The shedding of cytomegalovirus (CMV) in the genital tract is associated with a higher HIV viral load in semen, according to a small Canadian study published in the January 1st 2006 edition of the Journal of Infectious Diseases (now online).
HIV viral load in semen is strongly associated with viral load in blood. Although the amount of HIV in semen is generally lower than in blood, some men shed disproportionately large amounts of HIV in their semen, and many men who have low HIV levels in their semen have intermittent “spikes” in HIV viral load in their semen.
Increased shedding of HIV in semen has been associated with sexually transmitted infections. Although not classified as a sexually transmitted infection, CMV can be readily transmitted during sex and earlier studies have suggested that CMV is present in also present in the semen of men shedding large amounts of HIV in their genitals.
To gain a better understanding of the role of CMV activation in the genitals on HIV viral load in semen, investigators from Toronto obtained paired blood and semen samples from 26 HIV-positive men. All the men had been infected with HIV for over six months, none were taking antiretroviral therapy, and all had tested negative for sexually transmitted infections including gonorrhoea and Chlamydia. All 26 men were gay and all were infected with CMV.
Median CD4 cell count was 574 cells/mm3 and median HIV viral load in blood was 21,400 copies/ml. CMV was not detected in the blood of any of the men, but was being shed in the semen of 17 (65%).
Although men shedding CMV in their genitals had a lower median CD4 cell count than those who were not (509 cells/mm3 versus 696 cells/mm3), there was no difference in their HIV viral load in blood.
Overall, median HIV viral load was higher in blood than in semen (20,000 copies/ml vs 5,000 copies/ml), but the investigators found that in nine men (35%) more HIV was being shed in semen than in blood.
These nine men were defined as “disproportionately high semen HIV shedders” and their median blood HIV viral load was a little under 22,000 copies/ml and their median semen HIV viral load was 48,000 copies/ml. CMV was detected in the semen of all nine men who were shedding disproportionately large amounts of HIV in their semen, but was only found in the semen of eight of the other 17 men who had more HIV in their blood than semen. This difference was statistically significant (p = 0.002). The investigators also found that the mean CMV viral load in semen was over 100 fold greater in the men shedding disproportionately large amounts of HIV in their genitals.
The investigators also found a correlation between semen HIV load and the level of CMV shedding in the genitals (p = 0.05).
Further tests were performed by the investigators which indicated that CMV activation discreet to the genital tract was causing HIV viral load to be disproportionately increased in the semen of the nine men.
Tests were also performed to see if infection with herpes simplex virus-2 (HSV-2) increased HIV viral load in the genital tract. However, no relationship was found between infection with HSV-2 and disproportionately high shedding of HIV, even when CMV shedding was included in the analysis.
“We have confirmed that the presence of CMV in semen is associated with increased likelihood of HIV shedding”, write the investigators. They add, “more importantly, we have demonstrated that disproportionate shedding of HIV in semen, relative to that in blood, is strongly associated with local CMV reactivation in the male genital tract…these observations emphasize the potential for very substantial negative synergy between these two infections.”
The investigators’ findings could have very important implications for HIV prevention. Men with increased shedding of HIV in their genitals due to CMV are likely to be considerably more infectious for HIV and the investigators conclude, “prevention of CMV reactivation warrants study as a possible strategy to reduce semen shedding of HIV.”
Sheth PM et al. Disproportionately high semen shedding of HIV is associated with cytomegalovirus reactivation. J Infect Dis 193 (online edition), 2006.