Is most bird flu from poultry?

A new fatality from H5N1 infection in Indonesia has been confirmed by the WHO reference lab in Hong Kong. So what's new? Nothing in particular except it gives me an opportunity to point out something that has been bothering me about how health officials are characterizing the risk of getting infected from this virus. First the report (via Reuters):

The victim died on 16 Jun 2006 in Tulungagung in East Java province after being admitted to hospital on 8 Jun 2006, I Nyoman Kandun, director general for communicable disease control at the health ministry, told Reuters. The infection was confirmed to be [caused] by the H5N1 avian virus by a WHO laboratory in Hong Kong, he said. An official at the health ministry's bird flu center who declined to be identified said: "There was a dead chicken near his house." The chicken cage was 15 meters from the boy's home, the official added. (Reuters)

Notice what is offered as an exposure history here: "a dead chicken near his house." This time we have some additional information that the chicken's cage was 15 meters away. But at least a third of the cases from Indonesia and roughly an equal proportion from Vietnam (the two countries with the most human H5N1 infections) have no or little history of exposure to sick poultry. The Indonesian Ministry of Health will often claim sick birds somewhere in the neighborhood of a case, although whether this means literally in the neighborhood, next door, a block away or something else is not stated. The first human cases in Indonesia, a family cluster of three, had no known exposure to sick poultry. Despite this, WHO, CDC and many other "experts" consistently repeat the mantra that almost all cases have had close contact with sick or dead poultry. This is a false statement and has been a false statement for years.

It may be true that almost all cases have in fact contracted the infection from sick or dead poultry. It's certainly possible, perhaps even plausible. But the evidence doesn't establish it and one wonders how much of this is wishful thinking. So far this disease has been very hard to contract, even from poultry. Millions of exposures are occurring constantly from infected poultry in many countries. Yet few people have come down with the disease. We don't know why a few people become infected but most people don't. So it is also possible that the same people who are susceptible for reasons we don't understand can get it in other ways than from poultry, although so far most people can't.

It's time to stop repeating that close contact with poultry is required. We don't know if it is true or not.

Tags

More like this

What do you do when the chickens come home to roost and they look healthy (but might not be)? Ask the folks in Hong Kong. Except they don't know either. Update, 6/15/06: The Chinese Ministry of Health is confirming the diagnosis of H5N1 in the 31 year old truck driver from Shenzhen. He is now…
A survey of doctors specializing in the infectious diseases of children attending a conference showed over half weren't very worried about a bird flu pandemic. I guess they know something I don't. Or maybe I've been reading the wrong things. Things like this: The H5N1 bird flu virus in Indonesia…
Bird flu news tends to be episodic. For long stretches there seems to be little news (unless you deliberately go looking for it; then you find it). Then there are these little spurts as bird flu pops up here and there in the news. Human cases in places where they haven't been for a while tend to be…
So now we can all rest easy because we know where the 6 year old girl (not a boy as the Indon Ministry of Health previously reported) who died recently of H5N1 infection acquired the virus. The source of all H5N1 infections: birds. How do we know? Because an official of the Indonesian bird flu…

There is substantial evidence that human sequences in Indonesia, and perhaps elsewhere, more closely resemble sequences from other mammals than they do birds. It has been reported that there are widespread asymptomatic infections in pigs. Cats have also been shown to be infected. It seems obvious, to me, that taking samples from mammals in close proximity to infected people should be routinely done. The failure to do this suggests an astounding degree of incompetence.

By Monotreme (not verified) on 04 Jul 2006 #permalink

Monotreme is as usual quite correct. Given the degree of ignorance we suffer concering the transmission of this disease it seems highly irresponsible not to be pursuing vigorously all possible avenues, especially if their is already some evidence to suggest mammals as carriers. Are we seriously concerned about a possible pandemic or not? This is not just a problem of Indonesia but of the entire international public health research effort. It's just too much business as usual, lots of talk and meetings and not enough real research.

why can't we get some private investigator bureau to
investigate the cases independently ?

anon: They are medical records and not open. Who's the "we"? What if I paid someone to look into your medical records?

If a dead chicken can infect someone from 15 metres (50 feet) away, we are by no means talking about "intimate" or even "casual" contact being required. That's a far deadlier weapon than the chicken cannon used in our "Canadian Air Farce" TV comedy skits--and every bit as ridiculous.

Revere and readers: Dr. Niman's three most recent posts on the Recombinomics website discuss the novel cleavage sites on human and (one) cat isolates from Indonesia. According to him these do not match the avian h5n1 isolates. In looking for a mammalian source for the human infections, Dr. Niman also states that we cannot rule out a human source.
Questions: Could it be that a pandemic has already begun, and could it be a slow-fuse pandemic commencement in which asymptomatic infected people are infecting others, some of whom become symptomatic? If this is the case, and there are many asymptomatic people at large, wouldn't some widespread testing for antibodies reveal this?
As a chef, I'm just involved in cooking these darn birds and I don't know anything about the medical/scientific issues. Maybe some of you can enlighten me? Happy 4th!!

3 things:

1 I thought that blow flies were mechanical vectors.

2 How is water distributed to the households where infections are taking place in families? Is water limited? Could a sick person be using the same dipper in a gallon of water? Could enough virus be shed to infect a person by that means?

