".

In a standard text book, "Pesticides Studied in Man." W.J.

Hayes, one of the world's most eminent toxicologists, said:

In a feeding study, some increase in mortality may have been attributable to the dietary level of eight thousand parts per million diuron. There was no dosage related mortality at five

thousand parts per million or less. At two

thousand five hundred parts per million for two years ... no histological changes were seen and there was no evidence of carcinogenicity.840


The author nominates NOEL of 1.4 mg/kg day.

Thus at dosages several orders of magnitude higher than

those likely to have been received by soliders in Vietnam

and for the whole of the lifetime of the experimental

animal no carcinogenicity was found.


Nor has anyone been concerned enough to refer this

compound to WHO's Research bodies.

The Commission concludes that diuron is not carcinogenic

and that it has not and will not cause cancer amongst

Vietnam veterans as a result of exposure to it in Vietnam.


In its initial submission to the Commission VVAA did not

mention monuron.

In its submission related specifically to toxicology

monuron appears only on page 87 as part of a list over the

words "these chemicals appear to have been used in small

quantities in South Vietnam".

In its final submission the only words relating to monuron

appear at page 33 in the following sentence:

Monuron is oncogenic and diuron is suspect. Dr Shearer, Exhibit 1615. page 4 and report cited. Dr Shearer, Exhibit 1615, page 4 and reports cited. See also Exhibit 1654a and

Exhibit 1655.


Monuron was used only by the US forces. There was no

Australian use of this compound. Furthermore, only small

quantities of this chemical was used in Vietnam between


1962 and 1964, for testing purposes only.

Exhibit 1615 is a statement by Dr Shearer. It contains

the sentence, "Monuron is oncogenic." She cites as a

reference a document843 which is a paper by J.P. Seiler

dealing with a possible chemical basis for mutagenic

activity of monuron. Dr Shearers 1 s reference to the

article seems to be to an interpolation in the first line

of the head note. This is simply - "a recognised

carcinogen" In the body of the article this is

supported by reference to IARC monographs.

Exhibit 1655 is a document issued by the US EPA entitled

"Guidance for the Re-Registration of Manufacturing Use and

Certain End Use Pesticide Products containing Diuron as

the Active Ingredient." It does not refer to monuron.

As the Commission will deal with the source documents and

the IARC1s direct comments there is no further need to

refer to Dr Shearer's unqualified and unsupported



The Commonwealth Department of Health reviewed

toxicity of monuron and concluded:

No mutagenic, carcinogenic or teratogenic effects of monuron in man or animals are known to have been reported.844

In the JRB Review the carcinogenicity of monuron

evaluated, as follows:

Male and female mice of two hybrid strains, C57BL/6 x CBH/Anf and C57BL/6 x AKR. were fed 215 mg commercial monuron/kg by gavage daily until the mice were 4 weeks of age (Innes, 1969:

Exhibit 659; Bionetics, 1968). Thereafter, the mice were given 517 mg monuron per kg of diet until they were killed for examination at

approximately 18 months of age. The authors reported that a significantly increased incidence of pulmonary adenomas occurred among the treated animals when the combined groups of treated

animals were compared with untreated controls (10/66 versus 20/338). The incidence of total tumor types, hepatomas, and reticulum cell carcinomas was not significantly different between control and treated animals. However, I ARC (1976) reported that tumor incidence reported in this study was significant only for

lung adenomas in males of one strain (6/16 versus 9/90). It is not clear from examining the

original report (Innes, 1969: Exhibit 659; Bionetics, 1968), which did not contain a

statistical comparison of test and control groups by sex and strain, how the conclusion made by IARC (1976) was formulated.

Rubenchik et al (1970) fed 50 mixed-breed and 45 C57B1 mice 6 mg monuron in milk once per week for 15 weeks. The mice were held for 27 months. The number and survival rates of the control mice were not specified. The first tumour observed in

the mixed-breed mice was found after 16 weeks, and in C57B1 the first tumour was discovered




after four weeks. At these times, 23 (46%) mixed breed mice and 26 (58%) C57B1 mice had survived. A variety of "reactive changes1 1 reportedly was found early in the mice, including lymphocytic

infiltrates, catarrhal inflammation,

proliferation of epithelium of the bronchi, and focal necrosis of the liver. It was not stated whether these changes occurred in control or treated animals or at which point during the

investigation these changes were noted. A total of 13 tumours was reported in the mixed-breed mice and seven tumours occurred in the C57B1 strain. Most of the tumours consisted of benign hepatomas, hepatocellular carcinoma, alveolar

carcinoma, and kidney cancer. The survival rate of the treated mice at the end of the experiment was not reported and no statistical analysis of the data was provided.

Separate groups of both sexes of two hybrid strains of mice were given single subcutaneous injections of 10 mg monuron per kg on the 28th

day of life and observed until they were

approximately 18 months of age (Innes, 1969: Exhibit 659: Bionetics. 1968). No significant increase in tumours was noted. I ARC (1976) commented that a single injection may not be an

adequate basis for discounting monuron-induced carcinogenesis.

Rats Hodge et al (1958) maintained groups of 30 male and 30 female Rochester albino rats on diets containing 0.0025, 0.025, and 0.25 percent

monuron for 2 years. Because of a respiratory infection, 70-90 percent of all groups, including controls, died. Of the remaining animals, gross and microscopic examination of all major tissues

revealed no carcinogenic lesions. I ARC (1976) noted that a lack of detail in reporting data made this study difficult to evaluate.

Fifty random-bred rats received 450 mg monuron per kg daily in their diet for 18 months and were observed for an additional 9 months (Rubenchick et al, 1980). The first tumour was discovered after 18 weeks, at which time 32 rats were still alive. The number of rats surviving for the 27-month experimental and observation period was not stated. Fifteen rats were found to have


tumours with no specific target organ excessively affected. No tumours were found in any of the 30 control rats. which IARC (1976) found unusual.

Summary and Evaluation

Two studies in which mice of different strains were fed monuron were able to demonstrate the induction of tumours. with the lungs and liver being the major target organs (Innes, 1969; EXHIBIT 659; Rubenchick et al, 1970). One of

these studies (Rubenchick et al, 1970) also showed that mixed breed rats fed monuron

developed tumours, but no specific target organ was apparent. The report by Innes, 1969;

Exhibit 659 (details of which are provided in Bionetics, 1968) used a small number of animals in test groups. Innes (1969; Exhibit 659)

recommended that monuron be subjected to further study. Serious reporting deficiencies are evident in the publication by Rubenchick (1970), especially with regard to data on control animals, which make evaluation of the study difficult.

Based on the information presented for monuron and in the absence of any human data, IARC (1976) concluded that the data suggest that monuron is carcinogenic. In light of the deficiencies of the data presented here, final conclusions should await publication of a chronic feeding study in mice and rats that is currently under way by the National Cancer Institute.845

These studies are rather doubtful in the view of the

Commission's expert consultants.

The National Toxicology Program reported the results of

the chronic feeding study in the "NTP Annual Plan for

Fiscal Year 1984" (NTP-84-023). It stated:


Carcinogenesis study results for monuron. Rats fed 0, 750 or 1500 ppm and tumour incidences for the three groups are given respectively. Clear evidence of carcinogenicity for male F344/N rats: tubular cell adenomcarcinomas of

the kidney (0/50, 1/50, 8/50*) and combined (0/50, 3/50. 15/50*); neoplastic nodules (1/50, 6/49, 7/50*) and carcinomas (combined) of the liver (1/50, 6/49, 9/50*). No evidence of carcinogenicity for female rats or B6C3F1 mice of either sex fed 0, 5000 or 10,000

ppm.(‘indicates statistically significant results).846

Thus, monuron must be a candidate carcinogen. However, as

the agent was not used by or near Australian personnel,

the Commission concludes that no veteran has or will

suffer cancer by reason of exposure to it in Vietnam.

Bromacil (HYVAR-X)

This compound is not mentioned in WAA's first submission

except in a paragraph which reads, "Other herbicides used

- monuron, dalapon, diquat, paraquat, polybor chlorate,

distillate/cresote, borate chlorate, bromici1 (sic),

picloram, discoxol, diuron, monuron, tandex, tordon 50-0,

trinoxol.(at p 51.)

In WAA's submission specifically relating to toxicology

bromacil is mentioned in a list over the words, "These

chemicals appear to have been used in small quantities in

South Vietnam", (at p 87.)


In its final submission W A A mentions bromacil only in the

following sentence, "Diuron and bromacil have mutagenic

activity in multiple bacterial test systems." (at p 33.)


The Commonwealth Department of Health concluded that,

"No mutagenic, carcinogenic or teratogenic effects of

bromacil in man or animals are known to have been


In November 1981 the Standing Committee on Pesticides and

Agricultural Chemicals and the Standing Committee on

Poisons Schedule, two committees of the National Health

and Medical Research Council (NH & MRC) considered a

report of an oncogenicity study in mice conducted by

Dupont (Australia) Limited.

The Poisons Schedule Standing Committee's view was, inter


Although the results showed no oncogenic

response, the Committee requested the Secretary to seek details from the company of the in vitro mutagenicity assays.

At the present meeting the committee received details of the in vitro mutagenicity on

bromacil. The reviewing toxicologist commented that all the tests had been adequately designed and that in the short term mutagenic tests


bromacil had been cleared. However there still remains doubt over a possible hepato carcinogenic affect, testicular atrophy and an inability to establish a no effect level in chronic studies.

The committee decided that there was not enough evidence to consider changing the existing exempt status for bromacil but requested the secretariat to collect more data. to be evaluated at the next meeting. 848

The Pesticides and Agricultural Chemical Standing

Committee concluded that all the mutagenicity data

supplied showed bromacil to possess no mutagenic 849 potential. NOEL of 10 mg/kg/per day remains in


W A A produced no evidence to support any carcinogenic

potential for this compound. Those Assisting found none.

The Commission finds that no Australian Vietnam veteran

has or will suffer cancer as a result of exposure to this

compound in Vietnam.

Cacodvlic Acid

The two compounds sodium cacodylate and cacodylic acid

were the active ingredients in Agent Blue.

In WAA's inital submission it is listed with sodium

cacodylic above the words, "Carcinogen (or co-carcinogen).


and can cause chromosomal damage and chronic illnesses -see L 9, page 23 et seq and L 37." The L numbers are

references to documents lodged with that submission. L9 is

a paper by Dudley et al and relates to 2,4-D

poisoning.851 L 37 is an EPA Document referring to

2,4,5-T and TCDD.852 Neither mentions cacodylic acid or

sodium cacodylate.

In its submission specifically relating to toxicology

questions W A A commences a six page review of some

literature and makes a number of bald allegations, (at p

65.) This review makes an entirely inappropriate equation

between organic and inorganic arsenicals. Cacodylic acid

is, of course, an organic arsenical, as is sodium


The National Research Council of Canada in 1978 reported.

The literature provides little evidence

associating arsenic exposure and the induction of cancer in animals.853

The Commonwealth Department of Health concluded:

Organic arsenicals have no known mutagenic or teratogenic effects in man or animals and no conclusive evidence of carcinogenic activity in laboratory animals is known.854

V I 11-300

The International Agency for Research on Cancer reviewed

the carcinogenicity of cacodylic acid in 1980. After

discussing a number of studies performed by the National

Technical Service which observed no increased incidence of

tumours between treated mice and controls the Agency


There is inadequate evidence for the

carcinogenicity of (organic) arsenic compounds in animals.855

The International Program on Chemical Safety in 1981


No conclusive evidence of carcinogenic activity has been reported for any of the organoar senic compound tested in experimental animals.856

The JRB Review concluded that the only study considered

was negative but limited by the small number of animals

* „ ^ 857 tested.

Dr Philippe Shubik dealt with the carcinogenicity of

cacodylic acid as follows:

Although the only term study on this compound was undertaken in 1969 by Innes et al (Exhibit 659) and has been deemed inadequate, the compound was reported as negative. The particular study must

be viewed in the context of the overall study in


which other compounds including DDT were reported to be carcinogenic.

From a biochemical standpoint, cacodylic acid seems most unlikely to be carcinogenic since it is a detoxification product. It would be guite inappropriate to equate the possible activity of this compound with the inorganic arsenical said to be carcinogenic in man.858

In his oral evidence, which the Commission accepts. Dr

Shubik elaborated on the prepared statement and he

concluded by saying:

Organic arsenicals have been considered to be free from suspicion as carcinogenic over the years and as I say in this instance I think that the detoxification patterns is very recent

information on this and it would make it

extraordinarily unlikely to be a dangerous compound to man.859

Dr Ian Munro also dealt with the carcinogenicity of

cacodylic acid both in his statement and in his oral

evidence. He said:

There is no evidence that cacodylic acid is carcinogenic, having been tested in two strains of mice.860

The Commission also notes that Agent Blue was used in

limited quantities in Phuoc Tuy province.861 The

Commission accepts Dr Shubik's and Dr Munro1s evidence.

Dr Brusik's evidence that cacodylic acid has to be

considered to be a candidate mutagen862 is noted.


Further, the Commission has considered that part of the

final submission of W M where arsenic is mentioned as a

carcinogen and where, whilst making reference to the

distinction between inorganic and organic arsenic in

another context, (neuropathy), it makes no distinction in

that regard in relation to carcinogenicity. The documents

therein relied upon863 have also been studied.

The Commission concludes that cacodylic acid and sodium

cacodylate are not carcinogenic and that no Australian

Vietnam veteran has suffered or is likely to suffer cancer

by reason of exposure to Agent Blue in Vietnam.


In WAA's original submission distillate-creosote is

referred to only as one of the herbicides used in

Vietnam.(at p 51. )

In its submission relating to toxicology kerosene is

mentioned but distillate-creosote is not.

The WAA's final submission makes no mention of these

products at all.


The International Programme on Chemical Safety reviewed

petroleum products in 1982. In connection with

carcinogenicity this review said:

Benzene and aromatic extracts are the only well defined petroleum solvents for which

carcinogenicity has been reported.864

Dr Robert Baker in a personal communication to the

Commission865 said, inter alia:

Since creosote is a mixture of polycyclic

hydrocarbons, it is hard to conceive that

creosote would not be mutagenic in in vitro systems and also carcinogenic in appropriate animal tests.

Coal Tar is a known carcinogen in man.

W A A called no witness to establish a connection between

cancer suffered by Vietnam veterans and the use of

distillate-creosote in Vietnam.

A mixture of creosote and dieselene was used for base

perimeter defoliation prior to 24 August 1967. It caused

problems however because tar precipitate from the creosote

clogged the nozzles of the spray equipment and apart from

a slight irritation of skin or throat it seems to have

V I 1 1-304

caused no acute problems. The usage seems to have been

confined to 1966 and 1967 and to perimeter foliage


Accordingly the Commission finds that the dieselene

creosote herbicide mixture is unlikely to have caused any

increase in the incidence of cancer among Australian

Vietnam veterans and is not likely so to do.


This compound is mentioned in WAA's original submission

simply as one of the herbicides used in Vietnam, (at p 51.)

In the WAA's submission related specifically to

toxicology borate-chlorate is not mentioned. Nor does this

compound rate a mention in the WAA's final submission.

To the best of the Commission's researches it has never

been nominated as a carcinogen nor have any of the

Agencies sought its testing for carcinogenicity.

Dr Susan Tepi, a consultant to the Commission, dealt with

the mixture in the following way:


Sodium chlorate and sodium borate are often combined in a 3 to 1 ratio and used as a

non-specific herbicide. This reduces the fire and explosion risk of sodium chlorate alone.866

Sodium chlorate data is sparse. The data which do exist

focus on the compound's ability to produce


The OHM-TADS (1984) data bank reports that "No chronic

effects to humans are known".

There is a similar lack of data concerning sodium borate.

It is unlikely that combining sodium chlorate and sodium

borate would cause toxicity different from that observed

after exposure to the chemicals separately.

The Commission is thus left in a situation where no

evidence implicating the borate-chlorate mixture has been

led by WAA. The Commission's own researches indicate no

carcinogenic hazard. Accordingly, the Commission finds

that no Australian has suffered or is likely to suffer

from exposure to it in Vietnam.

V I 1 1-306

Η.2. The Insecticides


In WAA's initial submission DDT gets reference only as an

insecticide referred to in the Army Report.(at p 51)

In its toxicology submission DDT is dealt with at length,

(at p 74.) The submission begins by conceding that DDT has

been described as being of low toxicity but emphasises the

word "toxic."

As everything is toxic if the dose be sufficiently high,

the Commission is at a loss to understand this

submission. The submission then goes on to relate the

symptoms of DDT poisoning when ingested by drinking or

eating; continuing with references to immediate symptoms

of heavy exposure. In the last line of the submission

regarding DDT. one finds the words "DDT is ... a

carcinogen".(at p 75.) No reference is given.

In its final submission W A A points to Dr Shearer's

evidence that DDT is oncogenic in mouse liver. Dr Shearer

relied for that statement on research work performed by

Tomatis et al reported in 1972.867 In Tomatis's


experiments CF-1 inbred mice were given DDT at dose level

of 2, 10, 50 and 250 parts per million for the entire life

span of the mice for two consecutive generations. This

exposure produced an increase in liver tumours, the

percentage of tumour-bearing animals being slightly higher

in DDT-treated mice, being most evident in mice exposed to

the highest level of DDT, i.e ., 250 ppm of DDT.868

The WAA's submission also points to the evidence of Dr

Sielbergeld who says in her statement869 that DDT

amongst other chemicals is mutagenic and carcinogenic.

DDT has been widely used by the Australian community for

many years. It is a moot point whether Australian

personnel in Vietnam would be more exposed to DDT than

most living Australians.

The controversy about DDT has been substantially fuelled

by the fact that DDT remains in the environment for long

periods after spraying and has been found stored in the

body tissue of animals as remote from its use as polar

bears. Nonetheless, a careful analysis of the data

concerning DDT is required.

WHO reviewed the toxicity of DDT on numerous occasions

during the 1960s. In its 1969 evaluation of DDT,


available carcinogenicity data on DDT was summarised as


Rhesus monkeys of mixed sex (12 males and 12 females) were divided into groups and fed over periods up to 7.5 years or longer on diets

containing 0, 5, 50, 200 and 5000 ppm. Biopsies were performed on several organs. The

histopathology gives no report of tumour

formation in any animals (Durham et al. 1963).

Mouse A total of 683 mice of the BALB/o strain, spread over five generations, were fed dietary levels of 2.8 - 3.00 ppm of DDT (0.4 - 0.7 mg/kg bw/day)

for six months. The background DDT content of the diet given to 406 controls corresponded to an intake of 0.03 - 0.05 mg/kg bw/day. Altogether leukaemia occurred in 85 test animals (12.4%) but

only in 10 controls (2.5%). The significance became apparent in the F3 generation. The

incidence of tumours in the DDT group attained significance one generation earlier (in the F2 group). However, a suspected transgeneration cumulative effect resulting in progressive

increase in tumour incidence with each generation was not substantiated when the results from all five generations became available. A total of 196 tumour-bearing animals (28.7%) was recorded

in the DDT group compared to 13 in the controls (3.2%). The most frequent tumour types were leukaemia, carcinoma of the lung,

haemangioendothelioma and reticulum cell carcinoma (Kemeny and Tarjan, 1966; Tar jan and Kemeny, 1969).

Groups of 18 mice of each sex, from two hybrid strains of mice were given DDT for 18 months. The dose of 46 mg/kg bw was given by gavage from the seventh day of age to the time of weaning at four weeks of age and thereafter DDT was added to the diet in a corresponding amount of 140 ppm. The incidence of hepatomas was 18 out of 35

treated males (51%) and 5 out of 24 treated

females (21%) compared with the control value of 13 out of 162 males (18%) and 7 out of 158 (0.6%) females (Innes et al, 1969).


Two strains of mice namely BALB/oJ and

C HeB^FoJ, were divided into groups„ each of which contained 100 male and 100 female animals of each strain. The groups were fed 0 or 100 ppm

of DDT in the diet for periods up to two years. In the BALB/oJ strain, there was no significant increase in tumours in the DDT-treated group when compared to the controls, but because of the high

incidence of mortalities in both groups of this strain, the results were considered to be questionable. In the C3HeB/FoJ strain the number of deaths was much lower. The females of

this strain displayed a 24% incidence of

hepatomas in the group fed DDT compared to 9% in the controls. There was, however, a lower

incidence of tumours at other sites in this group when compared to the controls, resulting in no overall increase in the total incidence of tumours. The incidence of hepatocarcinomas was equally low in treated and control groups of both

sexes in both strains (Fitzhugh, 1969).

Rat Groups of 12 male rats were subjected for two years to diets containing 0, 100, 200, 400 and 800 ppm of DDT. In another experiment, groups each of 24 rats (12 males and 12 females) were given, during the same period, diets containing 0, 200, 400, 600 and 800 ppm. Also additional groups of 24 animals received 600 and 800 ppm

incorporated in their feed in a dry state. In the groups receiving 400 ppm and above, an

increase in the mortality rate was seen in

relation to the dose. Apart from nervous

symptoms at doses of 400 ppm and above, typical liver lesions were found at all concentrations. Hepatic cell tumours were seen in 4 out of 75 animals and 11 other rats showed nodular

adenomatoid hyperplasia. The authors concluded that a minimum tendency for the formation of hepatic cell tumours was evident and that this feature was apparent only after 18 months of feeding (Fitzhugh and Nelson, 1947).

Trout Rainbow trout were fed varying levels of DDT ranging from 0 to 9600 ppm for periods up to two years. At the 75 ppm level hepatomas were

present in 7 out of 19 fish after 15 months of

V I 11-310

feeding DDT. Mortality was high in the 2400 and 9600 ppm groups (Halver, 1967).

Studies (sponsored by WHO) on the investigation of the potential carcinogenicity of DDT to mice or rats are presently being conducted at the International Agency for Research on Cancer, Lyon, France; the National Institute for the

Study and Cure of Tumours, Milan, Italy; the Institute of Experimental and Clinical Oncology, Moscow, USSR and the Research Institute of Oncology, Leningrad, USSR. The results of these

studies should be avaible in 1971.870

The available experimental data do not provide sufficient information to allow a definite evaluation of the potential carcinogenicity of

DDT, however, they strongly indicate that DDT ought to be extensively tested. In fact, the 1967 Joint FAO/WHO Meeting already recognised the need for further studies. Following this

recommendation, work was initiated by WHO and I ARC. The results of these studies will not be available until 1971. For these reasons a

definite decision on the potential hazard of DDT to man cannot be taken now. However, because the hazard to man from DDT has not been ruled out, it is recommended that uses of DDT should be limited

to those situations where there are no

satisfactory substitutes.

In the light of this consideration, it was

decided to lower the ADJ and change it to a

conditional ADJ in order to limit the use of DDT except where it is absolutely necessary.871

In 1974 the International Agency for Research on Cancer

undertook a full review of the toxicity of DDT. This

included an extensive analysis of the animal

carcinogenicity data. The Agency reported:

The hepatocarcinogenicity of DDT by the oral route has been demonstrated in several strains of mice. Liver-cell tumours have been produced in


both sexes „ and in CF1 mice some were found to have metastasized. Increased tumour incidences have been reported in some other organs; however, this finding was not confirmed in two recent multigeneration studies using a wide range of doses.

