Evolution of Estimated Glomerular Filtration Rate in Human Immunodeficiency Virus and Hepatitis C Virus-coinfected Patients Receiving Sofosbuvir-based Direct-acting Antivirals and Antiretroviral Therapy

Chen-Hua Liu; Hsin-Yun Sun; Szu-Min Hsieh; Wen-Chun Liu; Wang-Hui Sheng; Chun-Jen Liu; Tung-Hung Su; Tai-Chung Tseng; Pei-Jer Chen; Chien-Ching Hung; Jia-Horng Kao


J Viral Hepat. 2021;28(6):887-896. 

In This Article


The risk of nephrotoxicity in HIV/HCV-coinfected patients receiving SOF-based DAAs and TFV-based ART has been controversial since the widespread use of both therapies. Most published studies had limited sample sizes and were retrospective in design which may affect the conclusions.[18–20,22] Soeiro et al. prospectively assessed the eGFR changes during SOF/LDV treatment in HIV/HCV-coinfected patients on TDF-based and TDF-free ART. Although the eGFR decline between baseline and end of treatment of SOF/LDV was numerically higher in patients on TDF-based ART (5.40 ml/min/1.73 m2 for non-boosted TDF regimen; 4.31 ml/min/1.73 m2 for non-boosted TDF regimen) than those on TDF-free ART (2.02 ml/min/1.73 m2 for non-TDF regimen), the differences did not reach statistical significance by paired t-test.[21] By applying the GEE model for repeated eGFR measurements at 10 different time points with only 0.7% missing data, our study indicated that patients on TFV-based ART had a slight but significant eGFR decline (adjusted slope difference: −0.82 ml/min/1.73 m2/month) compared to those on TFV-free ART during SOF-based DAAs. Based on this model, we expected a linear trend of eGFR differences of 2.46 ml/min/1.73 m2 during a 12-week course of SOF-based DAAs between patients on TFV-based ART and those on TFV-free ART, which was in line with Soeiro et. al's observation of 2.29–3.38 ml/min/1.73 m2 differences in eGFR decline between TDF and non-TDF groups.

Our participants with HCV genotype 1 infection were marginally associated with eGFR decline in univariate analysis, while this association was not significant in multivariate analysis. Although a recent large-scale cohort study observed the accelerated decline of renal function in patients with HCV genotype 1 infection, the causal relationship for this association was not clear, and therefore, more data are needed to confirm our findings.[28] During the 12-week course of off-treatment observation for SOF-based DAAs, no factors of interest were associated with significant eGFR declines. The different on- and off-treatment trends of eGFR between our participants on TFV-based and TFV-free ART receiving SOF-based DAAs implied the potential interplay of SOF and TFV on the nephrotoxicity. A recent cell-based model demonstrated that SOF strongly inhibited renally abundant carboxylesterase-2 (CES-2). Because TFV is a substrate of CES-2, the inhibition of CES-2 by SOF may increase the blood concentration of TFV and potentiate TFV-related nephrotoxicity by increasing the serum levels of hydrolysed metabolites, which are evident by the elevated urine retinol binding protein-4 (RBP-4) and β2 microglobulin (β2M) levels in proximal renal tubules, as well as by the plasma pharmacokinetic profiles.[29–31] On the contrary, the trends of eGFR tended to improve in participants on TFV-free ART after the initiation of SOF-based DAAs, indicating that the risk of nephrotoxicity by combination of TFV-free ART and SOF-based DAAs was low, and the renal damage was mitigated by potent HCV suppression.[32]

Because higher plasma levels of TFV may increase the risk of nephrotoxicity, TAF, the novel prodrug of TFV, results in higher intracellular concentration of active TFV compared to TDF, which allows a much lower dose for TAF to reach the therapeutic effects as well as to lessen the renal damage.[33] A recent pharmacokinetic study revealed that the plasma levels of TFV in terms of the area under concentration–time curve, the maximal observed concentration and the at-the-end dosing interval were lower in HIV/HCV-coinfected patients on TAF-based ART than those on TDF-based ART during SOF-based DAA therapy.[31] The favourable pharmacokinetic profiles of TAF may contribute to the observed beneficial effect of eGFR evolution in our participants on TAF-based ART over those on TDF-based ART.[34] Because COBI may block the renal tubular secretion of creatinine which raises serum creatinine levels, the use of novel TAF-based COBI-free ARTs, such as RPV/FTC/TAF or BIC/FTC/TAF, may further improve the renal safety in HIV/HCV-coinfected patients during SOF-based DAAs.[35] After discontinuing SOF-based DAAs, we also found that no factors of interest affected the 12-week interval of the eGFR evolution probably due to the lack of SOF-induced adverse effects on TFV plasma concentration.

In addition to evaluating the eGFR evolution by the GEE model, we further analysed the eGFR changes between the time points of baseline and on-treatment week 12, and between baseline and SVR12. We found that more participants on TFV-free ART had eGFR grade improvement than those on TFV-based ART during SOF-based DAA therapy. Although the grade of eGFR changes differed during SOF-based DAA treatment, there were no differences with grade changes between the time point of baseline and SVR12 in our participants on TFV-based or TFV-free ART. The dynamic changes during and after SOF-based DAA therapy further confirmed the adverse effect of TFV on eGFR under SOF treatment, which resolved after discontinuation of SOF.

Our study had several limitations. First, because our participants did not randomly receive TFV-based or TFV-free ART, the potential confounding factors, such as concomitant use of other nephrotoxic agents, as well as the disease control status for HTN or DM which might affect eGFR, might not be well balanced between groups. However, our participants were young with low reported rates for HTN and DM. Furthermore, most of them did not take agents commonly known to be nephrotoxic.[36] Second, because all participants had CKD stages 1 and 2, our findings cannot be extrapolated to HIV/HCV-coinfected patients with advanced CKD stages, in which the pharmacokinetics of TFV and SOF are significantly altered. Third, the other surrogate markers, such as urine albumin–creatinine ratio (UACR), cystatin C level, and β2M level, were not performed to confirm the risk of nephrotoxicity of TFV and SOF in our study.[37,38]

In summary, our prospective study confirmed HIV/HCV-coinfected patients on TFV-based ART had a slight eGFR decline compared to patients on TFV-free ART during SOF-based DAA therapy. The trends of eGFR declines were also observed during DAA therapy in participants on TDF-based ART and those on TAF-based ART. Periodic monitoring of the renal function and choosing TAF-based ART are needed for HIV/HCV-coinfected patients who are on TFV-based and SOF-based antiviral therapies in order to early detect and lessen renal dysfunction.[39] Because the increased risk of nephrotoxicity of TFV and SOF coadministration in HIV/HCV-coinfected patients with good renal reserve is limited, the physicians should not restrict the use of SOF-based DAA therapy in this special clinical setting on the basis of WHO targets for HCV microelimination.[40]