3 Are we placing pressure on virii (sp) to select for the remaining species population? Are we forcing polymorphisms that are adapted to more then one species?

I really doubt number 2. I just read that flu is infectious in lake water (to other fowl) for various amounts of days at different temps, PH and salinity.

By Trina Bashore (not verified) on 04 Jul 2006 #permalink

Ricardo: great point! I was going to bring it up myself. I am no expert, but as a biology teacher I have passing familiarity with what Dr. Niman is talking about, and to me it is THE smoking gun. Genetic sequencing is as good a scientific proof of how the virus has been passed on as you're going to get. It is like the trail of blood leading to the killer's hideout. Following the phylogenetic tree using the sequence of these exact dna "fingerprints" it would seem impossible to claim that those human cases in indonesia were individually each the result of direct infection from a sick bird unless somehow in every single case the bird virus, when it got into the human, mutated in exactly the same way at exactly the same cleavage site. Other than that, it seems obvious it was H2H infection.

By Mary from Hawaii (not verified) on 04 Jul 2006 #permalink

Ricardo and Mary from Hawaii,

I agree it is very possible that many of the cases are not bird-to-human. I also think there have been many cases of human-to-human spread, in the clusters. That does not mean the pandemic has started. Another possibility is that some of the cases represent mammal-to-human spread. My understanding is that some of the isolates from cats more closely resemble isolates from humans than isolates from birds. There may be a mammalian reservoir.

By Monotreme (not verified) on 04 Jul 2006 #permalink

Ricardo, etc.: I think mono's diagnosis is about right. We badly need some properly conducted seroprevalence surveys. What has been done so far does not suggest widespread asymptomatic infection but we need to nail this down. I find this quite surprising, but that's what the evidence we have indicates, although how reliable it is for inference more generally is an open question.

If I had to put money on one source for most of the cases I would say it was avian, but that kind of forced choice isn't required since I have no great confidence in it. It's just the most likely of the many plausible sources.

Genetic sequences, unfortunately, only sometimes tell us something about transmission. Maybe because I am an epidemiologist I prefer epidemiological data, but ideally one combines various data sources to make a judgment. The anomalous cleavage site in the Indon cases has been known for quite awhile. It is not clear what they mean.

If I'm correctly reading/understanding the philogenetic tree linked from Niman's site, it seems notable that only one cat and one "environmental" sample closely matched the majority of human samples in Indonesia. I wondered if the missing link, if there is one, might be mice (mice/rats would be nibbling at chicken feed, falling prey to cats, hanging around many households and have been implicated in the spread of other viral diseases).

Does anyone know if they've tried testing mice/rats in areas of outbreaks?

revere, we is we all - or some of us.
Just let's get organized. The investigator can interview
the victims, collect samples, interview neighbors,
do blood-tests, interview doctors.

strange that still no one from WHO, Hongkong,CDC,Indonesia
officially commented on the cleavage site.
Should be a good argument to release the sequences

anon: My understanding is that it is he opinion of WHO scientists (and I think this is unswayed by higher ups) that the cleavage site does not affect transmissibility or antigen binding. They know about it because I believe it is Peiris's lab that did the sequencing. Peter Roeder was supposed to start a study of cats but I don't know if it ever got underway. The scientists in Indon are3 aware of the issues as I understand it, but I am guessing the problem is getting cooperation of the Indon gov't, but I might be wrong.

If you wanted to send private people in to do the same kind of work you would also need the cooperation of the Gov. of Indonesia. Good luck.

I wondered if the missing link, if there is one, might be mice (mice/rats would be nibbling at chicken feed, falling prey to cats, hanging around many households and have been implicated in the spread of other viral diseases).

Does anyone know if they've tried testing mice/rats in areas of outbreaks?

Dead mice have been found at Uvs Nuur Lake, Russia (& Mongolia), where there has been a recent mass die-off of wild birds. H5 was confirmed in birds tested from the area.

http://www.flutrackers.com/forum/showpost.php?p=18488&postcount=27

no, for most things you need not even a permission
of the Indon. government or it's easy to get one.
Someone could phone the people, send mail, interview them
or just read the newspapers like Niman does.
We get such info almost randomly from some company
with other interests. Or occasional emails from
Andrew. If we were really interested in this stuff,
we certainly should do more. With little effort
($1000 per month, I guess) we could redouble our
info from Indonesia. I feel that it's completely disorganised and random.

Posted by: revere | July 4, 2006 06:32 PM "My understanding is that it is the opinion of WHO scientists (and I think this is unswayed by higher ups) that the cleavage site does not affect transmissibility or antigen binding. They know about it because I believe it is Peiris's lab that did the sequencing."

Revere, it is my understanding that the cleavage site is absolutley crucial to transmissibility. Changes in the cleavage site are what cause the virus to acquire the ability to jump species. This species jump is the crux of the problem. If bird flu would stay in birds---humans would not have an infection risk. The cleavage site changes in the bird flu virus allow it to infect humans and other species.

Please double check your info.

Earl E. , please confirm that you agree, that WHO
should finally comment about the cleavage-sites in
Indonesia ! This is known since Febr. or March
to the public, probably much longer known to WHO.
But no comment. Not by WHO, nor by other experts.

