Background: A lack of consensus exists as to the timing of kidney biopsy in children with steroid-dependent nephrotic syndrome (SDNS) where minimal change disease (MCD) predominates. This study aimed at examining the applicability of a biomarker-assisted risk score model to select SDNS patients at high risk of focal segmental glomerulosclerosis (FSGS) for biopsy.
Methods: Fifty-five patients with SDNS and biopsy-proven MCD (n = 40) or FSGS (n = 15) were studied. A risk score model was developed with variables consisting of age, sex, eGFR, suPAR levels and percentage of CD8+ memory T cells. Following multivariate regression analysis, total risk score was calculated as sum of the products of odds ratios and corresponding variables. Predictive cut-off point was determined using receiver operator characteristics (ROC) curve analysis.
Results: Plasma suPAR levels in FSGS patients were significantly higher, while percentage of CD45RO+CD8+CD3+ was significantly lower than in MCD patients and controls. ROC analysis suggests the risk score model with threshold score of 16.7 (AUC 0.84, 95% CI 0.72-0.96) was a good predictor of FSGS on biopsy. The 100% PPV cut-off was >24.0, while the 100% NPV was <13.3.
Conclusion: A suPAR and CD8+ memory T cell percentage-based risk score model was developed to stratify SDNS patients for biopsy and for predicting FSGS.