Validation of a Predictive Model for New Baseline Renal Function After Radical Nephrectomy or Robot-Assisted Partial Nephrectomy in Japanese Patients.

Abstract

Purpose: The study's aim was to externally validate a new predictive model for the new baseline glomerular filtration rate (NB-GFR) postnephrectomy among Japanese patients. Materials and Methods: Patients with renal tumors who underwent radical nephrectomy (RN) or robot-assisted laparoscopic partial nephrectomy (PN) at a single Japanese institution during the period 2000-2020 were retrospectively analyzed. The NB-GFR is defined as the final estimated glomerular filtration rate within postoperative 3-12 months. The correlation/bias/accuracy/precision of the equation was examined by comparing the calculated NB-GFR with the observed rate. Results: The study included 485 cases of RN and 1030 cases of robot-assisted laparoscopic PN. The correlation/bias/accuracy/precision of the new equation predicting the NB-GFR was 0.86/-0.92/95.9/-5.65-3.62 in robot-assisted laparoscopic PN and 0.79/-1.02/87.8/-6.26-3.91 in RN, respectively. The fractional polynomial regression line approximated zero and its pointwise 95% confidence interval (CI) was considerably tight for the majority of both cohorts. The 95% CI to discriminate NB-GFRs ≥45 mL/min/1.73 m2 from receiver operating curves was 0.96 (0.95-0.97) and 0.89 (0.87-0.92) in robot-assisted laparoscopic PN and RN, respectively. Various preoperative factors, including age, tumor size, complexity, body mass index, hypertension, and diabetes, did not affect the predictive ability (correlation >0.7) of the subgroup analysis. Conclusion: The novel simple equation can accurately predict NB-GFRs after radical and robot-assisted laparoscopic partial nephrectomies in Japanese patients. This model will help physicians choose surgical treatments for renal tumors in daily clinical practice.