Across the United States, clinicians and payor organizations rely heavily on the National Comprehensive Cancer Network® (NCCN®) guideline recommendations, which combine evidence with expert opinion, to support their decisions. These guideline recommendations include indications for drugs that received accelerated or regular approval by the US Food and Drug Administration (FDA). Accelerated approval enables certain drugs for serious conditions to be approved on the basis of clinical trials that show improvements in surrogate measures, with a commitment from the manufacturer that further, comparative trials will be conducted to achieve regular approval. With a rise in recent years of cancer drugs receiving accelerated approval by the FDA, Dr Edward Scheffer Cliff and colleagues conducted a cross-sectional study to determine how the NCCN guidelines’ assessments for cancer drug indications that received FDA accelerated approval compared with those for drugs that received FDA regular approval.
The authors identified 100 cancer drugs that received 315 indications from 1992 through June 2022. Of these, 23% of indications were for 1 of 6 checkpoint inhibitor drugs. The analysis found that cancer drug indications with accelerated approval were less likely to be assigned high-level evidence ratings (ie, category 1) and/or preferred status in the NCCN guidelines than indications with regular approval. However, they also found that as of October 2022, only about half (49%) of accelerated approvals were converted to regular approval, 13% were withdrawn, and 38% remained as accelerated approvals.
In some instances, regular approvals were supported by lower-level evidence; less than half of regular approvals and only 38% of accelerated approvals that were converted to regular approval were rated as having category 1 evidence. Despite low-quality evidence ratings, there were high levels of consensus among NCCN committee members. Among 21 accelerated approval indications that were withdrawn, over one-third continued to be recommended at a level representing uniform consensus even after the corresponding guidelines were updated post-withdrawal, reinforcing that the guidelines are frequently broader than FDA-approved labeling.
The FDA has increasingly relied on nonrandomized trials and studies using surrogate efficacy measures for confirmatory studies of drugs with accelerated approval and for drugs with regular approvals. The authors feel this is problematic, given the frequency with which oncology drugs that show promising efficacy in phase II studies fail to show benefit in phase III randomized clinical trials. The authors suggest that greater clarity on the thresholds and definitions of evidence levels is needed to make the NCCN guidelines more useful for optimal patient care.
High level
The authors suggest that better clinical evidence is needed to ensure optimal use of cancer drugs and ensure they offer clinically meaningful benefits for patients. They propose the addition of an intermediate category of evidence within the NCCN guideline recommendations to help clarify thresholds for high-level vs lower-level evidence. It should also be made clearer that only 85% of votes are required for a recommendation to be classified as “uniform consensus.” Policymakers are encouraged to think more broadly about how to align reimbursement with level of evidence, such as providing lower reimbursement for drugs approved via accelerated approval or excluding coverage for withdrawn indications.
Ground level
While the NCCN guideline recommendations remain a valuable resource for informing treatment decisions, providers should look deeper into the levels of evidence for specific recommendations, as well as current drug approval status, when considering options for their patients. Staying knowledgeable about the latest approvals and withdrawals can help clinicians make informed treatment decisions with their patients.