Oral administration studies in rats have provided no convincing evidence of carcinogenicity of DDT to this species.

In a single experiment, hamsters tolerated higher dietary intakes of DDT than did mice and rats and did not develop tumours in excess over the


The negative results obtained in feeding studies with dogs and monkeys cannot be regarded as conclusive, due to limitations in group size and duration of treatment.

Liver-cell tumour induction in the trout cannot be considered as conclusive until additional studies using properly controlled diets are reported.

Skin application and subcutaneous injection studies in mice were too limited in duration and/or group size to allow any conclusions to be made.

Two dose-response studies on liver-cell tumour response to DDT following oral administration in two strains of mice are available. In both CF1 and BALB/c mice, dietary intakes of 250 ppm

(corresponding to about 37.5 mg/kg bw/day) were highly effective in both sexes. The lowest dose used, 2 ppm in the diet (corresponding to an intake of about 0.3 mg/kg bw/day), induced a significant increase in the incidence of

hepatomas only in male CF1 mice.

The two DDT metabolites, p, p '-DDE and TDE (DDD), were tested by oral administration to mice. The latter produced a significant increase in lung tumours, while DDE was found to produce a high

incidence of liver-cell tumours.872

The WHO again dealt with the carcinogenicity of DDT in

1976. It said:


DDT has been shown to be carcinogenic in a two generation life-span study. (CF1 mice) at 2, 10, 50 and 250 ppm, all levels increased tumour incidence in females and two highest levels in males. While early work of DDT suggested that it might be tumorigenic in rats, no convincing evidence of carcinogenicity to this species has

been provided. The significance of the finding in mice for man cannot yet be assessed. No

increased incidence of tumours has been observed in formulators and spraymen exposed to excessive quantities of DDT for 20 or more years.873

WHO again reviewed the toxicity of DDT in 1979. In short,

it said:

If DDT is fed for long periods at dietary levels ranging from 2 mg/kg upwards for mice or 5 mg/kg upwards for rats, the changes in the liver

progress from hypertrophy, margination, and lipospheres in isolated, centrolobular

hepatocytes to the formation of nodules of affected cells. The first change has been

observed within 4 days of administration, the earliest time of observation. With

administration of dosages corresponding to those that people may encounter, changes in the livers of susceptible rodents require the entire lifetime of the animal to develop fully. At first, the nodules are microscopic in size, but some may become more than a centimetre in

diameter, particularly in mice, and show almost complete loss of lobular architecture. The same series of changes can be produced in rodents by other inducers of microsomal enzymes, including phenobarbital. Although there is persuasive evidence that these multinodular tumours of mice associated with changes in the endoplasmic reticulum are carcinomas, there is equally convincing evidence that they are not and the

views of some highly qualified pathologists in this matter remain diametrically opposed. More important than the question of classification is the fact that the entire continuum of changes

V I 11-313

from the prompt response in isolated cells to the eventual formation of tumours is peculiar to some rodents, and does not occur in other animals in which the endoplasmic reticulum does not respond morphologically in the same way.874

Cabral in "Carcinogenicity of Pesticides" (1980) dealt

with the carcinogenicity of DDT:

DDT is present everywhere in the human

environment and is stored by plants and animal tissues. As a consequence most food items are contaminated with it. In the US an extensive survey on ready-to-eat food commodities carried

out during 1964-1970 indicated that the daily intake of total DDT ranged between 0.0004 and 0.001 mg/kg/day, the higher values being

antecedent to 1968 (Duggan and Corneliussen, 1972). In 1969, the maximum acceptable daily intake (ADI) established by FAO/WHO was 0.005 mg/kg/day. Several studies in different

countries carried out during 1951-1969

demonstrated that the concentration of DDT in human milk ranges between 0.05 and 0.2 ppm (Laug et al, 1951, Egan et al, 1965; Heyndr ickx and Maes, 1969). Assuming an average concentration of 0.1 ppm total DDT in milk, a newborn baby

ingesting 700 g milk/day would have an intake in the order of 0.01-0.02 mg/kg/day. These intakes exceed, by 2-4 times. the DDT maximum ADI. The concentration of DDT in cow's milk is definitely

lower. The average daily intake in workers engaged in the manufacture of DDT has been

estimated to be 0.2 mg/kg/day. There are already some indications available that following the restrictive legislation approved in many countries DDT contamination of food is showing a steady decrease (Duggan and Corneliussen, 1972).

The evidence of carcinogenicity of DDT can be discussed in the light of data obtained in three types of situation: 1

(1) The concentration of DDT in tissues of

terminal cancer patients. One study showed an almost identical average concentration of


residues in human fat among 292 patients dying with cancer and 336 patients dying with other diseases (Hoffman et al, 1967). One other study (Radomski et al, 1968) showed that patients with

non-neoplastic liver disease had fat and liver concentrations of DDT residues higher than in controls.

(2) Studies on groups professionally exposed to DDT. No cases of tumour or blood dyscrasia were found in two separate investigations concerning respectively 40 men followed for 0.4-8 years

(Ortelee, 1958) and 63 men followed for at least 5 years (Laws et al, 1967) . The daily intakes were calculated to average respectively 200 and 50 times those of the general population. The

small number of workers studied, the short period of exposure and follow up render this study as of little relevance for asessing long-term risks of occupational exposure to DDT. There are no additional studies reported in the literature, designed to test the potential carcinogenicity of DDT in man.

(3) Long term studies in experimental animals. The first reference to the carcinogenicity of DDT was made by Fitzhugh and Nelson in 1947.

Technical DDT was fed to Osborne-Mendel rats during 24 months at dietary levels ranging from 100-800 ppm. Among the 75 rats alive after 18 months, 4 had "low-grade" hepatic cell-carcinomas and 11 had hyperplastic nodules. This rat study has been recently re-evaluated by Reuber (1978). The review of the histology has shown a

significant incidence of liver tumours, ranging form 36% to 54% in treated females and males. The incidence of liver tumours in controls was 0%.

Later studies by various authors with DDT in rats gave no evidence of carcinogenicity (Radomski et al, 1965) and others. Some of these studies were considered inadequate either because of low survival rate, insufficient data report or because the doses given were too low. More

recently Rossi et al, 1977, observed a high incidence of liver-cell tumours in Wistar rats fed 500 ppm DDT for their lifespan.

In our work with DDT in rats we have shown that concentrations of 125, 250 and 500 ppm DDT in the


diet induced an increase in liver-cell tumours predominantly in the females (Cabral et al, 1978). Ά study by Cabral et al (1977) used

Syrian Golden hamsters. The animals were fed for life a diet containing 125, 250 and 500 ppm DDT. The results of this experiment confirm that the hamster is resistant to the carcinogenic effects of doses of DDT which are carcinogenic for other species. In the previous hamster experiment

(Agthe et al, 1970) also did not give any signs of carcinogenicity.

Summary of Long-term Feeding Experiments with DDT Hepatocarcinogenicity

Species No of Experiments Positive Negative Mouse 8 6 2

Rat 7 3 4

Hams ter 3 0 3

The above table .. . shows a summary of long-term feeding s tudies with DDT: of a total of 18

experiments (9 were positive and 9 negative). The experimental evidence for carcinogenicity of DDT is actually based on the induction of

liver-cell tumours in mice (Tomatis et al, 1972; Terracini et al, 1973) multigeneration studies. The studies in rats have provided contradictory results. This discrepancy, as well as the

negative results with hamsters, suggest different metabolic pathways that we do not know yet the nature of. These could be important to

understand should we data to man. attempt to extrapolate the

Present knowledge on the carcinogenicity of DDT in animals does not allow us to predict with certainty that DDT will not have a carcinogenic effect in man. There are still many gaps to fill before an overall evaluation can be made, but on the whole i t seems that DDT presents a low

risk. As a matter of fact, a risk vs benefit

evaluation of DDT was made by WHO (1973).

Because of the use of DDT in world malaria

control programs, saving the lives of millions of people, it was concluded that the benefits obtained from the usage of DDT outweighed its possible risk, (emphasis added).875

V I 11-316

In 1981, the report of a Working Group of Experts prepared

for the Commission for the European Communities,

Directorate-General for Employment and Social Affairs,

Health and Safety Directorate also included a lengthy

review of the carcinogenicity data on DDT.876

The International Agency for Research on Cancer again

reviewed the carcinogenicity of DDT in 1982 and concluded

that there was "sufficient evidence" for the

carcinogenicity of DDT in mice. It said:

DDT is carcinogenic in mice following its oral administration, causing benign and malignant liver neoplasms (IARC Monographs. 5, 83-124, 1974; Kashyap et al, 1977) and lymphomas and

lung neoplasms (Kashyap et al, 1977). Following its subcutaneous injection to mice, it produced liver tumours (Kashyap, 1977). Oral

administration to rats caused liver neoplasms (Rossi et al, 1977; Cabral et al, 1982a). Two feeding studies with hamsters were negative (IARC Monographs, 5, 83-124, 1974; Cabral et al,

1982b), while feeding studies with dogs and monkeys were inconclusive (IARC Monographs 5, 83-124, 1974). DDT slightly increased the incidence of liver neoplasms in rats previously

exposed to N-nitrosodiethylamine (Nishizumi, 1979).877

Dr Philippe Shubik dealt with the carcinogenicity of DDT

as follows:

DDT induces liver tumors readily in mice and less so in rats and not at all in hamsters. There is a considerable amount of work on the metabolism

VI 11-317

of this compound encouraged by these interesting species differences.

There is a great deal of frustration present in many of the toxicologists that several decades after the introduction of this widely used, persistent and easily detectable compound, no one

is prepared to say whether or not it has proved to be safe in man. My view is that if hazard

were present there should have been some

indication of it by now.

Efforts to mount studies in which levels of DDT in human fat are related to tumor incidences have foundered. One can do no more than have a common sense opinion in this instance; my personal view is that it is most unlikely that DDT is a human health ..hazard in any respect. (emphasis added.)

Dr Shubik also dealt with the carcinogenicity of DDT in

his oral evidence during questioning by Counsel

Assisting. He said:

.... DDT is an extraordinarily difficult problem to pronounce upon. There is no question that DDT induced liver tumours readily in mice of a

variety of strains. It is not just a matter of enhancing the tumours in a specific strain where high instances of these tumours occur. The rat data with DDT is much less satisfactory and there are those who would argue with its validity at all. The hamster data I have to think is good because my name appears on the paper saying that it is negative with my old colleague„ Dr Cabral„ in our International Agency for Research in Cancer. We, and that is my group when I was in the States, did a lot of work on the metabolism of DDT because of the interesting species

differences that you can in fact do anything to a hamster with DDT, you cannot make it ill, you have to drown it in DDT to produce an effect. The frustration amongst toxicologists that exists here is one that illustrates, I think, in many ways the massive public pressures that many


people are subjected to. I was an adviser to the World Health Organization's Committee on Pesticide Residues last year and DDT was one of the compounds discussed and the group there were

bound and determined at the outset of the meeting to make a statement to the effect that they felt that there was not anything wrong with DDT and by

the end of ten days of hassling they were unable to formulate a statement that they felt they could put on paper. I find this is a rather sad state of affairs in a sense. I and many others believe that there would be now certainly have been some manifestation of a hazard in man from DDT should it indeed be carcinogenic in the

human. In view of the amounts of DDT used

particularly during the second World War and just after it and the amounts used in many countries since that time one would have believed that something should have emerged from cancer

registries, but there appears to be nothing there. There are those who say, well, we really cannot produce a negative and this is a difficult state of affairs. The International Agency for Research in Cancer at one stage of the game

subsequent to their own work on the

carcinogenicity in animals collected vast numbers of samples of human fat which were to be analysed and see whether or not one could relate the

incidence of cancers to the levels of DDT found in people; one would have thought that the fact that you could still find it many years later, that you would then have these medical records,

but somehow or other this particular effort foundered. I think to a great extent it was

because of some of the complexities in shipping specimens and doing these analyses. It was something I looked forward to seeing results from with great enthusiasm, but this thing was never done. I can only say that one can do no more

than to have a commonsense opinion in this instance. Here is a compound which persists in human fat, which has been used at an enormous level; at one stage of the game I can tell you

from personal investigations that I was involved in that the level of DDT in human milk in

Californians was over 7 parts per million. That was at a time when the tolerance level

established by the US Department of Agriculture was half a part per million. As a matter of

fact, the State of California submitted a


petition to the Federal Government saying that someone ought to do something about regularizing the situation between human milk and cows milk pointing out how impossible it was to meet those levels. Millions and millions of gallons of milk were thrown away about 20 years ago because of

this situation. At this stage of events. I can only say that from a commonsense point of view, it seems inconceivable that DDT really poses a carcinogenic hazard and no more than that.

(emphasis added).8/y

Dr Bernard Stewart in his written statement to the

Commission referred to the above IARC 1982 review of the

carcinogenicity data on DDT and the conclusions from the

World Health Organization 1979 Environmental Health

Criteria Document. namely:

The occurrence of tumours in some rodents only casts doubt on the significance of the phenomenon and on extrapolation of the findings to man.

There is not currently available evidence to suggest that DDT is tumorigenic or carcinogenic in man. (emphasis added)880

Dr Stewart then went on to say:

The fact that people who have been heavily

exposed to DDT have not developed cancer suggests that rodents are more sensitive to this chemical than humans. Careful consideration might be given to DDT exposure as a carcinogenic hazard if quantities used in Vietnam has led to a higher exposure than ever previously recorded for DDT. Since this does not appear to be the case, DDT cannot be judged as having offered a carcinogenic

hazard to Australian servicemen in Vietnam.88^


The Commission took Dr Stewart to mean that there was no

increase amongst those exposed. Of course DDT-exposed

people have contracted cancer but not in increased numbers.

Dr Stewart also dealt with the carcinogenicity of DDT in

his oral evidence before the Commission during guestioning

by Counsel Assisting. He said:

. . . DDT is almost the archetype of environmental chemicals which present the difficulty of extrapolating animal data to human beings. DDT is carcinogenic in mice following its oral administration, causing benign and indeed, malignant liver neoplasms and lymphomas and lung neoplasms in such mice. Following its

subcutaneous injection to mice, it produces liver tumours, oral administration to rats causes liver cancer. Two feeding studies in hamsters were negative, while feeding studies with dogs and monkeys were inconclusive - the distinction there

being these studies were restricted in their interpretation by some factors. DDT exhibited slight promoting activity in rats previously exposed to nitrosodiethylamine. I usually pronounce this as diethylnitrosomine nitrosodiethylamine is the correct chemical name. That test system is the same test system as used to demonstrate the presumed promoting activity of TCDD. Those data were considered sufficient evidence for DDT in experimental animals, that is to say, the Agency was of the opinion that DDT should be regarded as if it

caused cancer in humans. If we can leave that finding just for a moment, DDT was negative in a large range of short term tests to which it has been sub i ec ted . The available epidemiology are considered by the Agency inadeguate to evaluate

the carcinogenicity of DDT. Indeed, these studies include findings in human populations known to have been exposed to high levels of DDT. The World Health Organization through its Geneva arm also noted that tumours were produced


in rodents only and cast doubt on the

significance of these phenomena for extrapolation to human beings. There is not currently

available evidence to say that DDT is

carcinogenic for human beings. This is a

statement that is made not because there is no data but because there are data and those data do not indicate that DDT is carcinogenic. Again, we have the situation on the one hand with the

Agency saying it should be regarded as if it is but when the human data are actually available then these data do not provide indication of a carcinogenic effect. The theoretical inference would be that fortunately those human beings that

have been exposed have not been exposed to

sufficient quantity to produce a carcinogenic effect. The only other possibility is that human beings are simply resistant to the carcinogenic activity of this kind. The fact that the human beings who have been exposed to DDT have not developed any cancer, and bearing in mind the massive use of DDT in many countries of the world

for anti-malarial programmes and the like, suggest that indeed DDT is not carcinogenic in humans or indeed the amount of DDT that humans ingest by virtue of occupational exposure or environmental exposure is not sufficient to cause such a carcinogenic effect to be observed. It

therefore seems unlikely that DDT may be

anticipated to be causing cancer in personnel who were exposed to it in Vietnam." (emphasis


In October 1984 the FAO held a Joint Meeting on Pesticide

Residues in Rome. The joint meeting drafted a report

which concluded as follows:

Repeated exposures of workers for 25 years at an average dosage of .25 milligrams per kilogram per day is without any adverse effect and this may be taken as a no effect level for man. From

epidemiological observations of humans and a three generation study in dogs of dosages of up


to 10 milligrams per kilogram per day, together with other studies in rodents and rabbits there is no firm evidence that DDT has a reproductive or teratogenic effect. All epidemiological

studies in man have indicated that DDT is not carcinogenic for humans. ... (although the mouse is sensative to DDT) there is no significant risk of DDT producing tumours in man.883

An overall NOEL of 0.25 mg/kg/per day was confirmed.


The Commission therefore finds that DDT is not measurably

a carcinogenic hazard when used as an insecticide and that

no Vietnam veteran has suffered or is likely to suffer

cancer as a result of his exposure to this compound in



This insecticide was one to which every Australian serving

in Vietnam was exposed, and exposed in large quantities.

It is dealt with in VVAA's initial submission at page 56.

The submission recounts the effects of immediate and acute

poisoning by malathion.


The Merck Index, tenth edition, describes the toxic

symptoms of malathion as similar to but less serious than


In its submission related specifically to toxicology at

pages 79-81 W A A repeats the symptoms of acute malathion

poisoning, this time extracted from a NIOSH publication of

June 1976.884 This document makes no allegation of


In its final submission W A A at page 17 refers to Dr

Shearer's statement in which, after coupling malathion

with diazinon, she claims malathion is carcinogenic in

rats and mice.

The submission continues that malathion is also

potentially a mutagen.

Dr Silbergeld's evidence is referred to but she makes no

allegation that malathion is carcinogenic. The final

submission seems also to rely upon Dr Silbergeld's

argument that malathion inhibits antibody formation and

therefore interrupts immune suppression.


Amongst diseases caused by viruses (and therefore combated

by the immune system) are Kaposi's sarcoma, Burkitt1s

lymphoma and some leukaemia. Thus it is said Dr

Silbergeld's evidence supports an allegation that

malathion is carcinogenic.

Since malathion was undoubtedly absorbed by every

Australian in Vietnam, the Commission proposes to deal

with this suggestion in some detail.

WHO reviewed the toxicity of malathion on numerous

occasions during the 1960s and again in 1977 when it said:

There are no reports of an experiment primarily concerned with carcinogenicity. However, no increased incidence of cancers has been reported in several long-term dietary studies in rats.885

The Commonwealth Department of Health reported on the

carcinogenicity of malathion. in 1981, saying:

Examination of malathion for possible

carcinogenicity has proved negative in rats and mice (WHO, 1977; National Cancer Institute, 1978) .886

The IARC reviewed the carcinogenicity of malathion in 1982

and said:


Mouse: Groups of 50 male and 50 female weanling Charles River B6C3F^ mice were fed diets

containing 8000 or 16000 mg/kg malathion (purity 95%; impurities unspecified) for 80 weeks and observed 14-15 weeks. A matched control group comprising 10 animals of each sex was observed for 95 weeks; as the number of animals in the matched control group was small, an additional pooled control group of 50 animals of each sex was used for statistical analysis. In the

high-dose group, which received a maximum tolerated dose, 94% of the males and 88% of the females were still alive at the end of the

experiment: survival was somewhat lower in the low-dose and control groups. In female mice, no statistically significant increase in tumour incidence was found. In male mice the incidences of hepatocellular carcinomas plus neoplastic

nodules were: 2/10 in matched controls. 8/49 in pooled controls, 7/48 in the low-dose group and 17/49 in the high-dose group [Cochran-Armitage test for positive trend, p=0.041 (using matched controls) or p=0.019 (using pooled controls); Fisher exact test, high-dose versus pooled

controls, p=0.031]. When a time-adjusted analysis was performed, eliminating those male mice that died before week 52 of the study, the

following incidences resulted: matched controls, 2/9; pooled controls, 8/49; low-dose, 8/48; low-dose. 7/47; and high-dose, 17/49. Neither the Fisher exact test nor the Cochran-Armitage test showed these incidences to be significant when the matched controls were used; when the

incidences of hepatocellular carcinoma and neoplastic nodules were analysed separately, none of the results were significant (National Cancer

Institute, 1978). (The Working Group noted the short duration of treatment).

Rat: Groups of 20 male rats (age and strain not given) were administered malathion (purity 65%; impurities unspecified) in the diet at

concentrations of 0, 100, 1000 or 5000 mg/kg for two years. Survival was 60, 75. 90 and 65%

respectively. No compound-related lesion was reported by the authors (Hazleton & Holland, 1953). (The Working Group noted the small number of animals and the incompleteness of the



Groups of 50 male and 50 female weanling

Osborne-Mendel rats were fed diets containing various levels of malathion (purity, 95%; impurities unspecified): initially, animals were given 8000 and 12000 mg/kg of diet; then, because

the high dose appeared to be too toxic, the level given to the high-dose group was reduced after three weeks to 8000 mg/kg, and that to the

low-dose group to 4000 mg/kg after 14 weeks. The total dosing period was 80 weeks. Time-weighted average doses over this period were 4700 and 8150 mg/kg of diet. The low-dose group was observed

for 33 weeks and the high-dose group for 29

weeks. Two groups of matched controls comprising five or 10 animals of each sex were observed for 108-113 weeks; as the number of animals in the matched control groups was small, an additional pooled control group of 55 animals of each sex was used for statistical analyses. In the

high-dose group, which received a maximum tolerated dose, 58% of the males and 67% of the females were still alive at the end of the

experiment; survival was higher in the low-dose and control groups. In male rats, no

statistically significant increase in tumour incidence was found. In female rats a

statistically significant increase was observed only for the combination of thyroid

follicular-cell adenomas and follicular-cell carcinomas; 0/15 matched controls, 0/46 pooled controls, 0/48 . low-dose animals and 4/49 high-dose animals (Cochran-Armitage test for positive trend using pooled controls, p=0.026). Historical female control rats were reported to

have a spontaneous incidence of 2.2% for both types of tumour (National Cancer Institute, 1978). (The Working Group noted the short

duration of treatment and small matched control groups).

Three groups of 50 male and 50 female Fischer 344 rats, six weeks old, were fed diets containing malathion (purity 95%. impurities unspecified) at concentrations of 0, 2000 or 4000 mg/kg for 103 weeks. They were observed for a further two to

three weeks and then killed; surviving rats in the matched control group were killed after 105-106 weeks on study. Of the male rats, 88% of the control group, 86% of the low-dose group and 80% of the high-dose group survived the


experimental period; while of the females, 95% of the control group, 98% of the low-dose group and 90% of the high-dose group were still alive at

termination of the experiment. Females may not have received a maximum tolerated dose, as indicated by gain in body weight. No

statistically significant increase in tumour incidence was found in female rats. In male rats, the incidence of adrenal pheochromocytomas appeared to be increased in the low-dose group

(11/48) compared with the control group (2/49, p=0.006), whereas in the high-dose group only 6/49 pheochromocytomas were seen (National Cancer Institute, 1979a).

