HIV-associated nephropathy: outcome and prognosis factors. Groupe d' Etudes Néphrologiques d'Ile de France.

Records of 102 patients with biopsy-proven HIV-associated nephropathy (HIVAN) admitted to 18 hospitals in the Paris area from 1984 through 1996 were retrospectively reviewed. Demographics and clinical and laboratory features of the cohort were determined, and prognostic factors of renal and patient survival were analyzed. Renal and patient survival curves were estimated with the actuarial method. Prognostic factors were assessed by uni- and multidimensional analyses based on Cox regression models. Values were expressed as median with interquartile. The total population (median age 34) included 97% blacks and 71.5% males. Median patient follow-up was 165 d (range, 43 to 493). At the time of renal biopsy, median values of serum creatinine, proteinuria, and CD4+ cell count were 496 micromol/L, 6.5 g/24 h, and 48.5 cells/mm3, respectively. Fifteen patients were given steroids after the onset of HIVAN. Overall patient survival at 0.5, 1, and 3 yr was 73 +/- 5, 55 +/- 6, and 38 +/- 7%, respectively. The proportion of patients free of dialysis at 0.5, 1, and 3 yr was 73 +/- 5, 60 +/- 7, and 18 +/- 10%, respectively. Predictors of poor patient prognosis were a low CD4+ cell count (relative risk [RR; per 50 cells/mm3 decrease] 1.35; confidence interval [CI], 1.13 to 1.6) and antiretroviral therapy before the onset of HIVAN (RR 1.9; CI, 1.05 to 3.6). Main independent factors associated with better renal outcome were: steroid therapy (RR 0.29; CI, 0.1 to 0.9); low proteinuria level (RR [per 50% decrease] 0.7; CI, 0.5 to 0.98); low serum creatinine (RR [per 1.1 mg/dl decrease] 0.78; CI, 0.7 to 0.87); and hemoglobin level (RR [per g/dl increase] 0.76; CI, 0.58 to 1.00). HIVAN is not a rare nephropathy in Paris and its suburbs. Renal prognosis and patient survival are better than what was reported previously. Steroids may delay the downward course of HIVAN. It is not certain that in the new era of HIV therapy, the possible renal benefits of corticosteroids outweigh their potential risks. The only reliable predictor of patient survival is the intensity of immunodeficiency.