Efficacy and Safety of Two-drug Regimens for Treatment of HIV in the Central Nervous System

HIV can infect the central nervous system (CNS) soon after transmission, and early initiation of antiretroviral therapy (ART) reduces the establishment or persistence of viral reservoirs in the CNS compartment as well as HIV-associated CNS injury [1, 2]. Successful treatment and virus eradication will require approaches that are effective, safe, and durable in the CNS [3, 4]. Despite current ART regimens that suppress HIV RNA in blood to undetectable levels, people with HIV (PWH) differ substantially in HIV RNA concentrations in cerebrospinal fluid (CSF) [5–8]. In most PWH, HIV RNA concentrations in CSF are more than 1 log10 copies/mL lower in CSF than in blood. After initiation of ART, viral suppression in blood is the primary determinant of viral suppression in CSF [5, 7]. During therapy, a minority of PWH have detectable HIV RNA in CSF when HIV RNA in blood is undetectable, a condition termed CSF viral escape that can occur with or without symptoms [6, 9–12]. CSF viral escape is associated with ART drugs that do not achieve therapeutic concentrations in the CNS and/or have reduced efficacy as a consequence of drug resistance mutations, among other factors such as worse immune status [6, 8, 13–15]. Newer ART drugs with higher potency, high genetic barrier to resistance, and longer half-life, such as second-generation integrase strand transfer inhibitors (INSTIs), are more effective at suppressing HIV systemically and typically also do so in the CNS. These attributes have led to efforts to simplify ART regimens to reduce cost, improve medication adherence, and limit toxicities. However, concerns exist that simplification approaches could lead to HIV escape in the CNS or other tissues where drug distribution and efficacy may be suboptimal [3]. Moreover, if simplified regimens have suboptimal efficacy in the CNS, resurgent HIV replication and resulting immune responses could lead to greater risk for CNS complications of HIV disease, particularly cognitive and mood disorders [1, 10, 16, 17].

The study by Trunfio et al. [18] addresses these concerns by evaluating the efficacy of two-drug ART regimens in the CNS. Trunfio and colleagues analyzed data from a retrospective, cross-sectional study that compared effects of two-drug (n=19) and three-drug (n=78) ART regimens on CSF viral suppression, biomarkers of inflammation and neuronal injury, and neurocognitive performance in adults who had lumbar punctures performed between 2010–2019 [18]. All participants were on the same ART regimen for at least 30 months prior to lumbar puncture and were virally suppressed with plasma HIV RNA below 200 copies/ml for at least 6 months. Two-drug regimens were heterogeneous and included INSTI + protease inhibitor (PI), INSTI+NNRTI, PI+NNRTI, and PI+NRTI regimens. The authors found no statistically significant differences between the two-drug and three-drug regimens with respect to cognitive performance, CSF viral suppression, or biomarkers. These results are consistent with prior work indicating that ART regimens that suppress HIV in blood usually suppress HIV in CSF as well.

While valuable, these findings should be interpreted in the context of their limitations, including the retrospective cross-sectional design and small cohort with only 19 participants on two-drug regimens, some of which are not routinely used in the clinic. Generalizability of findings to the clinic population may be reduced, since lumbar punctures were performed for clinical indications (e.g., possible neurosyphilis). As such, participants may be more likely than the clinic population to have CNS inflammation and CSF viral escape. CSF viral escape was noted in about a fifth of participants irrespective of regimen type. This non-conservative bias may mean that two-drug regimens are even safer in PWH who do not have clinical indications for lumbar puncture. Other limitations include lack of drug resistance genotypes and measurement of HIV RNA only in cell-free CSF. Cell-associated HIV DNA and HIV RNA have been detected with higher prevalence than CSF cell-free HIV RNA among treated patients and the former has been associated with worse neurocognitive outcomes [16]. While Trunfio et al. found no difference in biomarkers of CNS injury and neuroinflammation, prospective studies of larger cohorts and long-term follow-up are needed to validate the findings and identify dual regimens with durable safety and efficacy in the CNS. Future studies should also include vertically-infected PWH, since neurocognitive outcomes are worse in those infected as adults.

Although earlier studies suggested that PWH receiving some older two-drug regimens were more likely to have detectable HIV RNA in CSF [3, 19], particularly those including a PI that can have sub-therapeutic concentrations in CSF (e.g. atazanavir) [8, 15, 20], existing trials support non-inferiority of two-drug regimens compared to three-drug regimens up to 48 or 96 weeks in treatment-experienced patients [21–25]. Dual regimens based on contemporary drugs consist of an INSTI (e.g., dolutegravir) and/or boosted PI (e.g., darunavir) with a reverse transcriptase inhibitor such as lamivudine or rilpivirine. While two studies comparing dual to three-drug regimens found no difference in neurocognitive effects [22, 26], there are few data on efficacy and safety of modern dual regimens in the CNS. Concentrations of most ART drugs are significantly lower in CSF than in blood, although drug concentrations in CSF may not accurately reflect those in brain [27]. Concerns about sub-therapeutic drug concentrations in the CNS are mitigated by the knowledge that regimens that suppress HIV RNA in blood typically suppress HIV RNA in CSF, and that reducing ART drug exposure could reduce the risk of neurotoxicity [3, 28]. Moving forward, identifying dual regimens with optimal efficacy and safety in different clinical settings will be important, particularly for PWH who struggle with medication adherence − those with psychiatric disease, substance abuse, or living with homelessness. Ongoing development and deployment of new longer-acting ART drugs [29], including long-acting oral drugs [30] and monthly injectables, may be useful in these populations. One study with informative results is the LATITUDE trial (NCT03635788), which compares long-acting injections of rilpivirine and cabotegravir vs. standard oral regimens in PWH with history of suboptimal adherence.

Overall, these reassuring findings reinforce the importance of future studies to evaluate the comparative efficacy and safety of simplified regimens, particularly long-acting injectables, for controlling HIV in the CNS and reducing the risk of ART neurotoxicity. Randomized controlled clinical trials to study simplified regimens for long-term safety and efficacy in the CNS and other compartments in naïve and experienced patients are awaited.