FDA BACKLOG OF "OVERDUE" NDAs HAS BEEN CUT, BUT FDA IS STILL FACING LARGE IN-BOX; 30 EFFICACY SUPPLEMENTS WERE APPROVED IN 1989, 41 WERE FILED

Executive Summary

FDA's most recent figures on the NDA review process show the lowest number of "overdue" applications pending at the agency since 1982. According to data from the end of 1989, FDA had cut the number of pending overdue NDAs to 67. In 1988, by comparison, the agency had a backlog of 71 overdue NDAs. Overdue NDAs are applications pending at the agency beyond FDA's statutory 180-day review period. Overall, however, the agency was facing a larger workload at the end of 1989 than a year previously. FDA had 239 pending full NDAs and 2,158 pending supplements at the end of last year compared to 219 full NDAs and 2,088 supplements on Dec. 31, 1988. The data on the NDA review process is contained in a recently issued annual publication from FDA, "Offices of Drug Evaluation: Statistical Report." The report is dated June 1990 but only recently has become available. The increase in the total number of NDAs pending developed despite a decline in original NDA filings in 1989. The number of original application filings (118) in 1989 was the lowest in at least 11 years. The next lowest year was 1986, with 120 filings. In 1988, 126 original NDA applications were filed. The top year in the recent decade was 1983 when 269 original NDAs were submitted to FDA. FDA continues to face a large group of antihypertensive NDAs. At the end of 1989, the agency had 13 NDAs since from 1986 in that class of products. More than half of those were filed in 1988 when FDA received seven for antihypertensives NDAs. Other drug classes with numerous NDAs pending are the antimicrobial (six filed from 1986 to 1989), anti-inflammatory (five) and hormone (five NDAs) categories. [EDITORS' NOTE: See chart on p. 5 of this issue for a listing of the pending NDAs that were filed with the agency between 1986-1989. The chart provides a quick overview of the types of products likely to emerge from FDA in the next several years]. The oldest new molecular entity applications still pending at FDA, according to the report, are two anti-inflammatories and an antihypertensive filed in August and September of 1982. In an apparent response to criticisms of the agency's handling of supplemental filings, FDA has added a new table to the report that describes actions taken on efficacy supplements. During 1989, a total of 30 efficacy supplements were approved and 41 supplements were received by FDA. The agency's record on efficacy supplements was one of the topics raised in a meeting between the board of the Pharmaceutical Manufacturers Association and FDA's Center for Drug Evaluation and Research Director Carl Peck on May 31. The report points out that "significant efficacy supplement approvals were atenolol [ICI Pharma's Tenoretic] for acute myocardial infarction and tamoxifen citrate [ICI Pharma's Nolvadex] for metastatic breast cancer in premenopausal women." Sixteen efficacy supplements were withdrawn in 1989; the same number were issued nonapprovable letters; and four were rejected with refusals to file. In an effort to collect data on the effect of end-of-Phase-II conferences on NDA processing, FDA has started to add the date of those conferences onto its tabulation of NDA review dates. (See p. 4 for FDA's 1989 new molecular entity NDA report card). The report compares the approval times for the NMEs approved from 1984 to 1989 where the sponsor and FDA had end-of-Phase II conferences to the approval times for products which were not the subject of conferences. The total mean approval time for the 41 NMEs that had conferences was 31.2 months, while the mean approval time for the 95 NMEs that did not have the meetings was 34.4 months. FDA took a little longer itself looking at applications for products that had conferences, but the agency apparently sent fewer of those applications back to sponsors for followup work. The data on 1989 approvals shows that several NME applications had protracted reviews because of applications returned to the sponsor. For example, Wyeth-Ayerst added almost 39 months to the total approval time for Dalgan (dezocine), according to FDA's calculations. The injectable analgesic, approved in December 1989, had an NDA pending at FDA for a total of 77.1 months. FDA attributed less than half of that time to its review (38.2 months). The company had the application for 38.9 months during the period from first submission to FDA until approval. The report notes that "total approval time is calculated from the NDA official receipt date to the approval date" and "FDA approval time is the total approval time minus the time required by the firm to provide information in response to an agency action (approvable, not approvable letter)." FDA approval time also omits the company time from withdrawal to resubmission. CDER Director Peck wrote to PMA President Mossinghoff in late July to point out the difference in review times if the amount of time that FDA waits for changes by a sponsor are excluded. Peck noted that the agency "found FDA's actual 'review' time [for the 23 new molecular entities approved in 1989] to be 25.4 months instead of 32.5 months [as calculated by PMA]." A number of applications for NMEs were relatively clean when initially filed with FDA, according to the 1989 figures. One NDA that had no approval time added on by the sponsor was Ciba-Geigy's drug for obsessive-compulsive disorder, Anfranil (clomipramine). The drug was approved in December 1989 with a total approval time/FDA approval time of 6.4 months. Another 1989-approved NDA that did not receive increased approval time due to additional company work was Bristol-Myers Squibb's Paraplatin (carboplatin), which is indicated for palliative treatment of recurrent ovarian cancer. The NME was cleared with a total approval time/FDA approval time of 8.1 months. Cancer drug NDAs often have the advantage of extensive clinical work under NCI sponsorship prior to actual NDA filings. Charts omitted.