Association Between Long term Tenofovir-Based therapy and the Incidence of Renal Dysfunction in Adult HIV-Positive Patients at Ronald Ross Hospital: A Retrospective Cohort Study

Abstract

Tenofovir disoproxil fumerate (TDF)-based antiretroviral therapy places HIV-infected patients at high risk of renal dysfunction. Therefore, we evaluated if the incidence of renal dysfunction in HIV positive adults patients on long term tenofovir-based regimen versus those on non-tenofovir use. We performed a retrospective cohort analysis of 834 HIV positive patients at the counselling and testing center (CTC) at Ronald Ross General Hospital. Patients were evaluated at baseline and with every follow up visit for creatinine and creatinine clearance (Cockcroft-Gault formula) calculated. Patients’ records in data management software called SMARTCARE from 2008 to 2014 were reviewed to compare renal function between patients on TDF-containing regimen (447 patients) with non-TDF containing regimen (387 patients). We evaluated Glomerular Filtration Rate (GFR) using creatinine clearance as defined by the Kidney Disease Outcomes Quality Initiative Classification (K/DOQI) by GFR.The effect of creatinine, urea and exposure to TB medication, Cotrimoxazole use and CD4 cell count on GFR were also follow up for 18 months. Univariate and Multivariate logistic regression was used to determine the factors associated with renal dysfunction. We report multivariable hazard ratios (Cox modeling) and binary outcomes with predictors retained if P < 0.05 and analyzed survival time. Renal function was categorized as CrCl <50 ml/min and ≥ 50 ml/min. Potential predictor variables for renal dysfunction included in the multivariable models were age, sex, weight, creatinine and treatment with tenofovir. The CD4 cell count, urea, anti-TB medication, Cotrimoxazole use, education level, employment status and creatinine clearance were similar at baseline between the two groups. There was no significant difference from exposure to TDF between groups on renal dysfunction development (OR: 2.52, 95% CI, 0.79-8.0, P-value=0.118) as a predictor of renal dysfunction in both univariate and multivariate analysis. Creatinine and creatinine clearance between groups at baseline and 18 months were not statistically different. The prevalence of mild and severe renal dysfunction among HIV-positive adults on TDF-based therapy was 18.4% and 0.2% respectively at 18 months during therapy. Those patients on TDF who were older than 50 years and presented with CrCl <50 ml/min and a CD4 cell count below 500 cells/uL at baseline were more likely to develop renal dysfunction. Twenty-four (5.5%) and 2 (0.5%) patients presented with mild and severe renal dysfunction respectively at baseline. 5 | P a g e Further, Univariate and multivariate analysis showed that ages <49 years than >50 (AOR: 6.35 (2.18-18.4), p-value=0.001), and higher CD4+ cell count >500 cells/uL (AOR: 1.24 (0.97-1.07), P-value=0.051) were less likely associated with renal dysfunction. Age per yearincrease (OR: 6.35, 95% CI -2.18-18.4, P<0.001) and CD4 cell count per 1 cell-increase/μl (OR: 3.03, 95% CI, 0.98-9.28, P=0.053) on univariate were important factors affecting renal dysfunction. We found no evidence to suggest that TDF-based therapy is associated with decline in renal function and that it does not affect creatinine clearance. However, there is need for close renal monitoring of patients initiated on TDF by using creatinine clearance.