Lower seminal viral load and hints of long-term viral control seen in Thai HIV vaccine recipients

This article is more than 12 years old. Click here for more recent articles on this topic

Men who received the Thai RV144 vaccine who nonetheless became infected with HIV had lower viral loads in their semen than men who received placebo, a new analysis of the vaccine study reports in the Journal of Infectious Diseases.   

The report also finds tantalising signs of long-term improvements in HIV viral load and CD4 count, starting four years after infection, in HIV-positive vaccine recipients.  However the researchers stress that this finding is only just statistically significant and could either be the result of the vaccine slowing long-term disease progression in general, or of it speeding up progression and therefore increasing mortality in a subset of ‘fast progressors’.

This study is encouraging because it revives the idea that a vaccine or other biomedical prevention technique could ‘work’ even in people who become infected, by slowing down or stopping progression to AIDS and/or by permanently reducing their viral load so they are less infectious. Vaccines such as the one in the STEP Study, which aimed to stimulate immune cells to kill cells already infected with HIV, were expected to do this, but proved disappointing.



A pill or liquid which looks and tastes exactly like a real drug, but contains no active substance.


The result of a statistical test which tells us whether the results of a study are likely to be due to chance and would not be confirmed if the study was repeated. All p-values are between 0 and 1; the most reliable studies have p-values very close to 0. A p-value of 0.001 means that there is a 1 in 1000 probability that the results are due to chance and do not reflect a real difference. A p-value of 0.05 means there is a 1 in 20 probability that the results are due to chance. When a p-value is 0.05 or below, the result is considered to be ‘statistically significant’. Confidence intervals give similar information to p-values but are easier to interpret. 


Moist layer of tissue lining the body’s openings, including the genital/urinary and anal tracts, the gut and the respiratory tract.

statistical significance

Statistical tests are used to judge whether the results of a study could be due to chance and would not be confirmed if the study was repeated. If result is probably not due to chance, the results are ‘statistically significant’. 


How well something works (in a research study). See also ‘effectiveness’.

Another study published this month found similar, though stronger effects in a study of monkeys who became infected with HIV despite taking pre-exposure prophylaxis (PrEP).

The RV144 vaccine trial was the first-ever HIV vaccine efficacy trial to report a positive result. However the efficacy observed was only 31% and hovered close to the boundary of statistical insignificance.

A subsequent hunt for the changes in the immune system that might have protected individuals from infection found very few differences between recipients of vaccine and placebo, but it did find two. Firstly, vaccine recipients had higher levels of antibodies to two specific parts of the HIV gp120 envelope protein, the V2 and V3 loops. Secondly, vaccine recipients had lower levels of a ubiquitous broad-spectrum antibody called immunoglobulin A (igA), which is secreted in large amounts by the mucous membranes. It is thought that too much igA may have interfered with a process called antibody-dependent cellular cytotoxicity (ADCC) whereby other broad-spectrum antibodies stimulated by the vaccine induce the anti-HIV activation of other parts of the immune system.

The current analysis looked at the course of HIV infection in 114 trial participants who were infected with HIV over a period of 5.5 years, and compared what happened to the 49 vaccine recipients with what happened to the 65 placebo recipients. This total was a ‘modified intent-to-treat’ group, consisting of any trial participant who was infected with HIV after receiving the first shot of vaccine or placebo, but excluding six subjects who were infected before receiving a shot. The study also looked at a ‘per-protocol’ group of 90 participants, 39 of them receiving vaccine, who received all four shots of the vaccine or placebo and who were infected al least six months after receiving their first shot.

It used a composite primary endpoint of disease progression which combined time to initiation of antiretroviral therapy (ART) or AIDS-defining illness or the CD4 count falling below 350 cells/mm3; 80% of people reached this endpoint because of this last criterion.

Possible long-term reductions in viral load

Overall, over the whole time period observed, there was no difference between vaccine and placebo recipients in the proportion of people who reached one of these endpoints, nor in the trajectory of individual CD4 counts or off-treatment viral loads.

However the researchers did start to see differences over time starting about 4.5 years after infection. Whereas in placebo recipients viral load steadily increased from about 4.2 logs to 5.0 logs (c. 16,000 to 100,000 copies/ml) over 5.5 years, in  the final year of observation the viral load in vaccine recipients dropped to about 3.8 logs (6300 copies/ml). Alongside this, average CD4 counts, which declined over the 5.5 year period from over 500 to under 300 cells/mm3 in non-ART treated placebo recipients, started to increase at 4.5 years in non-ART treated vaccine recipients and reached about 450 cells/mm3.  

Bu this time, however, a large proportion of participants had dropped out the study, partly because many had started ART. As a result although these analyses started with 109 participants, the last three six-monthly measurements of viral load and CD4 count only included 34, 22 and 12 participants respectively, which reduced the statistical power of these findings (p=0.2). As a result it’s hard to say more than this study hints at long-term efficacy, though whether this is because the vaccine slowed progression, or on the contrary speeded progression in the people whose viral loads declined fastest, is hard to say.

Immediate reductions in seminal viral load

More significant and also unexpected was the finding that seminal viral load in men who received the vaccine was significantly lower than in men who received placebo. In this case the viral loads were measured at the first study visit after HIV infection, generally around the time of peak viral load.

In the modified intent-to-treat group, mean seminal viral load was 1.75 logs (56 copies/ml) in vaccine recipients and 2.55 logs (355 copies/ml) in placebo recipients, and this was statistically significant (p=0.04). The proportion of men with undetectable seminal viral loads was 57% in vaccine recipients and 28% in placebo recipients. In the per-protocol analysis the difference was not so pronounced and lost statistical significance (p=0.14) but still 54% of vaccine recipients had an undetectable seminal viral load compared with 36% of placebo recipients.     

The viral load was also measured in cervicovaginal lavage fluid (CVL) is women. This was not statistically different (p=0.68) but the researchers say that this is probably because, unlike semen, CVL is a diluted fluid with very different levels of virus in it from sample to sample.  

In a separate editorial, Anna-Lena Spetz and Francesca Chiodi from Sweden’s Karolinska Institutet point out that the discovery of lower viral load in semen was almost accidental – it was not part of the original study – and that the mucosal effects of HIV vaccines remain very poorly understood. They speculate that whereas stimulation of high levels of igA in the blood increased the likelihood of infection, in the mucosa – where it is in a different, dimeric or ‘paired’ from that has very different immune effects – it could be helping to control the viral load in genital fluids.

They call for a renewed concentration on the analysis of mucosal immunity in HIV vaccine studies.


Rerks-Ngarm S et al. Extended evaluation of the virologic, immunologic and clinical course of volunteers who acquired HIV infection in a phase III vaccine trial of ALVAC-HIV and AIDSVAX® B/E. Journal of Infectious Diseases, early online edition, July 2012. See abstract here.

Spetz AL and Chiodi F. Reduction of HIV-1 load in semen during follow-up study of the RV144 vaccine trial boosts the interest for novel correlates of immune protection in genital mucosa. Journal of Infectious Diseases, early online edition, July 2012.