More on the science of the influenza "cytokine storm"

If you had a prowler at your house you'd call the police. You'd want them to come. But what if they sent a SWAT Team, surrounded the house and blasted it from all sides. Not good. That seems to be something like the scenario for a response to a class of virulent influenza viruses. They trip the alarm and the army descends and levels the house. The prowler is taken care of. So are you. The phenomenon has acquired the name "cytokine storm," although a better description might be immune system dysregulation. Your immune system has a lot of powerful weapons to protect you, but like a police force you want them used lawfully and appropriately. This week scientists at the St. Judes flu group in Memphis published an intriguing paper in the Proceedings of the National Academy of Sciences with the unintriguing title, "TNF/iNOS-producing dendritic cells are the necessary evil of lethal influenza virus infection" (Aldridge et al., PNAS; hat tip to Dave Fedson, with whom I enjoyed a long and informative lunch yesterday).

It is becoming increasingly clear that some of the nastiest effects of virulent flu strains like the 1918 H1N1 or the bird flu H5N1 are related to the violent and uncontrolled immediate response of the immune system, a response which destroys lung tissue by runaway inflammation. Helpless clinicians have tried to damp down the immune response in bird flu patients by using steroids which depress all parts of the immune system. It hasn't worked. The immune system is a very complex apparatus with many moving parts. Those parts have to be coordinated and regulated. Using steroids is like laying off the police force. It isn't going to protect people from attacks from pathogens like the flu virus. The intricate choreography of the immune system, like the cruder one of a police response, is based on signaling and communication between its different parts. Most of that signaling is done by chemical messengers called chemokines. When one kind of immune cell is ready to recruit others to help, it releases a chemical signal. When that signal is received (by another chemical, called a receptor), it activates a new set of biological events. And so on. I'd love to lay out the whole intricate system for you, but I can't. No one can. At the moment it's like a giant jig saw puzzle where we know some of the pieces and how they fit together, but not always where they sit in the Big Picture and there are lots of pieces still lying in the puzzle box. But we are making progress and the PNAS paper by Aldridge et al. is a step to putting some more in place.

The paper is quite technical, but the main ideas aren't too difficult. The immune system operates on two different time scales. There is an immediate response that is generic in character. Sound the alarm, there is a microbial intruder on the premises. Strike at it with generic weapons. That's called the innate response. But then there is a second, later response. The immune system has figured out what kind of intruder it is. Send in the specialized units designed for just that person. That's the adaptive response. Some of the adaptive response is via proteins made by immune cells that circulate in the blood. These antibodies are what vaccines are designed to elicit. Once you are familiar with different kinds of intruders you can react to them more quickly. The immune system learns how to react by seeing specific examples and training antibody making cells to respond quickly. The immune system also has its own Special Forces cells that go out to kill virally infected cells. These are a kind of T-cell, another one of the cells of the immune system. So what kinds of response do we see in the lung when its cells are infected with influenza virus? That's one of the important questions scientists are trying to puzzle out.

The Aldridge paper identifies a particular kind of immune cell, called a dendritic cell, as being centrally involved in the process. When a flu virus infects a lung cell it sounds an immune system alarm. One of the first things that happens is a generic immune cell called a macrophage, like a cop on the lung beat, sounds an alarm by making a protein, MCP-1 (monocyte chemotactic protein-1). MCP-1 combines with CCR=2 (chemotactic chemokine receptor-2) which calls to the area another cell called a monocyte. CCR-2 also induces the monocyte to perform a more special function: produce two more proteins, Tumor Necrosis Factor-alpha (TNFalpha) and inducible nitric oxide synthase (iNOS). This may sound complicated, but I'm simplifying a great deal. It's much more complicated than I'm letting on.

The monocytes have now changed into a cell called a dendritic cell (DC), or more precisely a special subset of dendritic cells called Tip-DCs (TNF-iNOS-producing DCs). There are many other kinds of immune cells in the mix: natural killer (NK) cells, neutrophils, other kinds of DCs, macrophages. Aldridge et al. found, however, that when a mouse is infected with the kind of virulent flu virus causing a runaway inflammatory response, the only kind of immune cell that ramps up compared to others is the Tip-DC. When a mouse was infected with a less virulent flu virus, there was no unusual increase in Tip-DC but with highly virulent strains, Tip-DC accumulated in the lungs. The Tip-DCs were also making a lot of TNFalpha and nitric oxide, damaging the lungs. The SWAT Team had arrived and was firing at will. This suggested the obvious thing: get rid of the Tip-DC response and you would protect against the runaway immune response.

