Renal Safety of Tenofovir Alafenamide vs. Tenofovir Disoproxil Fumarate

A Pooled Analysis of 26 Clinical Trials

Samir K. Gupta; Frank A. Post; José R. Arribas; Joseph J. Eron Jr; David A. Wohl; Amanda E. Clarke; Paul E. Sax; Hans-Jürgen Stellbrink; Stefan Esser; Anton L. Pozniak; Daniel Podzamczer; Laura Waters; Chloe Orkin; Jürgen K. Rockstroh; Tatiana Mudrikova; Eugenia Negredo; Richard A. Elion; Susan Guo; Lijie Zhong; Christoph Carter; Hal Martin; Diana Brainard; Devi SenGupta; Moupali Das


AIDS. 2019;33(9):1455-1465. 

In This Article


Previous studies have demonstrated more favourable renal biomarker profiles in TAF-containing regimens compared with TDF-containing regimens; however, the sample sizes of individual trials and the overall low rate of clinically significant renal adverse events in these trials limited the ability to detect differences in the rates of these events with the exception of the pooled pivotal EVG trials. In the present analysis, we integrated data from 26 individual trials and were able to demonstrate the renal safety of TAF over TDF across a broad range of PLH, including those who were treatment naive and those who were virologically suppressed at switch. After 12 519 person-years of exposure to TAF, there were no cases of PRT or Fanconi syndrome (identified objectively and independently by the primary investigator caring for the participant) and significantly fewer discontinuations due to renal adverse events in the TAF group compared with the TDF group. Notably, only three (0.02%) renal discontinuation events were reported in participants on TAF; none of these were reported as study drug-related by the investigators, and all had plausible alternative causes.

In treatment-naive participants, we observed fewer overall renal adverse events in participants taking TAF-containing regimens compared with those taking TDF-containing regimens. No difference in overall renal adverse events was observed in participants enrolled in switch studies; this may be explained by the fact that participants in those studies were already maintained on TDF at the time of enrolment, and thus self-selected as less likely to develop renal adverse events.

By using an integrated analysis, we were able to demonstrate favourable changes in renal biomarkers in participants taking TAF-containing regiments compared with those taking TDF, both in treatment-naive and treatment-experienced patients who switched to TAF-containing regimens. Our findings demonstrate favourable changes in CrCl as well as in proximal tubule function (RBP and β2M ratios). We also observed a lower incidence of treatment-emergent proteinuria in participants taking TAF-containing regimens. The observed incidences of proteinuria were high, but notably these are cumulative incidences over 96 weeks of follow-up, and are consistent with previously reported incidences of proteinuria in PLH.[44] These biomarker findings in combination with the clinical outcomes suggest that TAF does not induce proximal tubule dysfunction.

The mechanism for the improved renal safety profile of TAF is likely related to the approximately 90% lower plasma levels of TFV seen in participants receiving TAF compared with those receiving TDF. This mechanism is supported by the reported association between declines in renal tubular function and higher TFV plasma concentrations.[45–47]

Conversely, the use of boosting agents such as RTV and COBI increase TFV exposure, and accordingly the use of boosting agents has been associated with an increased risk of renal adverse events.[2,48] A recent meta-analysis sought to compare the renal safety profiles of TDF-containing regimens in the presence and absence of boosting agents, and suggested that unboosted TDF could have a similar renal safety profile as TAF.[48] However, the aforementioned meta-analysis is limited by a relatively small number of participants and short duration of follow-up. In the findings presented here, nine out of 10 PRT cases occurred in participants receiving boosted regimens; however, one severe case of PRT occurred in a participant receiving TDF without a boosting agent. Our data support the principle that boosting agents increase the risk of TFV-associated renal adverse events; however, our ability to make robust conclusions about the renal safety of unboosted TDF is limited by the comparatively small number of participants taking such regimens (of 9322 total participants, 2962 were on TDF, and of those 1101 were on TDF without a boosting agent). Although the question of renal safety of TDF in unboosted regimens warrants more evaluation, the available data indicate that TAF can be safely used with boosted as well as unboosted third agents with a very low incidence of clinically significant renal events.

We note several limitations to our analyses. It is challenging to diagnose PRT, and no commonly accepted single diagnostic exists in the clinic to confirm PRT. As such, we utilized investigator-reported events to document PRT, which may have underestimated the number of PRT cases. A reporting bias is possible given the use of investigator reported events, but is unlikely to have affected our findings as most of the included trials were double-blinded, and the majority of reported renal discontinuation events and PRT cases were reported during blinded trial phases. Our clinical trial participants may have been healthier than the general population of PLH due to the presence of inclusion and exclusion criteria in the trials, although TAF was found to safe in patients with impaired renal function (CrCl 30–70 ml/min, many of whom with diabetes mellitus, hypertension, and proteinuria), with no reported cases of PRT and overall stable renal function through 96 weeks of follow-up.[49] We also acknowledge that we did not have individual level data on the duration of prior TDF therapy in our trials and therefore could not adjust the rates accordingly.

Despite these limitations, the integrated analysis presented here is based on the large cumulative exposure in person-years to TAF, both in antiretroviral naïve and virally suppressed populations. Furthermore, the pooled data used for analysis includes a demographically diverse population with a wide age range, a large number of women, and diverse ethnic background. It is also notable that a proportion of participants had relatively low CrCl, with variable CrCl eligibility cut-offs of 30, 50, or 70 ml/min in the trials included in this analysis (Appendix Table 1, The clinical trial data are supported by experience from the postapproval use in PLH in which currently there has been no renal safety signal with 1.1 million cumulative person-years exposure to TAF.

In conclusion, the pooled data from 26 clinical studies, representing over 12 500 patient-years of follow-up in children and adults on TAF, suggests that the favourable renal biomarker profile observed with TAF vs. TDF in the individual trials translates into a lower rate of clinically significant renal events. These data support a comparative renal safety advantage of TAF over TDF in a broad range of PLH.