FDA Home Page | CBER A-Z Index | CBER Search | Contact CBER | CBER Home Page

DEPARTMENT OF HEALTH AND HUMAN SERVICES FOOD AND DRUG ADMINISTRATION CENTER FOR BIOLOGICS EVALUATION AND RESEARCH Bethesda Hyatt Regency Hotel Bethesda, Maryland Monday, November 4, 2002 8:00 a.m. PARTICIPANTS SPEAKERS AND MODERATORS: Bernadette L. Alford, Ph.D. Dr. Luiz Barbosa Celso Bianco, M.D. Robin Biswas, M.D. Margo Ann Brinton James Burdick, M.D. Michael P. Busch, M.D., Ph.D. Mary Chamberland, M.D. Dennis Confer, M.D. Andrew Conrad Laurence M. Corash, M.D. Mark A. Damario, M.D. George Dawson Mahmood Farshid, Ph.D. John Finlayson, Ph.D. Richard Freeman, M.D. Glen Freiberg Dr. James Gallarda Cristina Giachetti, Ph.D. Jesse Goodman, M.D. Dr. Ray Goodrich Dr. Albrecht Groener Melissa Greenwald, M.D. Lynell M. Grosso David Harlan, M.D. Liana Harvath, Ph.D. Indira Hewlett, Ph.D. Bill Hobson Cinnia Huang, Ph.D. Louis M. Katz, M.D. Steven Kleinman, M.D. Michael Kanaley Laura D. Kramer Thomas Kreil, Ph.D. Robert S. Lanciotti Jackie Malling Mark Manak, Ph.D. Paul Mied, Ph.D. Martin Mozes, M.D. Martin Munzer Robert A. Myers Hira Nakhasi, Ph.D. Lyle R. Petersen Bruce H. Phelps, Ph.D. Dominique Pifat Maria Rios, Ph.D. Martin Ruta, Ph.D. Laura St. Martin Susan Stramer, Ph.D. Edward Tabor, M.D. Charles Tackney Dr. Steven Wagner Darin Weber Alan Williams, Ph.D. Carolyn Wilson Susan J. Wong, Ph.D. CONTENTS Welcome and Overview of Workshop Goals: Jesse Goodman, CBER Historical Perspective of Panel Screening: John Finlayson, CBER/OBRR I. WNV Biology and Epidemiology: (Chairs: Hira Nakhasi and Jesse Goodman, FDA) West Nile Virus Biology: Margo Brinton, Georgia State University Current Status of the West Nile Virus Disease and Transmission Cases in the U.S. and Risk from Japanese Encephalitis Family of Flaviviruses: Lyle Petersen, CDC Questions II. Methodologies for Detection of WNV and Flexibility for WNV NAT Cross-Reaction with Japanese Encephalitis Family of Flaviviruses: Detection of Human Antibodies to WNV with a Recombinant Antigen Microsphere Immunofluorescence Assay: Susan J. Wong, NY State Department of Health Serological and Molecular Amplification Assays for the Detection of WNV Infection: Robert S. Lanciotti, CDC Utilizing Nucleic Acid Amplification Technologies in Arboviral Testing Programs: Practical Experiences from the Maryland Public Health: Robert A. Myers, Maryland DHMH Laboratories New York State Encephalitis Initiative: Cinnia Huang, New York State Department of Health Detection Issues in Testing Tissues from West Nile Virus: Laura K. Kramer, New York State Department of Health Controls, Panels and a Sensitive and Specific Real-Time RT-PCR Assay for West Nile Virus: Mark Manak, Boston Biomedica, Inc. General Discussion III. Human Cells, Tissues and Cellular and Tissue-Based Product and Organ Donor Transmission Issues: Overview of the Organ Procurement and Transplantation Network: Bill Hobson, HRSA Organ Donor Screening/Donor Acceptance in Extra-Renal Transplantation: Richard B. Freeman, M.D.,Tufts-New England Medical Center Donor Acceptance in Renal Transplantation Viral Disease Transmission in Transplant Recipients: James Burdick, M.D., Johns Hopkins School of Medicine Marrow Donation/Transplantation Issues: Dennis Confer, M.D.,National Marrow Donor Program HCT/P Transmission Issues: Melissa Greenwald, OCGT Discussion IV. Industry Perspectives on the Development of WNV Tests for Donor Screening: Update on National Genetics Institute WNV Screening Assays: Andrew Conrad, National Genetics Institute WNV Transcription Mediated Amplification (TMA) Assay: Cristina Giachetti, GenProbe No Denial About West Nile: Update on Roche's West Nile, Virus Program: Dr. James Gallarda, Roche Real-Time Nucleic Acid Sequence-Based Amplification for West Nile Virus Detection: Dr. Lynell M. Grosso, BioMuriex Target Enrichment Strategy: Proposal for Improvement of Sensitivity and Specificity of RT-PCR-Based WNV Blood Assays: Martin Munzer, Cygene Production of Materials in Support of WNV Assay Development: Bruce Phelps, Ph.D., Chiron Feasibility of an Improved Immunoassay for WNV: Charles Tackney, Ortho Clinical Diagnostics Update on Abbott Laboratories' Strategy for WNV Serological Testing: George Dawson, Ph.D., Abbott General Discussion PROCEEDINGS Welcome and Overview DR. GOODMAN: Why don't we get started? Good morning, everyone. I want to welcome you all to this workshop on West Nile Virus blood and tissue screening pathogen inactivation. I am Jesse Goodman, deputy director of CBER. [Slide] There are quite a number of people I wanted to thank for putting this meeting together. I think I left my list over there, but particularly Hira Nakhasi, from the Division of Emerging and Transfusion Transmitted Diseases, and Joe Wilczek, who put a tremendous amount of effort into this, and then the other people, Maria Rios, Indira Hewlett, Robin Biswas, Mahmood Farshid and Carolyn Wilson. Again, I thank everybody for putting this together in what for the government is a very short time period, probably for anybody for a meeting of this type. I would also like to thank the co-sponsors with FDA, the Centers for Disease Control, NIH, I believe that also NIAID has contributed, the Health Resources and Services Administration, the Office of the Secretary, the Office of Public Health and Science. Everybody has been very supportive of getting folks together quickly to move forward with this problem. Then, I thank all of you for coming. [Slide] Just to give a very brief overview of where we are going here, as everybody knows, West Nile virus transmission by transfusion and transplantation is a reality and it is something that we are all taking quite seriously. I think it is important to state that the degree of risk is still unknown, and you will hear some discussions of risk estimates and certainly an update on the various investigations. In the face of a continuing epidemic in this country the need for donor testing is anticipated. There is a continuing desire to have capability for screening available under IND by next summer. This is both real and this is also a test of our system. It is a test of flexibility and agility for the blood industry, for the diagnostics industry and certainly for the FDA. It is a test and a case study with lessons to be learned. Many people in this room, far more than me, have learned from previous issues of infectious disease testing and HIV and hepatitis, that there are lessons of the past that everyone has learned. I would also like people, as we go through this, to learn lessons from the present. These are relevant to future emerging infectious diseases, of which there shall be more, and also the unfortunate potential for bioterrorism. [Slide] I do really believe that so far the response has been very gratifying. There has been a strong collaborative and rapid flow of information among all the parties. I think that although nobody is ever quite satisfied with it, as many people say to the FDA, if everybody is mad at you you are probably doing the right thing. So, I think there has been an appropriate balance in this communication in the face of a rapidly evolving situation and a knowledge base, and I particularly want to thank our colleagues at the CDC and particularly Lyle Petersen and Mary Chamberland who have worked with us very closely on this. The communication in difficult circumstances has been extraordinary, and also the blood industry has really come through in this respect. All of those same people have come through with what I think are prompt and precautionary actions in a difficult situation, everybody doing their best. There have been, as I said, good inter-agency activities and public-private partnerships. [Slide] So, we need to move forward and we are mostly here for information sharing and discussion. This isn't a decisional meeting or a consensus meeting; it is a meeting to update each other from a variety of perspectives and help each other move forward. As you will see from the agenda, it is quite full of diverse areas but they all fit together to, I think, form what needs to be an ongoing public health response. We are going to review West Nile biology and recent events. We are going to hear about test methodologies. We are going to review issues that are very specific to an area we really haven't considered much before--cells, organs, tissue transplants. And, I would just like to mention that while this is also going, on HRSA regulates the organ transplants and FDA has put forward a plan for a framework to more effectively regulate tissues. The tissue industry is a growing industry meeting a tremendous medical need. We are going to hear diagnostic industry perspectives and results. Many of you here are from the diagnostic industry and we recognize that you bring critical expertise and ability to get things done here. We are going to hear about the related field of pathogen inactivation but specifically oriented towards West Nile virus as this is another strategy. We are going to hear from FDA about potential regulatory issues and pathways to try to make this more streamlined, to the extent possible. And, we are going to hear from the blood industry and others about what kinds of issues are anticipated in implementing screening. [Slide] Some of the kinds of questions that I think are pretty obvious but that we will consider would be what are the most promising tests? Are there special issues for tissues, cells and organs, and I believe there are? How can we move forward with the best kinds of tests? How can we better understand risk? What is the potential role of pathogen inactivation? Are there gaps here? Are there unmet scientific and resource needs that need to be addressed to move this faster or more efficiently or more effectively? I think a bigger, longer-term question is how can we accelerate development of robust platform technologies with the ability to rapidly alter the agents screened? Again, we are not necessarily going into that issue in detail here, but I would say that I and our colleagues at the FDA will really appreciate input and creative thinking about this issue because we sort of don't want to go through the same thing every time there is a new agent. [Slide] Some examples, just a few, of specific questions and issues would be what kind of sensitivity are we looking for? This is a very different situation than with the other infectious disease agents we have screened for. One I would like to raise relatively early in all this is should these tests be West Nile virus specific or cover multiple agents? We have had in the past in this country epidemics--St. Louis encephalitis. I think that Dr. Petersen will mention that. We have a global public health threat from Dengue fever. It may be possible to design strategies and have them in place that would protect from all these agents, and is that something desirable and feasible? Again, can blood screens truly rule out tissue infection? The answer may be no but we may then not have any real other choices. [Slide] There are some policy questions that are a bit broader that, again, we can't finish with today but which I think really need to be on the table. Again, we really want outside input on this. This shouldn't be any kind of unilateral decision-making. What about seasonal and regional application of screening tests? Can that ever be made to work? What is going to be the threshold of risk for testing for this and future agents? This is something that we may wish to bring to our advisory committees and again seek input on. And, how will the blood system