I cannot understand this--but others may. The Genesis of HPAI section talks about the cleavage site. Use Firefox to avoid an annoying popup from Microsoft Internet Explorer.

http://www.agnr.umd.edu/avianflu/

Earl: The cleavage site is a stretch of protein between HA1 and HA2. The HA protein is translated as one long protein, HA0. It needs to be cleaved into two pieces (although they remain connected via a cystine bond) and this happens in the bird respiratory or intestinal tract via a protease. In the low path strains other tissues don't have the ability to make this cleavage, but if the site is stretched out by basic amino acids then the cleavage can be effected by ubiquitous tissue proteases and hence the virus can infect many other tissues. Thus it is a question of tissue tropism.

The cleavage is required for uncoating the virus. HA1 grabs onto the host cell, the virus is then "taken in" via a pouch in the host cell membrane (endocytosis) but it is now segragated inside the cell in the endosome. The viral protein coat needs to fuse with the endosome membrane and in the process disgorges its genetic material into the interior of the cell (the cytoplasm). The M2 ion channel protein (the target of the adamantane antivirals) and cleavage are needed for this.

Thus cleavage is required for the ability to infect host cells and the ability to cleave is only present in some tissues in low path strains but in many tissues in high path strains.

One factor we know is important in species barriers is the receptor on HA1 (not the cleavage site). I am not aware of a known role for the cleavage site in host range. Maybe someone else is aware of it or you, Earl, know a source. It is likely that host range is made up of several factors that have to work together in certain ways. That is a story that still needs to be worked out. I am not aware of the role of the cleavage site in either transmissibility or host range. Perhaps you can correct me. Note that transmissibility and host range are completely different things, although you seem to have put them together.

Its importance in this instance, if it has any, is as a marker for origin, not an indication of host range competency.

You are right, revere. Most transmissibility and host range studies focus on the receptor binding domaion.

However, the cleavage site is crucial. It is the inherent
ease of "cleavability" that can cause host range jump, because it cleaves (splits) easily (like the 1918 pandemic virus). The transmissibility increases because a high pathology cleavage site produces systemic disease which can be transmitted and shed from all human organ systems.

Here is an interesting article:
See "HA Cleavability" heading

http://www.cdc.gov/ncidod/EID/vol12no06/05-1336.htm

There seems to be two issues that are bouncing around on this thread that are not controversial, but some of the comments suggest they are.

One has to do with what WHO does and doesn't know with regard to human sequences in Indonesia. The phylogenetic trees I have posted on the recombinomics site that show the human isolates are from a disc haded out to participants at the Jakarta meeting on human H5N1 in Indonesia. The disc did not have a "restricted" or "do not distribute" label, so I have posted the trees and commented on some of the slides on the disc.

The HA tree with the most information is here

http://www.recombinomics.com/phylo/Indo_Karo_HA1.html

The sequences in green are human and those shaded in pink are the Karo cluster. The green names in the lower branch of the tree have the RESRRRKKR cleavage site. These sequences not only have this novel cleavage site, but they have a number of polymorphisms in common which is why they are on a seperate branch. The only two sequences on this branch that are not green is an environmental isolate (faeces) and a cat isolate. I believe the environemnetal sample is from a human based on the number (3PH). The isolates on the tree are from Hong Kong and CDC in Atlanta. Both of these labs are WHO affiliated and both get samples from the lab in Jakarta. The CDC samples use "CDC" in the sample number. The human samples from Hong Kong end in H, which I believe stands for "human" and not "Hong Kong" (which is usually abbeviated HK for location or HKU for institutional affiliation).

In any event, the vast majority of the human isolates in Indonesia are on the lower branch and the vast majority of these sequences have the the RESRRKKR cleavages site (the isolate in red may have IESRRKKR).

Thus, WHO is WELL AWARE of the novel cleavage site and well aware of the fact that the avian sequences do NOT have this sequence.

WHO updates citing expoure to poultry are much more related to PR than science.

The second issue is the role of the HA cleavage site. H5 and H7 isolates can have a string of basic amino acids (R or K) at the HA clevage site. Many protease cleave after a basic amino acid (R or K), so a string of basic amino acids allows for cleavage of a number of different proteases. The proteases have various tissue distributions, so different sequences at the cleavage site can result in infection of different tissues. This tissue tropism can conceivabily affect transmission if the sequence allows for entry into more tissue types in the upper respiratory tract. Similarly, the cleavage site could affect clinical presentation, which can be altered by the tissue types that can be affected.

The HA cleavage site in H5N1 from the first isolate in Asia was from the Guandong goose in 1996. This sequences is RERRRKKR. It is the dominant cleavage site in H5N1 in Asia. There have been many variations on this theme. Most of the Qinghai isolates have GERRRKKR. The isolates from recent huamn cases in China have RERRRK_R (missing a K). Both of these cleavage sites have been found in H5N1 in Indonesia in poultry. However, the dominant cleavage site in Indonesia is the same as in other Asian countries and it is RERRRKKR.

