I'm not sure what to make of the report that scientists in Boston, California and the CDC in Atlanta have made monoclonal antibodies that protect mice against many different flu subtypes. Monoclonal antibodies are antibodies made by the descendants of a single immune cell (that is a single clone, hence monoclonal). Thus unlike natural antibodies, these are also monospecific, i.e., they are directed against one specific target. Our natural immune system "sees" a protein on the surface of the virus called hemagglutinin (HA), of which there are 16 broad subtypes and many, many variations within each subtype. Humans are infected with subtypes H1, H2 and H3, although several hundred cases of the H5 subtype (H5N1, "bird flu") have been reported and occasionally there are infections with some other subtypes (H7 and H9). Natural antibodies against a strain give us natural immunity. Vaccines aren't antibodies but they cause the body to produce antibodies, among them antibodies against the HA protein. The reason we need a different flu vaccine every season is that even within a subtype, e.g., H1N1 (the N1 is another protein that characterizes the subtype), there is enough variation (called strains) that a vaccine effective against one isn't effective against another. But the HA protein looks like a long stemmed flower and most or all of the variation is on the flower end, not the stem.
What the researchers did with monoclonal antibodies is make antibodies in the test tube directed at the stem part of the HA protein. It turns out that the stem doesn't vary much at all, so antibodies against it seem to attack at least 10 of the 16 subtypes and within those subtypes all the strains. Even though the virus can still enter the host cell, by glomming onto the stem, once inside it can't "uncoat" and release its genetic material (RNA) into the cell. The virus can't make copies of itself until its genetic material gets access to the protein making machinery of the host cell and the antibody hindered stalk prevents this.
When mice are protected with the new antibodies and then challenged with a lethal flu virus like H5N1 they survive. So at least under these conditions the antibodies "work." But if I had a nickel for every flu vaccine or drug that worked in mice but not in humans I'd be able to start a stimulus plan of my own. So for starters we don't know if it would similarly protect humans. For some reason we don't make much of this antibody naturally, so maybe there's a reason. Even if it did, it has to be administered intravenously, so this isn't much use in a pandemic or even an outbreak. It could conceivably save the lives of some severely ill patients (although we don't know the latest time it can be given to work since we don't even know if it will work given at any time) or be used to protect some critical workers (the ethical and triage problems here are even more severe than usual), but that's about it.
So in itself this isn't much of a therapeutic or preventive medicine advance. The question is whether there is gold in the underlying science. For example, maybe we can make a vaccine that targets the stalk, was effective and thus could be used against lots of flu viruses and wouldn't need to be changed every year. That's certainly worth exploring but we are very far from knowing the answer to that question.
Meanwhile, the producers of these monoclonals have already patented them. Since I presume they did this with tax supported research, if it works they want me to pay for it again. I'm not too happy about that. They could easily have put it in the public domain (where it belongs), thus lowering the cost. If it works and no company wants to develop it and make it, it can be made at public expense and will be far cheaper than if Big Pharma controlled it. If you want in object lesson in what that's like, take a look at another expensive monoclonal antibody, Avastin.
This is another in a string of stories about finding a "universal" or at least broad spectrum flu vaccine. Except this one is farther from that goal than most. It's mainly a tribute to the PR machines of Dana Farber Cancer Center in Boston and the other institutions that collaborated in the work.
You can find Sui et al., "Structural and functional bases for broad-spectrum neutralization of avian and human influenza A virusek" Nature Structural & Molecular Biology Published online: 22 February 2009 | doi:10.1038/nsmb.1566 here.
Your take on the paper is much the same as mine but I have a question I have asked many times before without any ever seeming to know the answer.
Why are mAbs so expensive?
There is the usually R&D and regulatory overhead but airlift fermentation seems a relatively cheap scale-up. If we could find a way to make this technology it is so generic we can make nAbs for anything and by using a variable cocktail keep moving the goalposts in a pathogens attempt to evolve away. Perhaps a 22nd century technology?
that's the question
maybe we can make a vaccine that targets the stalk, was effective and thus could be used against lots of flu viruses and wouldn't need to be changed every year. That's certainly worth exploring but we are very far from knowing the answer to that question.
don't know the answer.
let's call it "promising".
"If it works and no company wants to develop it and make it, it can be made at public expense and will be far cheaper than if Big Pharma controlled it."
Except that wouldn't happen, because unless a pharma company is willing to pay for the additional pre-clinical testing, and if that is successful the much more expensive clinical trials we'll never know if it works. Is there really an realistic chance that if these antibodies hadn't been patented their development would have been taken further with public or charity money?
It might be better to fix the system before criticizing researchers working within it.