Unfortunately there is nothing obvious about how the immune system operates. When Aldridge et al. repeated the experiments with mutant mice that couldn't make CCR-2, they also did badly, dying at the same rate as the mice that had an intact CCR-2 response. As expected, there wasn't an accumulation of Tip-DC in the lungs, but the mice died anyway. So more is involved. The "more" involves, at least in part, further activities of the Tip-DC cells. Dendritic cells also function in the later phase of the immune response. They are "antigen presenting" cells, essentially cells that take the pathogen and broadcast its description to other cells in the immune system, cells like the killer T cells. These T cells are like the specialized sniper team that comes in after the culprit is identified. Only the ones trained for that flu virus are called up. But if there are no Tip-DC cells to perform the description and broadcast function, there is a weak or ineffective T cell response. Aldridge et al. first established that the Tip-DCs were actually functioning as the flu virus antigen presenting cell (their method was technically clever, but the details would take us too far from the main story). Then they showed that the Tip-CD were necessary for the killer T cells to function properly in clearing virus. Here is their summary for where they had arrived at this point (APC means "antigen presenting cell," a cell that functions as the describer/broadcaster of the invader):

So far we have shown that the number of tipDCs is significantly elevated in mice infected with HP [highly pathogenic] influenza A viruses, that these tipDCs function locally in the respiratory tract as APCs, and that they are required for the full realization of protective CD8+ T cell-mediated immunity [i.e., functioning killer T cells]. In addition, their complete absence from the lungs of CCR2â/â mice is associated with severe disease, indicating that tipDC ablation [elimination] is not a viable therapeutic option. (Aldridge et al., PNAS)

In other words, when it comes to Tip-DC, you can't live with them and you can't live without them (in terms of infection with virulent flu virus). What to do? The Goldilocks Principle: Not too much, not too little. Just right.

They reasoned that this was yet another example of the balance needed in any complicated system. You want the police to come, but you don't want them to send the army. By searching the literature, they found a drug that affects CCR-2 but doesn't completely knock it out. The drug belongs to an interesting class of compounds called peroxisome proliferator-activated receptors (PPARs), of which there are three (alpha, beta and gamma). They were interested in the gamma subtype, one of which is a compound called pioglitazone. Another compound of this class is rosiglitazone, marketed under the tradename Avandia, as a treatment for type II diabetes. PPARs have many interesting biological effects and we have been studying them from the environmental angle and this connection with dendritic cells was new to me and extremely interesting. But back to the flu story. By pre-treating the mice with pioglitazone Aldridge et al. were able to reduce significantly the lethality and weight loss related to infection with the virulent strains of flu virus. The mice weren't completely protected, but the mortality went from 90% to 50%. There was also a reduction in the inflammatory response, with the most evident reason being a reduction in accumulation of Tip-DCs in the lungs of the mice.

Before you run out and buy Avandia, you should know that there are some significant questions about risk of heart attack for those taking it long term for diabetes. On the other hand, balancing that risk against death from a pandemic flu virus suggests it is worthwhile to consider as part of our arsenal should a pandemic develop. Unfortunately, I fear it will be put on the shelf with other possibly effective and low cost approaches, like statins. Antivirals and vaccines seem to be the only weapons that the infectious disease folks can imagine. Of course there is much to learn yet before this would make sense as a policy. But if a pandemic arrives before we've figured out how to handle it, statins and PPARs are something people would (or should) try in the urgency of the moment. Right now there is no effective treatment for a runaway response to flu.

Admittedly this is well out of the conventional approach to flu. The infectious disease community has been very slow to look in new directions. Maybe this paper by one of the premier flu groups in the world will change that. But it will also require the people at St. Judes to push their new findings. I hope they will.

It's not the only thing I wish they'd do. While they're at it, maybe they could release the flu sequences they're sitting on until they can publish papers. In that respect they are no different than the nutcase Indonesian Health Minister everyone likes to rant about (including us). They're not alone. I'll include CDC and Mt. Sinai and probably other elite flu groups as well. Just to be fair.

More like this

Interluken 9 (IL-9) was being studied by Medimmune (MEDI 528) regarding the body's opposition to the cytokine storm. Does your research deal with IL-9 and the part it appears to play, defensively?

Crap Revere.

This is one of the best explained shots you have ever written.

By M. Randolph Kruger (not verified) on 03 Apr 2009 #permalink

Please allow me to get to the point.My father is a physician and I personally have spent over 3000 hours of personal study on AI over the last 12 years.This study has included an extremely wide spectrum of the real-time implications of pandemic,social upheaval,and the vast array of ongoing scientific daily virological publishings.I would STRONGLY urge deeper investigation into the antiviral adhesion properties of lemon-based compounds.There may not be as much money in it as Avandia.But this area of study must be expanded.One part of nature can destroy what another part of nature can create.And it won't help the researchers and drug companies to continue novel approaches if the end result is that they don't live through a pandemic long enough to spend the money they made trying to stop it.PLEASE RESEARCH THIS ARENA OF STUDY.A partial answer to AI,imho,is understanding that looking too deep will lead to living less life.The answer is to be DISCOVERED...not created.A small microbe with a large appetite will be satiated and sent packing when faced with its NATURAL archenemy.Thank you..

For what its worth...which is is the opinion of one veterinarian.

First of all, I don't think the cytokine storm is the result of triggering an individually hypersensitive chemical response leading to a 'cytokine storm'.

We already know that H5N1 and other pandemic potential avian viruses have the ability to replicate at many, many times that of seasonal influenza.

Therefore, I believe that the first occurence in these infection is a seeding of a virus that replicates massively and infects a large number of cells inside and outside the lungs before the immune system has a chance to respond.

The cytokine response is then due to a normal immune response to too large a number of infected cells. This response results in an increasing circulation of dead cellular material that then stimulates the 'killing' response from the immune system. Therefore, the immune response in itself, is not that abnormal.

Secondly steroids have been around longer than most antibiotics and are used in a number of ways to counter shock and too temporarily adjust the immune system...

...I hope that a specific drug may be found in the future that works better...but at the moment there is no effective alternative.

