[1] for their analysis of the possible sexual transmission of HIV from patients whose viral load is <50 HIV-1 RNA copies/mL. The impetus for their work is the claim of the Swiss Federal Commission for HIV/AIDS that patients with undetectable plasma viral loads for six consecutive months are noninfectious provided that there are no concurrent sexually transmitted infections (STIs). Engsig et al. have found that regularly monitored HIV-infected

patients on highly active antiretroviral therapy (HAART) may present a greater risk of transmission than purported by the Swiss statement, particularly in the initial 12 months of therapy. This finding, inferred from their data about the dynamic nature of plasma viral loads, is important and extends our knowledge about HIV transmission risk. One of several concerns with the Swiss statement is its reliance on data almost exclusively Trichostatin A concentration from heterosexual couples and the lack of evidence on the magnitude of transmission

risks associated with low viral loads. Our recent work in Sydney [2] suggests that, despite the widespread availability of HAART, transmission rates among men who have sex with men (MSM) are now astonishingly similar to those seen in the pre-HAART era. Diagnosis rates have been increasing in Australia in an buy SP600125 era of increased HAART coverage and effectiveness. Similar findings have been reported from France [3]. Although HIV may be undetectable in blood, it may be present in semen or genital fluids at infectious levels. Indeed, the association between Methamphetamine plasma viral load and seminal viral load is far from perfect. For example, Lorello et al. [4] investigated

33 HIV-positive men who had plasma viral loads of <50 copies/mL for a mean of 3.96 years and who had been screened for STIs. Two of 33 men (6%) had detectable HIV in their semen. In another study, Sheth et al. [5] followed a prospective cohort of 25 men free of STIs initiating HAART. Despite their achieving a plasma viral load of <50 copies/mL, HIV was detectable in semen samples of 48% of the men on more than one occasion. In a control group of 13 other HIV-infected men who had undetectable plasma viral load at every 3-monthly assessment for the past 7 years, HIV was detected in semen samples in 31% of these men. Sheth et al. could not find any relationship between semen viral loads and the concentration of antiretroviral drugs in that compartment. HIV detected in semen samples was sensitive to drugs used by study participants. The degree of sexual infectiousness of MSM for given viral loads in plasma (or in semen or the rectum) is still not known. However, the results of Engsig et al., Lorello et al. and Sheth et al. underscore the possibility that, in some cases, HIV transmission may occur despite an undetectable plasma viral load. An undetectable plasma viral load does not imply an undetectable viral load in semen or rectal fluids.