(b) Carcinogenicity of metabolites Mouse: Groups of 50 male and 50 female B6C3F^ mice, seven weeks old, were fed diets containing malaoxon (0,0-dimethyl S-l,2-bis(ethoxycarbonyl)

ethyl phosphorothioate) (purity greater than 95%; impuritities unspecified) at concentrations of 0, 500 or 1000 mg/kg for 103 weeks. They were

observed for a further one to two weeks and then killed; surviving mice in the matched control group were killed after 103-105 weeks of study. The high-dose group received a maximum tolerated dose. Of the male mice, 90% of the control

group, 84% of the low-dose group and 74% of the high-dose group were still alive at week 103; while of the female rats, 78% of the control group, 76% of the low-dose group and 90% of the high-dose group were alive at that time. No statistically significant increase in tumour

incidence was observed in animals of either sex (National Cancer Institute, 1979b).

Rat: Groups of 50 male and 50 female Fischer 344 rats, six weeks old, were fed diets containing malaoxon (purity, greater than 95%, impurities unspecified) at concentrations of 0, 500 or 1000 for 103 weeks. They were observed for a further one to two weeks and then killed; surviving rats

in the matched control group were killed after 103-105 weeks of study. The high-dose group received a maximum tolerated dose. Of the male rats, 80% of the control group, 82% of the

low-dose group and 64% of the high-dose group were alive at week 90 of study; while of the

female rats, 82% of the control group, 90% of the low-dose group and 80% of the high-dose group


were alive at that time. No statistically

significant increase in tumour incidence was observed in male rats. In female rats„ there was a statistically significant increase in the incidence of thyroid C-cell adenomas and

carcinomas combined (National Cancer Institute, 1979b) . (The abnormally low incidence of these tumours in the matched controls, in comparison with that in historical controls (See Tarone et al, 1981) was noted).887

(Summary) The available data do not provide evidence that malathion or its metabolite. molaoxon is carcinogenic to experimental animals.(emphasis added).

The American Cyanamid Company supplied the Commission with

information concerning malathion. As an attachment to a

paper entitled "Understanding Malathion and the

Mediterranean Fruit Fly Eradication Program" (1981), the

Company said:

Malathion has been subjected to three bioassays for possible carcinogenicity in the

Carcinogenesis Testing Program conducted by the National Cancer Institute (NCI) and was not carcinogenic in any of these studies. In the first two bioassays (National Cancer Institute,

1978), the test animals were rats of the

Osborne-Mendel strain and B6C3F1 mice. The third bioassay (National Cancer Institute, 1979a) used only rats of the F344 strain. In all three of these bioassays, NCI concluded that malathion was not carcinogenic. A similar test was carried out, and a similar conclusion reached, with malaoxon fed to F344 rats and B6C3F1 mice

(National Cancer Institute, 1979b) . These tests were reviewed by the EPA and the EPA concurred with the NCI's conclusions. More recently, Reuber (1980) in excerpts of an unpublished

article has questioned the validity of the NCI conclusions, and contends that his interpretation of the same data shows malathion to be

carcinogenic in all three studies.


Reuber1s (1980) claims cannot be evaluated since the paper has not been published and only five pages of a typewritten manuscript are available

for examination. These pages do not contain any specific data on tumor incidence. but only statements to the effect that tumors of certain

types were found at a variety of sites, and that benign and malignant neoplasms increased in both male and female rats receiving malathion. His conclusions with respect to the carcinogenicity of malathion for mice appears to rest on the occurrence of higher numbers of liver tumors in treated male mice.

The simple fact that tumors were found in treated rats is of no significance, since aging animals of this species tend to develop a certain

incidence of neoplasms regardless of treatment. Either Reuber disagrees with the statistical analysis of the data performed by NCI or he

believes he has found a higher incidence of tumors as a result of his own examination of the histological material. Whether one or both of these factors applies cannot be known until more

information is available. However, for the present, we prefer to rely on the NCI conclusions since the histopathological examination conducted by the experts directly involved in these studies is very rigorous.

In the case of the mice, Reuber bases his

conclusions on a slightly higher incidence of liver tumors, compared to the "matched" controls, together with the occurrence of neoplastic nodules of the liver in the high-dose males. The NCI analysis (National Cancer Institute, 1978) of

these data finds that none of the direct

comparisons of treated groups with controls is statistically significant. Furthermore, NCI discounts the significance of the 22% incidence of hepatocellular carcinoma in the high-dose males by pointing out that several bioassays of other compounds in the same program had control groups with incidences of 35-40% of the same tumour.

At present, the weight of the evidence must be considered to rest on the conclusion that

malathion has not been shown to be carcinogenic for mice or rats.889


In a further paper "Malathion Research" (1984) American

Cyanamid said:

Malathion is not carcinogenic. There is no evidence, either from human populations exposed as a result of occupational exposure or aerial spraying programs, or from animal tests conducted

in laboratories, that malathion will initiate or promote cancer.

In 1982, the California Department of Health Services scrutinized all of the available data on the toxicology of malathion prior to its use in the Medfly eradication program in Santa Clara County, and reported:

1 Taking the premise that malathion may be

carcinogenic and using conservative (i.e. health-biased) assumptions and extreme

1 worst-case' examples, the risk of cancer from the malathion applications in the Santa Clara population has been estimated and found to be insignificant1.

The California epidemiologists estimated the actual risk of cancer to an exposed individual to be about one in a billion.

The National Cancer Institute (NCI) found malathion to be non-carcinogenic in three separate bioassavs conducted as part of its Carcinoqenesis Testinq Proqram. The US

Environmental Protection Aqencv (EPA) has reviewed these tests and has accepted the NCI's studies.

The first bioassay of malathion was conducted by the Gulf South Research Institute, under contract to NCI. In this study, 50 rats and 50 mice were

administered malathion in their feed for 80 weeks. They were then observed for several weeks prior to examination for cancerous growths.

The report concluded that "under the conditions of this bioassay, there was no clear evidence of


the association of the tumour incidence with the administration of malathion to Osborne-Mendel rats or B6C371 mice."

In the second bioassay, groups of 49 to 50 rats of each sex were fed diets containing 2000 or 4000 ppm malathion for 103 weeks. No tumors occurred in the dosed groups of rats of either sex at incidences which could be clearly related to malathion.

In the third bioassay, malaoxon, the oxygen analogue of malathion, was similarly evaluated, and once again, no evidence of carcinogenicity was discovered.

Each of three bioassays was subjected to intense scientific scrutiny. Eight scientists at Gulf South Research Institute designed, administered and supervised the bioassay. A statistical analysis of the results was performed by two additional scientists and reviewed by a third, also at Gulf South Research Institute.

The results were then analyzed by four members of the Mathematical Statistics and Applied

Mathematics Section of NCI. Nine more scientists at NCI were responsible for evaluating the bioassay experiment, interpreting the results and reporting the findings. None of these scientists found any reason to conclude that malathion, under the conditions of the bioassay, initiated or potentiated cancerous growths in the

laboratory animals.

The EPA then assigned a team of scientists to review NCI's findings. With one exception, this team of scientists also found no evidence of

carcinogenicity. Unfortunately, it is difficult to evaluate this single criticism since it has not been published or subjected to peer review.

More recently, an independent laboratory completed a two-year rat feeding study on behalf of American Cyanamid to assess the chronic toxicity of malathion.

It concluded:

'No correlation with clinical or

histopathological findings were evident. No


biologically or toxicologically important test material-related histopathological effects were noted. No tumorigenic effect of Cythionr(Malathion) was noted'.

Kim Hooper of California's Hazard Evaluation System and Informatioin Service stated that even if malathion were 'positive' as a carcinogen, 'it would be weaker than saccharin.' And, according

to Dr Bruce Ames of the University of California at Berkeley, noted researcher on environmental carcinoaens and mutagens, the risk of cancer from the spravinq of malathion is roughly ' the

equivalent of putting a can c>f diet soda on your front lawn.'

Recently, Dr Ames pointed out that large numbers of potent carcinogens arise naturally in most fruits and vegetables. 'Organic chemists have been characterizing them for 100 years,' he stated, 'and new plant chemicals are still being discovered.' He also stated that 'the human dietary intake of 'nature's pesticides' is likely

to be several grams per day - probably at least 10,000 times higher than the dietary intake of man-made pesticides.(emphasis added)°9Q

Dr Philippe Shubik in his written statement stated:

This compound has been bioassayed in the US by mouth in rats and mice and has been pronounced non-carcinogenic. The only suggestions to the contrary have not been based on any data.

It would seem most unlikely that this compound poses a potential carcinogenic hazard, and in my opinion, its use as an insecticide has and will cause no cancers in veterans.89-

In his oral evidence he said:

There again I have read the bioassay and believe it does certainly substantiate the view that malathion is non-carcinogenic. There has been a


lot of noise about malathion, particularly in the United States in the press and so on, after its use in California on a large-scale and so on, but I do personally believe all the evidence suggests

that this does not pose a carcinogenic hazard and that in fact it is amongst the most non-toxic in all respects of all pesticides.11892

Dr Bernard Stewart in his written statement stated:

"Malathion and its metabolite malaoxon were tested for carcinogenicity in mice and rats by administration in the diet. No evidence of carcinogenicity was found. The available data provide no evidence that malathion is likely to present a carcinogenic risk to humans.

Malathion may be excluded as contributing to any carcinogenic hazard hypothesized to exist in Vietnam.(emphasis added.)893

Dr Stewart in his oral evidence, said:

Malathion?--- Malathion has been tested quite extensively for carcinogenicity in rats and mice, no evidence of carcinogenicity has been found. This is one of the few

chemicals where the agency gives a

specifically negative statement. That is to say, it is judged to be not carcinogenic and therefore it is judged - the agency used the term, no evidence that malathion offers a carcinogenic hazard to humans. It is a

product of the fact that only chemicals that have some suspicion of carcinogenicity are ever tested, but there are so few compounds where a definitively negative statement can

be made.

But the view that you express--- ?----On the basis of that, I concur with their findings, I should say that of all the chemicals with which we have to be concerned, malathion is

the one that should be given the lowest

level of possibility of being involved.

V I11-334

Indeed, the view that you have expressed is that it can be excluded?--- Yes. I cannot

imagine stronger evidence. If the people manufacturing this material were

subsequently shown to develop cancer. well, obviously, such an evaluation would have to be changed. but on the basis of evidence

before us, it can be excluded. (emphasis added)894

Dr Ian Munro dealt with the carcinogenicity of malathion

in his oral evidence during questioning by Counsel

Assisting. He said:

Carcinogenicity of malathion, and of its

metabolite malaoxon have been studied very extensively in both rats and mice in many, many studies and no increase in tumors have been noted. In fact, the International Agency of

Research on Cancer had done an extensive

evaluation of malathion just two or three years ago, and it has come to the conclusion that all the evidence indicates that it is not


The Commission therefore concludes that malathion is not

carcinogenic. It follows that Australian personnel who

served in Vietnam have not suffered and are not likely to

suffer cancer by reason of their exposure to that compound

in Vietnam.


W A A 1 s initial submission includes a 2 1/2 page

description of the effects of dieldrin including a listing


of the symptoms of acute poisoning from dieldrin.(at p

58.) This list does not include an allegation that

dieldrin is carcinogenic.

In its submission related specifically to toxicology W A A

deals with dieldrin at pages 76-78. Quotations from the

major texts by Martindale and Wayland Hayes are included

but no allegation of carcinogenicity is made in those

quotations. The last sentence reads. "Dieldrin is, in

addition to other toxic effects, a carcinogen and a


In its final submission W A A says:

It damages human cells in culture and is oncogenic in mice and rats.

Dr Shearer, Exhibit 1615, page 3 and reports cited. See also Exhibit 1653. p. 3, paragraph 5. (at p 33.)

Exhibit 1615 is Dr Shearer's statement which baldly

asserts the proposition in the submission. It provides

three references to support the "oncogenic" statement.

These became Exhibits 1652, 1653 and 1654. The first,

Davis et al, recounts the feeding of dieldrin to mice at

10 ppm daily in the diet and the finding after a year of

an increase in benign liver tumours which was

statistically significant.

V I 1 1-336

The second is the EPA 1974 findings re Die ldr in Hazard

which of course reflect the influence of the Delaney


It is not a scientific paper.

The third does not relate to cancer but recounts the

production of enlarged livers in rats and fatty and

atrophied liver in dogs with regular daily doses of

dieldrin in the diet for 2 years.

Extrapolation from large oral doses for the life time of

experimental animals to humans environmentally exposed is


This material was however of sufficient concern to warrant

an intensive investigation by the Commission.

WHO reviewed the toxicity of dieldr in in 1965 and as to

its carcinogenicity said:

Mouse Groups of approximately 200 young CgHeB/Fe mice, equally divided by sex, were fed a diet containing 10 ppm of dieldrin for their life-span

(maximum 2 years). The dieldrin shortened their average life-span by 2 months, as compared with

VI11-3 37

an equal number of controls, and significantly- increased the incidence of hepatic tumours (Davis & Fitzhugh. 1962).

Rat Groups of 40 rats (20 male and 20 female) were fed diets containing 2.5, 12.5 and 25 ppm

dieldrin for 2 years. The liver/'body-weight ratio increased and characteristic histological liver damage was seen at all dosages (Treon & Cleveland, 1955).

With groups of 16 female rats each, dieldrin was incorporated in the diet for three generations at 2.5, 12.5 and 25 ppm. Two litters of offspring were taken from each generation of these groups. The presence of dieldrin in the diet at 2.5 and

12.5 ppm initially reduced the number of

pregnancies but this effect tended to disappear with continued feeding of the diet. All doses increased the mortality among the suckling young. The effect on survival during suckling was severe at 12.5 and 25 ppm dieldrin (Treon & Cleveland. 1955).

In a 2-year experiment, groups of 24 rats (12 male and 12 female) were given 0.5, 2, 10, 50, 100 and 150 ppm dieldrin. Concentrations of 50 ppm and above increased the mortality rate in a dose-response relationship. The

1iver/body-weight ratio increased and

characteristic histological lesions occurred in the liver at all levels; these were minimal at 0.5 ppm but increased in severity with increasing dose. There was an increase in the number of tumours in the experimental groups, especially at the lower levels of feeding, in contrast to the control group (Fitzhugh et al. 1964).

In other experiments, 40 males and 40 females were given dieldrin in the diet at a

concentration of 75 ppm for 440 days. Twenty animals of each sex were used as controls. All the females and 22 males of the experimental group and 5 control males died spontaneously before the end of the treatment. Seven males were killed in good conditions between 300 and

440 days and the last 11 males were killed after 440 days. The liver/body weight ratio was

markedly increased in the rats killed during

V I 1 1-338

exposure, but it returned to normal after the end of it. "Lesions of the hepatic parenchyma that have been considered typical of exposure to the organochlorine insecticides in rats" were observed only in healthy animals killed during the treatment. Rats dying spontaneously or killed after withdrawal of the insecticide did not show such type of change (Hunter et al. 1964).

Ewes Thirty-six ewes were given 0, 1. 5 or 25 ppm of dieldrin in their diets for a period of 40 months including 4 gestation periods. At 25 ppm lambs died shortly after birth. Liver function as well as other physiological tests on the ewes did not

show any changes related to the treatment with dieldrin (Shell. 1963 ).

Comments on experimental studies reported The primary mode of action of dieldrin is on the central nervous system. This is the mechanism of death in acute poisoning. Symptoms of central

nervous system stimulation are also seen after repeated high doses. Repeated doses at lower levels give rise to liver damage and in this respect young dogs are more susceptible than rats.

In one long-term rat feeding experiment, there was a general increase in tumour production in the experimental animals at the lower dosage

levels as compared to the controls. but the liver was not particularly affected.

Liver tumours, however, were significantly increased on a dosage of 10 ppm in one strain of mice susceptible to the development of these tumours. Some evidence has been presented that

the same effect can occur in another strain of mice, but this requires more information.

In rats, it has been suggested (Fitzhugh et al, 1964) that the apparent tumorigenic properties of dieldrin could be related to a 1 general ‘ type of


Then, again. at the WHO Joint Meeting on Pesticide

Residues in 1970, further dieldrin carcinogenicity studies

were reported, namely:


Mouse Groups of equal numbers (125-300) of male and female mice (CF1 strain) were fed dietary levels of 0, 0.1, 1.0 or 10 ppm of dieldrin for periods of up to 132 weeks. In addition, a positive

control group of 50 mice received 600 ppm of the carcinogen, 4-amino-2,3-dimethylazobenzene (butter yellow). Feeding was started at four weeks of age. The presence of liver tumours was detected initially by palpation, which was

started on all animals after 16 weeks of feeding, and the animals were sacrificed when the tumours becamae large enough to endanger health. No effect on health or behaviour was evident in the first nine months of feeding, and no liver

tumours were detectable before 37 weeks at any dose level of dieldrin. The morbidity of the mice receiving 10 ppm of dieldrin increased after nine months and at 15 months 50% of both sexes had either died or been sacrificed because of tumour size. The control group normally died after 20-24 months, and the life-span of the 0.1 and 1 ppm group was the same; these groups

displayed no palpable masses. All the positive control group died after 14 months, although feeding of butter yellow had been terminated after six months when the first hepatoma

appeared. Liver tumours increased in all test groups but not in other tissues. Tumours were of two types: type (a) consisted of solid cords of closely packed parenchymal cells with morphology and staining characteristics similar to the rest of the parenchyma. Little mitotic activity occurred, and growth appeared to be by

expansion. These growths were classified as benign tumours. In the second type, type (b) tumours, a much more abnormal structure was present with areas of cells proliferating in confluent sheets and often with foci of

necrosis. There were areas of papilliform or adnoid formation of liver cells with wide and irregular vascular channels within the growth. The mitotic activity was often increased and multinucleate forms were seen. On purely

morphological criteria, the authors classified these tumours as hepatocarcinomas. They were very uncommon in the control group of mice. In a few cases these type (b) tumours had emboli of

V I 1 1-340

cells in the lungs. Dieldrin increased the incidence of both types of tumour but did not produce the fibrosis and bile duct proliferation sometimes observed in the mice given butter yellow. In order to determine if

tumour-formation was reversible, certain mice from the group fed 10 ppm of dieldrin were

transferred to a normal diet after 0, 2, 4, 8. 16, 32 or 64 weeks and sacrificed after 104

weeks. Incidence of tumours indicated that removal of dieldr in from the diet did not cause the tumours to regress or disappear. - (Dieldrin also produced liver enlargement and cytoplasmic changes which were reversible). In companion studies, sterilizing the diet or the bedding did not influence the incidence of tumours in mice fed dieldrin (Walker et al, 1970a)

In connection with the increase in liver tumours arising in susceptible species of mice. it is noted that the incidence of tumour-format ion in C3H mice was also affected by variations in the diet (Tannenbaum and Silverstone, 1949a, 1949b) and in microbiological status (Roe and Grant,


Rat In groups of 25 male and 25 female rats fed 0, 0.1, 1.0 or 10 ppm of dieldr in for up to two

years, the only pathological findings

attributable to dieldrin were the existence of liver parenchymal cell changes in one male and four females fed 10 ppm. In one female,

microscopic intrahepatic nodules were evident, and in addition nodules were observed after two years in two females fed 10 ppm and in one female in the control group (Walker et al, 1969a).

Weanling rats were fed, for a lifetime, diets supplemented with aldrin, 0, 20, 30 or 50 ppm; dieldrin, 20, 30 or 50 ppm, or endrin, 2. 6 or 12 ppm. Benign and malignant tumours were observed

in 23 tissues or organs - in 199 of 800

experimental rats and in 79 of 163 controls, examined histologically. The highest number of tumours in all groups occurred in mammary and lymphatic tissues. When compared to the

controls, male and female rats fed aldrin 20, 30 and 50 ppm and dieldrin, 20, 30 and 50 ppm,

showed a dose-related significant decrease in the


incidence of all tumours, particularly those of the mammary and lymphatic tissues. In male rats fed aldrin and dieldrin 50 ppm, the tumour

incidence was reduced from 46 (in 75 control rats) to 13 (in 45 rats fed aldrin 50 ppm) and to one benign tumour (a skin papilloma) in 44 rats fed dieldrin. In the females fed aldrin and dieldrin 50 ppm, the total number of tumours was 23 in 41 and 16 in 31, respectively, while the female control rats showed 104 tumours in 88 rats. In all the 963 rats examined, no primary malignant hepatic tumours were found, only two benign hepatic hemangiomas, one in male control and one in a female fed endrin 6 ppm (Diechmann et al, 1970) .897

(Comment) The increased risk of development of liver tumours in mice exposed to dieldrin as compared with untreated control mice was regarded as the main matter for concern. This increased risk was seen even at an exposure level of 0.1 ppm in the diet. It was generally agreed that the tumours were neoplasms and that those seen in treated mice were similar in nature to those seen in untreated mice except insofar as the tumours in dieldrin-treated mice tended to show the same cytoplasmic changes as other liver parenchymal cells in the same animals.

There was no evidence that tumours regressed or disappeared after cessation of exposure to dieldrin, although the cytoplasmic changes did regress under such circumstances. It is

difficult to establish unequivocal criteria for distinguishing between benign and malignant neoplasms in mouse-liver. The incidence of mice

showing tumour emboli or metastases in the lungs was not significantly increased at any levels of exposure to dieldrin up to 10 ppm for

It was agreed that the effect of

132 weeks

dieldrin in

increasing the risk of tumour development in mice could not at the present time be regarded as sufficient evidence to categorize dieldrin as a carcinogen. Other work showed that nonspecific factors such as calorie intake, hormonal status and microbiological status mav profoundly

influence the risk of liver tumour development in mice, some strains of which are iin any case peculiarly susceptible to such tumours. Also it


was noted that long-term studies in rats and prolonged studies in dogs and monkeys had

revealed no evidence of hepatocarcinogenicity. It seems likely that the effect of dieldrin in mice is probably a species-specific phenomena.

Further information is urgently needed concerning the aetiology and pathogenesis of liver tumours in mice. A comparison of the effect of dieldrin in mice of high and low liver tumour

susceptiblity, with parallel metabolic studies, might provide a basis for relating observations of tumorigenicity in mice to other species,

including man. There are at present no

experimental data referable to teratogenicity.

Continued surveillance of persons occupationally exposed to dieldrin for chronic toxic effects and manifestation of possible carcinogenicity is recommended. In this connection, the extensive surveillance of occupationally exposed workers, which is currently being made. was noted.

