SGDEPARTMENT OF HEALTH AND HUMAN SERVICES
FOOD AND DRUG ADMINISTRATION
CENTER FOR BIOLOGICS EVALUATION AND RESEARCH
This transcript has not been edited or corrected, but appears as received from the commercial transcribing service. Accordingly the Food and Drug Administration makes no representation as to its accuracy.
BLOOD PRODUCTS ADVISORY COMMITTEE MEETING
Thursday, September 18, 2003
620 Perry Parkway
Kenrad E. Nelson, M.D., Chairman
Linda A. Smallwood, Ph.D., Executive Secretary
James R. Allen, M.D.
Kenneth Davis, Jr., M.D.
Samuel H. Doppelt, M.D.
F. Michael Fitzpatrick, Ph.D.
Judy F. Lew, M.D.
Paul J. Schmidt, M.D.
Robert J. Fallat, M.D.
NON-VOTING INDUSTRY REPRESENTATIVE:
D. Michael Strong, Ph.D.
TEMPORARY VOTING MEMBER:
Liana Harvath, Ph.D.
C O N T E N T S
Welcome, Statement of Conflict of Interest,
Linda Smallwood, Ph.D. 4
Open Committee Discussion
I. West Nile Virus - Informational
A. Introduction and Background
Hira Nakhasi, Ph.D. 14
B. Epidemiology and Surveillance
Hira Nakhasi, Ph.D. 31
C. Plasma and Plasma Derivatives Safety
FDA Current Thinking
Mahmood Farshid, Ph.D. 49
Presentation by PPTA:
Mary Gustafson, PPTA 56
Michael Gross, Ph.D., Aventis Behring 58
Thomas R. Kreil, Ph.D.,
Baxter BioScience 62
Open Public Hearing
Dr. Susan Stramer, American Red Cross 81
Dr. Michael Busch, Blood Systems
Research Institute 106
Dr. Michael Strong, Puget Sound Blood Center 137
Dr. James Gallarda, Roche Labs 142
Dr. John Saldanha, Canadian Blood Services 155
Dr. George Dawson, Abbott Laboratories 161
Dr. Jeffrey Linnen, Gen-Probe 170
Dr. Wayne Hogrefe, Focus Technologies 180
Dr. Steven Kleinman, AABB 188
Dr. Louis Katz, America's Blood Centers 193
Summary of Committee Updates
Dr. Jay Epstein 205
Open Public Hearing 212
P R O C E E D I N G S
Welcome, Statement of Conflict of Interest,
DR. SMALLWOOD: Good morning. Welcome to the 77th Meeting of the Blood Products Advisory Committee.
I am Linda Smallwood, the Executive Secretary. At this time, I will read the Conflict of Interest Statement that applies to this meeting.
This announcement is part of the public record for the Blood Products Advisory Committee meeting on September 18th, 2003.
Pursuant to the authority granted under the Committee Charter, the Director of FDA's Center for Biologics Evaluation and Research has appointed Dr. Liana Harvath as a temporary voting member.
Based on the agenda, it has been determined that there are no products being approved at this meeting. The committee participants have been screened for their financial interests. To determine if any conflicts of interest existed, the agency reviewed the agenda and all relevant financial interests reported by the meeting participants.
The Food and Drug Administration has prepared general matter waivers for the special government employees participating in this meeting who required a waiver under Title 18, United States Code, Section 208.
Because general topics impact on so many entities, it is not prudent to recite all potential conflicts of interest as they apply to each member. FDA acknowledges that there may be potential conflicts of interest, but because of the general nature of the discussions before the committee, those potential conflicts are mitigated.
We would like to note for the record that Dr. Michael Strong is participating in this meeting as the Non-Voting Industry Representative acting on behalf of regulated industry. Dr. Strong's appointment is not subject to Title 18, United States Code, Section 208.
He is employed by the Puget Sound Blood Center and Program and thus has a financial interest in his employer. He also is a researcher for Roche Molecular Diagnostics. In addition, in the interest of fairness, FDA is disclosing that his employer Puget Sound Blood Center has associations with regional hospitals and medical centers that could be affected by today's committee discussions.
With regard to FDA's invited guests, the Agency has determined that the services of these guests are essential. There are interests that are being made public to allow meeting participants to objectively evaluate any presentation and/or comments made by the guests.
For the discussions on Supplemental Testing for Human Immune Deficiency virus Type 1 and Hepatitis C virus, Dr. Miriam Alter was scheduled to speak, but she will not be here today. However, for the record, I will read the disclosure regarding her.
Dr. Miriam Alter is Chief of the Hepatitis Branch, Centers for Disease Control and Prevention in Atlanta, Georgia. She has a financial relationship with several firms that could be affected by the discussions. In addition, CDC purchases assays from firms to perform reference and research testing.
Dr. Michael Busch is Vice President, Blood Systems Research Institute. He receives consultant and speaker's fees from firms that could be affected by the discussions. He also has a financial and professional relationship with several other firms that could be affected by the discussions.
Dr. Steve McDougal is Director, Immunology Branch, Division of AIDS, STD, and TB Laboratory Research at the National Center for Infectious Diseases, Centers for Disease Control and Prevention in Atlanta, Georgia.
Dr. Susan Stramer is employed by the American Red Cross, National Reference Laboratory of Infectious Disease. She is a researcher, a scientific adviser, and has financial interests in firms that could be affected by the discussions.
In addition, there are speakers making industry presentations and speakers giving committee updates on regulated industry and other outside organizations. These speakers have financial interests associated with their employer and with other regulated firms. They were not screened for these conflicts of interests.
FDA participants are aware of the need to exclude themselves from the discussions involving specific products or firms for which they have not been screened for conflicts of interest. Their exclusion will be noted for the public record.
With respect to all other meeting participants, we ask in the interest of fairness that you state your name, affiliation, and address any current or previous financial involvement with any firm whose products you wish to comment upon. Waivers are available by written request under the Freedom of Information Act.
If there are any disclosures to be made by any of the committee members or participants that have not been addressed, would you please address them now.
DR. SMALLWOOD: Hearing none, I would like to inform the audience and those that are speaking in the open public hearing that there are statements that are required for you to make prior to you speaking to an open public hearing.
We have placed out on the table a listing that identifies open public hearing at FDA Advisory Committee meetings and frequently asked questions. There are questions that you may be concerned about when presenting during the open public hearing, however, we have a policy now which will be in effect as of this meeting, that when an individual is presenting during the open public hearing, the committee chair will read a statement, and the individuals we will ask to be able to respond accordingly.
I will just read this quickly in your hearing, so when you hear it, you will know what you are supposed to do. This is to be read by the chair for this meeting in particular in reference to guidance documents and policy review.
Both the Food and Drug Administration and the public believe in a transparent process for information gathering and decisionmaking. To ensure such transparency at the open public hearing session of the Advisory Committee meeting, FDA believes that it is important to understand the context of an individual's presentation.
For this reason, FDA encourages you, the open public hearing speaker, at the beginning of your written or oral statement to advise the committee of any financial relationship that you may have with any company or any group that is likely to be impacted by the topic of this meeting. For example, the financial information may include a company's or a group's payment of your travel, lodging, or other expenses in connection with your attendance at this meeting.
Likewise, FDA encourages you at the beginning of your statement to advise the committee if you do not have any such financial relationships. If you choose not to address this issue of financial relationships at the beginning of your statement, it will not preclude you from speaking.
Again, I will remind you that this will be read at the beginning of any open public hearing sessions by the committee chair.
I would just like to thank all of you for coming out with regard to the impending storm, our committee members in particular, our participants, and the audience.
I would just like to announce that there will be a workshop on Factor VIII inhibitors November 21st, 2003, at the Lister Hill Auditorium, National Institutes of Health. There is a flyer out on the desk.
At this time, I would like to introduce to you the members of the committee that are present today. As I call the committee's names, would you please raise your hand.
Dr. Kenrad Nelson, Chairman. Dr. Michael Fitzpatrick. Dr. Liana Harvath. Dr. James Allen. Dr. Judy Lew. Dr. Kenneth Davis. Dr. Paul Schmidt. Dr. Robert Fallat. Dr. Michael Strong. Dr. Samuel Doppelt.
We will also have members that will be retiring from the BPAC as of September 30th, 2003. Those members are Dr. Mary Chamberland, Dr. Michael Fitzpatrick, Dr. Daniel McGee, Dr. Paul Schmidt, Dr. Sherri Stuver, and Dr. Robert Fallat.
At this time, for those members that are leaving, that are retiring, that are present, I would like to ask Dr. Jay Epstein to come forward and we will present them with plaques for their service.
DR. EPSTEIN: Thank you, Linda.
Mike, with a moment of indulgence, I just want to add a personal word of thanks to the retiring or I should say outgoing members of the committee. I think it is worthy of just a moment to consider where we have been in the last year to two years, and the very exciting times in which we live.
This committee has helped us bring in the era of gene-based testing of donors to protect the safety of the blood supply with successful implementation and licensure of hepatitis C and HIV NAT tests.
You have guided us to develop the criteria for approving HIV rapid testing, to consider the suitability of waiver under CLIA. You have helped us review innumerable donor deferral criteria including the Uniform Donor History Questionnaire Initiative, but more specifically, the issue of tattoos and acupuncture, and other things.
We have been dealing with emerging infectious diseases. These have included the increasing impact of Chagas' disease, new awareness of significance and possible approaches to human parvovirus B-19, the outbreaks of SARS and West Nile virus, and against perhaps an historic background, but an unsolved problem, advancements in controlled bacterial contamination.
I could go on. That is just sort of a list of highlights. I am struck by the fact that the metaphor of coming in a hurricane is very apt to the business of our committee.
I would like to award these plaques and with that express both the gratitude of the Food and Drug Administration, as well as my own personal thanks to you for your exemplary service in the interests of the public health.
That said, first, Dr. Michael Fitzpatrick.
DR. EPSTEIN: Dr. Paul Schmidt.
DR. EPSTEIN: Dr. Robert Fallat.
DR. EPSTEIN: Thank you very much.
DR. SMALLWOOD: Again, I would just like to thank all the members of the Advisory Committee for coming out.
With regard to the agenda for today, we are all aware of the impending hurricane and we would like to make a modification, that is, that we will proceed with the first discussion issue, which is West Nile Virus, followed by the second issue, Supplemental Testing for HIV and HCV.
We will delay the presentations on Committee Updates. We will determine where we are at a significant time, but our purpose is to try to move the discussion as quickly as possible, so that you may make the best arrangements for your personal safety.
May I ask now, are the presenters for the first topic on West Nile Virus here? We are aware that Dr. Anthony Marfin will not be present today. All right.
That being the case, then, I will turn over the proceedings of the meeting to the chairman, Dr. Kenrad Nelson.
DR. NELSON: The first presenter on West Nile, Dr. Hira Nakhasi from FDA, or even also give Dr. Marfin's data?
DR. NAKHASI: I think I will just make a slide presentation. I may not be able to give exactly what he wanted to do, but I will try my best.
DR. NELSON: All right.
West Nile Virus - Informational
Introduction and Background
DR. NAKHASI: Good morning. Thanks for coming.
Let's talk about the West Nile virus and the recurring theme I think which we have been doing for two years, for updating you on the progress both in the testing and development of donor testing and donor screening.
Also, we would have done on epidemiology and surveillance, unfortunately, Dr. Tony Marfin cannot be here, and as we all know why.
The issues for discussion will be today basically the current status of West Nile epidemiology and surveillance.
Dr. Tony Marfin would have talked about it. I would talk about the progress in West Nile test development including lot release and validation panel development which is going in-house in the FDA, and progress on West Nile testing of blood donors using investigational tests.
Both the test manufacturers and the blood banks will talk about the progress in the testing, and I will just summarize in a slide how far we have come along, but the details will be presented by the blood establishments, as well as the manufacturers.
We will also be talking about the current thinking on the safety of West Nile infected plasma and plasma derivatives.
With that background laid out, let's move to the next slide.
I will be talking basically about the FDA's actions to date with regard to donor testing and the West Nile testing development including the lot release and validation panel.
Just to give you a brief background, but most of you know, because by now you should be remembering by heart, in fact, we have a quiz at the end of the session, so we will find out whether you remember things.
West Nile is an enveloped, single-stranded RNA virus. West Nile is a mosquito-borne flavivirus, primarily infects birds, occasionally infects humans and other animals.
As you all know, 80 percent of human infection is asymptomatic, and 20 percent develop mild febrile illness, flu-like illness.
Approximately 1 in 150 infections results in meningitis or encephalitis, and at advanced age, is the most significant risk.
Also, the viremic period can vary. It may vary up to two weeks prior to symptoms and can last for a month in some of the cases, however, these were the original data as we now get the real-time data, we will know more about how these infections proceed.
The 2002 U.S. outbreak showed that mosquitoes are not the only culprits. It can be transmitted through blood transmission, for example, RBCs, plasma, and platelets were shown to be implicated in transmission, transplantation, breast-feeding, transplacental and occupational by percutaneous injury.
That magnitude again it is the tip of the iceberg as you will see from Tony's slides. We do not know how much is through the transfusion transmission related West Nile risk.
Again, another component in the situation is that virus titer is low compared to other viruses, such as HIV and HCV, but the interesting thing is that viremia resolves rapidly in most cases after seroconversion to IgM.
IgM can persist up to two years, so that if somebody has an antibody, it doesn't mean that he is infectious, however, you need to make sure about that.
Fortunately, no chronic stage of West Nile infection has been reported so far.
I had removed this slide, but it came handy anyway because Tony was going to talk about it, and I will just basically give you an update. In 2002, there were approximately 4,000 reported cases of West Nile infections, 284 deaths, and approximately half of them were West Nile meningoencephalitis.
Forty-four states including D.C. was endemic. The overall rate in the United States was 0.4 per 10,000, however, in high-risk area it went all the way up to 10.46 per 10,000.
There were 61 possible transfusion reported cases last year. There was no testing, but it was retrospective testing of the 2002 epidemic. Out of that, 23 were confirmed from 16 blood donations, 19 were not transfusion related, 19, it could not be decided what was happening because of the incomplete donor follow-up.
There were 6 deaths reported in this cohort, however, it should be underscored that West Nile could not be established as the cause in most cases, so we do not know whether they were West Nile transfusion-related cases, deaths due to that.
The FDA actions to date. Starting last August, we started putting quite a bit of alert notices, then, we had a workshop, and we talked about the development. Soon after that, we talked about facilitating the development of donor screening and supplemental tests.
We had numerous congressional hearings and we had a workshop. We talked to industry about developing the test kit and to expedite development and implementation. We presented at several BPACs including this one.
In addition to that, more importantly, we also issued two guidance documents. One was in October 2002, and the other one is May 1st. In the May 1st, we added a pre-donation question.
We are asking people if they had a fever and a headache a week before the donation, because it seems, as you will see in one of the paper recently--in your package, you should have the paper from Tony Marfin's group--in a recent issue of The New England Journal of Medicine where they are talking about fever plus headache is a good indication they had developed West Nile infection.
This spring, following up on the test development, FDA approved three INDs, one for Gen-Probe, they would test 16 samples in a pool, and American Red Cross, which is using the same Gen-Probe test against pool of 16. We also approved Roche IND for West Nile NAT, which uses a pool of 6.
Investigational West Nile NAT testing started in mid-June 2003 using pooled or individual samples, with 95 percent screening since July 1st, because July 1st, most of it, greater than 95 percent screening had started with the whole nation.
FDA is participating in weekly meetings with the task force established by the blood banking community, which includes CDC and NIH, and we are weekly coordinating or monitoring West Nile infections and the outcome of the testing, so I think it has been a very, very fruitful interaction with industry, with the other PHS agencies, to really follow up.
I think the cooperation is excellent, and I would like to personally thank all the members of the committee who are really having an open discussion about it, and that's how we really know what is happening week by week.
With regard to the activities happening in-house, we are developing panels. This panel development is to monitor the sensitivity of assays to detect both viral nucleic acid and antibodies.
We also are trying to isolate and characterize West Nile strains from the epidemic samples from blood samples collected in 2002 and 2003. Why are we doing that? Because there is a possibility that there could be a genetic variation of the viral strains. We want to make sure that is not happening. We, also, with the currently available tests, can detect all those variations.
Then, we are planning with the other PHS agency and industry to look at infectivity studies basically, so that we want to see whether these samples, which have the low level of viremia, can be infectious or not.
For the assay standard so far, analytical sensitivity, as you know, we have 100 copies/ml for the individual donation, however, it can change as these tests become more and more sensitive, we started with that, and also the West Nile panel designed to monitor the sensitivity limit.
The two panels which we are developing are the ones for the West Nile virus strain isolated from flamingo strain, which is passed through the viral cells. We call it FDA NY99, which is New York 99, the flamingo strain, and the FDA-Hu2002 isolates, which was isolated from the plasma infected person.
These two panels have been tested for viral infectivity, are in a concentration, and also these are being distributed, these panels have been distributed now among the various collaborating partners to make sure that the number of copies/ml in these panels are right.
That is still going on, and maybe we will hear from some of the industry folks what the outcome of that study is.
Just to give you a flavor, the titration was done of these two isolates. That is the human isolate and the New York isolate, and now three different concentrations, five different labs, and it is gratifying to see that all of them came to the same conclusion that the final copy number was 1010/ml.
Also, there is a difference between the copy number and the infectivity, the infectivity as measured by plaque-forming units. You see the FDA-2002, the human strain, there is almost a difference of 1 plaque-forming unit/1,000 copies kind of a thing.
Here, New York 99 is 1/100, so there is a difference. Overall, what the standard now has been that maybe 1 plaque-forming unit is equal to 500 copies/ml, in that range.
It is important to know when we make these panels, we want to make sure, because these are inactivated viruses in there, because you are measuring the copies/ml, so you want to inactivate the virus. So, when you do the inactivation of this virus at 60 degrees/2 hours, you lose the infectivity.
However, at the same time, also, there is a reduction in the copy number, so even though we thought it was 1010 input, but when you inactivate it, you will come out with 107. Therefore, to make sure the numbers are right, so this number, we have done several times, that it comes out the same thing. The same way with the NY99, there is a reduction in the copy number upon heat inactivation.
This slide, I had not planned to keep it, however, because Tony is not here, I just wanted to briefly update you what is happening, how this epidemic is shaping this year, 2003.
As of last week, 3,000 cases, NMWR reported 3,000 human cases, 53 deaths, but this may be changed as I see the slides this morning, I will go there, their numbers are a little bit different, however, it gives you an idea that according to Tony, we are on the same rate as we were last year, so the epidemic is not moving faster than last year.
Again, as of today, there are 44 states and D.C. are endemic for West Nile.
Well, the good news is that, as I said in the beginning, that starting July 1st, all blood donations are screened by the minipool NAT under IND, and we have interdicted--"we" meaning the blood donor establishments--have interdicted several hundred donations which were from asymptomatic donors with confirmed or suspected West Nile infection. Approximately, more than 500 units of the approximately more than 2 million donations so far screened, more than 500 units have been interdicted.
So, the good news is that as of last year, when we did not have any testing in place, we have prevented quite a bit of going into the transfusion cases.
Basically, the minipool NAT, because remember this is done in a pool of 16 and 6, and we have removed effectively more than 75 percent infected donations entering the blood supply for transfusion, so that is the good news, because we are preventing and the blood supply is safe, however, there is some residue of risk because of the fact that, as we knew beforehand, that going into the minipool NAT and the virus having a low titer, that there will be some of the cases which will be missed because of the low titer.
This slide shows you what we are doing collectively. "We" means collectively, the industry, FDA, and other PHS agencies including CDC is doing, because as I said, there was a potential for transmission through the minipool NAT.
So, limited retrospective elevation of minipool NAT testing was initiated 2003 epidemic retrospectively in the areas where incidence was high regions, 1 in 250 minipool rate was there. ARC and others tested, and the retrospective studies using last year's epidemic sample to look back and seeing how if they did in a minipool test, and how many they missed there.
The bottom line in this case is that there are some samples with low level of viremia which may be missed by minipool testing, and there also are some putative, at this point, and maybe two confirmed cases of West Nile transfusion-related cases have been analyzed, but some cases are being analyzed at this time, confirmation by NAT, IgM reactive donor follow-up recipient even in spite of having a minipool testing.
Therefore, the question is should we undertake an individual NAT study, because, you know, as such, to do an individual NAT, there is a lot of problems, logistic problems, analyze the sample collections, and things like that, so I think those things will be discussed by the industry what the problems are.
The important thing is that industry has undertaken to study both the ID-NAT prospectively and retrospectively in high-incidence area, follow up the recipient, and to determine the infectivity rate in the units.
The important thing is we do not know, even though there is maybe low level of viremia, whether that is infectious or not, that is important, because we may say that, you know, but that can be done by some--unfortunately, that is done by if inadvertently some of these cases are transfused into people, and they seroconvert, then, that tells you that those samples are infectious, or it can be done in animal studies using these samples, minipool NAT negative and ID-NAT positive units.