deal with the economics of screening for this and future agents? Again getting back to the platform technologies, are there ways to make this more economical? [Slide] Finally, I would just like for people to think about bringing up a couple of sort of longer-term picture issues. As you can tell from how I am pitching this, we really need to be ahead of the curve. I think we acted very quickly here and by past standards maybe we are ahead of the curve, and some people may feel we are ahead of it now but I am not really sure we are, or that we have a paradigm that gets us there. You know, how can we be evaluating and getting tests ready before they are needed so they can be there exactly when they are needed? One very promising example of a technology is microarrays or pathogen chips. We could have a hundred or more pathogens on a chip, potentially have blood amplification in one reaction using random primers, etc. and potentially detect multiple pathogens, some of which we have no concern about at the moment but others which we are all familiar with. So, one idea which I hope isn't the Titanic but I think one needs to think about, for example, would be combining donor screening with surveillance, having a situation, which in a sense we do now. We obtain valuable data about hepatitis and HIV from donor screening. It is a specific well population, but can we obtain information and a "heads up" ahead of time about future threats by such multi-pathogen detection? That is using blood screening as another public health sentinel system for both natural and potentially intentional disease. [Slide] Anyhow, I think that many of you are going to touch on these areas in your presentations. We do want input and we want people to follow-up with FDA. I know the other federal agencies feel this way for their parts in the puzzle. I think we, so far, have done well and we can succeed in this and future efforts to address these infectious threats. Again, I thank you and all the coworkers, and really am eager to hear all your presentations. Again, welcome. DR. NAKHASI: The next speaker is John Finlayson. John will talk about the history of the testing. Historical Perspective of Donor Screening Testing DR. FINLAYSON: Good morning again. This talk should really be given by Jay Epstein, however, he is in Geneva and when we learned that he would be in Geneva it seemed logical that Jesse Goodman could give this talk. But it seems to have fallen to me so you, molecular biologists in the audience, recognize that we are operating here on a two-codon code-- [Laughter] --with a limited wobble hypothesis. That is to say, this talk can be given by anybody whose initials are JE, JF or JG. Presumably the one in the middle has the lowest energy state so I hope I don't fall asleep during the presentation. [Laughter] Maybe I should say I hope you don't fall asleep during the presentation. [Slide] This is supposed to be a historical perspective of donor screening. Quite clearly, as you all realize since you are in this business, in the time that is available to me I can't possibly go through all of the history of how we got to where we are today. However, as we used to say, ontogeny recapitulates phylogeny and, fortunately, if we look at the status quo we can get a pretty good idea of how we got here and the sequence in which we did. [Slide] Let's take a look at the first slide. Here we have the tests that are required or recommended, and you recognize these as being recommended or required for the screening of blood donors; those recommended or required for the screening of plasma, I should say source plasma, donors are a subset of these; and they are arranged in the order of implementation. If we were to have put the dates of implementation on these, we would see that there would be a very large gap between number one and number two and a very substantial gap between number two and number three, and the rest of these would be very closely, or relatively closely spaced together in time. Note that this is the order of implementation, not the order of recommendation. Were we to list these in the order of recommendation we would have to slice out the test for anti-HBC and reinsert it down here. [Slide] Are there any other tests? Yes. Here are a couple of tests whose implementation is voluntary. Now, voluntary means different things to different people, like, "John, you will voluntarily give this talk." [Laughter] What we mean here by voluntary is simply that these are used not at the behest of the FDA. [Slide] Then, as we come into the modern era we have the NAT assays, the nucleic acid technology-based assays. You see they are sort of grouped into three groups here if this is a group of one--I guess you can have a group of one; mathematicians would permit that. These are licensed tests which are essentially universally employed. These tests are selectively employed and you have heard about them a number of times at Blood Products Advisory Committee meetings and you will undoubtedly hear about them again at Blood Products Advisory Committee meetings. We have the hepatitis B virus DNA test which is also essentially universally employed for screening of source plasma donors but which is still considered experimental. [Slide] One of the questions that people have wrestled with a great deal over the years, indeed, over the decades, is that of at what point do we implement a new test? Many people have put a great deal of thought into this. I do not mean by any stretch of the imagination that these are the two papers that stand out head and shoulders above the others, but simply that these are representative of the thought that people have put into this. Clearly, people were thinking about this before 1990 and people were thinking about this long after 1993. [Slide] One can think about all the good things that a test does for one, but we can ask the question are there any downsides to initiating a new test? You perhaps remember this excitement a little over a decade ago. Both time constraints and modesty forbid me to go through all the arguments in here, but suffice it to say that in addition to those arguments, we could say, as far as downsides go, that there can be a negative economic impact. This is, of course, not something that the FDA is supposed to think about but it is one of these elephants that never goes away. One can also say there could be donor loss. Well, clearly if a donor is infectious you want to lose that donor, or at least you want to lose that donor during the time that he or she is infectious but, as we all know, sometimes donors will overreact and self-defer just in anticipation that they might come out reactive in a particular test. Now, are there any quick generalizations that we can make about the time at which one would implement a new test? And, the answer to that question is probably no because every test has its little nuances but, certainly, one can say a few things that would stand scrutiny. First, if you are going to implement a new test you have to have a test to implement. The next profundity will be uttered at exactly 9:14! [Laughter] Second, you have to know something about the test, about its characteristics, about its performance. Then, you should have a feeling of comfort--that perhaps is the best word that I can use--that there is going to be some positive benefit for implementing the test. Again, benefit can be defined in a number of ways but we are talking about benefit from the standpoint of the view of public health. And, one has to be comfortable that the adverse effects, if any, of this test that one is implementing do not outweigh the positive benefits thereof. [Slide] Now, are there some particular difficulties? In other words, which tests have given us the most trouble with the decision to implement or not to implement? Well, those of us who labor in the trenches sometimes feel that the most difficult tests are whatever ones we are working on right now, however, it might be interesting to see what Tom Zuck said in his review. He said the most difficult tests to deal with are those that offer only incremental improvement in the safety of the blood supply. In order to make this not too busy a slide, I truncated his statement but in the interest of telling the whole truth I should go on to say that what he said was those that offer only incremental improvement in the safety of the blood supply and whose increment is so small that it cannot be measured. That is, of course, part and parcel of the sort of thing that we are considering today, what do we know about these tests? As Jesse Goodman said, which are the appropriate tests? And, what are the other considerations that go along with them? I think you will agree that Jesse Goodman could have given that talk and Jay Epstein could have given it if he were here. I venture to say that 40 percent of the members of this audience could have given it. However, this was supposed to be a talk about a historical perspective so in the last two or three minutes let me give you something that within this field I will call, for want of a better term, pure history, if there is such a thing as pure history. Exactly 33 years and three days ago I was attending a meeting of the National Research Council. It was taking place at the National Academy of Sciences building on Constitution Avenue, in downtown Washington, D.C. and the subject was considerations in the implementation of tests for hepatitis. Foremost among these was a test for what in those days we were still calling the Australia antigen. This was the second day of a two-day symposium. The first day had been devoted to scientific presentations. The second day was devoted to panel discussions. The first panel discussion was on the subject of--are you ready for this?--sensitivity and specificity. The second panel discussion, however, was clearly the main event because the subject there was the impact on the blood banking community of the implementation of these tests. The panel members were all distinguished blood bankers. However, even though they were all distinguished blood bankers, some were more distinguished than others and, clearly, the senior statesman among these was Aaron Kelner who was at the time the head of the New York Blood Center. So, when he took the microphone there was an excitement that went through the audience. There was a certain electricity in the air. In other words, it was rather like a Michael Jordan press conference. Everybody is gathered to see what Michael has to say. Of course, it also was very much like a Michael Jordan press conference in that when Michael gets up to talk all the sports writers in the audience already know exactly what he is going to say. They have their stories written and they are just waiting for him to say them so they can file them. See? So, everybody knew that Aaron Kelner was going to say, "this is a very imperfect test." He was going to say that there can be negative units that slip through and they can cause hepatitis B and that is going to cause a great deal of consternation among the public and ourselves. It is going to give a false sense of security that we don't know how to deal with or even to recognize a false-positive. That it is going to cause chaos in the field because we haven't implemented a new test for a long time, and he was certainly right because it has been 30 years since the syphilis test had come to be. And, there was going to be this big economic impact. Everybody knew he was going to say that. So, what did he say? He said in effect that for the last