Information on the varying patterns of metabolism in the mouse, sheep, rat, monkey and man are as yet too incomplete to indicate whether the possible carcinogenic potential in the mouse is

likely to apply to man.(emphasis added)898

The International Agency for Research on Cancer reviewed

the carcinogenicity of dieldrin in 1974 and said:

Dieldrin was tested by the oral route only in mice and rats. The hepatocarcinogenicity of dieldrin in the mouse has been demonstrated and confirmed in several experiments and some of the

liver-cell tumours were found to metastasize. The dose response effect has been demonstrated in both sexes with an increased tumour incidence in females at the lowest dose tested, 0.1 ppm in the

diet (corresponding to about 0.015

mg/kg/bw/day). In mice, there is no evidence of carcinogenicity in organs other than liver.

The available data in rats has not provided evidence of carcinogenicity at levels of up to 50


ppm in the diet (corresponding to an intake of about 2.5 mg/kg/bw/day.)

The experiments in dogs and monkeys limited in duration and/or group sizes any conclusions to be made.8”

were too to allow

WHO in its "Data Sheet" on Dieldrin. issued in 1975. dealt

with the carcinogenicity of dieldrin in the following


For mice fed 10 ppm (1.5 mg/kg/day) of dieldrin for two years, there was a significant increase in tumours which were morphologically benign. Mice were fed dietary levels of 0.1, 1.0 or 10

ppm (0.015, 0.15 or 1.5 mg/kg/day) of dieldrin for periods up to 132 weeks. No tumours were evident before 37 weeks at any dose level. After this time, the incidence of tumours increased with the concentration of dieldrin that was fed.

Some tumours were classed as benign, the others as hepatocarcinomas; the latter kind were very rare in the controls. In a few cases emboli of tumour cells were found in the lungs.

In rats fed 20, 30 or 50 ppm (1.0. 1.5 or 2.5

mg/kg/day) of dieldrin there was a dose related increase in all tumours particularly in the mammary and lymphatic glands. There was no evidence of malignancy. In rats fed levels of dieldrin ranging from 0.5 to 150 ppm (0.025 to 7.5 mg/kg/day), there was an increase in the number of tumours when compared to a control group, but in this study, there did not appear to be a dose related increase".900

WHO again reviewed the toxicity of dieldrin at the Joint

Meeting on Pesticide Residues in 1977. Several new

dieldrin carcinogenicity studies were reported and

evaluated and the following comment was made:

V I 11-344

Results of various carcinogenicity studies in mice and other mammalian species indicate that there is a species -snecific effect of dieldrin and aldrin on the mouse liver resulting in an

increased frequency of liver tumors only in this animal species. In all species examined (mouse, rat, rabbit, rhesus monkey, chimpanzee)

12-hydroxydieldrin and 4, 5-aldrin-trans dihydrodiol were the major metabolites.

Regarding the ratios of these two metabolites, the mouse seems to metabolize the test compound predominantly by the opening of the epoxide ring. This metabolic pathway on the one hand and a high rate of metabolism compared with other

animal species examined on the other. could result in relatively high concentrations of 4, 5-aldrin-trans-dihydrodiol in the mouse liver.

Dieldrin, as well as other chlorinated

pesticides, is a powerful inducer of microsomal enzymes. Owing to their high rate of metabolism, mice are especially susceptible to such effects. Compared with other animal species, it__reacts

rather anomalously and the mouse, therefore,_may not be an appropriate model for humans in_th.i.s case. The results of prolonged enzyme induction are proliferation of the endoplasmic reticulum in

the cells, and hypertrophy and hyperplasia of the liver. These changes are fully reversible if treatment is ceased before tumours have developed on the chronically over-loaded and damaged cells. Similar effects are produced by


The results of the different experiments for the study of the possible mechanism of the

tumorigenic action of dieldrin on mouse liver at the sub-cellular level are inconsistent.

A number of in vitro and in vivo mutagenicity tests have been performed. In none of these studies did dieldrin reveal any mutagenic activity.

These new findings again support the view_that. dieldrin and aldrin are not carcinogens_on the basis of knowledge available to the Meeting. Therefore, no change of the existing ADI for

humans was considered necessary. (emphasis added)901


Cabral in "Carcinogenicity of Pesticides" (1980) reviewed

the carcinogenicity of dieldrin and said:

There have been at least 11 studies of

carcinogenicity in experimental animals.

The ability of dieldrin to produce liver-cell tumours in several strains of mice is evident. The data on rats have not provided conclusive evidence of carcinogenicity. Our own data in studies conducted in hamsters indicate that this species is resistant to the carcinogenic effects of doses of dieldrin higher than those producing a very marked incidence of liver-cell tumours in


The Commonwealth Department of Health concluded:

Evaluation of dieldrin for carcinogenicity proved to be controversial (Deichmann & Macdonald, 1977; Epstein. 1976; IARC, 1974; Kay. 1975; Ritper. 1979; van Raalte, 1977). In mice, dieldrin has been reproducably shown to induce liver tumours, some of which metastasised (IARC, 1974 and references cited therein). Studies with other mammals including rats, dogs and monkeys failed to show definite carcinogenicity (IARC, 1974 and references cited therein). Epidemiological studies of occupationally exposed workers are inconclusive (IARC, 1974) and there is no direct evidence that dieldrin produces cancer in man. Neveretheless. the mice studies led to the

banning of dieldrin manufacture in the USA and severe restriction of its usage there and

elsewhere (Kay, 1975). This decision has remained controversial but there is general agreement that the usage of dieldrin should have been restricted because of its environmental persistence (Anon., 1975).903


IARC again reviewed the carcinogenicity of dieldrin in

1982 and concluded that there was "limited" evidence for

the carcinogenicity of dieldrin to animals. In

particular, it said:

Dieldrin is carcinogenic to mice, producing benign and malignant liver neoplasms following its oral administration (IARC, 1974; NCI, 1978; Tennekes, H.A. et al 1981). No carcinogenic

effect was observed in feeding studies in rats (IARC, 1974; Stevenson et al, 1976; NCI, 1978). trout (Hendricks et al, 1979) or hamsters (Cabral et al, 1979). Feeding studies in dogs and

monkeys were considered to be inadequate (IARC, 1974).904

It is noted that the Commission contacted Shell Chemical

(Australia) Pty. Limited. the manufacturers of dieldrin,

inviting it to provide a submission to the Commission

concerning the toxicity of the compound. The Company

responded by supplying a reprint of the relevant series of

WHO Monographs 1967-1971, extracts from which have been

set out above.

Dr Philippe Shubik in his written statement to the

Commission said of dieldrin:

This chlorinated hydrocarbon, although more acutely toxic than DDT, presents similar problems from the standpoint of carcinogenesis. It enhances the incidence of hepatomas in mice but

is apparently inactive in rats and hamster.905


Dr Shubik in his oral evidence stated:

. . . Dieldrin, a similar sort of compound, more acutely toxic than DDT. However, in so far as carcinogenesis is concerned only I think it presents precisely the same sort of problems. It enhances the instance of hepatomas and again in this case appears to be inactive in rats and hamsters but one always has the same problem of negatives and I would again feel that it is again

like DDT unlikely to be carcinogenic in man, (emphasis added)y°6

Dr Bernard Stewart in his written statement to the

Commission referred to the IARC's review of dieldrin in

1982 and then said:

The situation regarding dieldrin is similar to that for chlordane. Presently available data are inadeguate to completely discount concern in response to the experimental findings.907

Dr Stewart in his oral evidence stated:

... Dieldrin is approximately similar in its activity, in its biological activity in

experimental animals. It produced benign and malignant liver neoplasms upon its oral

administration to mice No carcinogenic effect has been observed in feeding studies in rats or trout or hamsters, and feeding studies in dogs and monkeys that have been reported were

considered by the Agency to not contribute to the evaluation. The available evidence epitomises limited evidence, that is to say, there is only (sic) neoplasms in the mice, and the fact that

these chemicals have actually been tested in o ther species and produced negative results


indicates why these mice tumours must be regarded as having a distinct character. If you like, they provide an inadequate basis to predict whether the material will produce cancer in rats, which it does not. Nonetheless, in this

particular case, consistent with that, these data have been evaluated as inadequate, that is to say, they have been rated lower down the scale of

concern, not because the mouse tumours have been completely overlooked, but presumably because of the negative studies. So dieldrin must rank below chlordane in terms of. shall we say, our priority ranking of suspect carcinogens,

(emphasis added)908

Ian Munro in his oral evidence stated:

.... Dieldrin is an insecticide, one of the

chlorinated hydrocarbon insecticides; that is an oral LD50 in the range of 46 milligrams per kg, somewhat more toxic than 2,4,5-T. There have been studies done on the reproductive toxicity of dieldrin, and in animals administered high doses

ranging up to 20 parts per million in the diet in rats does not induce any birth defects at all, but did result in some reduced survival of the offspring of the young animals. At lower doses in the range of ..08 to .16 parts per million of dieldrin in the diet, no adverse effects were produced. The major concern about dieldrin, as I pointed out earlier on, is its ability to induce hepatic tumors in mice. The International Agency for Research on Cancer has concluded that doses

in the range of 1 to 10 parts per million in mice fed over the major portion of a life span will produce hepatic tumors, but this effect is not noted in rats or any other species that have been

tested, and dieldrin does not have any of the hallmarks of classic carcinogenicity. It is not active in short term tests for carcinogenicity, but it does produce marked changes in liver enzymes at those levels, namely 10 parts per million, and marked proliferation in liver cells, and as a result probably induces cancers through

some mechanism that is not related to that

normally considered to be of prime public health concern, and as I have mentioned previously. 1 1 1

V111 - 3 4 9

related to the classical activity for

carcinogenesis. In fact, the whole matter of hepatic tumour induction for mice is questionable in view of the fact that this is an isolated

observation confined only to mice. , and most public health authorities do not think that dieldrin in its use in agriculture represents a serious public health risk. This question of the role of hepatic tumors in mice. these liver

tumors in mice, has been discussed in detail in that report which I submitted to you from the Nutrition Foundation of the United States, and in accordance with the criteria outlined therein, dieldrin would not be considered to be a major concern from a public health point of view. I did not calculate a no effect level for this chemical, because I think one has to view this more or less as an all or none effect, and my

view is that I would have no concern over the prevailing uses_____of____dieldrin. (emphasis added)909

The Commission accepts the evidence of Drs Shubik, Stewart

and Munro and finds that the use of dieldrin in Vietnam

has not caused cancer amongst Australian personnel and is

not likely so to do. It is not carcinogenic used as a



In W A A 1s initial submission the bald statement appears at

page 58 "it is carcinagenic" (sic).

In its submission related specifically to toxicology the

following appears„ "Chlordane is in addition to other

toxic effects a carcinogen and neurotoxic", (at p 82.)


In its final submission under the heading. "Chlordane",

the following appears: "Dr Shearer, Exhibit 1615 Page 3.

refers to DNA damage to human cells in culture. Exhibit

1600 shows oncogenic in mice and probably rats. Exhibit

1652." (at p 33.)

Exhibit 1652 relates to aldrin and dieldrin and does not

mention chlordane.

Exhibit 1600 is a paper by Ahmed et al and relates to

induced DNA damage and its repair in cultured human cells

exposed to a number of pesticides. The conclusion that

Dr Shearer would draw from this paper is that agents which

produce DNA damage may be potential carcinogens. The

paper certainly does not positively support the bald

statement made by Dr Shearer.

The Commission conducted it own research on the issue of


WHO at the 1970 Joint Meeting on Pesticide Residues dealt

with the toxicity of chlordane and reported the following

study on its carcinogenicity, namely:


Rat: Groups, each of 20 male and 20 female rats were fed dietary levels of 0, 5, 15. 25 or 35 ppm of

cis-chlordane or 15, 25, 35 or 75 ppm of

trans-chlordane or 5, 15, 25, 35 or 50 ppm of a 1: 1 mixture of cis and trans-chlordane. In the group fed cis-chlordane. growth retardation became apparent in the group fed 35 ppm after four months in the males and after five months in the females; with trans-chlordane. the 75 ppm group of males only displayed growth retardation after eight months. With the mixture, growth retardation was evident in both sexes fed 50 ppm, beginning early in the males and later in the females. Growth retardation was not evident in any groups fed lower doses of either isomer. Food consumption bore a relationship to growth.

Increased mortality for both sexes became significant in the groups fed 35 ppm of

cis-chlordane, 75 ppm of trans-chlordane or 50 ppm of the cis-trans mixture.

Haematocrit was normal for all test groups. Autopsy revealed no gross pathological lesion, in particular, there was no evidence of tumours. Histological examination of all organs showed no

changes from feeding chlordane at any level, except in the case of the liver. Compression of sinusoids due to slight hepatic cell hypertrophy

in the centrolobular region and minimal bile duct proliferation were evident at 35 ppm of

cis-chlordane. The above were noted, but were minimal, at 25 ppm of the same isomer. Slight to moderate cytoplasmic homogeneity of the hepatic cells in the centrolobular region, minimal perinuclear vacuolation and minimal cell hypertrophy with constriction of sinusoids were noted at 75 ppm of the trans-chlordane. Slight cytoplasmic homogeneity of hepatic cells in the centrolobular region and occasional cytoplasmic margination were observed at 50 ppm of the

cis-trans mixture. The above alterations were minimal at 35 ppm of the same mixture. No liver abnormalities were evident after feeding lower

levels of the chlordanae isomers (Ingle,


(Summary) In the 78 weeks feeding study in rats .... no tumours were evident when dietary levels up to 3 5 ppm of the cis- or 75 ppm of the trans-


isomer were fed (Ingle. 1969). There is no

report that tumours have been looked for in any other species fed chlordane.

WHO Joint Meeting on Pesticide Residues in 1977 again

considered the toxicity of chlordane and reported on

special studies on its carcinogenicity. The Report said:

Mice The groups receiving high and low doses of chlordane consisted of 50 male and 50 female mice, the control group of 10 males and 10

females. In the final evaluation pooled control groups were formed by combining controls from other experiments. The doses which were used decreased during the study because of toxicity. For the males the time-weighted average doses were 30 ppm in the low and 56 ppm in the high dose group. For the females these were

respectively 30 and 64 ppm. The mice were administered chlordane for 80 weeks, then observed for 10 weeks. Hepatocellular carcinoma was induced at a highly significant rate of

incidence in mice given chlordane, particularly at the high dose (NCI, 1977).

Rats The groups receiving high and low doses of chlordane consisted of 50 male and 50 female rats, the control groups of 10 males and 10

females. The rats were administered chlordane for 80 weeks, then observed for 29 weeks. In the final evaluation pooled control groups were formed by combining controls from other experiments. The doses which were used were decreased during the study because of toxicity. For the males the time-weighted average doses were 204 ppm in the low and 407 ppm in the high

dose group. For the females these were

respectively 121 and 242 ppm. The body weights in the high dose group were lower in both males and females. The incidence of hepatocellular adenoma was extremely low. There was no

statistically significant increase in the number of neoplastic nodules. A dose-related trend in


follicular cell lesions of the thyroid was found„ which was significant for females. In the males a significant dose-related trend in malignant fibrous histiocytoma was found. However, the nature, incidence and severity of

the thyroid lesions and the histiocytomas were not sufficient to indicate a clearly

carcinogenic effect of chlordane in rats.

The sample of chlordane used in both

carcinogenicity studies contained 94.8% chlordane (71.7% cis-chlordane and 23.1% trans-chlordane) (NCI, 1977) ....

(Summary) In long-term studies, chlordane caused hepatocellular carcinoma in mice at a dose of 60 ppm in the diet, but not in rats at

doses as high as 400 ppm for males and 240 ppm for females. Both male and female rats were found to have an increase in proliferative lesions of follicular cells of the thyroid. Male rats only showed an increase of malignant

fibrous histiocytomas, but these observations were discounted because of their low incidence and the high variability of the number of

tumours in earlier experiments at the same research center.9- 1 -2

In 1978 WHO published a Data Sheet on chlordane which said

that in a number of long-term feeding experiments there

had been no evidence of increased tumorigenicity.

1ARC reviewed the carcinogenicity of chlordane in 1979 and

concluded that there was "sufficient evidence" that

chlordane is carcinogenic in mice. More particularly, the

Agency said:

Chlordane (analytical grade) was tested in one experiment in mice (NCI, 1977) and one in rats (NCI, 1977) by oral administration. It produced


hepatocellular carcinomas in mice of both sexes; in rats, the results were inconclusive. A

re-evaluation of unpublished studies involving the oral administration of technical grade chlordane to mice of another strain confirmed the hepatocarcinogenicity of chlordane for mice of both sexes (Epstein, 1976: Exhibit 1650).9- * - 4

The Agency again reviewed the carcinogenicity of chlordane

in 1982 and concluded that there was "limited" evidence

for carcinogenicity of chlordane to animals. It said:

Chlordane and heptachlor (which contains about 20% chlordane) were carcinogenic in mice, producing liver neoplasms following their oral administration (IARC, 1979; Becker & Sell,

1979). Data concerning rats were inconclusive (IARC. 1979).915

The International Program on Chemical Safety has recently

published an extensive review of the toxicology of

chlordane. In connection with the carcinogenicity of the

compound the report said:

Epstein (1976:)(Exhibit 1650) reported a previously unpublished study by the International Research and Development Corporation, carried out in 1973, in which groups of 100 male and 100

female Charles River CD-I mice, 6 weeks of age, were fed technical-grade chlordane (purity not given) 5, 25 and 50 mg/kg diet, for 18 months. Excluding 10 animals sacrificed from each group

for interim study at 6 months, mortality rates at 18 months ranged from 27 - 49%, except in males and females receiving the 50 mg/kg diet, in which the mortality rates were 86 and 75%

respectively. A relatively large number of the deceased animals were lost due to autolysis. A dose-related increased incidence of liver


hyperplastic nodules was reported in the 25 and 50 mg/kg diet test groups and a dose-related increased incidence of liver cell hypertrophy was found in all test groups. A significant

incidence of hepatocellular carcinomas compared with controls was also reported. In the males receiving chlordane at 0, 5, 25 or 50 mg/kg diet, hepatocellular carcinomas were found in 3/33, 5/55, 41/52 and 32/39 animals, respectively; in females, the respective incidences were 0/45, 0/61, 32/50 and 26/37.

Groups of 50 male and 50 female B6C3F1 hybrid mice, 5 weeks of age, were fed analytical-grade chlordane, consisting of 94.8% chlordane (71.7% cis-chlordane and 23.1% trans-chlordane), 0.3% heptachlor, 0.6% nonachlor, 1.1%

hexachlorocyc1opentadiene, 0.25% chlordane isomers, and other chlorinated compounds for 80 weeks (NCI, 1977). Males received initial levels of 20 or 40 mg/kg diet, and females 40 and 80

mg/kg diet; time-weighted average dietary concentrations were 30 and 56 mg/kg for males and 30 and 64 mg/kg diet for females. There were 20 male and 10 female matched controls and 100 male and 80 female pooled controls. Survival in all groups was relatively high, i. e . , over 60% in

treated males, over 80% in treated females, and over 90% in male and female controls. A

dose-related increase in the incidence of

hepatocellular carcinomas was found in males and females. The incidences were 43/49 and 34/49 in high-dose males and females, respectively, and 16/48 and 3/47 in low-dose males and females, respectively, compared with 2/18 and 0/19 in male and female matched controls, respectively.

Groups of 50 male and 50 female, 5-week-old

Osborne-Mendel rats were given analytical-grade chlordane in the diet for 80 weeks, at initial levels of 400 and 800 mg/kg for males and 200 and

400 mg/kg for females (NCI. 1977). The levels had to be reduced during the study because of adverse toxic effects; the time-weighted average dietary concentrations were 407 and 203 mg/kg for males and 241 and 121 mg/kg for females. There were 10 male and 10 female matched controls and

60 male and 60 female pooled controls. Survivors were killed at 80 weeks, at which time

approximately 50% of treated and control males


and 60% of treated females and 90% of control females were still alive. In all treated animals combined, there was an excess incidence of follicular-cell thyroid neoplasms (10/75 in

treated females and 7/65 in treated males versus 0/10, 3/58, 0/6 and 4/51 in matched and pooled female and male controls); there was an excess of malignant fibrous histiocytomas in the treated male groups (8/88 versus 0/8 and 2/58 in matched and pooled male controls).

A committee of the National Academy of Sciences (NAS) in the USA was asked to review all

available carcinogenicity data on chlordane as part of the cancellation hearings. Chlordane was not found to be carcinogenic in rats and the only target organ site for carcinogenic response in certain strains of mice was the liver. The committee concluded that "there are no adeguate data to show that these compounds are

carcinogenic in humans. but because of their carcinogenicity in certain mouse strains and the extensive similarity of the carcinogenic action of chemicals in animals and in humans. the

committee concluded that chlordane. heptachlor and/or their metabolites may be carcinogenic in humans.

Although the magnitude or risk is greater than if no carcinogenicity had been found in certain mouse strains, in the opinion of the committee the magnitude of risk cannot be reliably

estimated because of the uncertainties in the available data and in the extrapolation of carcinogenicity data from laboratory animals to humans" (US NAS, 1977).

IARC (1979), in its evaluation of the

carcinogenic risk of chlordane, concluded: "There is sufficient evidence that chlordane is carcinogenic in mice." In 1982, another IARC Working Group reviewed existing data on chlordane

and concluded that there was limited evidence of the carcinogenicity of chlordane for experimental animals (IARC. 1982). The group of Williams Telang et al (1982) suggested that chlordane had

the properties of many promoting agents.916

Dr Philippe Shubik in his written statement said:


This is another representative of the group of compounds giving rise to hepatomas in mice; data in rats is inconclusive. No epidemiological evidence is available. This compound is in my

view unlikely to be carcinogenic in humans.917

Dr Bernard Stewart said:

Chlordane must remain on the "suspect" list in respect of a putative carcinogenic hazard considered to have existed in Vietnam.918

Dr Stewart in his oral evidence said:

. . . Chlordane has been evaluated by the agency. They looked at chlordane and heptachlor together because these chemicals are closely related and chlordane is contaminated by heptachlor. They are carcinogenic in mice producing - that is, chlordane and heptachlor may be considered carcinogenic in mice producing liver cancer. Data from rats were inconclusive. These

findings consistent with the definition I described earlier were considered to be limited evidence of carcinogenic activity. Chlordane is negative in all the short term tests to which it has ever been subjected. Taking the

epidemiology, the human evidence in respect, I think, of people who were manufacturing chlordane do not provide an inference of a carcinogenic effect in humans. The WHO assessing the same

data concluded that although there was no direct evidence that chlordane is a human carcinogen the suspicion, the implication arising from its effects in mice cannot be completely ignored as again I said as a general principle. It is

suggested that further research should be done in an attempt to determine the way in which that mouse data might best be extrapolated to humans.