Other than the Karo cluster and the second reported H5N1 case in Indonesia, the human isolates have RESRRRKKR. This is associated with the other HA changes that distiguish the human isolates on the lower branch of the tree from avian isolates in Indonesia. This change may or may not affect transmission or clinical presentation, but the sequence is unique to Indonesia. No other reported H5N1 has this cleavage site.

Thus, this site is a marker for the version of H5N1 in the lower branch, and the isolates in the lower branch are NOT avian (with the possible, but unlikely, exception of the environmental sample).

Most of the human isolates on this branch are from western Java, but at least one is from EAST Java, indicating the "human" version of H5N1 in Indonesia is widespread, but not present in over 50 avian H5N1 isolates from many locations in Indoensia.

when WHO is well aware (probably since 6-9 months),
then can we conclude that they either consider it nonimportant or that they think the public shouldn't
be told about it ? Admitting the latter would be a
serious new twist in WHO-information policy.

Earl: I know the article in the link you provided. It does not say what you indicate. Where is the work that shows the cleavage site is crucial to host range? You present an argument for why it might be (because high virulence increases transmissibility) but I don't know of evidence that this is in fact true and would be glad to have a citation from you or elsewhere.

Henry: I don't know why you imply there is anything controversial or claimed to be controversial in this thread. You and I got the trees from the same source and in addition, I was reporting what scientists who are aware of the cleavage site differences thought about it. Your version does not differ from theirs nor does it differ from my responses here in this thread, so I am not sure what you are referring to when you say this. I say quite clearly that WHO has been aware of this and how they know.

As I said, the significance of the cleavage site difference, if any, is as a marker of origin, which is what you also say. Looking at the PAM matrix, the substitution isn't a particularly unlikely one. My comments were directed to Earl and his assertion that the cleavage site is a determinant of host range. Perhaps you would like to add your thoughts here on that subject and where you agree or differ with me on my response.

Some thoughts. First, we may have another unknown mammalian repository somewhere, non-avian. This kind of repository has been seen in other infectious diseases and can act as a long-term vehicle for genetic changes. Pigs discounted, we haven't explored this factor much in influenza. If such a repository exists, its host would not be significantly affected by HPAI. Dare I say non-mammalian? That would be the scary version, though I would be surprised--there's no precedence for flu. Second, Henry's data from a few weeks ago suggested in my mind another strain in play in Indo. If we had the HA/NA sequences we could try and test that hypothesis since the WHO's release of the polymorphisms that Henry reported gave us a substantial number.

Third, the cleavage sequence could be regionally modeled. I agree that the sequence could be linked to pleotropism; is it a question of chemical basicity, or is it a question of enzyme affinity or both? Fourth, there could be low-level undetected h-2-h transmission going on but if that's true a substantial number of individuals would be experiencing minimal symptoms. Serological tests conducted to date do not support this theory, though the testing would have to be much more comprehensive to rule this out.

Nearly all the human cases are rural. If there is a large asymptomtic human reservoir and H2H is the source of most infections, then I would have expected far more urban cases. It seems obvious that whatever the source is, it's something unique to the rural Indonesian environment.

revere, you didn't even mention the cleavage site in the main text. Later you wrote that you are not so much interested in it since you are an epidemiologist.
If you doubt the importance ("if any"), then please
quote a virologist's opinion about this.

anon: Please be accurate. The post was about something else. Earl brought up the cleavage site and the discussion about them is in the comments. My comment about being an epidemiologist was that I have a tendency to think that epidemiological information is important (I prefer it) not just sequence information. If two people get sick at the same time, the sequences don't mean anything. They didn't give it to each other. I did quote a virologist's opinion (the WHO scientists). Neither Henry nor I are virologists, nor are Earl or you. Henry is a molecular biologist and is qualified to discuss the sequences from that point of view and his not being a virologist doesn't disqualify him as far as I am concerned.

You seem to think that just because an "expert" says something you should believe it, including your insistence that if someone is an expert they should be expected to give an exact probability estimate of the chance of an outbreak. This is utter nonsense. When an expert says something you don't agree with, they aren't experts anymore, it seems.

Kevin: In Indon, I think more cases are urban (maybe someone can check; I am at my pesky day job).

Marissa: I agree with your comments. The idea of non-mammalian species isn't nice to contemplate. A problem with Indon is that there has been little done on other species there.

Reveres, Niman, Anon, Marissa, Earl E.: Thank you for having this conversation. Every insight into this situation is wonderful for people like me, who don't have a clue.

Each piece of information discussed gives a starting point to investigate and self educate. Not being in the loop you quickly come to a point where the information puzzle leads to a halt. You just don't know where to go to get what you need.

I for one wish a greater tech. understanding of what is happening. I am looking at this CHIME and trying to see what information my mind can grasp. Then I see a HA3. What is that? I thought it went: H was split into HA1 and HA2 that some kind of loop (purine ring?)polymorphism was carrying information near the site of where it joined the bodies cell that let furin and others unlock H in different organs.

I can get that 190 is a double important site from what I have read here and there. I see that 226, 227, 228 are adhering points (I think). Do you know of an example of PB2 like this so maybe I could get where 627 is?

Does something show all 8 segements interacting? How in the world does NS enter into the deal? Why is H3,H5,H9 a family because of the Twists? Ions Jumping, What?

