HIV TREATMENT SIMPLIFICATION: OUTCOMES OF SWITCHING HIV-1 PATIENTS TO PROTEASE-INHIBITOR MAINTENANCE MONOTHERAPY

The work in this thesis characterised the virological outcomes of HIV-1 patients on second-line ART switching to boosted-darunavir maintenance monotherapy in sub-Saharan Africa and explored their determinants, with the aim of providing evidence to inform practice and policy. Firstly, I took advantage of samples and data collected within a trial of maintenance monotherapy that was conducted in Yaoundé, Cameroon, between August 2014 and July 2015. The trial population was composed of HIV-1 positive adults who were receiving suppressive second-line ART with two NRTIs and a ritonavir-boosted protease inhibitor (PI/b). Patients were randomised to either a switch to maintenance monotherapy with ritonavir-boosted darunavir (DRV/r) for 48 weeks or to continue their current triple ART regimen. My first question was to investigate the virological outcomes and the relationship between viraemia and drug resistance. I used stored samples to investigate the presence of drug-resistance associated mutations (RAMs) in peripheral blood mononuclear cells (PBMC) collected at study entry (while patients were virologically suppressed on triple ART) and in follow-up plasma samples collected at the time of virological rebound on DRV/r monotherapy. I analysed the viral genomes by Sanger sequencing and ultra-deep sequencing (UDS) and used phylogenetics to assess their relatedness. The resistance analyses focused on reverse transcriptase and protease; in a subset of patients I also sequenced the gag gene to identify mutations in cleavage sites and other regions. I then used the resistance data alongside the available demographic, clinical and laboratory data to identify predictors of virological outcomes by statistical modelling. The results were interesting. I found that presence of RAMs at study entry affecting the NRTIs and the NNRTIs were predictive of a reduced (rather than increased) risk of virological rebound during follow-up on DRV/r monotherapy and the effect was independent of adherence levels. I also found that despite a high prevalence of viraemia during DRV/r monotherapy, there was no emergence of new protease resistance even by sensitive UDS, thus confirming data from previous studies conducted in Europe. While this observation was reassuring as it indicated no loss of treatment options in the monotherapy arm, I wondered whether there were other adverse consequences of frequent viraemia during DRV/r monotherapy. I thus investigated the kinetics of soluble CD27 (sCD27), a marker of immune activation and inflammation that a previous study from our group had linked to residual viraemia in subjects on suppressive ART. I tested prospective samples taken before and four weeks after the switch from triple ART to DRV/r monotherapy and observed that median sCD27 levels increased significantly after simplification, with the effect driven by a subset of the patients. I used various control populations to place the findings into context using different control populations. Hence, I propose that sCD27 may serve as an indicator of a risk of viraemia, a hypothesis that needs to be investigated prospectively. One additional research question was related to Hepatitis B Virus (HBV). My research hypothesis was that in a HBV hyperendemic setting like Cameroon, discontinuation of HBV active agents in people with HIV-1 could abolish a prophylactic effect against HBV acquisition and reactivation. To address this question, I measured markers of HBV infection prospectively in MANET trial participants. The results showed discontinuation of NRTIs was associated with a risk of both de-novo acquisition and reactivation of HBV. This represents another factor that makes DRV/r monotherapy undesirable for Cameroon and sub-Saharan Africa in general