Sexually transmitted infections as risk factors for HIV infection among MSMs: systematic review.

Abstract

To the Editor: We published a 1999 review of studies exploring whether sexually transmitted infections (STIs) are causally associated with human immunodeficiency virus (HIV) infection among men who have sex with men (MSMs),1 finding few studies and no convincing evidence for associations. This contrasted with evidence then emerging of several STIs (especially ulcerative infections) being risk factors for HIV among the general population. We speculated that our findings might be explained by higher exposure to HIV among MSM and relatively higher transmission of HIV in anal than vaginal sex. A more recent review/meta-analysis by Freeman et al. focusing on herpes simplex virus (HSV)-22 reported a significant association of HSV-2 with HIV among MSMs [RR = 1.7; 95% confidence interval (CI) 1.2–2.4] but smaller than among the general population. This difference is plausible for the reasons above but does suggest STI control may nonetheless be important in MSM HIV prevention efforts. Another recent review/meta-analysis3 suggests other STIs, including syphilis, gonorrhoea, and chlamydia, may be risk factors for HIV among the general population but reports no findings for MSMs. We therefore decided to update our review to assess whether other STIs not merely HSV-2 might be important to address within HIV prevention for MSMs. We searched PubMed for relevant articles published 1998 to 2007 using MeSH/non-MeSH search terms covering STIs, HIV, and MSM; abstracts from the most recent International Acquired Immune Deficiency Syndrome conference and International Society for Sexually Transmitted Disease Research conference; and reference lists of found articles. We identified 10,681 potentially relevant articles. Titles and abstracts were reviewed for possible pertinence and where this was the case, full articles obtained (46 in all) and reviewed against criteria of reporting association(s) between HIV and one or more STIs, or between exposure/allocation to an STI control programme and HIV; assessing incident HIV infections serologically/via medical case notes; assessing prior STI infections serologically/via medical case notes; and adjusting, matching, or otherwise controlling for age and measure(s) of sexual risk behavior. Four adequate studies were identified. Two reported on data from the same cohort,4,5 so only the more complete of these was included. Three reported on HSV-25–7 and one on HSV-1.6 None examined other STIs. None reported the effects of STI control. One study5 did not appear in Freeman et al.'s review and this reported a significant association between HIV and HSV-2 detected >24 months previously (hazard ratio = 1.5, 95% CI 1.1–2.1) but not HSV-2 detected <24 months previously (hazard ratio = 1.7, 95% CI 0.8–3.3), i.e., in line with Freeman et al.'s meta-analysis. A further study reviewed by Freeman et al. regarding its findings for HSV-2 also included data on HSV-16 finding no association with HIV. Thus, current evidence suggests HSV-2 is a risk factor for HIV infection among MSMs but does not tell us whether STIs other than HSV-2 exert effects on HIV that are negligible; smaller than those found for the general population but not negligible; or possibly even comparable to the general population. Our review also identified an absence of research on STI control programmes effects on HIV among MSMs. This is surprising given the high prevalence of many STIs among MSMs worldwide. To assess the importance of addressing STIs other than HSV-2 within HIV prevention for MSMs we recommend further research on this question. Observational studies could be nested within studies of behavioural interventions or repeat clinic attenders. Alternatively, given the variability in contents, targeting, and intensity of STI control for MSMs as well as the uncertainty regarding impact on HIV incidence, it would be ethical and useful for experimental evaluations of pilot-enhanced STI control programmes targeting MSM to assess effects on HIV incidence.