FDA arthritis advisory committee meetingPostmarketing surveillance of nonsteroidal antiinflammatory drugs.код для вставкиСкачать
I168 GOVERNMENT AFFAIRS FDA ARTHRITIS ADVISORY COMMITTEE MEETING: POSTMARKETING SURVEILLANCE OF NONSTERQIDAL ANTIINFLAMMATORY DRUGS HAROLD E. PAULUS This communication reports on the FDA Arthritis Advisory Committee meeting of April 29-30, 1985. Between August 16, 1983 and August 24, 1984, 16,911 adverse drug reactions (ADR) associated with nonsteroidal antiinflammatory drugs (NSAIDs) were entered into the United States voluntary adverse reaction reporting system or reported to the manufacturers. Summary tables were prepared by the Food and Drug Administration (FDA) reviewer with the cooperation of the manufacturers, after eliminating multiple adverse reaction terms and cases judged to have only a remote possibility of being causally related to the NSAID. With the understanding that the numerator data consist of reports of adverse reactions rather than the number of adverse reactions that actually occurred, and that the denominator data consist of millions of prescriptions filled rather than number of patients, some of the information presented to the Committee is included here because it is the best available on this subject and because it illustrates some of the problems associated with the voluntary adverse reaction reporting system (Tables 1 and 2). Additional information included summaries of ADRs related to specific organ systems. The FDA reviewer pointed out some “blips” in the data, which were then discussed by the manufacturers. For example, nonfatal acute allergic reactions were reported for tolmetin in 36.52 cases per million prescriptions, whereas each of the other NSAIDs had fewer than 9 ~Harold E. Paulus, MD, American Rheumatism Association Liaison to the FDA Arthritis Advisory Committee, UCLA School of Medicine, Department of Medicine, Division of Rheumatology, LOS Angeles, CA 90024. Arthritis and Rheumatism, Vol. 28, No. 10 (October 1985) similar reports per million prescriptions. The manufacturer presented Medicaid data from 3 states which indicated that the incidence of allergic reactions to tolmetin was lower than that for 5 comparison NSAIDs, suggesting that the “blip” was a reporting aberration that may have been stimulated by the withdrawal of a similar drug, zomepirac, because of allergic reactions. Diflunisal and piroxicam ranked at or near the top of most of the summaries for specific ADRs (Table 2). However, the manufacturers pointed out that these drugs were first introduced in the United States in April 1982, and the summaries were prepared from data collected during their second year of marketing. Information on ADR duriug the life-cycle of other Table 1. Summary of reported US ADR cases per million prescriptions (August 1983 to August 1984)* NSAID Phenylbutazone Diflunisal Piroxicam Sulindac Fenoprofen Indomethacin Tolmetin Meclqfenamate Naproxen Ibuprofen Aspirin Acetaminophen Fatal ADR Nonfatal Millions Fatal cases cases of Rxs ADR per lo6 Nonfatal per lo6 (or units) cases Rxs ADR cases Rxs 2.81 2.50 7.53 5.47 3.26 9.02 2.57 1.80 11.86 20.45 241.0t 211.0t 14 10 20 11 6 15 4 2 8 9 NA NA 4.99 4.00 2.66 2.01 1.84 1.66 1.55 1.11 0.67 0.44 NA NA 71 238 686 255 81 229 152 86 246 209 NA NA 25.29 95.31 91. I6 46.61 24.85 25.39 59.05 47.72 20.75 10.22 NA NA * ADR = adverse drug reaction; NSAID = nonsteroidal antiinflammatory drug; Rxs = prescriptions; NA = not available. t Number of 100-tablet units sold x 10’. 1169 GOVERNMENT AFFAIRS Table 2. Reported cases of nonfatal uppel-gastrointestinal perforations, bleeding, and ulcers per million daily doses* NSAID Difluni sal Piroxicam Phenylbutazone Indomethacln Sulindac Meclofenamate Naproxen Tolmetin Ibuprofen Fenoprofen Perforation Bleeding Ulcer 6 29 57 22 51 44 8 33 3 19 6 6 84 7 7 4 2 9 14 1 8 1 0 3 0 4 1 4 10 2 Cases per miI1ion daily doses 0.96 0.90 0.53 0.23 0. I7 0.15 0.14 0.08 0.06 0.05 * NSA.ID = nonsteroidal antiinflammatory drug; 1983 data for piroxicam, 1982 and 1983 data for all others. drugs indicates that the reporting rate during the second year is about 5 times greater than that after the fifth year of marketing. When this correction was applied, these drugs ranked near the bottom of the summaries comparing them with other, more mature NSAlDs. Attention was called to other apparent abberations in the data, but in each case the manufacturers were able to provide rational explanations, such as: more careful examination of the data might demonstrate that the same case was counted more than once, or that cases were incorrectly classified and should be excluded; some companies were more diligent or effective in their use of detail persons to identify ADRs; published case reports stimulated more ADR reports: etc. The explanations illustrate some of the many problems associated with the interpretation of data collected by the voluntary ADR reporting system. It must be noted, however, that this system provides the largest, and frequently the only, summary data of ADRs for marketed drugs. The FDA believes that major (threefold or greater) deviations from the norm, after appropriate corrections, provide useful signals of possibly important adverse reactions that should prompt more vigorous investigations using other approaches. In the past, 2 NSAIDs, meclofenamate and indomethacin, have been labeled “not for initial use” because of perceptions of increased risk of adverse effects associated with their use. Since their use now appears to entail no greater risk than that of the other NSAIDs, the Committee advised the FDA to remove the “not for initial use” restriction from the labels.