I quess all I know is what I can decipher since I do not have any background science education.

I honestly seek an understanding of what is going on. I feel like I am catching morsels dropped of the table of other people's knowledge.

Who wrote the classic flu papers? If I get the names of the people I will start digging. I know Webster, Webby, Sabbaro, Yu,Ya Ha, Shortridge. Give me some more.

Sorry if I am uneducated in things that you all know. I would like to think if I had more order my feeling of the internal interactions of influenza would be clearer.

My e-mail is momasun@earthink.net. Responses would be great. I know I am misunderstanding stuff so be constructive please.

Sorry for going on Reveres'. The subject fascinates me.

http://biology.kenyon.edu/BMB/Chime2/2005/
Cerchiara-Holsberry/FRAMES/
Hemagglutinin%20-%20Influenza%20Binding%20Protein.htm

Sorry that is the whole address of what I am looking at I do not know how to make a tiny url.

By Trina Bashore (not verified) on 05 Jul 2006 #permalink

revere, I believe that we are in agreement regarding the two main points: WHO is well aware of the cleavage site differences and the differences act as a marker for distinguishing the majority of the human cases from avain (or at least poultry) sources.

My comments were more in the fine tuning area. WHO knows not just because of sequences from the Hong Kong lab. The tree shows the CDC sequences also and shows that the are two sequences for the vast majority of the patients in the lower branch, indicating that the CDC is quite active in isolating and sequencing human H5N1 in Indonesia (and the data was present at the WHO sponsored meeting).

The other area was on the potential role of the cleavage site in transmission. Since the sequence determines which proteases can cleave, RESRRKKR could have an expanded tissue tropism that could be involved in transmission.

Tow additional points include the fact that the tree is dated June 12, 2006 indicating it was up to date, yet still had no avian isoaltes on the lower branch. This up to date tree includes two patients in the May 29th update, which included East Java indicating the "human" H5N1 was widespread, but there were no still avian sources listed.

The caveate not mentioned is the date of collection. The new tree does not have any 2006 avian isolates and many of the human isolates were from 2006, so recent avian outbreaks are under-represnted and the first human isolate is 2005, so the human sequence may simply be more recent.

Since it is July, the lack of 2006 avian sequences in and of itself raises significant surveillance issues because of the high number of human cases in 2006.

Revere, You say above, "Earl brought up the cleavage site and the discussion about them is in the comments."

NO, you brought up the cleavage site first above and I commented to your statement.:
Posted by: revere | July 4, 2006 06:32 PM "My understanding is that it is the opinion of WHO scientists (and I think this is unswayed by higher ups) that the cleavage site does not affect transmissibility or antigen binding. They know about it because I believe it is Peiris's lab that did the sequencing."

I reacted to your statement because I have read many references to the importance of the cleavage site.
One article I cannot find---and now have a headache from trying to find it. It was in Nature or Science in the fall of 2004---had comparison images of cleavage sites in swine, avian, and human influenza viruses.

This old 2004 article points out the importance of the cleavage site:
http://www-ssrl.slac.stanford.edu/research/highlights_archive/1918flu.h…

I am no expert---not in the least.

I just feel strongly that common sense dictates that the origin of the NOVEL Indonesian cleavage site that almost wiped out an entire family should be a MAJOR FOCUS of transparency and investigation into its origin IMMEDIATELY.

Earl: Yes, you are right. Sorry. I believe (again I am away from my stuff, at work) that the Karo cluster did not have the unusual cleavage site.

YES, Revere, the NOVEL Indonesian cleavage site is not in the Karo cluster. I am confused. The novel cleavage site is in most of the other fatal human infections in Indonesia, but not the large Karo cluster which did match a chicken isolate.

NO other released sequence from any other animal's H5N1 matches this novel cleavage site so far. This novel cleavage site has been around Indonesia in humans for a long time. Someone in a laboratory somewhere surely knows the answer to its origin----but
they are not telling, are they?

I guess we must wait for a large explosion of fatal cases before we get the bad news----the new WHO pandemic flu definition AND only stage left to go is called "Transmissible like a cold----but fatal".

2 questions:

1) What's a possible non-mammalian repository, Marissa/Revere and why would it be "scary"?

2) Have TPTB released any analysis of the Karo virus that might hint at why it caused an unusually large human cluster?

Earl: I don't think anyone knows what this cleavage site is all about. The R to S is not a hard substitution for a protein, however. It could be a random change with no advantage. Not every change is significant. In this case it might be a clue to origin, however.

Name: Non-mammalian could be reptiles, amphibians. I think insects or arachnids is unlikely, but who knows?

