Background:  Although the incidence of end stage renal disease has declined among HIV-1-infected persons in association with HAART, the effects of ART on kidney function remain incompletely understood.

Methods:  The association between changes in glomerular filtration rate (GFR) and changes in plasma HIV-1 RNA in response to HAART were examined using multivariable, longitudinal models in subjects beginning their first, or a new HAART regimen through an AIDS Clinical Trials Group (ACTG) randomized clinical trial, who were followed in the Longitudinal Linked Randomized Study (ALLRT), a large, multi-center, observational cohort study. GFR was calculated using the abbreviated modification of diet in renal disease equation. Models were adjusted for demographic factors, immune and viral baseline factors, ART history, and co-morbidities.

Results:  Serum creatinine and HIV-1 RNA measurements were obtained for 2159 subjects who enrolled in ALLRT, of whom 704 (33%) had mild to moderate renal insufficiency (<90 cc/minute/1.73 M²) at baseline. Additional median baseline characteristics included:  age 39 years, CD4 cells 248/mL, HIV RNA 4.64 log copies/mL.

The rate of GFR change depended on baseline renal function with significant GFR improvement observed in subjects with renal impairment that was sustained for >6 years (average slope = 2.12 cc/minute/year/1.73 M²), contrasting with gradual GFR declines in subjects with normal renal function (average slope = 1.42 cc/minute/year/1.73 M²). Models of GFR improvement within an observation interval demonstrated an inverse association with HIV-1 RNA changes that was accentuated in subjects with baseline renal impairment. Additional correlates of renal improvement included younger age and the absence of prior ART, but these predictors did not include race. In models of categorical GFR change among subjects with baseline renal impairment, the odds of a >10% GFR improvement from baseline within an observation interval increased in association with greater reductions of plasma viremia (OR 1.51, 95%CI 1.01 to 2.25; 2.60, 95%CI 2.15 to 3.15, with 1.0 to 2.0 and >2.0 log reductions, compared to <1.0 log reductions).

Conclusions:  Sustained improvement in renal function was apparent with initiation or change in HAART among persons with mild to moderate renal impairment at the start of therapy. This improvement was associated with reductions of plasma viremia, supporting a direct relationship between HIV-1 replication and kidney function.