Studies of HIV infectiousness and STIs

In vitro experiments have shown that both syphilis¹ and chlamydia² organisms cause mucosal and HIV-infected immune cells to secrete cytokines that cause ‘upregulation’ (increase) of the HIV co-receptor CCR5 on the surface of T-cells and increase HIV production from infected cells.

Moreover, in the 1990s a number of studies established that having STIs greatly enhanced HIV transmission, or at least viral load, in genital secretions. A number of studies concerned urethritis (inflammation of the urethra, the tube in the penis which transports urine out of the body), which may be caused by a number of different infections.

In 1995, Moss and colleagues³ found that, in a group of 106 HIV-positive men, having STI-associated urethritis increased the chances of HIV-infected cells being found in semen 2.4-fold and, when it was caused by gonorrhoea, nearly threefold. The only other factor associated with cell-associated HIV being found in semen was a CD4 count under 200 cells/mm³. Treating patients for gonorrhoea halved the proportion that had productively HIV-infected cells in their semen from 44% to 21%. Moss and colleagues comment that: “Treatment of gonococcal urethritis may be an effective strategy for reducing HIV transmission.”

In a study in Malawi in 1997,⁴ Cohen and colleagues measured the seminal viral load in 135 HIV-positive men: 86 had urethritis and 49 controls did not. The men with urethritis had very high HIV concentrations in their semen - eight times higher than those in HIV-positive men without urethritis (124,000 versus 15,100 copies/ml, p = 0.035). This was despite similar CD4 counts and similar blood plasma viral loads. Again, infection with gonorrhoea was associated with the highest seminal viral loads – 158,000 copies/ml. After the patients with urethritis received antibiotics for STIs, the HIV viral load in their semen decreased significantly to 89,100 copies/ml after a week and 41,200 copies/ml at two weeks. Blood plasma HIV concentrations did not change.

Cohen commented: “HIV…control programmes, which include detection and treatment of [STIs] in patients already infected with HIV-1, may help to curb the epidemic. Targeting of gonococcal urethritis may be a particularly effective strategy.”

Another Malawian study focused on genital ulcer disease (GUD), which is any infection causing ulceration in the genital area. The 1998 study recruited HIV-positive men who all had urethritis and measured the association between genital-ulcer disease and HIV viral load in blood and semen, after adjusting for infection with gonorrhoea.⁵ GUD was associated with lower CD4 cell counts (258 versus 348 cells/mm³) and increased blood plasma HIV viral load (median, 240,000 versus 79,400 copies/ml). Men with GUD had very much higher viral loads in their semen than men with gonorrhoea-associated urethritis who did not have genital ulcers. The seminal viral load in men with GUD was a very high 195,000 copies/ml, whereas in men without GUD it was a relatively modest 4000 copies/ml. Average levels in all men decreased fourfold following antibiotic therapy.

Dyer and colleagues commented: “Increased HIV-1 in semen was demonstrated in some men with GUD; such an increase could lead to increased transmission, thus complicating interpretation of the role of the genital ulcer itself in the infectiousness of HIV.” This indicated that there might be even more significant influences on HIV transmission in men than urethral inflammation.

Another study in The Lancet established that STIs in HIV-positive women also increased the viral load in their genital secretions.⁶ Mostad and colleagues investigated 318 women previously diagnosed with HIV who presented to an STI clinic in Mombasa, Kenya, during 1994-96. HIV-infected cells were detected in 51% of endocervical and 14% of vaginal-swab specimens. The most significant influence on HIV viral shedding was low CD4 count. After adjustment for CD4 count, three other strong associations were found with the likelihood of HIV being found in the genital secretions:

• One was treatment with hormonal contraceptives; hormones such as progesterone thin the vaginal epithelium so that it is only three to four cells thick rather than the usual twelve or so. High-dose oral contraceptives increased the chance of HIV being found in genital secretions 13-fold, low-dose oral ones 3.8-fold and injectable ones nearly threefold.
• Another was vitamin A deficiency, with moderate deficiency being associated with an eightfold increased chance of viral shedding (though this could be a marker for malnutrition and/or poor gut inflammation associated with HIV).
• Finally, gonorrhoea that caused inflammation of the cervix increased the chances of HIV being found in genital secretions more than threefold and vaginal candidiasis 2.6-fold.

In this study, however, trichomoniasis and chlamydia were not associated with increased HIV shedding.

In contrast, Ghys and colleagues did find chlamydia to be associated with increased likelihood of viral shedding, as well as genital ulcers and gonorrhoea.⁷ They investigated HIV shedding in genital secretions in 1201 female sex workers in Côte d’Ivoire, 404 of whom had HIV-1, 21 of whom had HIV-2 and 205 of whom were co-infected with both viruses. A quarter (96) of the HIV-1 positive women had detectable HIV in their cervical and vaginal secretions, as did a quarter of the dually infected women, but only one of the twenty-one women who had HIV-2 mono-infection. Low CD4 percentages were associated with cervicovaginal shedding of HIV. Infection with gonorrhoea nearly doubled the chance of HIV shedding (odds ratio 1.9), chlamydia multiplied it 2.5 times and the presence of a cervical or vaginal ulcer multiplied it fourfold. HIV-1 shedding decreased from 42 to 21% in women whose STIs were cured.

In a study of 1144 gay men with HIV in Brighton, Martin Fisher and colleagues were able, through phylogenetic analyses of clusters of infection, to find 41 men who had recently transmitted HIV to other patients.⁸ Seventy per cent of transmitters were not taking HIV treatment and 22% had been on treatment, but had recently interrupted it. Independent of this, the risk of transmitting HIV was 2.8-fold times greater in men who had had a recent STI.