The Need To Supplement Certain Listings In The Biennial Reports On Carcinogens In Order To Avoid Misleading The Public And Harming Public Health

January 28, 1999

A broad, cross-cutting issue of great public importance is how to ensure and maximize the quality and utility of information disseminated by the Federal government under its many programs and authorities, and particularly information intended to inform the general public, such as the Report on Carcinogens Program entrusted to the Department of Health and Human Services and administered by the National Toxicology Program ("NTP") of the National Institue of Environmental Health Sciences.

In May 1995, Congress enacted legislation on Federal information dissemination which established a policy to "ensure the greatest possible public benefit from and maximize the utility of information . . . disseminated by or for the Federal Government." The legislation vested the Director of the Office of Management and Budget with responsibility for overseeing the implementation of that legislative policy by individual agencies¹. Very recently, in the course of enacting the Omnibus Appropriations Act for FY'99, Congress became concerned with the lack of action to implement this policy, and it inserted strong language in the House and conference reports directing OMB to proceed to issue guidance to implement the policies of that legislation by September 1999, and to report to Congress on its progress by the end of the fiscal year². The OMB guidance will require all Federal agencies to issue their own guidance consistent with the OMB guidance within the following year and with public participation.

The current cycle of reviews of listing nominations for the 9th Report on Carcinogens raises serious issues under this policy. The principal issue that we believe should be addressed by the NTP Executive Committee, the NTP Director, and Secretary Shalala before final listing recommendations and decisions are made and disseminated to the public is whether certain listings should be qualified or expanded in a manner that, while remaining consistent with the Congressional mandate for the RoC Program, will greatly improve the accuracy and utility of the listing information for the public.

The Report on Carcinogens listings have always posed a public information challenge because they are hazard rather than risk evaluations, and probably most members of the general public are not aware of this distinction. When something is listed as "known to be carcinogenic to humans" in the Reports, it is very likely to be understood as causing cancer under all circumstances, although that is not what it is intended to convey. We are aware that the Reports attempt to convey the distinction by stating in the preface that "listing of a substance in the Report . . . does not establish that any such substance presents a risk to persons in their daily lives." Yet, the Reports are also clearly intended to provide information that is useful to the public in their daily lives, and they state in the Introduction that they are issued because "[t]he people of the United States have asked, through the U.S. Congress, for information about substances that cause or might cause cancer."

The public information challenge posed by the Reports, then, is how to make the information they provide as relevant and useful to the public as possible. With this in mind, we have followed with great interest the NTP reviews of proposed listings for the 9th Report. As discussed below, we believe that the accuracy and utility of the RoC information could be greatly improved, and unintended consequences of misleading the public could be avoided, by providing additional information in the listings for certain substances in instances where it is clear that providing such information would further the purposes of the RoC Program.

The broad purpose of the Report on Carcinogens program, as clearly reflected in its legislative history materials, is to provide the public and Congress with information on harmful substances so that Congress and Federal agencies can consider ways in which exposure to such substances might be reduced³. As stated in the Reports, it is presumed that any reduction of exposure will decrease the incidence of cancer and thereby improve public health⁴.

This basic premise is not always true, however, because some substances also have beneficial effects and uses, and therefore care must be exercised not to list substances in a manner that might actually cause harm. The reviews for the upcoming 9th Report on Carcinogens have brought to light two prominent examples of listing nominations -- tamoxifen and alcoholic beverages -- in which listing poses such a potential for harming public health rather than improving it if sufficient information is not provided as part of the listing. If scientific accuracy is also considered an important aspect of this public information, the issue is significantly broader and other substances could be given as examples. For present purposes, however, we will restrict the discussion to these two.

With regard to tamoxifen, at the October 1997 public meeting of the RoC Subcommittee it was made clear that both the Subcommittee members and NTP staff recognized the importance of tamoxifen as a beneficial anti-cancer drug approved by FDA. Because of this, while they have recommended that tamoxifen be listed as "known to be carcinogenic to humans" for endometrial cancer, it was also decided -- for the first time so far as we are aware -- to recommend that the listing be expanded to state that tamoxifen has also been demonstrated to be valuable for the treatment of contralateral breast cancer. It is important to note, though, that, while the Subcommittee seemed to be under the impression that this expanded listing language would convey that the anti-cancer benefits on tamoxifen greatly outweigh the cancer risks, the proposed listing language does not attempt to convey such benefit/risk relationship information in a manner that would clearly inform breast cancer patients and other concerned parties. As the proposed tamoxifen listing now stands, therefore, there is a clear potential for the listing to cause more harm than good by deterring women from a medical treatment that is known to have a net beneficial effect.