Nonetheless, WHO considered that as long as occupational hygiene procedures were maintained to ensure exposures were kept to a minimum, there

VI11-3 58

was little reason to believe that workers, who they considered to be the most highly exposed pesticide workers, which they considered would be the most highly exposed group, would be at risk

to cancer from their handling or contact with chlordane. Chlordane must remain on the suspect list because the mice tumours cannot be entirely discounted but again, it cannot be rated as highly as TCDD because when we compare

experimental evidence with experimental evidence, chlordane must be placed - it has not exhibited the spectrum of biological activity that TCDD has, or indeed that DDT has. Therefore, it is above malathion, if you like, but below TCDD.919

Dr Brusick in his paper "An Analysis of the Mutagenic

Activity of Chlordane and Heptachlor and its Implications

for Tumoriqenicity", said:

The conclusions derived from this review of the mutagenic and tumorigenicity data for chlordane and heptachlor are that neither of these agents produce tumors in mice via a genotoxic

mechanism. This conclusion, in conjunction with the indirect evidence supporting a promotional model for liver tumor enhancement places

chlordane and heptachlor among a group of related chemicals for which there is no precedent for a concern for human cancer, and a threshold model would be most appropriate to estimate a cancer risk coefficient from the animal data.920

Dr Brusick in his oral evidence said:

... Chlordane, heptachlor are closely related. They have been studied extensively for their genotoxic or mutagenic effects and to my knowledge, no data that can be supported by

careful evaluation shows positive mutagenic activity for these two agents.


So in your opinion?--- In my opinion, as far as genetic heritable genetic activity, they have posed no risk and again, as far as

initiation of malignancy, posed no risk in my opinion. however. cancer studies in animals have found both these to be able to produce liver tumours in rodents. But that appears to be through a promoting phenomena and not through a genetic initiation even,

(emphasis added)921


In its submission to the Commission, Velsicol

Australia Limited referred to the conclusion of the

International Agency for Research on Cancer that there was

"limited evidence" of the carcinogenicity of chlordane in

animals and then dealt, at some length, with the views

expressed by various witnesses before the Commission on

the carcinogenicity of this compound. The submission,

after guoting extracts of the evidence of Dr Stewart, Dr

Brusick and Dr Shubik before the Commission and from their

publications which have already been set out above, said:

C . Cancer Studies In fact, as discussed above, there is a

substantial body of human epidemiological evidence which confirms the view that chlordane is not carcinogenic in man. No significant

increase in deaths due to tumours of the liver or of any other site has been found among the groups most highly exposed to chlordane, the studies as updated now extending over a long

latency period. Further, during the period of about forty years in which chlordane has been extensively used in the US. the rate of liver cancer deaths in the US has in fact declined substantially (see Appendix 1). Even among other rodents the findings in relation to mice have not


been replicated: in the same study (conducted by the US National Cancer Institute) as that in which tumours in the livers of mice were

reported, there was no significant incidence of tumours in the livers of rats to which chlordane was also administered (NCI, 1977).

Dr F. Dost at p 2771 of the transcript expressed the view that chlordane was a weak carcinogen. He stated also that it was "almost out of use in the US" although he conceded that a little bit was still used for treating termites

underground. It is submitted that both the factual statements as to the quantities of chlordane in current use, and the view that it is a weak carcinogen, are erroneous and should be

rejected. There is no evidence to support such a view apart from the mice studies. The amount of chlordane sold in the United States for the

period 1978-1983 totals approximately 45 million pounds. It is to be borne in mind that Dr Dost did not come to Australia prepared to give

evidence about chlordane: the chemicals about which he had prepared to give evidence at the instigation of a party appearing before the Royal Commission, and to which his statement (Exhibit

1267) was directed, were the constituents of Agent Orange. His evidence about other chemical substances including chlordane was given on short notice and with no adequate opportunity for

preparation. His position is to be contrasted with those of Dr Stewart, Dr Shubik, Dr Aldridge and Dr Brusick who were specifically called by the Royal Commission to give evidence upon a

number of chemicals including chlordane, and who were prepared accordingly.923

(In summary) there is strong evidence against the proposition that chlordane is carcinogenic, and no evidence for the view that it is carcinogenic apart from the production of tumours in mice.924

The Commission adopts the views of Dr Shubik, Dr Brusick

and Dr Stewart and finds that chlordane is not

carcinogenic in humans and that no Australian Vietnam

veteran has suffered cancer or will suffer cancer as a

result of exposure to this compound in Vietnam.



Lindane is the last chemical mentioned in WAA's original

submission, (at p 60.) Its effects are said to be

" intoxication, irritation of nose, eyes, throat, nausea

and severe headache." No allegation of carcinogenicity is

made in that document.

In its submision related specifically to toxicology, W A A

at page 84 quotes from a review of the toxicity of this

compound found in the Archives of Dematology, Volume 113,

Page 353. The last paragraph on that page reads,

"Lindane is highly carcinogenic in rats and mice.

'Carcinogenicity of Lindane1" Melvin D. Reuber,

Environmental Research 19 Page 460-481."

In its final submission the following appears.

Dr Shearer, Exhibit 1615, reports that lindane is neurotoxic and causes persisting neurological and psychological dysfunction. It is oncogenic in mice.

A metabolite 2,4,6-trichlorophenol is oncogenic in rats and mice. It blocks proliferation of human lymphocytes in culture indicating possible immunotoxicity. (at p 32.)

Dr Shearer bases her suggestion of immunotoxicity on a

review by Solomon et al called Gamma Benzene Hexachloride


Toxicity. The authors postulate that effects on the

haemotological system may have occurred by an

immunological mechanism. The paper makes no statement

about carcinogenicity.

As far as oncogenesis in mice is concerned Dr Shearer

relies upon the US Department of Health and Human Services

Third Annual Report on Carcinogens, issued in December 926 1982. That paper relies entirely on I ARC data with

which the Commission will shortly deal. In a nutshell

however, it says that evidence of carcinogenicity of

lindane in humans is inadequate.

The other document relied upon by Dr Shearer is an EPA

9 27

determination in relation to lindane. That agency

expressed the belief that lindane should be considered to

have the potential for inducing carcinogenic effects in

humans. This led the agency to reduce the low benefit

uses of lindane in products. The EPA is of course

influenced by the Delaney Amendment.

The Commission conducted its own investigation of lindane.

WHO Joint Meeting on Pesticide Residues, considered the

toxicity of lindane at its 1966 meeting, and it reported

as follows:



With the aim of detecting the carcinogenic action, 3 groups of 20 young rats (10 males and 10

females) were fed diets containing 25, 50 and 100 ppm of lindane throughout their lifespan. At the two highest concentrations, slight hypertrophy of the liver was observed, while with 100 ppm there was a slight tendency to fatty degeneration of

this organ. At the lowest concentration, 25 ppm, there was no toxic effect, in particular, no histological lesions were detected in the liver or kidney. None of the concentrations used gave any significant increase in the percentage of tumours

as compared with the control group (Truhaut, 1954).928

WHO again considered the toxicity of lindane at the Joint

Meeting on Pesticide Residues held in 1973, and the

following study was reported, namely:

Groups of mice (30 male and 30 female per group) were fed lindane in the diet at levels of 0 and 400 ppm. (Be ta-BHC was also fed at 200 ppm to

another group) for 2 years. During the first 3 months, 10% of the males and 20% of the females died (12% of the males and 25% of the females fed Beta-BHC died). Liver enlargement with nodular

surfaces were described with both isomers as hyperplastic nodules and in some cases

neoplasms. There were no incidence of neoplasms in non-hepatic tissues (Thorpe and Walker, 1973).929

(In summary) it was noted that with mice at high dietary intake, liver enlargement and nodular formation was evident. Since the no-effect level for this effect on mouse liver has not been established and since neither of the 2 long-term studies in rats is fully adequate, the meeting considered that a further long-term

carcinogenicity study was now required. 930

A further carcinogenicity study was evaluated by the Joint

Meeting on Pesticide Residues held in 1977. In that study:


Special studies on carcinogenicity Three groups of 50 male and 50 female Chbi: NMRI (SPF) mice were fed 12.5, 25 or 50 ppm lindane (equivalent to 2.1, 4.1 or 8.2 mg/kg/day in males

and 2.0, 3.9 and 7.8 mg/kg/day in females) from 34 days of age for 80 weeks. One hundred mice of each sex served as convenient controls. A complete autopsy, light microscopic examination of brain, heart, lungs, liver, spleen, kidneys, suprarenal glands, gonads, bladder and all detected tumours was performed on all survivors. Electron microscopy of liver was performed on

10-15 samples from each of 4 male and 4 female mice from each group. Animals dying during the study were examined grossly post mortem.

Survival data were not available for individual groups. Combined data indicated an 84.2% overall survival. In animals dying during the study, 22.8% of the dead mice reported died owing to the

presence of tumours (mainly lymphocyte

leukaemia). In 19% (80 mice) of the survivors tumours were observed. In 6 of these animals two primary tumours were present in each animal. A breakdown of a total of 108 tumours recorded during the experiment by group and tumour type does not indicate either a site-specific or compound-related increase of tumour incidence in any group. The commonest tumours in survivors were lymphocyte leukaemia and primary lung

tumours. Electron microscopy did not reveal differences between treated and control groups (Weisse and Herbst, 1977).

Additional unpublished data on the above study indicate that mortality ranged from 14 to 19% of each group (Kollmer et al. 1975).

Two groups of 35 day old 50 male and 50 female B6C3Fi mice were fed 80 to 160 ppm lindane in a diet containing 2% corn oil for 80 weeks followed by a 10 week period on a diet containing 2% corn

oil without lindane. Controls comprised 10 male and 10 female mice fed 2% corn oil for 90 weeks. Non-concurrent controls from other experiments (pooled controls) comprised 50 males and 50

females. Body weights of females receiving 80 ppm lindane were constantly reduced. During the second year of the study, treated females were excitable and 6 treated males aggressive. During

the last 6 weeks of the study; all treated


animals were in poor physical condition. At least 88% of male mice and 80% of female mice survived to termination of the study. Random neoplastic, proliferative, degenerative and inflammatory lesions occurred in all groups. Neoplastic liver lesions occurred in 20% of

control males, 29% of low dose males and 20% of high dose males. In females, 0, 4% and 7%

neoplastic liver lesions occurred at 0, 80 and 160 ppm. Statistical analysis of the incidence of hepatocellular carcinoma showed that it was not significant for any group, on the basis of the contemporary control comparisons. Among

pooled controls, a significant increase was noted in low dose males as compared to pooled controls (NCI, 1977). Two groups of 50 male Osborne

Mendel rats (35 days old) were fed 320 or 640 ppm lindane in the diet for 38 weeks followed by 160 or 320 ppm for a further 42 weeks. A 30 week

withdrawal period preceded killing. Two groups of female Osborne Mendel rats (35 days old) were fed 320 or 640 ppm lindane in the diet for 2

weeks, followed by 49 weeks on 160 or 320 ppm, and then 29 weeks at 80 or 160 ppm. A 30 week

withdrawal period preceded killing. Controls comprised 10 males and 10 females maintained for 109 weeks. Body weight varied but was generally comparable between groups. Clinical signs were noted with increasing frequency in all treated groups as the study progressed. Mortality was 52%, 50% and 40% in high, low and control male rats and less than 40% in high and low dose

females but 60% in female controls (NCI, 1977).

(Comments) In view of the currently very low intake of lindane from a wide variety of sources, the relative rapidity with which it is degraded in mammalian organisms and in the environment to much less acutely toxic compounds and the relatively high levels required to produce adverse effects (especially carcinogenesis), in humans and laboratory animals there appear to be no further toxicological objections to the continued maintenance of the previously

established ADI for humans.

Neither of the new mouse studies reported here nor the new rat study both undertaken with lindane of much greater purity than previously tested provide any evidence of tumour or cancer induction following exposure to lindane. Since


the information required by the 1974 meeting has now been satisfied, the present Meeting decided that the previously allocated ADI for humans is no longer temporary. (emphasis added)931

IARC reviewed the carcinogenicity of lindane in 1979 and

concluded that there was "sufficient evidence" that

lindane was carcinogenic in mice. In particular, the

Agency said:

Lindane was tested in 6 experiments in mice by oral administration: it produced benign and malignant liver tumours in animals of both sexes in 2 experiments, one of which involved only

small groups of animals. The results of a third experiment suggested hepatocarcinogenicity but were inadequately reported. The results of a fourth experiment also suggested

hepatocarcinogenicity but were considered inadequate because of the low number of control animals used. The other experiments were

considered for an evaluation of carcinogenicity. Lindane was also tested in three feeding studies in rats: two were considered inadequate; in the other a slight excess of thyroid tumours was

observed in females. Lindane was tested

inadequately in mice by skin application and by subcutaneous and intraperitoneal


IARC again reviewed the carcinogenicity of lindane in 1982

and concluded that there was "limited" evidence of its

carcinogenicity to animals. In particular, the Agency


Technical gamma and beta hexachlorocyclohexane and lindane (the y isomer) are carcinogenic to


mice when administered orally, producing liver tumours (IARC, 1979; Kashyap et al. 1979); the technical grade also produced lymphoreticular neoplasms (Kashyap et al, 1979). Studies in rats

(IARC, 1979; Angsubhukorn et al. 1977; Hiasa et al, 1978) and dogs (Rivett et al, 1978) were considered to be inadequate. Technical

hexachlorocyclohexane and lindane were tested inadequately by skin application (IARC, 1979; Kashyap et al, 1979). Hexachlorocyclohexane increased the incidence of liver neoplasms in rats previously exposed to N-nitrosodiethylamine

(SchuIte-Hermann & Parzefall, 1981).933

Dr Philippe Shubik dealt with the carcinogenicity of

lindane. He said:

This chlorinated hydrocarbon again gives rise to hepatomas in mice; evidence in other species is said to be inconclusive. It is in my opinion unlikely to be carcinogenic in humans.93^

Dr Stewart in his oral evidence stated:

.... Lindane does produce tumours in mice of both sexes that were used for test purposes. These tumours were in the liver of the mice. Lindane was also tested by feeding to rats and mice but

these experiments had limitations. These findings were evaluated as limited evidence of carcinogenicity, again consistent with the protocols evaluation which I have already outlined. Lindane has given negative or

equivocal results in all the short term tests to which it has been subjected. and the available epidemiology does not provide any basis for inferring carcinogenic activity in

humans.(emphasis added)93b

The Commission adopts the views of Dr Philippe Shubik and

Dr Bernard Stewart and finds that lindane is not

V I 11-368

carcinogenic in humans. It follows that the Commission

finds that no Vietnam veteran from Australia has suffered

or will suffer cancer as a result of exposure to lindane

in Vietnam.


Diazinon is included in WAA's initial submission at page

51 amongst a list of insecticides. In the submission

related specifically to toxicology no allegation is made.

In its final submission it is included under the

organophosphates but no allegation that it is carcinogenic

is made.

Nonetheless the Commission conducted its own


The United States National Cancer Institute conducted a

bio-assay of diazinon for possible carcinogenicity in

1979. It found that diazinon was not carcinogenic in rats

or mice. In particular, the Institute said:

A two year carcinogenicity study in mice showed no significant incidences in female mice; liver tumours occurred in male mice but could not be related to diazinon.


No significant tumours occurred in female rats after a two year carcinogenicity study. Tumours occurred in male rats but not in dose-related

fashion. The evaluation was that there was 'no convincing evidence for carcinogenicity of diazinon'.936

Dr Philippe Shubik said:

This compound has been tested in a US

National Cancer Institute bioassay and pronounced negative for carcinogenicity in rats and mice.937

Dr Shubik also dealt with the carcinogenicity of diazinon

in his oral evidence. He said there was:

Ho evidence whatsoever to suspect the compound,

and that it was

unlikely to be carcinogenic (in humans.)938

Dr Bernard Stewart drew the same conclusion as Dr



The Commission therefore finds, adopting the evidence of

Dr Shubik and Dr Stewart, that diazinon is not

carcinogenic and that no Australian Vietnam veteran has

suffered or will suffer cancer as a result of exposure to

that compound in Vietnam.



In its initial submission W A A simply included pyrethrin

in its list of insecticides. (at p 51.)

In its submission related specifically to toxicology W A A

at page 88 refers to pyretherun (sic) but no allegation of

carcinogenicity is made (at p 88). In its final

submission W A A makes no reference to pyrethrins.

The Commission relies upon the WHO/FAO Data Sheets on

Pesticides, No. 11 Pyrethrins which reveal the following.

Chronic feeding studies with animals have shown only slight liver damage at high dietary levels (1000 parts per million). (There is no suggestion that there are mutagenic, carcinogenic or teratogenic effects in man whilst no reports of animal experiments to determine such effects are


In view of the lack of any allegation that pyrethr ins are

carcinogenic and in view of the lack of any suggestion of

carcinogenicity at an international level the Commission

has no hesitation in finding that pyrethrins are not

carcinogenic and that no Australian Vietnam veteran has

suffered or will suffer from cancer as a result of

exposure to this compound in Vietnam.

V I 11-371

Diethyl-m-ToIuamide (DEET)

None of WAA's submission refer to this compound at all.

Despite an extensive literature search the Commission has

found no suggestion that this product has any carcinogenic

potential. The Commission therefore finds that it is not

carcinogenic and that no Australian Vietnam veteran

suffered cancer or will suffer cancer as a result of his

exposure to it in Vietnam.

Dimethyl-Phthalate (DMP)

This insecticide is not mentioned in any of the W A A


Autian J . (1973) has said:

No animal data have yet demonstrated that any of the phthalate esters act as carcinogenic agents. Likewise, their role as possible co-carcinogens has not been established.941

Krauskopf L.G. has said:

Extensive testing has been reported concerning the effects of ingestion of phthalates. The literature contains studies including phthalates ranging from dimethyl to dietary ditridecyl. Acute toxicity tests define rather high LD50 values indicating a very low order of toxicity as compared to many common chemical substances.94^


Nearly all of the investigators were willing to conclude that the phthalates constitute a chemical family of very low order of toxicity, as measured by ingestion methods.....

The extensive feeding studies that have been reported appear to verify that the phthalates have a very low order of toxicity when

administered by oral ingestion.(emphasis added)943

No carcinogenic characteristics were found by any of the investigators. 944

Daniel W.J. has said:

Although the carcinogenic potential of the individual esters does not appear to have been investigated, chronic feeding studies provide no indication to suggest that they are


The Commission agrees with these eminent authors that DMP

is not carcinogenic and that no Australian Vietnam veteran

has suffered or will suffer cancer as a result of exposure

to this compound in Vietnam.

Di-n-Butyl Phthalate (DBP)

This compound is not mentioned in any of the W A A


Autian J . has said:

V I 11-373

No animal data have yet demonstrated that any of the phthalate esters act as carcinogenic agents. Likewise, their role as possible co-carcinogens has not been established.94^

Krauskopf L.G. said:

Extensive testing has been reported concerning the effects of ingestion of phthalates. The literature contains studies including phthalates ranging from dimethyl to dietary ditridecyl. Acute toxicity tests define rather high LD50 values indicating a very low order of toxicity as compared to many common chemical substances.94'

Nearly all of the investigators were willing to conclude that the phthalates constitute a chemical family of very low order of toxicity, as

measured by ingestion methods.....

The extensive feeding studies that have been reported appear to verify that the phthalates have a very low order of toxicity when

administered by oral ingestion.(emphasis added)948

No carcinogenic characteristics were found by any of the investigators.(emphasis added)949

Daniel W.J. said:

Although the carcinogenic potential of the individual esters does not appear to have been investigated, chronic feeding studies provide no indication to suggest that they are


The Commission therefore finds that this compound is not

carcinogenic and has not or will not cause cancer amongst

Vietnam veterans by reason of exposure to it in Vietnam.


Η.3. Drugs


9 51

As stated earlier in this Report Australian soldiers

in Vietnam were exposed to a most elaborate anti-malarial

prophylactic campaign. After 1968 that anti-malarial

campaign included the use of dapsone.

The use of dapsone involved a very substantial drop in the

number of cases of malaria.

Brigadier Rodgers, Senior Australian Medical Officer in

Vietnam in 1966-1967 and again 1969-1972 - presently the

Director of Medical Services for the Australian Army,


...Just before the application of dapsone there were up to 400 cases per thousand per year.

And after the dapsone regime was introduced?--- Virtually nil. It fell directly down to below ten or even below five.952

However, after dapsone was introduced it was found that a

number of Americans had died after using it. It was

believed to have been contaminated. Amongst the

V I11-375

Australian contingent some 11 soldiers developed a

disease, a serious disease, called agranulocytosis.

Because of this, dapsone was stopped for regular use but

was thereafter used only during a malaria epidemic. There

were but few epidemics and as soon as the epidemic

disappeared dapsone was discontinued.

As Brigadier Rodgers put it, "It was only when the risk of

dying of malaria became higher than the risk of dying of


dapsone (that it was used.)".

It is accordingly impossible to be precise about the

number of Australian soldiers who were exposed to

dapsone. Certainly none who were in Vietnam before 1968

would have been extensively exposed.

In its Initial submission W A A mentions dapsone under the

heading. Drugs. It does not there make any allegation

that dapsone is carcinogenic.

In its submission related specifically to toxicology W A A

quotes Martingale (sic) and adds the sentence, "Dapsone is

also a carcinogen and effects (sic) the blood", (at p 83.)

V I 11-376

In its final submission dapsone is not mentioned.

The Commission made its own investigations.

IARC reviewed the carcinogenicity of dapsone in 1980 and

concluded that there was "limited evidence" for the

carcinogenicity of this compound in experimental animals.

In particular„ the Agency said:

(al Oral administration

Mouse: Two groups of 35 male and 35 female

B6C3F1 mice, 5 weeks of age, were fed a diet containing 500 or 1000 mg/kg diet dapsone (USP grade) on 5 days/week for 78 weeks and observed up to 106-108 weeks. A group of 14 male and 14

female matched controls were untreated.

Seventy-three percent of the high-dose males and 67% of the low-dose males survived to the end of the study. Median survival rates were 92 weeks in treated females and 69 and 102 weeks in

untreated male and female controls. There was no difference in the incidence or types of tumours between treated and control animals (National Cancer Institute, 1977).

Forty pregnant C57B1/6 mice were given 100 mg/kg bw dapsone (purity not specified) by intragastric intubation on days 17 and 18 of gestation and when lactating, 5 times weekly from the 3rd day

after the birth of their offspring. Of their offspring, 50 males and 37 females were

subsequently given the same dose of dapsone 5 times weekly over 104 weeks starting 2 weeks after separation from their mothers; the total dose received after weaning was 52g/kg bw

(1.2-1.4 g/mouse) . The animals were observed for lifetime; no difference in survival was observed between treated and control mice. No significant difference in the incidence or site of tumours was seen between the 50 males and 37 females of

V I 11-377

the experimental group that survived over 24 weeks and the 47 male and 4 7 female untreated controls that survived for similar times (28/50 versus 20/47 in males, and 23/37 versus 24/47 in females) (Griciute & Tomatis. 1980).

