It has taken a while for me to get to this, because I have been waiting for the fallout / comment storm to settle a bit, so that I could get a good clear objective view.
And that is…that the recent Thai trial showed hints of promise, but was largely a failure. At least it did no harm…!
First things first: Nature News’ Elie Dolgin had this to say on 24th September:
Vaccine protects against HIV virus [!!! sic - I had something to say about this, see Comments]
The largest HIV vaccine trial to date has shown moderate success at preventing infection by the virus.
The experimental vaccine — a combination of two older shots that failed to work on their own — reduced the risk of someone contracting HIV by nearly a third. Scientists, however, are still scratching their heads as to how the double-shot approach blocks the virus….
The US$119 million study involved more than 16,000 HIV-negative men and women from Thailand aged 18–30. The trial was launched in October 2003, conducted by the Thai health ministry and sponsored by the US Army Surgeon General. It tested a two-shot infection-fighting strategy using drugs made by Sanofi-Pasteur of Lyon, France, and VaxGen of Brisbane, Australia. Over the course of 24 weeks, participants received four doses of a ‘primer’ vaccine — a disabled bird virus [canarypox - Ed] containing synthetic versions of three HIV genes [ALVAC, subtype B env, gag and pro - Ed] — and two doses of a ‘booster’, which consisted of a protein called gp120 [AIDSVAX subtypes B/E - Ed], a major component of HIV’s outer coat. [see here for link describing the components]. Clinicians tested for HIV infection every 6 months for 3 years….
Many HIV vaccine experts had previously criticized the approach as a waste of time because each of the vaccine components had a poor track record. The primer, called ALVAC, conferred little to no immune protection in multiple early-phase clinical trials, and the booster, called AIDSVAX, had flopped twice in high-profile, large-scale trials.
And here’s a thing: a high profile crew of scientists had, in 2004, written an open letter to Science magazine, stating in no uncertain terms that they thought the trial ought to be stopped. In their words:
“Concerns are expressed by a group of AIDS researchers about the U.S. government’s plans to conduct a phase III trial of a combination HIV-1 vaccine in Thailand despite the cancellation of a trial of a very similar combination vaccine in the U.S.A. last year. One of the vaccine components, recombinant monomeric gp120, has already been shown to be ineffective in phase III trials in Thailand and the United States; the other component, a recombinant canarypox vector, is also poorly immunogenic. The scientific rationale that has been offered for the new trial in Thailand is considered by the authors to be weak.”
And now we have Dan Barouch – not a signatory to the 2004 letter, I note – quoted by Dolgin as saying:
“I don’t think anybody knows why this worked the way it did,” says Dan Barouch, an immunologist at the Beth Israel Deaconess Medical Center in Boston, Massachusetts. “It’s the largest step forward that’s ever occurred in the HIV-vaccine field, but there’s a tremendous amount of more work that will need to be done.”
But exactly what is it that people are hailing as a breakthrough here? Dolgin again:
The two-pronged vaccine did not affect the amount of virus circulating in the blood of those who acquired HIV during the study. But it did show a protective effect — vaccinated individuals were 31% less likely to become infected. New infections occurred in 74 of the 8,198 people who received dummy shots, but only 51 of the 8,197 in the vaccine group [my emphasis - Ed], the researchers, led by Supachai Rerks-Ngarm of the Thai Ministry of Public Health’s Department of Disease Control, found.
Dorian McIlroy, a regular contributor to Viroblogy, had this to say on the 24th September in an email to me:
I just read the news story about the ALVAC/AIDSVAX trial results in Thailand. From the numbers on this press release:
The significance level is extremely slim. For example, if you go to this site
and type in the numbers you will find that p=0.048 by Fisher’s exact test.
If one more person in the vaccine arm had been infected, or if one less person in the placebo arm had been infected, the difference between the groups would not have been significant. [my emphasis - Ed]
None of the experts (Wayne Koff, Frances Gotch, for example) interviewed in different news stories seems to have noticed just how borderline the “statistical significance” really is, and seem to have accepted the bottom-line 30% reduction figure.
Ah well, I just thought I had to tell someone….
Lecturer in Microbiology and Cell Biology,
University of Nantes
Others have also picked up on this – which shows just how desperately slim the hope is. However, it does remain – although (pleasingly…B-) the pundits have been thrown into a state of confusion, as some strongly-held views have not been vindicated. Another Nature News article – from Erika Check Hayden, on October 1st - has this to say:
As the dust settles from last week’s surprising announcement that an HIV vaccine combination may protect some people from the virus, scientists are talking about what else the vaccine trial might tell them.
On 24 September, leaders of a US$119-million study of 16,000 people in Thailand reported that the combination of two shots had reduced the risk of HIV infection by one-third …. Now, the vaccine’s fate will depend on whether scientists can figure out its ‘correlate of protection’ — in other words, what caused it to partially protect some people from HIV. The key does not seem to be anything scientists had predicted, which has led to much head-scratching — and some unease.
“It’s a humbling thing, because for the first time we got a positive signal and it doesn’t jump out at us as being related to any classical parameters you would expect from a successful vaccine,” says Anthony Fauci, head of the National Institute of Allergy and Infectious Diseases in Bethesda, Maryland, which supported the trial. “That tells us maybe we were not measuring the right thing.” [my emphasis - Ed]
Amen, brother Tony…a clearer proof of Clarke’s First Law I have yet to see.
So what ARE the things that fall out from this? First, I would suspect, is that the value of a heterologous prime-boost combination seems to have been shown, albeit weakly. Second, the use of a poxvirus vaccine in particular in combination with a protein may be a good thing to chase. I note here that the South Africa / US joint Phase I human trial currently underway with the SAAVI DNA / SAAVI MVA (=modified vaccinia virus Ankara, a poxvirus) was almost certainly considerably more immunogenic in non-humanprimates than either of the ALVAC / AIDSVAX vaccines, so the gleam of hope may soon get brighter.
Third: take heed of Arthur C Clarke before you go sticking your neck out making predictions about HIV vaccines…B-)