day and a quarter we have been hearing particulars directed to the question of whether or not we should implement this test and I say that is the wrong question. The question is not whether we should implement this test but when, and I say the answer is now. For high historical drama it doesn't get any better than that. [Laughter] Who knows? Maybe we will have our own drama over the next two days. So, let the play begin. DR. NAKHASI: Thank you, John. That was a wonderful historical perspective and you did an excellent job. Even though Jay is not here, he would have said that, and this is an excellent presentation. Now we should start the scientific sessions, and the first scientific session is on West Nile biology and epidemiology. We have two excellent speakers today who know West Nile through and through and have been in the news lately, for the last couple of months. First I would like to invite Dr. Margo Brinton. She is a professor in the Department of Biology, Georgia State University, Atlanta, Georgia. She will talk to us about the West Nile biology, all you need to know. I. WNV Biology and Epidemiology West Nile Biology DR. BRINTON: First of all, I would like to apologize for my voice. I was in Washington about a week and a half ago for an NIH study section and picked up a bug during the time I was here. [Slide] The other thing I would like to do is thank the organizers for inviting me to speak about my favorite virus, West Nile. I started working with West Nile in 1968 as a model flavivirus for studying genetic resistance in mice. I chose West Nile because it grew well in mice and also it was considered a very safe flavivirus to work with. During the years in between then and now I have often been told that the virus I work with is fairly obscure. This is often granting agencies, and if I would work with a virus that was more relevant I would have a better chance of getting funds. [Slide] Obviously, that has all changed now and West Nile is definitely the virus that is going around and, although The New Yorker portrays it like this, it is obviously much different than that and here is an actual picture of the virus. It is a small enveloped virus, spherical, about 15 nanometers in diameter. It contains on the envelope the E protein and the M protein, and then there is a single capsid in multiple copies which makes up the core of the virion. Some very interesting structural studies recently have shown unusual features of this virus. One is that the envelope proteins form dimers, head to tail dimers, and that they sit very flat along the membrane. They don't stick up as spikes as is seen with many other viruses. The other is that the interaction between the membrane proteins is what enforces the icosahedral symmetry on the virion, not the capsid proteins. So, both of those are unusual features in the structure of the virus. [Slide] Most natural infections begin with a peripheral infection, usually by a mosquito bite, and there is initial replication at the site of inoculation. It is thought that that is probably mainly in dendritic cells. The virus then spreads to regional lymph nodes where it continues to replicate. The virus is in the lymph and then spreads from the lymph to the blood. From the blood it spreads to additional peripheral sites such as the spleen, the liver and the lungs and replicates in those organs as well. Then, in some cases, it crosses the blood-brain barrier and then replicates in brain neurons. There are certainly virus and host factors which determine the level of virus replication. [Slide] Looking at a recent analysis of different West Nile strains, first what is obvious is that when West Nile strains are compared on the basis of their sequence homology they can be divided into two groups, called lineage I and lineage II. But within each lineage they vary quite a bit in both their neurovirulence and neuroinvasiveness. So, there are lineage I viruses which are not very neuroinvasive, and in a sense not neuroinvasive in mice at all and still can replicate in the brains of these animals if inoculated directly, but usually it requires a higher titer of virus so that the LD50 is a much higher titer than with some of the more virulent ones. If you compare the ones from the U.S., they certainly fall in the category of the ones that are more virulent and neurovirulent. [Slide] There are host factors as well, and just comparing the viremia titers of different species of birds, you can see that the titers range from very low to very high. The exact host factors that determine the level of virus replication are not well determined and certainly haven't been studied in birds. There have been some studies in mice, which I will refer to briefly later. But there is quite a range and, in general, the lower the titers in the blood, the less chance there is for severe disease to develop. [Slide] Having said that, even with individuals that develop a low level viremia it still requires a functioning immune system in order to survive the infection. The different parts of the immune system are all important in helping the host resist the infection. In the innate response the interferon pathways are quite important. Interferon has been shown to be able to induce a PQR RNase L-dependent pathway and also OAS RNase L-dependent pathway as well. Both of these can certainly inhibit West Nile virus infection. However, interferons induce a large number of proteins and for most of these we don't really know whether they have antiviral function and, if so, what those functions are. So, I think this interaction is far more complicated than we appreciate at this time. It is also being found that there are ways that some of these pathways can be induced even without interferon. For instance, the nucleic capsid protein of some types of flaviviruses can actually induce the expression of OAS genes. It has also been suggested from a number of studies that there may be the existence of interferon-dependent alternative pathways, and these may involve some of those additional proteins that I mentioned. [Slide] The humoral response--it has been shown that antibodies to West Nile are directed mainly to the E protein, to the non-structural protein NS1 and the non-structural protein NS3. However, low levels of antibodies are found to all of the other viral proteins. The cytotoxic T-cell response is directed toward epitopes in a number of proteins, including NS3, E, NS1, NS2A, 4A and 4B. All three of these types of responses are really necessary for recovery, and the higher the viremic titer the more competent these responses have to be for the host to survive. [Slide] The genome of the virus consists of a single-stranded RNA. It is a messenger RNA. It encodes a single open reading frame that produces a single polyprotein. This polyprotein then is processed into the final proteins by the NS3 protein which contains a protease and also by cell proteases in some cases. There is a precursor of the M protein that is in virions inside cells, and then as the virion matures and exits the cell that is cleaved and then the virion maintains the M protein in the membrane and the pre part of it is lost. The proteins in the non-structural region--the NS1 protein is not a part of the replication complex but has been found associated with replicating RNA and there is evidence to suggest that it somehow regulates RNA replication, but exactly how it does that is not known. The NS3 protein, as I mentioned, has protease activity. NS5 and NS3 also have one of the capping enzyme activities. The four proteins listed as 2A, 2B and 4A, 4B are proteins that are thought to function as membrane anchors, holding the replication complex on the endoplasmic reticulum membranes, but these proteins probably also have additional functions that are not well characterized as yet. [Slide] The genome RNA has, in addition to the coding region, non-coding regions at both the 3' and 5' end and these have characteristic conserved RNA structures at each end. The sequences of these structures are not well conserved, but the structures themselves are well conserved among the flaviviruses and it is thought that they play important roles in the regulation of translation and transcription. [Slide] To just briefly go through the replication cycle, the virus attaches at the membrane of a cell. It is likely that it attaches to specific receptors. Those have not yet been characterized but it is thought that they are highly conserved because West Nile can infect so many different species of hosts that it is likely that it is a highly conserved protein. But it could be another type of molecule, other than a protein. Once the attachment occurs, the virus is taken in through an endocytic pathway and then is released. It is uncoded and the capsid is released and then the RNA is released. The first thing that happens is translation so that polyprotein is produced, and then it is processed to the final virion proteins. Some of these proteins are the non-structural proteins involved in RNA replication. So, the incoming genome then becomes a template for RNA transcription, a complementary strand called the minus strand. Then that strand, in turn, becomes the template for new genome RNA which is then packaged together with the structural proteins in the endoplasmic reticulum and the Golgi regions of the cell to produce new virions, which are then transported to the surface of the cell in vesicles and then the virus exists the cell. So, the entire replication cycle takes place in the cytoplasm of the cell. [Slide] During this replication cycle the virus has to interact with many different host proteins. I have sort of divided them into three classes. The first class are those which the virus requires for its replication. So, those would be, for instance, at the beginning of the cycle. If there is a protein that the virus interacts with, then that receptor would be required for the replication of the virus. But there may be additional proteins that are required through the entry process as well. This virus does not shut off host translation and so it is likely that there may be regions of the viral RNA that interact with proteins, and that these enhance the translation of the viral RNA over cell RNA. There are certainly interactions between viral non-structural proteins and the ER membranes and protein introduction to the ER membranes. These are important for assembling the replication complexes on those membranes. There are RNA-protein and protein-protein interactions which are involved in regulating and are possibly also required for the assembly of replication complexes. Those also are occurring in conjunction with the endoplasmic reticulum membranes. Then the various molecules that are made, the RNA and protein molecules of the virus need to be transported to the place where the assembly is occurring, and it is likely that cell proteins are involved in the transport of these molecules and also as chaperons during the assembly process. Finally, during the exit of the virus from the cells it is likely that cell proteins are probably facilitating that as well. [Slide] Because the viral RNA is messenger RNA, it most likely forms a closed loop conformation like cell RNAs have been shown to do. In the case of cell RNAs, proteins bind to the poly-A at the 3' end interact with proteins that bind with the cap at the 5' end. This interaction between the 3' end and 5' end enhances translation and helps to facilitate the recycling of ribosomes on the messenger RNA. For viral RNAs which don't contain a 