Impact of the Taylor Cancer Research Center (TCRC) and Tempus AI TIME Partnership on clinical trial start-up and accrual.

1565 Background: TCRC is a nonprofit oncology community practice serving patients in Northwestern Ohio. TCRC's partnership with the Tempus AI TIME program leverages proprietary software and nursing teams to support patient matching across a portfolio of industry-sponsored clinical trials. Trials may be rapidly opened "prospectively" before an eligible patient is identified, or "just-in-time" (JIT) once a patient is identified and ready to consent, allowing studies to be activated in response to patient need. The goal of this partnership is to utilize AI-enabled technology and clinical workforce support to alleviate site burden associated with activating and enrolling patients to oncology clinical trials, thereby increasing patient access to novel therapies and optimizing resource utilization. Methods: Tempus’ AI-enabled platform analyzes trial eligibility criteria and compares this against patient-level information from a combination of structured EMR data, natural language processing, and large language models. All patients are evaluated against the full TIME trial portfolio nightly, with potential matches reviewed by Tempus nurses. Patients likely to be eligible for a trial are communicated to TCRC for final confirmation. Trials are rapidly activated through a standardized operational model including a pre-negotiated clinical trial agreement, rate card, and central IRB. Program metrics include number of patients pre-screened, trial activation timelines, and trial consents. Results: From June 2024 to August 2025, Tempus nurses spent approximately 219 hours pre-screening 1,195 patients from TCRC, yielding 278 potential matches across 92 TIME trials. TCRC activated 16 TIME trials, leading to a total of 66 consents. Within the same time, TCRC activated 9 non-TIME trials resulting in 30 consents. TIME trials averaged 4.1 consents per trial, compared to 3.3 consents for non-TIME trials. TIME activations averaged 10.5 business days for JIT and 15.9 days for prospective trials, versus 55.4 days for non-TIME trials. Conclusions: TCRC's partnership with the Tempus AI TIME program, including enhanced patient pre-screening resources and rapid trial activation model, reduced trial start-up time with an increase in total consents and consents per trial compared with traditional trials. TCRC activated TIME trials 4x faster with a 2x increase in interventional consents and 120% increase in overall consents compared with non-TIME studies. This demonstrates how AI-enabled trial matching, combined with a streamlined operational model, can improve resource utilization and patient access to precision oncology trials in a community setting. Average activation time (days) Interventional trials Consents per trial (interventional) Observational trials Consents per trial (observational) Total consents Consents Per Trial TIME 13.7 12 2.3 4 9.8 66 4.1 Non-TIME 55.4 6 1.1 3 7.6 30 3.3