Objective: Factors that influence the infectivity of an individual and
the impact of antiviral treatment on infectivity are not well defined
. This study investigated the value of a sensitive method for detectin
g infectious HIV in semen for use as a marker for infectivity. Design:
A cross-sectional study of infectious HIV in the semen of 33 HIV-posi
tive men. Methods: A sensitive method for detecting infectious HIV in
semen was used. The correlation of culture in semen with clinical and
laboratory data was investigated. Biological phenotypes of isolates fr
om blood and semen were tested using an MT-2 assay. Results: HIV cultu
res from seminal cells were positive in 18 patients (55%) and in one p
atient from seminal plasma. Higher recovery rates of HIV from semen co
rrelated with a low CD4 count (80% in patients with a CD4 count > 100
x 10(6)/l versus 33% In patients with a CD4 count < 100 x 10(6) cells;
P < 0.025) and symptomatic disease (78 versus 27% in asymptomatic pat
ients; P < 0.01). Recovery of HIV from semen was independent of presen
ce or absence of plasma viremia and the biological phenotype of blood
isolates. Ten patients with syncytium-inducing (SI) isolates in their
blood had positive semen cultures for HIV. Seven of the 10 patients ha
d SI isolates recovered from their semen, whereas three had non-SI iso
lates only. Conclusion: Data from partner studies show higher rates of
HIV transmission for patients with low CD4 counts and symptomatic dis
ease. The compatibility of epidemiologic data with our finding that si
gnificantly more HIV is recovered in semen from patients with advanced
disease, suggests that HIV culture of semen samples may provide a use
ful surrogate marker to measure infectivity in clinical studies. Furth
er studies are needed to define the inoculum required to transmit HIV
and to study the impact of sexually transmitted diseases and HIV-1 phe
notype on semen infectivity.