Presenter: Jérôme LeGoff. Virology Unit and U743, HEGP, Inserm, Paris, France.

Objectives: There is conflicting evidence linking HSV-2 infection to HIV-1 transmissibility.We assessed HSV-2–HIV associations among women with genital ulcer disease (GUD) at baseline of a randomized placebo-controlled trial of episodic aciclovir treatment conducted in Ghana and Central African Republic.

Methods: Blood samples were analysed for HSV-2, syphilis and HIV serologies, CD4 lymphocyte count and plasma HIV-1 RNA load. GUD aetiology and lesional HIV-1 RNA were determined by polymerase chain reaction (PCR) from ulcer swabs. HIV-1 and HSV-2 cervico-vaginal (CV) shedding were assessed in CV lavages (CVL) by real-time PCR. CVL with positive Y-chromosome PCR were excluded. Women were defined as HIV-1 or HSV-2-infected if virus (HIV-1 RNA or HSV-2 DNA) was detected in a lesion or CVL, or they were seropositive.

Results: 388/441 (88%) women had Y-PCR-negative CVL, of whom 47% were HIV-1 infected and 79% were HSV-2-infected. 188 (50%) women had an HSV-2 ulcer, a further 14 women had detectable CV HSV-2 DNA only, and 177 women had no lesional or CV HSV-2 DNA. Among HIV-1-infected women, CV HIV-1 RNA was more frequently detected in women with HSV-2 ulcers (68%) or CV HSV-2 DNA only (73%) compared with women without HSV-2 infection (43%) (P=0.004). Similarly, CV HIV-1 RNA loads were higher among women with HSV-2 ulcers or CV HSV-2 DNA (3.13 and 3.17 log10 copies/mL) compared with women without HSV-2 genital infection (1.2 log10 copies/mL) (P=0.002). 180 women were co-infected with HIV-1 and HSV-2. Among these, median CV HIV-1 RNA load was significantly higher among women with detectable CV HSV-2 DNA than those without (3.43 vs 1.2 log10 copies/mL; P<0.001). Similarly, plasma HIV-1 RNA load was higher among women with CV HSV-2 DNA (median 5.13 vs 4.64 log10 copies/mL; P=0.03). Detectable CV HIV-1 RNA was also associated with lower CD4 counts, increased plasma HIV-1 RNA and presence of lesional HIV-1 RNA. There was significant correlation in the quantities of CV HIV-1 RNA and CV HSV-2 DNA detected among co-infected women (Spearman’s correlation 0.35, P<0.001), and after adjusting for CD4 count and plasma HIV-1 RNA load.

Conclusions: HSV-2 lesional and genital shedding was associated with increased genital HIV-1 replication in the population of HIV-1- and HSV-2-infected women with GUD. These findings provide biological plausibility for the role of HSV-2 as a co factor of sexual transmission of HIV-1.