Document jg4YY9pbDLNLmN00Om7BoR8jN

Teratogcnesis, Carcinogenesis, and Mutagenesis 5:251-318 (1985) Extrapolation of the Evidence on Teratogenicity of Chemicals Between Humans and Experimental Animals: Chemicals Other Than Drugs \ K. Hemminki and P, Vineis Institute of Occupational Health, Helsinki, Finland (K.H.) and Istituto di Anatomia e Istologia Palologica, Torino, Italy (P.V.) Epidemiologic literature regarding the possible association between malformations and 23 exposures or occupations other than pharmaceutical products, was ana lysed. The qualitative level of scientific evidence was classified into four cate gories: high (ethanol, methylmercury, PCBs, laboratory work), limited (anesthetic gases, carbon monoxide), low (hexachlorophene, LSD, nitrous oxide, smelter work, tobacco), and inadequate (all other exposures). Human data for exposures belonging to categories "high" and "limited" were quantitatively compared to results of animal teratogenicity tests of the relevant chemicals. Ethanol, methylmercury, and PCBs have caused malformations in experimental animals, and the effective doses have ranged from 0.2 to 8.0 times the effective human doses. Ethanol and PCBs caused similar types of lesions in some animal species as have been observed in humans. Key words: chemical exposures, malformations, epidemiology, teratogenicity testing, ethanol, methylmercury, PCBs FOREWORD This review was commissioned to the authors by the Advisory Subgroup in Toxicology of the European Medical Research Councils in 1981, The need was felt for a consideration, a posteriori, of the current practice of testing chemicals for teratogenicity and other embryotoxic effects in laboratory animals. This practice is based upon the belief that results of such tests will predict the ability (or lack of ability) of specific chemicals to induce similar effects in man. The financial support of the Finnish Medical Research Council and of the Italian National Research Council is acknowledged. Address reprint requests to K. Hemminki, Institute of Occupational Health, Haartmaninkatu 1, 00290 Helsinki 29, Finland. 1985 Alan R. Liss, Inc, ASI 00004708 252 Heniminki and Vineis The aim of the review was to assess the correspondence between epidemiologi cal and experimental observations. The rationale was that an analysis restricted to those chemicals for which human data are available would allow an estimation of the proportion of false positive and false negative results in experimental tests. It was soon apparent that this approach had several limitations: 1) The demonstration of lack of effects in man requires an amount and quality of epidemiological evidence which are very difficult to attain: Therefore, the identi fication of chemicals giving false positive findings in laboratory animals is problematic. 2) Reports of an association between exposures and embryotoxic effects in man are not equally convincing for the existence of a causal link. In fact, studies have covered a wide range of objectives, from hypothesis generation to hypothesis confir mation. Therefore, the epidemiological literature had to be thoroughly scrutinized in order to identify those chemicals for which an effect on man could be considered as firmly established. In addition, in many studies observations could lead to more than one interpretation, so that a description was needed of the essential findings in each study. The criteria and the interpretation by the authors are stated. 3) Epidemiological observations (of any quality) regarding teratogenicity fol lowing chemical exposures which have been reported in the scientific literature are very numerous. It was therefore decided to exclude medicaments from the first phase of the exercise. 4) The review was restricted to structural malformations for two reasons. Inclusion in the epidemiological review of studies of other embryotoxic effects would have further increased the size of the review. In addition, the majority of analytical epidemiological studies (ie, those allowing for the recognition of cause-to-effect relationships) are of the case-control type, in which only one effect is taken into consideration. The original objective thus turned out to be too ambitious. The review does not allow conclusions to be drawn about the usefulness of teratogenicity testing in terms of public health (ie, assessment of the predictivity of the results of tests in. laboratory animals), However, it has two elements of interest. On the one hand, it provides a thorough analysis of epidemiological findings with explicit evaluations of their relia bility. In addition, it pinpoints some established (or almost established) causal associ-. ations with teratogenicity in the human species, which require-investigation of their mechanism of action. No conclusion can be drawn on the validity of the practice of testing chemicals for teratogenicity. The many different types of embryotoxicity and the variation of doses and responses in different mammalian species reinforces the view of the Advisory Subgroup in Toxicology that only an increase in knowledge of the mecha nism of action of embryotoxins will lead to sound methods for the prediction of this hazard of human exposure. Advisory Subgroup in Toxicology of the European Medical Research Councils (*) (*)Composiiion of the Subgroup: W.N. Aldridge (Carshalton). E. Cohen (The Hague). D. Henschler (Wurzburg), E. Schenk (Sirasburg), B. Holmstcdt (Stockholm), K. Jager (Geneva), R. Lauwerys (Bmxelles), F. Lembcck (Graz), N. Lery (Lyon). 1. Purchase (Alderley Park). B. Terracini (Torino). . I Teratogenicity Data . 253 INTRODUCTION Tests for teratogenicity and other embryotoxic effects in experimental animals arc commonly used in the belief that they predict risks in humans. Yet, the validity of this approach has not been established; animals species vary considerably in their susceptibility to teratogenic agents 11]. In principle, experimental and epidemiological data should be systematically compared so as to enable the identification--if any--of "false positive" and "false negative" chemicals in experimental tests and thus to estimate the sensitivity and specificity of these. In fact, for only a limited number of environmental chemicals is information on their effects on the human fetus available, and the validity of this information requires a thorough scrutiny. In addition, because of the intrinsic limitations of the epidemiological approach, the identification of "true negatives" is questionable. An alternative, albeit reductive, approach is the identifi cation of "true positives" and the qualitative and quantitative comparison of epide miological and experimental data concerning these. Thus, the present review has two purposes: 1) to evaluate whether chemicals with evidence of teratogenicity in humans show, on a qualitative ground, similar effects in animals, and 2) to evaluate whether human arid animaleffective doses are comparable. A similar although qualitative approach regarding Carcinogenicity of chemicals turned out to be rewarding [2,3], since it revealed that almost all human carcinogens are carcinogenic in experimental animals also (although quantitative comparability is still uncertain). The exercise required a preliminary evaluation of the epidemiological findings regarding the effects of specific exposures on the human fetus. In view of the relatively large number of exposures for which some epidemiological studies or observations on man were available and required scrutiny, some limitations have been introduced: 1) The present review does not consider therapeutic agents; 2) only epidemiological studies regarding congenital anomaly (eg, structural malformations and neurological defects) as an end point are considered (thus excluding birth weight alterations, abortions, still births, and behavioral effects). Yet in experimental studies all expressions of teratogenicity and embryotoxicity detected during pregnancy or at birth are recorded. The limitation of epidemiological data to congenital anomalies is technical rather than scientific; it is realized, however, that in terms of public health, ie, for predicting effects in humans, all embryotoxic effects should be considered [1.4J. METHODS Chemical exposures, other than drugs, for which there are human data on teratogenicity have been identified through Medline and Toxline data banks using key words such as teratogenicity, malformations, and birth defects. Additional sources of literature were some reviews [1,5,6], and files maintained by the authors. No new articles were included after the summer of 1982. The chemical exposures identified from the literature are reported in Table I. It is obvious that for complex exposures and some mixtures comparison with animal data is problematic. The following criteria have been adopted in the evaluation of epidemiologic evidence for each chemical or exposure: A high level of evidence was achieved when information on malformations and exposures was reliable (ie, derived from objective sources), and 1) analytical (ie, controlled ) studies were available and no positive bias 254 Hemminki and Y'ineis TABLE I. Chemicals (Olher Than Drugs) for Which Epidemiologic Data Have Been Evaluated___________ Individual chemicals Carbon monoxide Fluoride (water fluoridation) Halothane Hexachlorophene LSD Lead Methylmercury Nitrous oxide 2,3,5-T" TCDD Vinyl chloride Mixtures, complex exposures Anesthetic gases Coffee Ethanol Gasoline - Cannabis PCBs (Kanechlor) Potato blight Smelting work Soft water Laboratory work (solvents) Spray adhesives Tobacco "Contaminated by TCDD and other dioxins. was identifiable (ie, systematic errors in selection of subjects, collection of informa tion, or comparability of groups); no positive confounder was identifiable (ie, a factor which was predictive of the occurrence of the disease and w_as associated with the exposure being investigated); the association was strong (say, relalive risk greater than two) or, if not strong, it was substantiated in two or more studies; the results were unlikely to be due to chance alone; ie, the statistical power of the study seemed to be sufficient as to detect the reported association; given the purpose of the review, we are not concerned with a low power as a cause of "false negative" results; in addition, several independent studies done under different circumstances make the cause-effect relationship more likely; or alternatively, 2) several case reports showed association between a very rare exposure and a very rare malformation. Limited evidence was achieved when several case reports or uncontrolled clini cal series were available; analytical studies were available and biases, if present, were unlikely to explain all of the reported association, and the association was weak and not shown to be reproducible. A law level of evidence was based on the following; few case reports or clinical studies, geographic and/or chronological correlation studies only, individual analytical studies showing a weak and biased and/or confounded association.. Inadequate evidence was based on studies showing serious bias or serious confounding or on insufficient available information. These criteria are only to be considered as guidelines to make the comparison with animal data easier. The categories of high, limited, low, and inadequate level of evidence do not indicate a different teratogenic potency of chemicals but different qualitative levels of available knowledge. The following criteria were applied while the original literature was reviewed: 1) Regardless of the terms used for the frequency of "malformations in the original papers, we use the term prevalence rate (number of malformed children per 100 live births or deliveries). 2) The literature search considered only maternal exposure during pregnancy. Thus, maternal exposure before conception and paternal exposure were not considered. 3) With some exceptions, individual case reports and geographic or historical correlation studies have been mentioned but not discussed in detail. Case reports may ASI 00004711 I Teratogenicity Data 255 reflect fortuitous observations, since defective children can be more easily reported than normal babies from mothers exposed to a chemical. Geographic correlation studies usually do not allow for the identification of the effect of the relevant agent separately from that of other exposures. 