HIV can be
intermittently detectable in the semen of men taking long-term antiretroviral therapy
that is achieving full suppression of viral load in the blood, French
investigators report in PLOS ONE.
They also found evidence that viral load in semen can fluctuate over a very
short period. However, it is unclear whether the levels of viral load detected in this study pose a substantial risk for the transmission of HIV.
therapy can suppress viral load in both the blood and semen. This is of real
significance to HIV prevention initiatives. It has been demonstrated that
starting antiretroviral therapy and the achievement of an undetectable viral
load in blood reduces the risk of sexual transmission of HIV in heterosexual couples
by 96%. Interim results from the ongoing PARTNER study showed that there were
no HIV transmissions in serodiscordant heterosexual and gay couples when the
HIV-positive partner was taking antiretroviral therapy and had an undetectable
viral load in blood.
several studies have shown that viral load can be intermittently detected in the semen of men who are taking treatment and who have
an undetectable blood viral load.
a unit in Paris offering assisted conception to HIV serodiscordant couples designed
a study involving 88 men living with HIV who received care at the clinic between
2006 and 2011. All were taking HIV therapy and had an undetectable viral load in blood
for at least six months. Using frozen sperm samples obtained from these men,
the investigators calculated the detection rate of viral load in semen and also
evaluated if the shedding of HIV in semen could change over a very short period
A total of 306
frozen semen samples were available for evaluation. The samples were obtained
by masturbation after two to seven days of sexual abstinence. If possible, each man provided two samples within a one-hour interval.
HIV was detected
in at least one semen sample for 17 men (19%) and in 23 samples in total (7.5%).
Median viral load
in these samples was 705 copies/ml, but in eleven samples it was above 1000
Of the 129 samples
in which two semen specimens were provided within one hour, twelve (9%) had
discordant results – viral load undetectable in one specimen but detectable in
the other. Median viral load in the detectable samples was 918 copies/ml, and
in six cases was above 1000 copies/ml.
“We show that
intermittent shedding of HIV-1 RNA [viral load] in the semen of patients given
efficient cART [combination antiretroviral therapy] could occur within a one-hour interval,” write
It is unclear whether these levels of viral load in semen are sufficiently high to pose a significant risk of HIV transmission. A study conducted in Rakai, Uganda, in serodiscordant couples prior to the introduction of antiretroviral therapy found no cases of HIV transmission where the HIV-positive partner had a blood viral load below 1500 copies/ml during a two-year follow-up period. A similar threshold has not been defined for viral load in semen.
The study found a trend towards a higher frequency of detectable HIV in the semen of men taking protease inhibitor-based treatment.
Over a quarter
(28.6%) of men taking HIV treatment based on a protease inhibitor (PI) had at
least one episode of detectable virus in their semen compared to 7.7% of men taking HIV treatment based on a non-nucleoside reverse transcriptase inhibitor (NNRTI) and 7.7% of men taking a combination based on another type of
anti-HIV drug. These differences were marginally short of significance (p =
towards a higher risk of a detectable sp VL [seminal plasma viral load] in
patients given PI-containing cART compared to a regimen containing an NNRTI
might be explained by the poor diffusion of most PIs in the male genital
tract,” suggest the authors.
believe their findings that approximately a fifth of men had HIV detected in
their semen and that virus was detected in 7.5% of all samples “should balance
messages on the individual risk of HIV transmission through unprotected sex as
an exclusive preventive strategy in serodifferent couples with procreation
The authors cite previous calculations suggesting that the risk of HIV transmission is in the order of three cases per 10,000 episodes of vaginal intercourse when seminal viral load is 1,000 copies/ml.
accumulating evidence of the extremely low risk of sexual
transmission of HIV in the context of virologically effective HIV treatment suggests that the possibility of small fluctuations in viral load in semen has not translated into cases of HIV infection in either a large clinical trial (HPTN 052) or a closely-monitored large cohort (the Partner study).