Most cancer drugs approved over the past two decades have offered only marginal impacts on patient survival, despite “hyped-up” trials suggesting they have substantially reduced the risks of death and tumour progression, a review has found.
Based on trial data supporting the FDA approval of 124 novel pharmaceuticals between 2003 and 2021, the study authors say the finding highlights the need for caution in evaluating initial approvals for drugs with non-robust clinical evidence which may overestimate efficacy outcomes.
Some 234 trials with available data were included in the meta-analysis – mostly open-label, phase III concurrent randomised studies, each with a median of 331 enrolled patients.
The trials assessed immune therapies such as checkpoint inhibitors such as PD-1/PD-L1 and CTLA-4 inhibitors, and targeted therapies such as TKIs.
Across these, the novel pharmaceutical was shown to reduce the risk of death by a mean of 27% (Overall Survival Hazard Ratio 0.73), and the risk of tumour progression by 43% compared to control.
However, this equated to a median increase of only 2.8 months in overall survival, and 3.3 months for progression-free survival, the researchers found.
They also noted a progressive loss of clinical benefit across indication extensions, with 78 of the 124 drugs approved across multiple indications for a combined 374 indications in total.
“This diluted benefit could be subject to the new indications’ line of therapy, clinical trial evidence, selected patient population, and treated disease,” they wrote in the Journal of Clinical Oncology.
“Cancer drugs are first approved as monotherapies in the second- or third-line setting, and then extended to first-line combination regimens.”
“First approving drugs for advanced-line treatments results in biased efficacy estimates as these studies select heavily pre-treated patients with more indolent diseases.”
The authors, from Heidelberg University in Germany, stressed the problem was not unique to the US, with studies showing similar results for cancer drugs approved in France, England and Australia.
Nevertheless, the FDA’s approval process was itself partly to blame, particularly its accelerated pathway for new drugs with indications targeting serious conditions with an unmet need.
This was because it allowed initial approval that was often only supported by single-arm trials, the authors said.
“Even initial indications that are supported by concurrent RCTs lack adequate comparators – merely 24% of RCTs for initial indications assess the new drug to an active comparator,” they wrote.
“By contrast, supplemental indications are more frequently supported by phase III RCTs with active comparators. The lower OS and tumour response benefits observed in confirmatory trials highlight that poorly designed trials may overestimate efficacy outcomes.”
“Yet, investors hype initial indications by posting aesthetic Kaplan-Meier curves and astonishing HRs, which are interpreted by patients— often without reviewing a trial’s robustness—who pressure the FDA to swiftly approve the new treatment.”
Some 45% of initial indications were approved via the agency’s accelerated pathway, although this fell to 22% for each drug’s second indication, and 19% for the third and fourth.
“This study highlights that initial drug approvals rely on non-robust evidence, which may overestimate treatment outcomes,” the researchers added.
“This differential evidence and efficacy a drug offers to patients and insurers across indications should be reflected in pricing, coverage, and reimbursement policies.”