In the Boston study, participants were on average 43 years old and three-quarters were white. They had all been on antiretroviral therapy (ART) for more than three months and 80% for over a year.
Nearly three-quarters were judged as being at high risk of acquiring an STI because they had had unprotected sex in the last three months. Nine of the men were diagnosed with an STI (gonorrhoea, syphilis, chlamydia or non-gonococcal urethritis) and 24 had leukocytospermia or white cells from the immune system in the sperm, indicative of urethral inflammation.
Eighteen of the 101 men had a detectable viral load in their blood; their median blood plasma viral load was 560 copies/ml and ranged from 80 to 640,000 copies/ml. Nine of these 18 men also had detectable HIV in their semen (50%).
Of the 83 men without detectable HIV in their blood, 21 (25%) had detectable HIV in their semen. The median seminal viral load in these men was 200 copies/ml and ranges from 80 to 2560 copies/ml.
As well as having an STI, in multivariate analysis, two other factors remained strongly associated with having detectable HIV in semen in men without it in blood. High levels of the inflammatory cytokine TNF-α were associated with a 14-fold greater risk of a discordant seminal viral load, and having had unprotected insertive anal sex (being ‘top’), which was associated with a more than sevenfold greater risk.
There were therefore in this study low but detectable levels of HIV in the semen of a quarter of men who would register as being virologically suppressed on a viral load blood test. To what extent might this be contributing to ongoing HIV transmission in gay men? This is unknown, but the researchers point out that although a viral load below 1000 had rarely been associated with transmission in heterosexual studies, some infections have occurred and animal models suggest that HIV is five times more transmissible via anal than vaginal sex – so a median viral load of 200 would imply a low but definite risk of transmission.
A 2008 study from San Francisco (Butler) found that the median seminal viral load in men transmitting HIV to partners was just 4300 copies/ml and the lowest was 110 copies/ml, while a 2009 study from Brighton in the UK (Fisher) that linked HIV infections in gay men genetically found that two out of 41 transmissions of HIV (5%) were from men with an apparently undetectable viral load.
However, studies of the link between viral load and transmission suffer from it being difficult to pin down transmitters in a cohort of gay men with multiple partners and where viral load may be measured months after the transmission (in the Butler study, the average gap between transmission and viral load test was 103 days).
One interesting aspect of this study was the higher risk of seminal viral load associated with unprotective insertive sex. The researchers suggest that urethritis in these HIV-positive gay men could be caused by infections with fecal bacteria acquired during sex or even that the virus detected could be passively carried virus from other HIV-positive gay men. Either way, this would tend to increase the infectiousness of HIV-positive men if they have insertive sex with negative men.