Salvage Antiretroviral Therapy: Time for “DeNUKElearization”?

(See the Major Article by Gandhi et al., on pages 1407–15.)

Nucleoside reverse transcriptase inhibitors (NRTIs) are important components of antiretroviral therapy (ART) when combined with other medications [1, 2]. The association of NRTIs with short- and long-term side effects such as nausea, diarrhea, and lipodystrophy and the emergence of virus resistance have led to the search for NRTI-sparing regimens [3, 4]. Current human immunodeficiency virus (HIV) treatment guidelines include several NRTI-sparing regimens that are recommended as initial therapy in special circumstances or NRTI-free regimens that could be considered as switch regimens in special circumstances [1, 2]. Although NRTI-sparing regimens have proven effective in ART-naive people with HIV type 1 (HIV-1) infection (PWH) [5] and as switch regimens in ART-experienced PWH who are virally suppressed [3], their role for salvage therapy in people with virologic failure on ART is less clear. For HIV salvage therapy among persons with virologic failure, treatment guidelines recommend starting at least 2, and preferably 3, new active antiretrovirals (ARVs) [1, 2]. When starting a new regimen in ART-experienced PWH, the standard of care still often includes NRTIs, even when ARV-experienced patients have HIV isolates with resistance mutations predicted to compromise NRTI activity. This practice follows observations that some NRTIs may retain moderate efficacy despite the presence of genotypic resistance mutations [6, 7]. Recycling NRTIs is also a common practice in many HIV salvage treatment trials to date [8, 9].

In the current issue of The Journal of Infectious Diseases, Gandhi and colleagues report the 96-week results of the AIDS Clinical Trials Group (ACTG) A5241 (OPTIONS) trial [10]. The OPTIONS trial enrolled PWH with triple-class ARV experience or viral resistance and plasma HIV RNA levels ≥1000 copies/mL while taking a protease inhibitor (PI) regimen who were then randomized take a new ART regimen that omitted or added NRTIs [11]. The primary outcome of the original OPTIONS trial demonstrated that the cumulative probability of regimen failure at 48 weeks was not significantly different between participants randomized to an optimized regimen with or without NRTIs (29.8% in the “omit NRTI” arm vs 25.9% in the “add NRTI” arm) [11]. In the current report of 96-week follow-up of the OPTIONS trial [10], investigators found that additional virologic failure was uncommon after week 48, with only 14% of virologic failure occurring after week 48. At week 96, 70% of those randomized to receive the optimized ART regimen without NRTIs and 65% of those randomized to add NRTIs to their regimen had HIV-1 RNA <200 copies/mL. These findings demonstrated that NRTI-free salvage therapy in the setting of virologic failure was safe and associated with durable virologic suppression so long as the new regimen contained >2 active drugs. Also, treatment-emergent resistance was uncommon (ranging from 3.4% to 24%) and explained only 3.4%, 11%, and 18%–24% of virologic failures by ARV class: PI (darunavir), INSTI (raltegravir), and NNRTI (most frequently etravirine), respectively, indicating that once virologic suppression was achieved, it was typically sustained.

The OPTIONS study also gives further insights into predictors of virologic failure across study arms. Compared to older participants (aged 47–69 years), younger participants (aged 16–46 years) were significantly more likely (adjusted odds ratio, 4.4) to experience virologic failure in each of the randomized groups [10]. Lower rates of ART initiation have been demonstrated among PWH aged 18–24 years compared with persons >25 years [12], and younger individuals have frequently demonstrated lower rates of engagement throughout the care continuum, from preexposure prophylaxis linkage and adherence [13–15] to HIV testing [16] and linkage to care and viral suppression [17, 18]. It remains unclear, however, whether a high frequency of virologic failure among participants in the lowest age quartile was driving this observation in the OPTIONS trial, or if there are other factors that contributed to the higher rate of failure in the younger population that remain unexplained.

