Thursday, March 30, 2006

Another vaccine trial, another disappointment

Another vaccine trial for H5N1. Another disappointment. The newly published vaccine trial, using a conventional vaccine prepared from a strain isolated in 2004 in Vietnam, only provided partial protection and that only at the highest doses (90 µg and 45 µg of protein antigen). The results just appeared in the New England Journal of Medicine and are available online gratis (thank you, NEJM; that's the way it should be done for H5N1 articles). Here is some explnation and a few of the dreary details.

When looking for a flu virus, what your immune system primarily "sees" are hemagglutinin and neuraminidase protein spikes on the outside of the virus (when the Flu Wiki is back up on its new servers you can read all about the virus there). Unfortunately your immune system has never seen one of the H5 subtype before, so it does not react right away to protect you. To help it along, we use a vaccine to show the immune system what the H5 protein looks like so that if it sees it during a natural infection it can react promptly. Something similar is done with the seasonal flu vaccines, which contain H3 subtype proteins. Although our bodies have seen H3 before (it is the prevalent flu virus since 1968), the H3 protein changes a little every year so it doesn't look exactly the same. It's like a person who changes clothes, sometimes wearing jeans, sometimes a suit. The immune system needs some help with the change, but more like a prompt than a whole new description. With H5, however, the clothes are so different (imagine a pirate or a bear costume) that the immune system doesn't recognize it at all until the infection is well underway. Hence the vaccine lesson for your immune system.

Conventional influenza vaccines are grown in fertilized chicken eggs and they grow in the amniotic sac of the developing chick. There is a problem with the Highly Pathogenic Avian Influenza H5N1 virus, however, as it is lethal to chickens, so some molecular surgery is first performed on the viral vaccine seed to eliminate the feature that causes the high pathogenicity (the polybasic amino acids at the cleavage site). Then the H5 and N1 genes of the Vietnam isolate are combined with other genes from a standard vaccine strain. When grown to high amounts ("high titer") in eggs, you have the relatively harmless persona of a standard vaccine virus wearing the clothes of H5N1.

The live virus is concentrated in a centrifuge and inactivated with formaldehyde (inactivated means it can no longer replicate). A series of purification steps then separates the viral antigen (the part you want the immune system to recognize) and it is put up in vials at different strengths. Strength here corresponds to how many micrograms (µg) of the antigen are in a single injection. The doses used were zero (placebo), 7.5, 15, 45 and 90 µg, administered twice in injections 28 days apart. For comparison, the amount of antigen in the usual seasonal flu vaccine is 15 µg, given once.

How do you know if the vaccine works? We can't vaccinate people and then infect them with H5N1 to see if they are protected. Instead the researchers used some data from the 1997 Hong Kong outbreak suggesting antibody titers of 1:40 might be protective. This, then, was selected as the level at which the vaccine was considered to protect people when their blood was assayed for anti-H5 antigen. Current seasonal vaccines achieve efficacy rates of about 70% against circulating H3 strains. Thus the current vaccine does not protect everyone, yet it is still a relatively high proportion. But the new vaccine didn't make that benchmark for H5.

The results, in fact, confirmed what had been found in other trials (see here and here for some of our many other posts on vaccines). A little over half the 102 vaccinees who received the highest (90 µg) dose achieved titers of 1:40 or greater. At the 45 µg dose about 45% were protected, with 22% and 9% being the corresponding levels for the lower doses. Thus only about half the subjects achieved protection when given a vaccine dose twelve times what is currently used. Bear in mind that we have a very small production capacity, the result of years of disinterest by Big Pharma in products that weren't obscenely profitable, although they were still quite reasonably profitable. What this means practically is that this vaccine is not currently practically feasible as a way to protect the population. It might be used in a highly targeted sub-population, but even there might not be that effective.

Despite the high dose, side effects seemed to be minimal, mainly pain and tenderness at the injection site. However the ability of this smallish study to detect serious side effects was relatively modest. It had an 88% chance of picking up three or more serious events as long as the true rate of serious events was at least 5%. Five percent is a pretty high serious adverse event rate when you are vaccinating tens of millions of people. So the jury on the safety of this vaccine is still out. So far, however, we don't see warning flags, although our vision is far from perfect.

NIH's Dr. Tony Fauci acknowledged the disappointing outcome, even as he tried to put the best face on it:
''We have a vaccine we know can spur an immune response, albeit at very high doses. We're going in the right direction. The sobering news is we have a long way to go. So it's muted good news,'' said Dr. Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases, which funded the study. (Chicago Tribune)
The lead researcher also put on a brave face, calling on a time worn cliché to take the heat for bad news:
"Every journey starts with a first step,'' said Dr. John Treanor, director of Rochester's Vaccine and Treatment Evaluation Unit and the overall leader of the project. ''We actually have a product. We're not there yet, but we're making progress."
The big question, of course, is whether we are traveling fast enough to get to where we need to before dark. Probably not.

Addendum: It's been pointed out to me that the NEJM article is just the published version of the earlier newspaper accounts (last August, I think). So this is not another disappointing vaccine trial result but the same one as earlier. My error. The paper has additional information, which is what I posted above.