From eligibility to enrollment: Evaluating barriers to clinical trial participation for leiomyosarcoma (LMS) at MD Anderson Cancer Center (MDACC).

Abstract

e23531 Background: LMS is a rare, aggressive malignant tumor of smooth muscle origin, and its clinical heterogeneity complicates diagnosis, treatment, and research. Clinical trial enrollment among adult sarcoma patients (pts) remains low, with many eligible pts not routinely identified or referred. Understanding enrollment barriers and improving screening are critical to increase trial access. This study evaluated a standardized manual pre-identification and referral process for LMS pts at MDACC and assessed its impact on screening, enrollment, and participation barriers. Methods: This prospective study implemented a systematic pre-identification and referral workflow for LMS pts (Sept-Dec 2025). Medical charts were reviewed 2-3 days before each outpatient sarcoma oncology visit in EPIC to document tumor site, histopathology, prior therapy, stage, and inclusion/exclusion criteria for 6 active LMS trials in pre- and post-intervention periods. Eligible pts were identified and referred to physicians and the trial team by email. Primary outcomes: % improvement in monthly screening and enrollment. Secondary outcomes: the average monthly pts screened and enrolled during pre-intervention (03/2025-08/2025: 6 months) vs post-intervention (09/2025-12/2025: 4 months) periods, and recording of reasons for non-inclusion. Results: A total of 13 pts were screened and 12 enrolled during the pre-intervention period, versus 15 screened and 12 enrolled during the post-intervention period. Implementation of the pre-identification and referral workflow led to 73% and 50% improvement in monthly screening and enrollment, respectively. The average monthly screening of pts increased from 2.2 (±1.2) to 3.8 (±1.9), and the enrollment from 2.0 (±1.4) to 3.0 (±1.4). Example from one LMS trial (protocol 2023-0710): no pts screened/enrolled pre-intervention; post-intervention: 25 pre-identified, 2 screened and enrolled. Reasons for non-enrollment across all trials: summarized in Table 1, with clinical/protocol ineligibility most common. Post-intervention data collection is ongoing; a full 6-month comparison to be included in the final presentation. Conclusions: The implementation of pre-identification and referral workflow showed a trend toward increased screening and enrollment in LMS trials, with trial-level examples suggesting improved pt identification. The study also identified LMS-specific barriers, defining targets for future interventions. Incorporating AI-assisted screening may further increase pt identification, reduce manual workload, and improve trial enrollment. Reasons for non-enrollment among screened pts (%). Reason for non-enrollment (%) Clinical/protocol ineligibility 58 Difficulty in traveling to MDACC 17 Patient preference 14 Alternative treatment selection 7 Communication/administrative barriers 4