Based on my field experience, I have found that relatively high dose steroid injections (those that do not have to be liver-activated, whether intramuscular or intravenous, modulate the immune response rather than suppress it.

The goal is to buy enough time to allow inherent sytems to successfully counteract the infection.

Like with headaches the pain is much less with earlier treatments. The treatment protocols must be started before the 'cytokine storm' gets fully in gear...obviously, this is a strategic problem with pandemic influenza...but even in this case, there are reasonable alternatives.

With any kind of infection, whether parasitic, bacterial or viral...steroid treatments after seven days are immunosuppressive and probably after five days.

Therefore, the way to handle it is to give large measured doses too early in series with non-steroidal anti-inflammatories with the goal to be weaned off the steroids by day five and certainly by day seven...

...often by modulating the response successfully, allows the steroids to be removed by day two or three.

Lastly, some individuals with seasonal influenza unfortunately die much in the same way as those with pandemic type influenza's.

It is my conclusion that this is not due to a hyper-reactive immune sytems primarilly but rather because these individuals have a chemical accelerant molecule that allows seasonal influenza's to replicate like a pandemic influenza...

...The end goal is the same...use harmless treatments before the cytokine storm gets going.

Thank you.. very excellent.

The viruses probable predilection to endothelial (blood vessel lining cells) and the resultant puching of holes and collapse of circulating blood volume can only be counteracted by steroids.

If you attempt massive intravenous fluids without the steroids having first sealed the holes in the blood vessel...then animal-human (yes, humans are animals) will die in short order.

For that reason alone, there are no alternatives to steroids in the first hours of the infection.



"The Goldilocks Principle"
*gasp* You've been going to the same innate immunity conferences I have, haven't you?!

So, if the Tip-DCs are functioning locally, were the NO + TNF-alpha serum levels up?
I'm wondering if this may explain differences in virulence between different strains of other microbes (Plasmodium, in particular), but if we're only looking at serum levels, would we miss it?
(granted, I doubt it's lung Tip-DCs, but maybe there are Tip-DCs that can cross a compromised blood-brain barrier and cause cerebral malaria? Or, given the NO production, they might even play an active role in breaking down the barrier...I should know more about this)

"their method was technically clever, but the details would take us too far from the main story"
awwwww. *pouts* more story time? PLEASE?!

hat tip to Dave Fedson, with whom I enjoyed a long and informative lunch yesterday

Glad you finally found each other! And, most excellent explanation, thanks!

becca: The Goldilocks Principle is used in other fields, too, so I didn't pick it up at an immunology conference (I'm not an immunologist but an epidemiologist). Describing the method would have been fun but a diversion and I don't have the paper in front of me to do it here (or the time; I'm about to run out the door).

Hi Susan: Yes, David and I had a terrific time. He provided me with the Aldridge paper. I also owe you an email. I am semi-overwhelmed and it deserved some thought instead of just a pro-forma response, hence my lack of any response. I hope I'll get to it soon (either that or I will forget, just like I forget everything else these days; the pleasures of age).

It seems to me that what causes the cytokine storm in the first place is the unfamiliarity of the immune system with Hemagglutinin 5. So why not vaccinate with H5 antigens to familiarize the immune system, possibly eliminating some or most of the cytokine storm? Granted, it will not stop the actual influenza strain, but it might also permit natural immune function to more rapidly respond to it as well.

I gather some animal testing has been done in this area, and it was determined that two exposures were required for the immune system to develop familiarity with H5. If anyone has insights into this area, please let us know.

By L.R. Davis (not verified) on 03 Apr 2009 #permalink

LR Davis: Alas, that isn't the answer. In the Aldridge et al. paper H1N1 (PR8 strain with seasonal H1 and N1 substituted for the usual lab versions) elicited the immune dysregulation quite efficiently. There are also H5 subtypes (e.g., H5N2) that seem fairly benign in human infections, as do some of the H7s and H9s. So it isn't immunological naivete that is the route of the virulence. There are several features of the virulent strains that have been discussed and it appears there are several routes to virulence but at the moment we cannot predict what makes this happen (the same goes for transmissibility). this is a very enigmatic and wily virus.

You'll find more here and elsewhere on this blog (we've done perhaps a thousand posts on bird flu):

This is a very interesting article and the possibility of reducing the human death rate due to cytokine storm to 50% from 90% with early treatment with glitizones is quite intriguing.

I share Tom's view that the storm follows the widespread infection of the lung with influenza virus. When the immune system is naive to the infecting agent, as is the case with H5N1, then its response to it is delayed. This delay permits the virus to reproduce at will within this tissue for several days until the immune system snap back occurs; an event we postulate today as cytokine storm.

Cytokines are terribly powerful biologic agents. They are incredibly destructive. The immune system employs them to destroy bacterial, viral, and fungal invaders of the body tissues. They also use them to eliminate aberrant cells that contain cancerous mutations.

When these agents are released by the immune cells there is an inevitable "friendly fire" effect that results in the death of nearby healthy cells. In essence, cytokine storm is the immune system equivalent of a military friendly fire incident that results in the death of the patient.

This study clearly suggests that the glitazone class of drugs used for type II diabetes has a profound effect upon the immune response of mice to infection with the type A H5N1 influenza virus. As Revere states, this is a very interesting discovery.

