Evaluating value in the age of accelerated approval: A pembrolizumab case study.

Abstract

11084 Background: The FDA accelerated approval (AA) program was established to expedite approval of promising drugs that meet unfulfilled clinical needs. AA is granted on the basis of surrogate endpoints, decreases time to market, and increases access to therapies. Pursuing multiple AA and label expansions can also be a significant component of a drug’s lifecycle strategy. Among 322 anticancer drug approvals (2009-2020), 59% were label expansions, with only 16% granted to new drugs with a novel mechanism of action. Our goal was to understand changes in value across multiple label-expanding AAs for high-cost anticancer medications. We studied pembrolizumab, a highly successful immune checkpoint inhibitor, evaluating value from its initial approval through multiple label expansions. Methods: We calculated value scores for different indications of pembrolizumab using the ASCO Value Framework Net Health Benefit Worksheet (2020) and ESMO-MCBS v2.0. The value calculations require mature survival outcomes, so we could not calculate scores for AA indications awaiting data maturity (2) or those eventually withdrawn from the market (2). We utilized the FDA drug database and primary literature to gather safety and efficacy outcomes for each indication. Analysis generated descriptive statistics about how the value of pembrolizumab evolved from 2014 to 2025. Results: We calculated value scores for 28 pembrolizumab indications, 12 with AA. Value scores varied widely across indications, with ASCO Net Health Benefit Score (NHBS) ranging from 17 (biliary tract cancer) to 59 (MSI/dMMR colorectal cancer) and ESMO MCBS scores ranging from 1 to 5. NHBS was not perfectly collinear with OS gains. The rolling average of NHBS across all indications over time remained fairly stable with a mean score of 37.67 (range: 30- 41). We also compared the NHBS for each AA indication to the cumulative NHBS average at the time each AA was granted. The NHBS for HCC, HNSCC, and cervical cancer were lower than respective pre-AA NHBS averages. The other 9 AAs yielded a final NHBS higher than each’s pre-AA average NHBS. Conclusions: The NHBS formula generated varying values for different pembrolizumab indications, distinct from simple measures such as OS-rank order. Although the average NHBS remained fairly constant over time, it was lower than the maximum NHBS (59) and less than the NHBS of other breakthrough approvals (osimertinib for NSCLC: 110, nivolumab for melanoma: 53). If the average value of multiple indications in a drug's indication portfolio is significantly lower than a few blockbuster successes, clinicians could be vulnerable to an availability heuristic and overestimate the value of new AAs, based on anecdotal understanding of prior successes. In an environment of fast-paced AAs and label expansions, clinicians, patients, and payers should develop and deploy rapid value assessment tools in order to sustainably maintain access to high-yield cancer therapies.