The simplest explaination may be that the WHO has unreleased avian sequences that do match the human sequences.
Rural vs Urban
A breakdown of Indonesian fatal cases at:
http://online.wsj.com/public/resources/documents/retro06-avfludeaths-co…
shows 16/40+ are from Jakarata area. 9 of these 16 had documented close association with live poultry. No data for the other 7. Not to say that anything outside of Jakarta is rural, or that poultry is *the* vector but whatever it is, it seems likely to be in some way related to animal husbandry. I'm sure that encompasses a large portion of the population of Indonesia, but I would still argue there is a significant bias in that direction.

but the RESRRKKR mutation occurs in almost all human
sequences and in almost no others. If it had no advantage
over RERRRKKR, then why doesn't RERRRKKR occur too
in West-Java ? It was probably originally RERRRKKR
and then mutated to RESRRKKR. At that point there were
both variants, but apparantly selection preferred
RESRRKKR. It has at least the one "advantage"
that it outperforms RERRRKKR in West Java
in Evolution. Either in humans or in the intermediate host.

anon: The cleavage mutation could be hitchhiking with another mutation that has a selective advantage. It is believed this phenomenon responsible for the adamantane resistance that suddenly developed in US flu strains when it wasn't under any selective pressure at all. Genetics is not just about selection. It is a complicated process in which selection plays a role in the successful lineages. Remember the Karo cluster doesn't have this mutation. Go figure.

revere:hitchhiking or not, it's still a selective
advantage as you say.
Are there other such clear distinctions bird-H5N1 vs. human H5N1 from any other geographic region ?
And that it's in the cleavage site by chance
is only 1:50.
Karo is in North Sumatra, not Java.

Wait did an R at 329 disappear in Karo?

By Trina Bashore (not verified) on 05 Jul 2006 #permalink

My friend Dr. Hideki who was on the ground in Indonesia for the quake relief specifically said that he thought that the local epidemic was underway with "pneumonia." but when they asked as a matter of course to test the people who were obviously sick, they were stopped by the local medical people. Those districts are medical warlordships in Indonesia. Jakarta has the power to make a phone call and thats about it. Everything else is treated with whatever local poultice they can concoct at the time. That includes Dr. Hideki said and no kidding, the local voodoo guy.

Because of this Dr. Hideki on arriving back in Japan wrote a letter to the Japanese Health Ministry. Now mind he was a volunteer MD but he specializes in infectious and he said there was no way that so many people could be walking around with pneumonia unless there was some sort of contagion going around.

Normally he said that you would see the post disaster diseases. Very few of those popped up because so many responded to the quake, but so many to have pneumonia in 80 degree weather is very unusual. His rant didnt go unheeded but what he got back was a statement that this was an internal Indonesian issue, they had to deal with it.

Okay, so we let it grow like we have a Petri dish growing over there until oops, it crawls out and eats everyone in the lab.....Yeah this makes a LOT of sense. Either way, cleavage sites novel or not will go undetected, untreated, and above all it will be unbelievable when it breaks out with rocket speed.

Remember Revere, that private email I sent you. There are developments on all fronts on that right now and none of them are good. Backup plans? Maybe I wasnt so off the mark last month on the old blog?

By M. Randolph Kruger (not verified) on 05 Jul 2006 #permalink

Randy: The central gov. is powerless. All power devolved to the locals. Regarding whether things are "underway," that's not something I can say one way or another. Another medical correspondent who was also on the ground doing relief after the earthquate didn't think so and he is an experienced flu person and no partisan of the gov. of Indon. That's all I know.

hi revere, you wrote:

>anon: Please be accurate. The post was about something else.

the subject was exactly this. The cleavage site gave the first
and strongest hints, that Indonesians maybe won't get it from poultry
but from some other source (pigs? cats?)

>Earl brought up the cleavage site and the discussion about
>them is in the comments. My comment about being an epidemiologist

I replied to your reply to Henry, not the one to Earl in case this was unclear.

>was that I have a tendency to think that epidemiological
>information is important (I prefer it) not just sequence
>information. If two people get sick at the same time, the
>sequences don't mean anything.They didn't give it to each
>other. I did quote a virologist's opinion (the WHO scientists).

you wrote:"my understanding is that it is the opinion"
that's a bit weaker than "quoting". And you said "scientists",
not "virologists".

>Neither Henry nor I are virologists, nor are Earl or you.

Henry very much is in virology aspects. ("recombinomics")

>Henry is a molecular biologist and is qualified to discuss
>the sequences from that point of view and his not being a
>virologist doesn't disqualify him as far as I am concerned.
>You seem to think that just because an "expert" says something
>you should believe it,

not always, of course. But better an expert than a layman or no
opinion at all. What were experts good for else ?
Reading your blog since long, I'm surprised by this your assumed
(unscientific) attitude now.

>including your insistence that if someone
>is an expert they should be expected to give an exact probability
>estimate of the chance of an outbreak. This is utter nonsense.

I didn't say anything about this in my post here. When you think you
can draw conclusions from the given email then I remind you to
your anonymity policy:
"email is required for authentication purposes only..."
Now you are trying to judge my argument here from previous arguments
of the apparant same author about a completely different issue ?

As for your last two sentences, it is utter nonsense anyway.
We discussed this before here and at fluwiki, and we can discuss
this anywhere if you want.
I have professors Neustadt,Fineberg,May,Steele, on my side here,
what experts can you quote ?
Anyone, expert or layman can give an "exact" (subjective) estimate
of his probability assignment to any future event. I said this earlier.
You may like it or not. Experts who say that they can't give an estimate
are just lying (why?). Quote me on this. Ask some mathematicians,
logicians,philosophers - but not epidemiologists, this is
not a medical but a logical,mathematical issue.

