Assessing antiretroviral therapy interruptions in HIV cure research
Richard Jefferys, TAG
Last month, researchers from the Laboratory of Immunoregulation at the National Institute of Allergy and Infectious Diseases (NIAID)  published a paper in PLoS Pathogens addressing the use of antiretroviral therapy (ART) interruptions in HIV cure research .
The joint lead authors were Katherine E. Clarridge and Jana Blazkova, and the focus of the study was on the effects of an analytical treatment interruption (ATI) performed during a clinical trial of the broadly neutralizing antibody VRC01. The main results offer reassurance that the ATI had no long-term negative effects for participants – as articulated in an accompanying NIAID press release titled NIH Study Supports Use of Short-Term HIV Treatment Interruption in Clinical Trials – but there are some possible safety concerns that the data does not address. 
Results of the VRC01 trial were published in 2016 in the New England Journal of Medicine. 
The protocol design involved an initial VRC01 infusion upon entering the study, followed by an ATI three days later. VRC01 infusions were subsequently administered every four weeks until week 24. Criteria for restarting ART included:
- A confirmed >30% decline in baseline CD4 cell count or an absolute CD4 cell count <350 cells/mm3
- A sustained (≥4 weeks) HIV RNA level of >5,000 copies/mL
- Any HIV-related symptoms
Ten participants were enrolled. VRC01 only led to a slight delay in viral load rebound, so ART was restarted in all cases with the duration of the ATI ranging from 22 to 115 days (median 57 days). Nine of the individuals restarted ART due to meeting the criteria for a sustained viral load increase to over 5,000 copies/mL, while the remaining participant experienced a confirmed >30% decline in baseline CD4 cell count. No participant had a CD4 T cell drop to less than 350 cells/mm3.
The researchers evaluated multiple HIV reservoir measures as well as biomarkers of immune activation and inflammation, comparing results obtained prior to the ATI with those observed after viral load was re-suppressed by ART. In almost all cases, there were no significant differences pre- and post-ATI. The lone exception was the chemokine RANTES, a potential inflammatory biomarker, which remained significantly elevated after ART resumption (at least at the time point measured, which was a median of 363 days after restarting).
Several parameters were evaluated during the ATI, revealing that there were increases in measures of the HIV reservoir and immune activation that subsequently declined to baseline levels after ART was resumed. However, data on inflammatory biomarkers during the ATI are not reported, making it unclear to what extent inflammation may have been temporarily elevated – it seems likely that it was, given the transient spike in markers of immune activation. The relationship between short-term inflammation and risk of clinical events is uncertain but potentially of concern, so this is an aspect of ATIs that deserves additional study. Among the steps taken to maximise safety in the trial, the exclusion criteria included evidence of heart disease, which might render an individual particularly susceptible to inflammation-related risks.
Another question that has been raised regarded the safety of ATIs – particularly by longtime treatment activist Jules Levin of the National AIDS Treatment Advocacy Project (NATAP.org) – is whether HIV levels in the central nervous system (CNS) might increase and have the potential to cause harm. CNS samples were not taken in this NIAID study.
Overall, the trial results are consistent with the idea that short-term ATIs can be performed safely in the context of HIV cure research, but they do not necessarily represent the last word on the topic. There was a robust dialogue regarding the use of ATIs at the recent Regulation of Clinical Research Related to HIV Cure meeting that took place in Bethesda on January 25, and a webcast of the session should soon be available on the Forum for Collaborative Research website (www.forumresearch.org).  TAG has also received support from the Elizabeth Taylor AIDS Foundation to survey community-based treatment activists regarding their views on ATIs, and a report will be published in the fall of 2018.
The Laboratory of Immunoregulation at NIAID has recently published a different study involving an ATI that highlights another important consideration in this type of HIV cure research: the inclusion of a placebo arm. 
The researchers conducted a clinical trial of a combination therapeutic HIV vaccine approach (DNA primes followed by a vesicular stomatitis virus vector boost) administered to individuals who initiated ART soon after infection. The vaccines were found to induce HIV-specific T cell responses, but this did not lead to superior control of viral load rebound during an ATI compared to placebo immunisation.
The researchers noted that, perhaps surprisingly, there were four cases of post-ATI control of viral load to low levels that occurred in the placebo arm of the trial. The finding emphasises that a subset of early-treated individuals can exhibit prolonged viral load suppression after an ATI, and that this needs to be considered when designing trials of therapeutic interventions. In the absence of a placebo arm, spontaneous control of HIV replication could be misinterpreted as evidence of a therapeutic effect. The authors specifically cite the open-label trial of therapeutic vaccination plus romidepsin that was presented by Beatriz Mothe at CROI 2017, pointing out that a similar proportion of participants in the placebo arm of their study maintained viral loads below 2000 copies/ml for at least 16 weeks.
The inclusion of placebo control arms in trials is not necessarily as simple as it might sound, because there are cost and logistical issues that are prohibitive for some research groups. However, it is clearly a necessary step for establishing that a candidate intervention has had a genuine effect, and results from open-label trials should be interpreted with caution until confirmed in a randomised controlled context.
Jefferys R. Assessing Antiretroviral Therapy Interruptions in HIV Cure Research. (22 February 2018).
- US NIH Laboratory of Immunoregulation.
- Clarridge KE. Effect of analytical treatment interruption and reinitiation of antiretroviral therapy on HIV reservoirs and immunologic parameters in infected individuals. PLoS Pathogens (2018). doi.org/10.1371/journal.ppat.1006792.
- NIH press release. NIH study supports use of short-term HIV treatment interruption in clinical trials. (11 January 2018).
- Effect of HIV antibody VRC01 on viral rebound after treatment interruption. N Engl J Med 2016; 375:2037-2050. DOI: 10.1056/NEJMoa1608243.
- Sneller MC et al. A randomized controlled safety/efficacy trial of therapeutic vaccination in HIV-infected individuals who initiated antiretroviral therapy early in infection. Science Translational Medicine (2017), 9(419):eaan8848. doi: 10.1126/scitranslmed.aan8848.
- Jefferys R. Capsules from CROI 2017. TAG basic science blog. (10 March 2017)