Why did the researchers test this compound in the first place is an interesting question?

Another area of interest would be to see if there is any efficacy in the addition of one of the many TNF inhibitors for the amelioration of cytokine storm. A study that compared the use of glitazones with and without a TNF inhibitor VS placebo in mice would be very interesting.

The Doctor

By The Doctor (not verified) on 03 Apr 2009 #permalink

The Doctor: When the immune system is naive to the infecting agent, as is the case with H5N1, then its response to it is delayed.

This is not correct. We are talking about the innate immune system which requires no prior exposure to the agent. The adaptive immune system comes later, in this case involving killer T cells.

Tom DVM... What is a good example of a non liver activated steroid? I use very low dose prednisone (5 mg) for a week following a week long surge pack for my knee that has a nice chunk of Russian origin metal buried deep in the kneecap. Would I be better off ceasing the use of it if it does come Revere/Tom? I can live with pain, but I cant live with my lungs full of gook. From previous posts I thought this kinda made it worse?

Seems to me that the UT statin thing thats being tested on sexed up H3 that Revere touts might be a big gun, but its all about the study money of course and its a cheaply funded one right now I think by the state.

I plan to get this well after everyone else has because I will be in a CBW suit and full mask for the duration, only shedding it to decon and sleep/eat in protected facilities. So, I would rather have a very mutated to the weaker side than the initial first wave or attenuate it by every means possible. I think they'll be trying most everything even with the use of untested vax after a month or two. But can anyone yet point a finger at something and say that this "XXX" incontrovertibly works on this shit?

By M. Randolph Kruger (not verified) on 03 Apr 2009 #permalink

Steroids have been shown worthless or worse for H5N1. I wouldn't touch them. Tom claims on the basis of his experience with animals they are the way to go. If that's good enough evidence for you, take them. I wouldn't.


One of the mysteries of the human innate immune system is how it is able to respond to an antigen it has never encountered before.

What might have made H1N1 in 1918 so virulent was because of its unique evolutionary path, one that did not involve mammals. H5N1 seems to be following the same path.

The human trials of H5N1 vaccines have been hampered by the fact that when killed virus is used, the immune response has been poor compared with similarly produced vaccines for H1N1 or H2N3.

For a protective antibody response to be obtained using killed H5N1 in humans has required the use of both significantly higher doses of the antigen, administered twice and the inclusion of an immune stimulant.

IMO, this suggests that the human response to live H5N1 is delayed due to a relatively slow recognition by the immune system of this unfamiliar pathogen. This does not mean that the innate immune system ignores the infection but that its response is sufficiently delayed. Since the reproductive rate of H5N1 is much faster than that of the activated immune cells, this fact may set the stage for cytokine storm.

The Doctor

By The Doctor (not verified) on 03 Apr 2009 #permalink

No, the innate immune system doesn't have to recognize it. It responds generically, and that is the case here. It is like the splinter in your foot. You don't need splinter vaccine. The apparent lack of immunogenicity for some H5 vaccines (not all) is the adaptive immune response part of the equation. It says nothing about the innate response. Indeed, all the data on the innate response suggests that the response is unusually vigorous for the virulent flu viruses.


Thanks for the opportunity as always.

I think we can agree that the only thing we know for sure is that we don't know much...either about immune system functioning or more importantly for this discussion influenza.

Yes, it is possible that a malfunctioning immune response is the reason for the cytokine storm...but then again there is another, just as reasonable explanation that I think needs to be discussed as well.

I happen to believe that you have the 'cart before the horse' because I don't think a malfunctioning immune system explains clusters of H5N1 infections within familly members or explain why some infected with seasonal influenza have the exact same course of infection of H5N1.

In discussion with Dr. Woodson, you make the statement that the immune response is inate requiring no priming at all. Well, there must be a time delay involved here. Are you saying that those with influenza present in a full blown cytokine storm. That doesn't seem to be the way with either H5N1 or seasonal influenza...

...Also, we must remember that the infection has been incubating for some time before the patient becomes sick enough to seek out medical help.

Dr. Webster has explained that H5N1 replicates at many powers of ten that of seasonal influenza and we know that some seasonal influenza patients follow the exact same course of events invariably to death within 2-3 days.

So scientifically, it is highly likely that the virus has a relatively long incubation period to hyper-replicate.

I understand that normal seasonal influenza does not extend beyond the lungs but in those unfortunately rare cases of seasonal influenza and avian influenza, it seems that the virus has escaped systemically. Something has caused this and I expect it is hyper-replication as explained by Dr. Webster.

Of course it is important to have a control for the cytokine storm but most probably the cytokine store, the immune response is secondary and not all that abnormal.

The point of the immune system is to remove infected cells. If it happens that functional threshold of organs have been overcome before we turn the immune system on...then the immune system will remove all of those affected cells...and this is self reinforcing because of the cellular junk that coats endothelial and lymphatic cells etc. etc. that may produce an autoimmune type response.

Now you made a statement above that, in my opinion and studies, is pharmacologically completely untrue. You said that...

..."Helpless clinicians have tried to damp down the immune response in bird flu patients by using steroids which depress all parts of the immune system. It hasn't worked. The immune system is a very complex apparatus with many moving parts. Those parts have to be coordinated and regulated. Using steroids is like laying off the police force. It isn't going to protect people from attacks from pathogens like the flu virus."