For the record the researchers in this case did found that the antibodies protected mice when administered 3 days after infection, if this is also seen in human patients it could be usefully given to vulnerable people who have been exposed during an outbreak.
The basic science of the paper is interesting but thats about it. I do applaud the PR dept at Harvard for running with the "Universal Vaccine" line. But they have probably hurt the authors chances at getting this funded next time. Any scientist that has seen the newspaper reports on this know its all hype and that will hit them hard when they are being reviewed at NIH.
Another reason why the public doesn't support scientists. Every week there is a new "cure" for something being pledged on the Today show or the evening news, and then nothing happens with that amazing finding. Atleast not that they will ever hear about.
The hype may be good for the PR hacks at Harvard but for everyone else it only hurts us.
considering the different volume of man compared to mice, how much antibody would be needed per patient? milligrams?
Revere-I have a question since it comes up often. To date, has there ever been much that wasnt patented by universities, researchers etc. when research that might turn a buck comes down the road?
I know your feelings on the matter but Salk and Sabin both apparently were on grants back in the fifties that came up with polio vax. They patented it.
Its a fine line we tread of course but the grants processes has been this way since they invented them I think. I personally think that they should rebate it by percentage to the government each year for say 10 or so years. Wouldnt work in flu vax of course as it changes, but it might work in other areas.
Hey, has anyone heard from Jonny Singleton in the last month? Revere?
Randy: Quite a bit. I think I blogged something about Corey's Tamiflu synthesis. I am funded by NIH. None of our work is patented and is licensed, if at all, under a Creative Commons license. I edit an Open Access journal. Copyright held by the author and unlimited copying permitted. We practice what we preach. You paid for it. It's yours.
The problems is that whether infected with hyper-virulent pandemic strains like H5N1 or seasonal influenza, you don't know which ones are going to be most seriously affected...and as a result, the tissue damage occurs so rapidly that by the time you pick it up and start treatment, the damage is irreversible.
So either you have to have enough to treat everyone who is presented with influenza or you have to selectively try to preventatively treat the most favoured...either way there are significant limitations.
Therefore, in my opinion, specifically during a pandemic,an effective treatment must be cheap,in an oral pill format, easily manufactured and be given to everyone who presents with influenza signs.
These comments answer some of the questions I had about the limitations of this proposed treatment. I have a couple other questions, if anyone has the time to answer them. If the MAb proved to be effective in humans, how often would the MAb need to be administered (I have no clue how often treatments like Herceptin are given) to maintain protection? I guess what I'm asking is what's the approx half-life of MAbs when administered to humans? And, is there any indication that (passive?) immunity from a MAb could stimulate active immunity and memory? Lastly, I wonder if it's already known whether the "stalk" is the slightest bit antigenic...I would think every trick in the book has been used to turn the conserved region into an effective vaccine. Thanks!
OK--found an answer to my frequency of administration and half-life questions. For Herceptin, the half-life is anywhere from approx 2 to 12 days depending on the amount administered; the frequency of administration ranges from weekly to every three weeks. That would be a lot of treatment for a pandemic...far short of the "pop a pill" approach Tom DVM advocates.
Brian: It wouldn't stimulate immunity except to the antibody itself, which is not what you want. So this is totally passive. We don't even know if it will work in humans. Its importance IMO has more to do with the biology than its promise as a therapy or prophylactic. Just my opinion and others may differ.
Salk did not patent his vaccine.
Thanks. I agree. I was initially excited about the possibility of a generalized anti-flu therapy but the applicable scenarios seem limited. As you say, it would be far better to be able to use the identified target region to develop a vaccine(s). Maybe I'm wrong, but I would think such strategies have been attempted previously without success. Perhaps, though, there is a way to induce a human immune response that mimics the passive immunity of the MAbs observed in mice. I will read the primary article at some point, but at least in the press releases, no mention was made of attempting to devise such a vaccine as a next step.
Couple of points:
1. Mabs can most definitely be made as an inhalation formulation. In particular for respiratory illnesses, this has been very well-researched.
There are quite a few other hits in PubMed about the formulation issues involved in formulating biologics for inhalation, but they are not often insurmountable if you're reasonably creative.
2. Re: JJackson's question, why are Mabs so expensive: Pharmaceutical pricing, including biologics, is an arcane and dark art. The most accurate description I ever heard from a pharma executive was, "It's like Hilton asking Mariott how much they should charge for a room." Yes, airlift fermentation is well-known, but there are several other technical constraints, starting with the super-dooper clean room the things have to be built in. However fixed equipment is actually not the largest expense anymore, as we now have disposable reactors that are way cheaper. No, the biggest technical expense is the especially careful aseptic technique all the technicians must use at all times, and a LOT of work goes into media formulations, which must be animal product-free for CJD/TSE reasons. Normally mammalian cells like to grow in fetal calf serum, and for drugs this is not possible, so it's a lot of work to find a good combination of other stuff.