4) The design of some studies (eg, on anesthetic gases) is unusual in epidemiol ogy, since they are based on a cross-sectional selection of (exposed and unexposed) groups at the time of the study, and a retrospective ascertainment of pregnancy outcomes and previous exposures; they are reported as "exposure-selective retrospec tive studies" [7], 5) Consideration of the potential confounding factors in studies on malforma tions is limited. It is reasonable to think that at least maternal age, parity, and socio economic status may be potential confounders [8], since they predict the occurrence of at leasfsome'malformations and are associated with many exposures. Also, alcohol consumption cotrid-confound association measures in some studies. EPIDEMIOLOGIC STDplES Anesthetic Gases^^ - Available analytical studies are summarised in Table II. These studies have been defined as "exposure selective studies" [7] since they were based on cross-sectional selection of exposed and-unexposed groups from professional or employment records or similar files. In general, groups were subsequently refined according to the information obtained in questionnaires. Exposure data were collected in most of the studies through mail questionnaires. Exceptions were the studies of Cohen et al [9], who directly interviewed nurses but not other groups, and of Ericson & Kallen [10] who used hospital records for ascertaining duration of exposure. In most studies, exposure was characterised by duration and period of pregnancy. In two studies, duration and levels of exposure to halothane were respectively scored by the interviewed subjects [11,12]. Cohen et al [13] report malformations of the offspring of chairside assistants exposed to nitrous oxide alone and to mixtures containing nitrous oxide. Lauwerys et al [14] collected information on the different types of anesthetic gases to which physicians and nurses were exposed. Pregnancy outcome was identified through mail questionnaires in all studies but ones in which use was made of a national delivery register [10]. Response rates ranged from 40 to 92%, with 70% as a common value. Among those studies reporting the response rate separately for exposed and control subjects, in two the rate was lower for the unexposed group. Results. Considering only females exposed at least in early pregnancy, the studies of Knill-Jones et al [15], Corbett et al [16], Cohen et al [17], Knill-Jones et al [18] (only when a 1:1 matching of unexposed to exposed subjects was introduced), Gothe and Hoffman [19] (only when nonsmokers were considered), Pharoah et al [20] (only for heart-great vessels defects) and Cohen et al [13] (only for light users of anesthetics) showed.statistically significant differences of prevalence rates between exposed and unexposed subjects. Of these studies, four controlled for potential confounders. Rate ratios can be calculated from these studies and ranged from 1.6 to 3.6 [18] (however, when 1:2 matching was introduced, the rate ratio became 1.1, not ASI 00004713 Finland Retrospective exposureselective study 159 Nurses, 50 female anesthetists, 81 physicians (controls) (respondents: 77% or anesthetists and physicians) 1241 Women anesthetists + 1678 women general physicians (controls) (respondents: exposed 82%, unexposed 80% 300 Married nurses (58 in anesthesiology) (respondents : 70%-75%) Mail questionnaire (all trimesters) Mail questionnaire (duration and level of exposure to | halothanc scored) Wives of male anesthetists: '3/137; female anesthetists: 0/26; anesthetists nurses: 1/229 Interviews to Female anesthetists nurses, mail 2%: physicians questionnaires 4.2% to other groups Mail questionnaire Mall question nairc 46/893 (5.2%) among exposed; 89/1835 (4.9%) among controls Exposed in I-II \ trimesters. 6.5%.I unexposed in I-II f trimester: 2.57o j No malformations Exposure to observed halothanc Corbett et al, U.S.A. 1974 [16] Cohen et al, 1974'[ 17] U.S.A. Knill-Joncs et al, 1975 [181 U.K. Retrospective exposureselective study Retrospective exposureselective study Retrospective exposureselective study 621 Nurse Mail questionnaire Mail 16.4% Among nurses < 0.05 anesthetists questionnaire practicing during Hemangiomas: (respondents: 85%) pregnancy: 12 in the ex 5.7% among nurses posed group vs. I not practicing in the unexposed during pregnancy inguinal hernias: 14 vs. 2 All dpcrjling mom Mail questionnaire Mad Exposed females: \ personnel in U.S.AI (49.585) (I trimester and 1 Questionnaire year before jf (title I' 9.6%; unexposed Jr, p * 0.001 Jj!females: 5.9%: 23,911 concept ion) cm phasizes wives of male 1 pediatricians and health risks of anesthetists: 5.4%; > p < 0.04 nurses (controls) anesthetics) wives of male J (respondents: 41- pediatricians: 4.2% Conf. control 76%) 7,949 Male Mail questionnaire on Mail 4.5% In children of Minor defects: c.r anesthetists (under male and female questionnaire exposed males; 3.09%"; c.f. 65) and controls exposure (I 5.5% in children of 3.19%; u.d. (general doctors trimester) exposed females; 2.35% and anesthetists not 3.6% in children of a = p< 0.02 working in first unexposed doctors trimester of their When matchin/; for Conf. control wives' potential pregnancies) con founders: c.m. p < 0.01--minor (respondents: 70%) 4.5%, unexposed \ delects 3.1% vs mules 3.2%; c.f. J 2 1% 5.5%; unexposed 1 p < 0.0I-/.7 females 1.5% / IllOtcilttlf; 5.3% c.f.; 4.9% p. value n.s.-- u.f. t:2 matching (continuer/) 257 fvirftlitX't? TABLE H. Epidemiologic Evidence on Teratogenicity of Anesthetic Gases* (Continued) Author(s) and year Country Type of study Subjects and sample size Source(s) of information-- exposurcJ Source(s) of information-- outcome Prevalence rates of defects at birth (per live births or pregnancies) Comments' Cohen et al. 1975 [21] U.S.A. Tomlin, 1976 U.K. 1152] Spence ct al. U.K.and 1977 [153] U.S.A. Gbthc and Hoffman, 11 1977 [19] 1' sUcdcn 1i Retrospective 2,798 Male oral Mail questionnaire Mail 4.7% in the exposed p = 0.26 Conf. exposure - surgeons + 4,797 questionnaire group; 4.1% in the control selective male dentists (title unexposed group study (exposed: subjects emphasizes having worked at health effects least 3 hours /week of with anesthetists anesthetics) the year before their wives' pregnancy) (respondents: 39- 65%) Case report Three malformed Types of defects not children from reported "approximately" 50 anesthetists Retrospective Combined analysis of Mail questionnaire Mail 5.5% In female ) exposure- Cohen (1974) and (only exposure in questionnaire anesthetists 4.0% i p < 0.04 selective Knill-Joncs (1972, early pregnancy In female controls 1 study 1975) studies considered) 5.0% In male j p < 0.01 (Greatest (response rate in anesthetists 3.7% \ difference in U.S. [} 'V. iJ ,i,);( 1 l x' Retrospective. All Swedish male U.S male controls: 45%) Mail questionnaire Mill i In male controls j Male anesthetists: data = low response rate) Conf. control Exposure-. anesthetists and questionnaire 1.7% among Jiclcctivc female anesthetist exposed 1.8% Study ( nurses vs. internal ' among unexposed physicians and female nurses: surgical nurses; index group: 1,382 3.0% exposed \ 3.2% unexposed/ smokers pregnancies; control group; 2.002 pregnancies 5.9% exposed 1 2.9% unexposed J nonsmokers-- p < 0.01 (response rate: 89.7%) Pharoah el a!> U.K. 1977 1201 Retrospective exposureselective study 7992 Women Mail questionnaire physicians (respondents: 72%) Mail Anesthetists: heart- questionnaire great vessels 1 13.8%", talipes . i 1* 1 5.2%; other ,1 , physicians': heart- a =* p < 0.05 L 1 great vessels i 3.6%", talipes'! ' 1.7%; National I ;t ' 1' ? 'V'ii'; i[ l Sjj1 i : 1 Child Develop. i! Study: herliit-great vessels 6.6% " i( Rosenberg and Finland Vanttinen, ' 1978 [12] Exposureselective study 2)2 Anesthetists + 356 Pediatricians (response rates: 85% and 64% Mail questionnaire Mail (subjective score of questionnaire halothane vapours) talipes 4.1 % 29/248 Pregnancies (11.6%) among anesthetists 28/266 pregnancies Six luxations of hip and 3 inguinal hernias m anesthetists, vs. (10.5%) among none in the pediatricians controls; but 2 vs. Tomlin, 1978 U.K. 11541 Uncontrolled exposureselective study 75 Anesthetists with 135 children Five musculoskeletal or nervous system defects among 135 children of 11 "other defects" Ependymoma included as a maiform approximately 75 Ericson and Kallcn, 1979 [10J Sweden Exposureselective study 669 Female operating Hospital records. room personnel. (duration of vs. all women exposure in Computerized delivery records (1973 anesthetists 22 Observed defects vs. 26 expected (from 986 children Conf control employed in pregnancy) and 1975) in the exposed medical work group and 19,127 in the unexposed) (ctmlwuetl) 259 TABLE II. Epidemiologic Evidence on Teratogenicity of Anesthetic Gases* (Continued) Author(s) and year Country Type of study Subjects and sample size Source(s) of information- exposure11 Source(s) of information- outcome Prevalence rates of defects at birth (per live births or pregnancies) Comments1* Tomlin, 1979 U.K. |22| Cohen ct al. 19B0 [ 131 U.S.A. Ii ' [ Retrospective exposureselective study 340 Anesthetists: ''exposed" were those spending at least 1/2 of their work time in anesthetic practice (respondents 92.4%) Mail questionnaire Mail questionnaire Retrospective exposureselective study i.| | i 107,771 Dentists 150,000 chatrside assistants (response rates: 70 and 75%) i'f.' Mail questionnaire (nonusers of anesthetics, light users ( = 1-8 hours/ week), heavy users ( = more than 8 h/ wk) in the same year and the year before concept) Mail questionnaire li ' 1 r' 9.3% For exposed in the last 20 years; 7.2% for exposed before; 4.3% for unexposed (children of male and female anesthetists not separately analysed) Excess due to CNS and muscu loskelcla! defects HenochSchonlein purpura. muscular dystrophy and diabetes included as malformations Wives of male Conf. comrot dentists: nonusers 4.9%, light users 4.6%, heavy users 4.8%, Female chairside assistants: nonusers 3.6%", light users 5,7%", a = p < 0.02 heavy users 5.2%, (nitrous oxide alone ('ii P exposure = 5.5%) 1 Exposed chairside 1 , assistants: 2,47%; I musculoskeletal unexposed c,a.: | defects--p < 0.01 1.15% I Lauwerys ei a!, 1981 [14} i Belgium Retrospective exposureselective study 121 Male and 28 Mail questionnaire female anaesthetists ' (duration of + 63 male and 176 . exposure, hours/ female operating day of exposure, theatre nurses vs. types of 62 + 27 anaesthetics, dermatologists, 290 ventilation system) + 22 occupational physicians;, 14+33 intensive c^rc 1 nurses, 20 female social nurses and 168 other doctors and nurses; Response rate: 47%, ranging from 41.5% (intensive care nurses) to 55.5% (social nurses) Mail questionnaire Exposure of mothers in pregnancy: fullterm malformed II 259 pregnancies; ' premature 'malformed 3/259: still-births: I malformcli 0/259. ,NCvcr-cxlposed V #nh4: lju ll-lernx ' ntalformcd 47/ 1,519 pregnancies; premature malformed 5/1.519; still-births, malformed 3/1519 Type of anaesthetics was ascertained. No differences among groups for smoking habits and |drug consumption. !No difference amoiip groups for )inInRefects was '(.tatistically 'significant *c.m., exposed males; e.f., exposed females, u.m., unexposed males; u.f., unexposed females; n.s,, not significant. 'Level and trimester of exposure in brackets, ''Comments include: p-values, types of malformations, specific exposures, serious identifiable bias, and confounding---when confounding factors were controlled in the study this is reported as Conf. control. 262 Hemrninki and Vineis statistically significant) or 3.8 (only for heart-great vessels defects) (20). Other studies [9,11,14,19,21, limited to smokers] showed either no or not statistically significant differences. The negative result obtained by Ericson and Kallen [10] is particularly relevant, since outcome was assesed through an objective source. Types of malformations. Excesses of specific malformations were shown by Corbett et al [16, haemangiomas and inguinal hernias], Knill-Jones et al [18, minor defects], Pharoah et al [20, heart-great vessels defects], Rosenberg & Vantlinen [12, luxation of hip and inguinal hernias], Tomlin [22, CNS and musculoskeletal defects] and Cohen et al [13, musculoskeletal defects]. Minor malformations are more prone to information bias [23], ie, an apparent excess may be due to more accurate reporting by the exposed subjects. This is particularly suggested when unexposed controls show an excess of unspecified "other" defects [12]. In addition, ependymoma, Henoch-Schonlein purpura, muscular dystro phy, diabetes, heart murmur, and malabsorption .syndrome were included among malformations in single studies. In some studies, the purpose of the investigation was mentioned in the question naire title ("health effects of anesthetics"); this may have induced more accurate answers by the exposed groups. Evaluation of anesthetic gases. Due to the aforementioned shortcomings (in consistency of results, nonobjective sources of information, possible recall bias), the available evidence of teratogenicity of anesthetic gases is limited. Nitrous oxide has been associated with an excess of birth defects among female chairside assistants in one study only [13] (see Table II), and the evidence of teratogenicity is low for it. The evidence for halothane is inadequate. --- Caffeine (Coffee) Ingestion of caffeine, mainly from coffee, tea, and cola beverages averages at 190 mg/d per adult population in the US [24]. Five epidemiologic studies on the as sociation between coffee and caffeine consumption during pregnancy and malforma tions in the offspring have been reported and are reviewed beKyw1; "results are sum marized in Table III. Also, other types of reproductive effects haye.been studied [25], In the prospective Boston Collaborative Perinatal Project, over 50,000 pregnant women were interviewed on their consumption of drugs [26}r"Over 5,000 women reported use of caffeine, usually as an ingredient in medication. "No specific type of malformation was related to caffeine intake. Yet, coffee drinking was not controlled - for and the authors recommend caution in interpretation of results. In a case-control study reported by Borlee and co-workers [27], the parents of 202 case children and of 175 control children were interviewed on the types of beverages consumed. Proportions of mothers drinking more than 8 cups of coffee per day were 23% among cases and 13% among controls (p < 0.05). The association did not involve specific types of malformations. The frequent coffee drinkers ap peared to be older than the women drinking less coffee. Teratogenicity Rata 263 TABLE III. Epidemiologic Evidence on Teratogenicity of Coffee Consumption* Information on coffee consumption Prospective I Sample size 50,300 Births Retrospective I- 490 Cases Retrospective l 2,030 Cases . - - ^.2- -Ir ~ - .-fe' Retrospective I 12,205 -__Women Retrospective I 482 Cases Malformations All All 380 Inguinal hernia, 299 cleft lip palate, 277 cardiac, 194 pyloric stenosis, 120 cleft palate, 101 CNS (I) All 83 CNS, 150 skeleton, 95 cardiovascular, 154 oral clefts Risk ratio (consumers vs. nonconsumers) - 1 (Use of caffeine medication) Significance (P) 1.5 (^ 8 Cups/ day) p < 0.05 Pyloric stenosis: 95% c.l.: 1.8 (> 200 mg 0.8-3.9 caffeine/day) -1.0 - 1.0 Reference Heinoncn et al, 1977 [26] Borlec et al, 1978 127] Rosenberg et al, 1982 [28] Linn et al, 1982 129] Kurppa et al, 1982 [30] *1, interview; Q, mail questionnaire; (I), for each group of defects, 712 other malformations were used as a comparison group. In a case-control study, 380 infants with inguinal hernia, 299 with cleft lip with or without cleft palate (CL + CP), 277 with cardiac defects, 194 with pyloric stenosis, 120 with isolated cleft palate, and 101 with neural tube defects were compared with 712 other malformed infants (control group). Maternal ingestion during pregnancy of lea, coffee, and cola, was investigated through personal interviews. Relative risks ranged up to 1.8 (0.8-3.9) for pyloric stenosis (total daily caffeine intake greater than 200 mg vs, no ingestion). The corresponding estimates for inguinal hernia, CLCP, cardiac defect, deft palate, and neural tube defect were 1.4, 1.3, 1.5, 0.8, and 0.8, respectively. None of the point estimates of relative risks was significantly greater than unity and no dose-response relationship was apparent [28] (Table III). Linn and co-workers [29] carried out a retrospective study on 14,500 women, among which compliance was 90%. There was no association between malformations in the offspring and coffee consumption even in the highest exposure category (> 4 cups of coffee per day). Finally, the mothers of 482 children with central nervous system, skeletal, cardiovascular, and oral cleft malformations from the Finnish Reg ister of Congenital Malformations were interviewed and compared to referent mothers of healthy children. No difference was found in the consumption of coffee [30]. The evidence oil the teratogenicity of coffee or caffeine intake is low. The one study reporting an effect does not rule out the results of other negative studies. 