Evaluation of the Safety and the Tolerability of a Combination of Two HIV Inducers in Patients With Undetectable Viral Load

Combination antiretroviral therapy (cART) is potent but not curative. cART requires lifelong adherence. Although multiple reservoirs may exist, the HIV-1 reservoirs containing stably-integrated, transcriptionally-silent but replication-competent proviruses, are recognized to predominate among infected CD4+ T cells. They are therefore a permanent source for virus reactivation and could be responsible for the rebound of plasma viremia observed after cART interruption. Persistence of truly latent (i.e. non-defective) HIV-1 proviruses represents a major obstacle to eradication, as suggested by the failure of cART intensification strategies at clearing the viral reservoirs. Indeed, the levels of HIV-1 reservoirs appear as one of the critical factors influencing the duration of a remission after cART cessation 4. Consequently, a decline of the HIV-1 latent reservoirs size to a level sufficient to permit an efficient control of the infection by the host immune system might allow interruptions in therapy (\"treatment-free windows\"). Reactivation of HIV gene expression in latently-infected cells together with an efficient or intensified cART could serve as an adjuvant therapy aimed at eliminating/decreasing the pool of latent viral reservoirs.

The chromatin organization and the epigenetic control of the HIV-1 promoter are key elements involved in transcriptional silencing. The repressive nucleosome nuc-1, located immediately downstream of the transcription start site, is maintained hypoacetylated by histone deacetylases (HDACs) in latent conditions. The use of HDAC inhibitors (HDACis) as latency reversing agents (LRAs) has been well characterized in several latency models and in ex vivo cART-treated HIV-1+ patient cell cultures. Several clinical studies and trials using HDACis were reported [VPA (by Margolis, Siliciano, Lambotte and Routy), SAHA (by Margolis and by Lewin), Panobinostat (by Rasmussen) and romidepsin (by Mellors (ClinicalTrials.gov NCT01933594) and Sogaard)]. The previous studies testing VPA did not show any benefit from this HDACi in reducing the number of latently-infected resting CD4+ T cells. Two clinical trials have demonstrated that administration to cART-treated patients of a single or multiple clinically tolerable dose(s) of SAHA was temporally associated with an increase expression of cell-associated unspliced (CA-US) HIV-1 RNA levels within resting CD4+ T cells in vivo. However, both trials could neither show an increase of residual viremia nor a decrease of the size of HIV-1 reservoirs. Another recent pilot clinical trial with the HDACi Panobinostat has shown a significant increase in CA-US RNA levels but also an increase of plasma HIV genomic RNA level analyzed by a transcription-mediated amplification (TMA) assay as well as a transient decrease in total HIV DNA level. Following cART interruption, this latter trial has shown in some patients a viral load rebound which was delayed compared to the rebound observed in most patients 2-3 weeks after stopping cART. Results from another recent pilot clinical trial including intravenous romidepsin injection, administrated once weekly for 3 weeks while maintaining cART, have shown that 5 out of 6 aviremic patients tested presented viral load levels quantifiable by standard commercial assays. Altogether, these studies are encouraging but question the efficiency of HDACi used alone to reduce the size of the HIV-1 reservoirs or to observe a delay in the viral rebound. The investigators propose some hypotheses in order to explain these partial results and some optimizations to improve therapeutic reactivation strategies:

1. The level of reactivation obtained in these clinical trials was too weak. Targeting simultaneously different mechanisms of latency should be more efficient when viral eradication is the objective since the combination of different classes of compounds could synergize (i.e. result in a higher reactivation level than the sum of the reactivations produced by each compound individually) to reactivate HIV expression in latently-infected cells.
2. A precise time schedule for LRA administration may be needed, suggesting that a sequential treatment could improve the effect of administration of two LRAs targeting different latency mechanisms.
3. HIV-1 latency is a heterogeneous phenomenon. Indeed, the investigators have previously observed in our reactivation studies performed ex vivo a high diversity among the patient cell cultures in terms of pattern of responses to the different LRAs tested. These findings suggest that HIV-1 transcriptional repression results from heterogeneous combinations of molecular mechanisms which vary from one patient to the other. The heterogeneity that the investigators have observed between patients emphasizes the need to evaluate the efficacy of an LRA first ex vivo in cell cultures from a given patient before the administration of this LRA to this given patient in vivo in the context of a clinical trial. Unfortunately, in the vast majority of the previous clinical trials aimed at reactivating HIV-1 from latency, a pre-selection based on ex vivo reactivation assays has not been performed.
4. Gene expression induction of latent HIV proviruses is stochastic. Indeed, Siliciano's group has shown that after one round of in vitro maximum T-cell activation, some proviruses remain silent but are potentially inducible after additional rounds of activation. In order to maximize the activity of a given LRA, it should thus be administered repeatedly at multiple time points.

Consequently, an optimization of the "shock" part, first part of the "shock and kill" strategy, would involve an individualized and combined treatment of LRAs, administered with a precise time schedule and repeatdly at multiple time points. In order to set up such a type of clinical trial, the investigators have investigated in vitro and ex vivo several types of LRA combination. One of them presented several advantages and was approved in human therapy.