It is also being planned that the two test kit manufacturers will test these samples among themselves because, as you know, one is using a pool of 16, another using a pool of 6, so is it necessary that we need to do the ID-NAT or smaller pool size will also do it, so that it will pick up at the smaller pool size.
So, what is happening? Since these studies are pending and we do not know what the outcome will be there, and it will be too late by the time we will know about it, so what industry has done, the ID-NAT is being performed prospectively in the high incidence areas. That is why the high infection, the positive rate of the minipool positive rate is much higher.
Again, this is done based on again the additional capacity the test manufacturers have and also the region by region. In addition to that, frozen plasma is being withdrawn in areas with high incidence of West Nile virus based on the minipool NAT.
In summary, we have now a test in place, even though it's an investigational test. We have whole blood donations. Most of the blood in the country is screened, and we are interdicting more than 75 percent of cases using minipool NAT and the studies are underway to see how many are we missing through this testing, which would help us in the future planning.
Thank you very much for your attention.
DR. NELSON: Questions or comments?
DR. LEW: Do you have some preliminary data on the areas where they are doing individual NAT, how many additional cases are actually seen?
DR. NAKHASI: Not yet, but I think the industry will present that data, but then there are some cases, you know, in one small study--is it okay, Mike, if I say?
DR. BUSCH: She is asking prospectively.
DR. NAKHASI: Prospectively?
DR. LEW: No, the data you already have.
DR. NAKHASI: There was a study of 7,000, you know, minipool-negative samples were tested individually, and initially, they found 27 reactives out of which 5 were repeat reactive or 6 were repeat reactive, so that is the rate at this point.
DR. NELSON: That seems high. If it's 500 per 2 million, that's only 2 per 10,000, and if 75 percent are picked up in the minipool, you would expect less than 1 per 10,000, maybe a couple per 100,000 in individual that were missed on the minipool, right? Did I do that wrong?
DR. BUSCH: The whole idea is we focused in on these very high minipool yield regions, so those regions had rates of pick-up of about 1 in 250, so we are looking at incremental pick-up approaching 1 in 1,000.
DR. NELSON: So, the 500 per 2 million is all the testing.
DR. BUSCH: That's the minipool, yes.
DR. NELSON: That was the basis for your estimate that said that the minipool would pick up 75 percent?
DR. BUSCH: Yes, based on that.
DR. ALLEN: In areas where you are now doing prospective ID-NAT testing, do you have a trigger for allowing that to drop back to minipool testing?
DR. NAKHASI: Yes, that's true, because once the rate falls below that 1 in 250, I think the 1 in 500 or something, and also the availability, I think there is a trigger because you cannot keep on doing that, because there is not enough extra testing material available.
DR. NELSON: This is really remarkable that this was implemented starting about this time last year.
DR. NAKHASI: That's right, I think that is a very, very interesting thing.
DR. NELSON: Are you going to present the--why don't you just show the slides?
DR. NAKHASI: I gave half of his information anyway, but I can show the other half.
Can we have Dr. Marfin's slides, please.
Epidemiology and Surveillance
DR. NAKHASI: Dr. Marfin would have talked about the epidemiology and surveillance update basically discussing the virology, clinical spectrum, epidemiology from 1999 to 2002, and some of these transfusion-associated cases, which he calls TAT, in 2002, surveillance update in 2003, and he was telling me the last time he gets always this question, what will happen in 2004, and a crystal ball.
I think we should skip these because I just mentioned these belong to the Japanese Encephalitis family, then, the mosquito transmission, humans are incidental hosts, and low level of viremia.
This is the tip of the iceberg. Again, the majority of the cases are asymptomatic when West Nile infection occurs and less than 1 percent CNS infection, and West Nile fever is 10 to 30 percent.
This is the update on last year and the take-home message is that older people are susceptible and now immune. He is also putting a question mark, immunosuppression and diabetes, so I don't know what information he has on that, but age is still the factor.
Meningitis is this age, average age is 46 years.
This is the cases of the fever, West Nile fever, I think, last year, which was done retrospectively because last year they did not really have anything, the information.
There are two information sets. One is the ArboNET, which is through the CDC and the State Public Health Department have reported that, and then the other thing is what we get from the testing donors and from the public health information.
There is a delay in the reporting into this system.
I guess in 1999, you can see the color coding of the slides. It started in the New England area, in New York, Long Island, Connecticut and New Jersey, and Maryland was part of that. Then, next, here it spread a little bit more, and then, you know, the next year a little bit more, and in 2002, the rest is history because it's all over the country now, and except still as of 2003, there are only two states where one can live peacefully is Oregon and Texas.
This is the human disease cases. I think this is as of last year, period ending December 2002, 704 counties, 39 states and D.C., and this many cases, which I just told you a few minutes ago.
This is now, what, 3,000 cases of neuroinvasion disease. This year is--which year is this, I don't know--2002, I think so.
The transfusion-associated transmission cases. Again, this is the manuscript which to which I was referring, I think it's on line now, and I guess this week it is going to be published in The New England Journal of Medicine, a very nice analysis of the story done there, basically, the one I mentioned earlier, 23 cases, came from 16 donations, and the follow-up of that, IgM and NAT testing, the disease correlation of how many had fever, how many had headache.
From that analysis, both fever and headache came out that it's correlating very well with the West Nile infection. Obviously, as you remember the case, the famous case was the organ donation from a West Nile person who had got into an accident, got a transfusion, and then donated and died and gave the donors, and the donors got, you know, through the organ donation.
Also, the cases where the woman got it through the blood transfusion after delivery.
This is the same slide.
This is basically talking about the median age and counties, and titer of these things, and higher mean titers in symptomatic was 4 pfu/ml in asymptomatic donors.
This gives you the number of meningoencephalitis cases by week of illness onset. There were cases collected by July 22nd to October 6th, and 65 percent of cases fall in middle of August to middle of September, which is still the case happening in the majority of cases, because the peak of viremia, in fact, epidemic is in that area.
The additional information since West Nile 2002 investigations. Titers of 0.1 pfu/ml identified in a donor of plasma, so I guess it can be as low as 0.1, and there have been more than the 23 cases estimated. There is 500 viremic donors may have occurred in 2002. We will know this year when the testing is going on, so we will know, and it seems there is a bet among the people how many cases we will see this year.
Surveillance in 2003. I don't know what joke he was going to say on that, so I don't know anything about that, so please forgive me for that.
The cases, as you can see what is happening when the reporting was done. It looks like, according to Tony, it's at the same rate as late year, so we will see, it's only the 11th of September, when we will know what is happening in the next few months.
This is the West Nile activity, as you saw what is happening this year, 2003. Interestingly, the non-human West Nile virus activity has spread all over the country except the West Coast, as Mike would say, the Rockies have stopped the West Nile over the West Coast, however, the human cases, you can see are now, which began in New York, then, the New England States.
Last year, the majority of cases were in the Michigan and in the Midwest area. This year it seems to be shifting to more in the bottom of the Rockies in the Oklahoma and Nebraska, Kansas, Colorado, Wyoming, and you can see the blue is the human disease cases. These are the high incidence areas where, some of the places, single-unit NAT testing has initiated.
So, it has been interesting to see how this thing has been moving along from east to west, and not many cases this year in East Coast. There are several factors, they were saying mild summer, rainy weather, a lot of rains, but who knows.
Again, this year, you know, it was very hot and not much rain in these areas, so could that be the possibility because last year we had the same weather as these folks here are experiencing according to Tony.
With regard to the West Nile meningoencephalitis cases, West Nile fever, and other--I don't know what other means--you can see that last year, around 50 to 55 cases with West Nile meningoencephalitis because fever was not monitored last time, this is a retrospective study, so one should not draw any conclusions from it because this year, fever has been monitored from the very beginning.
You can see still, you know, the year is not over yet, so 2003, we have around 30 percent of the West Nile meningoencephalitis cases, fever is more, and then I don't know what "other" means.
With regard to the cumulative number of West Nile virus meningoencephalitis cases by reported date, ArboNET, this is 2002 and 2003, in 2002, we had 727 cases as of September 17th, which is I guess yesterday, and we are around 153. Whether there is a slight increase, I don't know, but we should hold our judgment until the end, because according to him, we are still going on the same pace as of last year.
With regard to the deaths, again, the chart shows you the same thing, again reported to ArboNET, and I think we are in the same pace as last year.
With regard to the human disease cases by week of illness onset, again, you can see July, late July, it looks like it's peaking down here, but I don't know how much we need to put the weight on, we need to see, because the reporting data is delayed, there is delay in the reporting, so these figures could change here, but the trend look like that, it is the same as of today, the middle of August to the middle of September may be the epidemic.
This is the 44 states again, you know, the same thing which I mentioned to you, the number of cases, but this is the ArboNET. Remember, there are differences, sometimes I said 44, sometimes 36. This is because of the ArboNET and also the other reporting cases.
So, again, you can see this year, 88 percent of reported cases comes from this area, which is in the Rockies, in Colorado, South Dakota, Nebraska, Wyoming, Texas, Montana, North Dakota, and New Mexico. Colorado tops the list here.
Again, the epidemic greatly affects counties with small populations, it looks like according to his analysis. With less people, you see more cases incidence per million because much higher, but if it's in a population with more than 1 million, higher than 1 million, the incidence rate is different. So, on average, I don't know what the average would be around here. It would be still the same as last year, I think.
The culprit this year. Why we are seeing much higher incidence of this West Nile virus infection in Colorado and other areas, according to Tony, the mosquito Culex tarsalis is the one, which West Nile virus replicates very well into this insect, so this is very well adapted there, so that may be the reason there are more infections, you know, you are getting much more infections there.
According to him, the high infection rate in these mosquitoes is 36 percent, so that is very, very high.
Blood donations as surveillance events.
This is what we reported to you, that that testing has started, you know, that things are in place. It is much more better organized this year than last year because of the testing and follow-up.
There are approximately presumptive viremic donations, as I said, 500 presumptive viremic donations, and he has broken down on the basis of counties and where the maximum are, the age group, and also again there are seven states which have more than 20 presumptive viremic donations.
So, you can see again that 89 percent of the cases are coming from the 89 percent of states where the epidemic is very high, so nothing surprising in that.
This is giving you the map where these cases are located. Again, it matches with where the epidemic is.
As he says, 2.5 million donations have been screened as of July 1, and according to, not the ArboNET, but the other reporting the 601 presumptive viremic donors in 32 states. Again, the top 6 states are from those areas, so both cases, it is the same.
This is one case report where it is now confirmed that the West Nile virus was transfused into the recipient by transfusion. This is the case where the donor was donation, minipool NAT, non-reactive, and this is one of the samples which Mike Busch will be presenting out of those 7,000 units, minipool NAT-negative, one of those turned out to be individual NAT-positive, and then this individual donation was repeat-reactive on August 14, and became seroconverted on August 25th. So, obviously, the donor was infectious because it got seroconverted.
This is the same donation. This is the patient who had a surgery, I guess, aortic graft surgery. This is the famous Texas-2 case, they call it, and got 7 units from 7 donors from July 23rd to August 7th. Got into surgery on August 4, got the 5 units of packed red blood cells, and one of them, this individual donation, repeat-reactive unit was in that unit.
So, that person got that on 8-5. This person got symptom onset and PCR-positive and IgM-negative on 8-7. Then, it became unequal on the PCR, that is, it started losing PCR positivity on 8-12, and then by 8-16, not it got seroconverted IgM and the PCR-negative.
So, basically, it tells you that this unit, which the donor who became seroconverted, the donation from that person, and this person got also seroconverted, so obviously, the culprit was that donation. So, this is the first confirmed case of the transfusion-related transmission.
This is two, basically, putting the two slides together, and basically saying that this is what is the outcome of that.
Well, in summary, it says the second consecutive national West Nile epidemic, number of severe infections and deaths comparable to 2002, so not to worry, it is not going to be more than 2002.
But the important thing to remember is shift in geographic focus, southeast to west. The disease burden shift to rural counties. Traditional areas like New York City still reporting human West Nile meningoencephalitis cases.
Large number of presumed West Nile viremic donations in areas of highest West Nile meningoencephalitis rates. Risk of West Nile are not yet zero, so, obviously, that was the conclusion.
So, the question is, Tony, what is going to happen in 2004? I guess this is not Tony here.
2004. According to Tony's crystal ball, he thinks the St. Louis encephalitis virus transmission pattern versus, you know, we need to make sure, you know, history tell us, I don't know how the SLE transmission occurred, and the patterns are that, you know, now you get periodic, even though in 1997, there was a high epidemic, and then we have local transmission.
This doesn't seem to be the case at least for the last three years, we have a very sustained increase in the epidemic with regard to West Nile infection.
The question is whether next year it burns out, does not cross over to Rockies or does it cross over to Rockies, and what is happening, we will know, we will see. Maybe the epidemic may not be at the same rate as last year.
Then, the question is 40 to 50 percent of cases that already have illness onset are not yet reported to ArboNET. I guess the projecter decided not so say anything more.
But, in summary, what I will tell you that I think what he has been discussing--"he" means he and others--have been discussing with each other that the epidemic, as such, is at the same rate as 2002.
Also, we are following up on that, we are doing these prospective and retrospective studies and to see what the rate is, and we will see what happens down the road.
Thank you very much for your attention.
DR. NELSON: Thanks for presenting these slides on two minutes notice. You did a good job.
DR. NAKHASI: Thank you.
DR. NELSON: Any questions?
DR. NAKHASI: I will collect the questions and send it to Tony. No, I am just kidding, I will answer.
DR. NELSON: It is interesting that there seems to be a pretty high risk of an adult asymptomatic donor in a hyperendemic area as being viremic, and this contrasts a little bit with dengue and JE, which also has annual epidemics, but I guess the issue is that at least many of the dengue infections are in young children who aren't donors.
DR. NAKHASI: That is true.
DR. NELSON: But I am sure that as dengue expands, particularly in Latin America, that probably transfusion-transmitted dengue is being missed, and it would be interesting for some studies of this dengue in the last few years has expanded all the way into Argentina and even the southern Latin America now it is really pretty common, so many populations has high rates of infection without a large immune population, so it is possible that there are dengue donors that are transmitted, but I don't know if anybody is studying this, but it certainly would be worth looking at.
DR. NAKHASI: I just want to give at least a take-home message here, that is very important to remember that blood is safe and also compared to last year, when we didn't have any tests, we have a test in place, we are interdicting samples which are infectious, and we have still work to do, but we are in good shape at this time. That is the main take-home message here.
I think also we need to applaud the test manufacturers, the blood bankers, that the testing implementation within a short nine months, when we discussed this last year, that the test was developed, it is investigational use starting from when the epidemic started, and also that the studies are undergoing, so I think that we should not forget, and the blood banks had with all the other testing going on, to implement this one on top of that, it is commendable. Thank you.
DR. NELSON: I just wondered if the other prevention activity, that has been recommended, is interdicting or preventing transfusion of blood from donors who are febrile with headache, you know, in the couple weeks before or after donation, I wonder how many units have been interdicted.
DR. NAKHASI: Well, we may hear from--
DR. NELSON: From maybe some of the blood bank people?
DR. NAKHASI: Yes, we will hear from that.
DR. NELSON: It would be interesting to see how effective that is. I suspect with these rates of 500 donors, there probably have been some prevented with that.
DR. NAKHASI: Yes, I think there were. I don't have that many details. Go ahead.
DR. FALLAT: Could you comment a little bit more, did I read the slides right, that the infectivity rate in the East in non-humans is about the same as last year, but the infectivity rate in humans is way down? Why is that again, I mean you are explaining it on the basis of temperature or rainfall, could it be that a high percentage of the population is already immunized?
DR. NAKHASI: Nobody knows why it is, but those are the several possible explanations. One is that maybe people are immune and they have neutralizing antibodies, and therefore they don't get infected, one.
The second thing is that the weather was different this year, but the question is if, in fact, the mosquitoes, according to Tony, you know--maybe Matt will allude a little bit more in that, I see Matt here now--you know, I was presenting Tony's slides here in case you missed that, so that question is posed to Tony every time, and he basically says he has not really a good answer to that. Maybe, Matt, can you allude to that?
DR. NELSON: I doubt it's based on human immunity because, you know, in the Brooklyn area, whatever, there was only like 2 percent or 1 percent of the population had antibodies after the epidemic, but it may relate to immunity in birds. Matt?
DR. KUEHNERT: This is Matt Kuehnert from CDC. You did a much more entertaining job than I am sure I would have been able to do here, so I appreciate you presenting the slides.
I was just going to mention, actually came up to mention that the Texas case you described, as well as an update on activity in the United States, as well as the donor surveillance, is going to be presented at an NMWR dispatch that will come out today or tomorrow, so that will be further described.
Concerning the question you asked, I don't think I can answer that. I could make a guess. As you saw on the slide, there is a different mosquito species that is responsible for transmission of most of the cases in the Great Plains area of the United States. As was mentioned, there is likely a different dynamic of transmission amongst mosquitoes and birds compared to last year, so it is probably not the human side as prevalence overall is still quite low in the eastern United States, but rather dynamics between the mosquito and the avian population, so that is probably a more logical explanation.
DR. NAKHASI: Thank you, Matt.
DR. NELSON: Thanks very much.
The next presentation is Plasma and Plasma Derivatives. Mahmood Farshid.
Plasma and Plasma Derivatives Safety
FDA Current Thinking
DR. FARSHID: Good morning. The issue of West Nile and safety of plasma derivative has been discussed at BPAC in March of 2003, and today's presentation is intended to be an update since March and essentially to see where we have been, where we are currently, and where we intend to go. The presentation specifically deals with West Nile clearance studies.
Before going specifically to West Nile clearance, I would like to outline the current approach which is taken by CBER in evaluating viral validation studies in general.
Manufacturing processes must contain at least two orthogonal and effective steps for removal/inactivation of viruses, and we "effective" is defined as achieving more that 4 logs. This is not only CBER definition, this is also an accepted industry standard and also accepted by other regulatory agencies.
One of the clearance steps must be inactivation, therefore, complete reliance on removal would not be acceptable. However, two inactivation steps could be in the manufacturing process provided that one of the inactivation steps be effective against non-envelope viruses.
In response to how much total log reduction is required, the validated steps should remove or inactivate three to five orders of magnitude more virus than is estimated to be present in the starting material, and that is estimated to be more than 10 logs for relevant virus, i.e., HIV, HBV, HCV.
I would just like to indicate that what you saw here is not basically one format which deals with all viral validation studies. We look at the viral validation studies on a case-by-case basis. There are a number of variable components which need to be evaluated, and what we saw here, the assumption is that everything is done accurately and this is overall approach that we take.
With regard to the type of the test virus which are used to conduct such studies, the actual virus of concern should be used in viral validation studies if technically feasible, and the reason being because we want the data that are generated be as close to the real thing, if you would, as possible.
Specific model viruses can be used in validation studies if technical experimental limitations do not allow the use of relevant, and these being the case, for example, in case of HCV and HBV, where we do accept model viruses for viral clearance studies.
In March of 2003, we put specific question to the committee and we asked their advice. The question was whether viral validation data on model flaviviruses are sufficient to demonstrate West Nile clearance.
Essentially, we asked if the existing validation data, which are generated using model viruses, could be extrapolated to West Nile virus.
The second question was whether the clearance steps in the manufacture of plasma derivatives obviate the need to screen donations of source plasma.
In today's presentation, the second question will not be discussed because we are receiving data from the plasma industry which currently, some of it is under evaluation, some of it has not been received, and further discussion on this particular issue will be sometime in the future. Basically, the presentation will focus on the viral validation studies.
As you can see, BPAC's recommendation with respect of viral clearance, they advise that the viral clearance studies should be conducted using West Nile virus as opposed to model viruses.
Following the BPAC meeting in April, we met with PPTA, which is umbrella organization covering the plasma industry, and at that meeting, the West Nile clearance studies were presented and validation data submitted to the FDA.
Their presentation and the submission of data was done independently by three different manufacturers, and the data indicated that West Nile inactivation studies was done for multiple plasma-derived products, using a variety of inactivation methodologies to demonstrate the comparability of West Nile clearance to other flaviviruses.
This is summary of some of the data that we thought would be relevant to the discussion today, and in using two of the most common and frequently used inactivation methodologies which include heat treatment and solvent detergent treatment. The three different manufacturers using different product, as you can see, they compared the clearance of West Nile virus with that of BVDV, which is commonly used as model flavivirus, and is used as a model virus for hepatitis C also.
As the table indicates, using these two methodologies for inactivation, they generated significant viral kill with regard to West Nile virus, and the level of clearance which was obtained for West Nile was comparable to that of BVDV.
The data basically showed that certain common inactivation methodologies, which include specifically solvent detergent and heat, which include pasteurization or vapor heat, that have been used by multiple manufacturers and multiple product, consistently provided significant levels of inactivation. That is more than 5 logs for West Nile virus.
The level of West Nile inactivation is comparable to that of BVDV, as I indicated, an historical model for hepatitis C inactivation studies.
Therefore, the current thinking is that available viral clearance data comparing West Nile virus to BVDV have demonstrated comparable levels of clearance for these viruses following solvent detergent or heat treatment in multiple products.
