MBS Letter to the San Francisco Board of Supervisors and the Oakland City Council

December 29, 1998

San Francisco Board of Supervisors
401 Van Ness Ave., Rm. 308
San Francisco, CA 94102

Oakland City Council
City Hall
One City Hall Plaza
Oakland, CA 94612

Dear Supervisors and Council members:

Our firm follows and analyzes many Federal regulatory issues that have a public health or environmental dimension for various entities, including local governments and other entities with an interest in effective waste disposal. More generally, we have an active interest in the integrity of the scientific processes utilized by government agencies, and in maintaining scientific objectivity and absence of bias when assessing public health issues. One of those issues has been risk assessment of dioxin and related compounds and related regulatory proposals. Recently, we were asked to review for accuracy the proposed resolutions on dioxin, public health and the environment that are currently before the city governments for San Francisco and Oakland.

Broad Issues

It is clear even from a fairly cursory review that many of the principal scientific/health statements in the Whereas portions of the resolutions (which are almost identical) contain substantial inaccuracies, omissions, and out-of-date information. The scientific literature on dioxin and related compounds is extremely voluminous, and we will attempt to address only the more prominent issues raised by the resolutions, with the recommendations that the health issues be subjected to a more detailed review in a scientific forum by experts with appropriate scientific credentials and objectivity. We note that the relevant health issues are expected to soon be brought before such a forum, when the U.S. EPA submits a revised human health risk assessment to its Science Advisory Board.

Before focusing on specific statements in the resolutions, we note by way of general background that despite the huge amount of literature on dioxin (1) there is almost no data concerning health effects from dietary exposures, and almost all of the human health effects data comes from very highly-exposed workers in the chemical industry or the military who were exposed to complex mixtures of various substances (such as the production processes for chlorophenols) in which dioxin was present only as an impurity, and other substances to which the workers were exposed have been classified as potential carcinogens: (2) the data concerning exposure to dioxin outside of the complex chemical production environment comes from animals that were administered very high doses (in comparison to human dietary exposures) of pure dioxin; (3) dioxin and related compounds belong to a class of chemicals called receptor binders, which are notorious for variability in biological outcomes depending on species; (4) data relevant to potential human health effects of dioxin-related compounds (i.e., non-TCDD substances) is sparse and controversial, and the risk assessment approach of combining exposures from all dioxin-related substances to achieve a TEQ (toxic equivalency to dioxin) has a very high degree of uncertainty. There is very little about human health effects of dioxin that is known with any degree of certainty, other than that high human exposures can cause chloracne (a type of skin rash)¹, and that high doses given to animals can cause reproductive/developmental effects and increases in some types of cancer (and apparently reductions in other types at lower dosage levels).

The most obvious error in the resolutions is their repeated reliance on material contained in the U.S. EPA's 1994 preliminary ~ re-assessment of the human health risks of dioxin and related compounds. That re-assessment was undertaken in recognition that previous assessments were outdated. The draft which is cited and relied on in the resolutions states plainly on its cover: "Review Draft (Do Not Cite or Quote)".² After the draft was released for public comment, it was submitted to the Agency's Science Advisory Board, where it was reviewed by a special Dioxin Reassessment Review Committee comprised of 39 dioxin experts. 25 of whom were experts in health effects. The result of this review was that in September 1995, the Science Advisory Board rejected the EPA draft re-assessment.³ Two of the major flaws for which the SAB rejected the re-assessment were (1) that the draft was "not balanced", "with a tendency to overstate the possibility for danger", and (2) that "important uncertainties" were "not fully identified". (P.5 of SAB transmittal letter to EPA Administrator.) Since then, the U.S. EPA has been working on a revised re-assessment to resubmit to the Science Advisory Board. The current expectation is that this will occur sometime in 1999. The difficulties EPA is having in finalizing a re-assessment of risks demonstrate the great uncertainties involved. Despite these uncertainties and a clear lack of up-to-date and peer-reviewed U.S. EPA health risk estimates, both the San Francisco and Oakland resolutions repeatedly employ health effect levels purportedly developed by EPA, and they make no mention whatsoever of the Science Advisory Board rejection of those estimates and the current EPA efforts to develop a more scientifically supportable re-assessment. It is hard to believe that individuals familiar with the scientific data on dioxin were not aware of the above facts, all of which have been widely publicized.

For purposes of the central health issues involved in the two resolutions, it appears that the most significant conclusion reached by the EPA Science Advisory Board was its final one:

If it is intended to state [in the draft re-assessment] that adverse effects in humans may be occurring near current exposure levels, it is the Committee's judgment that EPA has not presented findings that support this conclusion adequately.

