Plasma and Genital Tract Viral Load and STDs in Early HIV Infection

Objective: To assess the relationship between plasma and genital tract viral load (VL) and semen VL and STDs during early HIV infection.

Design: Paired plasma and semen or cervicovaginal lavage (CVL) samples were obtained every 3 months from persons who enrolled in the HIVNET cohort of newly infected participants. For men, urine and rectal samples, and for women, cervical samples were obtained at enrollment for gonorrhea (GC) and chlamydia (CT) ligase chain reaction; all participants were asked about interim STDs in the prior 3 months at follow-up. Roche PCR was performed on plasma, semen and CVL samples; NASBA, which includes a silica extraction for inhibitors, was performed on a subset of semen samples. Limit of detection was <200 for Roche PCR and <400 for NASBA with >200 mcl of sample (or <1000, with <200 mcl of sample).

Results: 99 seroconverters (86 M, 13 F) were enrolled in the HIVNET early infection study an average of 6 months after seroconversion, 58 of whom took antiretroviral treatment during follow-up. 84%, 29% and 54% had detectable viral load by plasma PCR, semen PCR, and semen NASBA, respectively. Median viral load in plasma was 11,010 (range <200-608,549) compared to a median of <200 (range <200-92,840) in semen by PCR. For 96 semen samples with both NASBA and PCR results, 20 were positive by NASBA but undetectable by PCR and 9 were positive by PCR only. The rank correlation for semen and plasma viral load using the Roche PCR assay was 0.22 (p=0.018) and 0.3 (p=.00009) when men were on or off antiretroviral therapy, respectively. Among 150 paired semen and plasma, 94 had detectable semen VL by either PCR or NASBA. Of these 94, 5 (5%) had detectable semen VL when plasma VL <200, 17 (11%) when plasma VL was 200-10,000, and 72 (48%) when plasma VL was >10,000. The probability of detectable semen VL was significantly associated with plasma VL, both with and without adjustment for antiretroviral treatment (P<.001). Six seroconverters had confirmed STDs at enrollment (3 rectal GC, 1 rectal CT, 2 syphilis) and 25 participants reported an interim STD at follow-up. However, there was no association between semen VL and STDs, although power was limited given no confirmed urethral STDs at enrollment and the temporal difference in semen collection for VL testing vs. date of self-reported STDs.

Conclusions: Semen viral load in the first 1½ years after seroconversion is moderately correlated with plasma viremia, both when individuals were and were not on antiretroviral therapy. NASBA appears to be a more sensitive assay for detection of virus in semen. Viral load in cervicovaginal samples will also be presented. With the small number of confirmed genital tract infections at the time of semen collection, STDs were not associated with semen viral load. Importantly for counseling patients about potential transmission risk, plasma viral load correlates poorly with undetectable semen viral load.

Connie Celum
Pacific Medical Center
901 Boren Avenue, Suite 1300
Seattle, WA 98104