FDA Analysis of Screen Failures By Race in Clinical Trials of Acute Myeloid Leukemia

Abstract

Background: Ethnic and racial minorities are underrepresented in oncologic clinical trials. One potential barrier to participation may be narrow eligibility criteria which contribute to an increased rate of screen failure among minorities. Prior work in multiple myeloma demonstrated a higher rate of screen failure among black or African-American (AA) patients and lower rate of screen failure among Asian patients compared with white patients (1). Here, we examine screen failure rates by race in patients screened for participation in clinical trials of acute myeloid leukemia (AML). Methods: Trials submitted to the U.S. Food and Drug Administration (FDA) to support approval of new treatments for AML between 2016-2019 were examined. Trials which included information on screen failures and race were included. Trial, screen failure status, basic demographic information (age, gender, race, and ethnicity), country of participant, and reason for screen failure (if relevant) were abstracted from the trials. Screen failure rates were calculated based on this information. Results: Twenty-one trials of AML therapies were identified, of which screen failure and race information were available in fourteen. Of these, information on reason for screen failure was available in nine. A total of 6,471 patients were identified across fourteen trials (Table 1). Race was recorded as unknown for 2,934 (46%) of patients. A total of 3,372 (52%) screen failures occurred. Minorities were more likely to be screen failures at 201/603 (33%) compared to whites at 27%. Of minorities, screen failures occurred in 29% of black/AA patients, 36% of Asians, and 36% of patients of any other race. A total of 579 patients were included in the nine trials that included data on reasons for screen failures (Table 2). Lack of appropriate mutation (i.e. IDH1/2, FLT3) was the most common reason for screen failure overall and was more common in black/AA (41%) and Asian patients (66%) compared to white patients (29%). Other issues related to the disease course (e.g. not at the appropriate point in treatment, not treated within the time limit set in the protocol) led to screen failure in 28% black/AA patients versus 9-22% for other races (19% overall). White patients were more likely than patients of other races to be screen failures due to cardiac issues (4%) and laboratory abnormalities (6%). Conclusions: Minority patients were numerically more likely to be screen failures compared to whites in trials of AML therapeutics, which exemplifies the known issue of lower minority recruitment in clinical trials. Minority patients were more likely to be screen failures due to lack of appropriate mutation and less likely due to lab abnormalities or cardiac issues. The significance of this finding is uncertain given the limited information on reasons for screen failure and the small number of minority patients in the database. Further investigation into screen failures in minority patients with AML is indicated. These results underscore the importance of including racial and ethnic minorities in all phases of translational and clinical research to further precision oncology for all patients since oncogenic driver mutation frequencies may vary among different populations. (1) Kanapuru B, Fernandes L FDA analysis of multiple myeloma trials supporting approval; Presented at FDA-AACR workshop to examine under-representation of African Americans in multiple myeloma clinical trials 2/13/20 Disclosures No relevant conflicts of interest to declare.