Profound impact of ECOG performance status on clinical trial eligibility and access: Diverging effects by cancer type.

Abstract

97 Background: The Eastern Cooperative Oncology Group (ECOG) performance status is a widely used measure of a patient's (pt) physical function and is often used as a criterion for clinical trial eligibility. Strict use of this criterion may exclude pts with comorbidities or poorer performance status, limiting the generalizability of trial results to the broader cancer pt population. This study aimed to evaluate the impact of ECOG performance status on eligibility for solid and non-solid cancer clinical trials (CTs) among pts using the Leal web-based platform. Methods: Leal (formerly, Trialjectory) is an AI-based platform that matches cancer pts to CTs based on a self-reported questionnaire that generates an individual clinical profile. The profile includes parameters essential for CT matching such as disease status, stage, biomarkers, treatment history, comorbidities and demographics. Data from pts with solid and non-solid cancers were analyzed. The number of matched trials per pt was determined based on the pt’s profile, location and travel preference. To assess the impact of ECOG score, the data set was edited to reflect a better performance status per pt. We used Yuen’s robust paired samples analysis to assess the association between improving the ECOG score and the number of matched CTs. Moreover, we have collected the cumulative ECOG distribution as an eligibility criteria in CTs in various cancer types. Results: In a data set of 5,018 cancer pts, high ECOG score (3 or 2) was significantly associated with fewer matched clinical trials per pt (p<0.001). This relationship was independent of ethnicity and gender. The relationship varied by cancer type, with a 3-fold greater effect size in solid tumors compared to non-solid tumors. When ECOG score per pt was edited from 3 to 2 or from 2 to 1, the number of matched CTs significantly increased by 246% and 189%, respectively, for all cancer types (p<0.001). Notably, the prevalence of CTs that allow worse ECOG scores (2 or higher) was significantly higher in non-solid CTs compared to solid tumors (p<0.001). Conclusions: Our results show that strict ECOG scores are significantly associated with fewer matched CTs per pt. This relationship varies by cancer type, with a greater impact on solid tumors. These findings suggest that cancer type may influence the relationship between ECOG scores and access to CTs. Pts with solid tumors may face more barriers to trial enrollment due to strict ECOG requirements. Our results highlight the need for a more nuanced approach to using ECOG performance status as an eligibility criterion in CTs, taking into account the specific cancer type and its associated physical demands. Further research is needed to fully understand the impact of ECOG performance status on trial enrollment and to determine the optimal balance between pt safety and representation in clinical research.