BVDV inactivation data can be extrapolated to West Nile for the aforementioned methodologies, i.e., solvent detergent and heat treatment.
We further concluded that the level of inactivation shown in these studies, combined with the expected, but not yet determined--I would like to say that the validation data submitted to us by PPTA members were concerned with inactivation steps and did not do any removal step, therefore, we say that their contribution of the removal steps are expected, however, no experimental data similar to what we saw for inactivation have been submitted yet. So, I will read this again.
The level of inactivation shown in these studies, combined with the expected (but not yet determined) contribution of the removal steps in the manufacturing processes, are expected to be sufficient to demonstrate the safety of plasma derivatives with regard to West Nile virus.
Validation data with regard to the removal steps, comparing West Nile virus clearance to model flaviviruses, are needed to further establish West Nile virus safety of plasma-derived products.
Where we intend to go, FDA intends to request that the manufacturers document the adequacy of the clearance steps for all their products with respect to West Nile. This could be done either by submitting the clearance data for West Nile virus or they can submit the existing data which they have for BVDV data for the well established and, in this case, solvent detergent and heat treatment.
For novel or less established method, where we don't have experimental data comparing the two side by side, West Nile clearance data need to be submitted.
That concludes basically the updates since last March.
DR. NELSON: Thank you.
Any comments or questions?
DR. NELSON: Okay.
Mary Gustafson from the PPTA, Plasma Protein Therapeutics Association.
Presentation by PPTA
MS. GUSTAFSON: Thank you very much.
As described by Dr. Farshid, the issue of plasma derivative safety with respect to West Nile virus was presented at your meeting in March.
At that meeting, information from studies conducted by PPTA member companies were summarized. The information included studies using West Nile virus and model viruses, and showed that the model virus concept is predictive of West Nile virus clearance.
Following that meeting, PPTA requested a meeting with FDA. This meeting was held in late April. At that meeting, member companies supplied company-specific proprietary data that formed the basis of the March presentation.
In addition, PPTA presented a three-part approach for addressing the threat of West Nile virus to plasma therapies. The three-part approach includes the following.
The three-part approach presented by PPTA for addressing the West Nile virus threat includes: risk assessment, which is an assessment of relevant epidemiological factors and worst case challenge to the safety of plasma therapies; the verification studies that have been discussed by Dr. Farshid; and a plan to monitor existing donor screening and surveillance, and this continued surveillance of the West Nile virus epidemic for sentinel parameters and evaluating the industry actions.
What has happened since April includes cooperating with the American Red Cross, and we wish to thank them in sharing information collected during the 2003 epidemic of testing blood donors, and the industry has just completed a cooperative study, which includes retrospective testing of retention samples from plasma manufacturing pools prepared from plasma collected during the height of last year's epidemic.
The PPTA presentations for this morning include the PPTA cooperative study design and the results that will be presented by Dr. Michael Gross from Aventis Behring, and an update of the PPTA risk assessment for plasma derivatives that will be presented by Dr. Thomas Kreil from Baxter BioScience.
DR. GROSS: Good morning. I am going to present the results of our cooperative study which, as Mary said, we just concluded. We literally got the results yesterday and put this presentation together, so the ink is still wet on the page and we have not audited the data that I am about to present to you although we are confident that probably very little, if any, will change as we audit the data.
These are the companies that participated in the study, and I should go back just one moment and say, I wanted to add that, in fact, we had some independent studies that were done in conjunction with the study that was completed yesterday, and the results are consistent, which again gives us more confidence that probably very little will change.
These are the companies that participated in the studies. These are all obviously plasma derivative manufacturers.
The objective of the study was to basically try to determine the West Nile virus titers in worst case pools, and the way that we determined the worst case was to look at pools that were formed from donations during the peak period from geographic hot spots during the 2002 epidemic.
Each company submitted 50 samples, and these samples were from the first homogeneous plasma manufacturing pool after thaw and removal of cryoprecipitate, and because at least one company uses recovered plasma, the other companies used source plasma, there are some differences in the representation in those pools. It is fairly broadly represented in terms of the number of individual donations.
The samples were received frozen, shipped frozen, received frozen, manipulated occasionally for re-aliquotting, and so on, and certainly for assay, but during all other times, they were kept in a frozen state at minus 20 or below, and we did conduct a freeze/thaw study as part of this to ensure that these manipulations did not affect the titers of the samples.
The selection criteria, as I said, were from the peak time period 8-20 to 9-22-02 in states having at least 100 clinical cases of the virus last year.
The total number of samples represent over a million donations, and every pool contained at least some of the donations meeting those target criteria. In many cases, there was a very substantial number of donations, up to about 40 percent donations in some of those pools.
We also included both negative and positive controls formed from plasma that was known to be free of West Nile virus and then spiked.
The assay was conducted at NGI. We used the screening assay and then the assays that screened positive were titered.
The results are as follows: 93 percent of the samples were at or below the level of detection, 7 percent of the samples tested above the level of detection and were in the range of 200 to 400 or 420 copies/ml. The blinded controls performed as expected.
We conclude that the virus is not detected in the majority of manufacturing pools and when it is detected, the titers are low, consistent with our risk assessment. It needs to be kept in mind that further downstream processing results in the combining of pools and there can be further dilutions.
Now, I will turn the podium over to Thomas, who will update our risk assessment.
DR. KREIL: Good morning. I would like to give you an update on the risk assessment that initially, the industry has pulled together and shared with the Agency during a technical meeting in April of this year.
This is the major factors that we have taken into consideration in pulling together this risk assessment. It is information on plasma viremia levels, then, information on the prevalence of viremia in the donor population.
Both of these pieces of information would tell us about the resulting plasma manufacturing pool loads, and obviously, that information then needs to be seen in contrast to the reduction that we can achieve by our manufacturing processes.
This is the information that was available to us on plasma viremia, the first set here from symptomatic individuals. Obviously, this is not persons who would be eligible as plasma donors, one, an anecdotal observation from patients deliberately inoculated with West Nile virus, people with underlying malignant disease, and then an observation published during the 2002 epidemic with a person who had manifest neurological symptoms at the time of determination of West Nile virus titers.
Certainly, the more important and relevant set of data comes from otherwise healthy donors, and this is, first, information that was provided by the Agency to the BPAC for their meeting in March, suggesting that the typical levels of viremia would be anywhere between 1,000 and 5,000 copies/ml.
Then, we also have this one single observation that NGI publicized during the March BPAC meeting with the single donation at 200,000 copies/ml.
From the ongoing testing during the 2003 West Nile virus epidemic season, we know that one donation has been identified at a factor of 2, different from the 200,000 that we knew about earlier at 3.9 times 105.
What is certainly more important here, however, is that now with the pretty robust database and the total of 264 West Nile virus positive donations confirmed as of last week, the mean titer of these donations was a factor of 40 lower and there were around 11,000 copies/ml.
At this point, I would like to again thank Dr. Susan Stramer from the American Red Cross for sharing these data from their ongoing investigation with us.
The worst case for the 2003 epidemic then compares to the worst case of the 2002 epidemic by just a factor of 2.
The information that we have available on the prevalence of viremia in the population first came from a very elegant modeling study out of the CDC, which suggested that during the peak of the epidemic in 2002, and focusing on the hot spots of the epidemic, the prevalence would be between 10 and up to 20 per 10,000 or 1 to 2 per 1,000.
Also, in 2002, however, the American Red Cross did a PCR investigation targeting particularly high-risk areas and targeting also high-risk periods for collection of the donations.
For this particular population of donations, the before mentioned model would have estimated a risk of 8.2 in 10,000 roughly, and that is exactly where the later done PCR investigation came out, with 2.4 per 10,000 or roughly 1 in a 1,000. That is for the 2002 epidemic.
Now, from the ongoing investigation that ARC is performing for 2003, the numbers for the highest risk periods and the highest risk areas is 1 per 230 donations, which is a factor of 4, higher than we have seen from this limited investigation of 2002.
Now, if we translate that into resulting plasma pool loads following the model that we have tried to establish here, for 2002, the maximum viremic donor prevalence was 1 in 1,000, the maximum viremia levels 200,000 copies, and if one assumes that such a donation would be diluted 1/2000th in a plasma pool, then, the resulting manufacturing pool would have been at 200 copies/ml for 2002.
With the more recent information that we have now on the 2003 epidemic, the maximum risk areas in the U.S. may be as high as 1 in 230, the maximum numbers of copies found in a single blood donation was 390,000 copies, and these two numbers would then translate in a maximum of 1,700 copies/ml for a 2003 plasma pool.
I want to emphasize here that all of this obviously is worst case assumption and that we have assumed here the highest potential load and the highest prevalence with every donation at this prevalence coming in at the highest ever observed titer.
Now, this is probably a more realistic assumption here in that we would still assume 1 in 230, which again is the highest ever observed prevalence of viremic donations, but we would take the mean viremia levels out of the ARC investigation, which has a total database of some 260 donations so far of 11,000 copies/ml.
This would translate, by calculation, into this model estimated mean of 50 copies/ml for a plasma pool. Now, this, I would like to set into perspective to what you just heard from Dr. Gross, the retrospective investigation of 2002 manufacturing pools, and that if you take into consideration all the results generated here, then, this assay-based estimated mean would translate to roughly 40 copies/ml.
So, I think it is important to stress here that the retrospective testing data did verify the assumptions that we made during our risk assessment. Also, it would indicate to us that the retrospective testing of 2002 pools and the numbers generated for the 2003 epidemic are actually in quite good agreement.
Now, so far we have been talking obviously about PCR detectable virus which is not infectious virus. From an investigation that NGI has presented to the March BPAC, the minimum amount of virus that is PCR detectable corresponds to 3,000ths of an infectious unit per ml.
If one assumes that a single virus copy is already PCR detectable, then, that can be calculated backwards into 1 infectious virus particle corresponding to approximately 350 PCR detectable copies, and as we heard in the presentation from the FDA before, it seems that their investigation suggests that the range could be something like 1 in 500 PCR detectable copies, which I think is in quite good agreement.
Now, taking this ratio into consideration, then, the highest virus load measured in a plasma pool was 420/ml, which would then translate into a single infectious unit per ml for these plasma pools.
Now, this needs to be seen in relation to, as I mentioned before, the reduction capacity of our manufacturing processes for West Nile virus and similar viruses.
So far, all the dedicated virus inactivation steps which have been investigate with West Nile virus itself, and I have listed a number here, and all these treatments have so far been investigated and the results made available to the Agency, resulted in complete inactivation of any intentionally added levels of West Nile virus.
This inactivation was found to be very rapid in terms of kinetics of inactivation, and we feel that by comparing this type of inactivation between West Nile virus and the model viruses that we have earlier used to model such viruses like BVDV or TBD verifies that the behavior of West Nile virus was exactly like predicted from the model viruses earlier used.
In relation to the reduction capacity of our manufacturing procedures, it needs to be pointed out that the overall reduction capacity is a product of not only the dedicated virus inactivation steps, but there is further manufacturing steps which contribute to the virus reduction. These are typically purification steps which serve to remove impurities and also remove impurities such as viruses.
The dedicated steps, as I have told you before, have already been investigated with West Nile virus, and we would think that by establishing the comparability of West Nile virus and BVDV, then, the experience with BVDV for contributing steps would actually indicate a higher safety margin also against West Nile virus.
All the data generated were presented to the European regulatory body for plasma products, as well as the Agency, and has resulted in a CGMP position paper on West Nile virus where the CGMP has concluded that the data presented provide reassurance that the steps currently in place are adequate to assure viral safety of plasma-derived medicinal products.
We would therefore want to conclude by saying that if West Nile virus is present in plasma manufacturing pools, then, the loads are actually very low. Current manufacturing processes do adequately assure the safety of plasma derivatives, and we feel that West Nile virus PCR testing would therefore not add a meaningful safety margin for plasma derivatives.
DR. NELSON: Thank you very much.
Any questions, comments? Yes, Judy.
DR. LEW: Have there been any studies at all looking at the possibility of infectious batches of pools, though? I mean it is all theoretical, these numbers, that I am just thinking all of our plasma-derived products--I won't say all--but a good number of it goes to severely immunocompromised patients where one infectious unit per many cc's could be quite infectious.
DR. KREIL: I guess I should probably clarify here. The titers that we were talking about for the manufacturing plasma pools, that is obviously pools before they undergo any manufacturing processes.
Now, the manufacturing procedure, as the Agency has pointed out, and as we would like to also emphasize, have a very significant removal capacity for this virus, and if we are talking about removal of greater than 5 logs, then, that would mean a removal of greater than 100,000-fold. So, these manufacturing processes would be very capable to certainly deal with the levels of virus that you just mentioned.
DR. LEW: But you are saying there is no studies just to look in a number of pools, even after manufacturing process to see if there is anything infectious.
DR. KREIL: Well, I am afraid that would technically not be feasible because even for the manufacturing pool before any virus reduction has taken place, the majority of these pools is negative with even the PCR assays that we have used.
So, you have to translate, then, the low levels of PCR-detectable virus in the residual pools and translate that by down a factor of roughly 350 or 500 into infectious virus, which would make the detection of such low levels of virus probably not technically feasible.
DR. NELSON: The spiking experiments are certainly reassuring.
Any comments? Jay.
DR. EPSTEIN: Thank you. First of all, I would like to thank PPTA and its cooperating members for being proactive both in generating empirical data and in attempting models for risk assessment and models for risk management.
That said, I think that there are a few notes of caution that should be heard. The first is with regard to infectivity since that is the ultimate concern. Tissue culture infectivity may not represent human infectivity in the sense that humans may be more sensitive than tissue cultures and therefore we may be seriously underestimating the infectivity per number of genome equivalents. That is point one.
Secondly, the worst case estimate is being based on the data on the highest titers seen in surveillance of whole blood donors. Now, of course, that is the sample available for study, but when you consider the very large number of units that are pooled, even in a single lot it's typically, voluntarily restricted to no more than 60,000, and, of course, many, many lots of many, many products every year.
There is the possibility of rare infections that may have higher titers for brief periods and we simply haven't noticed that yet, at least in part because the studies are being done in the blood donor population where donors are sampled rather infrequently except for platelet apheresis, they are basically once every eight weeks or less on average, considerably less.
So, the likelihood to pick up a very high titer viremia of a very brief duration is kind of low, so again, it speaks to the issue of a possible underestimate.
A third point that I would make is that the dynamics of infection are different in the source plasma setting, not because the infections differ in the donors, but because the donors are collected at a much higher frequency, namely, as much as twice a week and for a series of weeks.
So, if there is a high titer viremia, (a), the likelihood of catching it with frequent collection is much higher than in the blood donor setting, and I would point out that there were no studies shown prospectively looking for frequency or titer viremia in the source plasma donor setting, those estimates all came from whole blood; and (b), the number of reactive units that would be captured in that setting is much higher because the viremia persists at least a week, maybe sometimes for two weeks, and you keep collecting from the donor.
Now, the units are in a voluntary hold of 60 days, but you won't end up interdicting them unless the donor gets sick and calls back, because you are not doing NAT testing, so you just simply don't know.
Eighty percent of infections are asymptomatic. You might have collected a donor multiple times, three times, four times, maybe even more. So, I just think that we need to be very cautious when we say we have done a worst case estimate, and as you, yourself, have implicitly pointed out, the estimates from 2002 data were, in fact, too low. Those were called worst case.
Now, you have taken some better data from 2003, but you are also calling them worst case, and I would say that we don't have the full spectrum of experience in hand yet.
Now, having said all of that, I think that the question is correctly framed, whether the clearance capacity of the manufacturing processes is adequate to handle the highest possible viral burden, and whether that can become a principle for deciding whether any additional donor screening is needed.
Now, the concept in donor screening is that it will serve the end of limiting the possible virus burden in the pool, therefore, adding robustness to the clearance capability, because you would know that the input virus is always lower than some level.
Now, another way to go about this, if the industry is highly confident that the burdens are always below some number, we could ask you to test the pools, and you would then reject manufacturing of high-titered pools as long as they were below some agreed limit.
We would then have confidence that the inactivation will clear that much virus instead of flying blind where you just don't know what was in the pool. You are presuming it was low, but you didn't test the pool, you didn't test the donors.
Then, a second point is that I think we would be looking for validation of the processes as described, I think very clearly, by Dr. Farshid, so that we have a high level assurance that there is redundancy and adequacy.
The redundancy is an important feature because we know that there can be at times unintentional errors, breaches of CGMP which could compromise the effectiveness of any given clearance step. So, part of the issue is wanting assurance that there are, in fact, multiple validated steps acting in concert, again suggesting that we have a very robust procedure.
So, again, my reaction to all this is that I think that it is movement in the correct direction, but that we need to be very cautious whether we have really looked at the worst case scenario and whether we have really put in place the adequate controls that are the thing that will provide assurance.
DR. KREIL: All these points are well taken. I would actually only want to respond to two of them.
You mentioned before that our worst case assumption for 2002 has been too low, and I would disagree with that because our worst case assumption for 2002 actually turned out to be true, not with the data from 2003, only indeed we hit the worst case.
The other point that I think I would like to make, about the blood donations versus the plasma donations, we feel that there is probably no fundamental reason why we should assume that blood donors are fundamentally different from plasma donors, and given the very high numbers of blood donations which have been tested, I think that that would at least provide us a more reasonable extrapolation on potential positivity in plasma donations as we had it last year.
Finally, with the GMP breaches, obviously, that is a long discussion between industry and the Agency, but we feel that there is strong mechanisms in place to assure that particularly for virus inactivation steps, that there is a double signoff on the adequacy of these, so the GMP breaches in these steps should never occur.
DR. NELSON: You have discussed a lot the viral inactivation steps, but I wondered, in production of plasma derivatives, if there are any viral concentration steps, in other words, does the virus sequester with Factor VIII, Factor IX, or a product.
We know that, for instance, HIV, there is more HIV in platelets and some blood products connection with concentrate virus, so is there any data on that? Is the West Nile virus uniformly distributed, and if we start with a certain titer, do we know that a product from that plasma also has the same titer?
DR. KREIL: Well, what you are talking about we would call partitioning of viruses, and this partitioning for several model viruses has been investigated. There is only a few examples where partitioning has led to effective virus removal and that the virus would end up in a waste fraction which is removed from the product stream, but I am not aware of any example that virus would be concentrated in a product.
DR. NELSON: Other comments or questions?
Okay. It has been suggested that we take a break now in case somebody is being blown around the expressway to try to get to make a presentation that has now been pushed up because of the change in the schedule.
Why don't we say 10:15 when we reconvene.
DR. SMALLWOOD: Please return to your seats and may I have the attention of the audience.
We are preparing to move in the open public hearing now for the West Nile discussion. For anyone who came in late, we did make an adjustment in the agenda. We initially omitted the committee updates, so that we could move into the discussion of issues.
If we move along at a pace that will permit us to deal with the updates, we will, but following the West Nile discussion, it is our intention to break for a quick lunch and then to reconvene for the second topic on the supplemental tests for HIV and HCV.
We would ask all of the presenters to help each other in terms of timing because we know that you are concerned about the weather. We did do a check. We haven't found that anything has changed at this point. It is still predicted that the storm will arrive a little later. Right now it's not raining.
However, I just wanted to inform everyone that public transportation is closing down at 11:00 a.m. There will be no Metro, no Ride-on. If you intend to get to the airport, hopefully, that you can, that you will have to take a taxi, and we will give you further updates as we become aware of them. Thank you.
Dr. Nelson, we are ready for the open public hearing. Again, I will just remind everyone that there is a statement that will be read by the committee chair prior to this open public hearing, and please respond accordingly. Thank you.
Open Public Hearing
DR. NELSON: We are now in the open public hearing. I have to read the statement that has already been read.
Both the Food and Drug Administration and the public believe in a transparent process for information gathering and decisionmaking. To ensure such transparency at the open public hearing session of the Advisory Committee meeting, FDA believes that it is important to understand the context of an individual's presentation.
For this reason, FDA encourages you, the open public hearing speaker, at the beginning of your written or oral statement to advise the committee of any financial relationship that you may have with any company or any group that is likely to be impacted by the topic of this meeting. For example, the financial information may include a company's or a group's payment of your travel, lodging, or other expenses in connection with your attendance at this meeting.
Likewise, FDA encourages you at the beginning of your statement to advise the committee if you do not have any such financial relationships. If you choose not to address this issue of financial relationships at the beginning of your statement, it will not preclude you from speaking.
Susan Stramer from the American Red Cross.
DR. STRAMER: Thank you. I have no finances with me. I have no finances in the my checking account and my savings account is rather low.
I have certain financial relationships with some of the companies that I will talk about today, and that has already been disclosed.
I am going to try to make this as painless as possible, but I do have a lot of information.
This is really an update of where we are today. The Red Cross began testing for one of its regions, Southwest, on June 23rd. We implemented testing systemwide on the 29th of June.
Based on data that has already been presented by Hira Nakhasi on high incidence area in Kansas and Nebraska, we initiated a 28-day inventory testing or rotation performed when we received our first group of hot positives. When we initiated testing, we initially saw positives on the 9th of July and then the 14th of July, so we did a market withdraw inventory rotation going back 28 days from the first positive to the initiation of testing, and we saw no positives that confirmed. During that period of time, the inventory testing was done by single-unit NAT, so it confirmed really that when we caught our first positives, we really hit the beginning of the epidemic.
