Development and internal validation of a nomogram predicting 3-year chronic kidney disease upstaging following robot-assisted partial nephrectomy

Abstract Purpose: Aim of the present study was to develop and validate a nomogram to accurately predict the risk of chronic kidney disease (CKD) upstaging at 3 years in patients undergoing robot-assisted partial nephrectomy (RAPN). Methods: A multi-institutional database was queried to identify patients treated with RAPN for localized renal tumor (cT1-cT2, cN0, cM0). Significant CKD upstaging (sCKD-upstaging) was defined as development of newly onset CKD stage 3a, 3b, and 4/5. Model accuracy was calculated according to Harrell C-index. Subsequently, internal validation using bootstrapping and calibration was performed. Then nomogram was depicted to graphically calculate the 3-year sCKD-upstaging risk. Finally, regression tree analysis identified potential cut-offs in nomogram-derived probability. Based on this cut-off, four risk classes were derived with Kaplan-Meier analysis tested this classification. Results: Overall, 965 patients were identified. At Kaplan-Meier analysis, 3-year sCKD-upstaging rate was 21.4%. The model included baseline (estimated glomerular filtration rate) eGFR, solitary kidney status, multiple lesions, R.E.N.A.L. nephrometry score, clamping technique, and postoperative acute kidney injury (AKI). The model accurately predicted 3-year sCKD-upstaging (C-index 84%). Based on identified nomogram cut-offs (7 vs 16 vs 26%), a statistically significant increase in sCKD-upstaging rates between low vs intermediate favorable vs intermediate unfavorable vs high-risk patients (1.3 vs 9.2 vs 22 vs 54.2%, respectively, p < 0.001) was observed. Conclusion: Herein we introduce a novel nomogram that can accurately predict the risk of sCKD-upstaging at 3 years. Based on this nomogram, it is possible to identify four risk categories. If externally validated, this nomogram may represent a useful tool to improve patient counseling and management.