Steroids do not shut down the immune system. They modulate the immune system. There is quite a difference between the two.

Your argument seems to be that it hasn't worked so it won't. I guess then a violin is a beautiful instrument no matter who's hands it is in.

Steroids, probably an unfortunate use of the word because they are not at all similar to anabollic steroids, have saved many lives since they were introduced in the 1950's including my own after a rare auto-immune response to an experimental sanofi rabies vaccine that ended up not being released due to abnormal reactions.

The fact is that your can't massive levels of steroids for weeks like they did when treating SARS. At those doses for that length of time, you will destroy the immune system with obvious results and sequelae. H5N1 is the same creature. You can't just throw shit at the barn and hope some of it sticks. Steroids, to be effective, must be used strategicall and judicously.

With respect to acute viral infections it seems that intramuscular treatments are far more effective than intravenous treatments. Also, the dosages need to be within reason (unlike those used in SARS) and must be given early, before what you call the cyotkine storm hits full effect. Permanent organ damage will occur within 24 hrs. if the immune system is not modulated.

Also, it seems that prednisolone (that does not require liver activation) is more effective. I had no luck at all with other steroids including prednisone.

You have said that statins etc are effective. The statin story has been around for a couple of years. So then why have they not been successful...we are still losing over 60% of those infected with H5N1.I could use the same argument you have used about steroids in that they won't work because they haven't.

I would suggest that you not throw the 'baby out with the bathwater'. You must know what a blessing steroids have been in the past five decades...and you must know that in the wrong hands all medications are suspect...

...and most importantly, I think you might want to reconsider that blanket statement about the cyokine storm being inate in light of the fact that influenza incubates for an extended period of time.

I am not in love with steroids. I am thankful that the protocol I used worked and saved many farmer's livelyhood. If I had used them like they were used in SARS, I would have had a hundred percent death rate...and that would not be the fault of a trusted, and fully explored medication since introduction in the 1950's.

They are also cheap as are statins.

I would say to everyone that if statins are preventable, then we should all be on statins and forget the pandemic.

Unfortunately, treatments that sound to good to be true...usually are.

Hyper replication is the only way you can explain familial clusters and seasonal influenza killing young people. That makes the immune response both a pretty much normal and secondary event.

There must be factor X, a chemical molecule that takes super charges the replication rate of seasonal influenza and super charges the already hypervirulent H5N1. The premary effect is allowing the virus to escape the lungs and become systemic attacking liver kidey and endothelial cells etc.

Some famillies may carry this factor explaining the clusters.


Chronic use of prednisone for arthritis can make the arthritis much worse. That is one of the problems with this medication. It makes you feel good so that you do things you maybe shouldn't. Also chronic usage makes you appear fat due to fluid retention.

You are obviously a highly intelligent google prednisone and arthritis for side-effects of better, talk to your doctor...

...and by the way, don't take advice from a veterinarian without discussing it with a healthcare worker first.

By the way, I'm glad that you got back okay from Afghanistan. Hope you met my fellow Canadians while you were there.

Yeah, thanks Tom I built a deep well there (centcom new feed-no pictures of me thank god) and was up all night visiting the locals if you know what I mean miles away. The Canucks were and are always a good lot to be with.

Hard nosed shit heads if someone tries to kill them...They take it personally. They can hang in my hooch anytime they want. Sides, they had cheap Johnny Walker. That makes them my friends automatically.

Affie-poo with the Canucks was pretty fun.Contracted for a bit after a general that I knew was looking for a mountain goat. You saw my handiwork on CBN if you looked and listened close. I was out and then bingo three days later........It only made the print in the US. You get what you pay for...I got paid and they got what they wanted.

Anyway, yeah my doc only lets me take it for the full week pack, then for an additional 5 dropping back to zero. Its not arthritis, its more like pure inflammation. Dont have the creaks and not a lot of pain either, more like "Dude, whats up with your knee" and you look down and its blown out like a balloon. I did Afghanistan with a full knee brace on with one of those cutouts and it was fine. I get back here, its all good and then three days later for no apparent reason, it just blew up. But its an old thing-30 years worth. They wanted to basically take a hole cutter and take the area out where the metal is on the kneecap. No guarantee they said that it would grow back in. No guarantee that I would be able to use it afterwards... I am into guarantees. The prednisone knocks it back for weeks at a time after taking it. I am good with it.

By M. Randolph Kruger (not verified) on 03 Apr 2009 #permalink

"This is not correct. We are talking about the innate immune system which requires no prior exposure to the agent. The adaptive immune system comes later, in this case involving killer T cells."
Oh oh oh, but what about the memory NK cells?
Sorry, tangential details. But prior exposure does ocassionally matter, even in what are traditionally classified as innate responses, we are learning.

"No, the innate immune system doesn't have to recognize it." Dur. So the Toll-like receptors (and NOD and all the other receptors du jour) are... completely irrelevant?
The innate immune response to a virus is different than to a bacteria, and both are different from a splinter (though how they are similar and how they are different is a very interesting topic in it's own right). I know you're trying to get across that it's the innate response itself-not the adaptive- that is participating in the cytokine storm (though I suppose one could argue about DCs being on the cusp of innate/adaptive and all; but that's semantics). However, I always thought that part of the reason the 1918 flu was so bad from an epidemiological perspective was that there were very few quasi-reactive antibodies floating around in the population (obviously, if you can get flu repeatedly, your antibodies from one strain don't protect fully against the others. But I always thought even a poorly-reactive memory b cell could be called into service and get a robust adaptive response up and running quicker than if there is no such basis to start with? Or am I missunderstanding the mechanics of how positive selection?). Is this incorrect? Is the only reason there was a global pandemic because that variant was so incredibly inflammatory?

