HIV Transmission Networks
HIV Transmission Networks
The chances of an HIV-negative individual acquiring HIV from having unprotected sexual intercourse with an HIV-positive individual ranges from 0.1% to 5% per act, depending on the act, the HIV RNA level of the infected person, the concomitant presence of other sexually transmitted diseases, and the extent of exposure to blood. But HIV is not spread at random. That is, new infections tend to cluster within “networks.” And within a particular network, 20% of the “members” may be responsible for as many as 80% of the new infections (the so-called 20/80 rule).1 These individuals often are referred to as “super-spreaders,” and they continue to play a major role in various epidemics, including tuberculosis and HIV infection.
Consequently, there has been tremendous interest at a public health level to better understand the dynamics of transmission networks, and to try to design interventions that are targeted at interrupting or reducing the transmission of various pathogens within these networks.
At CROI 2014, a number of sessions were devoted to these topics, including “How Transmission Networks Drive Epidemics.” At that session, for instance, Alexandra Ostra, from the CDC noted that 82% of all “dyads” (pairs of linked infections) live in the same state. In terms of race, blacks were linked most closely to blacks, but all other racial/ethnic groups were linked most closely to whites. Not surprisingly, men who have sex with men (MSM) were linked most strongly to other MSM. Specifically, 82% of dyads that contained at least 1 MSM contained another MSM. On the other hand, 33% of dyads that contained at least 1 injection drug user (IDU), 54% of dyads that contained at least 1 heterosexual male, and 44% of dyads that contained at least 1 heterosexual female also contained an MSM. The conclusion was that reducing transmission in the MSM community was likely to have the biggest impact on decreasing HIV spread in the United States.
Daniella Bezemer of the Netherlands made the point that “networks don’t stop.” In fact, 60% of the networks studied by her and her colleagues were present before 1996. Interestingly, 50% of the 4288 HIV sequences collected from MSM came from just 91 different networks. Networks are “rejuvenated” by an influx of young MSM. Furthermore, the widespread availability of antiretroviral therapy has not had an impact on reducing the spread of HIV in these networks—a sobering reality that calls into question the role of “treatment as prevention.”
Susan Little from University of California San Diego found that in the first year after presentation, plasma viral load (HIV RNA level), number of unique sexual partners, and high “transmission network score” (TNS) were the strongest predictors of transmission risk. Interestingly, high HIV RNA levels were a much stronger predictor of risk of transmission than stage of disease (early versus late). On the other hand, Dr Little noted that if individuals in these networks could be identified within the first 12 weeks following seroconversion, they were much less likely to spread the virus to others, compared with identifying network members more than 12 weeks after seroconversion. Dr Little speculated that if a TNS could be included with a laboratory’s antiretroviral resistance report sent to health care providers, it might help clinicians intervene with individuals at high risk of spreading the virus to others.
Both Teiichiro Shiino of Japan and Kim Tien Ng of Malaysia mentioned that the HIV epidemics in their countries are being driven primarily by MSM, although Dr Shiino noted that Japan also has large “local” networks of heterosexuals contributing to the spread of HIV. Nevertheless, he stated that “hyper-active male spreaders” are the major source of concern at the moment in Japan.
Finally, Gonzalo Yebra of Uganda reported on 3 distinct cohorts, with essentially no overlap, in Uganda. The 3 cohorts were (1) a female sex worker cohort; (2) a “fisher folk” cohort, and (3) a rural clinic cohort. Of note, specimens from the rural clinic cohort were much more likely to cluster than were specimens from the other two cohorts, suggesting that this very old rural southwestern Ugandan cohort may have been the major source of the Ugandan HIV epidemic. In the US, rates of HIV are particularly high in the rural southeastern US, which may be related to lack of access, or more difficult access, to clinical care facilities.
If we are doing to reduce the incidence of HIV in the US . . .
Taken as a whole, it seems clear that the data presented point to the urgency of intervening early in the MSM networks if we are going to have a chance to reduce the incidence of HIV in the United States. Exactly how to do so, and what strategies will likely be the most effective, are critically important areas for future research.
1. Woolhouse ME, Dye C, Etard JF, et al. Heterogeneities in the transmission of infectious agents: implications for the design of control programs. Proc Natl Acad Sci USA. 1997:94: 338–342.
2. Oster AM, Wertheim JO, Hernandez AL, et al. HIV transmission in the United States: the roles of risk group, race/ethnicity, and geography. Abstract 213. CROI 2014. March 5, 2014, Boston.
3. Bezemer D, Ratmann O, van Sighem A, et al. Ongoing HIV-1 subtype B transmission networks in the Netherlands. Abstract 205. CROI 2014. March 5, 2014, Boston.
4. Little SJ, Pond SLK, Anderson CM, et al. Using HIV networks to inform real time prevention interventions. Abstract 206. CROI 2014. March 5, 2014, Boston.
5. Shiino T, Sadamasu K, Hattori J, et al. Large MSM group and local heterosexual transmission are major concerns in the HIV epidemic in Japan. Abstract 215. CROI 2014, March 5, 2014, Boston.
6. Ng KT, Ng KY, Khong WX, et al. Transmission networks of HIV-1 subtype B, CRF01_AE and CRF51_01B among MSM in Singapore. Abstract 207. CROI 2014, March 5, 2014, Boston.
7. Yebra G, Ragonnet-Cronin M, Ssemwanga D, et al. HIV-1 phylodynamics and phylogeography among high-risk and general populations in Uganda. Abstract 208. CROI 2014. March 5, 2014, Boston.