Indeed, epigenetically, it is known that DNA methylation and histone deacetylation cooperate to establish and maintain a heterochromatin environment. Indeed, MethylBinding Domain (MBD) proteins bound to methylated DNA in cis-regulators regions can serve as bridges between DNA and chromatin-modifying factors (such as HDACs). In the case of HIV, the HIV-1 promoter has been previously shown to be hypermethylated ex vivo and resistant to reactivation in the latent reservoirs from aviremic HIV-1 infected individuals, as opposed to the hypomethylated 5\'LTR of integrated proviruses present in viremic patients. Interestingly, Trejbalova et al. have very recently reported that DNA methylation of the HIV-1 promoter increases progressively during cART treatment. Indeed, these authors have detected low levels of 5\' LTR DNA methylation in the resting CD4+ T cells of patients who were cART-treated for up to 3 years. However, after long-term cART, they have observed an accumulation of 5\' LTR DNA methylation in the latent reservoir. The DNA methylation status of the HIV-1 promoter could contribute to "lock" the silent state of the provirus in cooperation with histone repressive posttranslational modifications such as histone deacetylation, thereby making the return of the provirus to an active state more difficult. In this view, demethylating agents could represent promising candidate drugs in combination with HDACis for reducing the pool of latent HIV reservoirs of patients under cART during at least 3 years.

Moreover, it is known that the two epigenetic modifications, DNA methylation and histone deacetylation, are dynamically linked and result in the silencing of genes in cancers. Indeed, several preclinical studies support the view that pharmacologic targeting of both DNMT and HDAC may result in synergistic anticancer activity. Moreover, two clinical trials with the DNA methylation inhibitor Decitabine and the HDACi romidepsin are ongoing in cancer therapy (NCT00037817, NCT00114257). The DNA methylation inhibitor Decitabine approved in human therapy and HDACis present several advantages for HIV-1 purging strategies. Decitabine presents the advantage to cross the blood-brain barrier, to possess an excellent distribution in body fluids, to not induce global T-cell activation. HDACis do not induce global T-cell proliferation or activation, act in a broad spectrum of cell types and potently repress CXCR4 chemokine receptor expression and function. HDACis have also been identified as new P-TEFb-releasing agents. This is undoubtedly another crucial mechanism through which HDACis reactivate HIV expression. Indeed, the switch from promoter proximal pausing to productive elongation is mediated by the viral trans-activator Tat and P-TEFb, an essential cellular elongation transcription factor. In cells, P-TEFb exists in active and inactive forms. Its release from the inactive 7SK small nuclear ribonucleoprotein complex is a critical step for P-TEFb to activate transcription elongation. Thus, importantly, the use of the two drugs might impact other mechanisms (other than histone acetylation and DNA methylation) involved in HIV latency.

1.2 Research hypothesis Recently, our laboratory has evaluated the therapeutic potential of demethylating agents in combination with HDACis in reactivating HIV from latency in vitro and ex vivo. HIV-1 reactivation potential of LRAs was assessed first in vitro in two latently-infected T-cell lines (by quantification of HIV-1 transcription, expression and production) and next ex vivo in CD8+-depleted PBMCs and resting CD4+ T-cell cultures from 58 cART-treated aviremic HIV-1+ patients (by quantification of HIV-1 production). The investigators have shown that the DNA methylation inhibitor 5-aza-2\'deoxycytidine (Decitabine), but not 5-azacytidine, induced HIV expression in vitro. The investigators next demonstrated in vitro that a sequential treatment of Decitabine + HDAC is used at clinically tolerable doses synergistically induced HIV expression, highlighting for the first time the importance of a treatment schedule for LRAs combined treatments. Importantly, the investigators have shown the physiological relevance of this synergy and sequential aspect ex vivo. Interestingly, the sequential decitabine + romidepsin combinatory treatment tested induced HIV-1 recovery in a higher manner than the drugs alone in CD8+-depleted PBMCs cultures. In conclusion, the investigators reported for the first time that, in addition to the combinatory aspect, the sequential aspect of LRA administration might be critical for purging strategies aimed at decreasing the reservoir size.

The investigators hypothesize that such sequential combinatory reactivation approaches could lead to a decline in HIV-1 reservoir size to a level sufficient to allow control of the HIV infection by the host immune system and hopefully therapeutic interruptions. However, the first step in order to highlight the benefit of this type of intervention is to evaluate the feasibility, the safety and the tolerability of such sequential combinatory reactivation approaches.

Consequently, based on our ex vivo results and on our proposed optimizations of the "shock" part, first part of the "shock and kill" strategy, (involving an individualized and combined treatment of LRAs, administered with a precise time schedule and repeatedly at multiple time points), the investigators propose to set up a phase I trial in HIV-1 sub-type B positive patients under cART for at least 48 months and presenting, to the best knowledge of the clinical PI, no detectable viral blip for at least 12 months in order to evaluate the feasibility, safety and tolerability of a sequential treatment of Decitabine in combination with romidepsin delivered at multiple time points [one time (cohort 1), two times (cohort 2) or four times (cohort 3)]. As a secondary objective, the investigators will also evaluate the impact on viral reservoirs and HIV gene expression of the combinatory treatment (decitabine + romidepsin).