PSMA Radio-guided Surgery to Identify Lymph Node Metastases in Prostate Cancer Patients Undergoing Robot-assisted Radical Prostatectomy and Extended Pelvic Lymph Node Dissection: Results of a Phase 2 Prospective Study

We report the results of a prospective trial aimed at describing the feasibility and accuracy of prostate-specific membrane antigen (PSMA) radio-guided surgery (RGS) during robot-assisted radical prostatectomy (RARP) with extended pelvic lymph node dissection (ePLND) in patients with prostate cancer (PCa). This was a phase 2 study (NCT04832958) that enrolled 82 patients with localized PCa and a lymph node invasion (LNI) risk >5%. All patients underwent preoperative PSMA positron emission tomography (PET). [99mTc]Tc-PSMA-I&S was administered intravenously the day before surgery, followed by single-photon emission computed tomography/computed tomography. Side effects, perioperative outcomes, and the performance characteristics of PSMA-RGS for LNI detection were measured. A total of 62 patients completed all study procedures and were included in the final analyses. Median blood loss and length of stay were 50 ml and 4 d, respectively. No adverse events after tracer administration or intraoperative complications were recorded. Four patients experienced a Clavien-Dindo grade 3 complication within 30 d. Overall, PSMA-RGS identified two additional patients with high-risk PCa and pathologically node-positive (pN1) disease compared with preoperative PSMA PET, exhibiting a lower positive predictive value (PPV) and a similar negative predictive value (NPV) in the per-patient analysis (PPV: 50% vs 70%; NPV: 83% vs 83%). The PPV and NPV at per-region analysis in patients with molecular imaging node-positive (miN1) disease were higher for PSMA-RGS than for PSMA PET (PPV: 55% vs 42%; NPV: 85% vs 81%). In conclusion, PSMA-RGS identifies additional patients with pN1 disease missed by preoperative PSMA-PET among men with high-risk PCa, and extends the ePLND template in patients with miN1 disease, in whom PSMA PET underestimates the nodal burden. However, its NPV is suboptimal for avoiding ePLND in patients with PCa who have a LNI risk >5% and negative PSMA-RGS findings during RARP.