Tom DMV- I honestly don't know much about flu. If you asked me about cerebral malaria, however, I know that there are some strains of parasite that can cause massive intense innate immune responses/high inflammatory cytokine levels, despite a low overall parasitemia. In other words, some pathogens develop a set of virulence factors that do not increase replicative potential, but do increase immune response.
That other pathogens develop virulence factors which enable them to replicate faster that also lead to increased inflammation, I do not doubt. But there is no a priori reason to assume that. It could go either way.

Host differences also play a major role. If we knock out a central signaling molecule in parasite sensing pathways (MyD88) in two different mouse strains and infect with the same parasite, we get opposite results. In one mouse strain, making the parasite harder to identify (by removing MyD88) tends to kill the mouse (eventually, accompanied by high parasitemia). In another strain, making the parasite harder to identify (by removing MyD88) tends to protect the mouse, because the mouse is not dying early on with a hyper-inflammatory profile.

So, all in all, I don't see how this statement: "Hyper replication is the only way you can explain familial clusters and seasonal influenza killing young people." makes any sense. You get familial clusters and seasonal malaria killing young people with low parasitemia.


In my opinion...

What signals the fast instinction for other pathogens, genetic instability, is somehow a strength for the influenza virus.

Seasonal influenza harnesses this very successfully. However, pandemic influenza is a whole different animal... in part because of an extraordinary replication rate which further allows it to harness it's genetic instability.

The measure of its ability to harness genetic instability is the potential to spin mutations off in every direction from it's unstable vortex...spinning top so to speak.

Pandemic influenza more resembles animal disease than any human disease experienced in our generations lifetime...

...and the ongoing bad news makes H5N1 one of a kind: a monster that has the capability to do things never imagined let alone observed...and with now the real potential to develop into the equivalent of a biological asteroid at any point in time...and enough genetic material to spin off several unique pandemics over the next thirty years or so.

However at the same time we have the disturbing news of seasonal influenza acting H5N1-like in lethality to young children and teens primarily.

Influenza has a relatively long incubation period. That is why I believe the cytokine storm is secondary to real event in these children which is hyper-kinetic replication rates, H5N1 - H1N1 1918 type replication rates.

If the cytokine storm was purely an immune stimulus type event, then it would have to occur in the incubation period, sort of like anaphylactic shock...and its not.

So the question becomes why is the virus attaining high replication rates becoming systemic outside the lungs and overwhelming the immune system such that a functional threshold of cells in a variety of organs are taken out by the immune system response...such that organ failure occurs and death arrives quickly much the same as was recorded in 1918. Of course the difference is the higher numbers of those infected with pandemic influenza reaching this clinical state.

There are many kinds of genetic defects that are expressed as clincial I believe there is an X factor in these children. That X factor may be made up of more than one simultaneous sub-factor... individual body temperature for instance. We know that young animals have higher body temperatures than adults and we know that avian influenzas like the higher body temperatures of that could be one of the conditions.

However, I also believe there may be an abbherant circulating chemical molecule in these children that 1) stimulates hyper-replication and 2) allows the virus to escape systemically...which could be one and the same other words, could occur only as a result of a high replication rate.

We have had numerous familial clusters and few breakouts of H5N1 into the general public...and as said, we also have these similar but rarelethal cases of seasonal influenza in young people...too few to say that it is due to a certain subtype of a particular strain.

Other diseases and pathogens may be different, but with respect to influenza, the expression of the cytokine storm is the cart while the hyper-replication rate is the horse.

I also believe that the delayed cytokine storm that occurred with SARS was due to circulating cellular junk which of course would be both a functional load on human physiology to remove and a potpourri of foreign chemical molecules never meant to ciruclate in the blood stream to stimulate the cyotkine storm.

Of course all of this doesn't make me just gives another side of the argument to consider.



I also owe you an email. I am semi-overwhelmed and it deserved some thought instead of just a pro-forma response, hence my lack of any response.

So I'm guessing I didn't totally miss the mark, did I?

A timely topic. I'm not a doctor, but I hope you guys figure out a way to stop cytokine storms in the next day or two. ;)

Revere, let me remind you of Kwok-Yung Yuenâs paper in PNAS last year. They also reasoned that if the lethal event of avian flu is a cytokine storm, then anti-inflammatories should help, and they found that triple therapy -- two cytokine inhibitors, plus an antiviral drug -- worked in combination but not alone (unless you treated immediately after infection, not clinically feasible). So, more evidence that an over-exuberant immune response can be at least partly to blame for influenza mortality, though there are lots of caveats attached to the finding.

