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


A total of 231 patients receiving TDF were evaluated. Overall, the mean age was 46.7 years, the mean duration of HIV infection was around 16 years, the cumulative time on therapy was > 10 years, and nearly one-third were coinfected with hepatitis C virus (HCV). The median time on TDF was 63.6 months (1470.5 patient-years), and 151 (65%) patients received this drug with a nonnucleoside reverse transcriptase inhibitor (NNRTI; mostly in a single-tablet regimen with efavirenz, 92 patients; nevirapine, 35 patients; etravirine, 24 patients), whereas 79 received a boosted protease inhibitor (bPI; atazanavir and darunavir, 32 patients each; lopinavir, nine patients; fosamprenavir, six patients). No patient was receiving DTG, cobicistat or RPV when they were enrolled in the study.

After evaluation, 109 patients (47%) switched from TDF and 122 (53%) continued the same regimen. As shown in Table 1, patients who switched were older, were more heavily pretreated, had a higher rate of HCV coinfection, and had a lower eGFR at inclusion and at discontinuation than patients continuing TDF. Additionally, time on TDF was significantly longer in this group of patients who switched from TDF (80 vs. 50.9 months in those continuing TDF; P < 0.01). At the time of the switch, 14 patients (13%) had progressed to CKD, and the prevalence of tubular abnormalities or PRTD was significantly higher compared with patients who did not switch their regimen (mean number of tubular abnormalities, 1.49 vs. 0.91, respectively; P < 0.01). Of the patients who switched, 49 (45%) were receiving a bPI, mainly atazanavir (23 of 32), lopinavir (six of nine) or fosamprenavir (six of six).

Of the 109 patients who switched, 60 (55%) maintained triple therapy, changing TDF to abacavir, and 49 (45%) changed to an NRTI-sparing, dual regimen. As might have been expected, patients who changed to a dual therapy had a longer history of HIV infection, were more heavily pretreated and had mainly been receiving a PI-based combination (76%). Time on TDF was significantly longer in this group than in those who switched to abacavir, and there was a nonsignificant trend for more severe tubular renal dysfunction (proteinuria > 100 mg/g, 51% vs. 40%, respectively; glycosuria, 21% vs. 16%, respectively), although the prevalence of PRTD was similar, and eGFR was even better in this group, suggesting selection of patients (Table 2). Abacavir was used with efavirenz in 14 patients, DTG in 13, etravirine and nevirapine in 10 patients, darunavir boosted with ritonavir (DRV/r) in six patients, RPV in four patients, and raltegravir in three patients. The dual therapy chosen was lamivudine plus DRV/r in 20 patients, DTG plus RPV in eight patients, DRV/r plus raltegravir in eight patients, DRV/r plus ETR in seven patients, and DRV/r plus DTG in six patients. No patient received cobicistat. Thus, 28% and 29% of patients, respectively, received a drug known to alter tubular secretion of creatinine.

Renal Outcomes

Patients were re-evaluated after a median of 8.86 months [interquartile range (IQR) 2.6–14.1 months], although the evaluation was performed earlier in case of switching TDF (4.8 vs 13.3 months; P < 0.01). Overall, switching from TDF was associated with a significant improvement in eGFR (median change +0.3 vs. −2.91 mL/min/1.73 m2 in those who did not switch; P = 0.04), and the improvement was more significant for patients who were switched to abacavir than for those switched to dual therapy, compared with patients who continued a TDF regimen (Figure 1a). After excluding patients receiving DTG or RPV, this improvement was more significant compared with no TDF change (+2.96 vs. −2.91, respectively; P < 0.01). There was a small difference in the rate of improvement according to the use of abacavir or dual therapy (median change in eGFR +1.52 vs. −0.5, respectively; P = 0.12), which was less marked if those on DTG or RPV were excluded (+3.64 vs. +1.01, respectively; P = 0.35). As shown in Figure 1b, those patients receiving DTG or RPV had a median decrease of −6.17 (IQR −8.1, −1.7) mL/min/1.73 m2 (81% showed a decrease) at 2.73 months after switching the regimen, compared with +2.96 mL/min/1.73 m2 in patients who switched to other drugs (61% showed improvement) and a decrease of −2.91 mL/min/1.73 m2 in patients continuing a TDF-including regimen (60% showed a decrease). Thus, one patient receiving RPV met the CKD criteria (3%), but overall the CKD rate declined from 14 (13%) to eight cases (7%) in patients who switched from TDF.

Figure 1.