3' poly-A, and that is true of the West Nile virus RNA, other kinds of protein-protein interactions have been found to occur, and these are just two examples of that. In this case a different cell protein is binding to the 3' end but not to a poly-A. In this case it is a viral protein binding, and in each case these interact with one of the capsid binding proteins. So far with West Nile we haven't found any protein-protein interactions between proteins binding at the 3' end and the proteins binding at the 5' end, but there is good evidence that there are RNA-RNA interactions which could bring the two ends of the RNA together. The RNA of the flavivirus competes for three functions in the cell. As I mentioned, it is a template for translation; it is a template for transcription and replication, and then it is also encapsulated into virions. So, these three functions have to be regulated so that they are occurring in the proper time and at the proper rate. [Slide] This is just a little more realistic picture of what the viral RNAs look like. At the top, here, would be the virion RNA, and then the complementary strand here showing these conserved structures. There are not only secondary structures, but there are tertiary structures here which form what is called pseudo-knots. There are additional ones that have been predicted here, and it is very likely that the interaction of these structures with each other and the interaction of these structures with cell proteins are very important in regulating transcription and translation. Our lab has been involved in detecting these proteins and analyzing them, and recent data suggests that these interactions are definitely critical for the functioning of the virus, but exactly what these proteins are doing during the replication is still not known. The sequences shown here in red are sequences that are involved in that very long 3', 5' interaction. This is over 11 KB. This is a very long RNA, yet, these sequences are holding the 3' and 5' ends together. I mentioned that many of the processes carried out by the virus are occurring in association with the endoplasmic reticulum membranes. This just shows a nuclear region here, and these are cytoplasmic regions. The dark spots are the virus. It is known that translation, polyprotein processing, RNA transcription and also virion assembly are all occurring in association with these membranes. In order to make this environment one in which the virus can carry out all of these processes, the virus signals the cell to proliferate these membranes and also rearranges these membranes. Exactly how the virus does that is not really known. [Slide] It is known that the capsid protein, the NS4 protein and the NS5 protein can go to the nucleus for a virus that is replicating in the cytoplasm. When some virus proteins are found in the nucleus, that suggests that they are there in order to recruit some components from the nucleus or to signal some type of gene regulation in order to change the environment of the cell. So, this is an area which is just beginning to be studied but I think it will be a very interesting area. We don't know the protein partners that these proteins interact with or exactly what the roles are. There have been some studies that indicate that when cells are infected with West Nile the expression of some proteins, for instance ICAM-1 and MHC-1, are up-regulated. There is also some indication that NS1 can function on the surface of the cell as the beginning of a signaling pathway. So, there may be a number of ways that the virus infection signals the cell to make the changes that are needed for the replication of the virus. The other thing that has been observed is that infection can stimulate cell replication, but exactly how this is happening is not known. [Slide] There is a consequence for changing the cell environment, and that is that when the cell notices that the cytoplasmic environment is changing it has various defense mechanisms that come into play. Generally these are called ER. The infection is inducing an ER stress and that can lead to apoptosis pathways as well as to interferon type defense pathways. This is just a pathway here that has been recently published for a related virus, bovine diarrhea virus, and during a pathway like this you have proteins that are made which retard the development of apoptosis, and later on you also have proteins that lead to apoptosis. It has been shown that West Nile can cause apoptosis in cells using a Bax-dependent pathway. Exactly what the components are that lead to that pathway are not known, but it is very likely that when it does happen in infected cells, until 72 hours there are ways that the virus has for inhibiting the development of this pathway at least during the early part of the replication. So, it is thought that the interaction between the virus components and the components of the interferon pathways, and also the interferon-induced pathways and the apoptosis pathways, are probably complex and that there are virus proteins which are helping to first inhibit or delay these responses and then later on the cell may be able to overcome these mechanisms that the virus is using. [Slide] As part of this whole process, there is also the stress granule formation and then, as I mentioned, here is one of the interferon pathways, and a number of these pathways will converge on single molecules. I think probably a lot of the research in the next several years will be on trying to sort out all of these interactions and trying to figure out how these interactions affect virus production, and which ones are the most important in determining how much virus is made by a particular host. If there are changes in any of these interactions, either a change in the virus or a change in the host, then it could result in a different level of virus being produced. So, this may give us a new understanding of exactly how these factors regulate the level of virus in different host individuals within a population, as well as in different species. Just finally to mention that recently in our lab we have identified a mouse gene which is a type of OAS protein and differences in that protein can certainly change the level of virus that is made by a cell. So, a change in a single component can definitely change the level of virus replication that is produced by a cell. As I mentioned at the beginning, lower levels of virus produced by a host usually are the main determinant in whether that host survives, and higher levels of virus are usually not able to be contained by the immune responses of the host and can lead to severe infection and even death. So, I will leave it there and if you have any questions I will be happy to answer them. DR. NAKHASI: The next speaker is Dr. Lyle Petersen. He is the deputy director for science at Division of Vector-Borne Infectious Disease, CDC. He will talk to us about the current status of West Nile virus disease and transmission cases in the U.S. and risk from other Japanese encephalitis family of flaviviruses. Current Status of the WNV Disease and Transmission Cases in the U.S. and Risk from Japanese Encephalitis Family of Flaviviruses DR. PETERSEN: Good morning. [Slide] The topic of my talk is West Nile virus in North America, an update on transfusion transmission. What I will do this morning is first talk a little bit about the clinical epidemiology of the virus, then move on to what has been happening with the spread of the epidemic this year, and then move on to the organ donation and transfusion-related infections. Pictured on this slide is a picture of Culex quinquefasciatus, which is the southern house mosquito which is the principal vector of West Nile virus in the southern United States, and it is shown three times its normal size. [Laughter] [Slide] The incubation period of West Nile virus is not precisely known but generally falls within the range of 2 to 14 days, with most infections probably falling at the shorter end of this range. However, one interesting thing that we found is that in many immunocompromised people who have gotten West Nile infection the incubation period seems to be towards the longer end of this range, which is completely opposite to what one might expect. Advanced age is the primary risk factor for severe neurological disease and death. By far, this is the most prominent risk factor. We also know that immunosuppressive drugs and hematological malignancies are probable risk factors for severe disease and death. We have observed this both in this year's epidemic in conjunction with the transfusion-related cases, but also in experimental infections of humans in the 1950's it also showed that hematological malignancies, particularly lymphomas and leukemias, were very prominent risk factors for death. There is approximately ten percent mortality among persons with meningoencephalitis, and the long-term morbidity among people who do get neurological disease is quite substantial. [Slide] What do we know about how many people are actually getting infected? What we have done so far in the United States is four serosurveys, the most prominent being in 1999 which was centered in the epicenter in Queens during that outbreak, which showed that 2.6 percent of the entire population of the study zone in Queens has gotten infected with the West Nile virus. Three other serosurveys were done in Staten Island. In 2000 there was a small epidemic that occurred and a household-based serosurvey on that island showed that about 0.5 percent of the population had gotten infected, with lesser prevalence in areas which had less meningoencephalitis cases reported. [Slide] From those data, plus serosurvey data from Rumania, we showed this pyramid in which about 80 percent of the people who get infected with West Nile virus have no symptoms at all. About 20 percent of the people do develop what we call West Nile fever, which is a febrile illness of about three to six days duration, and less than one percent or about 1/150 total infections result in meningoencephalitis. This number of 1/150 is a very robust number and we have observed this in a number of serological surveys both here and abroad. [Slide] What has been happening with the epidemic this year? We have a national reporting system called ArboNet which tracks humans, birds, mosquito pools and horses, and these are the data from 1999 through 2002. As you can see, in 1999 there were four states affected, going out to 43 states plus the District of Columbia. What has happened in humans is quite interesting. In 1999 there were 62 human cases with six deaths. In 2000, despite the fact that the virus had actually spread throughout much of New England, there were only 21 cases and two deaths and about half of these occurred on Staten Island. In 2001, despite even larger spread of the virus, there were only 66 humans affected, with nine deaths and so far, as of August 25, this year we have had 3391 humans reported, with 188 deaths. These numbers, here, reflect primarily cases of meningoencephalitis since that is what we focus our surveillance efforts on. This year, of the 3391 persons reported, about 80 percent have meningoencephalitis. If you recall my last slide which showed that about 1/150 persons develop severe neurological disease, this number translates into a number much greater than 300,000 human infections occurring this year alone. [Slide] This map shows the spread of the virus. In 1999, the states affected are shown in red. Actually, this is a misleading slide because the whole state is covered in