Impact of learning curve on perioperative outcomes of off-clamp minimally invasive partial nephrectomy: propensity score matched comparison of outcomes between training versus expert series.

Background: Off-clamp robotic partial nephrectomy (Off C-RPN) is a challenging technique, hard to teach since bleeding control is not easily reproducible in training settings. We compared perioperative outcomes of two propensity score matched (PSM) cohorts of patients undergone Off C-RPN by either a training or an expert surgeon in the same Institution.

Methods: The prospectively maintained "renal cancer" database was queried for "off-clamp," "robotic," "partial nephrectomy" performed between January 2017 and June 2018. Achievement of main outcomes along the learning curve of training surgeon was assessed with logistic regression and Lowess analysis. A 1:1 PSM analysis generated two populations homogeneous for demographics, ASA score, tumor size, nephrometry score, baseline hemoglobin and estimated glomerular filtration rate (eGFR). Multiple tumors, and imperative indications were excluded. Categorical and continuous variables were compared by χ and t-test.

Results: Overall, 111 were treated by the expert, 51 by the training surgeon, respectively. Training surgeon experienced a significant decrease of console time (P=0.01). Patients treated by the expert surgeon had significantly larger tumors, higher PADUA and ASA scores (all P≤0.04). After applying the PSM, two cohorts of 29 patients, homogeneous for all baseline demographic and clinical variables (all P≥0.34) were selected. Hilar clamping was never necessary. Hospital stay, hemoglobin and eGFR at discharge, complication and positive surgical margins rates were comparable between the two cohorts (all P≥0.15).

Conclusions: Our results proved that the impact of learning curve on outcomes of Off C-RPN is negligible after completion of a proper training in minimally invasive surgery.