A novel ultra-low 6-month PSA cutoff for radiographic PFS stratification in high-volume mHSPC.

Abstract

e17099 Background: In metastatic hormone-sensitive prostate cancer (mHSPC), a 6-month PSA nadir ≤0.2 ng/mL is commonly applied as an early prognostic benchmark. Moreover, routine PSA reporting is often truncated at low values (e.g., < 0.1–0.2 ng/mL), limiting assessment of deeper PSA responses. Therefore, using ultrasensitive PSA measurements, we aimed to identify a novel ultra-low 6-month PSA nadir cutoff that better discriminates rPFS in high-volume mHSPC. Methods: A total of 159 patients with CHAARTED-defined high-volume mHSPC were included. The primary endpoint was rPFS; a 6-month landmark was used to mitigate immortal time bias. The optimal 6-month PSA cutoff for rPFS discrimination was identified using maximally selected rank statistics with log-rank testing across candidate thresholds, with multiplicity-adjusted p-values (R). rPFS was estimated by Kaplan–Meier, and hazard ratios were derived from univariable Cox regression. Results: Median age was 69.3 years (IQR 63.2-74.6), median baseline PSA was 100.5 ng/mL (IQR 25.6-306.3), and median follow-up was 32.8 months (IQR 20.2–44.7). Using maximally selected rank statistics, the optimal 6-month PSA cutoff was identified as 0.07 ng/mL. At month 6, 55/159 patients (34.6%) achieved PSA ≤0.2 ng/mL and 32/159 (20.1%) achieved PSA ≤0.07 ng/mL. Patients with PSA ≤0.07 ng/mL had significantly longer rPFS (median not reached vs 20.3 months; progression events 2/32 vs 84/127; log-rank p < 0.001). Compared with PSA ≤0.07 ng/mL, PSA > 0.07 ng/mL was associated with higher risk of radiographic progression (HR 14.27; 95% CI 3.51–58.05; p < 0.001). Conclusions: In high-volume mHSPC, the conventional 6-month PSA threshold of 0.2 ng/mL may be insufficient to capture the highest-risk patients. Our study identifies an optimal "ultra-low" cutoff of 0.07 ng/mL that provides superior prognostic discrimination for rPFS. Achieving PSA ≤0.07 ng/mL is associated with markedly superior outcomes, whereas PSA > 0.07 ng/mL identifies a population at imminent risk of radiographic progression. This deeper PSA response provides a pragmatic candidate threshold for assessing treatment efficacy and for risk stratification in high-volume disease, warranting validation in larger, well-designed studies. Baseline characteristics of the study population. Characteristic Overall population (n=159) Age, median (IQR), years 69.3 (63.2–74.6) ECOG performance status 0 69 (43.4%) ≥1 90 (56.6%) Metastatic status at diagnosis De novo 135 (84.9%) Recurrent 24 (15.1%) Visceral metastasis Yes 51 (32.1%) No 108 (67.9%) Charlson Comorbidity Index (CCI) Median (range) 3 (0–12) CCI <3 70 (44.0%) CCI ≥3 89 (56.0%) Baseline laboratory parameters ALP Normal Elevated 67 (42.1%)79 (49.7%) LDH Normal Elevated 68 (42.8%)72 (45.3%) PSA, median (range), ng/mL 100.5 (25.6-306.3) Systemic therapy ADT alone 39 (24.5%) ADT + ARPI 76 (47.8%) ADT + docetaxel 44 (27.7%)