1. Selecting the ‘who’ and ‘where’ components of HIV preventive interventions in rural South Africa using molecular and spatial epidemiological approaches (Tanser, de Oliveira, Bärnighausen, Imrie, Newell)
Research Question
In the 438km2 Africa Centre surveillance area in rural KwaZulu-Natal, considerable spatial heterogeneities in HIV prevalence are apparent with up to four-fold differences in levels of infection between adjacent communities. Using epidemiological and demographic data at an individual level, viral genetic information on a cohort of seroconverters and geographical information at an individual, household and community level, we aim to answer the following question: What is the most appropriate target group and the most effective combination of interventions to reduce the rate of new HIV infections given the epidemiological, socio-demographic and phylogeographical characteristics of HIV transmission networks within this population?
We seek to use genotypic information of HIV isolates taken from 940 recent seroconverters in a longitudinal population-based HIV surveillance in combination with geographical information system (GIS) technology and socio-demographic and behavioural data to understand the complex patterns of HIV-1 transmission dynamics in a widely disseminated rural epidemic in Northern KwaZulu-Natal. In the context of a generalized epidemic (where prevalence is >50% and incidence 8% per year in women in their upper twenties) we aim to assess the role of viral, individual and community characteristics to inform targeted intervention approaches. Using sequence information from seroconverters, we will estimate the timing and origin of transmission networks in the research area. To understand patterns of inter- and intra-community transmission patterns, we will identify transmission foci and clusters of new infections using a spatial statistical methodology.
Fit of the research question within the AC portfolio/strategic plan
This work fits within the population-based research of the Centre, and the HIV dynamics and identification of target groups objectives. Our proposed approach, which uses molecular and spatial epidemiology approaches, will inform strategies as to whom to target, where and when, for HIV preventive interventions, takes advantage of the unique setting of the Africa Centre. The work takes forward our recent results which show substantial geographical heterogeneity in HIV prevalence across the relatively small surveillance area, with identified clusters of higher than expected and lower than expected HIV prevalence.
Data sources and methods
We will produce HIV genomic data (genotypes) for all seroconverters in the surveillance area (N=940) identified by the Africa Centre between 2004 and 2009. We will sequence the POL gene (a 1,200 base pairs fragment that contains the complete protease and first 240 codons of the reverse involves the alignment of our population-sampled sequences (N=940) to a public available subtype C reference dataset (n»2400), which is constructed from all HIV-1 subtype C sequences at the Genbank (http://www.ncbi.nlm.nig.gov). The second step involves the application of phylodynamics analysis to determine the date of origin and growth rate of pockets of high-transmission in our local epidemic. In the third analysis step, spatial information will be used to reveal the location of infections and to infer the route of transmission and the rate of geographic spread. In the last step, we will use the Africa Centre’s demographic, socio-economic, behavioural and geographical data to develop profiles of “high transmitters” and communities with high numbers of transmission events.
Expected findings
We will define the molecular epidemiological characteristics of HIV transmission networks in the study population. More specifically, we expect to estimate the geographical and temporal origin of the local transmitted strains, and to determine the (strain-specific) number of secondary infections caused by each primary infected individual and the impact of “high-transmitters” in a sexual network in the overall transmission of HIV in the research area. The previous mentioned information will be used to guide interventions that aim to reduce the level of HIV-1 transmission in the study population.
Policy implications
We will demonstrate how the generated molecular and geographical knowledge of our local epidemic can provide essential guidance to deliver a target intervention strategy, which in turn can be evaluated using longitudinal surveillance platforms. To our knowledge, this is the first proposal of this type internationally, and therefore its demonstration value to the field should be substantial.
2. HIV Drug Resistance Surveillance (de Oliveira, Viljoen, Danaviah, Manasa)
We will use dried blood spot (DBS) samples isolated from recently infected individuals identified in the Africa Centre (AC) Demographic Surveillance to determine the incidence rate of drug resistance mutations at a population level. We will follow guidelines from the World Health Organization (WHO) related to sampling and analysis procedures. In total we will sequence 60 samples from young (< 25 years old), drug naïve, recently infected individuals sampled in 2009. The sequences (genotypes) will be produced using a commercial methodology (VIroSeq - Abbott). The genotypes will also be used to validate the development of a cheaper and in-house genotypic method in the Africa Centre laboratory.
Research Question
We hypothesize that due to the recent implementation of the Hlabisa ART Treatment and Care Programme (since 2004), HIV-1 drug resistance transmission is still low in our setting. However the high number of individuals on treatment (around 10,000), widespread use of short-term PMTCT (up to 80%) and high percentage of viraemic patients (23%) at 12 months post ART initiation, make it possible that resistance mutations reservoirs exist and are transmitted locally. A second question to answer is related to the applicability of WHO protocols for surveillance of transmitted resistance in our setting.
Fit of the research question within the AC portfolio/strategic plan
This project sits within the population-based research of the Centre, and starts routine surveillance for transmitted drug resistance in the Africa Centre, which is a survey that is also of relevance to the Centre’s clinical research portfolio.
Data sources and methods
We will use 60 DBS samples collected from recent individuals in the Africa Centre Surveillance 2009 round. Viral load, CD4 count and resistance genotyping will be performed on the dried blood spot samples. Genotyping and clinical assays (VL + CD4) will be linked via a unique identification number to individual clinics, and to treatment, adherence, socio-demographic and programmatic information stored in the Africa Centre databases.
The resulting sequence data will be analysed for known drug resistance mutations using a local copy of the RegaDB database, an open source, clinical and research HIV database http://www.rega.kuleuven.be/cev/regadb/. RegaDB includes three of the most used HIV drug resistance algorithms (REGA, ANRS and HIVDB-Stanford), which are used to identify resistance mutations and provide resistance interpretation for ARVs. Other existing databases and programs, such as the Stanford HIV Drug Resistance database and the Calibrated Population Resistance Tool, will be used to estimate transmitted HIV-1 drug resistance mutations by the application of a specific International AIDS Society (IAS) list for surveillance of drug resistance mutation (SDRM).
Expected findings
The work started in December 2009. We expect to find a relatively low level (< 5%) of HIV-1 drug resistance mutations in our cohort. However, there is a clear and increasing tendency for drug resistance strains to infect naïve individuals as treatment expands fast in the developing world and it would not be a surprise if the prevalence of transmitted resistance mutations increased above 5% in the near future. In this regard, we will be repeating this surveillance in 2010 and following years, as suggested by the WHO.
Policy implications
This surveillance of transmitted drug resistance study is necessary at the AC research area given that since the introduction of ARV treatment in 2004, no survey of resistance mutations has been done. In addition, this information is important to evaluate the impact of future ARV treatment guidelines (treatment with CD4 below 350 and TasP trial) on the transmission of resistance.