264 Hemntinki and Vineis Carbon Monoxide At least eight cases of documented maternal intoxication from coal gas used for illumination were associated with retarded mental or psychomotor development in the offspring [4,31,32]. In three of these cases autopsies were performed and showed softening of basal ganglia, cerebral atrophy, hydrocephalus, agd^picrocephalus [32). In nine other case reports [4], various types of brain pathology, including damage of the basal ganglia and microcephaly, were detected. The evidence is limited for a direct effect of carbon monoxide in inducing structural changes of the CNS in the offspring of exposed women. Ethanol Previous reviews indicated a causative role of alcohol in the "fetal alcohol syndrome, FAS" [33,34] (Table IV). The features of this syndrome are not unequiv ocal [35-37]; Sokol [38] has proposed that a reliable diagnosis should include at least one typical feature of each of three groups of defects: growth retardation, CNS effects, and facial characteristics. The presence of FAS in live born children in France, Sweden, and the US ranged between 1/600 and 1/1000 [37,39a,b,40]; it doubled when milder syndromes were included. Prevalenceirates of 19% were reported among children born from mothers consuming 56 oisrnore grams alcohol daily [37]; elsewhere, rates were in the order of 35% in children ofchronic alcoholics [40,41], In these studies, the quantity of alcohol consumptio'n~was estimated by interview and was probably subject to a.reporting bias. In three series of FAS children the prevalence of microcephaly, oral clefts, and heart defects were respectively 40-90%, 1-25%, and 25-50% [42^44], ie, greatly in excess of the corresponding rates in the general population. When considering the studies including information about alcohol consumption and the presence of major structural malformations, the association is less clear. In a prospective study, there was an increase in the frequency of major malformations (32 vs. 9% in the control group) in the offspring of heavy drinkers consuming a mean daily dose of 2,2 g alcohol/kg [45], Microcephaly was the main type of malformation noted, and it explained most of the difference between the groups. Hypospadias and genito-urinary and oral malformations were more common in infants of alcohol abusers as compared to the controls [39a,b]. No excess oral clefts and cardiovascular defects have been noted in a study of Tennes and Blackard [46]. --=- There is a high evidence of a teratogenic effect of alcofl&lism in humans in so far as it induces FAS. A causal relationship between maternal^alcohol consumption and other structural malformations in the offspring is less weTfestablished. Gasoline ^ Mental retardation, scaphocephaly, prominent occiput, poor head growth, and minor anomalies were found in two children whose mothers_.tcportedjnhalation of gasoline. Four children with a similar syndrome occurred in 1973-771n Shamattarere, Canada, but gasoline sniffing could not be ascertained The evidence is inadequate. Hexachlorophene Hailing [48] compared the occurrence of malformationsdn six groups of nurses using 1% hexachlorophene soap and in six hospital-matched groups not using the Teratogenicity Data 265 soap. Pregnancy outcomes were retrieved from the records of obstetric departments; the source of the exposure data was not stated. Exposure consisted of 20-60 washings per day, both before and during the first trimester of pregnancy. Loss to foliow-up was 15% in three hospitals. Major malformations were 0/233 live births of the controls {vs. 4-'5 expected) and 25/460 live births of exposed women. There was no adjustment forage and.parity, but it is unlikely that confounding could explain such a difference. " Baltzar et al [49] reported two studies. In the first study, 63 hospitals were asked to provide lists of women in medical occupations who had left work because of childbirthJnJ965-75. Only 31 hospitals replied and provided data on approximately 1,500 women. All-.^pmen had worked at least in the later part of pregnancy. Malformed children were identified through the Swedish Register of Congenital Malformations-in-196r72, and from the Medical Birth Record Register (MBRR) thereafter. Results-are-given separately for the 1965-72 period (5 malformed children born from 930 worfiSf^vs. 9-10 expected), and for the 1973-75 period. In 1973-75, observations- are reported separately for 200 women from three hospitals in the Gothenburg "area andjor 260 women from other hospitals. Observed and expected malformations were respectively 13 vs. 4 in the former group and 3 vs. 5 in the latter (this observation of an excess in the Gothenburg area could not be referred to any particular exposure). The second study reported by Baltzar et al [49] describes the outcome of 29,806 pregnancies among women employed in medical occupations in 1973-75. These women, identified from a register of medical personnel, were linked with the birth register through the personal identification number. Malformations were collected through the Medical Birth Record Register. Expected numbers were calculated from all deliveries in Sweden in 1973-75 (323,011 infants). A total of 1,551 defective children were born in the study group, vs, 1,491 expected; an excess in Gothenburg was confirmed (42 observed, 29 expected; p < 0.01). A comparison between the use of hexachloropherie "soaps in "periods presenting possibilities for exposure during pregnancy" and malformations in the offspring revealed no apparent association. Some aspects in these two studies reported by Baltzar et al require comments: exposure was ill-defined; in the first study only 31/63 hospitals responded, and the quality of information from these hospitals was probably different; there was no information on individual exposures to hexachlorophcne. There is a low level of evidence about hexachlorophene teratogenicity in man. Lead The data on the teratogenicity of lead to humans are historical and cannot be evaluated critically [50]. A case report on the association of lead intake in moonshine whiskey and fetal anomalies has been published, but it may also be a result of alcohol intake [51], No epidemiological evidence is available to evaluate the possible terato genicity of lead to humans. LSD and Cannabis Several case reports have been published regarding children bom to mothers taking LSD during or before pregnancy [52-55]. Most of them describe similar skeletal defects: a common feature was distal agenesis of limbs. In these reports the LSD doses were not estimated, but it is stated that they were sufficient for psychedelic TABLE tV. Epidemiological Evidence on Teratogenicity of Ethanol (Alcoholic Beverages)' Information on exposure______ Sample size Malformations Exposure level Risk ratio and statistical significance Reference Retrospective, I Retrospective, I (clinical observations) Prospective, I Retrospective, I (clinical observations) Prospective, I 127 Cases from alco holic mothers 11 Cases with "fetal alcohol syndrome" (FAS) 23 Chronic alcoholic women + 46 matched controls 6800 Births Facial characteristics, growth deficiency, psy chomotor disturbances Short palpebral fissures (11), microcephaly (10), maxillary hypoplasia (7), limb anomalies (8), al tered palmar crease pat tern (8), cardiac anomalies (7) 6 Babies with FAS in the alcoholic group (includ ing short palpebral fis sures, ptosis, strabismus, joint anomalies) Babies with FAS 633 Women "No specific pattern of anomalies" Prospective. 1 ill, i |H Ml Ml '! I1 1 70 Infants born from "alcoholic" moth- , 1 ers + 93!''con trol" infahts r} A priori definition of "fea tures compatible with FAS" ii -- -- Lemoineetal, 1967 ]35] Jones and Smith, 1975 and 1973 [33,36] Jones et al, 1976 [41] Severe cases 1/1000, less Severe 1/400 Dehaene et al, 1977 [39a] Nondrinkers; 326 women (I), C45 ml absolute alcohol daily: 249 (II), ^ 45,., ml daily averagei(a(M solute alcohol); 58 : (HI) i ' "Alcoholic" mothers: one ounce or'morc of absolute alcohol/ day or 5 + drinks on occasion 9% Congenital anomalies in Group I 11 % in Group II 32% in Group III (p < 0,001) 9/70 Infants with FAS from heavy drinkers (JI.O ounccs/day) 0/31 from moderate drinkers(0.1-0.9) 2/62 in control infants Ouellette et al, 1977 [45] i 1' Hanson et a|, 1978 [37] Retrospective for 2 infants, prospective for the other 2; 1 Prospective, I Retrospective 4 Dead infants Different features of brain malformation at neuropathology 23 Pregnancies, 2! live births 204 Abusers 11,923 controls FAS and other anotnalies v; i i>' ,171 i All1 `.i J , Prospective, Q 278 Births All Retrospective 109 Infants with FAS All Retrospective 71 Infants with FAS Alt *1, Interview; q, questionnaire; FAS, fetal alcohol syndrome. "Heavy ethanol in take" in pregnancy (2 women) + occa sional "binge" (2 women) 1 1 [! "alcohol problem" 7/21 FAS, other 7/21 thought to have al cohol-related ' i i . 11 pt'pblenii , Alcohol abuse w^ de FA^t2.5 iln abu- fined "clinica'lly" 1 * ' seri vs. 0% in con trol), hypospadias, genito-urinary and oral malformations increased 5 Exposure categories No excess risk during I and III trimester Anomalies in FAS infants: microccphalus (84%), var ious facia! anomalies (1174%) Anomalies in FAS infants; microccphalus (44%), various facial anomalies (7-68%) Clarren et ai, 1978 [42] Olegard et al, 1979 [40] , ,i ' ii | , ^kol^tal. 1980 [39bJ Tennes and Blackard, 1980 [46] Majewski, 1981 [44] Nestlcretal, 1981 |43| t 268 Hemminki and Vineis effects. LSD was either taken throughout pregnancy or in the first trimester. In at least three reports, there were multiple exposures (cannabis, drugs, or other strong hallucinogens). A further report [56] describes a baby with multiple malformations, including absence of phalanges, born to a mother who had taken 3 doses 9 months prior to conception. Finally, another report describes the birth of a normal baby from a woman who had been a high-level consumer up to 2 months before pregnancy [57]. No malformations were described in a follow-up study on ten pregnant women with an intake of LSD which entailed hallucinatory phenomena [58]. All women took LSD in the first trimester of pregnancy and some also later. Another uncontrolled follow-up study [59] included 247/300 women taking LSD and other drugs and responding to a questionnaire: during follow-up 148 of them had pregnancies (120 live births) and 14 malformations were detected, including six tumed-in feet, two tibial rotations, and one Unspecified bone deformity of legs. No birth defect was observed among the 12 women taking LSD during pregnancy. The median single dose for mothers of malformed babies was 100 gg (range 75-250 ng). Overall, the level of evidence for LSD is low since the positive findings derive from case reports, most of which describe multiple drug intake. However, the evidence cannot be entirely disregarded, because distal limb agenesis, which has been described in the offspring of six mothers taking LSD, is otherwise observed rarely. The epidemiologic literature on teratogenic effects of cannabis is scanty and overlaps with that on LSD. In two case reports [53,54], the mothers of babies with distal defects of limbs consumed both cannabis and LSD. In a further report [56] and in one follow-up study [59] concerning mainly LSD, cannabis is mentioned as one of several drugs concurrently consumed by the mothers. The level of evidence for cannabis is inadequate. Methylmercury Knowledge on the cmbryotoxicity of methylmercury derived from-studies re garding the ingestion of contaminated food [reviews by Koos and Longo, 60, Chang et al, 61]. These studies include the well-known epidemic in Minamata, Japan [62]. In the latter population severe-neurological disorders irkadults. ingesting methylmercury-contaminated fish were.descrihcd as early as in _1953. However, only a few years later children affected by symptoms resembling those of cerebral palsy were described. Between 1953 and 1971 at least 25 children were born with brain damage. After systematic research, a syndrome could be observed and this was also connected to alkylmercury ingested in contaminated fish by the mothers. A comparable epidemic of microcephaly and other neurological disorders took place in Iraq in 1972, as a result of ingestion of grain treated with methylmercury [63,64], Out of 32 infants exposed in utero, ten showed cerebral palsy at the age of 1 year. In Iraq as well as in Minamata, mercury was found in biological samples from the affected mothers and children. A dose-response relationship has been established [65], A total maternal exposure to 1-1.2 mg/kg corresponded to a 50% frequency of neurological signs in the offspring [65], In Iraq the total ingested doses were about 150 mg and the daily exposures about 80 /rg/kg, with large variations. Observations from Minamata and Iraq have also been confirmed in other small scries of cases from ASI 00004725 Teratogenicity Data 269 U.S.S.R., Sweden, and the U.S. [60,61,66]. Neuropathology of prenatal intoxication revealed the following in two cases from Matsumoto et al [67] and in two cases from Choi et al [68]: narrowing of gyri, hypoplasia of the basal ganglia and cerebellum, and heterotopic neurons in cerebral and cerebellar white matter. Methylmercury can be considered a teratogen with