Rat: Twenty 50-55-day-old Sprague-Dawley female rats were given single doses of 100 mg of dapsone (purity not specified) in sesame oil by gastric intubation. At the end of the 6-month

observation period, no tumours were reported in the 19 surviving rats. Another group of 20

Sprague-Dawley females, 40-days old at the beginning of the experiment, were given 30 mg/rat of dapsone in sesame oil by gastric intubation 10 times in a one-month period and then observed up to 9 months. No tumours were reported in the 19 surviving rats (Griswold et al, 1966, 1968).

(Attention is drawn to the short observation period due to the special design of the


A group of 25 male Wistar rats were fed a diet containing increasing concentrations of dapsone (purity not stated) . The experiment was started on 25-day-old rats with a concentration of 0.025%; the dose increased every 10th or 20th day, and from the 90th day after onset of the experiment until the 25th month the rats received 0.3% of dapsone in their food. During 17-25

months, 9 tumours were detected in 8/12 rats: 2 spleen tumours (fibrosarcoma and fibroangioma), 3 thyroid follicular adenocarcinomas, 1

fibrosarcoma of the retroperitoneum and 1 of the mesentery, 1 reticulosarcoma of the intestine and 1 liver cavernous angioma. One subcutaneous fibroma was observed in 13 control males observed up to 25 months (Bergel, 1973).

Two groups of 35 male and 35 female Fischer 344 rats (34 days of age) were fed a diet containing 600 or 1200 mg/kg diet dapsone (USP grade) on 5 days/week and control diets on 2 days/week for 78 weeks and observed for an additional 26-27 weeks. Mesenchymal tumours occurred in 13/35 males fed the lower concentration of dapsone, and

in 22/33 males fed the higher concentration. In the low-dose group, the tumours were: 6 splenic fibromas, 4 sarcomas (not otherwise specified), 1


fibrosarcoma and 1 fibroma of the peritoneum, 1 fibroma of the pancreas and 1 of the abdominal cavity. In the high-dose group, there were 3 sarcomas (not otherwise specified), 3

fibrosarcomas and 10 fibromas of the spleen and 3 sarcomas (not otherwise specified) and 3

fibrosarcomas of the peritoneum. None of these types of tumours occurred in male controls. The incidence of spleen tumours in males was

statistically significant in both experimental groups when compared with 43 pooled controls (low dose, P = 0.006; high dose, P less than 0.001) and in the high-dose group when compared with 14 matched controls (P = 0.002). The incidence of malignant mesenchymal tumours (sarcomas and

fibrosarcomas) in the peritoneum was also significant in males of the 2 treated groups: P = 0.014 and P = 0.005 respectively, when compared with 44 pooled controls. The authors detected fibrosis and osseous metaplasia in the spleen and peritoneum of some tumour-free males. No tumours or fibrosis of the spleen or peritoneum were observed in females (National Cancer Institute, 1977).

Twenty pregnant BD IV rats were given 100 mg/kg bw dapsone (purity not specified) by intragastric intubation on days 18 and 19 of gestation and 5 times weekly when lactating from the third day after the birth of their offspring. Of their offspring, 76 males and 72 females were

subsequently given dapsone 5 times weekly for 104 weeks, starting 2 weeks after separation from their mothers; the total dose received after weaning was 52 g/kg bw (10-16 g/rat). The

animals were observed for lifetime. Malignant fibrosarcomas and angiosarcomas of the spleen occurred in 4/44 males (P = 0.046) and 1/63

females that survived over 34 weeks, compared with 0/49 male and 0/47 female control rats that received olive-oil by gastric intubation and which survived for similar times. Partial or

complete fibro-sation of the spleen, in some cases with calcification and bone formation, was observed in 47% of males and 1.2% of females. Spleen sarcomas were topographically related to the lesions. Malignant C (parafollicular)-cell

tumours of the thyroid were observed in 8/44 males and 13/63 females of the treated group


compared with 2/49 males and 3/47 females of the control group (P = 0.014). Areas of C-cell

proliferation in the thyroid were observed equally in all groups of rats, including controls (Griciute & Tomatis, 1980).

(b) Intraperitonal administration Mouse: Three groups of 10 male and 10 female A/He mice, 6-8 weeks old, were given 12 i.p. injections (3 times weekly during 4 weeks) of dapsone (purity not specified) in

steriod-suspending vehicle. Total doses were 0. 525, 1.312 or 2.625 g/kg bw. All survivors (17, 15 and 13 mice of both sexes) were killed 24 weeks after the 1st injection. The numbers of mice with lung tumours (males and females

combined) were 6, 7 and 4, and average numbers of tumours per mouse were 0.35 + 0.09, 0.87 + 0.2 and 0.38 + 0.11. Of 30 male and 30 female

control mice injected with steriod-suspending vehicle alone, 6/28 and 6/30 developed lung tumours; the average numbers of lung tumours per mouse were 0.22 + 0.04 and 0.19 + 0.04 (Stoner et al, 1973). (Attention is drawn to the limitation of a negative result obtained in this test


(c) Other experimental systems Administration in conjunction with known carcinogens: Forty pregnant C57B1/6 mice received 100 mg/kg bw dapsone by intragastric

intubation on days 17 and 18 of gestation and 5 times weekly when lactating. Of their offspring, 35 males and 35 females were subsequently given 1. p. injections of urethane (dose and duration unspecified). A control group of 30 male and 30

female young adult C57B1 mice received urethane only. Multiple pulmonary adenomas, lachryhal gland adenomas and liver-cell tumours occurred with the same frequency in the dapsone-plus- urethane-treated group and in the

urethane-control group (Griciute & Tomatis, 1980).

A group of 15 male and 26 female BD IV rats were administered dapsone by gastric intubation 5 times a week during 104 weeks and benzo(a)pyrene by intratracheal instillations. The total dose

of dapsone was 10-16 g/rat. A control group of 16 male and 29 female rats received only

benzo(a)pyrene. Malignant tumours of the


respiratory tract were observed in 8/14 male and 6/24 female rats in the group that received the combined treatment, and in 3/16 male and 12/29 female rats in the group that received

benzo(a)pyrene only (Griciute & Tomatis, 1980) .95^

Dr Philippe Shubik in his written submission said:

Of all the compounds included in the list for discussion at the Royal Commission I would select Dapsone as the most likely compound to pose a potential carcinogenic hazard.

This compound is an aromatic amine which has been found to give rise to mesenchymal tumors of the spleen and thyroid tumors in rats.

The exposure levels are of a much higher order of magnitude from those encountered in dealing with herbicides or pesticides. I recommend an

epidemiological study of those individuals exposed to this drug in Vietnam and a control (group) from (those) who went to Vietnam who received other therapy.

On present data I am unable to say whether

dapsone is actually carcinogenic in humans. (emphasis added)955

Dr Shubik in his oral evidence said:

. . . . Dapsone as I have put down here perhaps being carried away with enthusiasm for feasible research projects. I would - and this is based on both an assessment of likelihood of

carcinogenicity from chemical structure and also on the basis of the available evidence - dapsone is an aromatic amine. It is a compound - when shown to carcinogenesis workers they say we must

test this compound. This is borne out by the fact that the tests recorded of dapsone were undertaken by a man in fact who is now director of the International Agency for Research in Cancer himself.


Although the I ARC concluded that this is limited evidence of carcinogenicity because of these unusual mesenchymal tumours of the spleen and thyroid tumours in rats, I would accord this compound of all the ones I have looked at the highest possibility of possibly potentially producing any damage and that is based on the

additional fact that this is not a pesticide or a herbicide in which we are talking about parts per billion exposure micrograms; we are talking about giving people many milligrams of this compound finishing up with total doses in the order of many, many grams of the material. I would

personally recommend that if it were possible to do this that an epidemiological study of the specific groups given this drug for malaria and as I understand it, this would not be a difficult matter because people given this drug are on record and they could be compared with

appropriate controls - you have all the other factors that they were exposed to - and it would be a situation in which, as I looked over this material this would be indeed the only compound that I could come up with where one could really mount a meaningful epidemiological study, I would have thought without enormous difficulty. It is a compound that has been of interest to people in the past, the leprosy and skin cases who have been treated, so there I would like to say that I feel it is potentially carcinogenic to many, in my opinion. I would think possibly additional animal studies would be interesting but here I would say that an epidemiological study would be the priority.956

Coming as it does from such a highly gualified expert, Dr

Shubik's queries about dapsone have great weight. There

is of course no evidence that dapsone is actually

carcinogenic in man. In summary his evidence is a plea

for a proper epidemiological study.


The Commission finds itself in a position where„ having

regard to the high oral dosages of dapsone, the chemical

structure of the compound and the limited evidence for

carcinogenicity in rats it must regard dapsone as being at

least a possible human carcinogen.

The Commission is reinforced in its opinion by a passage 957

in the transcript of Dr Shubik's evidence where the

doctor added. "Insofar as dapsone is concerned. compounds

that do produce leukopenia with white cell problems, are,

of course, suspect of being carcinogenic."

The Commission recommends the financing of further studies

of this chemical. Any Vietnam veteran suffering from

cancer who may have taken dapsone should have his claim

treated as showing that a reasonable hypothesis exists

connecting his incapacity with his war service.

Chloroquine, Primaquine & Paludrin

In its initial submission to this Commission W A A included

the following sentence:


DRUGS The possible effects of anti-malarial drugs. paludrine, dapsone and chlorquine (sic) and primaquine were given at the end of tours, (at p

51. )

In fact, anti-malarial drugs were used daily and as a

matter of compulsory routine on parades.

In its submission related specifically to toxicology these

drugs were not mentioned.

In its final submission these drugs are not mentioned.

The Commission has itself been unable to find any

suggestion in the literature that any of these three drugs

are carcinogenic.

In his evidence Dr Shubik said that he had actually tested

primaquine himself for carcinogenicity and found that it _ . 958

was negative.

In his wide experience and despite thorough searches he

was unable to find any data suggesting carcinogenicity of

paludrin or chloroquine.

Accordingly, the Commission finds that these compounds are

not carcinogenic and have not caused and will not cause

cancer amongst Vietnam veterans who took them as part of

the anti-malarial campaign in Vietnam.


Η. 4. Solvents

Dimethyl Sulphoxide (DMSO)

This solvent has been shown to be a useful addition to

pesticides in that it speeds up the intake of the

chemicals into the vegetation and the chemical process

within the plant system. As noted above in Chapter

IV.4.4. it was used in a very limited way in Vietnam.

during the Holt/Lugg trials,

Just as it speeded up the process of absorption of the

chemicals into the vegetation it accelerated intake of

chemicals by applicators when they became exposed to the


Health problems experienced in this regard are dealt with


in the Exposure Chapter. As a result of those health

effects the usage of DMSO was suspended and Major Holt's

Report recommended that it should not be used further in

herbicide trials. Accordingly only those engaged in the

Holt/Lugg trials were exposed to this solvent.

The solvent is not mentioned in WAA's initial submission

or in its submission related specifically to toxicological



Although there is a heading. Solvents, in the WAA's final

submission this solvent does not appear amongst those


No evidence was called by VVAA to suggest that this

solvent was carcinogenic.

The Commission's own researches into this solvent indicate

that some investigators suggest that the addition of this

solvent to material which is itself carcinogenic

potentiates or increases the activity of that

. 960


In view of the fact that no case is made against this

compound and in view of the fact that no evidence has been

called to suggest that any person who took part in the

Holt/Lugg trials has suffered cancer the Commission holds

that DMSO is not relevantly carcinogenic and that no

person has suffered cancer or will suffer cancer as a

result of exposure to it in Vietnam.


In its initial submission VVAA mentions distillate only as

a herbicide in the distillate creosote mix.

VI 11-386

In its submission related specifically to toxicological

matters distillate is not mentioned.

In its final submission distillate is mentioned at page 35

under the heading Solvents but no allegation of

carcinogenicity is made.

WHO concluded that, "Benzene and the aromatic extracts are

the only well-defined petroleum solvents for which

carcinogenicity has been reported".961

No allegation of carcinogenicity has been made against

distillate by W A A and no material has been tendered which

suggests that it is a carcinogen. The Commission is

conscious that aromatics are carcinogenic and that there

is some benzene in distillate. However, the Commission is

of the view that the risk of exposure to distillate or the

risk of cancer by exposure thereto was not increased by

service in Vietnam.


In its initial submission W A A makes no mention of



In its submission relating specifically to toxicology

kerosene is mentioned at page 86 but no allegation of

carcinogenicity is made.

In its final submission although there is a heading.

Solvents, kerosene is not mentioned.

The Commission's own researches indicate that kerosene has

never been nominated as a carcinogen and for that reason

no detailed testing has ever taken place. As well,

kerosene comes at the low boiling point end of the

petroleum product scale and it is therefore an unlikely

candidate for a carcinogen.962 The Commission therefore

holds that kerosene is not likely to be carcinogenic and

that no Australian Vietnam veteran suffered or will suffer

from cancer by reason of exposure to this product in



When the Commission analyses the allegations made by Dr

Shearer and Dr Silbergeld it is basic to their reasoning

that animal data are extrapolated to support a claim of

possible effects in humans.


The difficulties of such extrapolation have been well

analysed by Dr Ian Munro who said:

It is of particular concern to public health authorities that certain chemicals have been found to induce carcinogenic teratogenic, mutagenic, behavioural or other effects which may

be subtle in nature and may go undetected in the human population. It is, therefore, of critical importance that chemicals be tested carefully for these effects and that special care be taken in

the interpretation of animal tests designed to assess these potentials.

It is now well established that certain chemicals will produce cancer in animals and presumably, in man as well. In fact, the International Agency for Research on Cancer has noted that at least 20 chemicals or processes will induce cancer in man and animals alike. The presumption of human cancer induction based upon positive results in animal tests requires very special attention. While it is true that all or nearly all human

carcinogens are carcinogenic in animals the converse is not necessarily true. The question of the relevance of tumour induction by chemicals in animals to the human population requires

careful evaluation of scientific and medical facts and often requires expert judgment in assessing the human health significance of animal observations. Tumours may be induced or their natural background may be increased in animals

through a variety of mechanisms. Of greatest concern are those chemicals which are metabolized to highly reactive chemicals which attack DNA causing cellular alterations which lead to cancer development.

On the other end of the spectrum there are

chemicals which induce cancer in animals but only under very special conditions, such as very high (near lethal) doses, severe nutritional

deprivation or markedly altered physiological states. The relevance to humans of tumour

induction under these conditions is negligible at best. Likewise, many chemicals, notably some of


the chlorinated hydro-carbon insecticides have been found to increase the incidence of

naturally-occurring (spontaneous) tumours in the liver of certain strains of mice. This effect alone is of limited significance to the human population exposed to these chemicals. Thus, it can be seen that tumour induction in animals, and in particular, its relevance to exposed human

populations must be evaluated with great


The extrapolation of animal toxicity to the human setting must take cognizance of the limitations and nature of animal tests. Animal tests are designed to maximize the potential of producing a

toxicological response and therefore present a worst case analysis. This is done purposely so as to reduce the probability of not detecting dose-effect and dose-response relationships. Large groups of animals are therefore exposed to a range of chemical concentrations in their diet, air or drinking water over the duration of the test. It should be noted that minimum criteria for the design and conduct of tests acceptable for human risk assessment have been elaborated and that tests not validated as appropriate for human risk assessment are of questionable value for this purpose.

The extrapolation of animal data to humans begins with elaboration of the dose-response curve which describes in quantitative terms the relationship between the applied dose and the incidence and/or

severity of effects induced. Properly designed tests will have included at least one dose which does not induce any observable toxicity within the statistical limitations and power of the test.

This lowest "no-effect" dose is known as the "no-observed effect level" (NOEL) a term which is accepted internationally. The lowest effective dose may also be described as the "threshold dose" or the dose at which effects are first observed.

It is the accepted practice in toxicology to translate the NOEL to a maximum acceptable exposure for humans exposed in a similar manner as were the animals used for the test. Two


procedures are generally accepted for this purpose, one being the use of safety factors and a second being the use of mathematical

(statistical) procedures for risk assessment. The former procedure involves dividing the NOEL by a factor such as 100 or 5000 to obtain an

acceptable human exposure level.

Safety factors take into account the statistical limitations of animal tests, the nature of the dose response curve, the expected differences in susceptibility between animals and humans and variability (genetic differences) in the human

population. The size of the safety factor also will depend upon the nature and degree of

toxicity and as previously stated involves expert scientific judgment. Mathematical procedures are largely reserved for the extrapolation of carcinogenesis data from animals to man because of uncertainty in establishing; NOELs. These procedures, by and large. dispense with the

threshold concept which is not generally accepted in carcinogenesis and are used to establish the statistical probability of tumour induction at very low doses, doses to which humans may be


The final aspect of risk assessment involves the evaluation of human risk given the nature, degree, level, frequency and duration of human exposure. It must be kept in mind that in

carcinogenesis testing animals are exposed to high doses of the chemical in guestion daily for their entire lifetime while in most instances humans may be exposed intermittently or possibly only once or twice during a lifetime to much

lower doses. It is well established that cancer induction. particularly at low doses reguires very long term exposure, usually 20-30 years or more to elicit effects in humans. Brief or

intermittent exposures may Ee 5Έ no consequence because -(a) inadequate target tissue levels are reached to produce toxicity, or

(b) ongoing tissue repair processes counteract any induced damage.964


The Commission has kept these reservations and principles

well in mind in its analysis not only of the data

presented on behalf of W A A but also in its analysis of

the animal and other test data.


In its initial submission W A A claimed general synergistic

effects of the chemical agents that were used in Vietnam,

(at p 11). No particulars were supplied.

From its address it is clear that what is being suggested

is that some combination or combinations of the chemicals

used have produced the effect claimed, including cancer.

In the written submission related specifically to

toxicology synergism is not mentioned.

In its final submission the following submission is found:

The synergistic effects of the possible exposure to chemicals used in Vietnam are just not known. Regard must be given to the possibility however, of

combinations of chemicals used, as having synergistic effects, (at p 8)

Dr Shubik in his written statement to the Commission



I have been asked to speculate on the matter of possible synergism between the various

herbicides, pesticides, drugs and possible environmental factors on any potential cancer producing effects that may have occurred.

I can only respond that I have no knowledge of any studies that would suggest that any of the agents involved have ever been evaluated in any combinations to examine this possibility. Neither am I familiar with any studies that would

suggest theoretically that such a possibility exists.966

Dr Bernard Stewart in his written statement to the

Commission said:

There is not adequate experimental data to provide a basis for predicting the possible carcinogenic consequences to experimental animals of exposure to low doses of the suspect chemicals

in combination. Predicting such consequences for humans, on the basis of experimental data, must be regarded as speculation. If speculation upon the worst possible, but nonetheless credible, scenario is required, it would be predicted that some increase in environmentally-determined promotional influence might be experienced by the exposed population. Nonetheless, such a

prediction must be qualified. To be biologically effective, exposure to promoters normally involves a long period (that is a significant proportion of the individual's lifetime). Despite promotional activity being possible, it

is most likely that it will not be possible to demonstrate such an effect amongst those who served in Vietnam. Either there will, in fact, have been no effect or the effect will have been

so minor as to be not detected by current

investigative procedures.966

Dr Stewart also dealt with interactions between

carcinogens in his oral evidence. He said:


We know very little in terms of the mechanisms that I have described about interactions between carcinogens; and particularly we know very little about interactions between carcinogens of different structure. We know a lot more about

interactions between factors in the human environment because studies establish that human beings known to be exposed to more than one type of carcinogenic influence often exhibit a rate of cancer that suggests that both these stimuli are operative. The textbook example of this is the fact that cigarette smokers who are asbestos workers suffer a much greater incidence of lung

cancer than people who work solely with asbestos or people who smoke but have no known contact with asbestos. This has been exhibited by

uranium - this synergism say between cigarette smoking is seen in lung cancer in uranuim miners. nickel refinery mine workers and cadmium smelter workers, and it seems likely that the impact of atmospheric pollution on urban population is mainly restricted to that fraction of the urban population that already smokes. In other words, one has two factors interacting. So that one can talk about interaction between carcinogenic

influences in the human environment much more readily than one can talk about solid

experimental evidence to explain what is going on in these situations, not because it is not

interesting material for experimental work, but because the experiments are so complicated in light of our present limited knowledge as to be almost uninterpretable.967

Later, during cross-examination by Mclnnes Q.C., Senior

Counsel for WAA, Dr Stewart again commented on the

interaction between carcinogens and synergistic effects.

The following exchange occurred:

You referred this morning to the interaction between carcinogens?--- Yes I did.


Is that the same as a synergistic effect?--- No, it is not. A synergistic effect is a

specific type of interaction where the interaction is such as to produce a combined effect that is worse than either of the

contributing factors acting in isolation. Interactive effects I would consider to be a more general term and refer to any situation in which an individual was exposed to more

than one carcinogen and exhibited some response that suggested the two interacting in some way. It is theoretically possible -

it is not just theoretically possible, it is experimentally possible, to demonstrate a protective effect when apparent carcinogens of different types are given experimentally.

In these experiments with animals. sometimes the interaction increases the number of

tumours?--- That is correct.

DDT is one of the chemicals which have shown in mice and rats to interact with other

carcinogens?--- That is true.

And sometimes with DDT and the other carcinogens there is an increase in tumours?--- Yes.

And sometimes there is a decrease?--- Yes, I believe so. I must confess the usual effect is an increase. Most studies have been directed towards demonstrating an increase. Most of that has been directed towards

demonstrating the so-called promoting effect of DDT.

Is the increase purely additive in the sense that if two separate groups were each given a carcinogen, the total tumours would be the same as a group that is given both

carcinogens?--No, the increase is not -first of all, there is no simple description that would encompass all the experimental data, but the results are generally not additive. Usually, the compounds used in

these experiments are given in some optimal doses. That is to say if the contributing agents are given in isolation, the tumour

incidence would be very low - if not, no

V I 11-395

tumours at all, compared to a control group - whereas the combined treatment there is often a considerable incidence of tumours, where there is no statistical doubt that the

number of tumours is greater than control. Therefore, the effect is more than additive quite often.968


Dr Stewart In his oral evidence before the Commission

expressed his view as to the overall conclusions that

could be drawn from the experimental data dealt with

above. He said:

.... In reviewing the chemicals for which I was aware that carcinogenicity data were available and particularly when such data had been

evaluated by the International Agency for Research on Cancer, I concluded that TCDD and DDT are the chemicals whose activity as carcinogens is best established in experimental animals. When one attempts to extrapolate these data to

predict a hazard for humans. the normal

complications arise. In respect of DDT, the available epidemiology do not establish

carcinogenicity in humans, despite the relatively high level of exposure. Likewise, TCDD cannot be regarded as having been established as

carcinogenic in humans, although it has been implicated in certain studies. Certain of the remaining chemicals are capable of producing benign tumours in mice, and although these findings may be suggestive of human hazard. the activity of these chemicals seems to infer that

they might be active as promoting agents rather than exhibiting complete carcinogenic activity in humans. The epidemiology which one might like to use to confirm many of the hypotheses that I have lust raised are simply not available. I have discussed all of the chemicals in isolation. As


guidelines that provide a solid basis for inferences regarding the effect of small doses of these chemicals in combination with humans. It is - this discussion must be regarded as

speculation. If speculation upon the worst possible, but nonetheless what I believe to be credible scenario is required, it would be predicted that some increase in environmentally determined promotional influences might be experienced by the exposed population.