With the high rates that we are seeing in these two states that I will go over, we did initiate prospective screening in Nebraska on the 20th of August and in Kansas on the 25th, and based on continuing high rates, we did a frozen transfusible market withdrawal on the 5th of September. The period covered from the period when we had the first NAT-confirmed positive, which were these dates up here, to the initiation of single-unit testing. The same for Kansas.
To date, as of, I think this was Friday or Saturday, we have tested about 1.383 million donations. We have had a total of 592 NAT reactives, but not all of those are real or confirmed positives.
Our confirmed positives are nestled in these 286 positives, which we call hot cases, and they are hot cases based on the signal-to-cutoff ratio that is the output from the Gen-Probe test, which is the investigational test that we use.
So, when a signal-to-cutoff ratio is greater than 25 or equal to 25, it is considered presumptive and then we do confirmatory testing. So, all of our confirmed positives are within this 286, and to date, they include 192 of the 592 total. That actually is a frequency nationwide at least for Red Cross collections of about 1 in 5,000.
We have also had 306 cases that have had lower S to CO's, and to date, just about 60 percent of those, 178, have been negative by confirmatory testing. These are the parameters of our testing. We have had an invalid rate of just over 2 percent, which is pretty good, an invalid sample rate, which has been very low, and these are the reactivity rates for pools at just over 1 percent, and 1.67 percent for the single-unit testing that we have been doing.
In comparison to HIV/HCV NAT, if you just look at false positive rates, which are these percentages here from pools, these numbers all are very low in comparison, for example, to serology tests. This number now is lower than any serology tests that we perform.
The same thing is true for individual testing that occurs HIV/HCV. This is am extremely low number, that is, we defer only about 1 in 40,000 donors for HIV/HCV, which again is the lowest deferral we have for any infectious marker that we do testing for.
In comparison, West Nile--and these are only the false positives--have been running a little bit higher, at least 10-fold higher than it has for HIV/HCV.
We have seen this slide before. It just reminds us of where we are in the epidemic. This years's epidemic curve has been pretty similar to last year's, so we should be coming down now in the number of cases we are seeing.
This is the Red Cross version of our epidemic curve so far by week. The purple bars here indicate the number of confirmed or hot cases that we have had, and we did see a peak at the end of August, now it has been going down.
The eggplant-colored bars represent the cases that have less S to CO's, and in the beginning of testing, we saw a very high number of false positives. We implemented a number of measures to decrease that false positive rate, and as you can see, we were successful in that.
Now, as we went to single-unit testing, we have been seeing an increased number of low S to CO samples in those regions, higher numbers than we are seeing for high numbers in Nebraska and Kansas.
This is the same slide just showing it in a line graph rather than a bar graph.
Just to show some of the improvements we made. As we all know, the test was implemented very rapidly. Nine months after the developments, we implemented the testing. So, we had some problems, and one problem that we identified with West Nile, that we did not have with HIV-HCV NAT, was how we centrifuged the samples, the centrifugation conditions.
So, just showing for two of our NCLs over changing some of the parameters for centrifugation helped us eliminate some of the false positives early on in testing.
This summarizes our hot cases or those again confirmed positives and those that have very high signal-to-cutoff ratios. The individual unit S to Co's are here in eggplant, and you can see for the most part they are all very high. We have confirmed some cases with lower S to CO's, and the pool ratios, because these are diluted, do run with lower signal-to-cutoff ratios in the individual donations.
These are the corresponding cold cases, again, most with very low signal-to-cutoff ratios.
This is where our cases have occurred for Red Cross collection. Again, this has been highlighted. Our hot spots are right in the center of the United States, in Nebraska and Kansas, with Nebraska at 136, Kansas now with 76, and then these are our next highest winners.
We have had Minnesota reporting an increasing number of cases, and even out in our own area, we have had about 15. Hopefully, Isabel will take the mosquitoes away.
This gives you a detailed map of Nebraska. An earlier version of this is in an editorial by Roger Dodd in The New England Journal of Medicine that will come out with the Pealer paper.
Our first cases really occurred along the drainage basin of the Platt River, but as time went on, you can see all of these red dots represent affected counties, and the numbers in the boxes represent the number of donors that tested positive in each of these counties.
Also, please remember that Nebraska has a population of 1.7 million people, so this is rather a large impact on that sparsely populated state.
This is the same for Kansas. All the Kansas cases have been pretty much spread, even spread pretty evenly among the western part of the state. Again, the same, the numbers in the boxes indicate the number of cases per county.
Of those that have confirmed positive, the 192, this represents the viral loads of those samples. We have had a median of 5,000/ml, a mean of 25,000. As was pointed out earlier this morning, our maximum viral load was 390,000 copies/ml.
We have had some samples, five now, with lower viral loads that have also tested and index IgM positive.
If we look at the viral load testing that we are doing, and that is performed at National Genetics Institute, against the CDC TaqMan PCR results to look at the relationship, there is a very strong correlation although the CDC copies/ml are estimated from PFUs/ml from a standard that is run with their TaqMan PCR assay, and this assumes a conversion factor of 400 copies/ml, which is probably a little biased. It probably should be closer to 100 or 200.
I don't like showing line listings because they obviously can't be read, but this just shows you an example of the types of tests we do for each positive donation.
This is the index collection tube, a PPT. This is the test of record results, and you can see high S to CO's here. This is then the viral load testing from National Genetics. We retrieved the frozen plasma, repeat TMA, the NGI testing, and then we do the IgM test by Abbott, and in all of these except for one, that I show you here, they were all IgM negative.
This is the follow-up collection that we do, days post index. We again repeat the TMA, the NGI test, and the Abbott IgM. A couple of notes here that those donors on this slide and that we have seen ongoing, whose follow-up has been at five days or less, have been IgM negative when we have collected our first sample. So, we get very short follow-up times, we can still see IgM negativity, but most of the samples that we collect do have both--in the first sample--have both virus and IgM that coexist.
Here, we had a very interesting donor, stated at 1500, and actually increased in titer over the period of sample, and that doubling time then could be calculated at 13.7 hours.
If we used that donor that I talked about, is what I call an "anchor," and we plot the first time point and then the follow-up viral titer, and we draw a straight line here, we can use that as an anchor for which to estimate time from our other donors that have been included in the study for which we have follow-up.
So, if you plot each of the index donation viral loads here on this slide, you can use this, then, to determine time, so these are the follow-up virus samples from these donors and the IgM samples. So, you can create a pretty much makeshift so far, this is based on arbitrary time using our anchor time, but you can see about a seven-day viremic period following by another eight or so days where virus and IgM coexist.
This just adds more data to that slide to show the pattern continuing.
We have done a number of studies in 2002 that need to be reviewed. We have done three, a study of the case controlled plasma that was investigated as part of the CDC investigation, that is published in the Pealer paper, that is included, a total of 384 plasma that were associated with transfusion transmission last year. Out of those were a subset of the 16 that were published in the Pealer paper, and I will go through those.
We did a large prevalence study of about 7,900 samples in association with CDC. We looked at the first two weeks of September in two high-risk regions, Detroit and Cleveland. Testing was done using their TaqMan high input assay.
In the protocol, we took anything that was CDC-positive, and according to the criteria, it had to be positive times 3, anything that was equivocal, which was positive times 1, or anything that had an elevated negative.
Now, positivity here was based on positive PCR at 37 or less cycles, so CDC continued their PCR amplification to about 45 cycles, and anything we saw positive between 37 and 45 cycles was included in confirmatory testing, so, we took out of the study 173 samples for further testing.
We also did a larger study in association with the NHLBI and Gen-Probe, a prevalence study that included these samples in association with a total of 48,447 samples. That testing was performed at two Red Cross sites earlier this year, prior to the implementation of prospective screening.
We used Gen-Probe investigational reagents. Out of that study, we had 190 initial reactives that were investigated per the confirmatory algorithm, which was, because we only had a small amount of sample, our goal was to conserve volume.
We tested the samples, anything in this 173 or this 90, we tested at dilutions of 1 to 2, 1 to 8, and 1 to 16 by the Gen-Probe test to see if the reactivity could be detected in dilutions since all of this test was done neat.
We also tested each sample for IgM. If any dilution was reactive, it was sent for quantitative PCR. If dilutions were non-reactive, then, we only tested on diluted by the primary TMA and an alternate TMA developed at Gen-Probe using different primers and probes.
This is the first study, the plasma donor study, and all the samples that are in white indicate those that were included in the Pealer paper as part of the investigation, so there were 11 associated with transfusion transmission.
The last column here is IgM with the exception of three for which IgM results were not available. All samples were negative at index for IgM, and the plus here indicates that they seroconverted with donor follow-up.
The two yellow samples here indicate other samples that were part of the 384, but not involved in transfusion transmission. They were collected as part of the series, and it happens to also be RNA reactive.
The pluses here for the Gen-Probe test indicate triplicate positive results at various dilutions. Here, we have neat, here we have 1 to 8, and here we have 1 to 16. So, really, the conclusions that we had from this were that the cases last year, that were involved in transfusion transmission, were all able to be detected by pools, even pools using dilutions of 16, and that all of the samples involved in transfusion transmission were negative for IgM at index. So, we have relatively high titer, IgM negative samples.
Now, there was one sample here that was involved in transfusion transmission, negative by CDC's IgM test, but then when we used the Abbott IgM test, we did get an equivocal result.
These are the IgM results. This is the one sample that I just mentioned, that was CDC negative for IgM and then positive by an IgM at Abbott, and also positive in two out of three reps by Focus.
These were the two other viremic samples found in that study, both also IgM positive, one with very high titer IgM positive, which was a follow-up, a pheresis donor whose previous donation did transmit, so this was the follow-up donation that was strong IgM positive.
These were three other IgM positives found at CDC and also confirmed by Abbott, but with no reactivity in focus.
Moving on to the prevalence study, what we did with CDC was took the two highest regions, Michigan and Ohio. This was Cleveland and Detroit, and did the retrospective study of 7,915 samples tested at CDC individually.
The results of that, I told you 173 were further tested, 4 were clearly positive by CDC's test. They were all IgM negative, easily detected at CDC, reactive neat at Gen-Probe, and duplicate reactive at a 1 to 16 dilution with quantifiable viral loads. So, these are easy samples to detect, not a problem.
There was one other sample that was in the elevated negative category for CDC that is reactive, not at 37 cycles, but at 45 cycles, also IgM negative. It did react neat and easily at a 1 to 16 dilution, so again, not an issue.
But there were three IgM reactive samples that were not initially detected as positive by CDC, had very low copies of virus, or this one wasn't quantifiable and was also qualitatively negative. Of those, they were reactive neat, and you can see here lower signal-to-cutoff ratios. Only one was reactive in one rep, or actually in two out of three reps at a 1 to 16 dilution.
So, we can already see here from our first retrospective study that there are going to be some IgM-positive samples containing low level of virus.
So, when we put the prevalence data together, the overall prevalence was 1 per 1,000 in this study, Detroit being higher than Cleveland.
Then, when we did the larger study looking at multiple regions with Gen-Probe's reagents, out of the 90 reactives I mentioned, 46 were confirmed, and the total prevalence NAT study was similar to the CDC study of about 1 per 1,000. Of these 46, the samples can be grouped again into two categories, those that are IgM negative and easily detected in a pool of 16, there were 16 of those.
But then we had this other category that outnumbered this category almost by 2 to 1 of IgM containing samples that were low level viremic, and these two groups just divide them into the titer of IgM and the quantity of virus. So, we had 10 samples that could be detected at 1 to 2 dilutions, but no further, and had fair quantities of IgM.
Then, we had high-titer IgM samples that could only be detected by Gen-Probe's test, undiluted or neat, and actually 7 of 20 only could be detected as initial reactives and not repeat reactives.
This shows the breakdown. It's the same data, only showing the data over the regions, so Chicago had the highest prevalence followed by Detroit, Cleveland, and St. Louis.
If you put the data that we observed, which I just showed you, into a table using the data that was just published for the estimate of risk in these areas by Biggerstaff and Peterson, with the exception of Mississippi and Memphis, which when we did the study, the epidemic was over in the southern part of the United States, the point estimates from the CDC study or the larger study in both places of Cleveland and Detroit, or for the study as a whole, fits into the confidence intervals and fits almost identically into what was predicted by Biggerstaff and Peterson.
Looking at the three regions for which we had enough samples to plot data over time, this shows you the first week of September, 1 through 7 had the highest prevalence, because we really initiated the study through the month of September when the outbreak started high, but was coming down.
If you plot our samples over time trying to predict window periods--I just use this as a segue, we have our next slide--
--and this just shows, firstly, the number of samples further characterized. Here, we have the 16 that were IgM negative, high-titer, viremic, and then we had these 30 other samples divided into the first group of IgM positives, and then the higher titer IgM positives with lower virus.
So, if you add all of these time periods together using 6 1/2 from the studies of the 50s to assume that what we are detecting in pools of 16 would represent the same level of viremia that was detected in the 50s, we start with 6 1/2 days as our benchmark, and then we can do proportion of time for the number of positives we detected further in the study.
So, from our retrospective study, we would say that the IgM period that was also viremic was actually longer than just the viremic period in the absence of IgM.
I mentioned on the previous slide there were four samples of greatest risk because they had the lowest concentrations of IgM, but these four samples look similar to the samples I showed you from CDC, that they are detected neat, they are detected at a 1 to 2 dilution. We would only have sporadic detection at further dilutions, so even going down to a 1 to 8 dilution wouldn't really buy you anything o these types of samples. They all had quantifiable viral loads.
We are also doing some sequencing courtesy of Dr. Chang Fang, and of some of these samples that we are sequencing, interestingly enough, all the samples that have been sequenced that came to one region now fall into one very tight little group, and then our other sample from Cleveland falls into a different group. This will be expanded to see if we see any regional differences last year and we will continue this into this year's isolates.
I just want to end with a single-unit testing that we are doing prospectively at the Red Cross. It was initiated on the dates that I have mentioned, in Nebraska and Kansas, based on very high frequencies in these two regions.
We looked at whether we should initiate single-unit testing based on a trigger, but rather than a trigger, we really had to base it on capacity. So, we initiated individual unit testing as a feasibility study of logistics and yield on these dates. The testing is prospective. No products are released unless they test West Nile individual donation non-reactive, and we retain the pools that are made for HIV/HCV, so that we can compare the performance of neat and 1 to 16 dilution testing.
This just shows you the increase that we see, cumulative number of cases in Nebraska.
Same for Kansas, and we are starting to see a leveling off in Kansas.
Now, these are the results of the first 90 individual samples that we tested individually, and when we retested the pools, the pools were retested in duplicate, so the pool was repeat reactive, it is here and we have 42. We had 8 pools that were discarded, so they would not be consistently detected in pool testing, and 40 that were pool non-reactive.
But if you look at confirmatory data, and we don't have all the confirmatory results in as of yet, 29 of the 42 were PCR positive. I assume or expect all of these to confirm, but we had 5 here of these 48 that were not consistently detected in pools that were confirmed by PCR, and I am just going to end with showing you those 5 samples.
These are the 5 samples. The individual unit, S to CO's, you can see they all would have been classified as hot cases with the exception of this one. When we did retesting in duplicate on the pools, pools were sporadically reactive or not at all.
These were the viral loads at NGI, and we don't have all the IgM results back, but at least all ready 2 samples were IgM positive.
DR. NELSON: Interesting data.
Comments, questions? Paul.
DR. SCHMIDT: Susan, in your slide 31 there, two sources of virus from different parts of the country gave you I say genetic differences, now, does that tell us we have not only got different mosquitoes, but we have got a very rapidly changing virus, we don't know?
DR. STRAMER: Don't know. I just offer that as interesting data. There are only four samples, it is all very preliminary, but it's interesting that the three from Detroit all were virtually identical, and then the one from Cleveland was just a little bit different. So, as we test more samples, it may not prove to be anything, but it's just very preliminary data.
DR. SCHMIDT: Thank you.
DR. ALLEN: Just a follow-up, Paul, to your comment. Obviously, to be able to put this in total perspective, CDC or somebody I hope is doing similar genetic analysis on isolates from birds or mosquito pools and other large mammals.
DR. NAKHASI: Sue, I just wanted to follow up on that genetic diversity. I don't know what the data is because as you know very well in the literature, there are at least 5 to 10 percent differences between the remaining strain and the other strains, but whether that translates into differences in the infectivity, we do not know yet.
The question is how much difference, if at all, you know how much difference there is between the two groups.
DR. STRAMER: In the range of 3 to 5 percent. I mean it's not very different.
DR. NAKHASI: I just wanted to put that in perspective.
DR. STRAMER: They are always in the same lineage, it's just interesting to sort differently.
DR. NAKHASI: Of course, but it should not be alarmed that there are a lot of differences.
DR. KLEINMAN: Sue, do you know how long you are going to continue individual donation testing in those two regions?
DR. STRAMER: We are evaluating that now for when the number of cases decrease and when we can sort the noise from what would be real. Certainly, we have seen a decline, but even as of yesterday, we saw three, quote, unquote, "hot" cases out of Nebraska, so we can't stop testing yet.
I think when we go a week without seeing cases, we will probably terminate, but that has to be reviewed and we are doing that on a day-to-day basis.
DR. KLEINMAN: Thanks.
DR. KATZ: Sue, can you describe the proportion of your system that is being tested individually and give us some kind of sense for what this has involved in terms of source?
DR. STRAMER: When we were doing our calculations and when we were doing our systemwide testing all in pools, there was a split of about 95 percent in pools and 5 percent in single unit, single unit being resolution testing of reactive pools or just random, single units that are tested at the end of shifts.
Now, with just adding two relatively small regions that in total only collect 1,200 units per day from our entire system that collects 22,000 units per day, we have now shifted our pool to individual unit testing from 95 to 5, to 52 percent in pools and 48 percent in individual donation testing. So, it has had a huge impact on the operations, and we are at capacity the way we operate today.
We have had to add three staff for testing, one QC staff to St. Louis, and fortunately, we had trained staff from other NCLs that we could fill the void with, but beyond this, we are out of people.
DR. NAKHASI: Sue, in the last set of data which you showed, where you did the individual donation testing, and then you pooled them, and then you could not pick up on the pool? It was a pool of 16 or pool of 1 to 2, 1 to 8, what was it?
DR. STRAMER: The way testing is performed, we create a pool of 16, and from that pool of 16, both HIV, HCV, and West Nile are performed. So, when you do West Nile testing individually, you are still creating a pool of 16 to perform the HIV/HCV testing, so we retain those pools, the HIV/HCV pools, which would have been the pools that were tested for West Nile, but weren't, and then we save those frozen, and then when we get reactives individually, they are retained and tested.
DR. NAKHASI: So, a follow-up question is have you tested them in a pool of 1 to 2, 1 to 8, would you pick them up?
DR. STRAMER: No, we haven't done intermediate dilutions yet, but that is all part of future studies.
DR. NAKHASI: Thanks.
DR. STRAMER: But from the previous studies, 1 to 8 is a sporadic detection, it's just repeating the 1 to 16. It doesn't really make a difference unless you do a 1 to 2 dilution.
DR. NAKHASI: Yes, I was curious about that, you had tested in this sample, whether you picked it up.
DR. STRAMER: Well, we will do a 1 to 2, 1 to 4, and 1 to 8.
DR. NAKHASI: Thanks.
DR. NELSON: Actually, the two pool sizes or the two manufacturers are different, right, 1 to 6 and 1 to 16?
DR. STRAMER: Yes, but there are different testing systems, so pooling alone should not--
DR. NELSON: So, you think they are pretty equivalent.
DR. STRAMER: Yes, pooling alone should not be used to discriminate tests, I mean pool size, their sensitivity of the assay, et cetera, et cetera.
DR. NELSON: Right. Thank you.
Next is Mike Busch, Dr. Busch.
DR. BUSCH: Thank you. I am representing Blood Systems and particularly Sally Caglioti who manages the laboratory.
At Blood Systems Laboratory, the summary numbers are here, so 380,000 donations were tested from July 1st through September 6th. You can see the initial reactive 202. Of those, 167 were repeat reactive. As you will see, all but one of our repeat reactives have subsequently been corroborated as truly positive. This is simply the proportion of those repeat reactives that have been unequivocally confirmed. Again, all but one of the ones that we evaluated were confirmed.
There were 34 confirmed non-reactives and then 33 pending, so the rate of presumptive positives or repeat reactives is about 1 in 2,000.
This just then looks at the 202 initial reactives. Of the 202 initial reactives, 167 repeated reactive, our algorithm calls for an initial singlet Gen-Probe TMA assay coming out of the pool, of course, and then if that single sample is reactive, it is repeated on the very same assay, and 167 repeated reactive.
Of those 167, 119 have been confirmed, 1 was confirmed non-reactive, and 47 are still pending, working their way through this full algorithm.
