Executive Summary

Benzoyl peroxide should be allowed to remain on the market as an OTC treatment for acne pending the outcome of ongoing carcinogenicity studies, FDA's Dermatologic Drugs Advisory Committee unanimously recommended March 10. FDA brought the carcinogenicity issue to the advisory committee because some animal tests have shown benzoyl peroxide to be a tumor promoter in rodents. The agency reclassified the ingredient from Category I (generally recognized as safe and effective) to Category III (insufficient data available to classify as safe or effective) in August, saying "further study is necessary" to determine if benzoyl peroxide is tumorogenic ("The Pink Sheet" Aug. 12, 1991, T&G-8). While the committee voted four-to-three with one abstention to recommend against including a warning about benzoyl peroxide's potential carcinogenicity in OTC labeling, the committee nonetheless ended the meeting by unanimously recommending that some form of warning be included in labeling about the current lack of information on the ingredient's safety. The committee did not specify what form or language that warning should take. A Nonprescription Drug Manufacturers Association task force began dose-finding studies in mice on March 25 in preparation for a two-year carcinogenicity study in response to FDA's request for more data, NDMA Senior VP and Director of Science and Technology William Soller, PhD, told the committee. The study results should be available in "about three years," he said. The final protocol is still being discussed. The committee heard reviews of the animal and human carcinogenicity data on benzoyl peroxide from both FDA and NDMA consultants. Rich-Vicks' Albert Kraus, PhD, section head for human safety/toxicology, agreed that animal data on benzoyl peroxide shows the ingredient "promotes rodent tumors" but said the results were found "under conditions that lack relevance to human use," including high doses, acetone carriers for the benzoyl peroxide and "constant administration" to rodent skin. Because of these factors, he said, "the usefulness of tumor promotion data in [human] risk assessment has not been determined." Anthony Williams, MD, University of Toronto, reported on a Canadian epidemiologic study that showed "no increase in risk" of facial skin cancer between people using benzoyl peroxide products and those not using them. The study was undertaken in Saskatchewan province and involved 964 skin cancer victims and 3,856 controls. Conducted at the request of Canada's Health Protection Branch, the study was published in The British Journal of Dermatology in 1991. FDA Chief of Epidemiology Bruce Stidel, MD, called the Canadian study "a great epidemiological study" that showed "no increased risk in this population when [benzoyl peroxide] was used in this way." He noted, however, that he would like to see an analysis the "risk of skin cancer in relation to duration of use and time since completing various durations of use." Without that, "one might not get the whole story." While no committee member felt that benzoyl peroxide products should be taken off the market, the panel was divided on whether a warning on the tumor promotion results should be included in labeling. FDA Division of Anti-Infective Drugs Director Murray Lumpkin, MD, said: "We don't want to unnecessarily scare the American population." Committee member Joseph McGuire, MD, Stanford University, said that adding a warning that benzoyl peroxide might cause cancer "would be confusing, burdensome, and misleading." "The consumer deserves to know about this," countered committee member Pearl Grimes, MD, Department of Internal Medicine, Los Angeles. The committee concluded by voting unanimously that FDA should include some unspecified warning in labeling because "the bottom line is that the consumer should know something about these findings," invited-guest Harold Minus, MD, Howard University, Washington, D.C., commented. FDA Office of OTC Drugs Acting Director Paula Botstein, MD, suggested that benzoyl peroxide use along with "other [skin cancer] risk factors," such as sun exposure, sunburn or severity of acne could be added to the labeling.