Prognostic markers for overall survival and outcome to LuPSMA radionuclide treatment in patients with metastatic castration-resistant prostate cancer.

5548 Background: The aim of this international multicenter retrospective analysis was to identify prognostic markers for the clinical outcome in late-stage mCRPC patients treated with 177 Lutetium-prostate-specific membrane antigen (LuPSMA) radionuclide treatment. Methods: Patients with progressive mCRPC treated with LuPSMA at six centers in Germany, USA, and Australia were considered for inclusion. Eligible patients had 24 predefined, pretherapeutic covariates (demographics, prior mCRPC treatments, and PSMA PET/CT derived parameters) and survival data available. Endpoints included overall survival (OS) and PSA progression-free survival (PSA-PFS). Covariates were tested using univariate and mulitvariate proportional hazards regression Cox models. Results: 267/414 (64%) patients met inclusion criteria and were analyzed. 113 patients participated in clinical trials (ACTRN12615000912583, NCT03042312), while 154 were enrolled in compassionate-access programs. After a median follow-up of 22.5 months, median OS was 13.0 months (95%CI 11.6-14.4); 83% of the patients died. Median PSA-PFS was 4.0 months (95%CI 3.2-4.7). In the multivariate analysis, factors associated with shorter OS were: shorter time since diagnosis of prostate cancer (HR=2.04; p=0.002), lower number of prior systemic therapies (≤3; HR=1.56; p=0.006), prior exposure to chemotherapy (HR=1.42; p=0.05), lower hemoglobin levels (HR=1.13; p=0.002), higher number of lesions (≥20: HR=1.53; p=0.009), multiple sites of metastases (bone/LN only vs. bone + LN; HR=1.39; p=0.03) and visceral involvement (M1c) (HR=1.45; p=0.01). Factors associated with longer PSA-PFS were: longer time since diagnosis of prostate cancer (HR=0.44; p<0.001), higher hemoglobin levels (HR=0.32; p=0.03), presence of pelvic lymph nodes (LN) metastasis (N1) (HR=0.68; p=0.01), no distant lymph node metastases (M1a) (HR=0.66; p=0.01), no skeleton involvement (HR=0.44; p=0.01), no visceral metastases (M1c) (HR=0.51; p<0.001), higher PSMA-positive tumor volume (HR=0.87; p=0.04), and higher SUVmean (HR=0.94; p=0.002). Conclusions: This retrospective analysis identified prognostic factors for survival and treatment response to LuPSMA. Along with the conventional risk factors in mCRPC, PSMA PET/CT can be a useful tool for stratifying patients and guide patient’s selection for LuPSMA radionuclide treatment.