New Rap Against Fluoride: P&G Study

By Joel Griffiths

A hitherto unpublicized study of fluoride carcinogenicity in rodents, conducted in the laboratories of Proctor & Gamble (P&G) in the early 1980s but as yet unpublished, may carry significant weight in the Environmental Protection Agency (EPA) upcoming deliberations on whether fluoride should be classified a carcinogen.

The P&G study seems to be the only other rodent bioassay of fluoride carcinogenicity comparable to the widely publicized - and apparently positive - study by the National Toxicology Program (NTP, part of the Department of Health and Human Services). Company representatives term P&G's results "completely clean" - that is, free of significant malignant neoplasms. However, partial results are available, and discussions with oral pathologists, toxicologists, and risk-assessors suggest that some of the pathological changes observed could be interpreted as precancerous and that P&G's results may actually not conflict with the NTP's.

Meanwhile, official release of pathology results from the NTP study confirm the possibility of a causal link between fluoride and bone cancer (and perhaps oral cancer as well) - as first reported by this newspaper (MT, Dec. 28, 1989). For a discussion of the potential impact of the NTP's findings on the future of fluoride, see the accompanying article.

During the first half of the 1980s, P&G, a major manufacturer of toothpaste and other fluoridated products, conducted a numer of studies on both the mutagenic and carcinogenic potential of fluoride. This work followed a disturbing 1982 report by Takeki Tsutsui, et al., of the Nippon Dental College in Japan, which suggested that, in vitro, fluoride was both mutagenic and carcinogenic in hamsters. Several in vitro studies were subsequently published by P&G scientists, all of which tended to refute Tsutsui's findings (see, for example, Skare et al: Mutat Res 172:77, 1986). By contrast, P&G's rodent bioassay of fluoride carcinogenicity has not yet been published. Asked why a company spokesperson responded, "Just bureaucracy, I guess."

In June 1985, however, P&G investigator Michael Cheng, Ph.D., did present partial pathology results and other details of the company's study orally before federal scientists from the NTP and other interested agencies. A copy of the notes taken at this presentation was obtained under the Freedom of Information Act by John Yiamouyiannis, Ph.D., a biochemist and longtime water-fluoridation opponent, and furnished to th is newspaper. These data were submitted to P&G for verification and comment.

Like the NTP study, the P&G study bioassay was a two-year study in rats and mice, and the doses appear to have been roughly similar. In his oral presentation, P&G's Dr. Cheng appropriately presented only oral and jaw pathology results. Aside from hyperostosis of the maxilla and mandible - the bone overgrowth well known to be stimulated by fluoride - he reported dose-related increases in dysplasia (Table 1) and in squamous metaplasia, both of the ameloblasts (the tooth-enamel-forming) cells). The question MEDICAL TRIBUNE asked was: What is the neoplastic significance of those types of tissue changes?

In a letter of comment, an anonymous P&G pathologist stated flatly that "... ameloblastic dysplasia and ... squamous metaplasia are not neoplastic or precancerous responses." Rather, they are "terms used to describe the consequences of degeneration of the cells responsible for the production of dental enamel."

'A Precarcinogenic Change'

Others are not so sure. John R. Bucher, Ph.D., chemical director of the NTP study, who was present at Dr. Cheng's June 1985 presentation, told this newspaper that "some of the things they [P&G] found could be considered precursor lesions that might develop into neoplasia."

And David J. Zegarelli, D.D.S., director, stomatology, School of Dental and Oral Surgery, Columbia-Presbyterian Medical Center, New York City, explained that "dysplasia means atypical cells going toward malignancy. Dysplasia implies precarcinogenesis."

"Squamous metaplasia," continued Dr. Zegarelli, "may also be considered a precarcinogenic change, especially in certain contexts. For example, when it occurs along the respiratory tract of cigarette smokers, it signifies possible premalignant change."

"Context" would appear to be the key to interpretation of the P&G results. Summarized Scot Eustis, Ph.D., head of tumor pathology at the NTP, "The terms dysplasia and squamous metaplasia are used to cover a wide variety of cellular changes. In some cases they may be related to neoplasia, and in many other cases they are not. If they appear in an organ known to develop neoplasia, then they may be considered preneoplastic changes. Context is the key to their interpretation." Such a context should be provided this month with the scheduled release of validated pathology tables of the actual malignancies that occurred in the NTP study.

Added Ian Nisbett, Ph.D., an independent risk-assessment consultant in Lincoln, Mass.: "Taken by themselves, dysplasia or squamous metaplasias of ameloblasts could be neoplastic, but without actual neoplasms in the same study, it's so marginal you can't say Yes. Rather, your suspicions would be aroused that you were dealing with a weak carcinogen that might produce actual cancers if the study population were larger." P&G declined to specify the size of its study population or to compare it with the NTP's.

'A Gray Area'

"There's a genuine difference of opinion among pathologists on whether you can identify precancerous lesions," continued Dr. Nisbett, who works extensively with pathology data. "In the case of fluoride, there is a gray area between the enamel overgrowth you expect to get with high doses and a fullblown cancer. Dysplasia and squamous metaplasia are in that gray area, and which end of that area depends on the pathologist and his preconceptions."

P&G's letter also emphasized that "The changes... observed in the [P&G] study are not unique and have been reported in other published sodium fluoride studies." Several earlly investigations of high-dose fluoride toxicity from the 1930s and '40s were cited as examples, studies that did describe some apparently similar tissue changes. However, none were carcinogenicity studies, and none precluded the possibility that such changes were premalignant, Dr. Nisbett and others agreed. As the authors of one cited paper - from the pre-water-fluoridation days - remarked in conclusion: "Recent experimental as well as statistical surveys of clinical material indicate the apparently beneficial role of fluorine and the ensuing pathological changes when administered strongly suggest that very careful investigation and rationalization of findings be borne in mind to offset the misfortune that often follows the publications on 'cure-alls'" (Hoffman MM et al: J Dent Res 21:157, 1942).

P&G said it would present its final results soon to the EPA and the Food and Drug Administration and expected to publish them later this year.