Im really a Debbie Downer when it comes to HIV-1 vaccines/'cures'/etc.
But I swear to gawd, I was just about to write a post on the strategy employed by the recent HIV-1 vaccine trial, without knowing this vaccine trial was even going on, cause I just read a really cool paper that did something very similar:
Cross-clade neutralizing antibodies against HIV-1 induced in rabbits by focusing the immune response on a neutralizing epitope
The basic idea of this paper, and the basic idea of this recent vaccine trial is the same.
1. Broad Prime - DNA
First, you 'prime' the immune response with something really general via DNA. So, you either use a gene gun (paper) or canarypox (vaccine trial) to deliver the DNA for say, the gp120 portion of envelope. When this DNA is in your cells, it will make gp120 just like if HIV-1 had actually infected those cells. This means that the way the protein folds, how many/what kinds of sugars are added, is exactly what HIV-1 would normally do.
What you get out the end is a really generic antibody response to a specific kind of HIV-1 gp120 (whatever kind of DNA you used). These antibodies wont necessarily neutralize the virus. They just see it as non-self. But your immune system is already primed.
This by itself isnt particularly helpful as a vaccine, as the antibodies you make wont necessarily recognize anything but a match to what you were immunized with (there are tons of different subtypes of HIV-1, all with envelops with different structures), and wont necessarily neutralize the virus if you do get exposed to a near perfect match.
2. Narrow Boost - Protein
Using lab-made envelope proteins as vaccines hasnt worked out at all. You have to make modifications to the genes to make envelope soluble instead of membrane bound, more modifications to keep them as trimers instead of monomers, the bacteria you use to make the modified proteins doesnt fold right, or doesnt add sugar right... It just hasnt worked real well on its own for generating protective HIV-1 vaccines.
But! If you use these proteins as a booster-shot-- to take the broad immune response you got in #1, and use the normal B-cell somatic hypermutation/evolution process to hone that broad response into numerous specific neutralizing responses, that might work!
What they did in this paper is generate a broad immune response with HIV-1 Subtype C gp120 DNA, then gave rabbits booster-shots with peptides specific for the Variable-3 region of different subtypes of HIV-1 envelopes. The Variable-3 region is a great target for neutralizing antibodies, as this region must interact with your cells co-receptors to initiate infection. Know how I always say you cant force a person to create a specific antibody short of gene therapy? These folks got rabbits to at least make antibodies to the right region, thus increased the odds that the antibodies made would be neutralizing. Some of the rabbits made antibodies that could neutralize viruses they had never seen before! They didnt make a whole lot of antibody, and not all of the rabbits had the same kinds of responses, but this paper gave me hope that we might be able to figure out a 'good enough for now to help some people' strategy!
Turns out this vaccine trial was on the same page, kinda. For their boost, they used another failed vaccine, AIDSVAX, with was composed of gp120 proteins. These groups independently did the same thing-- DNA prime, protein boost, good news.
The problem with this vaccine is, it might just work in specific parts of the world. What DNA you chose, what proteins you chose, all depends on which subtypes of HIV-1 were the most prevalent in that region. This is highlighted by the fact that not everyone who got the vaccine was protected... only people who made the right response and were exposed to the right virus were helped (30% reduction in risk).
BUT! Its a GREAT start. This plan, this is doable. And now we can work on figuring out the best mix that can help the most people!
- Log in to post comments
YES! This is what my boss says all the time; apparently some people insist that prime-boost broadens the response, but to us it sure looks like narrowing (makes it really hard to study the sub-dominate epitopes).
It's so awesome to have one work at all (8cough* STEP trial *cough*) but to also have it show that prime-boost works, well, that's just super cool!
I swear I winced when I read this.
And here we were, pouring cold water on Phil Plait's happy-dance, and then you come and tell us that it's not such a bad idea after all.
Boo. I guess I need to make up for that.
Here we go! Pup/a> Py/a> Pictures.
He really was quite a darling with those brats pulling his tail and grabbing his nuts. And the owner had him properly leashed.
For good measure a picture of my cat being naughty. (Vey! Treif is mir!)
(No, that's not lube above my bed.)
I just don't buy the 30% figure. The actual numbers were 51 out of 8200 that got HIV while on the treatment, and 74 out of 8200 that got it while on placebo.
If you had injected all 16,400 people with HIV, the results would be meaningful, but that's not what happened (for obvious reasons). There's no way you can say that if they hadn't got the trial vaccine, ~20 more people in the first group would have contracted HIV
With those numbers and an unknown a priori infection rate, the difference is just noise.
Great post! You are awesome when it comes to explaining the logic behind the science!
Thanks,
JTD
RE: Thanny(#5)
Yes you can. You can do that because they had a freaking huge group to start with, so that minimizes the statistical error. That's why large studies are more authoritative than small studies. Your math is wrong.
