Background: IMbassador250 (IM250) was a prospective phase III trial which showed no overall survival (OS) benefit for adding atezolizumab to enzalutamide for men with mCRPC who had prior progression on abiraterone. Analyzing IM250, we previously demonstrated baseline TF and 75% reduction in TF (using a prototype TF) both to be prognostic. Here we hypothesized that early detection of and changes in ctDNA TF are associated with clinical responses. Methods: Pre-treatment (but post-abiraterone progression) plasma samples from IM250 were profiled using FoundationOne Liquid CDx (F1LCDx). To enable tumor-naïve ctDNA monitoring while avoiding non-tumor signal, we developed FoundationOne Monitor (F1M) which leverages the same sequencing platform as F1LCDx and enables quantification of and changes in ctDNA TF from prior F1LCDx or F1M results. Changes in status of TF (detected vs not detected) from baseline to cycle 3, day 1 (C3D1, 6 weeks), as well as detection at C3D1 alone, were compared vs radiographic progression-free survival (rPFS) and overall survival (OS), as well as changes in PSA (50% reduction). Results: A total of 418 patients with advanced mCRPC were included: median age: 70, median baseline PSA: 68.7 ng/ml, median baseline TF: 15.5%. In the cohort, 335 patients (80%) had detectable TF at baseline and 303 (72%) had detectable TF at C3D1. TF detected at C3D1 was associated with shorter rPFS (HR 3.24, 95% CI: 2.46-4.28, p<0.001) and OS (HR 5.03, 95% CI: 3.41-7.41, p<0.001). Patients with TF detected at both baseline and C3D1 had median rPFS: 4.1 months and median OS: 12.6 months. 88% of patients with “ctDNA detected” at C3D1 had a rPFS≤6 months. The positive predictive value (PPV) for ctDNA status in predicting non-durable response was 74% versus 67% for PSA. When TF and PSA reduction at C3D1 were discordant, patients with [TF undetected/PSA not reduced] had more favorable outcomes compared with [PSA reduced/TF detected] (median OS 22.1 months vs. 16 months, p<0.001). TF also provided additional stratification for patients with clinically ambiguous results who had no radiographic progression but lack of >50% PSA reduction, resulting in high and low risk populations (median OS 13.0 months vs 20.5 months, HR=3.80, p<0.001). Conclusions: Here we report on a new tumor-naïve monitoring assay (F1M) for molecular response assessment which is based on ctDNA TF detection and dynamics. TF detection at C3D1 was linked to unfavorable outcomes and identified early progression post-abiraterone, with increased information derived from TF detection at baseline. TF on F1M thus complements PSA testing, which had a lower PPV for identifying early progression than ctDNA. Together, we provide a non-invasive strategy that is independent of and additive to PSA and may refinepersonalized approaches tailored to the individual patient’s risk of progression. Clinical trial information: NCT03016312.