ANDA REFORMS SHOULD ADDRESS "ARBITRARY" REQUIREMENTS FOR MINIMUM BATCH SIZES -- AAPS, LEDERLE, GPIA RECOMMEND; ELKINS-SINN SUGGESTS "FAST-TRACK" ANDAS

FDA should "develop and publish guidelines for acceptable minimum batch size based on good science and technical considerations," the American Association of Pharmaceutical Scientists Task Force on Generic Drugs recommended in April 6 comments to FDA. "There is a perception that minimum batch size requirements vary among reviewers and appear arbitrary," AAPS noted. AAPS was among several groups that addressed the issue of batch size requirements in ANDA submissions in comments responding to a notice published by FDA in the Feb. 21 Federal Register that asked for input on how to reform the ANDA approval process. The Generic Pharmaceutical Industry Association, in April 6 comments, characterized some of FDA's changes in batch size regulations as "inconsistent, unfair, arbitrary, disruptive, and unnecessary." GPIA recommended that "policy changes should be prospective" and that the agency should "solicit and receive industry input before a change is implemented." Rugby-Darby subsidiary Chelsea ran into a number of problems with batch sizes and record-keeping when the company was recently inspected by FDA. In an Aug. 1 letter to FDA, Rugby-Darby complained that when FDA instituted a new batch requirement in January 1988, industry "operated under the premise that the minimum 100,000 dosage unit requirement related to designed batch size." Rugby affiliate Chelsea Labs subsequently received a letter in July 1989 from FDA's Division of Bioequivalence stating that the batch for its cyclobenzaprine tablets was less than 100,000 dosage units and that a waiver to the minimum requirement was required. When the company inquired about the discrepancy, FDA told them that the "intent' of the January, 1988 policy was that the minimum 100,000 dosage unit batch size was to mean actual batch yield and this was now their policy." Rugby said that Chelsea has about 18 products that were affected by the policy change and that the company has been asked to submit manufacturing information on batches of these products so that FDA could grant them waivers. In February, FDA proposed to withdraw five Chelsea ANDAs for untrue statements, discrepancies and omissions concerning the manufacturing of batches in support of the ANDAs and said it is reviewing other Chelsea ANDAs for similar problems ("The Pink Sheet" Feb. 5, p. 13). Lederle also found fault with FDA's requirement of a minimum batch size of 100,000 for oral dosage forms, in comments sent to the agency on April 5. "Minimum (BA/BE) batch size should be not less than 10% of planned production batch size (not an arbitrary number), as long as equipment size (not type) is changed only," Lederle maintained. In addition, American Home Products subsidiary Elkins-Sinn suggested that the batch size requirement for solid and liquid dosage should differ. * Elkins-Sinn also recommended that FDA establish a "fast track" review system for "certain" ANDA submissions. The company's suggestion echoed a similar recommendation from Rep. Dingell (D-Mich.) in comments to FDA. "We would like to have the generic group consider the possibility of giving "fast track" action to certain applications," with criteria for selection based on: "first generic copy"; "high priority based on medical values"; or "significant dollar volume of innovator's drug," Elkins-Sinn advised. GPIA also addressed the issue of when the Division of Generic Drugs should request validation samples and CGMP certification. "There is now no uniform time point for these requests," GPIA remarked, "a deficiency that could result in unfair treatment of applications in the review queue." GPIA recommended that "this uniform time point should be the clearance of a bioequivalence study by the Division of Bioequivalence." FDA's first-in first-out system for prioritizing generic reviews was criticized as discriminatory by Lederle in its comments. "By the present system, companies which delay filing their submission until they have completed a top quality ANDA, are 'punished' because they are put in the ANDA review queue behind those with lesser quality ANDAs," Lederle stated.