On-Clamp Versus Off-Clamp Robot-Assisted Partial Nephrectomy for Localized Renal Tumors: A Retrospective Single-Center Cohort Study

Background: Robot-assisted partial nephrectomy (RAPN) is an established nephron-sparing technique for localized renal tumors. It is performed using on-clamp (temporary renal artery clamping) or off-clamp (without hilar clamping) strategies. Comparative real-world evidence remains limited and is often confounded by non-randomized treatment allocation. Methods: This retrospective single-center study included 146 consecutive patients undergoing RAPN between 2020 and 2025. Patients were allocated to on-clamp (n = 108) or off-clamp (n = 38) groups based on tumor characteristics and intraoperative surgeon judgment. Perioperative, functional, and early oncological outcomes were analyzed. Tumor complexity was assessed using the RENAL nephrometry score. Surgical quality was evaluated using the Trifecta outcome (negative margins, warm ischemia time ≤25 min, and absence of Clavien–Dindo ≥III complications). Results: Off-clamp RAPN was more frequently applied in smaller tumors (p = 0.008), while RENAL scores were comparable between groups. Estimated blood loss was higher in the off-clamp group (260 ± 62 vs. 110 ± 35 mL; p < 0.0001), whereas transfusion rates and overall complication rates did not differ significantly. Trifecta achievement was similar between on-clamp and off-clamp RAPN (91.0% vs. 96.8%; p = 0.45). No significant differences were observed in early postoperative renal function (creatinine, hemoglobin, eGFR) or positive surgical margin rates. Conclusions: In this retrospective cohort, both on-clamp and off-clamp RAPN demonstrated comparable perioperative safety, functional outcomes, and early oncological efficacy. Differences in baseline tumor characteristics reflect selection bias rather than treatment effect. These findings support the feasibility of both techniques in appropriately selected patients, while highlighting the need for prospective comparative studies with adjustment for confounding factors.