Cytoreductive Radical Prostatectomy (cRP) might be an option in oligometastatic Prostate Cancer (PCa). However, no preoperative risk tool is available to select candidates for cRP. We aimed at developing a model to identify patients with oligometastatic PCa who are suitable for surgery.
152 oligometastatic patients treated with cRP with extended pelvic lymph node dissection (ePLND) in a multimodal setting at four referral centers were identified. Patients were eligible for cRP if they had: (1) resectable PCa; (2) ≤5 osseous metastases; (3) no retroperitoneal nodal metastases; (4) no bulky (>3 cm) pelvic nodal metastases; (5) no visceral metastases. CR was defined as de novo metastases at follow-up. Multivariable Cox regression analyses tested the impact of clinical features (biopsy grade, cT and cN stages, PSA and neoadjuvant systemic therapies) on CR. A regression tree analysis stratified patient into risk groups for CR according to preoperative characteristics.
Median follow-up was 46 months. Overall, 63 (41%), 41 (27%) and 82 (54%) patients had biopsy grade group 5, ≥cT3b and cN1. Neoadjuvant ADT was administered in 90 (59%) patients while the remaining proportion received systemic therapies after surgery. The 5-year CR- and cancer-specific mortality-free survival rates were 51 and 75%. At multivariable analyses, a clinical stage ≥T3b represented the strongest predictor of CR (Hazard ratio: 1.94; p=0.01) At regression tree analysis, patients were stratified in five risk group according to cT stage, cN stage and biopsy Gleason grade. This model depicted a moderate discrimination (AUC 76%).
Surgery might achieve cancer control in selected patients with oligometastatic PCa and favorable features (grade group 1-3 and absence of locally advanced disease). Local disease features have a prognostic role in the oligometastatic setting and should be considered for the selection of candidates for cRP.