Severe adhesive probability score (SAPS): an imaging scoring system for predicting adherent perinephric fat in patients with infectious nonfunctional kidney

Abstract

To identify and validate clinical and radiological predictors of perirenal fat adherence and develop a prediction model for perirenal fat adhesions in infectious nonfunctioning kidneys. A retrospective study was conducted on 277 patients who underwent robotic-assisted or laparoscopic nephrectomy from January 1, 2018, to July 1, 2024. The 277 patients were divided into two groups based on the presence or absence of intraoperative renal adhesions: the adherent perinephric fat (APF) and non-APF groups. Radiological elements were assessed by two readers blinded to the presence of APF. Perirenal fat on cross-sectional computed tomography (CT) was selected as the region of interest (ROI) for texture feature extraction and analysis. To assess APF risk factors, we developed a SAPS risk scoring system using univariate and multivariate logistic regression analyses. The predictive model’s discriminatory power, calibration, and clinical utility were evaluated using receiver operating characteristic curves (ROC), calibration plots, and decision curves. The APF was observed in 116 patients. Univariate analysis associated preoperative creatinine, perinephric fat stranding (PFS), posterior perinephric fat thickness (PPFT), and radscore with APF. Multivariate regression determined PFS, PPFT, and texture features as predictors of APF. Based on this, the SAPS score (0–6) was created to predict APF. Patients with a score of 0 had a 4% chance of APF, 32% for a score of 1, 66% for scores of 2–4, and 85% for scores of 5–6. The presence of APF in patients receiving robotic-assisted or laparoscopic nephrectomy was accurately predicted by SAPS.