Epidemiological studies

There was one characteristic which the Mwanza, Rakai and other trials of STI treatment all shared - none of them treated herpes. But, as the Ghys study had already shown,¹ this STI (the most prevalent STI of all and by far the most common cause of genital ulcers) might be a better target for the use of STI treatment to prevent HIV.

Having HSV-2 (which is most commonly associated with genital herpes), and particularly being infected recently, appears to greatly increase the risk of also acquiring HIV. These are ‘co-morbid’ conditions.

A 2002 meta-analysis² of nine studies that documented HSV-2 infection before HIV acquisition found that having herpes doubled the risk of subsequently catching HIV. For studies where the temporal sequence was less clear – in other words where the herpes could have been caught alongside or after catching HIV – herpes infection quadrupled the likelihood of having HIV infection. The researchers calculated that in populations with 80% herpes prevalence – typical of HIV-positive gay men – nearly half of all cases of HIV could be ‘attributed’ to herpes. In other words, if herpes did not exist in this population, HIV incidence would be halved.

A 2002 study followed HIV-negative gay men in San Francisco, Chicago, Denver, New York, Seattle and Boston.³ The prevalence of HSV-2 in the community was reported as 30 to 50% in HIV-negative gay men. HIV-negative men were recruited into a case-control study and followed every six months for 18 months. Blood testing for HSV infection was performed at regular intervals to determine which men already had the genital herpes virus, and which acquired it during the period of follow-up. Similarly, the men’s HIV status was monitored.

During the study period, 116 men acquired HIV infection. These men were compared to a control group of 342 men who remained HIV-negative. Overall, having prior genital-herpes infection increased the risk of acquiring HIV by 80%.

A carefully constructed study in Pune, India, added some new information to this.⁴ It found that recent, and especially primary, HSV-2 infection increased the risk of subsequently catching HIV much more than the secondary outbreaks or reactivations of chronic HSV-2 infection. Nonetheless, it also found an increased risk of HIV infection even when herpes infection appeared dormant or was causing no symptoms.

The study involved 2732 HIV-negative patients attending a sexual health clinic between 1993 and 2000. At baseline, 1175 patients (43%) tested positive for HSV-2 antibodies. During the course of the study, a further 217 patients seroconverted for HSV-2. A total of 224 patients were detected with recent HIV infection, of whom 28 also had recent HSV-2 infections, and in the majority of instances the two infections were diagnosed at the same visit to the clinic.

The investigators found that exposure to more recent HSV-2 infections substantially increased the risk of also being infected with HIV. People were 67% more likely to acquire HIV if they had previously acquired HSV-2 more than twelve months ago, nearly twice as likely if they acquired HSV-2 for between six months and a year ago, and nearly four times as likely (hazard ratio 3.81: 95% CI, 1.81-8.03) if they had had HSV-2 for less than six months or had acquired HIV and HSV-2 simultaneously. The investigators also found a strong link between visible genital ulcers caused by HSV-2 and recent HIV infection, with a fivefold increased risk of being infected with HIV for patients with visible ulcers, compared with a twofold risk with asymptomatic chronic HSV-2 infection.

An accompanying editorial suggested that anti-HSV therapy could prove a useful tool in controlling the spread of HIV, particularly in resource-limited countries. Cost would not pose a significant barrier. Generic aciclovir, the standard treatment for acute and chronic HSV-2, costs approximately 20 US cents a day, or $73 a year, and had a good safety and resistance profile, the authors pointed out.

Subsequent studies have reinforced the link between HSV-2 infection and HIV, especially the fact that primary HSV-2 infection represents a very high-risk time for HIV acquisition.

In a 2007 study of male and female US army veterans,⁵ having HSV-2 was associated with a fourfold higher risk of having HIV, though this was a cohort study which could not establish which infection came first or tease out other risk factors. The ‘population attributable risk’ (PAR) of HIV due to HSV-2 infection was 25% - in other words, there might be 25% fewer cases of HIV if HSV-2 did not exist.

In this study, each serving member of the US military diagnosed with HIV between 2000 and 2004 was matched with two members of the military who tested HIV-negative as controls. Demographic data and infection with HSV-2 were compared. The study involved 492 individuals with HIV and 984 controls. HSV infection was present in 30% of those diagnosed with HIV but only 10% of the controls.

A study from Moshi, Tanzania,⁶ among 689 initially HIV-negative female bar and hotel workers looked at the impact of HSV-2, as well as other STIs. During the twelve-month follow-up period, there were 32 HIV seroconversions – an annual incidence rate of 4.6%.

Women with chlamydia at baseline had 5.2 times the risk of subsequent HIV infection. Women with bacterial vaginosis had twice the risk. Women with HSV-2 at baseline had 4.3 times the risk, but 5.5 times the risk if their HSV-2 infection was recent. The authors estimated that approximately 63% of the HIV-1 incidence was attributable to prevalent HSV-2 infections.

An even stronger association between acquiring HSV-2 and acquiring HIV was found in a study of 4531 women seeking reproductive and general healthcare services at seven sites in Uganda and Zimbabwe.⁷

HSV-2 prevalence at enrolment was 52% in Uganda and 53% in Zimbabwe. HSV-2 annual incidence during follow-up was 9.6% and 8.8% in Uganda and Zimbabwe, respectively.

In Uganda, women with HSV-2 at baseline were 2.8 times more likely to subsequently become infected with HIV, and women who acquired HSV-2 during the study 4.6 times more likely. In Zimbabwe, women with HSV-2 at baseline were 4.4 times more likely to seroconvert for HIV and women who acquired HSV-2 during the study 8.6 times more likely. The population attributable risk for HIV infection due to HSV-2 infection was 42% in Uganda and 65% in Zimbabwe. Overall, approximately 84% of all HIV seroconversions (178/211) were in women with prior HSV-2 infection.

A substudy of the South African randomised controlled trial (RCT) of circumcision found that HIV and herpes simplex virus 2 (HSV-2) independently reinforced each others’ effects.⁸ People with one infection in the study were more likely to acquire the other infection. The study also found that circumcision protected men in the study against HSV-2 as well as HIV.

However, it did not find that HSV-2 infection had any effect, positive or negative, on the anti-HIV protective effect of circumcision: men who were circumcised during the study had the same decrease in the risk of HIV infection regardless of HSV-2 status.

The substudy examined 2974 young men out of the 3274 who enrolled in the circumcision trial. It also included 122 men (5.6%) who already had HIV at the time of enrolment and were screened out of the circumcision trial.

Having herpes had a highly significant effect on the chance of acquiring HIV. One point seven per cent of HSV-2 negative men acquired HIV during the study, 6.6% of men who were HSV-2 positive at enrolment, and 14.3% of men who acquired both HIV and HSV-2 during the study. This means that men already infected with HSV-2 were four times more likely to acquire HIV than HSV-2 negative men, and men who acquired HSV-2 8.8 times more likely.

Men already infected with HIV were four times more likely to acquire HSV-2 than HSV-2 negative men and men who acquired HIV 7.7 times more likely.

Circumcision protected against HSV-2 as well as HIV. In an ‘as treated’ analysis, HSV-2 annual incidence among men who remained uncircumcised was 3.8% and in circumcised men 2.1%. This was a 45% reduced risk.

Looking at the increased relative risk of HIV infection among HSV-2 study participants allowed the researchers to calculate that the proportion of HIV infections that were attributable to HSV-2 infection was 26%; in other words, a perfect anti-HSV-2 prophylactic might be expected to reduce HIV infection by this amount.