An unrevealed second analysis of the results from the initially lauded RV 144 HIV vaccine trial failed to show a statistical benefit over placebo, according to a ScienceInsider blog entry authored by longtime AIDS journalist Jon Cohen. 

Lead scientists from the RV 144 trial conducted in Thailand announced last week that there was a modest reduction in the HIV infection rate among those who received the vaccine. The results were heralded as a triumph by the general media, but were greeted with measured caution by some vaccine experts.

The skeptical experts pointed to the extremely narrow margin of efficacy. While the vaccine reduced the risk of HIV infection by 30 percent, only a handful of volunteers in the massive 16,000-patient study—51 patients in the vaccine group compared with 74 patients in the placebo group—actually became infected during the study, lending to a difference that was just barely statistically significant. 

There were also concerns about the fact that the vaccine failed to suppress viral loads in those who did become infected, with experts arguing that a vaccine capable of sparking the immune system to prevent HIV infection would help control HIV replication if infection did occur.

However, some expressed hope that while the vaccine itself would never be approved with such a low rate of success, it might point the way to other more successful preventive vaccines.

Now Cohen reports that a second analysis showed a trend toward benefit in those who were vaccinated, but failed to demonstrate a statistically significant difference. This means the difference was small enough that it could have occurred by chance.

The initial report was based on an intent-to-treat analysis that included all trial participants, including people who may not have received all of the vaccinations or who may have dropped out early. The unrevealed second analysis was per protocol, and only included people who received all their vaccinations on schedule. 

According to Cohen, the per-protocol analysis failed to demonstrate a statistically significant difference between the two groups of patients studied. The vaccine efficacy also dropped slightly—a confusing finding, given that a tested vaccine often performs better in per-protocol analyses, given that it eliminates people who weren't properly vaccinated and might not have developed an appropriate immune response to the virus. 

This new analysis, said the researchers interviewed by Cohen, casts some doubt on the possibility that the vaccine approach employed in the RV 144 study will have any positive influence on HIV vaccine development.