a high level of evidence, taking into account the morphologically detectable changes in the CNS. Kanechlor (PCBs) A "PCB-induced fetopathy" (Yusho) has been described as a feature of an episode of intoxication occurring in 1968 in Japan, in which approximately 1,000 persons of different ages were accidentally intoxicated by rice oil contaminated with Kanechlor 400, a technical tetrachlorobiphenyl product. The Yusho oil contained about 1,000 ppm of PCBs and 5 ppm of polychlorinated dibenzofurans [69]. The -etiologic role of the ingestion of this oil was ascertained for intoxicated persons through the following: 1) examination of oil containers, shipping records of the producing company, and records of oil dealers; 2) analysis of biological samples; 3) a case-control study on 121 adult patients and 121 adult controls concerning use of oil and fats [70]. . _ A fetotoxic effect of Kanechlor was first recognised when a group pf intoxicated mothers delivered 13 babies with a syndrome which included dark brown pigmenta tion of the skin and mucous membranes, retardation of growth, gingival hyperplasia, spotted calcification of the skull [70-72], Hyperkeratosis, atrophy of the skin, and cystic dilatation of hair follicles were described in stillborn babies delivered by exposed mothers [70,71], PCBs were detected in the skin by chromatography. One, five, and one females had been exposed respectively in the first, second, and third trimester only, two had been exposed throughout pregnancy, and the time of exposure was unknown for four. Miller [71] estimated an average exposure of 11 weeks. Intake was estimated to be 65-839 ng PCB/kg/day [73]. There is a high level of evidence that the contaminants found in the food oil are human teratogens. Whether PCBs, dibenzofurans, or possibly dioxins were the causative agents cannot be resolved. Potato Blight Renwick [74,75] reported a positive geographic correlation between potato blight and anencephaly-spina bifida (ASB). Subsequently to this, a number of geo graphical and temporal correlation studies have been undertaken; by and large they failed to confirm Renwick's findings [76-85]. Lorber et al [86] reported two cases of ASB bom to mothers on a potatoavoidance diet during pregnancy. A few analytical studies are available. Robert et al [87] interviewed mothers of 336 malformed babies (neural tube defects) and as many normal controls in Glamorghan. South Wales. Cases and controls were born in 1964-66 and their mothers were approached through a questionnaire several years later. One-third of both cases and controls did not reply or were not traced. Home-grown potato consumption did not differ among cases and controls. In another case-control study on potato consumption in pregnancy [88], 83 children with spina bifida, collected retrospectively up to 12 years before the study, were compared with 85 children chosen from the general population and matched to 270 Hemminki and Vineis the cases by sex, month, and year of birth, residence and social class. Mothers were interviewed by health visitors. No difference in maternal potato consumption during pregnancy was found between cases and controls. Potential confounders like maternal age and parity were not considered. In this study, a negative confounding may have been introduced by strictly matching for residence and date of birth, since blighted potato consumption varies according to social class, place, and time. In another case-control study in New York City [89], 265 mothers of babies with ASB diagnosed in 1970-72 and 549 control mothers agreed to be interviewed by telephone on their current potato consumption. Controls (2 per case) were chosen from birth certificate files and were matched to cases by parental residence and date of birth. No relative risks were reported, but they can be computed from the published figures. Most of them are lower than one. The study, however, had some shortcom ings: cases represented 70% of those who were eligible for interview, whereas the corresponding proportion for controls was 94%. In addition, although nonrespondent controls were replaced, the matched controls were not always dropped when the case refused the interview. Thus, a comparison bias (higher social class selection of respondent cases) cannot be excluded. Finally, the analysis did not take into account maternal age and parity. However, bias and confounding so strong as to mask a positive association are unlikely. Nevin and Merrett [90] reported on 88 pregnant women who had~a previous infant with ASB and who were advised to avoid eating potatoes in subsequent pregnancies. The advice was followed by 27 women who were compared with the other 61. ASB occurred respectively in 2/27 (8.7%) and in 2/61 (3.6%) (p > 0.05). This study leaves some open questions: The 88 women derived from a larger group but mechanisms of selection are not stated; inclusion of women in the groups (polatofree or not) was not random and confounding cannot be ruled out; numbers are small. The available studies do not indicate a role of potato blight in the etiology of ASB. Correlation studies were negative. Analytical studies concern potato (and not specifically blighted potato) consumption or avoidance; none of them suggested any association with ASB. Shortcomings in the design of these studies cast some doubt on the acceptability of such negative results but do not provide any suggestion of a positive association. Smelter Work In a recent study [91], a list of women employed as copper smelters in a company in 1955-1976 was compared with the records of pregnancies of-the local hospital. The completeness of either company records or follow-up is not discussed. Malformations were 17/291 among the offspring of women in the smelter as compared to 22/1,000 in controls (p < 0.005). Multiple malformations contributed to this difference. Sulphur dioxide, lead, arsenic, and cadmium have been suggested as the possible causative agents [92,93]. The results were not controlled for possible con founding factors. This individual study provides a low level of evidence on the teratogenic hazards of copper smelting. Soft Water Fedrick [94] reported a negative correlation between water hardness and prev alence rates of anencephaly in ten areas in the U.K. This contrasts with a similar correlation study [95] in Liverpool and south-west Lancashire where the highest ASX 00004727 | Teratogenicity Data 271 incidence rate of anencephalus (3.9/1,000/year) was found in a "hard water" area, and the lowest rate (2.9/ 1,000/year) in a "soft water" area. Lowe et al [96] correlated water hardness with a) CNS malformations (mortality data) in 48 South Wales areas and b) anencepahly in 58 country boroughs in England and Wales. Correlation coefficients were all negative. Analogously, negative correla tion coefficients were found by Crawford et al [97] comparing water hardness and anencephaly mortality in 58 English large towns (many comparisons were made and most correlation coefficients were statistically significant). Morton et al [98] carried out similar work in South Wales considering CNS malformations. The correlation coefficients were negative but low and statistical significance was weak (p < 0.05). Multiple comparisons were made. A case-control investigation in Cardiff included 108 cases of neural tube defects and as many controls matched for sex, age, and social class. Determination of 11 elements in the drinking water yielded similar results for cases and controls [99], Another case-control study in Canada included 468 deaths from anencephalus and a random sample of 4,129 live births [100]. Mothers lived in 142 localities for which the water concentration of 14 elements was available. No association was found for any of the 14 dements. Risk estimates were adjusted for potential confounding factors (parity, season'aTid year of birth, maternal age, and others). However, samples were not collected from case and control dwellings. Overall, these studies provide inadequate evidence on the teratogenicity of soft water. Laboratory Work (Solvents) Five studies suggest an association with teratogenicity (Table V). Kucera [101] reported that five out of nine mothers of infants with sacral agenesis had been exposed to miscellaneous industrial solvents, as ascertained retro spectively. No control group was used. Other malformations were not analysed. A Swedish group has published three studies on laboratory personnel possibly exposed to solvents for whom other exposures have not been ruled out. Meirik et al [102] identified from the university files the laboratory staff who had been working during pregnancy. They had given birth to a total of 245 newborn children. Malfor mations were recognised both through interviews and from the Swedish Malformation Register, Eleven children with serious malformations were found vs. four expected on the basis of the rates among deliveries in the same hospital birth wards, adjusting for maternal age and parity. The 11 malformations included four anal or oesophageal atresias or stenosis, and two children with pes equino-varus. The second study was similar in design but regarded offspring of women working in chemical laboratories of the pharmaceutical industry [103]. In the exposed group, prevalence of major malformations was 6/106 (5.8%) vs. 2/297 (0.7%) in the referent group. The malformations included two oral clefts and one multiple gut atresia. The exposed-referent comparison was uncontrolled for the potential con founding factors. It is reported, however, that the age distribution was similar in the two groups. The third study was similarly designed and covered laboratory workers of pulp and paper industry. Major malformations were 6 vs. 2.9 expected, standardized for age, parity, and hospital" [ 104]. Malformations included two gut atresias, two heart defects, one anencephaly, and one cleft palate. 1 TABLE V. Epidemiologic Evidence on Teratogenicity of Laboratory Work* Information on exposure Retrospective Q Sample size 9 Cases Malformations Sacral agenesis Retrospective (Uni versity lites + interview) 322 Women All Retrospective I, Q 405 Women All Retrospective, Q Retrospective, I 890 Women ! ` ' ''120 Cases | All CNS defects I, interview; Q, mail questionnaire. Risk ratio and significance (p) 5 Mothers exposed to or ganic solvents 4 Cases of anal or oesopha geal atresia in women working vs. none in. women not working in a laboratory during pregnancy 6 Major malformations out of 103 infants among "chemical laboratory workers" vs. 2/297 among the nonchcmical workers. One multiple gut atroid in the former group 6 Major malformations (vs. 2 9 expected) among women working in labora tories (2 gut atresia) RR = b.5 (p <0.01) for ex posure to organic solvents Reference Kuccra, 196011011 Mcirik ct al, 1979 |I02! Hansson ct at, 1980 11031 Blomqvist ct al, 1981 1104) 1 Hotmbcrg. 1979; Holmberg and Nurmincn 1105,1061 . Teratogenicity Data 273 In the three studies, ascertainment of both employment and malformation data were reliable. Even though small numbers of malformations could be collected in each study, the combined studies provide strong evidence for an excess of oral clefts and gastro-intestinal atresias over the rest of the population. The studies were not designed in order to identify the causal factors. These are thought to be various solvents by the authors. A Finnish case-control study has identified exposures to various organic solvents to be overpresented (14 vs. 3) among the case-mothers as compared to the controlmothers from the same maternity welfare district [105,106]. An excess of exposure to dusts was also reported. The study included 120 nervous system malformations reported to the Finnish Register of Congenital Malformations and a similar number of controls. Information on exposures was collected retrospectively in a personal interview. The evidence-on the teratogenicity associated with certain types of laboratory work is high. The evidence for solvents being the causal factor remains unestablished. Spray Adhesives Of the two studies reported, one is in abstract and gives insufficient information [107]. The other study reports use of spray adhesives more often in the homes of cases as compared to those of controls; it is limited because of very low returns in the postal interview [108]. Only 22% (N = 56) of the initial pairs could be subjected to analysis. Thus the evidence on teratogenic effects of spray adhesives is inadequate. The chemical composition of spray adhesives needs to be clarified. 