Tobacco Documents Online

Comments on EPA External Review Draft $=Qd, 5/17/90: HEALTH EFFECTS OF PASSIVE SMOKING: ASSESSMENT OF LUNG CANCER AND RESPIRATORY DISORDERS IN CHILDREN (EPA/600/6-90/006A) NOTE: These comments focus QD],y on the health implications of ETS exposure relative to lung cancer, as developed in the subject draft report (Draft Report). Human health protection has been, and is an important objective of several federal agencies. Cancer, and the regulation factors contributing to the incidence of its various forms, have occupied focal positions for actions undertaken by several key federal agencies, including the United States Environmental Protection Agency (EPA), the United States Occupational Safety and Health Administration (OSHA), the United States Food and Drug Administration (FDA), and the United States Consumer Product Safety Commission (CPSC). Significant interagency dialog examining the basis for carcinogen action and the development of adequate methodology to assess risks related to associated chemical exposures began in the 1970s. It culminated with a 1979 Interagency Regulatory Liaison Group policy (described in OTA, 1988) which played a key role in the ultimate development of EPA's "Risk Assessment Guidelines," published in 1986 (US EPA, 1986). The cancer assessment guidelines currently used by EPA were formally stated there, and risk assessment techniques, as used by the Agency, were put forth. [It is important to note that subsequent to the 1986 publication, EPA hasheld several workshops to evaluate practical application of the guidelines. The most recent (US EPA, 1989) discusses the use of human data for assessing risk and raises issues supportive of changing the current approach.] The EPA carcinogenic risk assessment guidelines recommend a'weight-of-evidence" approach for evaluating information relevant to classifying substances as carcinogens. The approach has four steps: 1) hazard identification (a qualitative review of relevant material to determine if an agent poses a cancer threat); 2) dose response assessment (a quantitative evaluation of the potency of an agent in causing cancer); 3) exposure assessment (a

quantitative estimation of exposure to a cancer-causing agent); and 4) risk characterization (a quantitation of the possible cancer threat to people, and a judgment of the significance of that quantitation). The successful application of these guidelines for the evaluation of complex mixtures poses significant challenges. Each step becomes difficult to perform, Md substantiate. I believe that each of the four steps EPA identified for carcinogenic risk assessment has inherent flaws in its application to ETS. I will briefly discuss my major concerns related to each step, on the following pages. Hazard Identification The decision by EPA to classify ETS as a Group A (known human) carcinogen rests principally upon epidemiologic studies. Twenty-one case control studies and three cohort studies are cited in the Draft Report. If we choose to focus only on U.S. studies as being relevant (because of potentially significant cross-cultural biases in different national studies) we have eight case control studies and one cohort study (EPA selects only one U.S. cohort study that it considers to be "major"). In each study, measurement of ETS exposure is indirect. The observed associations between spousal smoking (an indirect estimate of ETS exposure) and cancer effect are extremely weak (odds ratios below 2.0) and none of the U.S. studies is statistically significant. Chance alone cannot be eliminated as an explanation for the data. EPA attempts to overcome the lack of statistical significance by "meta•analyzing" the studies. Meta-analysis cannot overcome the inherent cultural differences in these individual studies. Bias and confounding appear to be poorly controlled for in many studies (I discuss a IIa~m confounding variable lola]y overlooked later in these comments). They cannot be excluded as explanations for the inconsistent, observed elevated, but inconsistent, elevations in risk. No supporting laboratory data were presented in the Draft Report. Shortly after the release of this Draft Report, EPA released another evaluating the potential carcinogenicity of electromagnetic fields (US EPA, 1990). The same EPA office prepared it. Over forty human studies were cited as being available and examining the relationship between electromagnetic (EM) field exposure and cancer incidence. Six were Page 2

cited as showing modest increases in cancer associated with EM exposure. Other occupational EM exposure studies indicated excess cancer risk. Laboratory study results described in the document (EPA, 1990) presented a plausible biological explanation for carcinogenicity in humans. The report concluded: The absence of key information...makes it difficult to make quantitative estimates of risk. Such quantitative estimates are necessary before judgments about the degree of safety or hazard of a given exposure can be made. This situation indicates the need to continue to evaluate the information from ongoing studies and to further evaluate the mechanisms of carcinogenic action and the characteristics of exposure that lead to these effects. No recommendation for the classification of EM field exposure as an EPA group A or B carcinogen was made! Evidence for classifying EM field exposure as carcinogenic is stronger than evidence for classifying ETS as carcinogenic. Application of EPA guidelines produced a different outcome for each of these alleged carcinogens. I submit that EPA needs a consistent approach to hazard evaluation, and such consistency is not evident in the Draft Report for ETS. The epidemiologic data clearly do not provide a sufficient basis for classifying ET'S as a group A carcinogen. os R tponse Assessment Adequate dose response assessment is = conducted in the Draft Report. This is a serious flaw. I will discuss the ]ack of appropriate characterization of ETS constituents, and the effect that has on quantitative risk assessment in the following sections of these comments. No animal data have been assembled that accurately reflect the alleged carcinogenic potency of ETS. Lacking a valid, appropriate animal model, EPA has relied upon imperfect epidemiological data to provide dose response information. EPA points out numerous potential biases in the case control studies and the cohort studies it uses in the Draft Report. Cross cultural confounding variables could pose significant problems for the use of dose response assessment in the U.S. Of the eight case control studies cited, only three were matched for ETS as a study variable. Misclassification of study subjects by smoking history is potentially a major source of concern. The major U.S: based cohort study, failed to show a statistically significant association Page 3