A somewhat similar situation has been presented by the proposed listing for alcoholic beverages, as evidenced by the public discussion at the December 1998 RoC Subcommittee meeting. The extensive evidence for beneficial health effects of alcoholic beverages when consumed in moderation was noted, and some members of the Subcommittee indicated discomfort with the implications of listing a substance as harmful under such circumstances. The possibility of adding some qualifying language to the listing was discussed briefly, and the tamoxifen listing proposal was even brought up as an example of an expanded listing, but the Subcommittee members appeared to conclude that their charge only allowed them to vote either yes or no on the proposed listing language and that they could not qualify the listing to indicate that cancer hazard was only "known" to exist at higher levels of consumption or was associated with an alcoholic lifestyle⁵. Also, as one Subcommittee member noted during this discussion, there were a number of other substances already listed in the Reports that arguably should be qualified to convey more accurately the nature and degree of hazard.

We understand that the NTP Director, Dr. Olden, has taken note of these issues, and has even written to the manufacturer of tamoxifen to express his opinion that the listing proposal raises interesting and difficult communication issues that should be brought before the NTP Executive Committee. We concur. Doing so is particularly important at this time because the Executive Committee review stage is apparently the first opportunity in the RoC review process for FDA to participate substantially, and FDA, and a U.S. delegation as well, have apparently taken the position in the OECD hazard harmonization negotiations that pharmaceuticals should be dealt with differently from other chemical exposures because they are subject to important risk/benefit considerations that could make hazard listings misleading and because they are regulated differently⁶.

While these are two current and important examples of the information communication issues raised by the RoC listings, as we noted previously the issues are broader. The basic issue is whether RoC listings should be qualified for some substances to provide, clearly but briefly, more information on the nature and degree of the hazard in a manner useful to non-expert members of the public.

We believe there are a variety of options for addressing this issue, but before we attempt to delineate them, we believe it necessary to discuss briefly the issue of the scope of the Secretary's discretion to address this issue under the legislation establishing the Report on Carcinogens program.

Currently, the NTP staff administering the RoC Program and the RoC review committees appear to be operating under the assumption that their authority is restricted to answering simply yes or no as to whether a substance is either "known" or "reasonably anticipated" to cause cancer, and that they do not have authority to recommend that listings go beyond this. We believe that a review of the legislative history materials will show clearly that Congress did not intend to confer such cramped authority (and the recent recommendation to expand the tamoxifen listing is, we believe, within the law and beneficial). In actuality, the legislative history materials suggest that Congress intended the listings to be considerably more informative than they are currently.

In particular, the legislative history materials indicate that the Reports should provide "where possible, estimates of the magnitude of the risk each [substance] poses", "the relative risk posed by each substance", the "relative toxicity" of an agent, and information on "subpopulations expected to be at higher than average risk". Neither the legislation nor its legislative history indicate an intent to restrict the pertinent information that can be provided in a listing in addition to a designation as "known" or "reasonably anticipated". Also, the legislative history as a whole indicates an intent that the Reports should provide information to assist in reducing risks from "harmful" substances in the environment. This intent is confirmed by the language of the legislation itself, which requires that the Reports provide information on the extent to which implementation of regulations "decreases the risk to public heath from exposure to the substance". Thus, the legislation and evidence of Congressional intent clearly raise the issue of how the Reports should address an exposure that appears to have substantial beneficial health effects under certain circumstances, or for which the evidence of carcinogenicity indicates that adverse health effects are restricted to certain circumstances.

In practice, the Reports have not always stuck to a strict division of listings into only two categories based on the "known" and "reasonably anticipated" classifications. In the Seventh Report (1994), a third category was added for certain "Occupational exposures associated with a technological process". In the 8th Report, this third category was not used, but what was effectively a third category was added in the form of an appendix for certain occupational exposures which had been reviewed by IARC but not by HHS, and it was stated that this information was provided "[i]n the interest of health and for completeness". Thus, there are instances in which HHS has previously recognized that there are circumstances under which the Reports should provide additional listing information in order to promote the Program's fundamental purpose of improving public health and providing accurate and useful information.