The paper is
Zheng, B., Chan, K., Lin, Y., Zhao, G., Chan, C., Zhang, H., Chen, H., Wong, S.S., Lau, S.K., Woo, P.C., Chan, K., Jin, D., Yuen, K. (2008). Delayed antiviral plus immunomodulator treatment still reduces mortality in mice infected by high inoculum of influenza A/H5N1 virus. Proceedings of the National Academy of Sciences, 105(23), 8091-8096. DOI: 10.1073/pnas.0711942105

And I talked about it at the time here.

iayork: It should still be there. I didn't remove it. This paper is about mechanisms but doesn't claim it is the only mechanism. This is a very complex dynamical system, or at least one with complex dynamics.

I wasn't trying to correct your point, just support it with a separate paper that also showed that anti-inflammatories have potential in flu treatment.

What if the cytokine storm is not a mistake, but in fact an ancient mammalian strategy for dealing with virulent pandemic flu? I figure that the flu has been around for longer than mammals- and that it has blown through animal and bird populations regularly for millions of years. Maybe when faced with this sort of viral challenge, the response is a sort of "apoptosis" a programmed person (as opposed to cell) death? Maybe by dying quickly and horribly, the rest of the population (group) is protected by (1) being warned that a plague has arrived (2) The person dies quickly and does not circulate amongst the group further spreading the infection?

This is just pure speculation, but what if the cytokine storm is exactly what it is supposed to be?

clark: One thing that is different about humans than other organisms: we have an additional defense and adaptation mechanism, technology. It allows us to live anywhere on the planet. It also allows us to remove certain population dynamic elements like epidemic disease (e.g., smallpox). If we had an effective universal flu vaccine, how would that affect your speculations?

This is an interesting article and post. One of the things that does regulate NFkB (which is the transcription factor that causes the expression of iNOS and other things) is NO. NO inhibits NFkB and reduces the expression of iNOS and so reduces the level of NO that is produced by that iNOS. PPPR-gamma agonists do have effects mediated through NO and NO has similar effects all by itself.

iNOS is not post transcriptionally regulated the way that eNOS and nNOS are regulated, it is pretty much open-loop control, the cells make a certain amount of iNOS and they then make NO levels commensurate with that iNOS level until the iNOS is degraded. If you make too much iNOS, you are SOL until it is degraded.

In effect low basal NO turns up the âgainâ on the immune system by disinhibiting NFkB so more iNOS and more cytokines are produced. Normally I think this is for local control, that where specific tissue compartments have a lower NO level, during a local infection that lower NO level is likely due to the local infection, due to proteolytic cleavage of xanthine oxidoreductase, due to increased metabolic activity and more superoxide, due to superoxide and inflammation from other immune cells. The respiratory burst on immune cell stimulation causes low NO and ratchets the immune system into the âonâ state. Low NO also potentiates mast cell degranulation.

I think that the normal systemic NO from moderate continuous immune system stimulation, as from parasitic worms, may be one of the things that normally modulates the immune system and reduces systemic inflammation and also things like inflammatory bowel disease. Parasitic worms are a great treatment for IBD and Crohnâs. The immune system deviation from being âtoo cleanâ and from not having parasitic worm infestations is thought to be important in allergies and asthma as in the hygiene hypothesis.

Allergies can be thought of (very simplistically) as from too high a âgainâ on the immune system. The immune system (like all important systems in physiology) has automatic gain control. When it isnât doing anything, it turns up the âgainâ until it is doing something. In the absence of an infection, that may be the source of the chronic inflammation that is observed in the metabolic syndrome. The basal NO level is an important part of that automatic gain control.

Turning up the âgainâ on the immune system might be expected to help resolve minor infections, but (I think) would tend to exacerbate chronic inflammatory disorders, maybe more cancer and cause more (and more severe) cytokine storm-type reactions. It is interesting that the epidemiology of the vaccine preventable infectious diseases showed a decrease in the death rate from them prior to the introduction of vaccines. It may be that there was some factor that was increasing the gain on the immune system during the period of time while the death rates from these diseases were decreasing prior to implementation of vaccination (which was virtually completely effective at preventing disease and death from those diseases).

The time constants of the various immune system sub-routines are very important. Getting them slightly out of sync or out of phase could cause big problems. Those time constants are necessarily coupled to fluid flow and turnover in the extravascular space which is not as consistent over the whole body as is blood flow.

I will be increasing my NO/NOx level as much as I can before I get the flu. My plan/hope is that it will then turn down the gain and turn any cytokine storm into a mild but effective drizzle.

@Clark 29: Maybe by dying quickly and horribly, the rest of the population (group) is protected

You're proposing population-level natural selection that is stronger than individual-level natural selection. It's debated whether population selection occurs at all, but no one knowledgeable about evolution claims that it is outweighs individual-level selection.

So, no.

If you think about it- every population is full of people who are prepared to sacrifice their personal safety and even their lives for the greater good of the community- think police, doctors and soldiers. What about all the people who stay at their stations during a pandemic taking care of the sick? If they were only concerned for their own lives- they would be long gone to a cabin (bought with their high salaries) in the woods. It is not ignorance on their parts- they know exactly what they are getting themselves into.

Most parents know what this feels like- you wouldn't think twice about risking or losing your life to save or protect your child. Maybe this impulse also happens on a physiological level? I am just wondering out loud.

daedalus2u, how do you plan on adjusting your NO rates appropriately? Arginine and NAC meddling, or some other method?

By takeitdown (not verified) on 07 Jul 2009 #permalink

takeitdown, arginine doesnât work to raise the basal NO level long term. Arginine is the substrate for nitric oxide synthase, but NO production from NOS is controlled independent of the arginine supply (within broad limits).