The SAB statement quoted above really goes to the core of the San Francisco Bay issues: What levels of contamination present a significant health risk? A multitude of "contaminants", including many natural substances in the environment and the human diet, may be considered a health danger at excessive levels, but harmless at normal exposure levels. In the case of dioxin, findings of potential adverse health effects derive almost exclusively from extremely high human exposures due to work in chemical factories or accidents, or from high-dose animal experiments. For example. in the recent NIEHS review of the potential cancer hazard (not risk) of dioxin, discussed below, the assessment focused almost exclusively on several relatively small groups of-very highly exposed chemical workers that had been isolated from larger study populations; and assertions regarding the "endocrine disruption" potential of dioxin have frequently focused on certain animal experiments in which the dose rate was many thousands of times higher than normal human exposures. On the other hand, advances in analytical chemistry have enabled us to detect dioxin in quantities as minute as fractions of parts per trillion (e.g., picograms), and even when dioxin has not been detected at these microscopic levels, it has often been assumed that it is present at a level at least one-half of the detect level.⁵

In connection with the expected submittal of a revised EPA dioxin re-assessment for further SAB review in 1999, we also note that the California Office of Environmental Health Hazard Assessment has also given notice that it expects to complete a draft risk assessment for dioxin in 1999.⁶ It seems appropriate to question whether local government officials who do not have expertise in human health risk assessment and the dioxin literature should be drafting and considering resolutions containing asserted scientific findings while the same health risk issues are still under review by California OEHHA and the U.S. EPA and will be reviewed by a large expert panel of the EPA Science Advisory Board.

Some Specific Inaccurate and Misleading Assertions in the Proposed Resolutions

Some more specific inaccuracies in both proposed resolutions include the following:

Carcinogenicity

• The statement that dioxin is a "known human carcinogen" is inaccurate. On December 2, 1998. the expert outside peer review committee of the Board of Scientific Counselors for the National Toxicology Program concluded that dioxin should ~ be classified as a "known human carcinogen". The resolutions contain references to an October 1997 review by the same committee that was later voided because it was determined that the review might have been inadequate, and the matter was scheduled for re-review at the recent December 1998 meeting.⁷ Even if the referenced 1997 vote had remained in effect, that vote was only one intermediate step in the Report on Carcinogens process administered by the National Institute of Environmental Health Sciences.

• Perhaps more importantly, the statements that dioxin is a "known human carcinogen" and is "believed to be the single most carcinogenic chemical known to science" could be tremendously misleading in implying that dioxin is known to cause cancer in humans at some level of exposure that is relevant to exposures resulting from San Francisco Bay and other everyday environmental exposures (as opposed to extremely high exposures used in animal experiments or accidental or unusual occupational exposures). Determinations such as those made by the National Toxicology Program and IARC are what are called "hazard" determinations, as opposed to "risk" determinations. Hazard determinations do not take level of exposure into account; basically they mean that an agent may, or is known to, cause cancer at some level of exposure, often extremely high compared to normal human exposure. For example, the Preamble to the recently-released 8th Report on Carcinogens⁸ states: "Risk assessments are not conducted by the National Toxicology Program for substances in the Report. The listing of a substance in the Report, therefore, does not establish that any such substance presents a risk to persons in their daily lives."

• Expert risk estimates for dioxin made more recently than EPA's 1994 draft and the 1995 EPA Science Advisory Board review indicate that the lowest level for cancer risk is well above normal environmental exposure levels. In 1996, a 20-member "International Expert Panel", convened by the American Health Foundation with support from the National Cancer Institute, which had reviewed quantitative cancer risk data on 10 chemicals, including dioxin, from 1990-95, published their findings. With regard to dioxin the Panel wrote:

  Further research is required to determine whether the tumorigenic effects of [dioxin) in rodents apply to humans. [T]he mechanisms producing tumors in rodents may not be effectively operative in humans at any plausible exposure level.⁹

  In a detailed analysis of dioxin cancer data accompanying the Panel comments, and reviewed by the Panel, it was concluded that the exposure levels below 0.01 microgram/kg/day indicated almost exclusively anti-cancer effects¹⁰, and no effects, either beneficial or potentially adverse, could be seen below 0.001 microgram/kg/day.¹¹ A level of 0.01 microgram/kg/day is more than 10,000 times higher than daily human exposures of 0.3-0.6 picograms/kg/day estimated by EPA in its 1994 draft re-assessment. More recently, a panel convened by the World Health Organization decided that 1-4 picograms/kg/day could be considered a "tolerable daily intake" (WHO's equivalent of a "safe" dose) for all health effects, a level which incorporates substantial safety factors. WHO also emphasized that recent exposure data indicated that there had been substantial declines in human intake during the last several years.¹² This point was also emphasized in the EPA SAB review. Therefore, the levels of human intake reflected in the 1994 EPA draft re-assessment are likely to be outdated and significantly overstated.¹³