I am not going to go into all the detail, but of these 119, only 97 could be confirmed initially on the index donation by the alternate PCR type assay, the confirmatory assay. An additional 7 on the index donation were alternate PCR assay negative, but were actually IgM reactive, and then an additional 15 cases were only confirmed through follow-up, so they were repeatedly reactive on the index donation, but that reactivity could not be confirmed either on the confirmatory RNA assay, nor were they IgM positive, but on follow-up, the donors had evidenced seroconversion with or without viremia.
On the other side of the coin, there were samples that were reactive initially, but were non-repeatable on repeat testing, 35 of these, and importantly, we actually confirmed 11 cases here.
So, this is a clue as to what Sue summarized, that there is low level viremia that is actually being detected, even coming out of pools, that is not repeatable on the index assay, but on further testing either using IgM on the index donation or follow-up, again, 11 confirmed cases were identified.
This is the rates over time for Blood Systems Laboratory, so fortunately we beat the epidemic, if you will. We had zero pick-ups in the first week of July, then 3, then 10, and again, you can see the rates peaked at about 1 in 1,200 and have declined, so we had 9 that first week of September. The subsequent week, there were 6, so this has continued to drop off.
So, it looks like, at least from our system, the blood screening yield has substantially decreased and it looks like we are through the worst of the epidemic.
This is detail that you don't need to look at obviously, but this is all the sites that have had repeat reactive again of which virtually all confirmed. We have rates in our different regions that have seen West Nile yield ranging as high as 1 in 180, and other regions that obviously have much lower rates, and many regions that have had zero pick-ups.
Now, there are stars here representing the regions that were part of what we will call our sentinel ID retest project, the reflex testing that we have performed retrospectively. Those regions represent 72 of our total confirmed cases, which is 63 percent of all reactives were represented in the sites that we have conducted our retrospective and now prospective testing on.
The next series of slides will just show the rate of cases over time, so we are looking at the week of July 7th, which was the second week of testing. We got a few hits, as shown here.
A few more beginning to see some clustering in Texas, which led us to be concerned about that region.
Now, you are actually beginning to see, not only the cases in green, which were detected by minipool NAT, but the increment of the number of cases that were subsequently detected based on the retrospective ID testing. So, here, you see we picked up a case retrospectively that was released as minipool negative, but then determined later to be viremic based on individual testing.
Now, you see more of these ID-only, more cases, so activity in Texas, but beginning with more activity up in the South and North Dakota regions.
Now, the activity is really expanding up here. We are beginning to see units that were missed by minipool, detected by ID.
Again, just continued accrual of cases.
So, that is what we were observing into August and early September. I want to just now sort of describe, working on what Susan presented, the same sort of concept of this viremic phase followed by IgM/IgG conversion.
In our program, the viral loads of the yield cases, the average was about 7,000, the median about 7,000. There were some high viremic units, one as high as 700,000, which pulled up the mean. Of our samples, I think it's about 10 percent of the minipool yield samples actually have detectable IgM, so they are coming from this, if you will, back end of the viremic phase that is detectable by minipool NAT.
Just conceptually, we know the individual donation NAT assays have sensitivity of around 5 copies/ml, the 50 percent limit of detection, but we are running them, of course, at pools of 6 to 16, which results in a dilution and operating reduction in the sensitivity.
In the case of a 16-member pool, that would reduce the sensitivity to about 80 copies/ml. So, we can sort of break up the viremic phase into periods that would be detected by minipool NAT which we are capturing, and then the samples that might be detectable by individual donation NAT.
Now, I am aware of Sue Stramer's findings from the retrospective work from last year with the surprisingly high rate of low viremic samples, most of which were seroreactive that were observed, as well as the concern about the high minipool yield rates in our region.
We felt it was important to further study this question, so we developed and implemented a study that involved focused ID-NAT retesting of samples collected from high-minipool yield regions.
So, to accomplish this, obviously, we didn't have the capacity to turn on prospective individual donation testing, so we continued to test and release blood based on the minipool work. We implemented PPT tubes, which allowed us to freeze away the residual plasma from these donations in a rapid and effective way in linked tubes.
Then, what we did was to implement retrospective testing of that stored plasma as quickly as possible following the donations. Beginning on July 15th, in these sentinel sites, we started with the date of the first minipool-positive yield and worked our way forward.
We eventually tested samples through August 13th, testing the majority of available samples from those sentinel sites.
Now, if the donation was initially reactive, we immediately trigger retrieval of product, so did a voluntary market withdrawal of all of the components from the initial reactive donations. The samples were then, over the next day, retested by the primary TMA assay, were sent for IgM and IgG, and then alternate TMA assay was performed on those samples, and we triggered donor follow-up to determine if these were real or not by donor seroconversion.
We decided, in conjunction with FDA and CDC, given what we knew about the initial reactive non-repeat rate, that no recipient notification would be triggered until the case was confirmed with "confirmed" defined as either repeatable reactivity on the index TMA assay or presence of IgM or IgG, or alternate TMA confirmation of the index sample, or donor follow-up.
So, in other words, if it was initial reactive- only with no corroborating antibody or alternate reactivity, we did not trigger recipient notification or CDC did not embark on recipient tracing unless the donor came back and was documented to seroconvert.
With the exception of one case, all of those initial reactive-only, the donors have not seroconverted on follow-up, so recipient notification is not warranted. The CDC again manages the recipient follow-up component.
So, here are summarized some numbers earlier that were based on a preliminary, what at the time had been called Phase I component of this study. The study is concluded, we are not embarking on any further retrospective testing, and as you will hear, we have converted what resources we have for individual sample testing to prospective screening in these high-risk regions.
So, we ended up testing 11,156 specimens that had been screened as minipool negative by individual donation samples, 43 of these were initial reactive. Of those, 10 were repeatedly reactive by individual donation testing, and 33 were repeat non-reactive.
All 10 of these repeat reactive have subsequently been corroborated as real based on either donor follow-up or alternate TMA or serology. Importantly, of these 10, 4 of them were IgM positive and 6 were negative, so they sorted into sort of front end specimens that had reproducible RNA detectable by ID that had been missed by minipool, "front end" meaning that there was no antibody present, and then these other ones which were similar to the large number that Sue observed of reactivity that was in the presence of antibody, repeat reactivity.
Of the 33 that were initial reactive, but non-repeatable, we ended up identifying 6 that were real, so the majority of these were nonspecific reactivities, initial reactive, very low S to C's, and on donor follow-up, a number of these have been followed and determined to have not seroconverted.
But we did among this group, which were exceedingly low viremia, only detected even erratically by the ID assay, we identified 6 true positives, 4 of them were identified on the index donation because they had IgM/IgG again. These were similar to those very low viremic back end cases that Sue saw so many of.
One donor, on follow-up, was determined to have seroconverted, initial reactive only. We were not able to get, because of volume limitations, a serology status on that particular donation. Then, there was one additional donor who was initial reactive only, negative by antibody, but on alternate TMA had reactivity right at the cutoff, so there is a very low viremic front end infection.
So, we had a total of 16 infected donations identified out of these over 11,000 units that had been screened and were negative by minipool.
This just shows by the sentinel sites, if we sort of focus on the bottom line here, within these regions during the overall time period in question, there were actually over 15,000 donations collected, so we only tested 11,500 or so.
There were 59 minipool yield pick-ups, repeat reactives in those regions for a rate of about 1 in 250, so this was how we focused in to do the ID-NAT testing based on minipool yield and were able to effectively test a sample set that had a minipool yield rate of about 1 in 250, and we picked up 16 additional infections, so we had a total of 75 viremic donations.
Because we didn't test all the samples, this, if you will, denominator of minipool yield needs to be adjusted down by dividing 11,156, which is the number that was tested over the denominator, so in essence, in picking up these 16 yield ID-only, there were the equivalent of 42 minipool positive cases during that period.
So, what we have conceived of as just kind of staging this early infection phase based on the kind of data that we are seeing and then we can populate these stages with the samples that we have picked up, so we talk about a stage I, which is such low level viremia that even individual donation NAT erratically detects it, so it is kind of a sample that is initial reactive, but non-repeatable, no antibody, but proves to be infected based on either additional testing of that sample or follow up the donor actually proves to be infected.
Then, we go to stage II, which is this period where individual donation NAT can detect the infection consistently, no antibody. Then, you go into the minipool yield stage, which we call stage III, so minipool picks it up.
Most of these are IgM negative, but there are some samples, as I mentioned, that are IgM positive and very low viremic. We are kind of dividing stage III into three phases, so there is a stage A, which is extremely low viremia, non-quantifiable, often negative by the confirmatory PCR assay that we use, very low viremia, but no antibody; stage B, which is the majority of the minipool yield, which has fairly high viremia, no antibody; and then stage C, which is very, very low viremia again, below quantitation, but with antibody, with IgM. This will come up a little bit later.
Then, we go into stage IV, which is antibody is present usually IgM-only, no detectable IgG, the low titer IgM such as Sue alluded to. These are detectable reproducibly by ID-NAT, but were missed by minipool.
Then stage V, which is again full conversion, often with IgG. In our work with CDC, wherever we see IgG, there is good plaque neutralizing activity, and in this phase, even the individual donation NAT is erratically positive.
So, what we have done then is to take these stages and then take our observed yield from our sentinel retest project. Again, we had 42 minipool yield cases or adjusted yield of 42. We picked up 1 sample in this very low level viremic, erratically detectable by ID-NAT, totally seronegative, 5 samples in stage II, which were reproducibly ID-positive, minipool negative, no antibody; 6 samples in stage IV, reproducibly ID-Nat positive with IgM; and 3 samples in stage V, which were erratically detectable even by ID-NAT, but had ample antibody.
For stages I and V, because these samples were only detected on 1 of 2 reps, and the initial screening was only done in singlet, just for a first step, we figured that, well, you must have missed 1 for every 1 you picked up, because had you screened the donor pool in duplicate, you probably would have picked up another one because of the low level viremia.
So, for just modeling purposes, we doubled the number of samples in stages I and V, which is illustrated by the parenthesis.
This is sort of populating the model, so the focus and sort of driving the model is our yield of minipool positives. So, we picked up 42 samples in stage III, which is the observed minipool yield.
The CDC prior estimates were about 6 1/2 days, and Sue's very nice new modeling work on her yield dynamics seems to confirm that there is about a 6- to 7-day, or 6 1/2 day for this purpose, estimate of the duration of the minipool positive window phase.
What we can do is then take the relative rates that we pick up samples in these other stages, the very low viremic, seronegative, the low level, but detectable by ID-NAT, minipool negative, seronegative stage II, and then after conversion after the minipool yield phase, you get the very low viremic, but IgM positive, and then the erratic viremia.
So, by then simply factoring the relative frequency of seeing these donations in these later stages to the yield of minipool NAT, we can estimate the length of these subsequent stages by simply taking 6.5 days and multiplying that by 2 over 42. That gives 0.3 days, 0.8 days, 0.9 days, 0.9 days for the lengths of these stages.
The other thing we can do is to assume that you have a minipool yield in a focal region at different levels, so 1 in 50 or 1 in 500, et cetera.
Knowing the duration of this stage, you can actually calculate the incidence per month that must be taking place within the donor pool in order to yield viremia at that rate, but more important for this purpose, you can estimate how many donations would be currently missed by minipool NAT, but detectable by ID-NAT in each of these stages relative to some assumed rate of infection in the donor pool.
So, for example, if the rate in the donor pool is 1 in 500, we would estimate that there would be about 1 in 10,000 donations in stage I, 2 per 10,000 in stage II, et cetera.
Now, what we know I think very clearly from the work of CDC and Red Cross last year, the paper in The New England Journal is that these kinds of units are transmitting, there is no question about that.
We also have transmissions in an NMWR that come out today absolute confirmed case that here is summarized from CDC. Thee is a second, virtually assured case that has been documented where ID-NAT is reproducibly positive, no IgM detectable, was missed by minipool. So, we are quite certain that transmission can occur with these units.
These very low level erratic viremia is barely detectable, no IgM, or in the presence of IgM, if there is reproducible viremia, but not yet any IgG, no neutralizing activity, these may well be infectious, but on the other hand, these very low level viremics in the presence of high titer IgM/IgG neutralizing activity, probably represent residual exceedingly low level virus in the setting of a mature immune response. Neutralization is present.
We didn't see any transmissions last year despite these being observed now fairly commonly, so we doubt these are infectious.
So, this sort of puts together the risk estimates as we have considered it, in a summary mode, so we have this minipool yield phase, 42 cases, 6 1/2 days long.
We had this stage II, which is probably infectious. These are the reproducible ID-NAT positives. We have confirmed one and probably a second transmission from such units, estimated at 0.8 days.
Then, we have stages I and IV which are possibly infectious. We had 8 cases estimated at 1.2 days. So, in terms of incremental pick-up, we would increase pick-up by about 12 percent relative to what we are getting in terms of stage II, and 18 percent in stages I and IV.
Really, this is sort of the bottom line. I think we are confident that minipool is picking up at least 75 percent of the infectious units. We are probably missing something in the range of 10 percent of probably infectious units by minipool screening.
Of course, this is all relative to the yield in each focal region, and it is possible that there may be as much as 25 percent total if these very low viremic stages are also infectious.
So, one thing we have done, again at the urging of FDA, is to take advantage of the fact that we have plasma unit volumes of specimens from these various stages, so we have developed a panel and have now gotten commitment from all of the companies to participate in a study.
These units are being compiled right now and will be distributed next week, a 20-member panel. Each organization will get 15 ml. Now, we actually have large volumes frozen away, we have multiple 50-ml aliquots, as well as 1 and 2 ml aliquots, which will allow us to go back and look at subsequent improvements in these assays, but also allow us to do infectivity work in both in vitro and animal systems, and we are beginning to talk about these studies.
These units represent samples that were collected either--and you will hear later about the ABC REDS plasma recall study that we identified several units that were from stage V.
From the BSL ID-NAT Reflex study, we have units available from stages I, II, IV, and V, and then we have also selected very low level viremic units, as controls if you will, from stage IIIa, which was that very early phase, so these were picked up by minipool NAT, but had extremely low level viremia, no IgM, or stage IIIc, which also had extremely low level viremia, but had IgM. So, these are serving as low level viremic positive controls and then negative controls.
We have gotten commitments from each of the companies to receive this coded panel. Gen-Probe will test with their both primary TMA assay. All this testing will be done in 10 replicates by all the assays, so they will test neat at 1 to 8, as well as 1 to 16. Roche will test in 10 reps neat and that they are operating 1 to 6 dilution.
Because of the issue of the relative sensitivity of the screening and confirmatory assays, we will also send this panel coded to the Chiron/Bayer group which has the confirmatory PCR assay that is used in the Gen-Probe IND, NGI, which does the confirmatory work for Roche and for the Red Cross, and Gen-Probe, which has this alternate region TMA assay that could potentially serve as a confirmatory, as well.
The final slide is just what Blood Systems has done in response to this data, and again in close discussion with FDA, CDC, the rest of the blood industry, we initiated a withdrawal of plasma from these high minipool yield regions, and eventually withdrew in the range of 1,500 units from these regions.
We continue to perform and release products by minipool everywhere in the system with the exception of several regions that have evidenced continued high minipool yield.
In deciding how and where to apply the ID-NAT testing, Sally Caglioti and her team did an evaluation of what the capacity was to add ID testing given both reagent availability and technical and space resources, and determined that they could do about 1,600 donations additional per week.
This more than doubles the number of tests that Blood Systems is currently performing relative to what has been done on minipool testing.
There is a weekly review of the minipool and the yield rates from the system in order to target the test capacity to the sites that have highest activity, and that has been initiated in the first week of September.
Now, our decision was also that if a region that initially was determined to warrant ID testing did not have any yield cases detectable for two full weeks, that we would revert back to minipool testing for that region.
This has, in fact, been the case, so whereas we were seeing early experience of yield in Texas, by the time we had turned on ID-NAT, that region no longer warranted the ID-NAT prospective testing, so we did not initiate it. It was only begun in our sort of Dakota type regions.
Then, of course, we are working closely with CDC to follow these donors and recipients. A total of 11 products were identified by the ID-NAT project for which the units were transfused. The majority, especially early on, were retrieved and not transfused.
There were a total of 12 recipients of these 11 units, and from that we have 1 and probably 2 infected recipients, several other, probably viremic transmitting donations where the recipient follow-up is not yet far enough along by CDC.
But then of the other cases, the other 8 cases, 4 of them were initial reactive, the donors did not seroconvert, there were false positives, and recipient activity is not going to be pursued. Then, there is 4 others which were initial reactive, no antibody, and the donors are still not back to make that determination.
So, there will be very limited, but important I think outcome information as to the infectivity from these look-back studies, and then the infectivity animal work, I think will be the other piece of how much of this low viremia is transmitting.
DR. NELSON: Thanks, Mike, that was really very interesting data.
Any questions or comments?
You kind of worry about outliers in terms of having a longer viremic period and maybe delayed IgM response, and what you have described, I think is a good model, but it looks like the outliers, if present, may not be common.
DR. BUSCH: Yes, I think our data mirrors Sue's and I think there is also beautiful data from the Bonfils Blood Center where they have had an incredible yield.
The follow-up data on these minipool yield cases is consistently demonstrating rapid conversion to antibody. What has been surprising, though, is the also consistent observation of persisting very low level viremia detected by the very sensitive ID-NAT tests for up to several weeks after the IgM has kicked in and the IgG.
CDC, a year ago would say you will never see both IgM and RNA in the same sample, but that was because the assays were relatively less sensitive, I think particularly the RNA tests, so now we are seeing this low viremia.
DR. NELSON: It appears, though, that IgG antibody is maybe a marker for clearance of even low level viremia. Can we conclude that?
DR. BUSCH: Obviously, the clearance of viremia is happening early, even, in fact, before we can detect IgM, so something else is helping mediate a nadir response, but what is clear in the work that I have seen from CDC and the samples we have sent them, is when we detect IgG, CDC detects high level neutralizing activity.
Even in the setting of an extremely low level viremia, there is 2 or 3 logs of neutralization capacity beyond that low level endogenous virus.
DR. NELSON: I was thinking, not so much of a mechanism, but a marker, you know, could IgG be a marked for non-infectivity, and IgG clearly can't from your data.
DR. BUSCH: Although again I don't believe--Sue, I think has one suspicious case, but no one has really confirmed transmission from a viremic sample with IgM in it, even in the absence of IgG.
Dr. : That was excellent, Mike, and so was Sue's presentation. I just noted one discrepancy in your conclusions, if I understood it correctly, and that was Sue seemed to conclude that IgM persisted for about 12 days. I guess that was IgM with viremia might persist for up to 12 days, and you concluded that if persisted for 0.9 days, so I don't know which is correct, but I think it just illustrates that we are modeling and we don't really know precisely what these time periods are, unless I have misinterpreted what either one of you have said.
That's a fairly big discrepancy and it just shows that we need more data to find out how the modeling proves out.
DR. BUSCH: Yes, and I think eventually kind of integrate. One of the differences that kind of led us to say we need to do this study very early in the epidemic is Sue's results were based on samples collected in September of last year, sort of on the back end, if you will, of the epidemic.
So, you would have predicted that she would be more likely to see people with that tail of waning viremia in the setting of seroconversion because the epidemic was more mature.
She, just by cross-sectional analysis, would be a little more likely to see back ends, whereas, our study focused on the earliest period of observable minipool yield within these regions, so our study was more front-end focused.
DR. STRAMER: Part of the reason I showed the 2003 data with the anchor donor was because we are reproducing what we saw last year with this year's donors when we plot them according to time, that we are seeing about a 10 to 12 day in follow-up of both viremia and IgM occurring in our 2003 donors. So, that number is almost virtually identically reproducing last year's data.
DR. BUSCH: And I agree, and we are seeing the same. When we follow our donors, we are seeing a tail of low viremia in the setting of antibody that is in that ball park. So, I think that phase is longer. Why we didn't see more of it in our study is perhaps related to the fact that we focused our testing on the very early stages of the infection coming in.
DR. : Well, not only that, but I think it points out that although the data is very good, it is still based on small numbers of observations, and obviously, you couldn't show the statistical variation, but to get precise percentages on such small actual detected number of donors, you know, there is room for statistical error.
So, I think we just should go away with the message that we have a good model, but we don't fully know how to populate it yet.
DR. NELSON: Has anybody tried to culture virus when IgM is present as opposed to PCR, because you made the comment that there has not been transmission, but clearly, the risk of transmission would be lower, but the issue is, is this just PCR DNA that we are detecting, or is actually they are culturable virus.
DR. BUSCH: We have had a lot of discussion obviously with CDC and FDA about this issue. I mean again until this year, they never had a sample that was thought to be viremic and IgM positive.
Now, we are getting lots of those samples. We have large volumes to look at this question, but the tissue culture systems are not sensitive below 400, 500. The reason they say 400 or 500 copies per plaque-forming unit is that is the lower end sensitivity of those assays.
These samples have something in the range of 5 copies/ml of viremia. I think the comments from the plasma industry earlier and use of that TCID or, in this case, plaque-forming unit per copy parameter to estimate infectivity, we have transmissions documented from units that have something in the range of 10 genome/ml, they infected humans.
