PENDING ANDAs WITH BIO WORK BY CONTRACT LAB BIODECISION BEING WITHHELD, HHS SECRETARY SULLIVAN TESTIFIES BEFORE REP. DINGELL; "POOR SCIENCE" FOUND AT LAB

Pending ANDAs that rely on bioequivalence work conducted for sponsors by the contract lab Biodecision Laboratories will be withheld from approval, HHS Secretary Sullivan testified at a July 18 hearing before the House Energy & Commerce/Oversight Subcommittee. It is not clear whether the hold will also include NDAs with bio work performed by Biodecision, based in Pittsburgh, Pa. Half the contract lab's business reportedly involves brandname companies. The decision was made after an FDA inspection of the contract lab, conducted from Oct. 4, 1989 through Jan. 11, found 11 studies (of 13 reviewed) with questionable data. FDA's March 7 Establishment Inspection Report (EIR) states that the data was "called into question based on such practices as manipulation of control data, selective reporting of sample values, improper calculation of results, etc." In a July 17 press release, Biodecision said it "has achieved regulatory compliance" since the inspection and that the tests it conducts are valid and the data and the conclusions are reliable." The company noted that it formed a task force Jan. 12 to review the inspection report and "prepare an action plan to bring about systemic correction" of deficiencies. On March 8, the firm met with FDA to present a schedule for correcting procedures and re-evaluating data in each of the 11 studies found faulty in the FDA inspection. The re-evaluations have since been completed and "confirm the original conclusion that the generic products are bioequivalent to the pioneers," the firm said. Biodecision "has revised and expanded its standard operating procedures" and made other improvements, including the hiring of additional staff and institution of "internal audits of each operation of the analytical laboratory." Subcommittee Chairman Dingell (D-Mich.) suggested at the hearing that revelations involving Biodecision constituted further proof that the generic drug crisis "continues to grow." Based on an estimate of the number of ANDAs filed at FDA involving Biodecision, Dingell estimated that "perhaps another 20% or so of the tablet and capsule generic drug applications may have been approved based on scientifically and legally deficient bioequivalency studies." In a July 16 letter to Sullivan, Dingell had requested that HHS "instruct FDA to withhold approvals of all" ANDAs and "abbreviated antibiotic drug applications dependent upon biostudies from Biodecision until all underlying data and methods have been reviewed by a qualified analytical chemist and a qualified biostatistician from that agency." Dingell's letter also urged that marketed generic drugs whose supporting data included work done by Biodecision be revalidated. The congressman suggested that revalidation be made a higher priority for drugs that have narrow therapeutic ranges. In addition, the congressman asked Sullivan to "instruct FDA to supply the subcommittee and the American public with a plan to determine the validity of [Biodecisions' bioequivalence] studies by July 25, 1990." A plan for revalidating the bioequivalence of drugs whose therapeutic ranges are broader should be drawn up by Aug. 8, the letter continues. Dingell asked that both revalidation efforts "be constructed to identify patterns of meaningful scientific errors in specific products and specific companies." The congressman noted that "the most important errors reported in the EIR involved analytical and statistical methods." Since "late 1989," he continued, FDA has known that Biodecision "has employed such poor science in performing [bioequivalence] studies that, absent further information, their results cannot be relied upon to demonstrate the safety and efficacy of generic equivalence." The letter acknowledged that the faulty bioequivalence work did not necessarily mean all products approved based on work by Biodecision are unsafe or ineffective. "Some of the errors may have had no effect on the measurement of relative bioavailability, and some may have actually biased the studies in such a way that bioequivalence was harder to demonstrate. However, clearly some of the errors were likely to produce a bioequivalent determination when scientifically proper methods would have not produced the result required by law." Therefore, in the context of the "corruption found to date," in the generic industry, Dingell said "this latest revelation of impropriety is truly calamitous." FDA's EIR credits Biodecision's management with "a willingness to expeditiously correct the problem found" in FDA's inspection. "In at least one case," involving a study of amoxicillin for the company Clonmel, the report states, Biodecision "plans to re-evaluate the reported study data after correcting for problems found in a review of the study." The report accuses Biodecision of "poor science"; however, it adds that the inspection was "unable to find evidence of fraud or conspiracy in the sense of manufacturing false data or manipulating data to steer the outcome of a study." FDA noted that Biodecision was founded in 1972 "as a bioequivalence testing facility for Mylan Pharmaceuticals." Recently, the report adds, Biodecision conducted approximately 150 bio studies per year for a client list of about 150 companies. FDA's inspection "included in-depth review of five chromatographic studies and eight micro-assay studies."