What you call "my insistance" in 2006, was common wisdom in 1976
with swine flu, see
http://birdflueconomics.blogspot.com/2006/06/lessons-from-1976-non-pand…
You wrote about it in your blog too, but didn't mention the aftermath
from Neumann,Fineberg,May. No surprise, since you don't like probability
estimates for some reason...

But as I said, that is another subject and should be separated IMO.
Me being right there (and you wrong) doesn't already imply, I'm right here too.

>When an expert says something you don't agree with, they aren't
>experts anymore, it seems.

it usually takes more than one statement. And others should disagree too.
Don't know your reference, though. ("it seems")
It's easier when you quote to what you answer, as I do here.

>Posted by: revere | July 5, 2006 09:33 AM

you can answer by email, if you prefer, or I can post it.

[no email received]

Henry and Andrew seem to agree that this is an important issue.
Yet WHO,HK,Indon.,CDC,... didn't comment or even mention it.
Although they knew it sonce long.

Can we do a little cutting with Occam's razor here?
1. FACT: the human sequences in indonesia differ from the avian sequences, hence the humans are not getting the flu from the birds...yes?

2. If 1 is true, then Humans are either passing the flu H2H, or they are getting it from an unknown reservoir (since the human sequence differs from the sequences found in other potential reservoirs they have tested such as swine, cats etc.)

3. The question that started this whole debate was whether it might be considered flawed, unwise, shortsighted, stupid or whatever term fits for the WHO epidemiologists and others in the field to make the presence of sick birds the first criteria before testing a sick human for avian influenza. (If #1 above is true, then the answer to 3 is yes.)

Also, I recall some news articles a month or two ago in which some scientists reported having done fairly extensive testing of healthy people in areas where there had been human avian influenza outbreaks to see if there was perhaps a number of H5N1 cases out there that were asymptomatic and thus going undetected.
They checked these people for antibodies to the virus, I believe, and the results were that no sign of H5N1 infection was found in any member of the general populace, thus no evidence of a hidden reservoir of H5N1 in the human population in the test area. I would appreciate it if anyone has a link to that study or some better and more direct knowledge of it than what I just recounted here. But the point is, if that study is true (and more studies like it should be done on a much broader scale over a much broader area to be sure) then Monotreme's hypothesis of a secret reservoir of the virus by which humans are being infected is a very real possibility.

By mary in Hawaii (not verified) on 05 Jul 2006 #permalink

mary:1) , looks very likely meanwhile that human and avian sequences are different in West Java (nothing official). This increases the chance that humans don't get it from birds. But this is unclear.
The mutations could also happen in the humans or only
the mutated viruses infect humans while others don't
enter the human cells and replicate.
2)see above
3) truth comes in degrees... But other sources should
be considered too. How much - hard to say as long as they
keep the data secret. It could be that they have data which makes their actions reasonable, but I'm skeptic.
It could be political that they aren't testing - they
are afraid of the results , or just lack of funding.

With the antibody testing, see here for an overview:
http://www.fluwikie.com/pmwiki.php?n=Science.Seroprevalence

Name: Why would a non-mammalian source be scary? Precisely because it would indicate that the virus has found a novel way to adapt to a host. I think this possibility is unlikely but I do not discount it. There are several viruses that are believed to have an ultimate repository (particularly in Africa) but we don't know what they are. From an evolutionary point of view, this would be an advantage. One of things I spend a lot of time doing in my day job is identifying the source of the disease in the viruses and bacteria I work with, and often this is no easy task. The scientific community has made one vast assumption in its study of H5N1: namely that to date except for perhaps the Karo cluster, every victim has caught the virus from an avian source, be it birds or infected derivatives. What I am seeing increasingly is a kind of desperateness to stick to that theory when the evidence is starting to say sometimes "no, that's not how it happened."

On the other hand, I am a loss myself to suggest a specific alternative. I'm going to do a little more research in other diseases to look at repositories, and if anything turns up, I'll post the results here.

Mary in Hawaii: Some testing for h5n1 antibodies was done in Vietnam. One discussion of this is in Effect Measure post of June 24, 2005 entitled "How much h5n1 infection in Vietnam?"

anon: I feel compelled to correct two things you said: "mutations could also happen in the humans,"

This misses the point: we are talking about a particular strain, if you will, that is for the moment consistent in humans. It may well have started as a strain coming originally from a bird and then mutated within a human host, or it may have come from a bird into another animal where it mutated before being passed to humans, but it is currently Not mutating (changing its sequence) in humans. That's why it's on its own little phylogenetic tree branch.
And that's the point of saying that, in indonesia, the majority of human cases of bird flu are NOT coming from birds. Actually, the fact that - as revere the epidemiologist points out - there are frequently sick birds linked to cases of human influenza and yet the genetic sequences are not the same, makes the puzzle more interesting. Like rats, fleas and bubonic plague...put the two disparate facts together and start looking for that non-mammalian vector monotreme proposed.

You went on to say:"or only mutated viruses infect humans while others don't enter the human cells and replicate."

This again misses the point that birds and other tested repositories of H5N1 do not have that sequence at all, so it cannot be a matter of these carriers passing a "mutated virus" to infect humans. Unless, as I hinted above, it mutates inside a vector between the first and second host.

