Switching to Abacavir Versus Use of a Nucleoside-Sparing Dual Regimen for HIV-Infected Patients With Tenofovir-Associated Renal Toxicity

JL Casado; C Santiuste; MJ Vivancos; M Monsalvo; A Moreno; MJ Perez-Elías; JM del Rey; S Moreno


HIV Medicine. 2018;19(8):541-550. 

In This Article


This was a prospective, nonrandomized cohort study that began in July 2012. We investigated the rates of tubular and renal toxicity observed in patients using TDF and their outcomes after discontinuation of TDF. After entry into the study, an increasing number of patients discontinued TDF because of tubular dysfunction and/or progressive eGFR decline. Here, we present data on the evolution of tubular and renal toxicity in patients who changed to a dual, NRTI-sparing regimen compared with patients who switched from TDF to abacavir, and control patients who continued on the same TDF regimen. The decision to interrupt TDF treatment and the choice of the alternative regimen were made by the responsible physician, on a case-by-case basis, considering previous resistance and toxicity.

Inclusion criteria were age > 18 years, receiving TDF at inclusion, and at least one serum and urine determination during follow-up, after switching from or continuation of TDF. Pregnant patients, those receiving other nephrotoxic drugs, and those being treated with drugs known to alter TDF tubular secretion (diclofenac and sildenafil) were excluded.

The present study was approved by our institutional review board (EC 218/12). Written informed consent was obtained from all patients before enrolment. The study was registered at clinicaltrials.gov (NCT02209740).


Demographic data [age, sex, race and body mass index (BMI)], HIV-related data (risk practice for HIV acquisition, time of infection, previous cART and reason for discontinuation, and time on cART), and data on hepatitis C virus (HCV) coinfection status, hypertension (HTA) and diabetes mellitus (DM) diagnosis were collected at inclusion. Routine laboratory test results, including serum creatinine and phosphate concentrations, CD4 cell counts and HIV-1 RNA levels [Versant HIV-1 RNA 1.0 (kPCR); Siemens Diagnostics, Erlangen, Germany; quantification limit 37 copies/mL], were available since therapy initiation.

As part of the study, measurements of fasting serum creatinine, glucose and phosphate, and urinary protein, albumin, creatinine, uric acid, phosphate, glucose and tubular proteinuria [beta-2-microglobulin (B2M) and retinol-binding protein (RBP)] were performed at inclusion and during follow-up. eGFR was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-epi) equation.[12]

Urine total protein was measured using the benzethonium chloride method and urine albumin was measured using an immunoturbimetric assay. They were expressed as the ratio of total protein or albumin to creatinine in urine in mg/g [urinary protein:creatinine ratio (uPCR) and urinary albumin:creatinine ratio (uACR), respectively]. The urine creatinine concentration was determined using an enzymatic method. Phosphaturia was determined as the fractional excretion (FE) of phosphate according to the equation [(urine phosphate × serum creatinine)/(urine creatinine × plasma phosphate)] × 100, and expressed as tubular reabsorption of phosphate (TRP; 100-FE). Uricosuria was measured as the FE of uric acid: [(urine uric acid × serum creatinine)/(urine creatinine × plasma uric acid)] × 100, and also expressed as tubular reabsorption of uric acid (TRU; 100-FE). Urinary concentrations of B2M and RBP were measured by immunonephelometry (BN II system; Siemens Healthcare Diagnostics, Eschborn, Germany). The concentrations of both proteins were standardized to a urinary creatinine concentration of 1 g/L (μg/g), and values were log transformed.


Hypertension was defined as a systolic and diastolic blood pressure ≥ 140 mmHg and ≥ 90 mmHg, respectively, or the use of antihypertensive agents. DM was defined as a diagnosis of DM before inclusion or use of antidiabetic agents. Proteinuria was defined as uPCR > 100 mg/g (11.3 mg/mmol), microalbuminuria as uACR 30–300 mg/g (approximately 3–30 mg/mmol), and macroalbuminuria as uACR > 300 mg/g (> 30 mg/mmol). Hyperphosphaturia of tubular origin was defined as an FE of phosphate > 20% or > 10% (TRP < 80% or 90%) in the presence of hypophosphataemia. Confirmed hypophosphataemia was defined as serum phosphate ≤ 2.5 mg/dL. Hyperuricosuria was defined as an FE of uric acid ≥ 10%.[13] Glycosuria was defined as a value > 100 mg/dL (5.56 mmol/L), which was the threshold in our laboratory for nonpregnant patients, in the presence of normal glycaemia (plasma glucose < 125 mg/dL).

Although there are no validated criteria, proximal renal tubular dysfunction (PRTD) was defined as the presence of at least two tubular abnormalities that are known to be part of Fanconi's syndrome: increased phosphaturia, nondiabetic glycosuria, hyperuricosuria and/or proteinuria. Because hypophosphataemia could be related to phosphaturia, serum phosphate was not included. Renal function change was defined as a sustained eGFR modification in mL/min/1.73 m2 compared with the value at inclusion. CKD was established if eGFR values were < 60 mL/min/1.73 m2 at two determinations (CKD stage 3 or higher) according to the National Kidney Foundation guidelines for CKD.[14]

Statistical Analysis

Characteristics of patients according to TDF switch, or the use of abacavir or dual therapy, were compared for statistical significance using the Mann-Whitney U-test for continuous variables and the χ 2 test for categorical variables. Analysis of paired observations during follow-up was performed using the Wilcoxon rank t-test. Correlations between tubular abnormalities were assessed using the Spearman correlation coefficient. A linear regression model was created using eGFR slopes during follow-up as the dependent variable, considered as a numerical value (mL/min/1.73 m2), and adjusted for potential confounders (age and eGFR at baseline), using the backward stepwise method. Entry and elimination criteria were set at a value of P < 0.10. All statistical analyses were performed using spss v. 18 (Chicago, IL, USA). All P-values were two-sided and a value of P < 0.05 was considered statistically significant.