Those are red cell components with maybe 20 ml of plasma, so those units, assuming plasma viremia represents total infectivity, may only have 200 genomes in them, and yet they are infecting. So, I don't think I agree with Jay's comments. You cannot use the TCID type parameter to judge human infectivity.
DR. NELSON: Hira.
DR. NAKHASI: Mike, even though they are trailing maybe 12 days or so, have you seen any of those cases transmitting?
DR. BUSCH: No. As I said, we have 12 recipients of products. Now, of those, 4 of them definitely did not get infected, there is no reason to even chase them, 4 of them are pending, so there is really only 4 units that are relevant. Two of them were minipool-negative, ID reproducible positive, antibody-negative, and transmission occurred.
We have two more. One of them has antibody, went into a patient, we will see, and one of them does not and is borderline reactive in two reps, not antibody. So, we have got two recipients, if you will, who will be very informative, but it is only two.
DR. NAKHASI: That is what I am saying, but at least that will give some information there, and obviously, the bottom line is in doing the infectivity studies with those samples will really nail it down.
DR. BUSCH: Right, and, you know, there is good discussion about other primates, et cetera, and we have the units that are human pedigreed transmitting that will let us test the sensitivity of those models. If we can't transmit into an animal or tissue culture system with stuff that we know infected humans, then, the model is not sensitive enough.
DR. NELSON: Maybe a crow would work.
DR. BUSCH: And you just can't put much volume in.
DR. KUEHNERT: Matt Kuehnert from CDC again. Nice presentation. I had a question, sort of a follow-up on the observation about culture really being the gold standard, if that is possible to do, but I was thinking about some of these sort of possible cases at the fringe stages at 1 or 5, where you are using IgM as the gold standard, and you have a situation where there are some activity, say, last year in the same region or where there is high activity going on, how to differentiate between an infection that may have happened before the donor donated.
Is there any way to differentiate that by other means? I guess I was thinking there has been some discussion about other immunoglobulins like IgA, and I just wondered if there are any samples being stored for those purposes for further testing.
DR. BUSCH: Well, yes, I mean there is both the kind of unit volumes that we have picked up cross-sectionally either from minipool screening that had been further characterized or from the ID retest project. There is also just beautiful follow-up samples now in the freezers from these donors, and I think we will see data later today about the duration of that IG and window, I think is being refined, so I think there is an opportunity obviously to better characterize West Nile pathogenesis using these donor materials.
DR. KUEHNERT: Because I think the IgM, you know, being positive, I mean in a worst case scenario, you have to assume that that is real, but also it could represent distant infection.
DR. BUSCH: Again, you know, the CDC has published work that suggests that IgM can persist in frank meningoencephalitis cases for as long as a year or more. I think we will see data later from a focus that shows that in their work, it doesn't seem to persist that long, probably two months, so probably what we are seeing is IgM reactivity in 03, particularly in regions that didn't even have West Nile a year ago is reflective of recent infection.
DR. KUEHNERT: I think it will get more complicated in future years definitely.
DR. NELSON: Thanks, Mike, that was very interesting.
Dr. Smallwood has an interesting and relevant announcement.
DR. SMALLWOOD: I have been informed that all flights have been canceled for the rest of the day, but it is our intention to move through the agenda and to finish early because the FDA staff does need to try to get home since we can't stay in the hotel.
So, we will still proceed with our original plan to try to finish as soon as we can, but so that you will be advised if you need to make other arrangements, that all flight have been canceled for this afternoon.
Next, is Dr. Strong.
DR. STRONG: As requested, I have multiple conflicts to divulge although some of those have already been divulged earlier this morning.
In this case, I am an employee of the Puget Sound Blood Center, and my travel was paid for by the FDA, and I will be presenting Roche data, so I think that covers most of my conflicts.
Actually, the Roche data will be presented in two parts. I will give you kind of an overview from an operational perspective.
As everyone knows by now, the Roche system was brand-new to our laboratories that are doing this particular assay. We switched to a completely different platform using the TaqMan system compared to what we were using with the other PCR assays.
That required, of course, completely new training, new equipment, making space for these. This piece of equipment here occupies about 12 feet across, so in the laboratory, such as ours, we have had to place three of these into the space that we had, which was no small matter.
These are the various information systems that operate to make this an integrated system.
These are the laboratories that are running the Roche system throughout the country, and we have included Canada in this system, as well. So, we have 14 laboratories with approximate volume of roughly 4 million which includes the Canadian laboratories, both HemaQuebec and the Canadian Blood Systems.
The rollout of these systems took into account the potential for infections early in the season, so the Midwest, Canada were implemented earlier than the others, so implementation started as early as June and completed, for example, for our particular laboratory, the 7th of July.
But everybody was up and running with approximate targets dates of the 1st of July for this epidemic.
This is just an example of what the TaqMan PCR provides to us. It is a real-time PCR, so that we can actually get data based on cycle times, and the readout really will be provided in terms of positive cycle targets.
This is the growth curve for the negative control and anything that shows up down here in the growth curve would invalidate the assay. This is the positive range, and this is a typical positive sample, and it shows the growth curve through amplification cycles, and a positive sample would give you a reading something like this.
This is the internal control. Once again, you would have curves generated from the assay with invalid rates if we had curve growth on this end, or if we had an internal control that was negative down here, it would be an invalid run. Again, the typical internal control growth cycle with a positive and this kind of value.
Here is the yield cases including both Canada and the U.S. in the Roche sites, and it is really the same pattern that has already been described by the previous two speakers with most of our cases running right up through the middle of the country.
Early on in the investigation, the cases were first described in Texas. More recently, in the last two or three weeks, we have seen a growing number of cases in the same region that has been reported, South Dakota and Nebraska, et cetera.
So, in terms of the Canadian results, they have screened about 250,000. They have 14 positives. If we break those down, there were 13 minipool positives, 12 reactive individuals from initial minipool, and 1 false positive due to an instrument problems.
Mean of the problems in the start-up phase for these laboratories have really been related to this new instrument and working out the parameters. These were instruments designed primarily for diagnostic purposes, and we are putting them through a real burn-in phase here with the volume that it is going through.
In terms of the reactive individuals, there is some individual testing going on now, particularly in Saskatchewan, where there have been higher numbers of cases and at this stage, we have two reactive ID-NAT, and both of those have been repeat positive.
In terms of the U.S., there have been 70 positive pools, of which 58 have confirmed. We have 8 that were not resolved to individual tests, and we still have 4 pending. There are also two centers that have begun to do some ID-NAT testing, Minnesota and Texas. There have been at this stage four individuals that have been picked up as ID-NAT positive, three pending, so we don't have confirmation of those, one that is probably going to be a false positive.
601,000 samples have been screened at this stage, so we have 58 that are, in fact, positive. You will see here a few sporadic cases on the East Coast. New York has 4, we have 3 in Florida. This was the hot spot last year, but now obviously, again, as has been described, everything has moved west.
This just summarizes for each of the individual centers what the results have been. This is Minnesota that has had the bulk of the cases, but more recently, initially, it was in Texas, this is Gulf Coast, who had 18.
We have 70 positive pools from this group, of which 8 have not been resolved in individual testing. This 62 includes the 4 that were on the previous slide that were still pending, so we have essentially 62 positive out of the 601,000 on the U.S. side.
That is the end. Now, the follow-ups on those particular donors will be provided later on.
DR. NELSON: Thank you, Dr. Strong.
Questions or comments?
Next is Dr. Jim Gallarda.
DR. GALLARDA: What I would like to do is follow up with some of the further testing that we have done in the clinical trials so far after Dr. Strong's presentation.
This is an example of the type of follow-up testing that we are doing in the clinical trial. I have just shown five examples from the interdicted units. We have an initial minipool result that identifies the individuals and then in order to say we have what appears to be a presumptive positive, we rely on at least a second test. In many cases, that comes from the plasma bag itself with the same test or with an alternate NAT assay.
We are also doing, for the first several samples that we have identified, sequencing, and to date, I think we have about 10 or so sequenced, and these are all confirmed to be West Nile by sequence analysis.
In some cases, if we have the follow-up samples coming in, we will do a follow-up on the initial screening assay, so you can see the ranges for four of these donors. They are negative upon follow-up, and this is by individual PCR. In this case, we see it was negative on day 2 after the initial index donation, and it was positive at day 9.
We also do quantitation, alternate NAT and quantitation using NGI. In this case, NGI confirms the index donation. Quantitation ranges in this case from 33,000 for this individual on the index bleed to 200 copies/ml on the index bleed.
We are also doing IgM and IgG and plaque neutralization, and for this donor we have a positive on the subsequent two follow-up donations.
This is a summary of statistics for 33 of the interdicted units. This is a histogram showing the distribution against viral titer. The mean for these 33 is 43,000 copies. This is not a normally distributed population, so the median is better representative of central tenancy, that is 12,000 copies. Our highest one was around 410,000, and the lowest one was at 200.
If you just look at the 95 percent confidence interval for this population, it ranges from 14,000 to 71,000.
So, no surprises. West Nile, as everyone expected, is a low-titered virus for the most part.
We have participated in a couple of retrospective studies. This one, Sue Stramer had talked about previously from September, I believe, of 2002, a collection of 383 frozen plasma units that were tested by CDC, as well as IgM.
At the time, this was presented early on, CDC's summary results were 10 were plasma reactive and another 2 were segment reactive. Of the 10 plasmas, 2 of them were non-reactive with the standard PCR. It required a high-input PCR to demonstrate reactivity.
Our summary of blinded testing is that 371 were negative, 12 of those 383 were positive as individual samples, and 10 were positive when diluted 1 to 6.
This is a table showing the outcome of the test results, so all 12, when tested as individual testing, were positive. Two of them, however, were negative, and for this tube No. 28, we actually tested it in four replicates and failed to detect it on all four.
Sample No. 281, we tested it--and this is at 1 to 6 again--and it was positive in one of three replicates with a rather delayed growth curve CT value.
We have also been testing Mike Busch's REDS 2002 population, which is still in process, we are done yet, but this source from roughly 60,000 components that were recalled in the summer of last year, 1,400 of these were sent out to various labs.
We have tested 494 to date, and we found 1 positive tested individually, 593 are negative. We are doing the continued testing on the individual samples, as well as 1 to 6 dilution. So, hopefully, our next meeting we will have closure of that population.
We have several dedicated individual NAT studies going on right now. As Mike referred to, CBS began a limited ID-NAT study for samples collected in Saskatchewan beginning September 2nd. John Saldanha, I believe will be providing some information in greater detail for this study.
To date, 1,355 samples have been tested. Two samples were identified that were ID-NAT positive and minipool negative. Both of those samples are positive by IgM. We were able to get volume from CBS, and we found out, in fact, that with replicate testing, we are getting about a 20 percent hit rate.
So, these appear to be extremely low titered individual samples consistent with the IgM status, which is controlling the viremia.
If we look at the replicate testing results and we plot that against the historical growth curves and Poisson distributions, those are less than 10 copies/ml.
Memorial Blood Center in Minnesota also began individual NAT testing for a select portion of their samples collected in Nebraska. They began on September 4th. To date, they have tested 690 samples. They have found 2 that were identified as individual NAT positive. They are only testing this particular population by individual NAT, so they do not have corresponding data on 1 to 6 dilutions, however, we will get those and we will be able to report those later.
Memorial also has a possible transfusion-transmitted infection from a West Nile negative minipool that is being investigated. This was a Nebraska donor and I believe that data will be provided on this situation sometime later.
Finally, Gulf Coast Regional Blood Center in Houston, Texas, also began, in addition to their normal minipool NAT testing, in parallel, they selected their hot spots to the north of Houston, and they have tested 2,551 samples. None of those were identified that were ID-NAT positive.
This was a protocol that they followed in parallel to their normal minipool. During this time, they did find one donor that was minipool positive, and, of course, that was resolved to the individual donor as individual positive, so that formally is included in this population, but in a separate protocol.
Mike Busch alluded to a panel that he is putting together that is comprised of samples that are early on in the viremic stage of growth, as well as in the late stage. This will be the first time I think there will be a nice panel that is controlled and well characterized, and can be sent out to multiple laboratories to look at the effectiveness of minipool testing with whatever pool size compared to individual NAT testing.
So, I believe we will have at the next BPAC meeting an update on that from Mike.
So, our conclusions, as Hira talked about early on this morning, is that clearly, minipool screening in North America has been a success, and we have identified at least according to the model 75 percent of the West Nile positive units.
The titers are low, which is a departure from HCV and HIV, as we all know. FDA is working on putting together reference panels, and I think there is a lot of variation in what is claimed as a copy number from one technology to another, so there will be an important process to standardize test methods.
There is a discussion about, well, what do we do next summer for single-unit testing, so it looks like California is next in line, Mike, for the hot spot if the trends continue.
I think, you know, this morning one of Tony's slides said that we have not yet reached zero risk. I think most of us would agree, and we probably never will reach zero risk.
Clearly, I think one of the more important things is what is an infectious dose for West Nile.
So, that's my talk. Thank you.
DR. NELSON: Thank you.
DR. ALLEN: You had that one specimen that you identified a copy level of about 410,000/ml.
DR. GALLARDA: Right.
DR. ALLEN: Any further information? I mean I assume that that was duplicate tested and confirmed at that level. Was that an asymptomatic person? Do you have any information about where that sample came from, because that is getting close to the worst case scenario that we discussed earlier this morning.
DR. GALLARDA: I don't have the information here, but normally, NGI does the quantitation on that sample. The testing we have done with our in-house assays confirm it's a high-titered sample. Where that sample came from, I don't recall the location, however, it is consistent with I think one Sue had seen, which was around 390,000 or so. So, those are our top picks right now for high-titered West Nile.
DR. ALLEN: It does apparently come from an asymptomatic person.
DR. GALLARDA: Right. I believe it is from an asymptomatic person, correct.
DR. STRONG: There are a few donors, though, that have become symptomatic after donation, within the week after donation, and those tend to be the higher titered ones.
DR. GALLARDA: Yes.
DR. NAKHASI: Jim, thanks for a nice update.
A couple of questions. One, the study where you showed last year's study, you missed two of them.
DR. GALLARDA: Right.
DR. NAKHASI: Were they IgM positives?
DR. GALLARDA: That was a blinded panel. Sue has got the data, and I think she will be sending the database. I don't know if those were IgM positive or not. I think perhaps they were not.
DR. STRAMER: They both were not in CDC's hands. One of them was positive by both Focus and Abbott in additional testing that we did in one of the slides that I showed.
DR. NAKHASI: So, one was positive and one was not?
DR. STRAMER: One was equivocal, low level IgM.
DR. NAKHASI: Thanks. The other question is in this Nebraska case which you said, this was minipool NAT negative, 1 to 6 in pool.
DR. GALLARDA: Right.
DR. NAKHASI: What was the viral load in that, viral copies number, do you have any idea how many copies you had?
DR. GALLARDA: We don't have quantitation on that, so I don't know. We are working with the local Public Health labs to try to get volume to do further testing on it, because that information is really limited.
DR. NAKHASI: That would be very interesting to see how low we can go.
DR. GALLARDA: That appears to be what Mike described as early Stage I/Stage II type sample, but we need to get some volume to do testing on it.
DR. NAKHASI: Thanks.
DR. NELSON: I am impressed about how valuable and interesting the blood donor screening data is for an epidemiologic surveillance tool for this epidemic.
I wonder if anybody has looked at rates of positivity by age of the donor. One might predict if, in fact, the older people are more likely to be symptomatic, that they might be less likely to be seropositive as donors.
I wonder if anybody has explored other aspects of this I think very valuable dataset. Now, we have got, you know, huge denominators and also pretty good numerators for positivity.
Do you have any data on seropositivity by age?
DR. GALLARDA: In the database, we are collecting gender, date of birth, other demographics, so when the full picture is in, I think that will be a very powerful dataset to look at the epidemiology of this virus.
DR. BUSCH: The ABC program, which is virtually all of the ABC centers including Roche, Gen-Probe users are participating in a collaboration funded by NHLBI, where all of the NAT yield cases and all of the denominator data, that includes breakouts by age and gender and zip code, down to the zip code level, will have numerator/denominator and be able to look at these issues.
We will also actually be able to run monthly incident rates by region and determine accrued incidence rates really throughout the country for the epidemic.
DR. NELSON: Also, the data that was on one of Marfin's slides on seroprevalence by rural and urban areas seem to be dramatic, I think it was 10- or 100-fold difference between a population of a million to a population of 10,000. These data, it would be very useful to look at that, as well.
DR. STRAMER: We are also collecting, as Mike said, for the ABC Centers, we have set up a database starting last year when we did the 2002 prevalence study with all the donor parameters down to zip code, county of residence, and also Red Cross, and I know other blood centers, as well, are using a donor survey to look at symptoms in the donor, travel and the donor, et cetera, and we are doing that for all NAT reactives, so we will have a built-in set of controls and cases, so those can be compared against age, sex, you know, all parameters, viral loads, entomology, et cetera.
DR. STRONG: At this stage, the age range is 16 to 82.
DR. BUSCH: Actually, Debbie Todd just gave me the summary of the ABC Surveillance, and it's 563 reactives. This haven't all been resolved, but they cluster actually, you know, like 142 are in the 35 to 45, and 156 are in the 45 to 55, so it does look as if we are seeing yield.
I mean this doesn't have the denominators and rates, but we are seeing yield in a relatively younger group than I would guess at the overall donor pool.
DR. GALLARDA: Mike, how does that compare with the normal population at large?
DR. BUSCH: That would be hard, but will have to try to do that.
DR. NELSON: Next is Dr. Saldanha from the Canadian Blood Services.
DR. SALDANHA: I am representing Canadian Blood Services.
Canadian Blood Services is one of the operators in Canada, and we collect blood for the whole of Canada except for the Province of Quebec. In that province, the blood operator is HemaQuebec.
I am going to give you a summary of the data we have collected to date using the Roche TaqScreen assay for donor screening.
Screening was started in the center in Toronto on the 23rd of June, and in a second center in Calgary on the 2nd of July, and since then, we have been screening all the donations we have collected.
So far, up until the 15th of September, we have tested just under 200,000 donations, and the number of NAT positives using the Roche TaqScreen assay, which is tested in the minipool format, has been 14.
Eleven samples were tests in minipools of 6 and 1 sample was tested individually, and this was accidental, because if we have samples that are not sufficient to make a minipool, then, under the protocol, we can test them singly.
As Jim mentioned earlier on, we started single donor testing in a high-prevalence area and identified 2 additional samples, and I will talk about these a bit later.
All the plasma that was identified from the positive donations was sent back to head office and tested using an in-house quantitative assay based on the CDC assay. This assay will only pick up West Nile virus.
Twelve of the 14 samples tested positive and, by definition, these are presumptive positives, and the RNA titers, using the in-house assay, varied from about 2,000 up to 70,000 copies/ml.
One of the samples had a very low titer, about 30 copies/ml, but I think really should be less than 100 copies/ml, which is the limit of quantitation.
This slide shows a graph of the donations and the titers and copies/ml, so they vary, as you can see, from 31 up to 70,000, and the numbers on the side are the numbers of the samples collected in chronological order.
The last two samples, 18 and 19, which were tested using single donations, were really at the threshold of detection.
Most of the quantitation using the assay was done either in duplicate or triplicate, and this slide shows the sort of the standard deviation. The replicates were pretty good on the quantitation apart from the very low samples.
As we have mentioned, we have been doing a limited study to determine assay sensitivity using minipools versus single donor testing. In Canada, we have been testing single donations for a limited number of collections done in Saskatchewan, and these are products for transfusion. This was started at the beginning of September.
And why did we choose Saskatchewan? This slide shows the prevalence of West Nile virus taken from a Health Canada web site. In brackets are the prevalence data for 2002. It is obvious that the main focus of the infection this year is in Saskatchewan, Alberta, and Manitoba.
We have had a few cases in Ontario. Last year, Ontario was, in fact, the focus of the infection with 395 cases. We have had 14 cases, as I mentioned, of West Nile virus in the blood donors.
This slide shows you the human cases per million population, and in Saskatchewan, you can see that the incidence is very high, it is 223 compared with Ontario, which was 34 last year, which is where the peak of infection occurred in Canada.
This slide just shows you a map of the country, and most of the infection, which mirrors what is happening in the U.S., is based in the western provinces, and it is centered around Saskatchewan with infections in Alberta and Manitoba.
Last year, the focus of the infection was in southern Ontario, so like the U.S., the infection is moving westward.
I will talk a little bit about the single donor testing which we have done in the high-prevalence area. As Jim Gallarda mentioned, we have tested a total of 355 samples. Two samples were at the threshold of detection using the in-house assay.
We tested the replicate samples of these two donations, in fact, we tested the plasma units, and in both cases, roughly about 1 in 40 of the replicates were positive, so it is a much lower rate than was obtained by Jim.
Both samples, as well, were positive for West Nile virus IgM using a commercial assay, the Panbio assay.
The final slide just shows you the number of positive donations on a time basis. Again, like the data we have seen before, it looks like we have the peak of infection about mid-August to the end of August.