2,4,5-T and TODD It is not possible to analyse separately exposure to dioxins from exposure to other chlorinated chemicals (such as 2,4.5-T) contaminated with dioxins. However, an effort has been made to specify for each study the agent to which people were mainly exposed. Case reports and informal investigations. This section is mainly based on the review by Young et al [109], Tung et al [110] described .four pregnant women who had been hit by herbicide sprays in Vietnam: two mothers had children with Down's syndrome. Other anecdotical reports about Vietnam have been published by Rose and Rose [111] and Honoroff [112]. A survey was conducted by a Committee of the U.S. National Academy of Science [113] in three Saigon hospitals, but the material was considered inadequate, and no conclusions could be reached. Roan and Morgan [114] carried out an epidemiologic study based on the hospital records in the areas near Globe, Arizona, where 2,4,5-T had been massively sprayed. Analytical measurements in tissues and fluids were all negative, and allegedly no trends in reproductive mortality and morbidity could be attributed to herbicide use. Two cases of malformations in human infants were described in Lapland and attrib uted to phenoxy herbicide spraying [115], In New Zealand, .a report by Sare and Forbes [116] described two babies with spina bifida, born to mothers having been exposed to 2,4,5-T spraying during pregnancy. The degree of exposure was not known, and in one case exposure occurred 274 Hemminki and Vineis after the neural tube closed [these comments were made by a subcommittee of the N.Z. Agricultural Chemicals Board: 117]. Geographic pathology studies. Cutting et al [118] surveyed retrospectively malformations occurring in 22 hospitals in South Vietnam, as reported on the daily summaries prepared by midwives. Private clinics were not surveyed^ some hospital records had been destroyed, home deliveries were excluded, as was the Chinese group of the population. Prevalence rates were 5.5/1,000 in 1960-65 (period of light spraying of Agent Orange) and 4.5/1,000 in 1966-69 (heavy spraying). Overall, 2,355 defects in 488,852 births were identified. Anencephaly, cleft lip/palate',,club foot, and hydrocephaly accounted for 80% of all malformations. There was no nonexposed category and only a dose-response relationship could be measured. It has also been pointed out [109] that heavily sprayed areas were underrepresented and there was no precise knowledge of exposure of pregnant women. Two geographic pathology studies on 2,4,5-T spraying in New Zealand have been reported. McQueen et al [119] collected cases of anencephaly and spina bifida in three areas where clusters of cases were allegedly present. Cases were collected in 1974-76 from hospital records, and a comparison was made between observed and expected numbers in each area. In Northland there were seven cases, which corre spond to the expected value. Five cases occurred in Taranaki, of which three were in the same street. In Waikato there were eight cases (one anencephaly, seven spina bifida), all diagnosed in 1 month. The mothers were interviewed about 2,4,5-T contacts in pregnancy: 3/7 mothers of cases in Northland, 3/4 mothers of cases in Taranaki (one is excluded because she conceived elsewhere) and possibly 4/8 mothers of cases in Waikato declared to have conceived during the period of 2.4,5-T use. However, the authors state that only in one or two cases was there "some exposure" to 2,4,5-T in the critical period, and they concluded that there is no evidence. This is not to be considered an analytical study, since it lacks controls and standardized interviews. Hanify et al [120] compared prevalence rates of malformations in seven New' Zealand areas in 1960-66 (when no exposure to 2,4,5-T occurred) and in 1972-76 (when 2,4,5-T spraying was common). Quantity, dates, and places of spraying were collected from the companies. No geographic or temporal correlation could be found, with the possible exceptions of talipes and hypospadias+epispadias (no mention is made of the changing attitudes in the registration of birth defects). Other correlation studies relating to 2,4,5-T spraying in agriculture have been reported in Australia [121], Arkansas [122], and Hungary [123].'These studies concentrated respectively on anencephaly and meningomyelocele in 1965-76, cleft lip/palate in 1943-74 and cleft lip/palate, cystic kidney, anencephaly, and spina bifida in 1970-76. They were all negative. A further study of geographic and time correla tions in Western Australia suggested a role of "insecticides" and""herbicidcs" in the occurrence of cleft lip/palate, but no firm conclusion could be reached [124]. A Birth Defect Registry has been established as a surveillance system in the Seveso area, which was highly polluted by 2,3,7,8-TCDD in 1976 [125]. It collects malformations (all types) in II towns (222,000 residents) from the lists of live and still-births: this is followed by home visits of paediatricians. From 1978 to 1980, Teratogenicity Data 275 newborns contacted increased from 86 to 94%. The Registry area was divided into four zones: A, B, R, and Out (allegedly not polluted) depending on the TCDD concentrations in the soil. Rates were compared with Lombardy hospital records and other birth defect registries. High rate ratios (RRs) were found, in comparison with Lombardy hospital records, for spina bifida (RR = 2, p = 0.1) and hypospadias (RR = 13.5, p < 0.0001). In addition, the RR for "other anomalies" in the zone A + B vs. R + Out was 6.6. These preliminary results must be interpreted cautiously: an excess of-hypospadia and "other anomalies" in the A + B zones could result from a detection bias due,to greater diagnostic accuracy and reporting in highly polluted areas. - Analytical studies. The only available studies concern paternal exposures. Smith et al [126] published a preliminary report of reproductive outcomes among male pesticide applicators using 2,4,5-T in New Zealand. They were postally interviewed about exposure, "'whereas pregnancy outcomes were obtained from their wives. Twenty-four malforrgations (2.0%) were found in children of the exposed group, and 18 (1.6%) in childreiipf the unexposed group. The RR was 1.25 (90% c.l. 0.8-2.0). Four heart defects found in the exposed group, vs. I among the unexposed. In a further paper [127], these data are reanalysed, dividing the exposed group into men who were not spraying during their wives' pregnancies, those who sprayed chemicals other than 2,4,"5-T, and those who sprayed 2,4,5-T during the pregnancy: The rates of congenital defects were respectively 9/401 births, 9/113 births, and 13/486 births (the differences were not statistically significant). Townsend et al [128] described a cohort study on wives of 930 employees of Dow Chemical (Michigan Division), working at least 1 month in 1939-75 in chloro- phenol processes. These were considered "potentially" exposed to dioxins. The same number of men working at least 1 month in the same plant with no chlorophenol exposure formed the control group. The compliance rates were low (62-63%). For all congenital malformations, the crude and age-adjusted rate ratios were 0.96 and 0.85, respectively, when exposure to any dioxin w'as considered. Ratios were 1.18 and 1.08 when exposure to TCDD alone was considered. This study does not report observations separately for an unspecified number of workers developing chloracne following an accident in 1963. The available analytical studies on 2,4,5-T and TCDD refer only to paternal exposures. Evidence of teratogenicity of maternal exposure is inadequaie. Tobacco The possible teratogenicity of cigarette smoking has been investigated in several well-designed studies as reviewed by Landesman-Dwyer and Emanuel [129] and Johnston [130]. The designs and results of recent studies are summarised in Table VI. Fedrick et al [131] reported an increase in congenital heart defects among smokers in material based on 17,000 pregnancies of the British Perinatal Mortality Survey. The findings were not confirmed by several other studies [26,132-135]. In some studies a small effect (less than 1.5-fold) was observed [136], In some investigations, heavy smoking was associated with the occurrence of central nervous system (CNS) malfor mations [26,133,i34]~but the effect has been considered modest at most. In some ASI 00004732 TABLE VI. Effect of Smoking During Pregnancy on the Malformations in the Offspring: Epidemiologic Evidence* Information on smoking Sample size Malformations No. cigarettes smoked Risk ratio s moke r/non-smokcr Significance (p> Retrospective Q Retrospective I Retrospective 1 16,200 Births 16,000 Births 18,600 Births All Heart All Oral clefts > 1/d > 1/d > I/d <1 1.5 1.2 2 No <0.001 No <0.05 Prospective? 14,600 Births Heart ~ i No Prospective I 50,300 Births All > I/d < i No Retrospective I Retrospective Q Prospective I 1,400 Cases 10,500 Births 66 Cases All All Oral clefts > 1/d > 20/d > 1/d > 1/d i.i 1.6 ~ 1.5 1.7 No <0.05 <0.05 <0.01 Retrospective 1 Prospective I ,1 Retrospective . 66 Cases CNS 69,000 Births All Anencdphalia 14,700 Births' ! ! 320 Cases i 1 j Aii 1 ; c!ns 420 Cases j. j Ora! clefts i20 Cases i' Musculo- skeletal *1, interview, Q, mailed questionnaire > 1/d > 1/d >20/d > I/d >20/d ^ 1/d l/d ^ 1/d ~1 -1 1.8 ~1 ~'Vi 2 1.3 1.3 i t No <0.5 No <0.05 No No No Reference Underwood et a!.. 1965 [155] Fed rick ct al, 1971 [131J Andrews and McGarry, 1972 [137] Ycrushalmy 1973 |I32[ Heinoncn ct al. 1977 126J Kelsey ctal, 1978 H33| Himmclbcrgcr ct al, 1978 11361 Ericson et al, 1979 [1381- Evans et al. 1979 [134J ' Christianson, 1980 I135J t Hemminki ct al. 1983 [139] Teratogenicity Data 277 studies an association between maternal smoking and oral clefts has been observed [133,137,138], but again this has not been confirmed [26,134,139]. Taken together, the literature on the effects of maternal smoking on congenital malformations in the offspring fails to show any consistent strong effect. The modest association seen in several studies between heavy smoking (> 20 cigarettes/d) and central nervous system malformations may show a weak effect or confounding. The designs of the positive studies have not been superior to those of the negative studies. " The evidenceis low. Vinyl Chloride ~ ~' Infante [140] analysed birth certificates in 1970-73 in a group of Ohio commu nities, three of which had at least one vinyl chloride polymerisation plant in operation. Prevalence-rates of birth defects (per 1,000 live .births) were as follows: 10.14 in the entire .Ohio .state, 17.37 in Ashtabula, .18.10 in Painesvile, and 20.33 in Avon Lake (the three communities with PVC plants). In the three communities, the excess was statistically significant (p < 0.01). However," similar prevalence rates were found also in two othe'r (Sties "withourpVC plants. A major contribution of CNS defects to the excess was observed in Painesville (with PVC facilities) and in North Ridgeville (without PVC facilities). On this basis, a case-control study was conducted by Edmonds, et al [141] in Painesville. All babies born with CNS malformations in the Painesville hospital in 1970-74 (six with anencephaly and nine with spina bifida) were compared with a group of 30 control babies (the first bom immediately before and the first immediately after the case). Parents of the cases were interviewed, whereas information for the controls was collected from the clinical records. None of the 15 case parents and two of the 30 control parents worked in PVC plants; more controls than cases had mothers working within 10 miles from Painesville's PVC plant. Another case-control study was conducted by Edmonds et al [ 142] in Kanawha County (West Virginia, U.S.): All but one incident cases of CNS malformations occurring in 1970-74 were compared with a group of control babies identified as in the previous study and matched to the cases by paternal education, maternal age, month of birth, race, and social status. Parents were interviewed by telephone on their work in PVC manufacturing plants and residence place. Two out of 41 fathers of cases and 2/41 fathers of controls worked in PVC plants during the conception time; 0/41 case fathers and 3/41 control fathers worked in PVC plants with a contract work before conception. More cases than controls lived within 3 miles from the PVC facilities (p < 0.02) Matching by paternal education (which is strictly associated with job) could have introduced negative confounding. Reviews of the available studies are provided by Wagoner and Infante [143] and Clemmesen [144], We conclude that the three studies provide inadequate evidence of the association between environmen tal or paternal exposure to vinyl chloride and birth defects in humans. Water Fluoridation Mainly, correlation studies are available. Rapaport [145] estimated prevalence rates of Down's syndrome in Illinois towns divided into three categories of levels of water fluoridation and reported a positive correlation. Two other investigations, however, showed absence of such geographic correlations [146,147], Only Erickson's paper took maternal age into account. In another study [148], prevalence rates of 278 Hemming, ..id Vineis TABLE VH, Evaluation of Levels of Evidence on Teratogenicity in Humans (According to the Criteria Listed in the Text) High level Ethanol Methylmcrcury PCBs (Kanechlor) Laboratory work (solvents?) Limited Anesthetic gases Carbon monoxide Low level Coffee Hexachlorophcne LSD Nitrous oxide Tobacco Smelter work Inadequate 2,4,5-T Fluoridation* Gasoline Halothane Lead Potato blight* Soft water* Spray adhesives Marihuana TCDD Vinyl chloride There is some evidence that potato consumption, water fluoridation, and soft water ate not associated with human malformations. Down's syndrome were very similar in 30 towns before fluoridation, in the same towns after fluoridation, and in 321 other communities with levels of fluoride lower than 0.03 ppm. Confounders like maternal age and temporal trends of Down's syndrome were taken into account in the pre-/after-fluoridation comparison. No short term effect of fluoride on risk of Down's syndrome was apparent (relative risk = 0.95; upper 95% confidence limit 1.2). A further study dealt with congenital malformation patterns before and after fluoride introduction in Birmingham, U.K. [149]. The minor changes in (he preva lence of malformations were attributed to demographic changes or to modified ascertainment and classification. The authors' conclusion is that there was no evidence of an association. An American study shows prevalence rates of birth defects respectively in cities which had and cities which did not have fluorinated water supplies, and it takes maternal age into account [150]. The result is negative. Since the design of negative studies is superior to that of the one positive study, there is some evidence that water fluoridation is not a determinant of malformations. The level of epidemiologic evidence, reviewed above, is summarised in Table VII. EVALUATION OF EPIDEMIOLOGICAL AND ANIMAL EVIDENCE Experimental findings regarding compounds with high or limited evidence of teratogenicity in humans (Table VII) have been analyzed and summarized in the Appendix. In the experimental studies, all expressions of embryotoxicity detected during pregnancy or at birth were taken into consideration. Below are compared the experimental and human findings in qualitative and quantitative terms. References to the articles summarized in the Appendix are indicated by A and the number of the reference (A 1, A2, etc.). Oral or injected doses are compared directly between experimental animals and humans. For inhaled doses the calculations assume an inhalation volume of 15 m3/24 TABLE VIII. Animal Tests on Chemicals With High and Limited Evidence on Teratogenicity in Man After a Quantitated Exposure (for details, see text)_________ 11 i .