And they also consider how much the market will bear, as evil as that is.
3. Re: Paul's point about needing a pharma to pay for clinical testing, not necessarily. University of California has an excellent clinical trial system, and NIH often runs clinical trials themselves. IP that is developed past Phase II is WAY WAY WAY more valuable than mere proof of concept IP, and a pharma will pay several times more for a drug that has passed Phase II than they would for something only demonstrated in mice.
4. Re: Brian's question, "And, is there any indication that (passive?) immunity from a MAb could stimulate active immunity and memory?" Look up "epitope spreading" on PubMed. The short answer is, yes, quite probably. You have to be rather cautious about dosing to some extent for this reason. It's sort of weird, there's a lot of debate on this particular issue--is immunity or tolerability due to epitope spreading or particle size or...? No one knows for sure, we just think odd things.
Hope that helps!
Revere, are you funded 100% or does the university pick up part of your tab? Here is my understanding of the law. IF its funded 100% and I mean 100% then yes you are right, its the property of the people. On the other hand if state funding is involved then the people of the state of XXXX have a partial claim.
This was explored in the 60/70's when funding for war research was involved (better killing through smart bombs) when several anti-war groups sued. They lost. I believe that the idea was that the "benefit" was more from the defense of the nation as a whole rather than the single part. The government funds for a "result" rather than "property."
I believe its just the law Revere, not that you are wrong. We know that there is a huge chasm between the two. In fact the law is probably specific in that. Have to check that one out a bit more.
Info... seems there is BIG cluster jumping up down in Indon....
I didn't say it was illegal. I just think it's wrong and I won't do it. I am not funded 100% for my research because I do a lot of other things: I teach, I advise, I am on lots of committees, I represent the University various places, I write more grants and papers. But I am paid to do research and I don't ask for even more pay for what i produce when I do it. I am not alone. Most scientists don't patent things or charge for their products. And they shouldn't. It's not a question of the law.
As for a cluster in Indonesia, we'll have to see what is happening. I hope not.
OK, perhaps there's more publically funded clinical trials being done that I'd thought.
In their paper the authors do refer to using the stem epitope in vaccine design, so it's certainly something that they are considering (probably worth checking the patent).
"It is not surprising that many
viruses are highly adept at keeping their most crucial (and conserved)
determinants of pathogenesis cryptic, in which case subunit-based
vaccines, using properly presented fragments of F10 or F10-like
epitopes, may offer distinct advantages over whole virusâbased
approaches to induce broad spectrum nAbs in vivo."
I have to say that I'm a little disappointed by the reaction here to this work, which isn't by a long stretch trivial or unpromising, in fact compared to a lot of the "gene for X" stories that run in the press it's pretty darn promising. I'm not saying that noer hyping results is not a problem, but sometimes it seems to me that when it comes to publicising your research you're pretty much damned if you do and damned if you don't.
Paul: No one said this wasn't good work. But this paper got a lot of press on the basis that a universal flu vaccine was "on the horizon" (I'm quoting a headline). If so, the horizon is a very long way away nor is it clear it is really there. Natural antibodies to this epitope are very rare, suggesting that it may not be accessible in humans. There are many basic studies and even some Phase I trials for universal vaccines much closer than this. Much, much closer. There is a lot of good work being done on flu, these days. The "hook" that was used to get this one a lot of publicity seems pretty weak. It's interesting work. It's good work. It is just not that big a deal in the landscape. It's another piece of the puzzle. Not a break through, or no more than a ton of other stuff.
So if you are disappointed that we didn't accord it a special level of interest, then that's a difference of opinion about this work compared to a lot of other good work.
Lora thank you very much for taking the time to answer my question.
I thought the paper was very tight and a staggering amount of work must have gone into it producing it. I just wonder, if the technology could be shown to work economically, is it best to try and attack a conserved, but inaccessible, pocket where inaccessibility will probably need more mAbs. Will this not just push the virus into finding a solution by selecting for viable stalk changes? Why not opt for targeting exposed sites on the head and achieve both wide serotype/strain coverage, and avoid resistance, by using a cocktail which is regularly varied?
Vaccines, antivirals, mAbs all any of them do is either speed up humoral immune response or slow down viral replication. If the hostâs immune system is severely compromised they can not do the whole job for it. The question is which technology can most rapidly, and affordably, give the most hosts the best edge should battle be joined with a virulent pandemic flu?
