A Summary of Four Presentations:

High Levels of HIV-1 in Semen and Blood in Men in Malawi
Presenter: Joe Eron, University of North Carolina, Chapel Hill

High Viral Load in Semen of HIV-1 Infected Men at All Stages of Disease and Reduction by Antiretroviral Therapy
Presenter: Phalgani Gupta, University of Pittsburgh, Pittsburgh PA

Recent HIV Infection Treated with AZT, 3TC and a Potent Protease Inhibitor
Presenter: Martin Markowitz, Aaron Diamon AIDS Research Center, NY, NY

The Effect of HIV Protease Inhibitors on Seminal Proviral DNA Presenter: SL Boswell, Fenway Community Health Center, Boston MA

How much HIV is present in semen and whether antiretroviral therapy can reduce the viral load in semen is of pressing concern (N.B. I have edited all of my bad puns from this report. I leave slippery slope jokes to King Holmes who can get away with it). The obvious question is whether effective antiviral therapy can render a person "noninfectious" reducing the risk to their partners and contributing to control of the epidemic. The other question is more subtle. Are the testicles a "sanctuary site" where drug levels remain low and virus can persist and replicate. In some types of cancer, the testicles are a potential sanctuary site. Measuring the relationship between viral RNA in the blood and semen is one way to look at this. The ability to measure free virus in seminal plasma (the sperm and white cell fluid portion of the ejaculate) has been complicated by the fact that PCR reactions are inhibited by semen. However, NASBA seems to give accurarate and reproducible results.

A study by Joe Eron and colleagues compared the levels of viral RNA in the blood plasma and the seminal plasma relative to CD4 count in a group of 53 men in the US and Switzerland and compared it to a group of 49 men in the African country of Malawi. HIV RNA could be easily detected by NASBA in 60-80% of the men. The mean viral load in the blood plasma was 4.45 log10 (28,182 copies) in the western men and their mean viral load in semen was slightly lower at 3.68 log10 (4,780 copies). In contrast, both the viral load in blood and semen was higher in the men from Malawi, despite similar CD4 counts. The mean viral load in the blood plasma was 5.23 log10 (170,000 copies) in the western men and their mean viral load in semen was 4.18 log10 (15,136 copies). The investigators speculate that the higher viral loads in semen might help explain the greater efficiency of sexual transmission in Africa.

Phalgani Gupta and his colleagues at Pittsburgh also used NASBA to examne viral RNA in semen in 34 gay men at various stages of disease. The median viral load in semen was 11,000 copies per mL, ranging from <400 to almost 3 million. High levels were seen in all disease stages, and there was no correlation of seminal viral load and the CD4 count. The blood viral RNA was fairly well correlated with seminal viral RNA, as in the study by Eron. In 6 patients treated with indinavir or indinavir plus DMP 266, the viral load in the blood and seminal plasma fell below the limits of detection.

However, the absence of free virus in the seminal plasma may not mean that a person is not infectious for two reasons. First, semen contains lymhocytes and other cells which may be infected with HIV and may play a key role in infection. Sexually transmitted diseases increase the number of cells in semen and likely increase the number of infected cells. Second, the assays for viral RNA used in semen are able to detect, at best 400 or more copies per mL. The average ejaculation could contain more than 2000 virus copies despite measuring "undetectable."

As part of a late breaker abstract about ongoing experiments to eradicate HIV, Marty Markowitz of the Aaron Diamond presented the semen analysis on the 20/24 acute seroconverters who remain on the study. All showed no replicating virus in the cellular compartment as measured by RNA PCR, but all of the men still had proviral DNA. Seminal plasma results are in progress. A study by Boswell and collaborators demonstrated that proviral DNA (indicating infected cells) was detectable in the majority of men. Neither CD4 count nor plasma viral load (using the first generation bDNA) predicted which men had detectable proviral DNA in their semen. After beginning indinavir, proviral DNA became undectable in some but not all of the men, but several had detectable virus in semen after having been undetectable before treatment.

Taken as a whole, these studies suggest that

1. The amount of virus in the semen is substantial at all stages of infection.
   
2. The higher the viral load in blood, the higher the seminal viral load.
   
3. In general, the amount of viral RNA in the seminal plasma drops with effective therapy, which suggests that the testicles are not a sanctuary.
   
4. As long as infected cells are clearly present in semen regardless of treatment, semen should be considered infectious even in patients with extremely low viral loads in blood or even in seminal plasma.