Race and ethnicity reporting by US federal standards in high-impact phase 2/3 oncology clinical trial publications

1595 Background: There are 5 races defined by the US Office of Management & Budget (OMB): White, Black, Asian, American Indian and Alaska Native (AIAN), and Native Hawaiian and other Pacific Islander (NHPI) with Hispanic ethnicity. Despite the existing 1997 race standards, race reporting in oncology clinical trials is still dismal. Here we aim to 1) quantify cancer trial publications reporting race/ethnicity, 2) determine representation in published trials, and 3) identify geographic patterns of reporting race/ethnicity by 1st author. Methods: PubMed/Embase were queried for phase 2/3 clinical trials of the 6 most common cancers, published in 4 high-impact journals between 2012-2022, each reviewed by 2 independent reviewers. Trials were excluded for non-US 1st author and those including multiple cancers. Representation quotients (RQ) for each race were calculated by dividing the percent of clinical trial patients by the proportion of year-matched, site-specific incident cancers in the US (Cancer in North America data). RQ of 1 represents equal clinical trial representation compared to US cancer incidence. Trial characteristics were recorded. Descriptive statistics were calculated. RQ statistics were calculated using Kruskal-Wallis tests with Bonferroni Corrections ( P BC ) for multiple testing. Results: Of 1,149 publications assessed, 342 were included (224 JCO, 84 NEJM, 18 Lancet, and 16 JAMA) with 258,123 patients, 60% phase 3, and 45% US Northeast 1st author. By cancer site, 31% breast, 24% lung, 18% prostate, 18% kidney/bladder, 5% colorectal, and 4% endometrial publications were included. Race was reported in 245 publications (72%) varying by journal: 81% JAMA, 78% Lancet, 72% JCO, and 67% NEJM. Specific race/ethnicity were reported in 239 publications (70%) for White, 201 (59%) for Black, 185 (54%) for Asian, 62 (18%) for Hispanic, 48 (14%) for AIAN, and 29 (9%) for NHPI patients. Among publications reporting race, RQ varied after adjusting for multiple testing ( P BC <.001) with median (IQR) RQ of 1.0 (0.9-1.1) for White, 0.6 (0.1-2.7) for Asian, 0.4 (0.1-0.7) for Black, 0.0 (0.0-0.0) for AIAN, and 0.00 (0.0-0.0) for NHPI patients. Stratified by cancer, RQ significantly differed by race for all cancers ( P BC <.001), except colorectal ( P BC =.004) and endometrial cancer ( P BC >.99). NHPI was the only race with median RQ of 0.00 for all cancers. AIAN and NHPI race reporting by 1 st author was observed most in CA, while Hispanic was observed most in TX. Conclusions: While 72% of phase 2/3 cancer clinical trials published in high-impact journals report race, 91% do not report NHPI race. Indigenous patients have the lowest reporting in cancer clinical trials. Our findings support a call to action to mandate consistent race/ethnicity reporting for publication. US clinical trialists and publishers should comply with mandated US OMB federal reporting guidelines in oncology publications.