Robotic Posterior Retroperitoneoscopic Adrenalectomy: 1-Year Experience at a High-Volume Endocrine Referral Center.

INTRODUCTION

Evidence on multiplatform implementation of robotic posterior retroperitoneoscopic adrenalectomy remains limited. We assessed first-year operative performance in a high-volume endocrine referral center experienced in laparoscopic posterior retroperitoneoscopic and robotic adrenal surgery.

METHODS

All consecutive adults undergoing unilateral total robotic posterior retroperitoneoscopic adrenalectomy between July 2025 and July 2026 were retrospectively analyzed. Operative times, perioperative outcomes, chronological trends, cumulative sum curves, and multivariable Gamma generalized linear models were assessed.

RESULTS

Thirty-four procedures were performed: 7 with Toumai, 25 with da Vinci Xi, and 2 with da Vinci SP. Median total operative, docking, console, and procedure times were 59.0, 6.0, 19.0, and 54.0 min, respectively. One conversion to laparoscopic posterior retroperitoneoscopic adrenalectomy occurred (2.9%). Two intraoperative complications (5.9%) and two 30-day postoperative complications (5.9%) were recorded, including one death after postdischarge acute myocardial infarction. Across the cohort, docking time decreased with consecutive case number (Spearman ρ = -0.623; p < 0.001), whereas total operative and console times did not. After adjustment for platform, body mass index, and lesion size in 32 multiport procedures, only docking time remained associated with case number (time ratio 0.980, 95% confidence interval 0.963-0.997; p = 0.021). Cumulative sum inflections occurred around Cases 7-8 but coincided with platform transition.

CONCLUSION

In this limited first-year experience in an expert adrenal surgery program, robotic posterior retroperitoneoscopic adrenalectomy implementation was feasible across three platforms, with adaptation reflected mainly by docking optimization. Platform-specific time differences were descriptive and cannot establish comparative efficiency because platform use was sequential and instrumentation differed across systems.