Association of adverse events and tumor response in melanoma patients treated with tremelimumab.

Abstract

e13104 Background: Several analyses of data from trials of anti-CTLA4 mAbs have suggested a relationship between antitumor activity and certain AEs. Such associations are confounded by the fact that responding patients receive more doses of drug and have longer follow- up. An analysis was done using data from 2 pivotal tremelimumab melanoma studies to determine whether certain AEs in the first treatment cycle (90 days) predict response to tremelimumab. This time period was chosen in order to decrease the difference in follow up time between responders and nonresponders. Methods: Tremelimumab was administered at 15 mg/kg Q90days in studies A3671008 and A3671009. Adverse events were reported for 90 days post last dose or until start of new therapy. RECIST responders were determined by independent radiological review for A3671008 and by sponsor using investigator measurements for A3671009. The following AE categories were analyzed: diarrhea/colitis, all grades and grade ≥3; endocrine events, rash, and pruritus. Results: 37 of 325 treated pts in A3671009 and 16 of 246 treated pts in A3671008 achieved response. All responders and 67% of nonresponders remained on study at least 80 days. Most tumor responders received ≥4 doses and were on study for ≥1 year. Most nonresponders received only a single dose. Rash but not pruritus was more frequent in the first cycle in responders compared to nonresponders. Rates of diarrhea and of severe diarrhea in the first cycle were similar in responders and nonresponders. None of the responders reported endocrine events in the first cycle. Conclusions: Endocrine and GI AEs occurring in the first cycle did not predict response. The higher incidence of endocrine and GI events AEs reported on study in pts with tumor response is due to the greater number of doses received and longer time on study. Responders were more likely than nonresponders to report rash, but not pruritus, in the first cycle; however, about half of the responders did not experience rash in the first cycle, so appearance of rash should not be used to decide whether to continue therapy. Author Disclosure Employment or Leadership Position Consultant or Advisory Role Stock Ownership Honoraria Research Funding Expert Testimony Other Remuneration Pfizer Pfizer