We haven't, since the first week in September, picked up any positive donations in the population. What I would like to say again is using these quantitative data, which are preliminary, I support completely what Jim said, and we really need to standardize some of these quantitative assays.
DR. NELSON: Thank you very much.
It is sort of very similar geographic data from Canada and the U.S. That is interesting.
DR. GALLARDA: John, I have heard there has been a frost up in Canada now.
DR. SALDANHA: Yes, we have had I think our first snowfall in one of the provinces.
DR. BUSCH: Just so I am clear, in Saskatchewan, when you converted to IDT, you picked up only 2 yield cases?
DR. SALDANHA: That's right.
DR. BUSCH: And both of them were extraordinarily low viremia, so there weren't, you know, a proportionately high number that would have been detected by minipool.
DR. SALDANHA: No, we haven't seen that at all.
DR. NELSON: Dr. George Dawson from Abbott.
DR. DAWSON: I am here today representing Abbott Laboratories.
Abbott will continue to work on West Nile really looking at the evolution of markers associated with infection, and in that vein, we have developed a research assay, an IgM-based test. We have set a cutoff based on human populations and experimental infection in primates.
We have tried to characterize the repeatably reactive specimens to find out how we should set our cutoff.
Many observations have occurred in the last several months, one of which there is considerable overlap of RNA and IgM detection. These often are low level viremics that would be negative by minipool and IgM positive, and you may need, as we have heard from other speakers, infectivity studies to determine if these type of virus can be transmitted.
As IgM tests aim to be more sensitive, there will be greater overlap with RNA detection, and these more sensitive tests will be more difficult to confirm the reactivity because there is no other test that could confirm it. So, we tried to address some of these issues today.
Two specificity concerns, and there are others, but there are two specificity concerns that I would like to just address briefly.
One is that West Nile virus has a high degree of amino acid homology with other flaviviruses. You are going to see cross-reactivity with St. Louis Encephalitis, Murray Valley Encephalitis, and Japanese Encephalitis viruses.
The second is confirmatory testing. The low level reactives will be difficult to confirm, and when we have brought this up, even to Dr. Rob Lanciati at CDC, he said, well, when you have got low IgM positives, you may often need follow-up samples in order to confirm those.
Here is a table put together by molecular biologists in my group wherein we looked at different regions of the genome of West Nile virus and you see here depicted the percent amino acid identity of West Nile virus with other flaviviruses here.
I am just going to point out to you the envelope the pre-membrane, membrane, gene products that are showing 75 to almost 80 percent amino acid identity between West Nile, St. Louis Encephalitis, Murray Valley, Japanese Encephalitis virus. There is a lot of expected cross-reactivity.
Here is a slide from the CDC web site actually showing four sera that were West Nile virus IgM positive, tested in a West Nile virus ELISA. The S to N values are shown here, showing they are strong in the West Nile ELISA, but they are also cross-reactive with Japanese Encephalitis, St. Louis Encephalitis, and in some cases even with dengue, keeping in mind that S to N of 3 or above is positive.
So, you will see as these studies go on in the next years, more confounding results as you begin to detect individuals infected with viruses other than West Nile virus.
Here is a depiction of our assay format. We have been making IgM capture EIAs for probably 25 years, so we have a lot of experience in how to make that quarter-inch polystyrene bead just shine. We have put a lot of work into that, and we put a lot of work, in addition, into the diluents and that sort of thing, so we have got some experience in this area.
We set our cutoff last spring, a kind of provisional cutoff of the sample to negative value of 5, and I am going to show you some recent data that we have obtained applying that sample to negative value of 5 cutoff to a volunteer donor study we did with Mike Busch and Leslie Tobler on experimentally infected macaques, a study we did with Dr. Tesh at the University of Texas Medical Branch at Galveston, and Marion Ratterree at Tulane University.
In this study, we were able to take daily bleeds for two weeks and to therefore look at the natural history of West Nile virus infection.
Here is a graph showing postinoculation evolution of marks of this monkey that was inoculated with 105 plaque-forming units of West Nile virus, New York 99 strain.
On the X axis are days post-infection, on the Y axis are the S to N's, sample to negative values for the IgM and IgG ELISAs we have. I will first point you to the top that we have done PCR, our in-house RT-PCR, showing detection on days 1 through 10 with the exception of day 8 being negative, and here is virus isolation data performed down in Texas showing that infectious virus could be detected at days 2 through 5.
IgM was first detected at an S to N of about 8 at 10 days post-inoculation, IgG first detected at about day 21. You can see in this case, it is probably different than in the human case, is probably in several areas.
One is the IgM seems to come up very quickly and decline. I am not sure we do see overlap of IgM and RNA, I expect that in the human cases we have seen described, there is a lot more overlap than we are seeing in this study.
Here is another depiction of that same data starting on the left with days post-inoculation through day 63. Here is the virus isolation in terms of the log of plaque-forming units/ml, so between 10 and 100 of plaque-forming units/ml at the peak of infectivity, days 1 through 10, day 8 being negative.
Here is the IgM values. I have bolded the IgM at an S to N of 8 as the first positive bleed date. It's IgG negative, and IgG, as you know, didn't come up until day 21.
Now, we have included also three other serologic tests. These are classic virological serum tests for determining antibodies to West Nile virus - hemagglutination inhibitor, complement of fixation, and plaque reduction neutralization testing.
You can see on the first bleed date we do get, while this is plus/minus by PCR, we do confirm it with plaque reduction testing.
Here is the second animal inoculated with the same dose. Infectious virus was only found on one day. RT-PCR was positive on five different days out to day 6 post-inoculation. IgM again comes up on around day 6, IgG at around day 14.
Here is a table showing the same data, 170 plaque-forming units seen on day 2 post-inoculation, PCR showing positivity here, the IgM being positive on day 12, and this sample was confirmed by the classic serologic tests of hemagglutination inhibition and complement fixation.
In this case, the plaque reduction was negative, but the HI and CF were positive. So, using a combination of supplemental tests, we felt that setting the cutoff where we set it is probably a decent place to set it.
Here is a composite of the five different animals that were tested showing the log plaque-forming units, the PCR detection. You can see all animals being viremic for the first 4 days and then dropping off in subsequent days.
This is the median day for the serologic tests, an S to N of 11.9 on day 10, and confirmed by HI or plaque reduction, so we felt that setting the cutoff where we did on experimental studies, an S to N of 5 looked pretty decent.
So, we confirmed with the supplemental assay 38 of the 41 positives. We did not confirm three of the first bleed dates on the macaques. So, we set the provisional cutoff of S to N of 5.
We did this study in conjunction with Mike Busch and Leslie Tobler where these were volunteer blood donations from 2002, from centers considered at high risk. Applying our S to N of 5, we see a pretty good separation between the negatives and the positives.
So, you can see when you hear someone say "equivocal," it is usually between 4 and 5. There is very few samples that usually fall into this range, and you can see that most of the bulk of the samples are far away from where our prospective cutoff is set.
There were 15 initial reactive samples and 13 repeatably reactive, 7 of these 8 out here were confirmed by plaque reduction neutralization tests at CDC. One was not confirmed. So, we have taken away the confirmed samples and looked at setting a cutoff up here, our specificity would be about 99.6 percent, which is not too bad at this stage.
We are doing additional testing on samples in this area. We have sent them under code to investigators who look at HI, complement of fixation, and plaque reduction neutralization testing.
Here is a depiction of those 13 reactives sorted from the highest IgM positivity to the lowest that were repeatably reactive, and we can see that they did repeat pretty similar S to N values.
We have a second ELISA we have developed internally that uses a different recombinant protein produced in drosophila cells, and we have again showed that our other internal IgM tests using a different recombinant protein confirms pretty much the same positive samples.
The first 7 samples were confirmed by plaque reduction, the next 6 were not. IgG was also positive in these first 7, so the supplemental testing took these 7 and called them as confirmed.
We are doing some additional tests on these remaining 6 samples that I think at least 1 or 2 of these may confirm.
Summarize. We fixed our assay pretty much where it is as a 3.5 hour assay, 37 degrees. We can see from our study that IgM was detected around day 10 post-infection in the animal study. A cutoff of S to N of 5 seemed to provide adequate specificity separating most of the negatives from the confirmed positives with a specificity of about 99.6 percent.
We are looking at potential specificity improvements. It may just entail moving the cutoff higher, but we are waiting yet to see the results of the additional supplemental testing we are running.
So, we will continue doing these research studies and using our tests to support some of the studies being done in the blood banks.
DR. NELSON: Thanks, Dr. Dawson.
Questions, comments? Okay. Thank you.
Dr. Jeffrey Linnen from Gen-Probe.
DR. LINNEN: I am going to give an update on the Procleix West Nile Virus assay. As was mentioned, I am representing Gen-Probe.
I want to just go over the topics I am going to cover. I am also going to talk about the performance of our assay in the ABC/REDS panel. This is recalled plasma components from 2002.
I will update the clinical performance we have been seeing so far and talk about the performance of the assay in a high-prevalence region, then finish by talking about some future options for this assay.
You have heard a little about this so far. We were sent 1,464 blinded samples, and this was a study that was organized by Leslie Tobler and Mike Busch, so these came from the Blood Centers of the Pacific.
Unknown to us when we were doing the testing, these included 5 positive controls from BBI at 10,000, 1,000, 300, 100, and 30 copies/ml.
The way this was carried out is we screened these individual samples with our screening assay and then reactive samples were confirmed with a Gen-Probe alternative TMA assay that I think has been mentioned in a few other presentations.
All of the initial reactives--this is actually a correction--were then tested at 1 to 8 and 1 to 16 dilutions to determine the efficacy of pooling, so this is still when the samples are blinded.
This has a lot of numbers and I guess some of the points that I wanted to make here have already been covered in Sue's talk. What I would like to focus on first is the one real reactive sample. Now, this is the one reactive clinical specimen.
So, you can see the neat result here had a high S to CO above 30, the repeat neat was similar. We were also able to detect 3 replicates at a 1 to 8 dilution, and 3 replicates had a 1 to 16 dilution. Similar results with the alternative TMA assay.
Now, this sample was quantitated at I think at least two sites, Chiron and NGI, and the estimated titer is about 440 copies, so you can see, at the 1 to 16 dilution, this would correspond to less than 30 copies, and we were able to detect this in all 3 replicates with relatively high S to CO's.
The other thing I want to point out, and I guess I will just focus on the lower copy levels, now, these show the BBI panel members. As you can see here, at 300 copies, we were able to detect with both assays diluted 1 to 16, so that is a pretty low titer. In 100 copies, we are able to detect at 1 to 8.
The thing I want to point out here is what the S to CO's look like in the screening assay. These are relatively low. It really represents the low titer of the samples.
Now, in this particular study, the alternative TMA assay performed a little bit better, detecting 2 out of 3 replicates at the 1 to 8 and 1 to 16 dilutions of the 30 copy/ml panel. We have done a number of studies comparing the sensitivity of these two assays, and they really are very similar, so we attribute this to the low number of replicates in the study.
So, in summary, one reactive donation was identified and we were able to depict the 5 blinded controls.
After that initial testing was done, we were sent 13 IgM or IgM/IgG reactive samples that were non-reactive in that initial work that we did. For this testing, we tested 10 replicates using the screening assay, and we identified 2 additional NAT-reactive samples.
The thing I want to point out here then is that the reactivity is pretty sporadic, 7 out of 10 were detected for this sample, and you can see what the S to CO is. Once you get below 30 copies/ml, the S to CO numbers become very variable. You see it ranging from 2.3 to almost 29 here.
Also, a similar case here where, in this case, we detected 2 out of 10, and the S to CO's for the 2 positives were 3.3 and about 24.
To summarize, we detected the 1 sample on initial screening. We also detected 2 in further testing, so there are a total of 3 reactive samples in this collection, we were able to detect all the controls, 3 samples failed to be confirmed with additional NAT testing, and you saw those at the bottom in the previous slide.
The other point to make is that at a 1 to 8 and 1 to 16 dilution, we showed consistent detection at what turns out to be 12.5 copies/ml, and we saw some detection at as low as 2 copies/ml.
Now, I want to turn to my update on the clinical performance of the assay. As you have heard, the live testing under IND for this assay started June 19th, and most sites started testing by July 1st. We are testing at a total of 24 sites.
Most of the testing is being done in pools, about 3 percent of the donations are being tested as individual donations. Actually, this refers to the non-ARC sites in terms of the percent of donations being tested as individual donations.
The initial reactives are confirmed in a number of ways. There is an alternative NAT assay, which is a TaqMan PCR assay that was developed by Chiron. It is being run at the Bayer Reference Lab in Berkeley. There is an IgM test which is being run by Focus.
After the start of clinical testing, we added an additional confirmatory test, which is the alternative TMA assay which I showed in the previous slides.
As of August 26th, we have tested 2.1 million donations, and we have over 400 confirmed West Nile virus reactive donations. That is about 2 per 10,000 donations. Again, "confirmed" means either by alternative NAT reactivity or IgM testing.
The confirmed reactive rates range from zero--and an example of a site like that is in Central California where they tested over 13,000 donations--to a high of about 0.37 percent. I will go into detail about the Bonfils Blood Center in Denver where, for this data, they had tested almost 36,000 donations.
The other thing to point out is that confirmatory results for many initial reactive donations are still pending, so this is really in progress.
Just to introduce the results at Bonfils, I just wanted to show this map that comes from the CDC, and it is as of September 16th, just to point out that there are over 1,200 human cases in Colorado, so I will move on and talk a little bit in detail.
This is a very interesting slide. One thing it points out is that the implementation timing was nearly perfect for the people in Colorado. As can see, when they started at the beginning of July, if you look at the positive pool rate, it was zero percent, and it has peaked towards the beginning and middle of August, and it looks like it is pretty much decreasing and hopefully we are not going to see another peak, but I think they have had frost a lot of places in Colorado.
To summarize the experience at the Bonfils Blood Center, they have tested 35,648 total donations. Now, this data is as of 8-27. That is about 2,200 pools and some individual donations, of course, when the pool can't be completed. There have been 159 initial reactive results, so that is about 0.5 percent of the total donations, and confirmatory testing for 18 of these 159 initial reactives are still pending, but we do have confirmation for 133 donations, and this is by alternative NAT or IgM, and I have to say the majority of them are by alternative NAT.
There are 3 repeat reactive results which we haven't been able to confirm with either one of those tests.
Overall, for this site, the specificity in pools has been 99.8 percent. I have to mention this site does spin the samples with a high RCF. Of 134 index specimens tested for IgM, 16 have been reactive, so that is about 12 percent. I think that might be what we might expect at this point.
I think what we would expect to see as we go on, there are going to be a lot more IgM positives that we are going to have a difficult time confirming. Based on the confirmed results through the end of August, the prevalence at the Bonfils site is about 37 per 10,000 donors.
I want to finish by talking about some of the options for this assay. One obvious one I think based on what we have heard earlier today is to test in smaller pool sizes, and the software is under development for pools of 4 and 8 donations, and that should be available next year for next mosquito season.
We have also demonstrated preliminary feasibility for this assay on our fully automated TIGRIS system, so that is also something that is under investigation.
I have one final slide.
It is important to acknowledge that this project is funded in part with federal funds from the National Heart, Lung, and Blood Institute, so we thank them for their support.
Thank you very much.
DR. NELSON: Thank you.
Questions, comments? Steve.
DR. KLEINMAN: Jeff, do you have estimated numbers for 50 percent and 95 percent detection by your primary assay, copies per ml?
DR. LINNEN: Yes. It is an interesting question because based on the confidence intervals for the analytical sensitivity data, the 95 percent detection level for the assay probably runs 6 copies to maybe 15 copies.
We started to not take that too seriously because we are very interested in how the assay performs with the clinical specimens, and we think in the data where the clinical specimens are able to be quantitated, we can typically detect less than 30 copies/ml consistently.
DR. KLEINMAN: But can people really quantitate down to 30 copies per ml accurately?
DR. LINNEN: We can make an estimation by doing serial dilutions of the samples and come up with a rough number for the copy per ml.
DR. KLEINMAN: So, you can detect reliably meaning approximately 100 percent down to 30 copies per ml?
DR. LINNEN: Right, right. In fact, in our testing in-house, that is the routine way that we QC reagents is to test that low copy, and we typically never miss.
DR. NELSON: Thank you.
Next is Chiron. Is there somebody from Chiron here who is going to make a presentation? No.
Dr. Wayne Hogrefe from Focus Technologies.
DR. HOGREFE: I am going to be very brief as some of this data has already come up, you are just going to see it from a different way from a diagnostic perspective. Quite frankly, the Focus Technology assay has been developed over the last two years as using a diagnostic, so the offshoot was looking in blood supplies quite different for us.
Just to give you an idea of where we are at with this, last year, we based the performance of the assay on about 35,000 samples we ran through our laboratory, so that is how we were able to set our cutoffs and decide where our performance would be.
For diagnostics, obviously, we want to, if we have to err in any direction, it is going to be on the side of specificity. Right now the assay is being used in some selected reference labs in the U.S. This is the kit itself as an entire product is front of another branch of the FDA, hopefully, we will hear from them in the next 10 days.
Public Health has also been supplied with the product. We are also supplying Public Health with antigens since the CDC got out of that business last year, they are not manufacturers.
It is also being used in a couple of studies at Gen-Probe, IND study, you have already heard, several research projects going on donor screening.
This is going to be very brief. I just want to hit on a couple points which really probably are biased more to a diagnostic approach, but hopefully, there will be some tidbits of information here that will be useful looking at blood screening.
This is based on a couple publications, some abstracts that are coming out. A publication came out this month, and some of the studies that will be presented at ASH by Leslie Tobler.
One key finding, as you know, CSF testing in patients is considered diagnostic, and how this would impact blood screening, maybe it won't at all, but I think an important finding that we had, last year we were able to run nearly 1,200 serum and CSF pairs that were submitted from patients.
Eighty-five percent of them were negative for both. We had about 4 percent that were just serum positive, and we ran nearly 11 percent of the samples submitted. Now, these were just blindly submitted to us, were positive in both the CSF and sera.
What is interesting is we didn't see one case where IgM is only detected in the CSF. From a diagnostic perspective, that is a very nice thing to see. Again, from a blood screening perspective, I am not really sure that has much impact, but if you have infectivity or CNS involvement, CSF is what you need.
Just as an aside to that, we just finished this study, and we ran about 75 PCR positives from other viruses. About a third of those were from enteroviruses, and as you know, that is an issue going on right now, do enteroviruses, can you have concomitant or the rates for concomitant enterovirus in West Nile.
Just to illustrate a point, none of these PCR positives to other viruses ever come up with IgM in the CSF to West Nile, so specificity is good.
A little more important to this group, and I think we can follow up from the data that we have been hearing, is IgM persistence, and as you know, John Roehrig and the group in Fort Collins earlier this year, when they went out 500 days on patients from New York, encephalitis patients, half of them were still IgM positive. This is in sera now.
What we just published this month, we have a pretty good patient population to look at. Now, these are not clinically defined, so you have got to assume most of these are going to be West Nile fever that we are seeing.
We were able to follow--and I will show the next slide in just a second--a number of these had paired samples sent at subsequent times, but we reached 50 percent as quick as 60 days of these folks still being IgM positive. The next slide will break this down.
We had 92 samples that were a second sample sent to us within 20 days, and as you would expect, only 1 percent of them became negative. As you go out, at 40 days, it is still less than 15 percent, but when we took samples--now, these numbers are small, I will grant you that, but I think it has been pretty consistent--but 35 percent of the samples at two months have already seroreverted to become IgM negative.
The only samples we had, the farthest one we went out was 111 days, and at this time we are getting very few numbers, but about half are still seropositive, so the blood donors that are being followed now, I think will give a lot of valuable information, not only for the blood screening purposes, but also for diagnostic purposes, because that is the issue we have been running into this year in some of the cases coming back from Ohio and Illinois.
I don't have to go through all the details, this has been brought up several times now, but I am just looking at the Focus component of the study that is being done through the REDS. Again, this is the 1,468 plasmas from last year from highly endemic areas.
Just to give you an idea of the breakdown, 53 percent are from Illinois, the rest from Louisiana, Michigan, Mississippi, Ohio, Texas, Nebraska.
What was found here, out of these 1,468 patients, only 8 in the Focus assay came up as positive for IgM. I want to look at the IgG issue here just real quickly, because there has been some discussion about using index numbers and values to try to follow patients and what does it mean.
Now, our assay was not meant to be quantitative, so these index numbers are just that, they are an indication of reactivity, but they are not meant to be linear by any stretch.
Just look at the IgG's, first, on the next slide because about--well, let's stay here for a second. This has been a pretty consistent number for us in all the studies we have done. We have seen about 1.5 percent seroprevalence for IgG in all the, quote, unquote "normal" populations we have looked at. That has been quite consistent.
Just to give you an idea, this is again on the how do you use index numbers. I think you have to be very careful and not go down that path without thinking about this. Of the 23 G-positives and that were M-negative in this pool, many of them were very weakly reactive, 57 percent, and in our hands, that is very low, weak reactive sample. Only 13 percent had index values of greater than 3.