In the discussion of NRTI-associated toxicities provided by Gandhi and colleagues, it is worth noting that 95% of those randomized to add NRTIs received tenofovir disoproxil fumarate (TDF) [10]. Thus, the toxicity results presented in this study are relatively specific for TDF and may not apply to the more widely used and newer formulation, tenofovir alafenamide (TAF), or to other NRTIs. Long-term toxicities of NRTIs include lipodystrophy, decreases in bone mineral density, and nephrotoxicity [19–21] and vary widely among NRTIs. Even between the different tenofovir formulations, worsening of the lipid profile is more frequently observed in individuals receiving TAF, whereas TDF has been shown to lower lipids [19, 20, 22]. In contrast, reductions of bone mineral density [23] and nephrotoxicity (occurring in ~2.5% [24]) are some of the most frequent adverse outcomes observed in persons receiving TDF, with improvements in hip and spine bone mineral density and urine protein or albumin to creatinine ratio observed in individuals who switch to TAF [25]. In this study by Gandhi et al, participants randomized to omit NRTIs had a greater increase from baseline to total, non–high-density lipoprotein, and low-density lipoprotein cholesterol compared to those randomized to add NRTIs, potentially driven by TDF-associated decreases in lipid levels in the vast majority of the NRTI study participants who added TDF to their regimen [10]. Consistent with previously observed data for TDF, there was a greater decline in creatinine clearance observed in those randomized to add NRTIs whereas bone mineral density was not investigated.

Even after achieving viral suppression, mortality remains higher in PWH than in the general population [26], driven mostly by age-related noninfectious comorbidities [27], a fact well recognized by the investigators of OPTIONS when designing their study [10]. Virally suppressed persons with HIV on ART have persistently elevated levels of systemic inflammation, immune dysfunction, and hypercoagulation, which is driven in part by microbial translocation and immune activation [28, 29]. The presence of increased levels of inflammatory markers has been linked to an increased risk of age-related noninfectious comorbidities including cardiovascular disease, non-AIDS malignancies, diabetes and hypertension, neurocognitive disorders, and renal, liver, and bone disease, as well as functional impairments, frailty, and mortality [27–33]. Of note, higher cumulative TDF-based ART exposure has been shown to be associated with increased biomarkers of endothelial, monocyte, and lymphocyte activation and inflammation, potentially explained by the known immunomodulatory effects of tenofovir in vitro and in vivo [34]. These data are worthy of further evaluation in light of the trend toward fewer deaths in the group that did not add a NRTI compared to the group that added a NRTI (ie, predominantly TDF, with 1 vs 10 deaths, respectively) [10]. In contrast, the proportion of individuals with moderate-to-high cardiovascular risk (ie, Framingham risk score) increased over time to a greater extent in the group randomized to omit NRTIs compared with those randomized to add NRTIs [10, 35]. This finding, however, should be interpreted with caution, as it could have been driven by worsening of lipid profiles in those randomized to omit NRTIs compared to improved lipid levels in persons who added NRTIs (ie, a TDF-specific outcome) [35].

There are still important gaps of knowledge not addressed by the OPTIONS study. One would be to evaluate potential effects of NRTI-sparing or NRTI-free regimens on anatomically privileged reservoirs such as the central nervous system (CNS) [36]. A potential concern with the use of NRTI-free ART regimens is CNS viral escape [37–39]. However, there are currently insufficient data related to CSF concentrations of ARV and corresponding CSF HIV RNA levels in persons receiving NRTI-free regimens [40] to determine if NRTI-free (compared to NRTI-containing) regimens deliver clinically comparable concentrations of ARV to the CNS or other anatomically privileged reservoirs.

In conclusion, the study by Gandhi and colleagues showed that HIV-1 salvage therapy can safely omit NRTIs without compromising efficacy or durability of response so long as the new regimen has a cumulative activity of >2 active drugs. These results will likely inform HIV treatment guidelines. The study also showed that younger people require careful monitoring given their greater likelihood of virologic failure. Finally, this study showed higher cardiovascular risk and worsening lipid profiles in persons not adding NRTIs to their ART regimens over 96 weeks, a finding that may be specific for TDF. Future studies should evaluate the potential impact of NRTI regimens on mitochondrial function and persistent inflammation, as well as outcomes beyond 96 weeks, to assess the contribution of ART regimens on long-term non-AIDS outcomes. These studies evaluating longer-term cardiovascular and other non-AIDS outcomes may also be needed to answer the question of whether there is a potential cardioprotective effect of sparing NRTIs.

Notes

Financial support. This work was supported by the National Institutes of Health (grant numbers AI106039, MH113477, and AI036214).

Potential conflicts of interest. M. H. and S. J. L. have received research grants awarded to their institution from Gilead Sciences.

Both authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.