but the extent of West Nile virus activity was a very small area around New York City, except for one dead bird found in the State of Maryland that year. In 2000 virus had spread throughout the upper eastern seaboard and by 2001 the virus had spread throughout the eastern United States. In yellow is shown the spread as of 2002, and you can see that it has basically gone from coast to coast. California is a little bit misleading since there has only been one human case in California and that was a woman in Los Angeles who, interestingly, worked for an express mail delivery company. We have found no other evidence of West Nile virus activity in California. This woman had not traveled so we are not sure how she got infected. Washington State--there was one dead bird found up there, in that corner of the state, but that really is the western extent of known West Nile virus transmission. So, as you can see, in four short years this virus has gone from here all the way over to here. [Slide] This is an epidemic curve of the epidemic this year. The epidemic--and this is mostly meningoencephalitis cases--started back here in early July, actually late June. There was one case in June which was the first case this year in the southern United States. In grey is the southern U.S.; in green is the northern U.S. Now, the cycle of this epidemic peaking in late summer is due to the fact that this is really a virus of birds, and there is a bird-mosquito cycle that sort of winds up throughout the year and eventually, as the season goes on more and more birds get infected; more and more mosquitoes get infected and after a point in time there are so many mosquitoes infected that humans become at increased risk for getting infected and that accounts for the late summertime peak of human incidents. The other thing you might note is that the peak in the south occurred before the peak in the north, which is accounted for by the earlier emergence of mosquitoes in spring. [Slide] Transfusion transmission of West Nile virus has always been a consideration because there is a transient viremia after infection and, as I showed you earlier, most persons actually are asymptomatic. So, before this year occurred the risk was viewed as small but not zero. The risk was small because there is no chronic carrier state and no cases of West Nile virus transmission, or any other Japanese encephalitis complex of flavivirus had been reported in previous years or in endemic countries. The risk of transfusion-related transmission is related to the prevalence of viremia among donors which is the incidence of infection times the duration of viremia. [Slide] So this raised some concern about is there a cause for concern about blood transfusion? This cartoon says, "it's okay, Gus, you tested negative for the West Nile virus." [Slide] Earlier this year Brad Biggerstaff developed a model to estimate the risk of transfusion-related transmission and I will attempt to go through this model. For more details of the model, it is published in the August issue of Transfusion. The first step of the model is to determine the incidence and temporal distribution of viremia in the population at large. First we start out with the incidence of meningoencephalitis cases reported through ArboNet or our national surveillance system. We have national data by county of all the meningoencephalitis cases that have occurred. If we assume the onset of viremia is one to five days before the time of symptom onset, which is what most of the literature seems to indicate, and the total duration of viremia has a mean of 6.2 days., with a range of 1-11, and this figure is derived from experimental infections of humans done in the 1950's, taking these two assumptions plus this, using a Monte Carlo simulation, we estimated the temporal viremia distribution in the population at large. As I mentioned to you earlier, there are 140 more infections for every case of meningoencephalitis reported, so you assume the incidence of infection in the whole population is about 140 times the incidence of meningoencephalitis. So, all of this gets you the incidence and temporal distribution of viremia in the population at large. [Slide] Step two in developing this model is to determine the incidence and temporal distribution of viremia in the donor population or the people from which blood donors are drawn. We assumed for the model that the incidence in donors was similar to the incidence in the population at large. We know from our serosurveys that actually the incidence of infection in the population is independent of age, which makes sense since everybody can be exposed to mosquitoes. So, this assumption is probably a valid one. Step three is to account for exclusions. We took a conservative approach and we assumed that 32 percent of the viremic donors would be symptomatic and would be deferred. This number comes from population data from the New York City serosurvey done in Queens. Finally, to estimate risk we assumed 100 percent transmission rate from viremic donors to recipients. In the paper in Transfusion the estimated risk was 1.8 to 2.7 per 10,000 donations in Queens during the New York City epidemic in 1999. [Slide] We have extended this model to data from this year, and this is what we got. I would like to emphasize that these data are very preliminary but I think the numbers probably end up in the right ball park. On this slide I indicate some locations. The lower 48 states is indicated below, the counties from which these estimates were made, and these are the average risks estimated per 10,000 donors and the maximum risk during the epidemic per 10,000 donors. For the entire U.S., the average risk per 10,000 donors was about 0.33, with a maximum risk of about 0.96. You can see that in selected cities which were hot spots for the epidemic the risks were much higher. [Slide] This is the temporal distribution of the risk. This is corresponding to August 29, about here, which was the peak of the epidemic. As you can see, for St. Louis the peak was actually earlier than in some of the more northern cities, which corresponds to the fact that the epidemic was earlier in the southern United States than the northern United States. The other point that I would like to bring out of this slide is that the risk of transmission according to this model was highly time limited. From the peak of about August 29 till two weeks later the risk was already about half. [Slide] To date we have 47 possible cases of transfusion-related transmission reported from August 28 to October 26. Of these 47 investigations that we have embarked on this year, 14 were not transfusion related and 33 are still under investigation. [Slide] All of these investigations began with the case of the organ donor transmission. Briefly, what happened with this person, this was a young person who got into an automobile wreck, was hospitalized for two or three days, received blood products from 63 donors, and donated her organs to four recipients. Now, a blood sample obtained at the time of hospital admission, before any blood products, was West Nile virus PCR negative and IgM negative. She got the 63 blood products over the next two days and on the day that her organs were harvested blood samples were PCR positive, culture positive and IgM negative. Her organs went to four recipients, two kidneys, liver and heart. Three of these people developed meningoencephalitis and one developed West Nile fever. [Slide] This incident raised the specter of transfusion-related transmission. To date, out of all these investigations that we are doing we have six that we feel very confident now were related to transfusion. Patient 1 was a postpartum woman who developed West Nile virus meningoencephalitis after receiving blood products from 18 donors following obstetrical complication. She lived in an area where there was active West Nile virus transmission going which raised the question of whether this was transfusion versus mosquito borne. However, a recovered unit of plasma from a PCR positive donor of a unit given to this woman grew out West Nile virus. Patient 2 and 3 I will go into more detail in a minute. Patient 4 was a patient who had a prolonged hospitalization, for more than 60 days, before developing West Nile virus meningoencephalitis; had a donor that was PCR positive which was subsequently shown to seroconvert. So, this person really had no possibility of mosquito-borne transmission. Patient 5 and 6 I will detail in a second. [Slide] This slide shows patients 2 and 3. Patient 2 was a recipient of a liver transplant. Patient 3 was a patient who received blood transfusions postpartum. These two people had a common donor. The donor donated around August 15 and two days later developed fever, weakness and rash. So, the donor actually became sick after the time of donation. Patient 3 received two units of red cells from this donor from September 2 and 3, and about ten days later developed onset of symptoms. Blood that was available five days later--her CSF was IgM positive in a sample taken five days later. The donor was subsequently shown to seroconvert. [Slide] These are two other patients, both of whom have cancer. Both of these developed West Nile virus meningoencephalitis. The donor for these two persons had fever, headache and eye pain developing about five days before the donation and developed a rash two days after the donation. Patient 5 received red cells. Patient 6 received FFP from this donor. Patient 6 was interesting because serial samples were available from this patient. Before the FFP a serum sample was IgM negative and PCR negative. The FFP was transfused on October 6. Two days later the patient developed fever, and a sample taken seven days later was IgM negative and PCR positive. Subsequently, seven days after that, the serum sample was also shown to be IgM positive and PCR positive. [Slide] In summary, out of these six cases, the illness onsets were from August 1 to October 9. The days from transfusion to illness onset ranged from 3-13, with a median of 7.5. Red cells were implicated in three of these; platelets in two of these; and FFP in one of these. [Slide] For the four implicated donors of these six persons, the dates of donation ranged from July 22 to September 6. The symptoms of these people--two donors had symptoms two days post donation. One donor had symptoms five days pre to four days post donation. One donor had symptoms three to ten days pre donation. The maximum interval from donation to transfusion for platelets was five days, which corresponds to the shelf-life of those platelets. For red cells it was 26 days, and for FFP 44 days. [Slide] From these four donors, which all were shown to seroconvert which were shown to be TaqMan positive at the time of donation, as I mentioned, all four of these people had illnesses shortly before or shortly after the time of donation. They all had low levels of viremia, less than 20 PFU/ml. For other investigations we have TaqMan positive donors identified. However, on follow-up they were IgM negative. We are not quite sure what this represents. [Slide] In summary, the six cases provide evidence for West Nile virus transmission via blood transfusion. The donations all tested positive by PCR. West Nile virus was isolated from a unit of FFP. The donors had seroconverted after donation, and the donors had West Nile virus compatible illness before and after donation. So, I think all of these facts provide pretty strong evidence that West Nile virus is, in fact, transmitted via blood