Changes in estimated glomerular filtration rate (eGFR). (a) Differences in median eGFR (mL/min/1.73 m2) after switching from tenofovir disoproxil fumarate (TDF) to abacavir (ABC), switching to nucleoside reverse transcriptase inhibitor (NRTI)-sparing dual therapy, or continuation of the same TDF-containing regimen. (b) Change in median eGFR (mL/min/1.73 m2) according to the use of dolutegravir (DTG) and/or rilpivirine (RPV) in the new regimen, in comparison with TDF continuation. CI, confidence interval.

A progressive decline in eGFR was associated with a longer time on TDF (Spearman rho = −0.16; P = 0.02) and higher B2M level (ρ = −0.17; P = 0.08), whereas an improvement in eGFR was linearly associated with improvements in phosphaturia (ρ = 0.19; P = 0.01), uricosuria (ρ = 0.5; P = 0.05), RBP (ρ = 0.28; P < 0.01) and B2M (ρ = 0.18; P = 0.05). In a linear regression analysis, change in eGFR was associated with age (β = −0.14; P = 0.04), time on TDF (β = −0.19; P = 0.04), and eGFR decline from therapy initiation (β = −0.42; P < 0.01), after adjusting for HTA, HCV coinfection, nadir of CD4 count and antiretroviral strategy. These results were unchanged after excluding those patients receiving DTG or RPV.

Tubular Renal Outcome

Switching from TDF was associated with an improvement in the rate of various tubular abnormalities, which was significantly different from that observed in patients who continued TDF. There was an improvement in phosphataemia (median change +0.3 vs. −0.1 mg/dL in those who did not switch from TDF; P < 0.01), phosphaturia (−2.5% vs. +1.26%, respectively; P = 0.02), glycosuria (−7% vs. +3%, respectively; P = 0.02) and urinary B2M (−0.38 vs. −0.03 log μg/g, respectively; P < 0.01), with a trend for an improvement in urinary RBP (−0.05 vs. −0.01 log μg/g, respectively; P = 0.17) and proteinuria (−19.2 vs. −6.7 mg/g, respectively; P = 0.13), and no change in albuminuria (P = 0.95). Thus, the number of tubular abnormalities decreased (median change −0.35 vs. 0 in those who did not switch from TDF; P < 0.01), and the rate of tubular dysfunction decreased from 44 cases before the switch to 26 cases after it (40% to 24%, respectively; P < 0.01). However, there was an increase of 26% in the rate of tubular disease in patients continuing TDF (+30%; P < 0.01).

There were no differences in tubular renal outcome according to the strategy of switch. PRTD fell from 24 to 12 cases in those switching to abacavir (−50%), compared with 20–14 cases (−29%) for the dual regimen. However, the mean decrease in tubular abnormalities was similar in patients receiving abacavir and those receiving dual therapy (−0.35 vs. −0.36, respectively; P = 0.87), as were median improvements in phosphataemia (+0.3 mg/dL in both; +10.1% vs. +9.4%, respectively; P = 0.94), phosphaturia (−3.4% vs. −2.4%, respectively; P = 0.78), uricosuria (−0.05% vs. −0.2%, respectively; P = 0.9), urinary B2M (−0.32 vs. −0.6 log μg/g; −9% vs. −15%, respectively; P = 0.42), RBP (0.04 vs. 0.08 log μg/g; +1.7% vs. +3.4%, respectively; P = 0.95); uPCR (−17.5 vs. −23 mg/g; −22.5% vs. −19.1%, respectively; P = 0.44) and uACR (−0.15 vs. −0.2; −3.25% vs. −1.22%, respectively; P = 0.73). This improvement in the mean number of tubular abnormalities or the change in different tubular parameters was independent of the use of DTG or RPV in both groups. Thus, there was a significant improvement in phosphataemia and phosphaturia in comparison with continuing TDF (P < 0.01 and P = 0.02, respectively; Figure 2), and significant reductions in proteinuria, albuminuria and B2M (P < 0.05), with a nonsignificant reduction in RBP (P = 0.57; Figure 3).

Figure 2.

Percentage of change [median and 95% confidence interval (CI)] in phosphataemia (dark boxes) and phosphaturia [fractional excretion (FE) of phosphate; grey boxes] after tenofovir disoproxil fumarate (TDF) switch, according to the use of dolutegravir (DTG) and/or rilpivirine (RPV). P-value < 0.05 for the comparison between change and TDF continuation. ABC, abacavir.

Figure 3.

Percentage of change [median and 95% confidence interval (CI)] in total proteinuria [urinary protein:creatinine ratio (uPCR); dark boxes], albuminuria [urinary albumin:creatinine ratio (uACR); grey boxes], the beta-2-microglobulin to creatinine ratio (boxes with horizontal lines) and the retinol-binding protein to creatinine ratio (boxes with skewed lines). The P-value was < 0.05 for the comparison between switching groups and TDF continuation, except for retinol-binding protein. DTG, dolutegravir; RPV, rilpivirine.