Impact of physicians’ awareness of prostate-specific antigen doubling time (PSADT) on treatment (Tx) decisions in high-risk (HR) biochemically recurrent (BCR) prostate cancer (PC).

Abstract

354 Background: PSADT is one of the strongest predictors of outcomes in patients (pts) with BCR PC and a criterion for HR BCR definition. As such, it is crucial to determine whether physicians are aware of pts’ PSADT and how this influences Tx in routine practice. We compared Tx patterns in pts with HR BCR whose PSADT was known (kPSADT) or unknown (uPSADT) by the physician. Methods: This physician-abstracted chart review used data of pts with HR BCR from the Cardinal Health Oncology Provider Extended Network (2018–2020) in the US. HR BCR definition: PSADT ≤9 months (mo) with PSA at or above the threshold per the EMBARK trial. Follow-up was from the index date (date on which HR BCR definition was met) until disease progression, last follow-up, or death through 2022. Physicians reported PSADT in case report forms (CRFs) using doubling time from labs, clinical judgement, or an online calculator (kPSADT). If not provided in the CRF, PSADT was calculated retrospectively based on PSA values up to the index date that the physician had entered in the CRF (uPSADT). We analyzed time to treatment after index via the Kaplan–Meier method. Results: Among 284 pts with HR BCR, median time from initial PC diagnosis to the end of follow-up was 39.1 mo; most pts had kPSADT). There were differences between in age, time from localized PC diagnosis to BCR, Gleason score, and PSA at initial diagnosis and at BCR (Table). A higher proportion of pts with uPSADT had a fast PSADT (≤3 mo) (61% vs 20%, P < 0.001). A higher proportion of pts with kPSADT (64%) vs uPSADT (17%) received Tx within 60 days after index, with a shorter median time to Tx (1.0 vs 6.7 mo; HR: 3.4, 95% CI: 2.6–4.4, P < 0.0001). Conclusions: Most physicians did not know their pts’ PSADT. Even though pts with HR BCR PC who had kPSADT were older and had slower PSADT, they were over three times more likely to receive Tx vs pts with uPSADT. The results suggest that many pts with HR BCR PC may be missed in clinical practice, which limits these pts’ opportunity to receive guideline-concordant Tx to delay progression. Characteristics of index kPSADTn = 104 uPSADTn = 180 P value Age (years), mean (SD) 70.0 (7.6) 65.6 (7.0) <0.001 Primary definitive PC Tx, n (%) Radiotherapy 28 (27) 51 (28) Prostatectomy 76 (73) 129 (72) Time from localized PC diagnosis to BCR ≤2 years, n (%) 97 (93) 143 (79) 0.002 Gleason score ≥8, n (%) 76 (73) 81 (45) <0.001 PSA before initial localized PC diagnosis (ng/mL), median (IQR) 9.0 (6.7) 7.9 (8.6) 0.019 PSA at BCR (ng/mL), median (IQR) 3.4 (9.0) 2.6 (2.8) <0.001