transfusion. [Slide] There are a number of other unanswered questions. One is a better definition of the clinical course of West Nile virus infection and viremia. Second is to define the scope and magnitude of transfusion transmission; to determine the prevalence of viremia in donors; the rate of transmission from viremic donors and associated risk factors; look at the seroprevalence in frequently transfused persons; and also to look at transmission via other flaviviruses such as St. Louis encephalitis and Dengue. [Slide] So, the question is really where are we at right now in the spectrum of West Nile virus illness in the country. One thing that we haven't discussed very much is the relationship of St. Louis encephalitis and West Nile virus infection but these two viruses appear to have a very similar ecology and probable behavior. So, if we want to predict the future maybe we should look at the past, and one way to do this is to look at what has happened with St. Louis encephalitis in the last 34 years. As you can see, this chart shows the incidence of meningoencephalitis due to St. Louis encephalitis virus by year, and you can see that St. Louis encephalitis causes periodic outbreaks which are very difficult to predict and varying in location. In 1975 there was a very large outbreak of St. Louis encephalitis in the Midwest which caused about 2000 cases. [Slide] If you start comparing St. Louis encephalitis in 1975 with West Nile virus in 2002--these are maps showing the incidence of meningoencephalitis per million population and you can see that both of these epidemics centered in the Midwest. The other thing I would like you to take note of is this map which shows the incidence of meningoencephalitis for West Nile virus. What we would estimate is that the incidence would directly correspond to the risk of viremia among the donors taken from those areas. You can see that the high risk donors would be those predicted to be in the Midwest this last year. [Slide] I would like to end on this slide. This is the carrier of the West Nile virus and this is the carrier of West Nile virus hysteria! [Laughter] DR. GOODMAN: First of all, thanks for those great talks. We have a crammed schedule here and we are way behind already but I think we could take one or two questions if they are really good ones. If they are not, I am going to laugh at you! So, please, come up with one or two good questions, or were those talks so clear that there are no questions? Yes? Please, when you get up identify yourself so if I want to take retribution I can do it! Questions DR. KLEINMAN: Steve Kleinman, from UBC. I have a question for Lyle. The serosurvey data that you showed from the 1999 and 2000 series seemed to be very sparse actually. I mean you have estimates of 2.6 percent but only 800 people were tested. So, they must have wide confidence intervals. So, my question is have there been similar serosurveys done this year? If so, do we have any meaningful information and do we have large enough sample sizes to really pin this thing down? DR. PETERSEN: The answer is household-based serosurveys are extremely difficult to do, as you might imagine, particularly in New York City. So, it is a monumental task and one of the limitations is that you just can't get huge sample sizes out of a household-based serosurvey. So, the confidence intervals are wide, as you pointed out. A serosurvey was done last week in Louisiana, in an area just north of Lake Pontchatrain, where we estimate that the seroprevalence in that population should be about 20 percent. So, I think out of those data this year we will get much more precise estimates both of the asymptomatic to symptomatic ratio but also the prevalence at least in that population. DR. BUSCH: Mike Busch, both for Lyle and Dr. Brinton. What do we know about the distribution of virus in blood, especially is the virus strictly in plasma or might there be a cell-associated component? Secondly, the infectious titer of virus, do we have any understanding as to the concentration or dose of virus required to transmit, and whether that dose is reflected in probability of disease? DR. BRINTON: It has been estimated that you need 105 PFU/ml in the blood in order to have transmission. Obviously, statistically you could have less than that and have the mosquito just by chance pick up the virus. DR. BUSCH: Is that dependent on the inoculum size? DR. BRINTON: Yes, it is a very small drop of blood. DR. BUSCH: So, if you put a blood unit in a much smaller concentration it would likely transmit. DR. BRINTON: Obviously, the larger the volume, the more chance of transmission. As far as tissue associated, certainly the virus in the periphery is replicating in tissues. Then, when it gets to the brain it is in neurons so there would be virus in tissues as well as in the blood. DR. BUSCH: How about leukocytes? DR. BRINTON: There is certainly evidence the virus can replicate in macrophages. DR. PETERSEN: There is one piece of data from these experimental infections of cancer patients in the 1950's which showed that inoculum size actually had no relationship at all to the clinical symptomatology. DR. NAKHASI: Lyle, this question is for you. We know from the last epidemics in West Nile virus throughout the world in 1974 in South Africa and in 1976 in Rumania, what would you predict will happen in the future? I am asking you to look at the crystal ball. Are we over the epidemic now, or are we going to have future epidemics? Also, because this virus, as many of you have told us, is changing, you know, the virulence has changed from what we have seen in the past versus now, what are we predicting? DR. PETERSEN: Well, one thing that Rob Lanciotti, who is here in the audience, has done is look at the genetic variability of the virus here, in the United States, and has shown that over the past years the virus has remained very genetically stable. So, it is unclear whether the virus genetics are going to change anytime soon that would cause a change in virulence. About what we might predict in the future, I think the best model is St. Louis encephalitis virus in which we would predict that the incidence will remain fairly low in most years, as it has in previous years here in the U.S., with periodic outbreaks of varying sizes in various parts of the country, and these outbreaks will be very difficult to predict. The one caveat is that if you look at St. Louis encephalitis in birds versus West Nile virus in birds, West Nile virus produces a much higher viremia, several orders of magnitude higher in most species of birds than St. Louis encephalitis, indicating that the potential for more outbreaks or higher incidence throughout the country is probably there. So, if I had a crystal ball I would say the incidence of West Nile virus is going to be higher on average than St. Louis encephalitis and there will be periodic outbreaks, sometimes very large as occurred this year. DR. GOODMAN: One last question. DR. TABOR: Ed Tabor, from FDA. This is for Dr. Petersen. Of the six cases that are confirmed transfusion-transmitted West Nile virus infections, how many of the donors had symptoms at the time of donation? In the morbidity and mortality weekly report it looked like at least two of the donors had symptoms before and after donation. But is it possible that some of the donors had symptoms at the time of donation and that they slipped through the currently implemented screening process? DR. PETERSEN: It is possible that they slipped through the donor screening. Now, as you mentioned, a couple of these people clearly did not develop symptoms until after the time of donation. A couple of people did have symptoms before the time of donation, and whether or not they felt okay at the exact moment they donated I don't know. It is difficult to tell from the donor histories. We have gone back and talked to these people again and again and tried to elicit good histories out of these people, but the symptoms that many of these people had were rather minor and it is difficult to really determine how symptomatic, if at all, they were at the time of donation, at least on the day of donation. DR. GOODMAN: Well, we will have more opportunity for discussion. Since Ed brought that up and Lyle commented, I think it is very interesting that it is a very small N, four. Of four donors in these proven cases, none of them were the majority of individuals who we believe get West Nile and were totally asymptomatic. That is very interesting because we were quite skeptical certainly of how efficacious either pre or post donation illness screening would be, and at least in these four cases it could potentially have been efficacious. So, whether there is something different about symptomatic donors that makes them more likely to transmit disease such as level of viremia, although the preliminary data doesn't indicate much about that--so, I think it is something to keep in mind as the investigations of the other cases go on and we learn more. We can move on with the next session which Maria Rios and Bob Lanciotti, from CDC, will oversee. We will just start right up with Susan Wong, from New York State, who will be talking about some of the antibody assays that they have developed. Susan? Thanks very much, Susan. II. Methodologies for Detection of WNV and Flexibility for WNV NAT Cross-Reaction with Japanese Encephalitis Family of Flaviviruses Detection of Human Antibodies to WNV with a Recombinant Antigen Microsphere Immunofluorescence Assay DR. WONG: I would like to thank the FDA for this morning for the opportunity to be able to present to you a new immunoassay which detects antibodies to West Nile virus that was developed over the last several months at the Wadsworth Center. [Slide] I plan to tell you first what properties I consider important for a diagnostic serology assay for West Nile virus, and then what other properties might be desirable in an assay that is used to test for antibodies to West Nile in the blood supply. [Slide] Certainly, the quality of the antigen is very important for any sensitive and specific immunoassay. With respect to testing for antibodies to flavivirus, it is very important to have a highly pure antigen in a native confirmation and, hopefully, with the glycosylation as would occur in vivo. Many of the antigens which are currently being used have been polyethylene glycol precipitated and acetone extracted or lyophilized. Such procedures have the potential to partially disrupt this fairly complex tertiary structure of a homodimer of the West Nile envelope protein on the surface of a cell. Moreover, many of the antigens currently being used have some contamination from other cell proteins or from growth medium, leading to a phenomenon known as non-specific binding. It is also important to have a high affinity detector antibody, and also to have a signal amplification mechanism so that you can have a sensitive assay. Certainly, West Nile virus is not the only flavivirus which may be encountered in donors to the blood supply in North America. Before West Nile virus had arrived in North America we had encountered, as you have heard, previous outbreaks of St. Louis encephalitis virus. It is also possible that individuals who have been on their winter holidays in the Caribbean returned to the United States while they were incubating Dengue virus. Certainly, an antigen that is predictably cross-reactive to these other flaviviruses could actually be advantageous for screening the blood supply. It is also important that we have an accurate test and that we have a