( Daily dose liuman data i ', 'Ji 1 r ' Animal dgta i f.l ,,Da.ily dose i- i Substance (mg/kg) Risk Ref. Species (mg/kg) ' Risk Carbon monoxide Ethanol Halothane1 Mcthyimercury Nitrous oxidec Polychlorinated biphenyls 7 2,000 60 ppm = 0.6 ml/kg t.0b CNS defects 33% (FAS) 1.5 times 25% microcephaly 14,31,32} (37,40,41] [9-23,156] (65) 3,000 ppm = 0.3 1/kg 0.5J 5.5 vs. 3.6% 1.5 risk 100% (syniptoms) 30% i (small for ' date) [13,1561 [70,71,73] Mouse, rat. rabbit Mouse Mouse Mouse, rat Cat Rat Rat Monkey 30-250 ppm 3,000-5.000 1.10 1/kg 5-10 0.25 300 i/kg 0.2 i/kg 0.10 Resorptions ~30% (FAS-type) Threefold (skeletal) 50% Malformations 32% Malformations Mai format ions sixfold Fetal deaths 13% (Skin symp- toms) 50% (low birth weight' 40% (lethality) "It is assumed that halothunc is the responsible agent in anesthetics; in fact the evidence is inadequate (sec Table VII). bFor mcthyimercury i mg/kg is the total ingested dose; the daily doses were in the order of 80 /rg/kg. cJt is assumed that nitrous oxide is the responsible agent in anesthetics; in fact the evidence is low (see Table VII). dEstimated intake of Kancchlor. 1 Ref, from Appendix A12, A15, AI9 A28, A29 Al A40-A47 A48 A8.AI0 A9 A59.A60 280 Hemminki and Vineis h by a person of 50 kg by weight (= 0.3 nrVkg/d). The weights of mice are taken as 20 g and those of rats as 300 g. Oxygen consumption by weight is assumed to be two (rat) and three (mouse) times higher per time unit in rodents than in man. This equals daily inhalation volumes of 0.6 m3/kg/d for the rat and 0.9 m3/kg/d for the mouse. Complete absorption of the inhaled compound is assumed in the airways of experi mental animals and man. The quantitative comparisons between epidemiological and animal results are shown in Table VIII. Anaesthetic Gases, Halothane Cleft palates and skeletal malformations were induced in mice at a 3 h inhaled dose of 1% halothane [A 1 ]. In one study using a similar dose no effects were seen in rats and rabbits [A6]. Lower doses, about 0.2% halothane, were investigated only in rats exposed by inhalation for 8 or 12 h; fetal deaths and decreased fetal weights were noted [A3-A5]. When these results are translated into an inhaled dose, halothane caused malformations in mice after inhalation of 110 L of 1% halothane/kg, ie 1.10 L of halothane/kg. Rats experienced fetotoxicity and depressed weight gain, when inhaling 300 L of 0.2% of halothane/kg, ie, 0.6 L of halothane/kg. - - The limited epidemiological evidence for teratogenicity of occupational expo sure in the operating rooms suggested an excess of malformations at different sites in different studies and was not consistently related to skeletal malformations,-The halothane concentration in operating rooms ranges between 0.1 and 60 ppm f I56J: assuming a breathing volume of 5 m3 during an 8-h working day, these would correspond to intakes in the order of 0.0005-0.03 L of halothane, ie, 0.01-0.6 ml of halothane/kg. This is several orders of magnitude lower than the effective doses in mice and rats. However, the meaning of this comparison is doubtful, since the contribution of halothane to the teratogenicity of anaesthetic gases--if any--is not demonstrated. Anesthetic Gases, Nitrous Oxide Many types of malformations were induced in rats inhaling either 45-50% or 70-75% nitrous oxide for 24 h [A8.A10]. This translates to about 300 L of nitrous oxide/kg. Fetal deaths have been observed at a lower dose of I'.tXX) ppm inhaled for 8 h [A9], This translates to 0.2 L of nitrous oxide/kg. The concentrations of nitrous oxide in operating rooms have ranged between 30 and 3,000 ppm [156]. This would involve inhalation of 0.15-15 L of nitrous oxide in a working day, and a dose of 0.003-0.3 L of nitrous oxide/kg. It can be observed that the upper dose limit is in the range where fetal deaths have been observed in experimental animals. This may be related to the assumed role of nitrous oxide to account for the excess of spontaneous abortions in the female anesthesia staff [9,13,18,22,156]. Carbon Monoxide Several experimental studies are available indicating embryotoxic effects other than major structural malformations. In mice, 250 ppm of carbon monoxide induced some resorptions and lumbar ribs [A 12]. In rats, 150 ppm decreased the physical activity of the pups [A14], 230 ppm decreased weights of the pups [A 15], and 30 and 90 ppm caused preimplantation losses [A16]. In rabbits exposed to 90 and 180 ppm, stillbirths were increased and body weights were decreased [A 19]; another study with Teratogenicity Data 281 rabbits was negative-{A 12]. Fetal deaths and stillbirth were observed in sheep and pigs [A20,A21J. An important corollary to the findings in human, 1-3 h exposure to 1,000-3,000 ppm at near-term caused neurological defects and necrotized areas of brain in 5/9 Rhesus monkeys [A22]. In the case of exposed humans, neurological defects and brain damage have been observed. Unfortunately, no sound data are available on the underlying levels of exposure. " Ethanol Malformations at differnt organs have been induced in two studies on mice, including, features similar to FAS such as maxillary hypoplasia, oral clefts, micro phthalmia, and shortened palpebral fissures [A28,A29]. The doses used were about 3-5 g/kg/d. With the exception of one study mentioning microcephaly, which was inadequately reported^A30], studies on rats have been negative as far as structural malformations are concerned. On the other hand, in teratogenicity tests microcephaly can be easily missedif-it not specifically looked for. Nevertheless, several studies in rats report some form of developmental toxicity such as increased resorptions and reduced weight gain. Alcohol was also reported to induce brain immaturity in guinea pigs [A37], multiple types of malformations in piglets [A38], and oral clefts, absence of kidneys, and kinky tails in dogs [ A39]. The study on rabbits was negative in terms of malformations: However, loss of implanta tion was increased [A26], The risk of a chronic alcoholist to bear a child with FAS is about 30% (see epidemiologic section). The daily doses consumed by such persons would be above 100 g ethanol, ie, above 1,800 mg/kg/d. This is not too far from effective doses in mice. However, there is a need to emphasize that even though ethanol has been embryotoxic in the rat (at doses of the same order as in mice), malformations and particularly FAS-type anomalies have not been observed in this species. Species other than rats and mice have not been investigated extensively, Methylmercury In mice, rats, and hamsters, single doses of less than 10 mg/kg, given at appropriate times, induced structural malformations (mainly oral clefts and cleft palate in mice and heterogeneous malformations in rats and hamsters). No malfor mations were observed in the miniature swine and dogs at doses which produced toxic effects in the mothers. In some studies, particular attention was given to histological changes in the brain. Cerebellar changes were detected in 8/65 mice given 1 mg/kg/d methylmercury chloride throughout pregnancy vs 0/111 in controls [A43). In rats, cellular necrosis of the CNS was found following treatments corre sponding to maternal liver concentration above 15 pg/g [A49]. In cals, prolonged treatment with 0.08-0.2 mg/kg/d produced dose-related cerebellar histological changes as well as reduced weight at birth [A48], In rats and mice, daily subcutaneous doses of 5-10 mg/kg or oral doses of 20 mg/kg caused malformations of various types in about half the pups, including cerebellar changes but not microcephaly. The subcutaneous route seems 2-3 times more effective than the oral one. Based on the data from Iraq, Clarkson et al [65] have estimated that a maternal exposure to about 1 mg/kg caused a 50% probability for the child to have CNS effects typical of poisoning, half of which are microce- 282 Hemminki and Vineis phalic. Thus, effective chronic doses in humans seem to be in the same order of magnitude and possibly somewhat lower than effective acute doses in rodents. Both in humans and rodents, histological damage to the CNS has beemnoted at a 1-mg/kg dose. PCBs (Kanechlor 400) PCB compounds have reduced implantations in mice at very low doses of 0.025 mg/kg in one study [A53] and at a dose of 0.5 mg/kg in another [A55], Up to 50% malformations were seen in mice exposed to 8 and 16 mg/kg of 3,3'. 4,4', 5,5'hexachlorobiphenyl [A56]. Studies on rat and rabbits have not noted malformations [A57] but have observed a decrease in birth weight and an increase in stillbirth at 50100 mg/kg [A58J. Feeding studies with monkeys at 5 ppm have shown weight reduction in the offspring and some fetal mortality. This corresponds to a dose in the order of 0.1 mg/kg. Thus, PCBs appear to be developmental toxins and cause hyperpigmentation of the skin in monkeys similar to their effect in humans. In humans, Kanechlor 400 has caused 100% hyperpigmenfation symptoms and 30% babies small for date at estimated daily doses of 0.5 mg7kg PCB, the range being 0.06-0.8 mg/kg [70,71,73]. The contribution of dibenzofurans or other contam inants in the Yusho oil cannot be ruled out. Skin changes and low birth weight can be taken as a basis for comparison between effective human and monkey doses.. Since 13% monkeys were affected by skin lesions at a dose of 0.1 mg/kg, linear extrapola tion would yield 100% response at a dose of 0.8 mg/kg. Thus, the ratio of human/ animal dose would be 0.65. This, however, is based on two monkeys. This is a dose range (0.025-0.5 mg/kg) where reduced implantations were observed in mice. Un fortunately, no related observations are available in humans. CONCLUSIONS Testing for teratogenicity is required by many national agencies responsible for the safety of food, drugs, and pesticides. Yet a quantitative evaluation of the predictive value of teratogenicity tests had been lacking, although arguments have been pre- __ sented on qualitative grounds. In the present exercise, a qualitative evaluation was made of the level of evidence of the ability of several chemicaFexposures to induce malformations in humans. Effective doses and the magnitude of effects were recorded as much as it was possible from the original studies. A number of exposures definitely or plausibly associated to human malformations were so identified. Qualitative and quantitative findings were then compared to observations resulting from animal experiments. Results are summarized in Table VIII. The accuracy (and therefore the validity) of these comparisons is impaired by several shortcomings: 1) some substances induced different types of lesions in humans and experimental-animals,-eg, micro cephaly induced by methylmercury is only seen in humans; 2) exposure data in humans are often crude; 3) dose-response data in humans (and often in experimental conditions) were largely unsatisfactory, so that comparisons were restricted to the order of magnitude of dose levels giving a similar magnitude of effects; 4) animal studies were by no means consistent, and observations under the most sensitive conditions were used for comparisons. In the case of ethanol, MeHg, and PCBs, the evidence fora causal association with human malformations was high. These chemicals also'induced malformations ASI 00004739 Teratogenicity Data 283 under some experimental conditions. However, the experimental effects neither con sistently reproduced those observed in humans nor were similar in the different species in which they yvere investigated. In mice, ethanol produced features similar to FAS as well as other malformations hitherto associated with neither ethanol nor alcoholism in human. No malformations either similar to FAS or of any types were reported in rats or rabbits exposed to ethanol. Had the studies been limited to these two species, the ability of alcohol to induce malformations would have gone unnot iced. However, in the two species, other forms of developmental toxiciy would have been apparent. __ __ In human's.'mefhylmercury produced morphological changes in the brain and these were detected also in mice, rats, and cats when appropriate methods were used. On the other hand, other malformations were seen in laboratory animals which would have hardly "escapedjjttention in humans, had they occurred in this species. For PCBs, an experimental counterpart to the changes induced in human fetuses was detected in monkeys only, whereas structural malformations at other sites were found in rodents. These comparisons should be considered very cautiously. The application of very rigid criteria of analysis of the epidemiological studies has led to the recognition of a very small number of cause-effect relationships. In addition, these conditions referred to malformations which are veiy rare in people who are not exposed to the relevant agent. On the one hand, at least one experimental counterpart has been identified for the human malformations shown to be associated respectively to ethanol, MeHg, and PCBs and the order of magnitude of the effective doses was not too different in the human and in the experimental conditions. In addition, malformations at sites different from those described in humans have been seen in laboratory animals given each of the three chemicals. A small part of this difference can be ascribed to the low sensitivity of epidemiological studies in picking up associations. Other reasons may relate to differences in time of exposure and to differences in developmental biology of species. Finally, one should wonder whether these three chemicals convincingly shown to be human teratogens would have been detected on the basis of conventional routine testing. Had ethanol and PCBs been tested only in rats and rabbits, their ability to induce malformations would have gone undetected. In strict terms, they might be considered as "false negatives." On the other hand, embryotoxic changes in other than structural malformations would have been recognized in the two species. Fur thermore, elements treated with agents such as halothane, nitrous oxide, ethanol, and PCBs that caused both malformations and other types of embryotoxicity were more sensitive to the latter than to the former lesions. Carbon monoxide caused no struc tural malformations but it induced resorptions, stillbirth, and decreased weight gain. On this basis, in pragmatic terms, the practice of considering any experimental embryotoxic effect as a possible predictor of teratogenicity in the human species seems justified. REFERENCES 1. Wilson JG: "Environment and Birth Defects." New York: Academic Press, 1973, p 305. 