BRISTOL-MYERS' CARBOPLATIN NDA SUBMISSION FOR OVARIAN CANCER

BRISTOL-MYERS' CARBOPLATIN NDA SUBMISSION FOR OVARIAN CANCER will be submitted based on preliminary results from the firms' two pivotol studies comparing cytotoxin and carboplatin to cytotoxin and cisplatin, Bristol-Myers Senior VP-Anticancer Research Stephen Carter, MD, indicated at the National Cancer Institute's Division of Cancer Treatment Board of Scientific Counselors meeting Oct. 2. Bristol-Myers is hoping that an NDA based on preliminary data using response rate as an efficacy endpoint may be acceptable. In a presentation to analysts last spring, the company said that it would be filing an NDA for the cisplatin (Platinol) analog "soon" ("The Pink Sheet" April 13, p. 12). Carter noted that if the company was to "wait for those two pivotol studies to reach their fruition of 300 patients with adequate follow-up, we would be into the early 1990s before [we could make] this submission." FDA Office of Drug Research and Review Director Robert Temple, MD, commented that the caroboplatin pivotol studies might not have to be completed for an NDA submission if preliminary analysis showed complete response in a significant portion of the patient population. The carboplatin data "has not been reviewed [by FDA]," Temple said, "but you may find that everyone will agree that evidence of complete histologically confirmed response in a reasonable fraction of patients would be taken as relevant evidence." NCI officials at the meeting cited carboplatin as an example of FDA efficacy requirements, such as survival data, holding back approval of new chemotherapeutic agents. NCI Division of Cancer Treatment Director Bruce Chabner, MD, maintained that carboplatin should be approved based on currently available data showing comparable efficacy with cisplatin and a lower toxicity profile. "It seems to me with the lesser toxicity that [carboplatin] has, you should approve it based on the [comparable] response rates [with cisplatin]," Chabner said. He noted that one study of "a comparison of CBDCA [carboplatin] to cisplatin in ovarian cancer [found that] response rates were the same; it was 53% [for carboplatin] v. 64% [for cisplatin], or something like that." Carter reported that the company has "evidence from all the studies that have been done by the NCI and sponsored by [Bristol-Myers] that carboplatin is clearly active in first line treatment of stage 3/4 [ovarian cancer] . . . and significantly active in second line treatment." The Bristol-Myers VP added that in "a variety of studies of carboplatin in ovarian cancer [patients] who had received prior cisplatin containing regimens [found] response rates . . . from 15%-to-35%, and in every study . . . some complete responses." The "toxicity differences in these studies -- vomiting, renal toxicity, ototoxicity, neuropathy, and very critically, therapy being stopped as a result of side effects -- was highly statistically significant in favor of carboplatin," Carter said. Using the difficulty involved with obtaining survival data on carboplatin as an example, Carter suggested that FDA consider revising its guidelines for the evaluation of oncologic drugs to allow response rate as a primary efficacy endpoint for new agents. Carter maintained that FDA should be more flexible with its efficacy requirements for analogs in particular, because they tend to have improved toxicity profiles compared with the parent compound. At the same meeting, Temple told NCI that FDA will be reviewing its efficacy requirements for anticancer agents by tumor type, beginning with ovarian cancer ("The Pink Sheet" Oct. 5, T&G-9).