A study of factors associated with HIV superinfection was recently published from a prospective study of HIV positive gay men.
While superinfection has been associated with increased disease progression, specific risk factors for superinfection have not been defined. It is unclear, for example, whether superinfection results from continued high-risk sexual behavior or incomplete immune responses during early infection.
This study, from Jouni Vesa and colleagues at the University of California, San Diego prospectively followed a high-risk cohort of 96 HIV positive MSM to investigate demographic, behavioral, clinical and immunogenic determinants of superinfection. 
Superinfection is defined as infection with a second HIV virus after the immune response to an initial HIV infection has already developed. Superinfection most often occurs during primary infection when humoral and cell mediated immune responses have not fully matured, enabling re-infection. By contrast, co-infection with two different viruses can also occur either simultaneously or before an immune response has been mounted to the initial infection. 
Men were eligible for the study if they were newly diagnosed with recent HIV and had at least one longitudinal follow up visit. Viral loads and CD4 counts were measured at each follow up visit, and evidence of superinfection was determined using ultra-deep sequencing of partial coding regions in HIV gag, pol and env. Participants with evidence of dual infection at enrollment were excluded. Participants were also followed while antiretroviral-naive and were censored upon initiating ART. In order to assess if immunogenic determinants influence susceptibility to superinfection, HLA types were assessed by genotyping HLA-A, HLA-B, HLA-C and HLA-DRB1 alleles. As the number of individuals with superinfection was limited, two-digit codes for HLA lineage groups were used in the analysis.
Overall, men with documented HIV mono-infection were monitored for a median of 15.6 months (range 0.7 to 73.9 months). The median time from the estimated date of infection to presentation was 70 days (range: 10 to 170 days). During follow up, ten people became superinfected, each with subtype B, at a median time from initial infection of 10.4 months (range: 3.5 to 21.1 months).
There were no associated between risk of superinfection and age, race, sex, higher viral load, lower CD4 count and high-risk sexual behavior. In contrast, a higher number of sexual partners was associated with acquisition of superinfection at alpha = 0.10 (hazard ratio 4.74, 95%CI: 0.87 to 25.97; p = 0.073).
Specific HLA alleles have previously been associated with increased HIV disease progression, whereas others have been shown to have a protective effective. In order to determine if HLA-alleles influence susceptibility to superinfection, at a time during which immune responses to the initial infecting strain have not yet fully matured, the researchers determined the hazard ratios for each HLA allele.
Increased superinfection risk at alpha = 0.10 was observed for: HLA-A*29 (HR 4.10, 95%CI: 0.88 to 14.76, p = 0.069); HLA-B*35 (HR 4.64, 95%CI: 1.33 to 18.17, p = 0.017); HLA-C*04 (HR 5.30, 95%CI: 1.51 to 20.77, p = 0.010); HLA-C*16 (HR 4.05, 95% CI: 0.87 to 14.62, p = 0.071); HLA-DRB1*07 (HR 3.29, 95% CI: 0.94 to 12.90, p = 0.062); HLA-DRB1*08 (HR 15.37, 95% CI: 2.11 to 79.80, p = 0.011). Decreased superinfection risk was observed for HLA-DRB1*11 (HR 0.13, 95%CI: 0.00 to 1.03, p = 0.054).
In multivariate analysis, assessing significant HLA alleles and number of sexual partners, the following alleles retained significance regardless of the number of sexual partners: HLA-B*35 (p = 0,020); HLA-C*04 (p = 0.033); and HLA-DRB1*08. In contrast, the increased risk associated with HLA-A*29, HLA-C*16 and HLA-DRB1*07, and the decreased risk associated with HLA-DRB1*11, were not retained in the multivariate model with number of sexual partners.
HIV is known to downregulate expression of HLA alleles. It is possible that primary HIV infection may downregulate expression of protective HLA alleles, thereby reducing the efficiency of antigen presentation to cytotoxic T cells and in turn increasing the risk of superinfection. In contrast, mechanisms that upregulate protective alleles may reduce risk of superinfection.
The authors concluded that further analysis is needed in larger cohorts to fully characterise the significance of HLA alleles associated with superinfection.
By Gareth Hardy
- Vesa J et al. Increased HIV-1 superinfection risk in carriers of specific human leukocyte antigen alleles. AIDS. 2017. 31:1149-1158.
- Smith DM et al. HIV superinfection. J Infect Dis. 2005. 192. 438-444.