Answers are grounded in this paper and cite their sources.
The International IgA Nephropathy Prediction Tool (IIgANPT) was incorporated into the 2021 Kidney Disease: Improving Global Outcomes (KDIGO) guidelines as a key criterion for assessing the prognosis of IgA nephropathy (IgAN). This study aimed to determine the most suitable model for the Chinese population, with or without the inclusion of race, and to conduct exploratory analyses examining whether the association between crescents and renal outcomes differs across IIgANPT-defined risk groups. Data from 998 IgAN patients diagnosed via renal biopsy between 2014 and 2019 at Hangzhou TCM Hospital Affiliated to Zhejiang Chinese Medical University were analyzed. The primary outcome was a sustained decline in eGFR of ≥ 50% or progression to end-stage kidney disease (ESKD). Prognostic risk estimates were generated using the IIgANPT. Model evaluation employed Kaplan–Meier survival analysis, time-dependent receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis. Exploratory subgroup and interaction analyses of crescents across IIgANPT-defined risk groups were performed using Kaplan–Meier survival analysis and Cox proportional hazards regression models. Over a median follow-up period of 4.2 years, 7.1% of patients reached the primary outcome. Both the race-inclusive and race-exclusive IIgANPT models demonstrated clear separation of survival curves, particularly between the higher- and highest-risk groups. The AUC values were 0.835 (95% CI: 0.787–0.883) for the race-inclusive model and 0.888 (95% CI: 0.855–0.920) for the race-exclusive model. The race-exclusive model had a calibration slope closer to 1 and lower ICI, E50, and E90 values than the race-inclusive model. Visual inspection of the calibration curves showed some differences between the two models across risk groups. Decision curve analysis showed greater net clinical benefit for the race-exclusive model across most threshold probabilities. In the crescent subgroup analysis, patients with C2 crescents exhibited significant survival curve separation from those with C0 and C1 crescents in the higher-highest risk group; however, Cox regression analysis did not demonstrate statistically significant interactions between crescent classification and risk group outcomes (race-inclusive model: HR = 0.15, 95% CI 0.02–1.2; race-exclusive model: HR = 2.3, 95% CI 0.17–31.0). Both the race-inclusive and race-exclusive IIgANPT models demonstrated good predictive performance in this Chinese cohort, with different strengths across discrimination, calibration, and clinical utility. The race-exclusive model showed a numerically higher AUC discrimination and greater net clinical benefit, while quantitative calibration metrics were also more favorable for the race-exclusive model. However, the relatively small number of outcome events limits the precision of comparisons between the two models. No statistically significant interaction was detected between crescent classification and IIgANPT-defined risk groups in this cohort, although the limited number of outcome events reduced the precision and statistical power of these exploratory analyses.
No takes yet. Share an insight, caveat, or question.
Xiang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: