ABSTRACT
BACKGROUND: We aimed to describe the reasons for failure of experimental anticancer drugs in late-stage clinical development. MATERIAL AND METHODS: We searched the PharmaProjects database (https://citeline.com/products/pharmaprojects/) for anticancer drugs discontinued between 01/01/2009 and 06/30/2014. Drug programs that reached phase III trials, but never gained Food and Drug Administration (FDA) approval were compared to 37 anti-cancer drugs achieving FDA approval in this time period. RESULTS: Forty-two drugs fit our criteria for development failures. These failed drugs (49% targeted, 23% cytotoxics, and 28% other) were tested in 43 cancer indications (drug programs). Only 16% (7/43) of failed drug programs adopted a biomarker-driven rationale for patient selection versus 57% (21/37) of successful drug programs (P<0.001). Phase II trial information was available in 32 of 43 failed drug programs and in 32 of 37 successful programs. Nine of the 32 trials (28%) of failed drugs versus 28 of 32 trials (87%) of successful drugs (P<0.001) achieved proof of concept (single agent response rate (RR) ⩾20% or combination therapy showing a ⩾20% RR increase above the median historical RR without the experimental agent (with a minimal absolute increase of 5%) or a randomized phase II trial showing significance (P⩽0.05) for its primary outcome). No pattern of study sites, trial design or funding characteristics emerged from the failed drug analysis. CONCLUSION: For drugs that reached Phase III, lack of a biomarker-driven strategy and failure to attain proof of concept in phase II are potential risk factors for later discontinuation, especially for targeted agents.
