ECOG score as a barrier for clinical trial eligibility in the patient population with genitourinary cancer: Is it time to adapt this criterion to real-world data?

Abstract

78 Background: ECOG score is a common eligibility criteria used in clinical trials (CT). Many CT recruit only patients in good physical condition (ie. minimal comorbidities and good performance status), which may misrepresent the real-world cancer population.This gap may result in safety and effectiveness data that may not be applicable to the real-world patient population. Furthermore, applying strict eligibility criteria augment a barrier for patient accrual to CT. In this study we analyzed the impact of ECOG performance status on Genitourinary (GU) oncology trials eligibility of patients who used the Trialjectory web-based platform. Methods: Trialjectory is an AI- based platform that matches cancer patients to CT based on a patient reported questionnaire that translates to an individual clinical profile. The profile includes disease status and stage, biomarkers, treatment history and comorbidities and demographics. Data from patients with prostate and urinary bladder cancer was analyzed. The number of matched trials per patient was determined based on the patient’s profile, location and travel preference. To assess the impact of ECOG score, we edited the data set to reflect a better performance status per patient. We used a paired samples Wilcoxon test to assess the association between decreasing the ECOG threshold and the number of matched CT. Results: Between May 2021 and August 2022 we analyzed 153 bladder cancer profiles with ECOG=2 and 71 with ECOG=3, and 90 and 27 prostate cancer profiles with ECOG=2 and ECOG=3, respectively. In a multiple regression analysis, high ECOG score was inversely and significantly associated with the number of matched trials per patient in prostate cancer (p=0.009), independent of cancer stage, patient ethnicity and gender; and in urinary bladder cancer (p=0.003), after adjustment for cancer stage, gender and ethnicity. Editing ECOG score from 2 to 1 resulted in a profound increase (233% for prostate cancer patients and 300% for urinary bladder cancer patients respectively) in the number of matched trials per patient, from a median of 3 trials to 7 and from 2 to 6 trials per patient respectively (V = 426, p-value <0.0001 and V = 14, p-value = 0.0002 respectively). Among Bladder Cancer patients, editing the ECOG score from 3 to 2 resulted in an increase of 200%, from a median 1 to 2 matched CT (V = 291, p-value = 0.0002). Conclusions: The results reflect a high magnitude reduction in eligibility for patients with higher ECOG scores. Excluding these patients may result in skewed reporting on effectiveness-toxicity of drugs that may not be applicable for the real-world patient population augmenting a significant barrier for CT accrual amongst the proactive population who seek it.