IgA Nephropathy Prediction Model Aids 5-Year Survival Estimation

A new study has developed and validated a predictive model to estimate 5-year renal survival and support early risk stratification for personalized management in IgA nephropathy (IgAN).1

Designed with potential application in primary care settings, the model integrates routinely available clinical and pathological variables to enable early identification of patients at increased risk of disease progression, including in resource-limited environments outside of specialty nephrology centers.1

“By facilitating timely risk stratification and supporting evidence-based decision-making, this model may contribute to more precise, proactive, and resource-efficient care in IgAN, ultimately improving long-term outcomes,” wrote study investigator Yongqiang Lin, from the Wenzhou Hospital of Integrated Traditional Chinese and Western Medicine, and colleagues. “Further multicenter, prospective studies are warranted to validate its generalizability and support its integration into routine clinical practice.”1

Several prediction models have been developed to estimate renal outcomes in IgA nephropathy, including the International IgA Nephropathy Prediction Tool, which is widely used in clinical research settings. In this study, investigators compared the performance of their XGBoost-based model with the IIgANPT using an external validation cohort.2

To assess 5-year renal outcomes of IgAN patients, investigators conducted a retrospective cohort study with single-center model development and multicenter external validation, developing an XGBoost-based survival model and a corresponding nomogram.1

They collected 1000 bootstrap samples from the training set for variable selection. Variables with non-zero coefficients were recorded, and the 5 most frequently selected were used to construct both the XGBoost survival model and a corresponding nomogram. These include anemia status, estimated glomerular filtration rate stage, hypertension status, Oxford classification T score, and 24-hour urinary total protein.1

Investigators randomly assigned patients in a 7:3 ratio to training and internal validation cohorts. They assessed model performance based on discrimination, calibration, and clinical utility using ROC curves, Brier scores, calibration curves, and decision curve analysis. To enable direct comparison with the IIgANPT, investigators performed an additional external validation analysis redefining the endpoint as a ≥50% decline in eGFR.1

A total of 1135 patients with biopsy-confirmed IgAN were initially identified at Hangzhou Hospital of Traditional Chinese Medicine, of whom 723 met the inclusion criteria and were included in the training and internal validation cohorts. The external validation cohort comprised 352 patients from 3 independent centers.1

Upon analysis, the area under the curve (AUC) was 0.951 (95% CI, 0.914–0.988) in the training cohort, 0.927 (95% CI, 0.877–0.978) in the internal validation cohort, and 0.913 (95% CI, 0.870–0.955) in the external validation cohort. Brier scores were 0.029 and 0.045 for the internal and external validation cohorts, respectively.1

When investigators compared the IIgANPT using the ≥50% eGFR decline endpoint with the XGBoost, their model demonstrated increased discriminative performance, achieving an AUC of 0.915 compared to 0.715. Using a ≥40% decline in eGFR as the primary endpoint, investigators noted the model demonstrated increased sensitivity to early disease progression, potentially enabling identification of high-risk patients at an earlier stage.1

“An intuitive nomogram derived from the model enables individualized risk assessment at diagnosis, helping primary care physicians recognize high-risk individuals, optimize follow-up strategies, and guide early, personalized management,” concluded investigators.1

References:

1. Wang J, Chen T, Fu Y, et al. Development and preliminary validation of a predictive model for IgA nephropathy progression. Scientific Reports. Published online December 14, 2025. doi:https://doi.org/10.1038/s41598-025-32280-8

2. The International IgA NEPHROPATHY Prediction Tool. UK Kidney Association. Published 2021. Accessed December 17, 2025. https://www.ukkidney.org/health-professionals/information-resources/international-iga-nephropathy-prediction-tool