NSAID ADVERSE REACTION REPORT COMPARISONS ARE NOT SUITABLE AS BASIS FOR THERAPEUTIC RANKING OF INDIVIDUAL DRUGS, FDA ARTHRITIS ADVISORY CMTE. AGREES

Non-steroidal anti-inflammatory drug (NSAID) adverse reaction report comparisons should not be used to designate individual drugs within the class as first or second line usage, FDA's Arthritis Drugs Advisory Cmte. agreed at its Oct. 26 meeting. Cmte. discussion focused on a compilation of adverse reaction reports received by FDA and tabulated by FDA Anti-inflammatory Drugs Group Leader John Harter, MD. Summarizing the cmte.'s view of the data, cmte. consultant Michael Weisman, MD, declared: "I think that we should look at labeling and dosage as issues in interpreting this data rather than automatically approving the idea of stratifying drugs." Weisman added that "the more drug you use, the more effective it becomes and the more toxic." Harter's analysis reviewed 11 marketed NSAIDs by the proportion of adverse reactions er mil. Rxs. (see chart on p.4). Harter ranked the drugs by the rate of fatal and non-fatal reactions as well as reactions by body systems. Additional analysis performed by the FDAer included foreign data as well. Based on the Harter data, FDA proposed for discussion the possibility that several NSAIDs might be designated as "not for initial use." Among those drugs were Merck's Indocin (indomethacin), and Ciba-Geigy's Butazolidin (phenylbutazone) and Tandearil (oxyphenbutazone). FDA and Ciba-Geity have already revised phenylbutazone and oxyphenbutazone labeling to state that the drugs are not for initial therapy (see related story, p. 8). Because the cmte. included six new members, much of the discussion was carried by cmte. consultants (some of whom are former cmte. members), industry representatives, and FDA staffers. Commenting on the Harter data, FDA Office of Epidemiology and Biometrics Div. Director Gerald Faich, MD/MPH, said: "Fully a third of the adverse reaction reports we receive are for nonsteroidal anti-inflammatories." Faich noted, however, that data from adverse reaction reports "is not necessarily a direct measure of the rate of occurrence," but "the rate of reporting." The "number of spontaneous reports that come into a system peaks at year two or three after marketing," Faich said. "Given that phenomenon, it's probably more appropriate to look at a drug . . . in the year three of its marketing." Concurring that "you have to compare them at the same point in their life cycle," Pfizer representative Roger Sachs told the cmte. "this needs to be considered strongly in making any comparison." Additional variables and biases such as differences in mfrs. adverse reaction reporting, drug usuage, Rx size and compliance were also cited by Sachs and other industry representatives. In addition to citing biases that may occur in adverse reaction reporting systems, McNeil's Sam Fried told the cmte. that comparisons based on such data could compromise FDA's voluntary system of adverse reaction reporting. Stating that "the data becomes biased when reporters flood the system with reports in the hopes of avoiding any possibility of missing even the most trivial event," Fried asserted: "They become biased also . . . when the threat of using the accumulated data for competitive purposes becomes fully appreciated." In either case, "the validity of data accumulated is diminished, the significance of the reported number is questionable, and the value of the . . . voluntary reporting system . . . may well be compromised," the McNeil representative declared. Differences in the mechanisms of action and side effects of different NSAIDs would also make a distinction between "not for initial therapy" and initial therapy inappropriate, Evelyn Hess, MD, Director of Immunology at the University of Cincinnati, told the cmte. A former chairman of the advisory cmte., Hess said she "would not rest happy if in fact there was to be such labeling based upon the reactions presented. "I think we have to remind ourselves that the NSAIDs are six quite different chemical classes" with "their own perhaps different set of side effects," she said. Chart omitted.