Nonetheless, such a prediction must be

qualified. To be biologically effective. to promoters normally involves a long exposure period, that is a significant proportion of either the animals or the humans' lifetime, and, of course, we have not got that in the present human situation. Despite this activity being not inconsistent

promotional with the

experimental data, it seems to me most likely that it will not be possible to demonstrate such an effect amongst the population serving in Vietnam. Either there will in fact have been no

effect or the effect will have been so small as to be not detectable bv current investigative procedures and current methods of statistical analysis. (emphasis added).969

Dr Ian Munro during questioning by Counsel Assisting was

asked to assume a "worst case" for Australian Vietnam

Veterans exposure to chemicals in Vietnam and then to

express an opinion on the likelihood of long term toxic

effects from such exposure. The following exchange


Would you expect long term toxicological effects of any sort from that level of

exposure?--- No, I would not.

Why not?--- Well, I think I have to pre-empt any statement by saying under the conditions you


have outlined which are rather extensive it is almost impossible to estimate with any precision the precise quantitative risk of exposure because you did not present any

information with respect to the dosage that would be received. I think there are

certain features in relation to the exposure as you have outlined it that I can comment on, that is, based at least upon my

knowledge of the animal toxicity studies. I would not expect long term effects such as cancer from such exposure because in animal studies it is fairly clear now that lonq term hicrh dose exposure, certainly to near toxic doses of dioxin is required over a very long period of time, essentially the

lifetime in order to produce toxic effects like cancer and therefore I would1 conclude that intermittent and low level exposure over a shor t period, let us say in the ranqe of six months to two years. I would not

expect would induce anv measurable lonq term effects .(emphasis added)y/0

Having regard to the eminence and high qualification of

Doctors Munro, Shubik and Stewart. the Commission

concludes that no long term effect in cancer induction has

been caused to Australian personnel (as defined) by

chemical exposure in Vietnam whether to any particular

chemical or by reason of any synergistic effect of

exposure in Vietnam to any combination of chemicals.

The Commission is of the view that cancer induction by

reason of dapsone ingestion in Vietnam is unlikely but is

not fanciful and a claim based on such allegation should

be treated on the basis that a reasonable hypothesis

exists connecting the cancer with service in Vietnam.


It regards the suggestion of Soft Tissue Sarcoma and

Lymphoma (Non-Hodgkins) induction by exposure to TCDD in

2,4,5-T (in Agent Orange) as unlikely but not fanciful. A

Determining Authority might well be reasonably

that a reasonable hypothesis linking incapacity

such inductions with service in Vietnam exists.





1. Exhibit 1959.

2. Exhibit 1745.

3. This exchange of letters became Exhibit 1161 and Exhibit 1871.

4 . Lancet May 8, 1982, p 1057.

5. See Ch IV.

6. Exhibit 1354 p 3.

7. Ibid p 3.

8 . Ibid pp 3-4.

9. Ibid.

10. Ibid p 6.

11. Exhibit 1377 p 1; Exhibit 1268 p 31;

P 9.

Exhibit 1354

12. Exhibit 1268 pp 31-40 especially 31

pp 3914-3924, Exhibit 1379. ; Transcript

13 . Yamagiwa, K., Ichikawa, K., Jap.

(1915) Vol 5. 142.

Path. Ges.

14 . Deelman, H.T. "Some Remarks on Experimental

Cancer from Tar. A. The Scarification Method. B . Chemical Investigations on Tar." Bulletin de L 'Association Francais Pour L 'Etude du Cancer, Vol 12, (1923). pp 24-30.

15. Oppenheimer, B.S., Oppenheimer, E.T.,

Danishefsky, I., Stout, A.P., Eirich. F.R., "Further Studies of Polymers as Carcinogenic Agents in Animals" Cancer Research Vol 15,(1955), pp 333-340.

16. Barnes, J.M., Denz, F.A., "Experimental Methods Used in Determining Chronic Toxicity, Critical Review." Pharmacol. Rev. Vol 6, (1954), pp 191-242. 1 7

17. Druckrey. H. "Quantitative Aspects in Chemical Carcinogenesis" In Truhaut, R. (ed) Potential Carcinogenic Hazards from Drugs, UICC Monograph (Footnote continued on next page)


(Footnote continued from previous page) Series, 7: 60-77, Springer, Berlin, N.Y., 1967; "Production of Colonic Carcinomas by

1,2-Dialkylhydraz:ines and Azoxyalkanes" In:Burdett, W.J. (ed). Carcinoma of the Colon and Antecedent Epithelium, 1970, Thomas. Springfield.

18. Sir Richard Doll, in Risk Assessment, Report of a Study Group 1983 Royal Society, discussed the concept of negligible risk. An acceptable neglible annual risk of 1 x 10-5 was nominated. One notes that the annual risk of death by motor

car usage in NSW, Australia, is one in 10~4 and most people take that risk every day!

19. Exhibit 1959 p 1193.

20. Doll, Sir Richard, Personal Communication to the Commission at Oxford, October 1984. The reasons for the distinction do not seem to be rational.

21. Exhibit 1965 pp 682-687.

22. Exhibit 1354 p 11.

23. Exhibit 1348.

24. Transcript p 3541.

25. Transcript pp 3552-3.

26. Transcript p 3640.

27. Exhibit 1761.

28. Transcript p 6069 and see also Ch VII.

29. Exhibit 1387

30. Exhibit 1761 p v.

31. Ibid p vii.

32. Ibid p ix.

33. Ibid p 62 Table 4.3.

34. Transcript p 5971.

35. Exhibit 1761, p 56, Table 4.1; Transcript p 5979.

36. Exhibit 1761, para 4.3.1, p 61.



38 .

39 .

4 0 .

4 1 .

42 .

43 .

44 .

4 5 .

4 6 .

4 7 .

48 .

49 .

5 0 .

5 1 .

52 .

53 .

54 .

55 .

56 .

57 .

58 .

59 .

6 0 .

Ibid p 61 para 4.3.1.

Exhibit 1773. p 9.

Transcript p 6103.

Exhibit 1761 p 44.

Ibid p 45.

Transcript p 6108.

Transcript p 6109.

Ch VI.

Transcript pp 2616, 6224.

Transcript p 6217.

Transcript p 6218.

Delegate's Determination, 19 June 1984, Exhibit 1790.

Transcript p 6217.

Transcript p 6200.

Transcript p 6201.

Exhibit 892. 3-54 Para 61. 3-56 Para 70.

See Ch IV.

Transcript p 6201.

Transcript pp 6203, 6203A. and 6204.

Davies et al 1981. Exhibit 1786.

Exhibits 1753 and 1754.

Exhibit 906 pp 1-3. Exhibit 897 pp 3-31. 3-93: Exhibit 1448 p 54.

Exhibit 897 pp 3-93, Exhibit 1448 pp 39-43.

Ibid p 62 table 4.3 .



62 .

63 .

64 .

65 .

66 .


68 .

69 .



72 .

73 .

74 .


76 .


78 .

79 .


Dr Bernard Stewart Transcript p 3589.

Exhibit 1761 p 177 para viii 5.1.

Exhibit 897 pp 3-93.

Exhibit 906 p 1-3, Exhibit 897 pp 3-31: 3-93.

SEER data is American data elaborately collected and collated by a simple technique designed to mirror the American population. An appropriate percentage of those surveyed lived in major

cities and inland; an appropriate percentage in coastal cities; an appropriate percentage in country towns, an appropriate percentage on farms and so on. Australian data is more limited but has been found to be generally consistent with

the SEER data.

Exhibit 852, p 8, Reference 15.

See Exhibit 1888.

Exhibit 906 pp 1-2, 1-14, Exhibit 1448 p 23.

Exhibit 1755. p 1.

See Ch VII and Exhibit 1394 - Lathrop et al 1984 at pp 1-2).

See Ch IV.

Exhibit 1755.

Ibid p 1.



Exhibit 1755 Table 18 p 17.

Exhibit 1755 p 18 Table 20.

Exhibit 1755 p 19 Table 21.

Exhibit 1755 p 31, Table 39.



81. Exhibit 1755 p 33.

82. Exhibit 1755 p 39.

83. Ibid.

84. Exhibit 1394.

85. Ibid p 11-8.

86. Ibid p II-l.

87. Ibid pp III-4. III-65.

88. Ibid p 111-6 paragraph 5.

89. Ibid Table X-l. p X-3.

90. Exhibit 1394 Table X-5 p X-6.

91. Ibid Table X-5.

92. Ibid pp X8-X9. In both Ranch Hand I and Ranch

Hand II. the I CD code 9 th Edition was generally used but for morphological analyses the ICD-0 Code was used. The ICD-0 Code classifications for non-Hodgkin1s lymphoma are M9 5 9 to M963

(Exhibit 1755 Table 21 p 19). The ICD-0 Code classifications for Hodgkin's Disease are M965 to

93 .

M966 (Exhibit 1394 p X-9).

Exhibit 1394 Table X-6.

94 . Exhibit 1394 p X-4.

95. Ibid Table Z-l p X-3.

96 . Ibid p X-4. paragraph 2.

97 . Ibid p X-4.

98 . Ibid Table X-3.

99 . Ibid X

1 σι

100. Ibid p X-6 .

101. Ibid p X-10.

102 . Ibid Table X-10 and p X-13


103. Ibid pp XIX-5 •

104. Ibid pp XIX-9 •

105 . Exhibit 1669 para 18.

106 . Transcript p 5407.

107. Transcript p 5405; 5474.

108 . Exhibit 1394 Table X-6.

109 . Transcript p 5477.

110. Transcript p 5477.

111. Transcript pp 5474, 5478.

112 . Transcript pp 5487-5488.

113. Exhibit 1671 p 8.

114 . Transcript p 5480.

115 . Transcript p 5488 .

116 . Exhibit 1669 p 8 paragraph 18

117. Transcript p 5492

118 . Transcript p 5490-5491.

119 . Exhibit 1394 p X-6.

120. Exhibit 1671 P 9.

121. Transcript p 6564 .

122 . Exhibit 1512.

123. Exhibit 1512 p 6.

124. Ibid.

125. Ibid.

126 . Exhibit 1872.

127. Ibid p 1.


128. Ibid p 21.

129 . Ibid p 21.

130. Ibid p 22.

131. Ibid p 23.

132. Ibid p 23.

133. Transcript p 6582.

134. Exhibit 1874A.

135. Exhibit 906 p 1-3; Exhibit 897 p 3-31.

Exhibit 1448 p 54.

136. See Exhibit 1874B.

137. Dr Mathews Transcript 6586.

138. Exhibit 1165.

139. Exhibit 1165 paras 6 and 7.

140. Dr Mathews: Transcript p 6569 .

141. Exhibit 1165 para 15.

142. Exhibit 1873.

143. Transcript pp 1026-1040.

144. Transcript p 1027

145. Transcript p 1029.

146. Transcript p 6574.

147. Transcript p 6574.

148. Exhibit 1239: 1540.

149. Exhibit 852.

150. Exhibit 1764.

151. Exhibits 1226, 1239. 1244. & 1540.


V I 1 1-406

152. Exhibit 1226 para 13.

153 . Ibid 1226 para 11.

154. Transcript p 2250.

155 . Transcript p 2253.

156 . Exhibit 1244.

157 . Transcript pp 2260-2261.

158 . Exhibit 1226 para 6.

159 . Transcript p 2262.

160 . Transcript pp 2267-2268.

161. Transcript p 2270.

162 . Transcript pp 2260-2261.

163 . Exhibit 1226 para 16.

164 . Exhibit 1239 p 570.

165 . Transcript p 2262.

166 . SF Transcript pp 354-357.

167 . Exhibit 1540 p 12.

168 . Ibid p 15.

169 . Exhibit 852.

170. Ibid p 2.

171. Figures derived from age rates for males. specific national death

172 . Exhibit 852 Table 3.

173 . Ibid p 5.

174 . Table 5 to Exhibit 852


is interesting in this

175 . Transcript p 2262.


176. Exhibit 1922 p 781.

177 . Exhibit 852.

178. Exhibit 1922.

179 . Ibid p 782.

180. Ibid.

181. Exhibit 1271A p 11-99.

182 . Exhibit 1764.

183 . Ibid Table I.

184 . Ibid Table III.

185. Ibid Table III.

186. Ibid p 7.

187 . Ibid Table V.

188. Ibid p 8.

189. Exhibit 763.

190. Table III to Exhibit 763

191. Exhibit 763 pp 23 and 24

192 . Exhibit 1515.

193 . Exhibit 1362.

194 . Ibid p 1.

195 . Ibid p 5.

196 . Ibid p 6.

197 . Exhibit 1515.

198 . Exhibits 721, 128IB.

199 . Exhibit 1516 p 8.

200. Exhibit 1517 p 9.


201 .

202 .

203 .

204 .

205 .

206 .

207 .

208 .

209 .

210 .


212 .

213 .

214 .

215 .

216 .

217 .


219 .

220 .

221 .

222 .

223 .

224 .

225 .

Ibid p 18.

Exhibit 537.

Exhibit 537 p 55.

Exhibit 1518 Table 1.

Ibid p 12.


Ibid p 15.

Exhibit 763. p 21.

Transcript p 3741.

Exhibit 763 p 23 and 24.

Transcript p 3693.

Transcript p 3693.

Transcript p 3694.

Exhibit 1362 p 1.


Ibid p 5.

Ibid p 5

Ibid p 6.

Ibid p 5.

Ibid p 6.

Exhibit 11 and Exhibit 1929.

Exhibit 11 p 78, Exhibit 1929 p 6.

Exhibit 763 p 24.

Exhibit 1929 abstract.

Ibid p 6.

V I 11-409

226 .

227 .

228 .

229 .



232 .

233 .

234 .

235 .

236 .

237 .

238 .

239 .

240 .


242 .

243 .

244 .

245 .

246 .

247 .

248 .

249 .


Exhibit 11 p 78.


Transcript p 3692.

Transcript pp 3756-3757.

Transcript p 3757.

See Ch II.

Transcript pp 4045-4046, Dr J.D. Mathews.

Transcript p 3762.

Transcript p 3764.

Transcript p 3757.

Transcript p 3760.

Cf Dr Gaffey: Exhibit 1543 p 5; STL Transcript 60/61.

Transcript pp 3779-3780.

Transcript p 3780.

Transcript p 3749; Exhibit 1928 p 8822.

Exhibit 1928 p 9071; Transcript p 3748.

Transcript p 3751.

Transcript p 3751.

Transcript 3765; Exhibit 1928 p 8843.

Exhibit 1928 p 8841-8842.

Transcript p 3761.

Exhibit 763 p 21.

Exhibit 1928 p 8997.

Exhibit 1369. formerly ΜΕI 05.

Transcript p 3872; Exhibit 1369.


251. Transcript p 3790.

252. Transcript p 3791.

253. Transcript pj) 3873-3874.

254 . Transcript pj> 3777-3778.

255. Exhibit 1928 p 8957; cf Transcript p 3875

256. Transcript p 3875.

257. Exhibit 1369.

258 . Transcript p 3875 .

259 . Transcript p 3875.

260. Exhibit 1928 p 8957.

261. Transcript p 3700.

262 . Transcript p 3767 .

263 . Transcript p 3767 .

264 . Transcript p 3768 .

265. Transcript p 3772 .

266. Transcript p 3782.

267. Transcript p 3782 .

268 . Transcript p 3782 .

269 . Transcript p 3784.

270. Transcript p 3786 .

271. Exhibit 1928 pp 9082-3.

272 . Transcript p 3787 .

273 . Transcript p 3768.

274. Transcript p 3787 .

275. Transcript p 3700.


276 . Transcript P 3702.

277. Transcript P 3779.

278 . Transcript P 3778 .

279 . Transcript PI) 3783-3784.

280 . Transcript P 3800.

281. Exhibit 1370.

282 . Transcript P 3802.

283 . Transcript P 3802 .

284 . Transcript P 3802 .

285 . Transcript P 3802a.

286 . Transcript P 3808 .

287. Transcript P 3809 .

288 . Transcript P 3805 .

289 . Transcript P 3806 .

290. Transcript P 3830.

291. Transcript P 3707 .

292. Transcript P 3706.

293 . Transcript P 3707 .

294 . Transcript P 3708 .

295 . Transcript P 3709 .

296 . Transcript P 3709 .

297. Transcript P 3830.

298 . Exhibit 763 Tables 3. 4. 5 and

299 .


Exhibit 11,

Exhibit 11.

Table 4. footnote.



302 .

303 .


305 .

306 .

307 .

308 .

309 .



312 .

313 .

314 .

315 .

316 .

317 .

318 .

319 .



322 .

323 .

324 .

325 .

301. Transcript p 3831.

Transcript p 3827.

Transcript p 3864.

Exhibit 11 p 73.

Exhibit 11 p 74.

Exhibit 1766 p 8.

Transcript p 3821.

Transcript p 3826.

Transcript p 3847-3863.

Transcript p 3832.

Transcript p 3863.

Transcript p 3864.

Exhibit 1362.

Exhibit 11 Table 2.

Transcript p 3837.

Transcript p 3832„ 3833

Transcript p 3839.

Transcript p 3840.

Transcript p 3836.

Exhibit 11 p 77.

Transcript p 3693.

Exhibits 716. part 721,

Ibid and Exhibit 1281C.

Exhibits 537, part 721,

Exhibit 1509.

1281A and 1516.

128 ID, 1508 . and 1518.


326 .

327 .

328 .

329 .



332 .

333 .

334 .




338 .

339 .



342 .

343 .

344 .

345 . 346 .

347 .

348 .

Exhibits 128IB; 1519 and part 721.

San Francisco Transcript p 224.

Exhibit 1515.

Exhibit 537; San Francisco Transcript p 278.

8 May 1985, Gibbs et al v. Dow et al; US District Court, Eastern District of New York MDL 381 at p 48 .

Palmer & Ors v. Stora Kopparberqs Berqslaqs Aktiebolag trading as Nova Scotia Industries (1983) 60 Nova Scotia Reports (2d) 271.

Supra at p 352.

Exhibit 1515; also described as Study 1 in

Exhibit 721.

San Francisco Transcript p 152.

Exhibit 1515.

Exhibit 1515 p 4.

Ibid p 4.

Ibid p 1.

Exhibit 1516 p 1.

San Francisco Transcript p 129.

Exhibit 1515 pp 2 and 3.

San Francisco Transcript p 158.

San Francisco Transcript p 71.

San Francisco Transcript pp 72-73.

San Francisco Transcript pp 132-134. San Francisco Transcript p 134; 157.

Exhibit 1516; San Francisco Transcript p 172.

Exhibits 1244. 1540.


349 . San Francisco Transcript p 11.

350. Exhibit 1515 p 3.

351. San Francisco Transcript p 141.

352 . San Francisco Transcript p 141.

353 . San Francisco Transcript pp 141-142.

354 . San Francisco Transcript p 130.

355 . San Francisco Transcript p 134.

356 . Exhibit 1515 p 4.

357 . Ibid.

358 . San Francisco Transcript p 132.

359. Exhibit 1515 p 4.

360. From Exhibit 1515.

361. San Francisco Transcript p 130.

362 . Exhibit 1515 p 4.

363 . San Francisco Transcript pp 138-9.

364 . San Francisco Transcript p 139.

365 . San Francisco Transcript pp 155-6.

366 . San Francisco Transcript p 161.

367. San Francisco Transcript p 165 and Exhibit 1523.

368 . San Francisco Transcript p 87.

369 . San Francisco Transcript pp 87-88.

370. Exhibit 1523; San Francisco Transcript pp 164-165.

371. Exhibit 1521, 1522.

372 . San Francisco Transcript pp 90-92.

373 . San Francisco Transcript pp 90-92 and Exhibits 1520, 1521, 1522, 1372 and 1373.


374. Exhibits 1372. 1373, 1523 and 1524; San Francisco Transcript pp 94, 136, 164.

375. Transcript pp 1827-1828.

376. San Francisco Transcript pp 114-115, 117; compare Exhibit 1875.

377. San Francisco Transcript p 100.

378. San Francisco Transcript pp 90, 101-102; Exhibit 1372.

379. See North Sweden Questionnaire, Exhibit 1533; North Sweden Questionnaire, Exhibit 1367; Malignant Lymphoma Questionnaire, Exhibit 1534; San Francisco Transcript p 194.

380. San Francisco Transcript p 122; Exhibit 1366 -(formerly MFI-02 English translation). Exhibit 1533 (Original Swedish).

381. Exhibits 1524


, 1524A; San Francisco Transcript p

382 . Exhibit 1533.

383 . 80-90,000 people; San Francisco Transcript p 80.

384 . San Francisco Transcript p 80.

385 . San Francisco Transcript p 81.

386 . San Francisco Transcript p 82.

387 . San Francisco Transcript p 175.

388 . Exhibit 1515; San Francisco Transcript pp 174-175.

389 . San Francisco Transcript p 173.

390. San Francisco Transcript p 178.

391. San Francisco Transcript p 97.

392 . Exhibit 1515.

393 . San Francisco Transcript p 185.

394 . San Francisco Transcript 186.


395 . San Francisco Transcript P 96.

396 . San Francisco Transcript 179-180.

397 . San Francisco Transcript PP 186-187

398 . San Francisco Transcript P 187.

399 . San Francisco Transcript P 216.

400. Exhibit 1533.

401. Exhibit 1524, 1524A.

402 . Exhibit 1516 : P 4.

403 . San Francisco Transcript P 210.

404 . Exhibit 1516 P 3.

405 . San Francisco Transcript PP 222-223

406 . San Francisco Transcript PP- 214-215

407 . San Francisco Transcript P 197.

408 . Exhibit 1515.

409 . Exhibit 1281A p 713.

410. San Francisco Transcript P 188 .

411. San Francisco Transcript P 188 .

412. Exhibit 1281A •

413 . Exhibit 1515.

414 . San Francisco Transcript P 202.

415. San Francisco Transcript P 278 .

416 . San Francisco Transcript P 201.

417 . Transcript p 2969 .

418 . San Francisco Transcript P 202.

419 . San Francisco Transcript P 205.











432 .

433 .

434 .

435 .

436 .


438 .

439 .

440 .


442 .

443 .

444 .

San Francisco Transcript p 203.

Transcript 2971.

San Francisco Transcript p 205.

San Francisco Transcript p 206.

San Francisco Transcript p 207.

San Francisco Transcript p 211.

Exhibit 1985.

Exhibit 1516 p 5.

Exhibit 1516 p 8.

Exhibit 1516 p 5.

Exhibit 1516 p 5.

Exhibit 1516 Table 3.

Exhibit 1281A.

San Francisco Transcript p 215-216.