"Endocrine Disruption"

• Statements regarding "endocrine disruption", like references to "carcinogenicity", are misleading in a context where risk at normal exposure levels is implied. Both terms are hazard rather than risk terms, meaning that they do not take level of exposure into account. Under all circumstances, the maxim of toxicology that "the dose makes the poison" applies. The word "disruption" is also misleading, since it implies damages, when in many cases the data only indicate some change in biological parameters rather than an adverse effect. A vast number of substances, including a multitude of components of ordinary foods, such as apples, soybeans and rice, affect human and animal endocrine systems and could also be termed "endocrine disruptors" if exposure levels are ignored.

• As indicated by the discussion above, as with the issue of carcinogenicity, it is the level of exposure that is important in defining "endocrine disruption".¹⁴ As noted previously, the recent WHO-determined "safe" level, which incorporates substantial safety factors, is above even normal exposure levels that are likely outdated and too high. Much has been made of potential reproductive/developmental effects of dioxin. Such effects were considered in the WHO determination. In addition, the referenced EPA 1994 draft risk re-assessment. and the unreferenced Science Advisory Board review of that re-assessment, indicated a likely lowest adverse effects level of somewhere between 125 and 1.000 picograms/kg/day, but with substantial uncertainties.¹⁵ For example, the SAB (and many commentators) have focused on the experiments of Mably et al., which indicated reduced spermatogenesis in rat offspring following administration of a single acute dose of 64 ng/kg to the mother on a single day of pregnancy. Such a single-day dose is many thousands of times higher than a normal human daily exposure to dioxin.

• Assertions in the proposed resolutions that dioxin has been "linked"¹⁶ to various non-cancer adverse health effects, in addition from suffering from the distortions discussed above, also clearly contain a number of outright inaccuracies or very questionable assertions. For example, the reference to "testicular atrophy", for which no supporting reference is given, appears to be a reference to a mythical finding in the Air Force Ranch Hand Study. This myth was laid to rest in a 1996 peer-reviewed publication.¹⁷

Positions of State and Local Government Agencies

• The resolutions fail to mention that the California State Water Resources Control Board declined to include dioxins in its June 1998 recommendations for listing of pollutants under Section 3 03(d) of the Clean Water Act.

• On the other hand, the resolutions state erroneously that "the San Francisco Bay Regional Water Quality Control Board has resolved that dioxin is a high priority for immediate action to restore water quality and protect public health". We contacted the Board and were informed that such a resolution was never passed and no such document exists; the subject was discussed, but the Board was informed by its legal counsel that no action could be taken on such a resolution because it was not given public notice on the meeting agenda.

• The resolutions also state that dioxin contamination in fish reaches health advisory levels throughout the San Francisco Bay. The reference given in support of this statement is a December 1994 "interim" advisory issued by the Office of Environmental Health Hazard Assessment (OEHHA) of the California EPA. This interim advisory stated explicitly that it was based on only a preliminary review of data and that more specific recommendations would be issued when a thorough evaluation of the study data was completed by OEHHA in conjunction with other public agencies. In a May 27, 1998 memorandum to the State Water Resources Control Board, OEHHA stated that the December 1994 interim advisory was based mainly on concerns regarding PCBs and mercury, and that they would not be using current data on tissue levels of dioxins in fish "as a basis for issuing fish consumption advisories" due to the inadequacy of the data. In the same memorandum, OEH.HA informed the Board that it was working on an in-depth risk assessment for the purpose of revising the interim advisory as needed, and that a draft reported was expected to be completed soon. None of this is mentioned in the proposed resolutions, and one is left with the impression that State and local agencies have determined that dioxins in fish represent a health threat.

Conclusions

• There are substantial inaccuracies and misleading statements in the proposed resolutions.

• There are also major omissions of relevant facts and data.

• It is clear that the proposed resolutions are not based on a thorough and objective review of the relevant data.

Recommendation: Issues of scientific health effects analysis should be removed from the political arena. The State, local, and Federal health scientists should be allowed to do their job by completing their ongoing risk assessments.

Thank you for considering these comments.

Sincerely,
Jim J.Tozzi
Director

¹ See the EPA Science Advisory Board report. infra, at 15 (quoting from the EPA charge to the Board, which stated "Chloracne is the only clearly adverse health effect which is known to occur in dioxin exposed humans.")

