Effect of biopsy requirement on patient enrollment to phase I trials in cancer

Background: A recent trend in clinical trials is the requirement of biopsies (Bx), sometimes mandatory. The main reason is to better understand cancer biology pharmacodynamic effect. We sought to determine whether the requirement for Bx among this vulnerable population had any detrimental effect. Methods: We included patients (pts) enrolled in phase I trials from June 2022-August 2025 at a single NCI-designated comprehensive cancer center. Data collected included sex, race, age, body mass index (BMI), dates of consent/treatment start/last dose/off treatment/last contact/death, serum lab values, number of prior treatments, metastatic sites, and Bx site/approach/complications. An interventional radiologist reviewed images to assess safety, with targets deemed appropriate biopsied under image guidance. Outcomes were analyzed by Mantel-Cox test in Prism GraphPad v10. Results: 228 pts [male (n=104, 45.6%)], age 60.5, 23-82 (median, range), NHW-130 (57%), NHB-37 (16.2%), Hispanic-34 (14.9%), and Asian-27 (11.8%) consented to 25 clinical trials. Of these, 8 were mandated paired tumor Bx, 15 mandatory or optional Bx, and 2 did not require Bx. The most common diagnoses were colorectal (54, 23.7%), pancreas (29,12.7%), other GI (26, 11.4%), lung (18, 7.9%), breast (12, 5.3%), prostate (5, 2.2%), and others (84, 36.8%). Image guidance included ultrasound 74 (57.8%), CT scans 49 (38.3%), and others 5 (3.9%). Overall, 91 (39.9%) pts provided 128 Bx samples; 37 paired Bx, 49 only pre-dose Bx, and 5 only on-study Bx. Five pts did not undergo Bx because it was deemed unsafe/high risk. Sites of Bx included liver 63 (49.2%), lung 13 (10.2%), lymph node 15 (11.7%), peritoneum 8 (6.3%), bone 1 (0.8%), and others 28 (21.9%). Two patients developed pneumothorax and recovered without sequelae. The median duration from consent to start of study treatment was 18.5 days among Bx pts vs. 14 among non Bx pts (p = 0.001). The median duration of time on study was 62 days among Bx pts vs. 66.5 days among non Bx pts (p = 0.08). The overall survival (OS), calculated as time from first dose of treatment to death/last follow up, was 123 days among Bx pts and 163 days among non Bx pts (p = 0.049). In a univariable model, sex (males/women; HR 1.37; p=0.02), albumin (high/low, HR 0.47, p=0.0001), neutrophil-lymphocyte ratio (low/high, HR 0.68, p=0.003), hemoglobin (high/low, HR 0.71, p=0.01), BMI (high/low, HR 0.72, p=0.02), and AST (high/low, HR 0.68, p=0.004), were significant, while age, race/ethnicity, total bilirubin, platelet, ALT, # sites of metastases, and # prior lines of therapy were not. Conclusions: 40% of pts entering phase I trials underwent study specific Bx, had a median delay of 4.5 days in receiving the first dose of study medication and experienced lower OS by 41 days. Multivariable modeling will be presented to fully understand the intriguing observation of a compromise in OS among patients who underwent biopsies.