Oh, and re the danger of officials stuck on the idea that H5N1 is predominantly B2H and thus if there are no sick birds, people with severe flu symptoms don't need to be tested for H5N1: That appears to be exactly what has just happened in Egypt. Hopefully the link below will take you to the story. If not, follow the link in the birdflubreakingnews site regarding Egypt denies bird flu case. Gets you to a Bahrain news site, click on news and you'll find the little news blurb.

http://www.birdflubreakingnews.com/templates/birdflu/window.php?url=htt…

By mary in hawaii (not verified) on 06 Jul 2006 #permalink

mary in hawaii wrote:

>anon: I feel compelled to correct two things you said:
>"mutations could also happen in the humans,"
>This misses the point: we are talking about a particular strain,
>if you will, that is for the moment consistent in humans.
>It may well have started as a strain coming originally from
>a bird and then mutated within a human host, or it may have
>come from a bird into another animal where it mutated before
>being passed to humans, but it is currently Not mutating
>(changing its sequence) in humans.

why not ? That's exactly what I meant. That it comes into the human
as RERRRKKR and leaves it as RESRRKKR most of the times in Java.
Just as a possibility, is it so unlikely ? Well, thinking at it again,
we usually only have a few mutations inside a host and that
always the R->S should be among them seems unlikely.
Although selection might favour it, once it occurs. I'm not sure,
how this works.
The other possibility is, that there could be many viruses in a heap
of H5N1 when the human is infected. Most viruses have RERRRKKR,
but some few also have RESRRKKR and only these are able to
infect the human cells for this particular strain.

>That's why it's on its own little phylogenetic tree branch.

yes, that makes the above possibilities less likely.You would need
an unknown avian reservoir in quails or such which remained
unrecognized and phylogenetically similar to the human viruses
but with RERRRKKR. But when we assume this, then we could
as well assume that this unknown reservoir has already RESRRKKR.

>And that's the point of saying that, in indonesia, the majority
>of human cases of bird flu are NOT coming from birds.

or from special birds, not poultry.

>Actually, the fact that - as revere the epidemiologist
>points out - there are frequently sick birds linked to cases
>of human influenza and yet the genetic sequences are not the same,
>makes the puzzle more interesting. Like rats, fleas and bubonic plague...
>put the two disparate facts together and start looking for that
>non-mammalian vector monotreme proposed.
>You went on to say:"or only mutated viruses infect humans while
>others don't enter the human cells and replicate."

yes, see above.

>This again misses the point that birds and other tested
>repositories of H5N1 do not have that sequence at all,

maybe they could have it but in rare quantities ?

>so it cannot be a matter of these carriers passing a
>"mutated virus" to infect humans. Unless, as I hinted above,
>it mutates inside a vector between the first and second host.

you'd have the same problem to explain it then in the intermediate host

>Oh, and re the danger of officials stuck on the idea that H5N1
>is predominantly B2H and thus if there are no sick birds, people
>with severe flu symptoms don't need to be tested for H5N1:

yes, that could make the two above possibilities again a bit
more likely

>That appears to be exactly what has just happened in Egypt.
>Hopefully the link below will take you to the story.
>If not, follow the link in the birdflubreakingnews site
>regarding Egypt denies bird flu case. Gets you to a Bahrain
>news site, click on news and you'll find the little news blurb.
>http://www.birdflubreakingnews.com/templates/birdflu/window.php?url=htt…
>Posted by: mary in hawaii | July 6, 2006 02:40 PM

they should try to replace the S by an R and leave everything else
and then try to infect human cells with that strain. If it infects
them, then the two alternatives above can be ruled out and
we definitely have a silent reservoir of RESRRKKR.

I thought the cleavage site affected the reproduction of the virus in the rest of the body: that standard cleavage sites require an enzyme in the lung lining, while the multi-basic cleavage sites have less dependence on these enzymes.

By Ground Zero Homeboy (not verified) on 06 Jul 2006 #permalink

GZB: Yes, that's correct, more or less. The cleavage is needed for the virus to uncoat and dump its genes into the cytoplasm (see my comment above at 7 am on 7/5). Thus it allows replication to occur in other tissues besides the lung and the intestines in birds. The H5N1s that infect humans are almost always (some minor exceptions) hi path strains of H5, H7 or H9.

Revere and others. You probably know about this, but WSJ posts a monthly breakdown of H5N1 deaths, which goes all the way back to the first cases. The list is quite helpful in looking at, among other things, which cases were related to contact with birds. What they don't say is what the "blank" spots in the description column mean...is it simply no information, or does it mean no contact with birds? And if there is no detailed information regarding source, onset etc. available in so many cases - even recent ones - that is also problematic. Seems by now we should have much greater transparency and care in tracking these.

Here's the URL link: http://online.wsj.com/public/resources/documents/retro06-avfludeaths-da…

By mary in hawaii (not verified) on 09 Jul 2006 #permalink

mary: Yes, this is a helpful source. The blanks are (as we understand it) places where there is no released information and probably no information at all. In Indon, especially, they do a very bad job of investigating cases. In China there may be some information management, but for the most part those blanks are lack of info and even the info that's registered in this compilation may be of doubtful accuracy in some cases.