More than half of cancer drugs granted accelerated approval between 2013 and 2017 did not demonstrate a significant improvement in overall survival (OS) or quality of life (QoL) within five years of approval, shows an analysis of U.S. Food and Drug Administration (FDA) data.
The study by postdoctoral research fellow Ian Liu and colleagues at Brigham and Women’s Hospital and Harvard Medical School also found that approvals for conversion from accelerated to regular approval were typically granted based on surrogate outcome measures, such as response rates or progression-free survival (PFS), rather than those showing clear clinical benefit like OS.
This can leave patients and clinicians with “uncertainty as to the drug’s actual benefit, especially after such a long period of time,” Liu told Inside Precision Medicine. He said that how well a drug works is an important question for patients and “if at the end of this regulatory pathway, we still don’t know then is not good for patients or for the clinician to take care of them, and we should really be doing our best to resolve those uncertainties.”
Liu explained that more than 80% accelerated approvals granted by the FDA are for cancer drugs, and approximately one-third of all oncology drugs take this pathway, which was originally implemented in the 1980s in response to the HIV epidemic. It provides a way for promising drugs treating unmet medical needs to reach the market sooner based on changes to unvalidated surrogate measures considered “reasonably likely” to predict clinical benefit.
Once accelerated approval is granted, manufacturers are then required to confirm clinical benefit and drug indications in postapproval confirmatory trials. The treatments may then be converted to regular approval or withdrawn depending on the results of these trials.
For their study, Liu and co-workers identified 129 cancer drug–indication pairs that were granted accelerated approval between 2013 and 2023. Of these, 46 indications were approved between 2013 and 2017 and thus had at least five years of follow-up data.
During the follow-up period, 29 (63%) indications were converted to regular approval, of which 20 (69%) demonstrated clinical benefit: Seven (24%) showed improvements in both OS and QoL; Seven (24%) improved OS but did not demonstrate QoL benefit; six (21%) improved QoL without improving OS.
The remaining nine (31%) were converted without showing improvements in either OS or QoL in their confirmatory trials.
A further 10 (22%) of the 46 indications were withdrawn, and seven (15%) remained ongoing with no conversion decision after a median of 6.3 years.
This meant that, in total, 26 (57%) of 46 drug–indication pairs granted accelerated approval failed to demonstrate clinical benefit, i.e. improved OS or QoL, after at least five years of follow-up, the researchers report in JAMA.
The team also looked at the characteristics used in pivotal trials to determine conversion from accelerated to regular approval for 48 drug–indication pairs with suitable data.
They found that 19 (40%) were converted based on a pivotal trial demonstrating an OS benefit, 21 (44%) were converted based on an improvement in PFS, event-free survival, or disease-free survival, five (10%) on response rate plus response duration, two (4%) on response rate alone, while one (2%) conversion was granted despite a negative confirmatory trial that showed no improvement in either overall survival or PFS.
Furthermore, the proportion of conversions based on response rate has increased over time, from 0% (0 of 28) between 2013 and 2020 to 40% (7 of 19) from 2021 to 2023.
Liu et al. say that although response rate is a logical surrogate measure to use for accelerated approval because many tumor types are unlikely to spontaneously regress, “it cannot by itself determine whether a drug offers patients a clinical benefit, and—as highlighted by the FDA—should therefore not be used without additional patient-centered information to establish regular approval.”
Liu believes that many of the precision medicines currently in development for cancer will benefit from accelerated approval based on surrogate endpoints, but stresses that “confirmatory trials should be based on endpoints that matter to patients.”
He hopes that the study findings will help clinicians to have nuanced conversations with patients about the risks and the benefits of each treatment. “This hopefully already happens,” he said, but noted “it is especially important for these accelerated approval drugs, where we hope that a lot of the uncertainty as to how well a drug improves outcomes that patients care about–feeling better, living longer–is actually conveyed to them, clearly and honestly.”