However, when you look at the opposite, you look at the samples that were both G and M positive, it is just the opposite. The vast majority of them--well, there aren't many numbers--but 4 out of the 7 had very high values.
So, that makes sense. How this is all going to play out in the long run, we really don't know.
If we look at this group here a little closer, I sorted this just by the index number of the IgG. As has been discussed several times, this is just over the cutoff for us. That is as weak an M-positive as you are going to see.
It is hard to find PRNT titers on those samples. We have a fair number, about 10 percent of our samples last year were M-only positives. When they do confirm, the PRNT titers are rarely greater than 1 to 40, so it's in the 1 to 10, the 1 to 40 range.
I highlighted these two samples here in red because the last speaker mentioned those. Those were samples that were looked back. Well, all of the M positives--step back here a second--all these M positives, we went back and looked at individually by PCR or TMA, I should say, the ones in black were negative, the ones in red were positive, but not repeatably.
This was the data you just saw. Two of these samples that were M-positive, 7 out of 10 replicates, and 2 out of 10 replicates came up positive at Gen-Probe. There was 1 TMA-positive out of the entire population, which again you just heard, was both Focus G and M and PRNT negative.
What I was trying to show here, I think there are some points that IgM persistence really needs to be looked at. I think it is going to satisfy a lot of needs especially in the diagnostic side.
The issue of weak positive IgM's, this is a theme you have heard for--I think I am the 9th speaker now--of where does the IgM and the copy number for West Nile, where does that break point for infectivity occur, and hopefully, some of the studies that will come up will show that.
DR. NELSON: Thank you.
Questions or comments? Thanks very much.
Steve Kleinman from AABB.
DR. KLEINMAN: With the airports closed, this is truly a definition of a captive audience, but I know people must be getting hungry. I won't show any data, but I want to briefly tell you about the activities of the AABB in the area of West Nile virus.
Throughout 2003, the AABB has facilitated communication concerning West Nile virus within the blood banking community through its publication of association bulletins to membership.
In a bulletin published in April, that informed membership of actions that would be require of blood collection agencies and hospital transfusion services in anticipation of implementation of West Nile virus minipool NAT testing for donor screening.
Then, there was another bulletin issued in May, which updated membership as to the requirements for donor deferral, quarantine and retrieval of products, and other items that were set forth in the FDA's guidance that was issued in early May.
In addition to communications within the blood banking industry, AABB has also played a major role in coordinating the flow and exchange of information between blood banks and the appropriate Public Health Service agencies.
This has been accomplished through the AABB establishing and chairing an interagency task force that was referred to earlier this morning by Dr. Nakhasi with membership from FDA, CDC, AABB, America's Blood Centers, and American Red Cross.
This task force began having weekly telephone conferences in the spring of 2003 in preparation for the 2003 transmission season. Now, the task force has several purposes for these conferences. One was to ensure that all involved parties were informed of the epidemiology of West Nile virus in 2003, in as timely a fashion as possible.
The second was to monitor the implementation of minipool NAT testing. A third was to review up-to-date performance and yield data from the minipool NAT testing programs, and fourth was to identify important scientific and public policy issues related to preventing West Nile virus this year.
Initially, the task force identified two major issues that needed to be addressed on a national basis, that is, in addition to the implementation of testing.
The first of these issues was the exchange of information between blood banks and state health departments with regard to donors who tested positive by West Nile virus, as well as patients with West Nile illness who may have recently donated blood.
So, a task force subcommittee consisting of people from blood banking and CDC epidemiologists developed suggested criteria for such reporting, the chief feature of which was to preserve the anonymity of the positive blood donor, and we have heard that subsequently, this reporting to state health departments has actually evolved into reporting from state health departments to the ArboNET system.
We developed the document, which was published as an association bulletin by the AABB, and was also distributed by the CDC to all state health departments, and we would like to think that that helped to facilitate some of the early reporting that needed to be worked out between blood agencies and local state health departments.
Secondly, the task force realized that cases of transfusion-transmitted West Nile might still occur in 2003, despite minipool NAT screening. In order to effectively assess such cases, another similarly constituted task force developed suggested recommendations for conducting post-transfusion investigations, and this was published. It is a rather detailed guide in a recent association bulletin.
We now, in fact, see that some of these transfusion cases, suspected transfusion cases are being investigated, and we have heard about one or two today that seem to link transmission to transfusion.
Now, in the last several months as the clinical epidemic has grown, the task force has continually reviewed current data on the yield of testing, and then some of the data, the task force provided data to CDC that was reported in their NMWR that was published on August 15th, the first NMWR on the issue of donor screening.
Also, the task force identified that due to inherent delays in confirming that donors who are reactive on NAT screening are truly infected with West Nile virus, it was actually necessary for each IND holder to establish definitions for West Nile virus presumptive positive donors based on data that were available at the time of NAT screening. Up until then, we were talking about screening test positives and confirmed positives.
We have evolved definitions of presumptive positives that hopefully are being used somewhat consistently in reporting to state health departments.
Finally, based on limited scientific data that we have heard today, that individual donation NAT could identify some viremic donations that were missed by minipool NAT, the task force convened a meeting several weeks ago to discuss this issue in depth.
You have heard a lot of that data that was discussed a few weeks ago here today. After reviewing all available data, it was concluded that one very important issue that needs further investigation was whether units that are individual donation NAT reactive, minipool NAT non-reactive, and IgM antibody positive, the so-called phase IV units that Mike talked about, whether these units could, in fact, transmit West Nile infection to transfusion recipients.
The task force has currently established working groups to explore the design of animal infectivity studies and also to explore collecting data from look-back studies of transfusion recipients to attempt to get a more definitive answer to this question.
I think, in part, an answer to this question may have some importance in deciding West Nile screening policies in 2004.
DR. NELSON: Thanks, Steve.
The final presentation is America's Blood Centers, Lou Katz.
DR. KATZ: I just was going to describe what Mike alluded to, which is the ABC/REDS NHLBI study. America's Blood Centers is the other half of the blood supply, the 70-some independent "kittens," as we call them, other than the Red Cross, so getting ABC together to produce a unified dataset is a little more work perhaps than it needs to be.
This describes what is going on. We have 88 sites. Most of our members are reporting with data on West Nile virus yield submitted every two weeks, analyzed and tabulated by Westat, and provided to the members about a week later. So, by the time we get the reports, the data is about three weeks old.
They are being reported by state of origin to the Public Health authorities, although the demographics that we are collecting will allow us to drill down, as I think Mike said, to the zip code level.
This data, the study set will be complete in the first quarter of 2004 and made available to the centers that participated and analyzed for interesting epidemiologic data.
This is by two-week reporting period, the number of donations screened in the ABC centers that have reported, and the prevalence during each of those intervals through August 31st. As you can see, as of our last dataset, the prevalence was still increasing.
This slide shows you just a line graph of those prevalences.
This is where they are coming from, and this is after the prior 15 presentations, not a surprise, that were clustered in the Rocky Mountain and upper Midwest region. These are raw numbers, so they are not corrected for the denominator number tested, and you have seen this data.
We have identified a substantial number in ABC, 563 that are presumptive positives, as you have heard, these are not all confirmed, that could have transmitted the virus in this approximately two-month interval, and assuming that each donation is transfused into 1.5 recipients as components, we have probably interdicted as many as 800 transmissions as a result of this rapid implementation.
I think this slide is the people who are doing most of the work, and Mike Busch's name should be on there, as well.
DR. NELSON: Thanks.
DR. BUSCH: I just want to correct one point because it is actually quite important. Those were initial reactives. The study is capturing initial reactive rates and then following up with the confirmatory data that hasn't been folded in.
The reason that is important is actually, there was data on Oklahoma there that is important. It implied that Oklahoma had 120 cases. They actually only have, to my knowledge, 12 real ones, and that is the one center in the country that is screening by individual donation testing.
Importantly, these tests, although very specific when they are run through these algorithms of pooled testing to individual, if they are applied directly individually, they have reactive rates in the range of 1 in 700, to 1 in 1,000, and because of the low viremia, one needs to trash those units and defer those donors.
So, OBI has driven--that's why there was a disparity between the numbers that you have seen--and other sites that would screen individually will have a fairly high rate of false positivity, and I think that is important for policy going forward.
I think the approach of targeting ID testing in regions that have high minipool yield makes sense, but if you were to nationalize individual donation West Nile testing, we would be throwing away something in the range of 20,000 units a year and deferring these donors as a result of the false positive rate if you react to initial reactivity on these tests.
DR. NELSON: Is the proportion of false positives greater in individual than in minipool, or are you just saying because of the fact that the numbers are larger--
DR. BUSCH: No, no. Obviously, you see from each of the programs, it kinds of disappears, but they have minipools that are reactive that do not yield reactive individual donations, and that represents false positive minipools.
In order to have a false positive individual donation coming out of minipool testing, you have to have both a false positive minipool and a false positive individual on resolution, so it is the factor of those two rates.
DR. NELSON: Right, it's the duplicate testing essentially.
DR. ALLEN: Mike, can I just follow up on that? Clarify for me if you have got a minipool positive on an initial test, it is quarantined, and then what happens with all of those 16 units that are under quarantine?
DR. BUSCH: Right. Products are not released until there is a negative determination either because the minipool was negative, then, the products in that minipool are all free to go, or if the minipool is reactive, the individual 16 members are all tested individually, and if they are all negative, they are released. If one is positive, it is further worked up, and the other 15 are released.
DR. ALLEN: Thank you.
DR. NELSON: If you have got an individual donation that was positive, and it was repeat testing and it was negative, I guess you would still have to not use that unit, is that right?
DR. BUSCH: With West Nile, an individual reactive donation is trashed or used for research, the donor is followed, and we have seen--as I summarized--we have seen individual reactives that didn't repeat, that proved to be real, the donor seroconverted.
In contrast with HIV/HCV, the algorithm allows if you do individual testing, and there is two discriminatory tests to sort it out, if they are both negative, then, you can repeat the multiplex, and if that is negative, the result is considered negative and the unit is releasable.
So, there is a difference with West Nile because of the dynamics. We are concerned about very low level viremia and are operating on initial reactives as if they are real, whereas, with the other viruses, even if you do ID-NAT with the other viruses, there is an algorithm that allows you to not have to trash a lot of blood.
DR. NELSON: Steve.
DR. KLEINMAN: Just to emphasize something Mike said, I am not sure if it came across. The reason it is different for West Nile with initial reactives is that although most of those are false positive, it is clear that some of those are true positives, so you are still seeing, in the experience of somebody doing individual donation testing, that most of these don't turn out to be real, but since you don't know which ones are real, you have to defer all those donors, follow up, notify them, follow them, et cetera. So, it is not a very specific way to go if you go initially to individual donation testing.
DR. NELSON: The only way you can resolve it is with further samples from that positive donor, right, I got you.
Dr. Nakhasi, did you want to make any summary statements or any pronouncements?
DR. NAKHASI: Yes, thank you, Dr. Nelson.
I think the purpose of this session was really to inform the committee the progress, and basically telling the committee where last year we were and where we are now, and what the issues were and what we encountered during the testing which started July 1st, and therefore, what we are collectively doing with industry in moving the tests forward.
I think Dr. Steven Kleinman really summarized the issue very nicely, which I was going to do, basically highlight a few points, is that even though the testing is in place, even though it is investigational, more than 75 percent are interdicted, infectious cases are interdicted, there are low levels which you heard from the discussion which could be infectious in some cases, one or two cases so far in the human study have been shown they are, but there are others which we do not know.
Therefore, the studies are planned to answer that question, and I think that is where we will go, and having said that, and what the outcome of the studies would be, that will allow FDA to formulate the policy in the future.
So, basically, this was the purpose of this whole informational session and I don't want to take more than that time because we are already running late. We were supposed to be starting this session at 10:30. So, we started at 8 o'clock, and we are at 1 o'clock time here.
So, I think if some questions have to be answered, I will be ready here and the other members will be ready, too.
DR. NELSON: Paul.
DR. SCHMIDT: I would like to make a comment based on what you said before, it is really clinical epidemiology. As some of you may remember, it's on dengue, I got into a discussion here with Dr. Petersen. They had a case of dengue in Puerto Rico, which they claimed was from transplantation, and therefore dengue is transmissible by transfusion.
I challenged that. I was in Puerto Rico at the time. It was all on the basis of was the recipient of this bone marrow in the hospital or at home at the beginning of the incubation period. He was right, they checked the incubation periods, hospital records. This was a transmission of dengue by transplantation.
What I want to comment on here is I am sure this is happening all the time in countries that have a lot of dengue and nobody pays any attention to it. So, we get a virus that moves in and whoops, you know, it is transmitted by transfusion, we didn't know that before.
When the virus is omnipresent, everybody, you know, you just don't pay any attention to it. I think the lesson to be learned, and the big picture, we have got to work on killing the bugs, and not doing tests for them, because there are a lot of them in the waterfall off China, and they are on their way. We will go through this every time.
DR. NAKHASI: I fully concur with you, however, we have to maintain the safety of the blood supply, so we cannot just ignore the fact that there are these things out there.
DR. NELSON: I think the progress and the data that was presented is very impressive, it has been interesting looking at this.
It is now 10 of 1:00. Could we meet at 1:30, break for lunch until 1:30?
DR. SMALLWOOD: I would just like to announce you may want to check with your airline just to be safe, even though I made the announcement before, but we keep hearing different versions, but just check with your airline, but we think for the most part, they are canceled, but to be safe, check. Thank you.
DR. ALLEN: Before we totally break, I would like to just make one very brief comment and ask one question. The comment is I am extremely impressed with how the blood industry has responded to this issue. This is superb, the amount of wonderful research data and practical application data is enormous.
The question is does anybody from any of the blood collection organizations have a quick statement to make in terms of how this has actually been applied. We have heard all of the laboratory testing data.
Obviously, every time you collect additional data on every unit of blood that is going into production, into the blood supply, every time you add a new test, you have got to make changes to the computer software system, you have increased the complexity of the quarantine period and all, have there been problems that have occurred because of the application of this new test.
DR. NAKHASI: Susan asked me to say no.
DR. DAVIS: Except it costs money, and that's an enormous problem in this country.
DR. NELSON: Yes, but it saved probably 1,600 people from getting West Nile virus, so it is probably worth it.
[Whereupon, at 12:54 p.m., the proceedings were recessed, to be resumed at 1:40 p.m.]
DR. SMALLWOOD: We are ready to reconvene at this time.
Our plans for the afternoon are as follows. Dr. Epstein will give a brief summary of the updates. If there is anyone that came prepared to speak in the open public hearing regarding the updates, if you have given us a copy of your presentation, we would be happy to place that on the web site, however, if you would like to make a brief presentation, we will permit that.
After the summary of the updates, we will close out the meeting. The second topic on Supplemental Testing for HCV and HIV will be discussed at another meeting.
DR. NELSON: Dr. Epstein is going to summarize whatever it is, half a dozen updates, and then we have a number of people that wanted to testify, make a statement about the updates.
Summary of Committee Updates
DR. EPSTEIN: Thank you very much, Kenrad.
As Linda said, we will be posting all of the materials on these update items on the FDA web site, and will post any materials provided to us by persons who wish to be heard as part of the open public hearing, whether they speak here today or not.
What I would like to do is just give you the thumbnail sketch of the content that will be posted just so everybody knows what the issues are. You will not hear any detailed presentations. Of course, the Agency will be interested in whatever feedback we get through other venues.
In regard to the PHS Advisory Committee on Blood Safety and Availability, the last meeting addressed primarily the possible use of frozen and liquid blood reserves as a strategy to deal with disasters and terrorism.
I would say that the conclusion in brief was that we really cannot lean on frozen reserves to solve our problems, but that there is a need for more development to figure out how one would utilize and design a system to make sure of liquid reserves, and there is the implicit suggestion that the AABB and Organizational Task Force on Domestic Disasters and Terrorism will continue to take the lead and bring forward an actual proposal or at least concept for where we might go managing reserves.
The Committee also dealt with issues of blood product reimbursement and made recommendations for consideration by CMS.
With regard to the Approval of HIV-1 Group O Sensitive Assays, we are pleased to make known that the FDA has approved two enzyme immunoassays which have sensitivity for HIV-1 Group O.
The Genetic Systems HIV-1/2 plus O combination test is approved as a donor screen, as well as medical diagnostic.
The bioMerieux Vironostika HIV-1 plus O test is approved only as a clinical diagnostic, and not as a blood screen. Fortunately, cases of HIV-1 Group O have remained rare in this country, and nevertheless, we have continued to feel that it is important to have available these more broadly sensitive assays, so we do think that this is useful technology advancement both for the blood screening arena and also for medical diagnostics in general.
With regard to the Revised Guidance on Severe Acute Respiratory Syndrome, or SARS, the FDA posted on its web site two days ago an updated guidance which modifies the previously issued guidance from I think it was April 03, to provide criteria for discontinuing donor questions.
In specific, we are recommending that blood establishments can discontinue travel questions for identified geographic areas 14 days after the Centers for Disease Control lifts the travel alert for that specific geographic area. That information is then, of course, available on the CDC web site, so you would not have to ask about travel or residence in that area that was previously listed.
Ninety days after CDC has lifted all travel alerts for any and all SARS-affected areas, we recommend that blood establishments can discontinue asking questions about history of SARS, suspected SARS, treatment for SARS, or close contact in the previous 14 days with a person with SARS or suspected SARS.
Now, the nature of this guidance then is that it remains in effect and should it ever transpire that SARS re-emerges, that areas are thought to be at risk for community spread, and then they are relisted, the guidance in effect would cause the questions to then resume.
So, these are discontinuation criteria, but blood establishments would need to continue to check the CDC web site. Of course, I believe also that it would be highly newsworthy if and when SARS re-emerges, people will know to go look.
We also will post on the web site the update related to Current Thinking on the Use of Chlorhexidine as an Arm Prep for Donor Phlebotomy. We have already posted a notice indicating that FDA does find acceptable the use of the chlorhexidine arm prep as the primary procedure. This is based on the review of the literature that was done at a previous BPAC meeting where we examined available arm preps.
The lead conclusion from that study was that green soap is inappropriate, that there did not appear to be superiority to the tincture of iodine, isopropyl alcohol procedure compared with povidone iodine procedure, but that additionally, the data equally supported chlorhexidine.
We have been asked whether chlorhexidine could be used as the routine preparation, and what the web notice makes clear is that for products already marketed as surgical scrubs, which contain instructions for use, those specific products are suitable for use for the donor phlebotomy, and that is not all chlorhexidine products, so it is geared to how they are labeled based on their prior validation.
The update on Labeling and Storage of Blood and Blood Components is an alert that the FDA did publish a proposed rule, it published on July 30th. There is an open comment period until October the 28th.
The most important features of that for this group are the fact that it proposes to remove previously existing restrictions that may have made it difficult to implement the ISBT standard bar coding system.
Additionally, the proposed rulemaking contains revised storage times and temperatures for various blood components, and we are aware that there is a lot of interest in commenting on the rulemaking and look forward to receiving the public comments.
Lastly, we had prepared a summary Overview of the Counterterrorism Exercise, so called TOPOFF II, that took place a month or two ago. Those of you who tuned in are aware that it was a simulated outbreak of plague in Chicago and a nearly concurrent radiation exposure incident in Seattle.
What we would like to post is a brief sketch of lessons learned, and I think that the two most important lessons learned were, first of all, the need to be able to link on some kind of geospatial map the locations of blood and other manufacturing facilities that are part of the blood system, you know, where do you get reagents, where are components shipped from, et cetera, where is blood collected, with the ability to map whatever the hazard is, so that we can put 2 and 2 together and help guide us how we might take responsive measures.
Then, in addition, I think it became very clear that one cannot wait for these incidents in order to have policies and plans in hand to deal, not only with emerging infectious agents including bioterrorism agents, but also chemical and radiation hazards.
So, we will increase our effort, which has been ongoing, but we will be stepping up our effort to try to anticipate such possible outbreaks and develop the response policies that would need to be put in place very rapidly at such times.
I am going to close and return it to Dr. Kenrad Nelson regarding the open public hearing.
Open Public Hearing
DR. NELSON: I think I won't read that statement again since I read it already, and so did Dr. Smallwood, but the open public hearing, Kay Gregory from the American Association of Blood Banks wanted to speak on three of the issues that Jay just summarized.
MS. GREGORY: I will be happy to postpone my comments.
DR. NELSON: Okay. The second person, Lou Katz from America's Blood Centers. No? Celso, no? Okay.
Plasma Protein Therapeutics Association. Somebody wanted to make a statement on labeling and storage? No? Okay.
Anybody else want to make a statement?
The statement is that it is raining outside.
I guess that we will postpone the supplemental testing for another meeting, and actually, two of the four people that were going to speak aren't here anyway, and we won't have time really for committee discussion unless we stay here through the night.
I guess if nobody else wants to make any statements, the meeting is closed until the next meeting, which is in December, I guess.
[Whereupon, at 1:54 p.m., the meeting was adjourned.]
- - -