231 Background: The 18F-Fluciclovine (FACBC) PET/CT scan has higher sensitivity than traditional imaging modalities for detection of prostate cancer (PCa) recurrence. We aim to identify the PSA characteristics as predictors of positive FACBC scan. Methods: Seventy-eight patients who underwent FACBC scan in 2018 at our institution were identified. Patient demographics, prior PSA, clinical information including prior treatment was recorded. Scans were deemed positive if definitive lesions were noted at the prostate bed, pelvic lymph nodes, or skeletal level. Detection rate was the ratio of positive over total scans and was calculated for PSA ranges. PSA velocity (PSAV) was calculated for patients who did not initiate androgen-deprivation therapy 12 months prior to the scan. The baseline PSA and PSAV were modeled as predictors of having a positive scan using logistic regression. Results: The median baseline PSA is 2.7 (range 0.2 - 226.5). The rates of positivity increased with increasing baseline PSA (Table 1). Positive scans had higher median baseline PSA (3.0, range 0.2-226.5 vs. 1.2, range 0.2-26.0; p = 0.0015) and higher PSAV (median 2.3, range -0.6-1478.7 vs. 1.0, range -0.5-31.9; p = 0.025). Baseline PSA (AUC = 0.712) was better predictor than PSAV (AUC = 0.656) of a positive scan. Combining the two variables does not improve their predictive ability (AUC = 0.719). There is a 50% detection rate in post-radical prostatectomy (RP) patients (Table 2). Conclusions: FACBC PET’s detection rate increases with increasing baseline PSA. While higher PSAV is associated with higher rate of positive scan, it did not increase the predictive ability of baseline PSA for a positive scan in prostate cancer recurrence patients. [Table: see text][Table: see text]
Combining radiomics and mathematical modeling to elucidate mechanisms of resistance to immune checkpoint blockade in non-small cell lung cancer