Real-world effectiveness of approved anticancer agents among Medicare beneficiaries.

Abstract

6589 Background: The overall survival (OS) among elderly patients (pts) treated with novel anticancer agents in the real world may be inferior to the OS reported in pivotal trials for drug approval. Pts ≥ 65 yrs old constitute 60% of cancer pts, yet only 40% of cancer clinical trial participants. Elderly pts have greater comorbidities and experience a higher risk of toxicity from cancer drugs. Using the Surveillance Epidemiology and End Results-Medicare database (SEER-Medicare), we compared the OS among elderly pts treated with FDA-approved cancer drugs for advanced solid tumors to the OS reported in the clinical trial. Methods: We identified cancer drugs FDA-approved for metastatic solid tumors between 1/1/08-12/31/12. In a retrospective analysis, for each indication we identified pts in SEER-Medicare meeting disease eligibility criteria (stage, histology, prior therapies) in the trial associated with approval. Pts were diagnosed with cancer from 2010-2013 with follow-up through 2014. Treatment (tx) was determined from national drug codes in Medicare Part B for intravenous (IV) drugs and Part D for oral drugs. Indications were included if ≥ 30 pts receiving tx met eligibility. Kaplan-Meier methods were used to calculate median OS and cancer-specific survival (CSS) in Medicare pts. Median duration of therapy (DOT) was estimated from date of the first through completion of the last prescription claim for oral drugs and number of cycles for IV drugs. Results: OS data were available and sample size parameters met for 14 drug indications. The median OS among SEER-Medicare pts was shorter than the reported trial OS of tx arms for 13 of 14 drug indications (median difference -7.6 mos, range +3.4 to -28.7 mos). CSS was similar to OS among Medicare pts. Median DOT among SEER-Medicare pts was shorter than the reported trial DOT of tx arms for 13 of 14 indications (median difference -2.9 mos, range 0 to -8.9 mos). Conclusions: The OS and DOT among SEER-Medicare pts treated with FDA-approved cancer drugs was inferior to the reported OS in pivotal clinical trials for nearly all indications analyzed. The shorter DOT among Medicare pts may explain survival differences. Trials leading to regulatory approval may not be generalizable to cancer pts ≥ 65 yrs.