Study confirms STIs make no difference to undetectability and infectiousness in people taking HIV treatment

A study conducted with gay men in Thailand has found that people who are diagnosed with HIV and start antiretroviral therapy (ART) are no less likely than others to have an undetectable viral load if they are diagnosed with a sexually transmitted infection (STI).

This confirms the generalisability of the “Undetectable = Untransmittable” (U=U) message. It shows that the original Swiss Statement on undetectability and transmission, issued ten years ago, was too cautious when it excluded people with STIs.

However, the study also found that viral loads in blood, semen and rectal secretions were considerably higher in people with STIs at the time of diagnosis, before they started ART,  especially in semen. This adds to the evidence that STIs may be a major amplifying factor when it comes to the speed with which HIV spreads in different populations.

The study looked at the 143 gay men diagnosed with HIV out of 492 gay men and transgender women enrolled into the Test and Treat Demonstration Project in four clinics in Thailand. At baseline 106 participants tested HIV positive (21.5%) and 37 became HIV positive during the two-year study (6.4% annual incidence). At baseline, the mean age was 26, and 74% said they did not use condoms consistently.

The high prevalence of HIV was matched by a high prevalence of STIs. Fifty-two per cent of men had an STI diagnosed at baseline – more than in most other studies, but about the same as in the PROUD study of pre-exposure prophylaxis (PrEP). Chlamydia was found in 33% of participants, gonorrhoea in 23% and syphilis in 16% (the highest prevalence of syphilis seen in any study apart from one in Taiwan).

At baseline, the average HIV viral load in blood was 80,000 copies/ml. It was four times higher in men with any baseline STI (158,000 copies/ml) than in men without an STI (40,000 copies/ml). In rectal secretions, there was an eightfold difference between men with an STI and men without (12,500 vs 1580 copies/ml).

These may sound like large differences, but it takes differences of orders of magnitude in viral load to make a difference to transmission and disease progression. More significant perhaps was the fact that there was a 77-fold difference in the average viral load in semen; it was 3100 copies/ml in men with STIs and only 40 copies/ml in men without – almost undetectable, and almost certainly non-infectious. In fact, more than a third of men had undetectable viral loads in both rectal secretions and semen when diagnosed, and most of those did not have STIs, underlining the fact that STIs may be a significant factor in transmission.

One hundred and thirty-three people (93%) started ART, taking a standard, World Health Organization-prequalified generic fixed dose combination of tenofovir, lamivudine and efavirenz. Two men with efavirenz-related side-effects switched to a lopinavir-based regimen. Adherence as measured by drug levels was over 95% at months 12 and 24. 

During follow-up, men who had STIs were no more likely to have a detectable viral load than other men. Of 114 people followed up at month 12, five (4.4%) had detectable viral loads in blood – two with STIs and three without. Only 68 people were followed up to month 24, but the lower numbers were not due to drop-out, but to the study ending sooner than 24 months after recruitment in some cases. Of these, two men with STIs and none without an STI had a detectable viral load, but this difference was not statistically significant.

Detectable viral load in semen was found in only one person at month 12 and two people at month 24 and only one of these three had an STI. Detectable viral load in rectal secretions was only found in one person at any time point, and he did not have an STI.

The previous data relate to having an STI at the same time as a detectable viral load. It may be notable, however, that all five men who were detectable at month 12 were people who had had syphilis at baseline but had been cured.

This may not be because STIs are causing people to be detectable, but because both STIs and poor adherence may be signs of other difficulties such as depression. For instance, the only person who had both blood and seminal viral detectability at month 24 had gonorrhoea and chlamydia at month 12. This was the one patient who developed drug resistance, to efavirenz and lamivudine, by month 12 and was put on darunavir – despite which, he still had a blood viral load of 200 copies/ml at month 24. This participant clearly had issues with adherence and his STIs may have been an indicator of that.

This study’s location in Thailand is significant. The only previous study of undetectability and infectiousness in Thailand was Opposites Attract, in which 105 Thai gay couples took part, and this adds to our knowledge of adherence and infectiousness in the region. The use of a standard World Health Organization-prequalified regimen implies that the findings may be generalisable to other lower-income settings. And, although the researchers do not mention this, the predominant viral strain in Thailand is the CRF-AE subtype, which is suspected of being more virulent than most others.

The study is not the first to look at whether STIs raise the risk of viral detectability in people on ART, or whether undetectability in blood also means undetectability in semen and rectal secretions. A US study reported in 2012 found that a quarter of US gay men with undetectable viral loads in their blood still had low, but detectable, viral loads in semen – though 40% of men in this study had been on ART for less than a year. But a small London study reported in 2015 found that the one in three gay men who had an STI when starting ART were no less likely to become virally undetectable in rectal secretions than the majority who did not.

This study adds to the evidence that there is no evidence that having an STI will turn someone who cannot transmit HIV into someone who can.