Background: The Food and Drug Administration’s (FDA's) accelerated approval (AA) pathway allows drugs to be approved using surrogate endpoints, while requiring that sponsors complete postmarketing requirement (PMR) studies verifying clinical benefit. Subsequently, FDA determines whether AA can be converted to standard approval or be withdrawn. It is unclear whether clinical practice guidelines reflect PMR results or FDA regulatory decisions. As most drugs granted AA are cancer therapies, we investigated whether and how National Comprehensive Cancer Network (NCCN) guidelines reflect underlying evidence supporting AAs and subsequent FDA decisions. Methods: In this cross-sectional study, we used Drugs@FDA to identify drug indications granted AA by FDA for oncologic indications from 2016-2018, PMR status and results, and whether AA was converted to regular approval or withdrawn. We excluded pediatric and tumor-agnostic indications. PMR studies were classified as positive if the results from primary efficacy endpoint were statistically significant (P<0.05) favoring the drug. For each AA indication, we searched relevant NCCN guidelines for their recommendations and examined if AA status, PMR results, and updated FDA decisions were reflected. Results: Among 36 oncology drug indications granted AA from 2016 to 2018, 29 (81%) NCCN guidelines did not mention that the drug was approved via AA pathway or had not yet verified clinical benefit. For all 36 (100%) indications, NCCN guidelines did not state that approval was based on surrogate endpoints. As of December 31, 2022, among 16 drug indications that were converted to standard approval, 13 (81%) guidelines reflected FDA’s updated decisions. Among 8 indications withdrawn after AA, 5 (63%) guidelines reflected the withdrawn status. Of 6 drug indications with negative PMR study results that were withdrawn, 2 (33%) were still recommended in NCCN guidelines thereafter. Conclusions: NCCN guidelines do not consistently reflect FDA regulatory status or PMR study results, highlighting the need for more up-to-date and accurate information to guide clinical practice.

*12 drug indications have not yet been converted or withdrawn, 2 (17%) of which had negative PMR results. †1 drug indication withdrawal was based on a terminated PMR; withdrawal was reflected in guidelines.