Broadening the gates: Analysis of potentially modifiable study entry criteria in pancreatic and biliary tract cancer trials.

BACKGROUND: Eligibility criteria for clinical trials are crucial for maintaining safety and study integrity. However, overly restrictive criteria can result in unrepresentative trial populations, leading to gaps in understanding real-world treatment efficacy. Addressing potentially modifiable exclusions (PMEs) could enhance trial accessibility and participation without compromising safety. METHODS: The authors retrospectively analyzed screen-failed patients with biliary tract cancer or pancreatic cancer between August 2019 and November 2024 at The University of Texas MD Anderson Cancer Center. Screen-failed patients were those who provided informed consent but did not participate for any reason. Clinical data were obtained from screening logs and electronic health records. PMEs were identified and validated by two independent medical oncologists. RESULTS: Of 585 screen-failed patients from 18 trials, 509 were analyzed (367 with pancreatic cancer and 142 with biliary tract cancer) after excluding 76 because of incomplete data. Leading causes of screen failure were declined participation (19%), comorbidities (12%), suboptimal organ function (11%), absence of a biomarker (10%), and insufficient biospecimens (7%). Reasons for declining included preference for standard care (23%), travel (22%), and competing trials (5%). The authors identified 69 patients (13.6%) who had PMEs (primarily borderline laboratory abnormalities), including liver function (23%), kidney function (20%), platelet count (12%), hemoglobin (7%), and white blood cell count (6%) abnormalities. Other PMEs included previous or concurrent malignancies (9%) and viral hepatitis (4%). PMEs were evenly distributed across trials. CONCLUSIONS: Rigid eligibility criteria exclude stable patients who might benefit from investigational treatments. Easing criteria related to incidental or asymptomatic laboratory abnormalities could broaden trial accessibility and improve enrollment in populations that are more representative of real-world use.