San Francisco Transcript p 180.

San Francisco Transcript p 216.

San Francisco Transcript p 217.

San Francisco Transcript p 218.

San Francisco Transcript p 222.

Exhibit 1281A p 716.

Exhibit 1251A p 716.

San Francisco Transcript p 261.

San Francisco Transcript p 273.

Exhibit 1281 p 716.

Exhibit 1281C p 28.


445 . San Francisco Transcript P 162.

446 . San Francisco Transcript P 85.

447. Exhibit 1281C Table 1.

448 . Exhibit 1281C p 29.

449 . Exhibit 1281C p 29.

450. Exhibit 1281C Table 3.

451. Exhibit 12 81C,

452. Exhibit 1281C p 32.

453 . San Francisco Transcript P 278; Exhibit

454 . San Francisco Transcript P 16.

455 . San Francisco Transcript P 241; Exhibit

456 . San Francisco Transcript P 257-258.

457 . Exhibit 1518 Tables 2.3 and 4.

458 . San Francisco Transcript PP' 17-18.

459 . San Francisco Transcript P 18.

460 . San Francisco Transcript P 18.

461. San Francisco Transcript P 255.

462 . San Francisco Transcript P 256.

463 . San Francisco Transcript P 257.

464 . San Francisco Transcript P 257.

465 . San Francisco Transcript P 244-5.

466 . San Francisco Transcript P 245 .

467 . San Francisco Transcript P 246 .

468 . San Francisco Transcript P 247 .

469 . San Francisco Transcript P 247 .

V I 11-419

470. San Francisco Transcript p 248-9.

471. San Francisco Transcript p 247.

472 . San Francisco Transcript p 248.

473 . San Francisco Transcript p 248.

474 . San Francisco Transcript p 249.

475 . San Francisco Transcript p 250.

476 . San Francisco Transcript p 251.

477 . San Francisco Transcript p 252.

478 . San Francisco Transcript p 252.

479 . San Francisco Transcript p 253 .

480. Exhibit 1509.

481. Exhibit 1509 Table III and p 251.

482 . Exhibit 1509 p 249.

483 . Exhibit 1509 p 250.

484 . Exhibits 128IB and 1519.

485 . San Francisco Transcript 224; 279.

486 . San Francisco Transcript pp 279-280

487. San Francisco Transcript p 231.

488 . San Francisco Transcript p 232.

489 . San Francisco Transcript p 232.

490. San Francisco Transcript p 232.

491. San Francisco Transcript p 233 .

492 . San Francisco Transcript p 234 .

493 . San Francisco Transcript p 235.

494 . San Francisco Transcript p 235.

V I 1 1-420


496 .

497 .

498 .

499 .



502 .

503 .

504 .

505 .

506 .

507 .

508 .

509 .



512 .

513 .

514 .

515 .

516 .

517 .

518 .

San Francisco Transcript p 235.

San Francisco Transcript p 235.

San Francisco Transcript p 236.

San Francisco Transcript p 236.

San Francisco Transcript p 237.

San Francisco Transcript p 237.

Exhibit 1519; San Francisco Transcript p 258.

San Francisco Transcript

San Francisco Transcript

San Francisco Transcript

San Francisco Transcript

San Francisco Transcript

San Francisco Transcript

San Francisco Transcript

San Francisco Transcript

San Francisco Transcript

San Francisco Transcript

See The Exhibit 1539 para 7 Axelson Technique.

San Francisco Transcript

San Francisco Transcript

San Francisco Transcript

San Francisco Transcript

Exhibit 1539 P 3.

Exhibit 1539 p 4.

p 258.

p 259 .

p 260.

p 261.

p 261.

p 262.

p 263 .

p 265.

p 273.

p 308-9.

et seq. and Exhibit 1360;

pp 285-288.

p 286 .

p 287 .

p 287.


519 . See Endnotes 330, 331 and 332.

520. Exhibit 1738 p 10, Exhibit 1283 p 131.

521. Exhibit 250.

522. Exhibit 250; Table 2.

523 . Exhibit 250 p 13.

524. Exhibit 250 p 14.

525 . Exhibit 1916.

526 . Exhibit 1916 Table 8.

527 . Exhibit 1916 Table 9.

528 . Exhibit 1916 Tables 9 & 10.

529 . Exhibit 1916 p 3.

530. Exhibit 1916 p 13.

531. Exhibit 1917.

532 . Ibid.

533 . Ibid p 176.

534 . Exhibit 906 p 1-23 but see


also Exposure, Chapter

535 . Exhibit 1917 p 179.

536 . Exhibit 1918.

537 . Ibid p 2379.

538 . Ibid p 2373.

539 . Ibid p 2379.

540. Ibid p 2375.

541. Ibid p 2374.

542 . Ibid p 2376.

543 . See Exhibit 1394 p XIX-5.


544. Exhibit 673.

545 . Exhibit 147.

546 . Exhibit 673 .

547 . Ibid p 1 531.

548 . Ibid at pp 531-2.

549 . Exhibit 1541 Slides 5 and 6.

550 . Exhibit 673 p 532.

551. Exhibit 147.

552 . Ibid Table 3.

553 . Ibid Table 4.

554 . Ibid Tables 3 and 4.

555 . Exhibit 147 p 49.

556 . Ibid p 48 .

557 . Ibid p 49 .

558 . Exhibit 736.

559 . Exhibit 250.

560. Exhibit 673 .

561. Exhibit 1916 .

562 . Exhibit 736 at p 269.

563 . Exhibit 1543 paragraph 2.

564 . Exhibit

to 29. 1547 para 2 and St Louis transcript pp 21

565 . St Louis transcript at p 28.

566 . An Epidemiologist from Western Australia.

567 . Transcript 6026.

V I 11-423

568 .

569 .

570 .


572 .

573 .

574 .

575 .

576 .

577 .

578 .

579 .

580 .


582 .

583 .

584 .

585 .

586 .


588 .

589 .



592 .

Exhibit 1543 para 3[a].

STL Transcript 21.

Exhibit 1544.

Exhibit 748.

Exhibit 849.

Exhibit 843.

Exhibit 1541.

See Slide 6 to Exhibit 1541.

Exhibit 1541 Slide 9.

St Louis transcript at pp 51-58

Exhibit 1543 p 4.

Exhibit 1546 p 4.

St Louis transcript at p 54.

Exhibit 1543 para 12.

Exhibit 849.

Exhibit 1765.

Ibid p 180.

Ibid at p 181.

Exhibit 1283 p 104.

Transcript at p 2937.

Exhibit 1920.

Exhibit 198, 1919.

Exhibit 1920.

Exhibit 1921.

Ibid at p 5 and Tables 3a and 3b

VI 11-424

593. Exhibit 1919.

594. Exhibit 1765.

595 . Ibid p 183.

596. Ibid p 188.

597. Exhibit 1921.

598 . See Dr Hay Transcript at p 2876.

599 . Exhibit 1755 Table 20.

600. Exhibit 1394 Table X-5.

601. Exhibit 1761 Table 4.1 and para 4.3.1 p 61

602 . Exhibit 1921 p 5.

603 . Exhibit 1258.

604 . Exhibit 1258 p 288 .

605 . Exhibit 1258 p 289.

606 . Transcript at pp 763-5.

607. Exhibit 1258 p 288.

608 . Exhibit 1258 Table II.

609 . Exhibit 1258 p 298.

610 . Exhibit 1258 p 300, Table XIV.

611. Exhibit 1258 p 322.

612 . Exhibit 763 p 23.

613 . Exhibit 1362 p 6.

614 . Chapter IV.

615 . Dr D.G. Crosby at transcript p 790 et seq.

616 . Dr D.G. Crosby at transcript pp 770 et seq

617 . Exhibit 1377 p 2; Transcript 3922-3924.

V I 11-425






623 .

624 .


626 .


628 .

629 .




633 .

634 .

635 .

636 .

637 .

638 .

639 .



642 .

Transcript 3948.

Exhibit 1354 p 10.

Exhibit 1268 p 33.

Exhibit 1351.

Ibid p 101.

Ibid p 13. .

Exhibit 180.

Exhibit 1377 p 4.

Ibid p 4.

Exhibit 1371 p 7.

Exhibit 1377 p 7.

Ibid p 7.

Transcript 3923-3924.

Exhibit 1379 p 2.

Exhibit 1377 p 9.

Exhibit 606.

Exhibit 1377 p 7.

Ibid p 9.

Transcript 3968.

Exhibit 1379 p 1.

Exhibit 1379B Table 3; p 3.

Exhibit 103 pp 124-126.

Exhibit 1577A p 5.

Transcript 5290.

Exhibit 1377 p 8.


643 .

644 .

645 .

646 .

647 .

648 .

649 .



652 .

653 .

654 .


656 .

657 .

658 .

659 .



662 .

663 .

664 .

665 .

666 .

Ibid p 8.

Exhibit 1599 p 226.

Transcript p 5297-5298.

Transcript p 5308.

Exhibit 1577A p 5.

Transcript 5310.

Transcript 5293.

Transcript 5310-5311.

Transcript 5230.

Transcript 5319.

Transcript 5246.

Transcript 5267.

Exhibit 1591.

Exhibit 1577 p 2 paragraph 5.

Exhibit 1312 Appendix 1 part III section C p 24 and Table 2 to Exhibit 1312.

Exhibit 1312 Appendix I Part IV pp 32-37.

Exhibit 1312 Appendix I Part IV p 32.

Exhibit 1312 Appendix 1 Part IV p 36.

Extracted from Exhibits 1268, 130, 180, 1351, 1377.

Exhibit 1268 p 35.

Exhibit 1377 p 4.

Exhibit 1377 p 5.

Exhibit 1377 p 5.

Exhibit 1377, p 4.


667. Exhibit 1379B.

668 . Exhibit 103 pp 287-288.

669 . Exhibit 1377 p 10.

670. Exhibit 1351 at p 236.

671. Exhibit 1676.

672. Exhibit 1676 p 6.

673 . Exhibit 1577A p 7.

674 . Exhibit 1577A. p 7, final paragraph

675 . Exhibit 1312 Appendix 1 p 19.

676 . Exhibit 711.

677 . Exhibit 1312 Appendix 1 p 19.

678 . Transcript 3239 - 3244.

679 . Transcript 3242.

680. Transcript 3242.

681. Transcript 3243.

682 . Transcript 3243.

683 . Exhibit 1301.

684 . Exhibit 1351 at p 240.

685 . Exhibit 1377 p 13.

686 . Exhibit 1268 p 58.

687 . Exhibit 1268 p 59.

688 . Exhibit 1271A.

689 . Exhibit 1271A p 10-1.

690. Exhibit 1271A p 10-12.

691. Exhibit 1577A p 7 final paragraph.


692 .

693 .

694 .

695 .

696 .

697 .

698 .

699 .

700 .


702 .

703 .

704 .

705 .

706 .

707 .

708 .

709 .

710 .


712 .

713 .

714 .

715 .

716 .

Exhibit 1607.

Exhibit 1607 pp 217. 221.

Exhibit 1283.

Exhibit 1280 p 9 para 11-5.

Exhibit no 1286.

Exhibit 1280.

Exhibit 1283.

Exhibit 1280.

Transcript 2932.

Transcript 2935.

Transcript 2935.

Transcript 2892.

Exhibit no 1286.

Transcript 2893.

Transcript 2894.

Exhibit 1286.

Transcript 3941.

Transcript 3941.


Transcript 3947.

As depicted in Exhibit 1286.

Transcript 3942 -3944.

Transcript 3945.

Transcript 3945.

Transcript 3945.


717. Transcript 3947.

718. Exhibit 906 pp 1-23.

719 . Exhibit 1379 p 3.

720. Exhibit 659.

721. Exhibits 82 and 1592.

722. Exhibit 1592 p 122.

723 . Exhibit 1592 p 126.

724 . Exhibit 1592 p 126.

725. Exhibit 753.

726 . Exhibit 1268 p 35.

727 . Exhibit 1268 p 35.

728. Exhibit 1268 p 35.

729 . Exhibit 1377, p 19.

730. Exhibit 1354, p 11A.

731. Exhibit 103 p 114.

732 . Exhibit 82 .

733 . Exhibit 659 .

734 . Transcript 5128.

735 . Transcript 5129.

736 . Exhibit 1352.

737 . Exhibit 1577, p 3.

738 . Exhibit 753, Transcript 5128

739 . Exhibit 1352.

740 . Transcript 5316.

741. Exhibit 1354 p 11A.



743 .

744 .

745 .

746 .

747 .

748 .

749 .



752 .

753 .

754 .

755 .

756 .

757 .

758 .

759 .



762 .

763 .

764 .

765 .

766 .

Transcript 5330.

Exhibit 1663.

Transcript 5337.

Transcript 5337.

Exhibit 1354 p 12.

Exhibit 659.

Exhibit 114.

Exhibit 133.

Exhibit 1354.

Exhibit 1354 p 12.

Ibid p 12.

Transcript 3607.

Transcript 3606.

Transcript 3607.

Exhibit 1354 p 12.

Exhibit 130 pp 279-282. 292.

Exhibit 1351 pp 235-236.

Exhibit 1377 pp 19.

Ibid p 21.

Exhibit 1349 p 11.

Exhibit 1676 p 3, para 6.

Exhibit 1280 p 17 para 18.

Exhibit 1363.

Exhibit 116.

Exhibit 116 p 302.


767. Exhibit 1268 p 60.

768. Exhibit 166 p 302.

769 . Exhibit 116 p 302.

770. Exhibit 556.

771. Exhibit 1268 p 62.

772 . Ibid p i 62.

773 . Transcript 3607.

774 . Transcript 3608-9.

775 . Exhibit 1354 .

776 . Exhibit 1354. p 13.

777. Exhibit 1354 p 14.

778. Exhibit 1354 p 14.

779 . Ibid p ] .4 .

780. Exhibit 1351 p 239.

781. Ibid pp 238-239.

782 . Exhibit 1349 pp 12-13.

783 . Exhibit 1676 pp 6 para 12

784 . Transcript 5629.

785 . Transcript 5573.

786 . Exhibit 1686 .

787. Transcript 5573.

788 . Transcript 5574.

789 . Transcript 5574.

790. Transcript 5575-5576.

791. Exhibit 1271.


792 .

793 .

794 .

795 .

796 .

797 .

798 .

799 .



802 .

803 .

804 .

805 .

806 .

807 .

808 .

809 .



812 .

813 .

814 .

Transcript 5696.

Exhibit 1687 p 222.

Exhibit 1705.

Transcript 5697, 5703-5704.

As U.S. Court Exhibit 4387 in the Agent Orange Product Litigation.

Exhibit 1705 p 4.

Ibid Addendum p 2.

Transcript 5570.

Exhibit 232.

Exhibit 1271 p 11-4.

Exhibit 232 p 631.


Exhibit 1271 p 11-4.

Exhibit 1271 p 11-49.

Transcript 3636.

Transcript 3636.

Exhibit 1268 p 61.

Exhibit 1676.

Exhibit 1676A.

Exhibit 1687.

FAO/WHO, Joint Meeting of the FAO Panel and Experts on Pesticide Residues and the Environment and the WHO Expert Committee on Pesticide

Residues, 1970, FAO Rome, pp 253-254.

Exhibit 733.

Exhibit 828.


815. Exhibit 1245.

816 . 1970 Report of Joint Meeting of Pesticides.

817. Exhibit 732.

818. Exhibit 828 .

819. Exhibit 969.

820. Ibid at p 9-10.

821. Exhibit 1895 p 64.

822 . Exhibit 1615.

823 . Exhibit 1896 at p 62.

824 . Exhibit 894 Ch 10 at pp 16-17.

825 . Exhibit 1617 at p 207.

826 . Exhibit 1616 at pp 216-218.

827 . Ibid at pp 219-220.

828 . Exhibit 1224 at p 625.

829 . Ibid at p 628 .

830. Exhibit 894 Ch 6 at pp 2-3.

831. Ibid Ch 6 at p 6.

832 . Exhibit 934 .

833 . Exhibit 1354 at pp 14-15, see also transcript at p 3612 to the same effect. (The Dow Study has not yet been reported on although the Commission contacted Dow, Dow's reply dated 14/6/85 is Exhibit 1961).

834 . 835 .

Exhibit Exhibit 1349 at p 13. 1616.

836 . Transcript at p 3562.

837. Transcript at pp 2375-6.

838 . See Transcript p 5200-5204, 5378.


839 . Exhibits 779 and 1288.




843 .

844 .

845 .

846 .

847 .

848 .

849 .



852 .

853 .

854 .

855 .

856 .

857 .

858 .

859 .

860 .


862 .

Williams & Wilkins Baltimore/London (1982) at p 540.

Ibid at p 541.

Exhibit 892.

Exhibit 1654A.

Exhibit 779 and 1288 p 12.

Exhibit 894 Vol 1, Ch 10 at pp 17-18.

At p 102.

Exhibit 779 p 12. and 1288.

Exhibit 1224A at p 621.

Ibid at p 647.

Ibid at p 622.

Arch. Path. Vol 94 p 270.

Notice dated August 2, 1978.

Exhibit 1897.

Exhibit 779 and 1288 at p 11.

IARC Monographs volume 23. 39-141 Lyon 1980.

Exhibit 1908 at p 22.

Exhibit 894 Volume 3 Chapter 7 p 1.

Exhibit 1354 at p 15.

Transcript at p 3614.

Exhibit 1245 p 15 and Transcript p 2376.

See Ch IV Exposure.

Transcript p 3928.


863. Exhibit 1622. 1623.

864 .


866 .


868 .

869 .



872 .

873 .

874 .

875 .

876 .


878 .

879 .




883 .

884 .

885 .

886 .


Exhibit 1909 at p 14.

19 October 1984.

Personal Communication. November 1984.

Exhibit 1643.

Exhibit 1643.

Exhibit 1676 p 10 paragraph 24.

Exhibit 1970 at pp 61-62.

Ibid at p 62.

Exhibit 1981 p 108-109.

Exhibit 1980 p 4.

Exhibit 1898 at p 17.

Exhibit 1353 at pp 165-167.

See Orqanochlorine Pesticides (1981) at pp 17-22.

Exhibit 1351 at p 106.

Exhibit 1354 at p 15.

Transcript at pp 3614-15.

Exhibit 1349 at p 14, Environmental Health Criteria document at p 147.

Ibid at p 14.

Transcript at pp 3563-4.

Report 4.18-DDT. p 35.

Occupational Exposure to Malathion.

Exhibit 1982 para 2.1.7.

Exhibit 717 at p 13.

Exhibit 1978 pp 113-114.


888 .

889 .



892 .

893 .

894 .

895 .

896 .

897 .

898 .

899 .

900 .


902 .

903 .

904 .

905 .

906 .


908 .

909 .



Ibid at p 120.

Exhibit 1900 at pp IV.6-7.

Exhibit 1354 at p 15.

Transcript at p 3614.

Exhibit 1349 at p 13.

Transcript at p 3563.

Transcript at p 2380.

Exhibit 1969 pp 93-95.

Exhibit 1971 pp 183-188.

Ibid pp 196-197.

Exhibit 1979 pp 143-4.

Exhibit 1974 p 4.

Exhibit 1975 p 7. The lengthy discussion of recent dieldrin carcinogenicity studies in publication is found at pp 3-6 thereof.

Exhibit 1353 at pp 169-170.

Exhibit 730 at pp 5-6.

Exhibit 1351 p 112.

Exhibit 1354 at p 16.

Transcript at p 3619.

Exhibit 1349 p 15.

Transcript p 3566.

Transcript pp 2381-2.

Exhibit 1971 pp 15-16.

Ibid p 12.

Exhibit No 1903 at pp 7-9.

more this










923 .

924 .


926 .


928 .

929 .



932 .

933 .

934 .

935 .

Exhibit 1975 at p 80.

Exhibit 1976 p 3.

IARC Monographs: Vol 20 - Some Haloqenated Hydrocarbons (1979) p 57.

Exhibit 1351 at p 81.

Exhibit 1902 pp 32. 33. 34.

Exhibit 1354 at p 16.

Exhibit 1349 at p 15.

Transcript at pp 3564-66.

Exhibit 1378 at p 18.

Transcript at pp 3926-7.

Exhibit 1904.

Ibid pp 12-13.

Ibid at pp 17 & 18.

Exhibit 1612.

Exhibit 1648.

Exhibit 1649.

Exhibit 1969 at p 122-123.

Exhibit 1972 at p 312.

Ibid at p 313.

Exhibit 1975 at pp 334-335.

Exhibit 1977 at p 222.

Exhibit 1351 at p 133.

Exhibit 1354 at p 16. see also transcript at p 3619 to the same effect.

Transcript at pp 3566-67.


937 .

938 .

939 .



942 .

943 .

944 .

945 .

946 .

947 .

948 .

949 .



952 .

936. Bioassay_____ of_____ Diazinon_____ for_____ Possible

Carcinogenicity. National Cancer Institute, Bethesda Md. Carcinogenesis, Technical Report Series. No 137, 1979.

Exhibit 1354 at p 16.

Transcript at p 3619.

See also transcript at p 3566 and Exhibit 1349 p 15.

Exhibit 1973 at pp 4-6.

"Toxicity and Health Threats of Phthalate Esters: Review of the Literature". Env. Health Persp. 4, 3-26, at pp 17-18.

"Studies on the Toxicity of Phthalates via Ingestion". 1973 Env. Health Persp. 3, 61-72 (at P 65).

Ibid at p 66.

Ibid at p 62.

Toxicity and Metabolism of Phthalate Esters, Clinical Toxicology 1978 pp 257-268 at p 261.

"Toxicity and Health Threats of Phthalate Esters: Review of the Literature". 1973 in " Env. Health Persp. 4, 3-26 at pp 17-18.

"Studies on the Toxicity of Phthalates via Ingestion". 1973 Env. Health Persp. 3, 61-72, at p 65.

Ibid at p 66.

Ibid at p 62.

"Toxicity and Metabolism of Phthalate Esters". Clinical Toxicology 1978 in 13 (2), pp 257-268, at p 261.

Ch IV.

Transcript p 280.


953 .


955 .

956 .


958 .

959 .



962 .

963 .

964 .

965 .



968 .

969 .


Transcript p 355.

Exhibit 861 at pp 64-66.

Exhibit 1354 at p 16.

Transcript at pp 3615-16.

Transcript at p 3618.

Transcript p 3616.

Ch IV.

Patty's Industrial Hygiene and Toxicology,(eds) Clayton, G.D. & Florence E . , N.Y. (1981) Wiley, Revised Third Edition.

Exhibit 1909 at p 14.

Dr Bernard, Stewart, Personal Communication, May 1985.

Exhibit 1245 at pp 2-4.

Exhibit 1245 at pp 6-8. See also transcript at pp 2346-50; and pp 2355-59 to the same effect.

Exhibit 1354 at p 17.

Exhibit 1349 at p 16.

Transcript at pp 3550-51.

Transcript at pp 3590-90a.

Transcript at p 3567.

Transcript at p 2374.