2. Tomatis L, AgthtTC," Bartsch H, Huff J, Montesano R, Saiacci R, Walker E, Wilboum J: Evaluation of the carcinogenicity of chemicals: A review of the monograph program of the international agency for research on cancer. Cancer Res 38:877-885, 1978. 3. IARC working group: An evaluation of chemicals and industrial processes associated with cancer in humans based on human and animal data. IARC Monographs, Vol. 1 to 20. Cancer Res 40:1- 284 Hemminki and Vineis 12, 1980. 4. Barlow SM, Sullivan FM. "Reproductive Hazards of Industrial Chemicals." New York-. Academic Press, 1982. 5. Wilson JG: Environmental chemicals. In Wilson JG, Fraser FC: "Handbook of Teratology, Vol. 1." New York: Plenum Press, 1977, pp 357-385. 6. Shephard TH: "Catalog of Teratogenic Agents, 3rd Ed." Baltimore: John Hopkins University Press. 1980. 7. Miellinen OS: Design options in epidemiologic research. Scand J Work Environ Hlth 8 (suppl. 1):7-14, 1982. 8. Leek I: Correlations of malformations frequency with environmental and genetic attributes in man. In Wilson JG, Clark FC: "Handbook of Teratology, Vol. 3." New York: Plenum Press, 1977, pp 243-324. 9. Cohen EN, Bellvillc JW, Brown BW: Anesthesia, pregnancy, and miscarriage: A study of operating room nurses and anesthetists. Anesthesiology 35:343-347, 1971. 10. Ericson A, Kallcn B: Survey of infants born in 1973 or 1975 to Swedish women working in operating rooms during their pregnancies. Anesth Analg 58:302-305, 1979. 11. Rosenberg P, Kirves A: Miscarriages among operating theatre staff. Acta Ancsthesiol Scand 53:37-42, 1973. 12. Rosenberg PH, Vantlinen H: Occupational hazards to reproduction and health in anesthetists and paediatricians. Acta Anaesthesiol Scand 22:202-207, 1978. 13. Cohen EN, Brown BW, Wu ML, Whitcher CE, Brodsky JB, Gift HC, Greenfield W. Jones TW, Driscoll EJ: Occupational disease in dentistry and chronic exposure to trace anesthetic gases. JADA 101:21-31, 1980. 14. Lauwerys R, Siddons M, Misson CB, Borlee I, Bouckacrt A, Lechat MF, De Temmerman P: Anesthetic health hazards among Belgian nurses and physicians. Int Arch Occup Environ Health 48:195-203, 1981. 15. Knill-Jones RP, Moir DD, Rodrigues LV. Spence AA: Anesthetic practice and pregnancy. Lancet 1:1326-1328, 1972. 16. Corbett TH, Cornell RG, Endres JL. Licding K: Birth defects among children of nurse-anesthetists. Anesthesiology 41:341-344, 1974. 17. Cohen EN, Brown BW, Bruce DL, Cascorbi HF, Corbett TH, Jones TW. Whitcher CE: Occupa tional disease among operating room personnel: A national study. Anesthesiology 41:321-340, 1974. 18. Knill-Jones RP, Newman BJ, Spence AA: Anesthetic practice and pregnancy. Lancet 11:807-809, 1975. 19. Golhe CJ, Hoffman O: The consequences on pregnancy of smoking and occupational exposure to anesthetic gases. Proceedings of the rv Swedish-Yugoslav days of occupational medicine. Skopje 1977, pp 11-27. ------ . '1 . . 20. Pharoah POD, Alberman E, Doyle P, Chamberlain G: Outcome of pregnancy among women in anesthetic practice. Lancet 1:34 -36, 1977. 2L .Cohen EN, Brown BW, Bruce DL, Cascorbi HF, Corbett TH, Jones TW, Whitcher CE: A survey of anesthetic health hazards among dentists. JADA 90:1291-1296, 1975. 22. Tomlin PJ: Health problems of anesthetists and their families in the West Midlands. Br Med J {Clin Res] 1:779-784, 1979. 23. Axelsson G, Rylander R: Exposure to anaesthetic gases and spontaneous abortions: Response bias in a postal questionnaire study, Int J Epidemiol 11:250-256, 1982. 24. Morris MB, Weinstein L: Caffeine and the-fetus: Is trouble brewing? Am J. Obstet-Gynecol 140:607-610, 1981. 25. Weathersbec PS, Olsen LK, Lodge JR: Caffeine and_pregnancy: A retrospective survey. Postgrad Med 62:64-69, 1977. ' 26. Heinonen OP, Slone D, Shapiro S: Birth defects and drugs in pregnancy. Littleton, MA: Publishing Sciences Group Inc, 1977. 27. Borlee L, Lechat MD, Bouckaen A, Misson C: Le cafe, facieur de risque pendant la grossesse? Louvain Med 97:279-284, 1978. 28. Rosenberg L, Mitchell AA, Shapiro S, Slone D: Selected birth defects in relation to caffeine- containing beverages JAMA 247:1429-1432, 1982. 29. Linn S, Schoenbaum SC, Monson RR, Rosner B. Stubblefield PG. Ryan KJ: No association between coffee consumption and adverse outcomes of pregnancy. N Engl J Med 306:141-145, 1982. ` ASX 00004741 Teratogenicity Data 285 30. Kurppa K. Holmberg P, Kuosma E. Saxcn L: Coffee consumption during pregnancy. N Engl J Med 306-1548, 1982. 31. Grabowski CT; Atmospheric gases. In Wilson JG, Fraser FC: "Handbook of Teratology, Vol. 1." New York: Plenum Press, 1979. 32. Longo LL: Carbon monoxide in the pregnant mother and fetus and its exchange across the placenta. Ann NY Acad Sci 174:313-341, 1970. 33. Jones KL, Smith DW; The fetal alcohol syndrome. Teratology 12:1-10, 1975. 34. Streissguth AP, Landesman-Dwyer S, Martin JC, Smith DW: Teratogenic effects of alcohol in humans and laboratory animals. Science 209:353-361, 1980. 35. Lemoine P.Harousseau H, Borteyru IP, Mcnuet JC: Les enfants des parents alcooliques. Anomalies observees & propos de 1127 cas. Arch Fr Pediatr 25:830-832, 1967. 36. Jones KL, Smith t>Wr Recognition of the fetal alcohol syndrome in early infancy. Lancet 11:999- 1001, 1973. 37. Hanson JW, Streissguth AP, Smith DW: The effects of moderate alcohol consumption on fetal growth and morphogenesis. J Pediatr 92:457-460, 1981. 38. Sokol RJ: Alcohol and.abnormal outcome of pregnancy. Can Med Assoc J 125:147-152, 1981. 39a. Dehaene P, Samaille-Villette D, Samaille P, Crepin G, Walbaum R, Deroubaix P, Blanc-Garin AP: Lc syndrome d'alcoolismc foetal dans lc nord de la France. Rev Alcoolismc 23:145-158, 1917. r 39b- Sokol RJ, Miller SI,"Reed G: Alcohol abuse during pregnancy: An epidemiologic study. Alcohol ism: Clin Exp Res. 4". 135-145, 1980. 40. Oleglrd R, Sabel K-GTAronsson M, Sandin B, Johansson PR, Carlsson C, Kyllerman A, Iverscn K, Hbek A: Effects on the child of alcoholic abuse during pregnancy. Acta Paediatr Scand (suppl) 275, 1979. 41. Jones KL, Smith DW, Streissguth AP, Myrianthopoulos NC: Outcome in offspring of alcoholic women. Lancet 1:1076-1078, 1976. 42. Clarren SK, Smith DW: The fetal alcohol syndrome. N Engl J Med 298:1063-1067, 1978. 43. Nestler VM, Spohr H-L, Steinhausen H-C: "Die Alkohol-embryopathie, Mehrdimensionale Studien zu den Folgen des Alkoholismus in der Schwangerschaft." Stuttgart: Enke Verlag, 1981, pp 95. 44. Majewski F: Alcohol embryopathy: Some facts and speculations about pathogenesis. Ncurobehav Toxicol Teratol 3:129-144. 1981. 45. Ouellette EM, Rosett HL, Rosman NP, Weiner L: Adverse effects on offspring of alcohol abuse during pregnancy. N Engl J Med 297:528-530, 1977. 46. Tennes K, Blackard C: Maternal alcohol consumption, birth weight, and minor physical anomalies. Am J Obstet Gynecol 138:774-780, 1980. 47. Hunter AGW, Thompson P, Evans JA: Is there a feta! gasoline syndrome? Teratology 20:75-80, 1979. -- 48. Hailing H: Suspected link between exposure to hexachlorophcne and malformed infants. Ann NY Acad Sci 320:426^35, 1979. 49. Baltzar B, Ericson "A, -Kallen B: Delivery outcome in women employed in medical occupations in Sweden. J Occup Med 21:543-548, 1979. 50. Rom WN: Effects of lead on the female and reproduction. A review. Mt Sinai J Med (NY) 43:542- 552,1976. 51. Palmisano PA, Sneed RC, Cassady G: Untaxed whiskey and fetal lead exposure. J Pediatrics 75:869-872, 1969. 52. Zellweger H, McDonald JS, Abbo G: Is lysergic-acid diethylamide a teratogen? Lancet H: 1066-- 1068, 1967. 53. Hecht F, Beals RK, Lees MH, Jolly H, Roberts P: Lysergic-acid-diethylamide and cannabis as possible teratogens in man. Lancet 11:1087, 1968. 54. Carakushansky G, Neu RL, Gardner LI: Lysergide and cannabis as possible teratogens in man. Lancet 1:150-151, 1969. 55. Assemany SR, Neu RL, Gardner LI: Deformities in a child whose mother took LSD. Lancet 1:1290, 1970. 56. Hsu LY, Strauss L, Hirshhorn K: Chromosome abnormality in offspring of LSD user. JAMA 211:987-990, 1970, .s- 57. Sato H, Pergament E: Is lysergide a teratogen? Lancet 1:639-640, 1968. 58. Aasc JM, Laestadius -N, Smith DW: Children of mothers who took LSD in pregnancy. Lancet C: 100-101, 1970. jinu,jJiLHwgJ11 'iBtaw.im i Bpg!egeegggwBBP 286 Hemminki and Vineis 59. McGlothlin WH, Sparkes RS, Arnold DO: Effect of LSD on human pregnancy. JAMA 212:1483- 1487, 1970. 60. Koos BJ, Longo LD: Mercury toxicity in the pregnant woman, fetus and newborn infant. Am J Obstet Gynecol 126:390-490, 1976. 61. Chang LW, Wade PR, Pounds JG, Reuhl KR: Prenatal and neonatal toxicology and pathology of heavy metals. Adv Pharmacol Chemothcr 17:195-231. 1980. 62. Harada M: Congenital minamata disease: Intrauterine mcthylmercury poisoning. Teratology 18.285- 288, 1978. 63. Amin-Zaki L, Elhassani SB, Majeed MA, Clarkson TW, Doherty RA, Greenwood MR: Intrauterine methylmercury poisoning in Iraq. Pediatrics 54:587-595, 1974. 64. Amin-Zaki L, Majeed MA, Elhassani SB, Clarkson TW, Greenwood MR, Doherty RA: Prenatal methylmercury poisoning, clinical observations over five years. Am J Dis Child 133:172-177, 1979. 65. Clarkson TW, Cox C, Marsh DO, Myers GJ, Al-Tikriti SK, Amin-Zaki L. Dabbagh AR: Dose- response relationship for adult and prenatal exposures to methylmercury. In Barg GG, Maillie HD: "Measurement of Risks." Plenum Publishing Co., New.York, 1981, pp 111-130. 66. Snyder RD: Congenital mercury poisoning. N Engl J Med 284:1014-1016. 1971. 67. Matsumoto H, Koya G, Takeuchi T: Fetal Minamata disease: A ncuropathological study of two cases of intrauterine intoxication by a methyl mercury compound. J Neuropathol Exp Neurol 24:563-574, 1965. 68. Choi BH, Lepham LW, Amin-Zaki L: Abnormal neuronal migration, deranged cerebral cortical organization and diffuse white matter astrocytosis of human fetal brain: A major effect of methyl mercury in utero. J Neuropathol Exp Neurol 37:719-733, 1978. 69. Reggiani G: Toxicology of TCDD and related compunds. Observations in man. In Hutzmger O: "Chlorinated Dioxins and Related Compounds." Impact on the environment, Oxford: Pcrgamon Press, 1982, pp 463-493. 70. Kuratsune M, Yoshimura T, Matsuzaka J, Yamaguchi A: Epidemiologic study on Yusho a poisoning caused by ingestion of rice oil contaminated with a commercial brand of polychlorinated biphenyls. Environ Health Perspect 1:119-128, 1972. 71. Miller RW: Cola-colored babies Chlorobiphenyl poisoning in Japan. Teratology 4:211-212, 1971. 72. Yamashita F: Clinical features of chlorobiphenyls (PCBs) induced fetopathy. Paediatrician 6:20-27. 1977. 73. Shiota K: Fetal accumulation of polychlorinated biphenyls (PCBs). A consideration on the problem of human prenatal exposure to environmental pollutants. Cong Anom 16:9-16, 1976. 74. Renwick JH: Hypothesis--anencephaly and spina bifida are usually preventable by avoidance of a specific but unidentified substance present in certain potato tubers Br J Prev Soc Med 26:67-88, 1972. 75. Renwick JH: Anencephaly and potatoes. Lancet 1:96-97, 1973. 76. Emanuel I: Non-tuberous neural-tube defects. Lancet 11:879, 1973. 77. Smith C, Wan M, Boyd AEW. Holmes JC: Anencephaly, spina bifidirs-and potato blight in the Edinburgh area. Lancet 1:269. 1973. - - 78. MacMahon B. Yen S, Rothman KJ: Potato blight and neural-tube defectssd-ancet 1:589-599, 1973. 79. Elwood JH, Nevin NC: Factors associated with anencephalus and spina bifida in Belfast. Br J Prev Soc Med 27:73-80, 1973. 80. Elwood JH, Mackenzie G: Associations between the incidence of neurological malformations and potato blight outbreaks over 50 years in Ireland. Nature 243:476-477, 1973. 81. Cruz-Coke R: Anencephaly and potatoes in Chile. Lancet 1:269-270, 1973. 82. Elwood JM: Anencephaly and potato blight in Eastern Canada. Lancet 1:769. 1973. 83. Field B, Kerr C: Potato blight and neural-tube defects. Lancet 11:507-508, 1973. 84. Kinlen L, Hewitt A: Potato blight and anencephalus in Scotland. Br J Prcv -Soc Med 27:208-213, 1973. 85. Masterson JG, Frost C, Bourke GJ, Joyce NM, Hcrity B, Wilson-Davis Kr-Anencephaly and potato blight in the Republic of Ireland. Br J Prev Soc Med 28:81-84, 1974. 86. Lorber J, Stewart CR, Milford Ward A: Alpha-fetoprotein in antenatal diagnosis of anencephaly and spina bifida. Lancet 1:1187, 1973. 87. Roberts CJ, Revington CJ, Loyd S: Potato cultivation and storage in SoutlTWales and its relation to neural tube malformation prevalence. Br J Prev Soc Med 27:214-216, 1973. 88. Clarke CA, McKendrick OM. Sheppard PM: Spina bifida and potatoes. Br Med J (Clin Res] 3.251- 254, 1973. Teratogenicity Data . 287 89. Spiers PS, Pietrzyk J, Piper JM, Glehatis DM: Human potato consumption and neural-tube malfor mation. Teratology 10:125-128, 1974, 90. Nevin NC, Merrctt JD: -Potato avoidance during pregnancy in women with a previous infant with either anencephaTy and/or spina bifida. Br J Prev Soc Med 29:111-115, 1975. 91. Nordstrom S, Beckman L, Nordenson I: Occupational and environmental risks in and around a smelter in northern Sweden, VI. Congenital malformations. Heredilas 90:297-302, 1979. 92. Nordenson I, Beckman G, Beckman L, Nordstrom S: Occupational and environmental risks in and around a smelter in northern Sweden. II. Chromosomal aberrations in workers exposed to arsenic. Hereditas 88:47-507T5T8a. .93, Nordenson I, Beckman Gt Beckman L, Nordenson I: Occupational and environmental risks in and around a smelter iniAtfthfcm Sweden: IV. Chromosomal aberrations in workers exposed to lead. Hereditas 88:263-267, 1978b. 94. Fedrick J: Ancncephalus and the local water supply. Nature 227:176-177, 1970. 95. Fielding DW, Smithells RW: Anencephalus and water hardness in South-West Lancashire. Br J Prev Soc Med. 25:211-219, 1971. 96. Lowe CR, Roberts.CJ^Lloyd S: Malformations of central nervous system and softness of local water supplies. Br Med J [Clin Res] 2:357-361, 1971. 97. Crawford MD, GardncmMJ, Sedgwick PA: Infant mortality and hardness of local water supplies. Lancet 1:988-992, 197^ 98. Morton MS, ElwOSd PCTAberncthy M: Trace elements in water and congenital malformations of the central nervous system in South Wales. Br J Prev Soc Med 30:36-3, 1976. 