Molecular Epidemiological Investigations of HIV Transmission Patterns Among People Who Inject Drugs in Kenya

Abstract

In Kenya people who inject drugs (PWID) have a 4-fold greater prevalence of HIV and at least 10-fold greater prevalence of hepatitis C compared to the general population. While preliminary epidemiological evidence suggests needle-sharing, and associated parenteral HIV transmission, is decreasing, more evidence is needed to evaluate these conclusions and to understand factors contributing to sexual HIV acquisition and transmission in this population. Phylogenetic and molecular epidemiology approaches can reveal transmission trends within and between populations to help determine the most impactful interventions; however, such studies must also be mindful of the vulnerability of populations, like PWID, to stigma or other group harms. To characterize the HIV epidemic among PWID in Kenya, we first used cluster analysis and ancestral state reconstruction to estimate the following relative frequencies of HIV transmission: transmission within the PWID population; between the PWID population and other key populations (female sex workers and men who have sex with men); between the PWID population and the general population; and between PWID from the coastal region and Nairobi. Second, we investigated whether PWID shared an HIV or hepatitis C transmission network with the sexual or injecting partners they identified through assisted partner services, based on genetic distances of the virus sequences. We found substantial mixing between HIV-1 sequences from PWID with those from other populations and no excess similarity between the HIV-1 sequences from pairs of individuals identified as injecting partners. These results support prior evidence of the effectiveness of needle-syringe programs and the shift towards sexual acquisition and transmission as an important factor in this epidemic for PWID. We propose a renewed emphasis on addressing risk factors for sexual transmission and on understanding the environment within which sexual HIV acquisition and transmission occurs for PWID. Further research on hepatitis C incidence and transmission risk factors could strengthen conclusions about changes in injecting behaviors. Finally, we explored ethical issues in molecular epidemiological research of pathogen transmission trends, noting a lack of productive change coming from the current discourse centered around privacy risks to research subjects. We recommend that by treating pathogen sequence data as primarily a community resource, we can both increase the responsiveness of researchers to community concerns and improve community acceptance of molecular epidemiology as a public health and research tool.