A nomogram for the prediction of renal outcomes among patients with idiopathic membranous nephropathy

Abstract

The early prediction of renal outcomes in patients with idiopathic membranous nephropathy (iMN) remains challenging. The present retrospective study evaluated patients with iMN confirmed by renal biopsy. An optimized Cox regression model and a nomogram were constructed for the early prediction of renal outcomes. A total of 141 patients who met the inclusion criteria were evaluated in the present study. In total 18 (12.8%) patients eventually progressed to the endpoint, 6 of whom developed end-stage renal disease, and one patient died during follow-up. The optimized model demonstrated that 24-h proteinuria [hazard ratio (HR) 1.24; 95% CI, 1.10-1.40; P-value <0.001] and chronic tubulointerstitial injury [referred to as grade 0, grade 1 (HR), 5.12; 95% CI, 1.33-19.75; P-value=0.02] or grade 2 (HR, 6.43; 95% CI, 1.35-30.59; P-value=0.02) were independent risk factors for a poor renal outcome. Patients with an estimated three-year renal survival rate (ETR) less than 0.87 had a high risk of a poor renal outcome. In addition, patients with an ETR of 0.87 to 0.98 more quickly developed a decreased estimated glomerular filtration rate after two years of follow-up. In the present study a nomogram for the early prediction of renal outcomes in patients with iMN was developed. This nonogram suggested that patients with an ETR of 0.87-0.98 should receive greater attention during follow-up.

Keywords: chronic kidney disease; idiopathic membranous nephropathy; prognostic model; renal outcomes.

Figure 1

Nomogram of a model for estimation of the probability of renal survival and bootstrap calibration. (A) Nomogram of the optimized model. Points were assigned to parameters by drawing lines upward from the corresponding values to the ‘Points’ line. The sum of these points, plotted on the ‘Total points’ line, corresponds to the predicted three-year survival as the primary endpoint. (B) Calibration plot of the nomogram. The predictive line (solid line) overlaps well with the ideal line (dotted line), indicating that the predictive value approximates the actual value.

Figure 2

Comparison of the differences in the AUC between the ETR and traditional risk factors (24-h proteinuria, eGFR, and serum uric acid) for the identification of poor renal outcomes. AUC, area under the ROC curve; CI, confidence interval; eGFR, estimated glomerular filtration rate; ROC, receiver operating characteristic.

Figure 3

Trends of proteinuria and the eGFR for different ETRs. Cases were divided into four groups according to the quartiles of the ETR. (A) Changes in the urine protein-to-creatinine ratio. (B) Changes in the eGFR during the follow-up. Points and bars represent the mean and standard error, respectively. eGFR, estimated glomerular filtration rate; ETR, estimated three-year renal survival rate.