99. St. Leger AS, Elwood PC, Morton MS: Neural tube malformations and trace elements in water. J Epidemiol Community H^lth 34:186-187, 1980. 100. Elwood JM, Goldman AJ: Water composition in the etiology of anencephalus. Am J Epidemiol 113:681-690, 3981. 101. Kucera K: Exposure to fat solvents: A possible cause of sacral agenesis in man. J Pediatr 72:857859, 1960. 102. Meirik O, Kallen B, Gauffin U, Ericson A: Major malformations in infants born of women who worked in laboratories while pregnant. Lancet 11:91. 1979. 103. Hansson E, Jansa S, Wande H, Kalln B, Ostlund E: Pregnancy outcome for women working in laboratories in some of the pharmaceutical industries in Sweden. Scand J Work Environ Health 6:131-134, 1980. 104. Blomqvist O, Ericson A, Kallen B, Westerholm P: Delivery outcome for women working in the pulp and paper industry. Scand J Work Environ Health 7:114-118, 1981. 105. Holmberg P: Central-nervous-system defects in children born to mothers exposed to organic solvents during pregnancy. Lancet II: 177-179, 1979. 106. Holmberg PC, Nurminen M: Congenital defects of the central nervous system and occupational factors during pregnancy. A case-referent study. Am J Ind Med 1:167-176, 1980. 107. Oakley GP, Nissun-'Jf Hanson JW, Boyce JM, Roberts M: Epidemiologic investigations of possible teratogenicity of spray adhesives. Teratology 9:31A-32A, 1974. 108. Silberg SL, Rafison DR, Lyon JA, Anderson PS: Relationship between spray adhesives and congenita] malformations. South Med J 72:1170-1173, 1979. 109. Young AL, Calcagni JA, Thalken CE, Tremblay JW: The toxicology, environmental fate, and human risk of herbicide orange and its associated dioxin. USAF OEHL Technical Report TR-7892. USAF Occupational and Environmental Health Laboratory, 1978. 110. Tung TT, Anh TK, Tuyen BQ, Tra DX, Hugen NX: Clinical effects of massive and continuous utilization of defoliants on civilians. Vietnamese Studies 29:53-81, 1971, 111. Rose HA, Rose SPR: Chemical spraying as reported by refugees from South Vietnam. Science 177:710-712, 1972. 112. Honoroff I: Down's syndrome--it can happen here. A report to the consumer III (50), 1-4. Sherman Oaks, California: 1973. 113. Committee on the effects of herbicides in South Vietnam: "Part A. Summary and conclusions." Washington DC: National Academy of Science, 1974. 114. Roan CC, Morgan-BPr*1 Alleged Effects on Human Health of the Use of Herbicides in the Area Around Globe, Arizona." Arizona community pesticide study project. Tucson: University of Arizona, 1972. 288 Hemminki and Vineis 115. Advisory Committee on 2,4,5-T: "Report to the administrator of the Environmental Protection Agency." Washington DC: U.S. Environmental Protection Agency, 1971. 116. Sare WH, Forbes PI: Possible dismorfogenic effects of an agricultural chemical: 2,4,5-T. NZ Med J 75:37-38. 1972. 117. New Zealand Agricultural Chemical Board: "Report of the Subcommitfee'on 2,4.5-T." Wellington, New Zealand: 1972. 118. Cutting RK, Phuoc TH, Ballo JM, Bencnson MW, Evans CH: "Congenital Malformations, Hydatidiform Moles and Stillbirths in the Republic of Vietnam 1960-1969." Washington DC: Department of Defense. US Government Printing Office No. 903.233, (abstract), 1970. 119. McQueen EG, Veale AMO, Alewander WS, Bates MN: "2,4,5-T and Human Birth Defects." New 2iealand: Department of Health, June, 1977. 120. Hanify JA, Metcalf P, Nobbs CL, Worsley KJ: Aerial spraying of 2,4,5-T and human birth malformations; An epidemiological investigation. Science 212:349-351, 1981. 121. Field B, Kerr C: Herbicide use and incidence of neural-tube defects. Lancet 1:1341-1342, 1979. 122. Nelson CJ, Holson JF, Green HG, Gaylor DW: Retrospective study of the relationship between agricultural use of 2.4,5-T and deft palate occurrence in Arkansas. Teratology 19:377-384, 1979. 123. Thomas HF: 2,4,5-T use and congenital malformation rates in Hungary. Lancet 11:214-215, 1980. 124. Brogan WF, Brogan CE, Dadd JT: Herbicides and cleft lip and palate. Lancet 11:597 (letter), 1980. 125. Bruzzi P, Bisanti L, Borgna-Pignatti C, Marni E, Abate L, Basso P, Formigaro F, Strigini P, Sami L: Birth defects in the TCDD polluted area of Scveso. Results of a 4-year follow-up. Provisional draft, August, 1981. -~_ 126. Smith AH, Malhcson DP, Fisher DO, Chapman CJ: Preliminary report of reproductive outcomes among pesticide applicators using 2,4,5-T. NZ Med J 680'177-179, 1981. -- 127. Smith AH, Fisher DD, Pearce N, Chapman CJ: Congenita! defects and miscarriages among New Zealand 2,4,5-T sprayers. Arch Environ-Health 37:197-200. 1982. ____ _ 128. Townsend JC, Bodner KM, Van Pecnen PFD, Olson RD, Cook RR: Survey of reproductive events of wives of employees exposed to chlorinated dioxins. Am J Epidemiol 115:695-713, 1982. 129. Landesman-Dwyer S, Emmanuel 1: Smoking during pregnancy. Tertology 19:119-126, 1979. 130. Johnston C: Cigarette smoking and the outcome of human pregnancies: A status report on the consequences Clin Toxicol 18.189-209, 1981. 131. Fedrick J, Alberman ED, Goldstein H; Possible teratogenic effect of cigarette smoking. Nature 231:529-530. 197). 132. Yerushalmy J: Congenital heart disease and maternal smoking habbits. Nature 242:262, 1973. 133. Kelsey JL, Dwyer T, Holford TR, Bracken MD: Maternal smoking and congenital malformations: An epidemiological study. J Epidemiol Community Health 32:102-107, 1978. 134. Evans DR, Newcombe RG, Campbell H: Maternal smoking habits and congenital malformations: A population study. Br Med ) [Clin Res] 2:171-173, 1979. _-*sr 135. Christianson RE: The relationship between maternal smoking anduhe,incidence of congenital anomalies. Am J Epidemiol 112:684-695, 1980. - .. 136. Himmelberger DU, Brown BW, Cohen EN: Cigarette smoking during,pregnancy andthe occur rence of spontaneous abortion and congenital abnormality. Am J Epidemiol 108:470-479, 1978. 137. Andrews J. McGarry JM: A community study of smoking in pregnancy. J Obstet Gynaecol Br Commonwealth 79:1057-1073, 1972. ~ ___ ' 138- Ericson A, Kallen B, Westerholm D: Cigarette smoking as an etiologic factor in cleft lip and palate. Am J Obstet Gynecol 135:348-351, 1979. ... 139. Hemminki K, Muianen P, Salonicmi I: Smoking and the occurrence of congenital malformations and spontaneous abortions: Multivariate analysis. Am J Obstet Gynecol 145:61-66, 1983. 140. Infante PF: Oncogenic and mutagenic risks in communities with polyvinyl chloride production facilities. Ann NY Acad Sci 271:49-57, 1976. . 141. Edmonds LD, Falk H, Nissim JE: Congenital malformations and .vinyl chloride. Lancet 11:1098. 1975. 142. Edmonds LD, Anderson CE, Flynt JW Jr, James LM: Congential central nervous system malfor mations and vinyl chloride monomer exposure: A community study. Teratology 17; 137-142, 1978. 143. Wagoner JK, Infante PF: A review of the methodological approaches in-the assessment of an association between vinyl chloride exposure and reproductive hazards Jn Infante PF, Legator MS: "Proceedings of a Workshop on Methodology for Assessing Reproductive Hazards in the Work place." NIOSH-DHHS Publication No. 81-100. Washington DC, 1980, pp 43^19. 144. Clemmescn J: Mutagenicity and teratogenicity of vinyl chloride monomer (VCM). Epidemiologic evidence. Mutat Res 98:97-100, 1982. -- -- - Teratogenicity Data 289 145. Rapaport I: Nouvelles recherches sur Ic mongolisme. A propos du role pathogcniquc du fluov. Bull Acad Nall Med (Paris) 143.367-370, 1959. 146. Berry WTC; A study on ihc incidence of mongolism in relation to the fluoride content of water. Am J Ment Defic 62:634-636, 1958. 147. Erickson JD, Oakley GP, Flynt J\V, Hay S: Water fluoridation and congenital malformations. No association. JADA 93:981-984. 1976. 148. Needleman HL, Pueschel SM. Rothman KJ: Flouridation and the occurrence of Down's syndrome. N EnglJ Med 291:821-823, 1974. 149. Knox EGt Armstrong E, Lancashire R; Flouridation and the prevalence of congenital malforma tions. Community Medicine 2:190-194, 1980. 150. Erickson JD: Down syndrome, water fluoridation, and maternal age. Teratology 21:177-180, 1980, 151. Askrog V, Harvald B: Teratogen effect of inhalationsanaestetika Nord Med 83:498 -500, 1970. 152. Tomlin PJ: Pollution by anesthetic gases. Lancet 11:142, 1976. 153. Spence AA, Cohen EN, Brown BW, Knill-Joncs RP, Himmelberger DU: Occupational hazards for operating room-based physicians. JAMA 238:955-959, 1977. 154. Tomlin PL Teratogenic effects of waste anesthetic gases. Br Med J 1:108, 1978. 155. Underwood P, Hester LL, Lafine T, Gregg KV: The relationship of smoking to the outcome of pregnancy. Am JjSbstet Gynecol 91:270-276, 1965. 156. Vessey MP, Nunn JF: Occupational hazards of anesthesia. Br Med J (Clin Res] 281:696-698, 1980. ---sal APPENDIX A Experimental studies regarding compounds with high and limited evidence of teratogenicity in humans (Table VIII in preceding section) are summarized below in Tables A-I--A-V. All expressions of embryotoxicity detected during pregnancy or shortly after birth were considered. The results of these tests were used in the previous section, when human and animal data were compared (Table VIII). (Appendix begins on page 290) 4 User:004736 12feb86 PR 22/5/1-150 (Items 1-8) DIALOG 2 SDI PRINTS for UD*8603 DIALOG File 154: MEDLINE - 80-86/MAR 1854100 86071100. Extrapolation of the evidence on teratogenicity of chemicals between humans and experimental animals: chemicals other than drugs. Hemminkl K; Vine Is P Institute of Occupational Health, Helsinki, Finland. Teratogenes1s Carclnog Mutagen (UNITED STATES) ,1985, 5 (4) p251-318. ISSN 0270-3211 Journal Code: VM9 Languages: ENGL ISH Doc Typ : Review Journa1 Announcement: 8603 Subfile: INDEX MEDICUS Epidemiologic literature regarding the possible association between malformations and 23 exposures or occupations other than pharmaceut1ca1 products, was analysed. The qualitative level of scientific evidence was classified Into four categor1es: high (ethanol, methyl mercury, PCBs laboratory work), limited (anesthetic gases, carbon monoxide). low (hexachlorophene, LSD, nitrous oxide, smelter work, tobacco), and 1nadequate (all qther exposures). Human data for exposures be 1onging to categories "high" and "limited" were quantitatively compared to results of animal teratogenicity tests of the nelevant chemicals. Ethanol, methyl mercury, and PCBs have caused ma1 format ions in exper1menta1 animals, and the effective doses have ranged from 0.2 to 8,0 times the effective human doses. Ethanol and PCBs caused similar types of lesions In some animal species as have been observed in humans, (60 Refs. ) Tags: Animal; Comparative Study; Human; Support, Non-U.S. Gov' t Descriptors: Abnormalities, Drug-Induced--Occurrence (DC); A 1 coho 1, Ethyl --Adverse Effects (AE); Anesthetics--Adverse Effects (AE): Caffeine--Adverse Effects (AE); Cannab1 no 1ds--Adverse Effects (AE); Carbon Monoxide--Toxicity (TO); Copper--Adverse Effects (AE); Dioxins--Adverse Effects (A E); Env1ronmenta1 Exposure; Fluorides--Toxicity (TO); Gasoline--Toxtc1ty (TO); Hexachlorophene--Adverse Effects (AE) ; Lead--Tox1c1ty (TO): Lysergic Acid Dtethyl amide--Adverse Effects (AE); MethyI mercury Compounds--Toxicity (TO); METHYLMERCURY CPOS--Toxleity (TO); Occupational D1seases--Eti- ology (ET); Po1ychlorob1 phenyl Compounds--Tox1city (TO); POLYCHLOROBIPHENYL CPOS--Tax 1c1ty (TO); Potatoes-- Adverse Effects (AE); Sol vents; Species Specif1c1ty; `Teratogens; Tobacco; Vinyl Chlor1de--Toxlcity (TO); Water CAS Registry No.; 12737-87-0 .(KC 400): 50-37-3 .(Lysergic Acid Diethylamide); 58-08-2 .(Caffeine); 630-08-0 .(Carbon Monoxide); 64-17-5 .(Alcohol, Ethyl); 70-30-4 / . (Hexachl orophene ) -. --24^9-92- 1 .(Lead); 7440-50-8 .(Copper); 75-01-4 f. (Vinyl Chloride^: 7732- 18-5 .(Water) (1) p85-7, ISSN 0300-8738 Journal Code: SHP Languages; ROMANIAN Summary Languages: ENGLISH Doc Type: English Abstract Journal Announcement: 8603 Subflie: INDEX MEDICUS Tags: Animal Descriptors: `Aceta1dehyde--Po1 son 1ng (PO): Aceta1dehyde--T- oxlclty (TO); `Death, Sudden--Et1o1ogy (ET ) ; - Heart--Drug Effects (DE); Microscopy. Electron; Myocardium--U1trastrueture (UL): Rats; Rats, Inbred Strains CAS Registry No.: 75-07-0 .(Acetaldehyde) 1847874 86064874 [Toxic effect of volatile products of thermooxidfzfng decomposition of conveyor belts] Badanla toksyernego dz1aLania lotnych produk tow termooksydacyjnego rozkLadu pasow przenosn1kowych. Faff J; Dudklewlcz J; Kauczak M; Wantewskl E: Tokarzewski E WoJskowego Instytutu Higieny 1 Ep1demio1og11. Med Pr (POLAND) ,1985, 36 (3) p191-8. ISSN 0465-5393 Journa1 Code: MOJ Languages: POLISH Summary Languages: ENGLISH Doc Type: English Abstract Journal Announcement: 8603 Subf t ie: INDEX MEDICUS Toxic effects of tliermo-ox 1diz1ng decomposition products of conveying belts made of PVC and rubber have been tested. The resultant smokes were passed through a protective absorber. During 14-days' pos t-exposure observa tion, an increased activity of asparagine, ami not ransferase and In some animat alanine aminotransferase was found. Moderately increased pathomorphologica1 changes were found In lungs (emphysema, oedema Interstitial and bronchogen1c Inflammations!, along with slight degenerative changes in the liver and kidneys. In the peripheral blood. increased amounts of erythrocytes and 1eukocytes were found. Tags: Animal; Comparative Study; Male Descriptors: A1r Pollutants. Occupati ona1 --Tox i c i ty ^TO): Bronchopneumon i a--Chemica11y Induced (Cl); `Coal Min1ng--Instrumenta11 on (IS): `Heat--Adverse Effects (AE ) ; Leukocytos1s--Chem1ca11y Induced (Cl ) ; Lung-- Pa tho 1ogy ( PA); Po1ycythernia--Chem1ca11y Induced (Cl): `Polyvinyl Chi or 1de--Tox1c1ty (TO); >Po1yv1ny1s--Tox1c1ty (TO); Pulmonary Edema --Chemlea 1 Iy Induced (Cl): Pulmonary Emphysema--Chemical- ly Induced (Cl): Rats: Rats. Inbred Strains; `Rubber--Toxicity (TO): Time Factors; Volatilization CAS Registry No.: 9002-86-2 .(PolyvinylChloride):9006-04-6 .{Rubber) / 1852646 86069646 f [Role of acetaldehyde In sudden cardiac death] Rolul aceta1deh1de1 in moartea suolta prln cord. Constantinescu V; Scrlpcaru G; Constantin LV; Manoilescu E; Seripcaru C; Istrate E; Popescu E; Cotutlu C; Stoica L Rev Med Chir Soc Med Nat Iasi (ROMANIA) ,Jan-Mar 1985. 89 / 1840634 86057634 V Aceta1dehyde. IARC Monogr Eval Carcmoci R i sk 36 p101-32, Journa1 Code: GE4 ASI 00004747 (con t nex t Chem Hum page) (FRANCE) .feb 1985.5 gSPdihlog i. i \ 4 1 t