long shelf-life of the antigen reagent under normal storage conditions to decrease the problems with maintaining quality control in an assay. Fewer technical steps and a small number of reagents also leads to a better assay. We want to have low variation and a defined specificity of our assay in other infections and diagnostic conditions. [Slide] This slide shows an SDS polyacrylamide gel of four different product lots of recombinant West Nile envelope protein. This recombinant West Nile envelope protein is a proprietary product and is expressed in a eukaryotic cell line, purified by column chromatography. [Slide] Our antigen was covalently linked to the surface of fluorescent polystyrene microspheres by a two-step carbodiimide process. The assay protocol included a 30-minute incubation with a serum dilution, then filtration, washing and a microfilter plate, followed by a second 30-minute incubation for the fluorescent anti-human antibody. After more filter washes, the bound signal was interrogated by a laser on a Luminex bench-top flow cytometer, and we obtained a quantitative result. Fifty micrograms of pure antigen, when coupled to six million beads, which is our normal lot for making antigen beads, provides enough reagent for 2500 patient test results. For comparison, the standard CDC MAC ELISA and IgG ELISA currently being used in most public health labs take 24-48 hours to obtain the results and, moreover, the laboratories have to frequently make coated plates every single week, which leads to increased problems for us with performing the assays. [Slide] The test protocol that we developed initially was to test for total antibodies with a polyvalent detector antibody. Our strategy to detect IgM was actually to deplete IgG with a goat anti-human IgG and retest the sample. One value here that you haven't heard yet this morning is the P to N ratio. The P to N ratio is the patient's optical density against West Nile antigen, or the patient's fluorescence against West Nile antigen measured over the negative control serum being tested against West Nile antigen. In our assay, a P to N ratio of greater than 4 indicates evidence of flavivirus infection at an undetermined time, and P to N of less than 4 is no evidence of past occurrence of infection. So, our current test protocol is to test for polyvalent total antibodies. Then, if greater than 4 to test with an IgG depletion step. If the P/N stays greater than 4 we say current or recent infection, that is, IgM has been detected. If the P/N has now decreased to less than 4, it means that the sample was mostly IgG and that is an indicator of past infection. However, for confirmatory testing samples are referred for cross-neutralization plaque reduction titers. [Slide] Our assay has shown very good linearity over two logarithms in dilution series, for example shown here, on four positive samples. This indicates that our assay has a relatively broad dynamic range. Our chosen starting dilution was 1:100 because we had shown in other studies that frequently if you started at 1:25 the other serum proteins were inhibitory of the sensitivity of the assay and, moreover, 1:100 dilution gives near optical median fluorescence intensity, which is the fluorescence measurement on the surface of each bead when interrogated by the laser and 100 beads are counted. [Slide] Certainly, when we made a product lot of antigen and we used our positive control serum which gave a median fluorescence intensity of 7000, when we tested every day for approximately four months the value was still 7000. Our antigen on the surface of the beads is stable for approximately four months. However, we did some thermal denaturation studies to see how rapidly it would degrade at 25 degrees centigrade when the antigen was left on the bench and when heated to 37 and 50 degrees centigrade. You will notice that the antigen-coated beads were actually fairly stable when left on the bench at 25 degrees centigrade--but certainly by having an antigen with a long self-life that decreases lot-to-lot variation and your ability to maintain good quality control from assay to assay. [Slide] We were challenged by a blinded serum panel of 19 human sera from CDC. After we had performed our test and sent the results back to my colleagues in Fort Collins, it became apparent that our polyvalent assay to total antibodies with microsphere immunoassay, which is the black solid bar, had correctly identified all seven negative sera. This grey bar is actually the West Nile MAC ELISA, IgM capture ELISA as performed at New York State and at CDC, which is the striped bar. You will notice that one of those seven negative sera for West Nile actually was falsely positive in both New York and in CDC with the IgM capture antibody. You will also notice, moreover, that the total antibody microsphere immunoassay was able to detect as positive above our cut-off of P/N of 4 10/12 sera from patients with confirmed flavivirus infection by plaque reduction neutralization. Moreover, you will notice that we have in seven of these 12 sera a considerably higher signal to background level with the polyvalent microsphere immunoassay compared to the IgM capture ELISA and the IgG ELISA. The two sera that we missed I was told were collected on day zero onset of infection and day one after onset of infection. I am convinced that if we had received another sample within a few days we would have, indeed, identified these patients as well. I think it is important also to point out the ability of this microsphere immunoassay to pick up antibodies to Dengue. We had a very, very significantly strong signal in the sera from patients who had Dengue. Perhaps this is associated with the high viremia that can occur in Dengue patients. [Slide] We also challenged this microsphere immunoassay with a panel of sera from employees of the state health department who had received the Japanese encephalitis vaccine. These samples were received in our laboratory blinded, and after unblinding it became apparent that we had correctly identified the four new employees, who had not received the vaccine, as negative, and in the employees where we had a sample prior to vaccination to compare it to, those employees who had a neutralizing titer that developed as determined by plaque reduction neutralization testing, there was a significant boost in the antibodies to Japanese encephalitis virus as detected with this microsphere immunoassay. So, not only does the microsphere immunoassay detect antibodies to West Nile, St. Louis, Dengue but also to Japanese encephalitis virus. We also did an IgM depletion step to repeat this assay as an IgG version of the assay. The data is not shown here. [Slide] Certainly, one of the things that is very important is to compare a new assay to what is currently available. This was a very early experiment, about two weeks after we set up this assay. It shows a comparison of about 101 samples of our polyvalent assay or total antibodies detected compared to the West Nile IgG ELISA using a CDC reagent and format. In this instance our correlation gave a value of 0.92 and a slope of 2.09 [Slide] It was very important, however, for us to be able to develop an antibody test that could pick up IgM because IgM is a measure of infection. The top curves are the ELISA with an IgM antibody capture ELISA, the CDC reagents as performed at New York State. The bottom curve is a series of the same five samples of a West Nile encephalitis patient, going out to 260 days, and that is the IgG ELISA curve. Notice that the IgG ELISA curve never got higher than about 7 or 8 P/N. The bottom curves are from the microsphere immunoassay. The top curve is the total antibodies picked up by the polyvalent detector antibody. You will notice that the peak of antibodies detected in this assay was at about 58 days after onset, which correlates with the peak of the IgG reactivity in the IgG ELISA. You will also notice that we did an IgM depletion. The curve parallels the total antibody curve, again with the peak of IgG in the 58-day sample of the samples we tested. However, when we did IgG depletion, you will notice that the peak of IgM reactivity as detected with a polyvalent conjugate and with an IgM specific conjugate gave almost identical curves and the peak of reactivity was on day 17 after onset, as shown above in the samples tested with the IgM antibody capture ELISA. So, this was one of the examples of the protocol that we can actually get sensitive detection of IgM with both the polyvalent antibody conjugate and the IgM conjugate. [Slide] One of the things that is very important when one is developing a new assay is to make sure that the people performing the assay can replicate the work of one of their other laboratory technologists performing the test. This shows the cumulative inter-operator specificity study with our West Nile microsphere assay performed by two different technologists in my lab independently on the same day. You will notice that we had correlation on the results of 0.995 when 91 samples were tested on a plate, with a slope of 1.125. So, this is very good inter-operator performance for lab techs working at the bench. [Slide] Of course, one cannot develop a diagnostic assay, nor a test for the blood supply, without understanding the specificity of performance of your assay. We challenged our assay with sera from a variety of infectious disease conditions, such as Lime disease, granulocytic erlichiosis, syphilis, HIV, herpes simplex virus, EB virus, CMV infection, and also from patients with indicators of autoimmunity, such as patients with high antinuclear antibody titers and rheumatoid factor. Unfortunately, we did find one really surprising outlier. Greater than 50 percent of the sera from confirmed cases of syphilis actually related to our West Nile envelope protein that is very pure recombinant. This was a big surprise and I am certain that none of us want to put on our questionnaires to encephalitis patients have you ever had or do you currently have syphilis. However, it is important to know that we reflect these positive sera to the CDC IgM antibody capture ELISA and, indeed, found out that the same thing happened. This phenomenon was primarily a phenomenon of IgG reactivity, not IgM, however it was very reproducible and we don't know the reason for it. My colleague, Dr. Ray Koski searched the T. pallidum genome compared to the envelope protein gene sequence and we could find no significant areas of similarity. So, I did, indeed, notify my colleague at Fort Collins that IgG to West Nile envelope protein does not necessarily mean past flavivirus infection; it could mean past syphilis or current syphilis. [Slide] One of the things that I did was try to develop a cost estimate of what it costs us to perform the new microsphere immunoassay. Every day we have to spend about $6.74 to do the daily QC. About once every four months it costs us about $57.37, excluding staff costs, just to covalently link the antigen to the microspheres. However, to actually perform the test where we can put 94 samples on a plate, we can do a patient polyvalent result at a cost of 24 cents and a patient IgM result at a cost of 24 cents. So, we can get patient total antibodies and IgM at a cost of about 50 cents. This is excluding, of course, the cost of antigen. However, it costs us about $4.84 to get an IgM result with the CDC assays and about $5.25 to get an IgG result. So, we can decrease by 95 percent the cost to New York Health Department when we implement this for patient testing. Our time