I bet this vaccine will work for around two years and then simply not work anymore at all as HIV evolves it's way out of our defenses again. Seriously, it's like those old Road Runner cartoons. We're Wile E. Coyote and HIV is the Road Runner. We come up with elaborate and seemingly foolproof traps to catch that HIV, but it simply changes the rules at the moment of glory and speeds off into the distance.
Just got pointed to this blog; now working at the NCI in one of the labs that apparently was helping run the trial, and it's good that someone out there is explaining just how HARD it is to create a vaccine for this virus, and the difficulties in getting a trial like this set up when dealing with a disease like AIDS!
There certainly seems to be a disconnect between the actual prevention data and the hype that surrounded the announcement of these vaccine trials.
For there to be a real advance, there'd need to be something in the blood (e.g., truly neutralizing, or the new catch phrase, 'sterilizing' antibodies) that would truly protective. The absence of any such findings makes me deeply suspicious.
Also, I was surprised to see Fauci praise this work so uncritically. Possibly he's under a lot of pressure from the Obama administration to show some results.
Actually, the statistical significance level isn't very good. A similar difference will occur in slightly less than 5% of random trials just by chance. The result is encouraging but tentative.
Disclaimer: IANAS and I don't play one on TV. While the above does not contain any deliberately counterfactual claims, it is provided for entertainment purposes only and must not be construed as actual statistical advice. For indoor and outdoor use only.
JustaTech-- At least the way they did it in this paper, its totally 'narrowing' the response. Instead of forming Ab to superfluous epitopes, the rabbits made Ab to V3, thus were more likely to be neutralizing. We will have to wait to see what happened in the vaccine trial!
DC-- What do you mean? Read my HIV-1 posts, I really had little hope for a viable preventative HIV-1 vaccine with todays technology. But then this study AND this paper came out at the same time, and gave me some hope. *shrug*
Sili-- This vaccine trial on its own, I would be hesitant. But this vaccine trial + this independent paper that did the same thing (though technically 'better') makes me happy.
Also, that pup looks just like Arnie-man! Arnie also doesnt mind tail-pulling. It means hes getting attention, so he doesnt care :)
Jonathan-- Read the paper (its open access). Their strategy is smarter than the vaccine strategy (but the vaccine people were doing what they could with vaccines that were already approved)-- They tried to get neutralizing antibodies to a rather conserved and necessary portion of HIV-1 gp120, V3. Assuming they perfect this strategy and send it to trials, it would basically be taking an unlikely event (HIV-1 transmission) and make it even less likely (transmission of an HIV-1 with a different 'key' to our locks, when there is no selective pressure for this new key in the original host). Theoretically once it really got chugging, it would be an issue, but I think it will buy us more time than 2 years :)
ArchTeryx-- *high-five!*
LG-- Of course there is 'something in the blood' of individuals who got the vaccine. The vaccine induced antibodies to HIV-1. Read the paper I linked to (again, its open access). They did look for broadly neutralizing antibodies, and they found them. The scientists in the vaccine trials have 8200 patients to screen (as opposed to 50 rabbits), and I doubt they arent looking at individual antibody responses to all subtypes present in Thailand, the magnitude of this response, as well as CTL response in every patient. This is stock standard.
Youre also confused on terminology. "Sterilizing immunity" is the theoretical goal of HIV-1 vaccine research. You want a vaccine that prevents infection, because once HIV-1 gets a foothold, it doesnt leave (perfectly sterilizing vaccines are not necessary for other viruses). Personally, I would be happy with a "therapeutic vaccine" for HIV-1-- one that wouldnt 'cure' the patient, but allow them to live a normal happy life without any drugs, but hell if someone can figure out a viable preventative vaccine, YAY!
The other word you are looking for is "broadly neutralizing antibodies". This also is not a 'new catch phrase'. Every paper in HIV-1 research that generates HIV-1 antibodies checks them to see if they are broadly neutralizing. This is another goal of HIV-1 vaccine research, as of course, youd like to have one shot that protected people against the most subtypes of HIV-1 possible.
Also, the scientific community, including the HIV-1 research community, doesnt 'report to Obama'. We report to each other. And if you read this blog for 30 fucking seconds you would see that I am immensely more critical of research within my field than 'Obama'. Get your head out of your ass.
How reasonable is the proposition of using Phylogenetic HIV-1 mapping to identify the most prevalent subtypes? How geography specific are the subtypes?
I am thinking about profiles for primary and secondary resource allocation. France goes after its most prevalent subtype and the most prevalent sub-saharan subtype and so on until you develop therapeutic vaccines to contend with as many major branches as possible.
Does (theoretically) a therapeutic vaccine effect transmission rates?
Because I've been watching your roller-coaster-of-hope and seeing you write about hope again did the deja vu all over again thing to me.
Nothing remotely rational about it. Just me taking things personally at one remove.