IgA nephropathy: multivariate statistical analysis aimed at predicting outcome.

Background: Several risk factors of IgA nephropathy (IgAN) have been identified, but their importance in predicting outcome is still controversial.
Methods: We conducted a retrospective study on 119 patients (pts) with IgAN. All had a follow-up of over five years (mean 134+/-56 months). For each patient we recorded age, 24h proteinuria, hematuria, renal function (RF), arterial hypertension (AH) and histological features. Multivariate analysis was done for predictive purposes (segmentation, using Chi-squared Automatic Interaction Detection-CHAID).
Results: AH at the time of renal biopsy was the principal and independent predicting factor: 30/50 (60%) hypertensive pts had serum creatinine > or =1.5 mg/dL at the end of follow-up compared to 9/69 (13%) pts with normal blood pressure. Age was a further predictive parameter: 21/28 (75%) pts with AH and age over 39 years had reduced RF at the last examination. In this subgroup, 18/19 (95%) with evidence of tubulo-interstitial lesions showed a decline of RF.
Conclusions: AH and age alone are significant prognostic factors; tubulo-interstitial lesions are an additional pointer to poor outcome in these pts. The algorithm obtained with segmentation analysis may be a guideline for prognosis in single patients with IgA nephropathy.