Well it's true that there are other "universal" vaccines that are further down the pipeline, the Acambis and Oxford University vaccines spring to mind as being at Phase I or beyond, but they may yet fail so I'm glad to see that other avenues being explored. These other vaccines have also got quite a lot of press over the past few years, especially the Acambis vaccine.
I agree that the press coverage was a little over enthusiastic, but when is it not like that? I'd love to see the day when the "this is cool...this is cool...oops it might be shit...no it's cool again" approach to news items on science changes but I'm not optimistic that journalism in general will ever make such an exception for science.
I suppose that I'm just glad to see good research being reported, I've seen plenty of good research that the press have inexplicibly missed in the past. My point would be that while this study has been over-hyped (though not as badly as many) it is still worthy of being reported in the news.
JJackson "I just wonder, if the technology could be shown to work economically, is it best to try and attack a conserved, but inaccessible, pocket where inaccessibility will probably need more mAbs. Will this not just push the virus into finding a solution by selecting for viable stalk changes?"
They looked at this in the paper, in the section "Prospects for immune escape", and found that there are some fairly severe structural constraints that should slow down evolution of stalk mutants that can evade the mAbs, and then showed that in vitro mutants to evading mAbs that targeted the stalk didn't evolve while mutants evading mAbs that target the head did. This was only a preliminary experiment but at least it points in the right direction.
Paul: I think we are in substantial agreement. My point was essentially to ask what was so special about this good paper compared to a lot of other good papers (some of which we cover here). I think the answer is mainly they had a good press office. That doesn't take away from how good their work is or whether it will be significant in some way far down the line, but it does go to how it was reported. The PR flacks have something to do with that and we scientists go along with it (I have quite a bit of personal experience along that line).
While reading the immune escape section I was not sure that the 3 passages with VN04 giving a head mutation but no stalk mutations was strong enough to base much on. My other thought was to do with mutation mechanisms. The authors talked about drift but said nothing about recombination. If this is a significant mechanism in flu viral evolution then it will quickly find viable combinations of SNPs under selective pressure if the SNPs already exist. At present there are plenty in the head but fewer in the stalk but if you start targeting the stalk that will change. As they say if the resistance emerged you could change targets but that brings me back to the option of a cocktail which is regularly varied so as not to provide selective pressure in any one direction, a randomised combination therapy, targeting the most accessible surface antigens. All this is based on the assumption that the reason most B-cell product is targeting the head, not stalk, is that you need fewer antibodies because the target is more accessible.
Is there really an realistic chance that if these antibodies hadn't been patented their development would have been taken further with public or charity money?
Yes, the research and development would likely have been picked up by someone somewhere else in the world where the government considers medical R&D a priority instead of something it funds to keep up cultural appearances ("good old American know-how" and all that). Just off the top of my head, both Canada and Israel do a lot of good medical R&D, basically entirely on public money. The world doesn't end at the US border, and if someone in the US doesn't invent something, chances are good someone else will...
"I just wonder, if the technology could be shown to work economically, is it best to try and attack a conserved, but inaccessible, pocket where inaccessibility will probably need more mAbs."
Just because it's not readily accessible doesn't mean you need more to get there. It's a spatial problem, not a concentration problem. There are smaller antibodies being generated now from camelids and sharks that can fit in significantly smaller spaces than the traditional mAb and maintain immunogenic qualities. Or, we can fuss around joining designed peptides and Fc fragments together.
"Will this not just push the virus into finding a solution by selecting for viable stalk changes?"
Yes, but that usually takes some time.
"Why not opt for targeting exposed sites on the head and achieve both wide serotype/strain coverage, and avoid resistance, by using a cocktail which is regularly varied?"
Well, that's essentially what we're doing now with vaccines, using sentinel populations to figure out which is the most appropriate cocktail. It sorta kinda works most of the time, but the bottom line is that the head of the molecule is extremely variable--so variable we can't get any good common epitopes, and predicting how it should be varied isn't super-accurate. Imagine having to predict today's weather in your town based on the weather 2000 miles away. You can hazard a pretty good guess on satellite images, but it's not going to be perfect, know what I mean?
One pharma I know of is currently working on multi-branched engineered mAbs that have two different epitopes/molecule, but putting the things together and getting the host cells to produce them is apparently a giant pain in the ass.
Thanks to everyone for the fantastic discussion. Lora--thanks for the specific answer to my inquiry...I will indeed read up on "epitope spreading". I look forward to keeping informed about the world of science by way of this awesome blog.
Thanks again Lora