Effects of Highly Active Antiretroviral Therapy on Albuminuria in HIV-infected Persons

Background: Highly active antiretroviral therapy (HAART), especially tenofovir DFcontaining regimens, has been implicated in albuminuria. Objective: We prospectively evaluated the effects of HAART on albumin-to-creatinine ratios (ACRs) in antiretroviral-naïve HIV-infected individuals. Methods: One hundred and two (102) newly diagnosed, antiretroviral-naïve, human immunodeficiency virus (HIV)-infected persons were treated with Tenofovir disoproxil fumarate/ Emtricitabine/Efavirenz (TDF/FTC/EFV), n=33; Zidovudine/Lamivudine/Nevirapine (ZDV/3TC/NVP), n=53; and Zidovudine/Lamivudine/Efavirenz (ZDV/3TC/EFV), n=16. Diabetes mellitus and hypertension were excluded. ACRs and glomerular filtration rates (eGFR) were estimated at baseline, and at 1, 3, 6 and 9 months post-therapy; the prevalence of albuminuria (ACR ≥ 300mg/g), and microalbuminuria (ACR 30-300mg/g) were similarly estimated. HAART effects on normal ACR (0-30mg/g) were also monitored. Results: At baseline, one patient (0.9%) had nephrotic-range albuminuria with ACR of 2450mg/g. Overall, 8 (7.8%) patients had albuminuria; 53 (51.9%) had microalbuminuria; while 41 (40.2%) had normal ACRs, 28 (27.5% of 102) of which had nonalbuminuric renal insufficiency. eGFR and ACRs improved concurrently on HAART (ACR, Wilks’ lambda 0.439, power 0.763, p=0.032); albuminuria improved significantly on all the 3 regimens at 9 months (p=0.006, 0.012 and <0.001 respectively). Microalbuminuria resolved earlier (1 month) with ZDV/3TC/NVP than with TDF/FTC/EFV and ZDV/3TC/EFV (24.31mg/g versus 76.51mg/g and 63.59mg/g; p=0.028, 0.016 respectively). Microalbuminuria relapsed on TDF/FTC/EFV and ZDV/3TC/EFV at 6 months but resolved again at 9 months (66.7 versus 29 mg/g, p=0.006; and 51.2 versus 9.5mg/g, p=0.001 respectively); no relapse on ZDV/3TC/NVP. At 9 months, ZDV/3TC/EFV caused the greatest resolution of microalbuminuria (85.7% decline in ACR from baseline) compared with ZDV/3TC/NVP (72.5% decline) and TDF/FTC/EFV (63.9% decline). In multivariate analyses, predictors of ACR include older age (Odds ratio OR 2.8, p= 0.025); female gender (OR, 3.4, p =0.014); CD4+ (OR 0.99, p=0.002). Conclusions: HIV induces renal impairment. Thus, albuminuria, microalbuminuria and nonalbuminuric renal insufficiency are highly prevalent in antiretroviral-naïve HIV-infected persons but nephrotic-range albuminuria is uncommon. Albuminuria and/microalbuminuria and eGFR improve concurrently on HAART (with/without tenofovir DF). Zidovudine-based HAART (ZDV/3TC/NVP) resolves microalbuminuria earlier, and without relapse, unlike Tenovofir-based regimen and zidovudine with efavirenz (ZDV/3TC/EFV).