The basal NO level is too important for physiology to allow it to be regulated by dietary levels of arginine.

The method I use is to have a biofilm of ammonia oxidizing bacteria on my skin. I think that is the natural way to regulate basal NO levels, the way that our bodies evolved to do so. Sweating during a fever is a way to supplement NO/NOx levels by turning the ammonia in sweat into NO and nitrite which is rapidly absorbed. The most important time to have a high NO level is before the infection happens. That increases basal mitochondria number, which increases the capacity to tolerate mitochondria destruction during the cytokine storm.

This is all too deep for me but one small comment; it is St. Jude not St. Judes. I heard the phrase "cytokine storm" while watching ReGenesis.

By Deborah Marr (not verified) on 03 Aug 2009 #permalink

Hi Revere,
What about ACE Inhibitors & ARBs. Reduction of Angiotensis II is also known to reduce the backdrop for a cytokine storm.
Your Comments??

Sudesh: The back drop to a cytokine storm, as you put it, is so complicated that I think one could come with numerous scenario (read speculations) about what might or might not work. The experience with steroids suggests that it isn't an easy "immune system in overdrive" picture. The advantage of the statin proposal is that there is some observational data in pneumonia and sepsis patients but beyond that we still have to figure this out. It's hard to experiment on people as sick as these folks are so I don't foresee a quick resolution.

Hi Revere,
Thanks ofr the quick reply. The RAS (Renin Angiotensin System) has been implicated in Hypercytokinemia. The publications:
1) Moldobaeva, A; EM Wagner (December 2003). "Angiotensin-converting enzyme activity in ovine bronchial vasculature". J Appl Physiol (Department of Medicine, Johns Hopkins University)
2) Shigehara, K; N Shijubo et al. (April 2003). "Increased circulating interleukin-12 (IL-12) p40 in pulmonary sarcoidosis". Clin Exp Immunol (Sapporo Medical University School of Medicine)
3) Marshall, RP; P Gohlke et al. (January 2004). "Angiotensin II and the fibroproliferative response to acute lung injury". Am J Physiol Lung Cell Mol Physiol (Royal Free and University College London Medical School)
4) Wang, R; G Alam et al. (November 2000). "Apoptosis of lung epithelial cells in response to TNF-alpha requires angiotensin II generation de novo". J Cell Physiol (The Cardiovascular Institute, Michael Reese Hospital and Medical Center)

... and lastly by DAS
5)Das UN (May 2005). "Angiotensin-II behaves as an endogenous pro-inflammatory molecule". The Journal of the Association of Physicians of India

These studies indicate that ARBs & ACE Inhibitors severly dampen the pulmonary related cytokinemia disorders...

Maybe a combination of PPAR's & Angiotensin II inhibitors might get a better response ??? (at least in mouse related models)....Just a thought.

Would Avandia (or a statin, Like Zocor) need to be taken weeks prior to infection or could it be started at onset of symptoms and still be effective at reducing flu symptoms / mortality rate? ? ?

gilmore: We're not even sure it works, much less how (several mechanisms have been suggested). So your guess is as good as mine. I'd go for statins before Avandia, just on the basis of adverse effects. But I already take a statin.

Thanks Revere(s)

I too take a statin. . . Just trying to get as many clubs in my golf bag before the big tourney.

Perhaps what is needed is a two-pronged approach. That is, subdivide the viral infection and the cytokine storm as if they are independent, but interactive problems.

Beginning with the virus, the use of neuraminidase antivirals seems to be too far along the process. Instead, a preventative measure of viral cell adhesion inhibitors, viral reproduction inhibitors, such as ionic metals, and agents that increase cathelicidin production, such as Vitamin D, should be used to slow down the initial infection. This can take place even before the onset of symptoms, and can continue through the course of the virus, finally assisted by the neuraminidase antivirals.

Then at the onset of symptoms, in the lead up to the cytokine storm, there would be the use of immunomodulators, attempting to inhibit enough of the major inflammatory mediators to prevent the chain reaction. Importantly, no single modulator would seem to be enough, but are much more effective in combination. Blockers for ACE, Histamine-1, Histamine-2, prostaglandin, TNF-1, and many others.

In the former case, it doesn't stop the virus, just slows it down enough for a more moderate immuno-recognition and response, and in the latter case insures that the immune response remains at appropriate levels.

By Christopher A… (not verified) on 10 Sep 2009 #permalink

I believe that the way in which the cytokine storms result in morbidty/mortality is ultimately through sepsis and multiple organ dysfuntion syndrome (MODS) (??).

Other diseases/insults to the body such as pancreatitis, or simply trauma etc. can also lead to sepsis or systemic inflammatory response syndrome (SIRS) and MODS, is cytokine storm always the cause of sepsis/SIRS?? Or is cytokine storm a phenomena limited to viral infection exculsively?


Mark: It seems to me you claim boils down to the this: that what is inaptly called cytokine storm from flu is not specific to flu but is part of a common pathway that is also used in sepsis events that are clinically similar. I think that's the working assumption of most of us but there is as yet no good evidence this is the case since we don't yet know the details of the mechanism of either. So your first sentence is really the claim that these are different names for the same thing. Possibly true. Possibly not.