Effect of a novel artificial intelligence (AI) –enabled multi-trial matching system on patient matching using real-world data.

Abstract

e13501 Background: Currently clinical research teams at sites are manually matching patients to clinical trials, where each trial in our experience takes at least 25 minutes using highly trained dedicated resources. This limits cancer patients’ clinical trial options and enrollment rates. Moreover, the lack of NGS testing results may also negatively impact matching to clinical trials in targeted therapies. PCC, a consortium of 7 of the Top 20 Pharmaceutical companies, collaborated with Massive Bio to demonstrate how cancer patients’ matching rates can be increased if patients are prescreened to multiple trials simultaneously (i.e. multi-trial matching) instead of a single trial and NGS testing results are integrated into the matching process. Methods: Massive Bio has developed a first-in-class AI-enabled matching system to extract 180 structured clinical data from medical records using computer vision and natural language processing and a recommendation system for matching patients to parameterized inclusion/exclusion (I/E) criteria from over 14,000 actively recruiting interventional clinical trials.Our analysis included a sample of 5,707 cancer patients from the Massive Bio Synergy AI real-world database, corresponding to tumor types in 23 selected trials. We developed a decision-tree algorithm for retrospective matching analysis using modeled I/E criteria and compared these results to the theoretical matching rate (TMR) for each trial based on tumor type and biomarker prevalence, extent of disease at diagnosis and prior treatment history. Results: 690 patients matched to at least one trial were matched to 1.82 times more trials in multi-trial matching; 2.10 times more potential matches were obtained with hypothetical 100% NGS testing; and 99.9% matching time was reduced with AI enabled system; key findings are summarized (Table). Conclusions: This study confirms the underutilization of NGS technology and substantial benefits of AI and NGS in cancer trial matching. Their use resulted in a nearly two-fold increase in potential patient eligibility for trials, and an unprecedented reduction in manual effort, underscoring a significant advancement in the efficiency of the clinical trial matching process. A prospective country-wide analysis is under consideration. [Table: see text]