THE DYNAMICS OF LOCAL HIV-1 EPIDEMICS: The Colombian and Belgian cohorts

THE DYNAMICS OF LOCAL HIV-1 EPIDEMICS: The Colombian and Belgian Cohorts Approximately 35 million of people were living with HumanImmunodeficiency Virus (HIV) and 2 million were newly infected in 2012.Prevention measures and a higher coverage of antiretroviral therapy (ART) resulted in a decline of HIV incidence by more than 50% in 26 countries between 2001 and 2012. In addition, the observed impact of ART on community-level HIV incidence mainly on clinical trials has influenced clinical guidelines to start treatment early and to set up test-and-treat strategies or treatment as prevention (TasP) in some countries. However, it is a matter of debate if these policies would be cost-effectively in different settings. It is imperative that local governments become betterinformed with timely data about the characteristics of the local HIV epidemics, because they could influence the development of effective preventive policies. In this project we wanted to gain a deeper understandingof the dynamics of the HIV-1 epidemic in Belgium and Colombia, using viral sequences, socio-demographic and clinical data, complemented with statistical and phylogenetic approaches. These two countries share an increasing incidence and prevalence of HIV-1 infection in recent years, but differ in the HIV health care and availability of relevant data. Consequently, different research questions were triggered for each country. Unlike Belgium, there was limited information available about the molecular epidemiology of HIV-1 in Colombia and the spread of subtypes. In chapter 3, we focused on viral diversity and how the virus spread geographically and temporally in Colombia. Subtype B still predominated between 2002 and 2007 in the country. Additionally, phylogenetic analyses showed that multiple introductions occurred in the early 80s and suggested a link between the Colombian and Spanish epidemics. Bogotá, the capital, was the main exporter of the HIV epidemic, and other important cities within Colombia acted as sinks. These findings suggest an important role of tourism in the spread of HIV, but further studies should investigate the complete dynamics of this epidemic. Since Transmitted drug resistance (TDR) can impact the efficacy of first-line ART and its use in prevention strategies such as the prevention of mother-to-child transmission or TasP, we focused our efforts to retrieve any data on TDR in the Colombian epidemic and zoomed out our evaluation at regional level in chapter 4. A systematic review was performed using published literature and viral sequences available from Colombia and other Latin American and Caribbean countries. The overall TDR prevalence was stable and some countries had more than 5%. The reported prevalence of TDR was around 6% in Colombia, but there were no data about trends or related factors. These findings imply that surveillance of TDR should be frequently performed in countries such as Brazil or Venezuela or warranted insettings with scarce data such as Colombia, Central America and the Andean region. We investigated the evolution in TDR prevalencefor the different drug classes and its association with other socio- demographic factors at the AIDS Reference Center and AIDS Reference Laboratory in UZ Leuven as a first step in the study of the epidemic in Belgium. The TDR and viral diversity prevalence were stable in Leuven between 1998 and 2012, while a parabolic trend was observed for TDR against non-nucleoside reverse transcriptase inhibitors. We were not able to identify a specific population significantly associated with TDR or with spread of TDR, but men who have sex with men and originating from Belgium seemed to play an important role. The high number of chronically infected individuals in the cohort and the number of patients included in transmission clusters suggest that national collaborative studies should beprioritized to formulate policies that target earlier HIV diagnosis andprevention of transmission. As a methodological step, we evaluated the performance of automated subtyping tools with the aim to classify quickly the large amount of sequences used in different analyses in the Belgian and Colombian cohorts. Since our laboratory developed REGA version 3 (REGAv3), we compared our tool with the previous version REGA version 2 and six other tools. We found that most of the tools accurately identified subtype C and B, the most frequent clades worldwide. However, the performance was variable for other subtypes. COMET, jpHMM, REGAv3, and SCUEAL had good performance for pure subtypes using the pol region whereas COMET and REGAv3 also had good performance when analyzing most of the CRFs. Consequently, we recommended the use of two subtyping tools for surveillance and clinical purposes, and this was the strategy we also used to classify subtypes in other chapters.Finally, this thesis constitutes the initial step to characterize both epidemics. The complete understanding of these two local HIV epidemics will be useful for designing prevention strategies as well as for improving HIV health care programs.status: publishe