² The proposed San Francisco Board of Supervisors resolution does not contain references, but it is clearly based on the September 8, 1998, San Francisco Commission on the Environment resolution which does contain references. (Resol. No. 021-98-COE). In the COE resolution, there appears to be a reference to the U.S. EPA 1994 draft risk re-assessment in endnote 5: however, the reference includes a curious reference to the Bureau of National Affairs, and there is no reference anywhere to the EPA Science Advisory Board review.

³ "An SA.B Report: A Second Look at Dioxin (Review of the Office of Research and Development's Reassessment of Dioxin and Dioxin-Like Compounds by the Dioxin Reassessment Review Committee" (EPA-SAB-EC-95-021, September 1995). In 1996. EPA issued a separate cancer risk assessment for PCBs, which had been included in the 1994 dioxin re-assessment as dioxin-like compounds.

⁴ See the Letter to the Editor on "The Dose-Response Model for Dioxin" from 20 academic and non-profit experts in receptor-binding substances in Risk Analysis, Vol. 18, No. 1, pp. 1-2, 1998.

⁵ Memorandum dated May 27, 1998, from GA Pollock, OEHHA, to Walt Pettit. Executive Officer of the State Water Resources Control Board.

⁶ Id.

⁷ The reference to a 1997 IARC review is very problematical, as reflected by the NIEHS review. First, IARC does not use the term "known". Second, the referenced IARC working group determined that there was not sufficient evidence of carcinogenicity in humans, and made inferences based on hypotheses concerning dioxin's biological mechanism of action, when the IARC criteria should have allowed such considerations only when the knowledge of mechanism is very convincing. Finally, the decision was made by a relatively narrow majority vote (14-10), with the voting being dominated by a group of U.S. agency scientists. Since then, even IA.RC staff present during the deliberations have indirectly criticized the vote in a peer-reviewed article as not being based on actual scientific knowledge that the dioxin mechanism of action leads to cancer. McGregor DB, Partensky C, Wilbourn J, and Rice J, "An IARC Evaluation of Polychlorinated Dibenzo-p-dioxins and Polychlorinated Dibenzofurans as Risk Factors in Human Carinogenesis", Env. Hlth. Persp. 106, Supp. 2:755-60, at 758 (April 1998).

⁸ U.S. Dept. of Health and Human Services, National Toxicology Program (1998).

⁹ International Expert Panel on Carcinogen Risk Assessment, -"The Use of Mechanistic Data in the Risk Assessment of Ten Chemicals: An Introduction to the Chemical Specific Reviews", Pharmacol. Ther., Vol. 71, Nos. 1/2, at p. 4(1996).

¹⁰ See also the 1995 EPA SAB report at 87, and "Chemical Hormesis: Beneficial Effects at Low Exposures, Adverse Effects at High Exposures", Insights, Vol. 8 (No. 1), at 37-39 (1998, Tex. Inst. for Advancement of Chem. Technology).

¹¹ Whysner J and Williams GM, "2,3,7,8-Tetrachlorodibenzo-p-Dioxin: Mechanistic Data and Risk Assessment: Gene Regulation, Citotoxicity, Enhanced Cell Proliferation, and Tumor Promotion", Id. at 217.

¹² WHO Press Release WHO/45, June 3, 1998.

¹³ The proposed resolutions also state that dioxin is thousands of times more potent as a carcinogen than DDT. In addition from suffering from all of the inaccuracies discussed above with regard to the possible carcinogenicity of dioxin, DDT has not been recognized as a human carcinogen, as is clearly implied by this statement. The 8th Report on Carcinogens issued by the Department of Health and Human Services (1998), reports DDT to show "sufficient" evidence of carcinogenicity in high-dose animal experiments, but "inadequate evidence" of carcinogenicity in humans. At 86-87. DDT was banned for most agricultural uses in the U.S., but an exception remains for health-related purposes, since DDT is still the most effective pesticide for controlling malarial mosquitoes, and its continued use in developing nations has been credited with saving hundreds of thousands, if not millions, of lives. As a health control measure, DDT is sprayed on the interior walls of dwellings.

¹⁴ This is especially so in the case of dioxin, where advances in analytical chemistry have enabled detection down to fractions of a parts per trillion (or less than a millionth of a millionth).

¹⁵ EPA SAB report, supra, at 68.

¹⁶ This is a vague, and potentially misleading, term, since it could mean anything from an non-expert assertion reported in a news article or in an advocacy organization pamphlet to a convincing relationship observed in scientific studies. This term should be banned from any intelligent discussion of health effects data.

¹⁷ Hennksen GL et al., "Serum dioxin, testosterone, and gonadotropins in veterans of Operation Ranch Hand", Epidem. 7:352-57 (1996).