Sunday, August 07, 2005

Bird flu vaccine? Taking the (very) long view

While No News can be Good News, this weekend's stories about a new human bird flu vaccine is Little News that at least isn't Bad News. But what, really, is the News?

It is interesting there is a story at all. Apparently the New York Times got an exclusive from NIAID Director Tony Fauci over the weekend and the other MSM outlets are playing catch-up, citing the NYT story. But why now? The results are only partial, preliminary and very hard to interpret. It appears that the Administration has finally awakened to the bird flu threat--as a PR problem. With stories all over the media about an impending pandemic, many noting the shocking lack of preparation on the ground, the spin machine was put into action, with Fauci again the water-carrier (it is sad to see a good scientists brought so low).

So what do we have? Let's start with the headlines. They are all over the place:
First story was in last night's NYT, by Lawrence Altman, their long-time medical writer and a physician: To Create a Vaccine, a Virus Is Tweaked, Then Replanted

It was picked up by Reuters and appeared in the LA Times like this: Bird Flu Vaccine Takes Step Forward
Operating off the same Reuters story, other headlines are a little less restrained, but still defensible:
Thanh Nien (Vietnam), Sydney Morning Herald (Australia): Tests show promise for bird flu vaccine in humans
These are modest, and by and large, accurate descriptions of what we know. But things start to go downhill from here.
Health Central: U.S. Researchers Develop Avian Flu Vaccine

Herald Sun, Courier-Mail, Sunday Tasmanian (Australia): Bird flu vaccine effective in humans
Then come the truly dreadful:
Baltimore Sun: Test of avian flu vaccine succeeds
WaPo (Rob Stein): Vaccine Appears to Ward Off Bird Flu

Not to be outdone, the NYT and Altman return on Sunday morning with an expanded story (that is significantly different in some respects from the night before): Avian Flu Vaccine Called Effective in Human Testing
Now to content. From last night's NYT story:
Dr. Fauci's institute received the first batch of 8,000 doses from Sanofi-Pasteur in the spring. In April, doctors injected the vaccine into 452 healthy volunteers at three centers: The University of Maryland in Baltimore; the University of Rochester in New York; and the University of California, Los Angeles.

The recipients received one of four different doses of the vaccine and the results confirmed the scientists' prediction that a larger dose of A(H5N1) virus than used in standard influenza vaccines would be needed to produce the strongest immune responses, Dr. Fauci said.

Evidence that the vaccine could be protective came from analysis of 113 of the 452 participants. These findings have been discussed only among the doctors who conducted the studies and the scientists at his institute, Dr. Fauci said.

Though the numbers were small, they were enough to show scientifically that the vaccine elicited a strong immune reaction - a critical point of the study.
Yes, obviously a critical point, since if the vaccine didn't raise antibodies, Game Over. But in itself it doesn't tell us much (maybe more is known but Fauci didn't reveal any of it to Altman). In last night's story, Fauci makes another assertion:
The only way scientists can prove the effectiveness of the vaccine is if an epidemic of the A(H5N1) strain occurs among people. Ethics prohibit deliberately infecting people for experimental purposes, in part because that could lead to the very pandemic scientists are trying to prevent.

Under such circumstances, scientists must show that as the dose of the vaccine is progressively increased, the immune response increases proportionately.

Dr. Fauci said he was "encouraged" about the findings because "many other experimental vaccines have produced flat dose-response curves," meaning that the vaccines were ineffective because they did not stimulate the necessary immune response.

"What confirms the fact that you have an immunologically potent vaccine is when you get a good dose-response curve, and unquestionably we got one with this vaccine," he said.
This is rather curious, since last fall, when it looked like there would be a significant shortfall in flu vaccine availability, the New England Journal reported two studies that showed the vaccine supply could be stretched by using a reduced dose delivered into the skin instead of into muscle. The results were that 40% of the dose produced the same antibody response (i.e., a flat dose - response), showing that half the dose was still effective. Researchers went on to say:
Though striking, these studies had some serious limitations. First, both studies combined included a total of only 338 people, a relatively small number. Second, blood tests were used to measure immune response, rather than actual cases of the flu.
So in this case we have reports from 113 of 452 participants, less than a third of those in studies considered to have serious limitations because of small numbers. Moreover the dose - response argument has been reversed. Here, lack of a dose - response was considered significant.

As if to underline this interpretive difficulty, the dose - response argument is missing from this morning's NYT story and all the other stories as far as I can see. Instead, today's WaPo has this:
The vaccine was administered in four dosages to 452 healthy adults beginning in April. They received a booster shot four weeks later. Blood samples showed the level of antibodies produced by the subjects' immune systems rose in direct proportion to the dose they received, with the strongest dose producing what is considered an adequate response, Fauci said. [NB: no mention that the results are from only 113 of the 452].

"That put the antibody level in the range that you would predict would be protective. Obviously, you never know until you test it in the field. But generally you can pretty much gauge that if you get the antibody level up to a certain value, that pretty much guarantees protection," Fauci said.
This is rather different. Instead of saying the existence of a dose - response shows the vaccine is effective, it says that lower doses wouldn't be effective because they didn't provide an adequate response. Nor do we know from this whether the antibody response is actually protective, only that it reached the level of "what is considered an adequate response." But we don't know if these are neutralizing antibodies or some other kind of less effective immune response (e.g., T-cell dependent).

That's just for starters. Henry Niman also raises the crucial issue whether the virus has now changed sufficiently that the vaccine will no longer be effective against the current strain. He notes that there were sufficient changes between 1997 and 2004 that an earlier vaccine was no longer effective, and that the strain currently circulating in southeast Asia is quite close genetically to the one used in the vaccine. So that's good. But he goes on to say that there are changes in the Qinghai virus (western China) and presumably the Siberian and Kazakh versions headed toward Europe, that are as far from the 2004 vaccine strain as the latter was from the 1997 strain. So Fauci's declaration: "It's good news, We have a vaccine" is at best misleading and at worst a lie.

Fauci does go on to point out that even if the vaccine would be effective against a pandemic H5N1, there wouldn't be enough of it if the shit hit the fan in the next six months to a year. The WaPo story quotes Dr. Michael Osterholm on the obvious:
"These are very important studies," said Michael T. Osterholm of the University of Minnesota. "This is confirmation of what we hoped would be the case. We all had anticipated that two doses were going to work. If it didn't, we would be in trouble."

But Osterholm said the world would still be woefully unprepared if a pandemic occurred. The capability to produce and distribute a large amount of vaccine quickly, for example, is far from adequate.

"That's the Achilles' heel of our public health response," Osterholm said. "These data are encouraging and supportive, but in the end, a lot of people would be left unvaccinated in the event of a pandemic."
The US is now ordering 2 million doses. I assume that means enough for 1 million people (2 shots each). That only leaves another 296 million doses needed for the US and about six and half billion for the rest of our fellow creatures.

So this is a little news, but mainly the spin machine has gone into action. Still, I guess its progress. You have to take the Longview. The very, very, very Longview.