of lung cancer with ETS exposure. EPA clings to the report of mortality ratios of lung cancer deaths being greater than one as an indication that it: "...weakly indicates an increased lung cancer risk from ETS exposure...." This belief exists in spite of the concomitant observation that: "...the data have an observed inversion in dose-response, i.e., lower response ai high e sure Ja =usal smoking ~= at moderate c=sure [emphasis addedl,." There is na biological explanation for carcinogenic behavior of ETS md a protective effect by increased exposure to it. Fxnocnre Assessment ~ ETS is a complex mixture. It differs from main stream smoke and from side stream smoke, as the latter is released from a burning cigarette. Accurate composition information, integrated over potential exposure periods, is needed if realistic risks are to be estimated. Characterization of component chemicals, and their kinetic behavior, within the "smoke matrix" is needed. In the absence of that information, EPA relies upon an indirect approach based upon "cigarette equivalents." This method uses cotinine levels in the urine of non-smokers as an indicator of exposure to cigarettes. Cotinine is then equated in smokers and non-smokers. The "equivalent number of cigarettes smoked by the non-smoker" becomes the number equal to those smoked by a smoker to get the same level of urinary cotinine. No correction is made for the difference between main stream smoke and ETS. No correction is made for the difference in dosing between the two populations (smokers and non-smokers). This treatment is unsatisfactory and a major flaw in the Draft Report. Risk Characterization Information within the Draft Report is synthesized into risk estimates for the development of lung cancer in non-smokers. The cigarette-equivalent approach was carefully qualified by EPA. The Surgeon General (U.S. Surgeon General, 1986), Hoffmann et aL (1989) and an IARC expert panel (Saracci,1989) all conclude that this approach does not give a viable estimate for human ETS exposure. EPA considers Repace and Lowrey's 'phenomenologic" approach to lung cancer risk. They rely upon "background" incidence data Page 4

derived from lung cancer rates in a population of Seventh Day Adventists compared to a group of "matched" non-smokers in the general population. The major problems considered to date with this approach relate to confounding effects from lifestyle differences between the two cohorts. Neither method appears to be viable and neither should be relied upon. The epidemiologic studies do not exclude chance, bias, or confounding. Confounding may play amaj12I role in explaining observed associations. An important finding relevant for all studies relating to lung cancer incidence has just been reported at the most recent American Chemical Society Meeting (August 26-31, 1990) by Mushrush and Mose (Mushrush, 1990). Their study shows a relationship between radon levels in the water supply of private homes in Virginia and Maryland and higher cancer risk in residents (Chemical and Engineering News, 1990). They claimed an increase of five=to-ten-fold increase in well water-related cancer cases over airborne-associated radon in the geographic region of the study. Furthermore, when they turned on the shower, in a closed bathroom, they could measure airborne radon levels at 350 picocuries per liter almost 90 times above the EPA standard for radon! Radon cancer deaths were "estimated" by Repace and Lowrey (1985) to be 3600 per year, compared to 5000 per year-for ET'S. If the water-associated radon deaths found by Mushrush and Mose are Dj31X three or four-fold above air-associated projections, the number of radon related lung cancer deaths could dwarf those allegedly attributable to ETS. Any study which did not control for both air and water-borne radon could be seriously compromised. Conclusion: ' The epidemiological studies used by EPA do not show consistent statistical significance or biologically significant dose response, and since there is no valid animal model for ET'S dose response with a lung cancer endpoint, and there is no accurate exposure characterization data base, and there exists a major confounding variable for human exposure, i.e. radon (in water. and also air), capable of generating meaningless study results (if unaccounted for), I must conclude that EPA does not have sufficient evidence, by its own guidelines, to classify ETS as a human carcinogen. More investigations must be performed to clarify the matter. Page 5

REFERENCES Chemical and Engineering News (1990) Waterborne Radon Linked to Higher Cancer Risk, September 17, 1990, pp. 23-29. Mushrush, G.W. and D.G. Mose, Variation of Indoor Radon Levels Over a'Iwo Summer Period, Presented at the 200th American Chemical Society Meeting, August 26-31, 1990, Washington, D.C. OTA (1988) Identifying and Regulating Carcinogens, Office of Technology Assessment Task Force, Lewis Publishers, Chelsea, Michigan. Repace, J.LL and A.H. Lowrey (1985) A Quantitative Estimate of Nonsmokers' Lung Cancer Risk from Passive Smoking, Environ. jntgmai, 11: 3-22. Saracci, R. (1989) Passive Smoking and Cancer Risk. IARC Report of Panel of Experts. Prepared at the request of the European School of Oncology through the Europe Against Cancer Program of the European Economic Community. U.S. EPA (1990) Evaluation of the Potential Carcinogenicity of Electromagnetic Fields EPA/600/6-90/005A, Workshop Draft. U.S EPA (1989) Workshop Report on EPA Guidelines fo Carcinogen Risk Assessment: Use of Human Evidence, EPA/625/3-90/017 U.S. EPA (1986) Guidelines for Carcinogen Risk Assessment, S1 Federal Register, 33992- 34003. U.S. Surgeon General (1986) The Health Consequences of Involuntary Smoking. U.S. ~ Department of Health and Human Services, Public Health Service. 4M N1 t1t O .A. Page 6 w