As noted earlier, we agree with NTP Director Dr. Olden that this issue of providing supplemental qualifying information to listings to avoid the likelihood of misleading the public or harming public health should be addressed by the NTP Executive Committee. To do so is consistent with the Report of Carcinogens legislation and with recently-expressed Congressional policy on the quality and utility of information disseminated by Federal agencies. The issue should also be addressed by the NTP Director and Secretary Shalala. In addressing the issue, we recommend that the following options be considered at the next NTP Executive Committee meeting:

1. Utilize the Recommended "Reviewed But Not Listed" Category: In essence, a separate portion of the 9th Report has already been recommended which could provide a framework for providing public information while HHS/NTP moves ahead to address the broad issue more thoroughly. Two of the three RoC scientific review committees have recommended that information on potential hazards from boot and shoe manufacturing be set out in a new appendix entitled "Reviewed But Not Listed", so that existing information can be provided without definitive listing until additional scientific and technical information is obtained⁷. Information on agents such as tamoxifen and alcoholic beverages could be presented in this portion of the Report along with an explanation of the issues posed by potential listing and the advisability of waiting for OMB guidance (due by September 1999) and public input.

2. Add a new listing category for "Qualified Listings": Agents would be listed in this category when HHS determines that it is in the public interest and consistent with Congressional intent to do so. This special category, while still utilizing the statutory hazard descriptors, would provide supplemental listing language designed to more fully inform the public in order to avoid misleading impressions and adverse impacts on public health.

3. Defer action to the 10th Report: The issues, and action being taken to address them, would be described in the Introduction, but no listing conclusions or detailed scientific information would be provided in the body of the Report as in the first two options.

We recommend utilizing either option 1 or 3 as a short-term approach, and considering option 2 as a long-term approach in view of the need to obtain public comment and consult with OMB. Whether option 1 or 3 would be utilized in the short-term would depend on the information available for a particular substance -- for some, option 3 might be more appropriate than option 1 if there is uncertainty about the relevant information or how to best characterize it.

We do not see any legal impediment or public interest detriment from any of these options, since the Report would still provide pertinent information, while striving to achieve consistency with the goals of improving public health and the clear Congressional policies of ensuring and maximizing the quality and utility of information provided to the public by Federal agencies, and allowing opportunities for public notice and comment when issues such as this arise. We believe that public notice and comment on this issue is particularly important since Executive Committee meetings are not open to the public.

¹ Public Law 104-13, 109 Stat. 169, 35 U.S.C. 3501 et seq. particular emphasis on issues that involve multiple agencies.

² H.R.Rep. No. 592, 105th Cong., 2d Sess. at 49-50 (1998), and conference report Joint Explanatory Statement, Cong. Rec. Oct. 19, 1998 at H11508, 1st col.

³ H.R.Rep. No. 1192, 95th Cong., 2d Sess., 28, 45 (May 15, 1978); 124 Cong. Rec. H34938, Oct. 10, 1978;124 Cong. Rec. H13566, Oct. 14, 1978.

⁴ 8th Report on Carcinogens at 5 (1998 Summary).

⁵ The expanded listing for tamoxifen was proposed to the Subcommittee in the nomination (i.e., it had been recommended by RG1 and RG2), whereas the nomination for alcoholic beverages presented to the Subcommittee did not contain any additional language. The Subcommittee did recommend modifying the narrative discussion in the Report, which is separate from the listings and is more technical and far less likely to be consulted by the public, to indicate that risk is more pronounced at higher levels of consumption. But the fact that this information would be "buried" in the body of the report and would not be made a part of the listing clearly bothered some Subcommittee members. It was also mentioned during the discussion that this issue had been raised during the first non-public internal review of the listing proposals (the RG1 review).

⁶ FDA has also expressed the view, outside of the Report on Carcinogens Program, that it does not consider the evidence sufficient to classify tamoxifen as a "known" human carcinogen. That FDA scientists familiar with specific nominated drugs apparently do not participate substantially in the RoC scientific reviews prior to the Executive Committee stage raises another significant process issue of whether the RoC review committees comprise sufficient relevant expertise for recommending such important public health decisions. By the time a listing proposal reaches the Executive Committee, it has undergone three separate scientific committee reviews, and the listing recommendation might therefore by that time be effectively a fait accompli. Additionally, we note that the legislative history of the RoC mandate never mentions drugs as types of substances intended to be covered, although it specifies numerous other types of exposures. Nevertheless, HHS has interpreted the legislation to cover drugs for many years, based on a broad 1981 OTA interpretation of "environmental" exposures.

⁷ The RG1 and RG2 committees recommended the use of this new appendix. The RoC Subcommittee of the Board of Scientific Counselors, the third scientific review committee, did not have an opportunity to vote on a motion to utilize this approach, and instead voted in favor of a motion to defer action.