Perirenal adipose tissue imaging characteristics as predictors of systemic inflammatory response syndrome following percutaneous nephrolithotomy.

BACKGROUND

Perirenal adipose tissue (PAT) is implicated in inflammatory responses, yet its potential value in predicting systemic inflammatory response syndrome (SIRS) after percutaneous nephrolithotomy (PCNL) remains underexplored. This study evaluated the predictive role of perirenal fat characteristics for post-PCNL SIRS.

METHODS

A retrospective analysis included clinical data from 723 patients undergoing PCNL.Patients were stratified into non-SIRS (n = 592) and SIRS (n = 131) groups based on postoperative outcomes. PAT imaging characteristics were extracted from imaging databases. Clinical characteristics and biochemical parameters were compared. Least Absolute Shrinkage and Selection Operator (LASSO) regression was used for variable selection, and logistic regression was employed to develop a risk model. Model performance was evaluated using receiver operating characteristic (ROC) analysis. Subgroup analyses and interaction tests were conducted to assess model stability. Calibration curves and internal bootstrap validation (1,000 resamples) were applied to evaluate model reliability.

RESULTS

LASSO regression confirmed six facto: Urine nitrite(NIT), Hydronephrosis, Operation time, Lateral Perirenal Fat Thickness(LPrFT); Posterior Perirenal Fat thickness(PPrFT); The ratio of perirenal fat area to renal parenchyma area(PFA/RPA ratio). The multivariate logistic model demonstrated strong discriminative ability (AUC = 0.827,95%CI:0.783 − 0.871). Subgroup analyses showed consistent effect directions and magnitudes. Bootstrap validation confirmed model stability (calibrated AUC = 0.827;Hosmer-Lemeshow p = 0.387). An interactive web-based risk calculator was developed via shinyapps.io to enhance accessibility for clinicians.

CONCLUSION

LPrFT, PPrFT and PFA/RPA ratio independently predict post-PCNL SIRS. The findings underscore the significant role of quantitative PAT characteristics as novel and accessible imaging biomarkers for postoperative inflammatory risk.