commitment in terms of technologist time is about four hours, start to finish, with 94 patient specimens on a plate. That means that we can do polyvalent screening in the morning and turn around in the afternoon actually to tell which patients have IgM antibodies. [Slide] Certainly, we have done parallel testing with the sera that came into the Wadsworth Center in year 2002. To date, 669 sera have been tested by the polyvalent microsphere assay and where they were positive, and we tested them by the IgM version assay and we compared these data to the results from the IgM antibody capture ELISA and the IgG ELISA. As anybody doing West Nile testing this year knows, data analysis is incomplete because convalescent sera and plaque reduction neutralization test results are pending on many of the patients who actually were positive in August, September and October in New York State. Moreover, my institutional review board approval is for blinded testing. So, whenever I get a result that is positive in my assay and which is negative in the CDC assay I am not allowed to request the follow-up serum. [Slide] However, some preliminary results from our parallel test are that on 669 sera there is approximately 94 or 95 percent concordance between our new, more rapid assay and the assays currently being used. There are approximately 43 sera where we are hoping that we do get follow-up sera, perhaps clinical information, neutralization testing results. Of these 43, 29 are sera that are positive in the microsphere immunoassays while negative by the CDC assays. One of the larger areas of concern is that with the IgM antibody capture ELISA approximately 1.1 percent of these samples have to be reported out as non-specific or "I don't know" because of high background binding to what is called the control antigen, which is a tissue culture supernatant. In the IgG ELISA 16.7 percent of the sera, or 112 out of 669, are reported as non-specific. Certainly, in New York State I recognize that if you ever send out to a clinician a result of a non-specific laboratory test he is very unlikely to give you the follow-up sample that, for example, would have been useful for plaque reduction neutralization testing. [Slide] So, I think we have some advantages with the relatively new assay we have developed at the Wadsworth Center. Our pure antigen on beads is stable for months. The microsphere immunoassay has a higher throughput than the ELISAs, with 94 patients per plate. We have a low specimen volume requirement. We use 10 mcl of serum to do our initial dilution. And we have a spinal fluid version of the test, that I can't talk about today, which uses 30 mcl of spinal fluid. We have relatively good precision and accuracy, sensitivity and quality control. I forgot to mention that in our covariate studies for intra-assay variation we have CVs of approximately seven percent. Our inter-assay CVs are approximately 15 percent on positive samples. We have a low reagent supply cost compared to enzyme immunoassays. As I mentioned previously, a little antigen that is pure goes a long way in terms of producing patient results. [Slide] Some further advantages that should be considered are the fact that our assay seems to be predictably quite cross-reactive to the other flaviviruses that would be important for monitoring the blood supply. However, if we consider the other flaviviruses which may be encountered in North America, then perhaps by putting NS1 or pre-M or M or some other flavivirus protein on a different bead in this Luminex technology we could multiplex and actually get a more specific result for an individual flavivirus without having reflex to neutralization testing. Moreover, the assay could be multiplexed with antigens for other arbovirus infections. The turn-around time of our assay certainly would be good for organ donor investigations in a cadaveric donor setting. Certainly, the cold ischemic time for solid organs prior to transplantation, I believe, is somewhere between 24 to a little bit over 24 hours and we can turn around the result in three hours. Certainly, many transplant labs have on board a flow cytometer and often it is used for T and B cell cross matches. So, the tech could actually do these assays while they are doing the T and B cell cross matches in the lab. Also, the technology with antigen on beads can be adapted to other platforms by changing the bead or the detector. [Slide] The nucleic acid tests and the serologic tests are complementary. Serology, when it is sensitive, can approach the viremic period decreasing the window between when viremia is present and when IgM or IgG are detected. Sensitive serologic tests would give confidence to findings of nucleic acid testing, especially if pools are tested in nucleic acid testing. [Slide] My future plans are to, hopefully, characterize ranges for an indeterminate zone for polyvalent and IgM microsphere assay; set up ranges for our spinal fluid assay for which over 600 samples have been tested to date; and go live with applying with our new testing protocol for residents in New York State, hopefully, in the near future. I think it would be helpful if we could work together with people in the audience to make the technology available for screening, for diagnostic testing and also for surveillance applications. [Slide] This assay could not have been developed without a fairly unique and very productive collaboration between the biotech industry, academia and public health at the state and at the federal levels. My colleague, Dr. Ray Koski is in the audience today. He and Michelle and Kali actually produced the antigen used in my assay. The plasmid and the construct were designed by Dr. Fitrig and Tina Wong at Yale University. Valerie and Rebecca did all the work. Dr. Laura Kramer's lab did the neutralization testing, and Karen kept all of the rest of the diagnostic work ongoing while we developed this assay. Kathy Kellar, at the CDC in Atlanta, advised us on how to use the Luminex effectively. Dr. John Roehrig, Bob Lanciotti and Jane Johnson, I thank them for very kindly providing the challenge panel to me, and Dr. Tony Marfin, at Fort Collins, for saying, "Susan, don't give up" during minutes when it was frustrating to work with the technology. So, thank you very much. DR. RIOS: Thank you, Dr. Wong, for a beautiful presentation. I would like to call Robert Lanciotti, from CDC. He is at the Division of Vector-Borne Infectious Diseases and he is at Fort Collins. He will be talking to us about serological and molecular applications for detection of West Nile infection. I would just like to make an announcement. When you hear the beep, you have five minutes to finish up your presentation. We have a very tight schedule and we would like to keep up with all the presenters. Thank you for your consideration and cooperation. Serological and Molecular Amplification Assays for the Detection of WNV Infection DR. LANCIOTTI: You could also just tell me to shut up and sit down, and I will be happy to do that. [Laughter] [Slide] What I would like to do this morning is talk to you about the kinds of tests that we use at the CDC in Fort Collins. As you can see, I am with the Arbovirus Diseases Branch in the diagnostic and reference laboratory in Fort Collins. First of all, I would like to mention that these are the recommended tests or the tests that we use at CDC. As you can see, basically we divide serology tests for human serum or CSF. [Slide] I am going to talk about the ELISA testing that we do, and Dr. Wong has already mentioned the IgM and IgG ELISAs that we currently use. But I would just like to point out that our nucleic acid-based tests are not really the tests of choice for human serum and/or CSF. We have found over the past three years that a positive result by PCR in CSF or serum is quite meaningful, however, a negative really doesn't mean anything, and this gets into the whole discussion of the duration of viremia and the titer of virus in serum and in CSF and there are many times when there are true West Nile cases and we are not able to detect West Nile virus by PCR. So, we don't really rely on the nucleic acid tests when we are dealing with acute human serum and CSF. So, for serum and CSF we rely primarily on the serologically-based tests, the ELISA and the plaque reduction neutralization. When we get into discussion of tissue and fatal cases, tissues, whether they are from humans or from the dead birds in our surveillance or mosquito pools, in all of these environmental samples we find that the nucleic acid-based tests are, in fact, the most sensitive. So, again, for the environment testing of mosquito pools and dead birds, and so forth, the real-time PCR tests are our tests of choice. We can process many samples that way. We also do virus isolation on all these samples, and we found over the past three years, looking at thousands of samples, that the TaqMan test or the NASBA test is, in fact, more sensitive than virus isolation. There is a point at which we can still detect viral RNA by TaqMan or NASBA and we can no longer isolate virus. So these, again, are our tests of choice for looking at those types of samples. It is recently that we have gotten looking into human serum in a more comprehensive way by TaqMan, and I will talk about that in a few minutes. [Slide] So, this is our testing algorithm for serum, whether from humans or if we are looking for seroconversion among sentinel chickens. We do the IgM and IgG ELISA. If negative, we are finished. If we get a positive result, we then need to do a plaque reduction neutralization test with some of the other related flaviviruses. I will show you in a minute that there is a cross-reactivity of the immune response so we need to do the plaque reduction test. [Slide] Very quickly, this is what the IgM capture ELISA looks like. We coat plates with goat anti-human IgM. Then we add patient serum diluted 1:400. We incubate for an hour at 37 degrees. We then add our West Nile recombinant antigen, and the final step is that we add horseradish peroxidase labeled anti-flavivirus monoclonal antibody. So, that is our IgM specific test. For IgG it is a little different format. We coat with anti-flavivirus monoclonal antibody. Again, it is an overnight incubation. We add our West Nile recombinant antigen. We then add patient serum and then we follow that with an anti-human conjugate that is alkaline phosphatase labeled. You can see from these two slides, as Dr. Wong mentioned, there is overnight incubation involved. That is one of the things that we are trying to move away from in terms of developing more rapid diagnostic tests for serological purposes, to get away from those overnight incubations because the test does take at least 24 hours to complete. [Slide] This was also mentioned so I will just go quickly through this, the way we determine positivity is we divide the optical density of the patient divided by the optical density of the negative control. We do our assay in triplicate. Our cut-off is anything greater than 3 is positive; less than 2 is negative. Then we have this equivocal range of 2-3 and we find greater than 95 percent of the time those turn out to be negative. It is pretty rare that something between 2 and 3 will hold up to be a positive. So, we repeat all of our equivocals. [Slide] This is to demonstrate the cross-reactivity that everyone that is familiar with flaviviruses would be aware of. These are four West Nile confirmed cases. You can see that the P to N to West Nile is, in fact, highest. However, these are all positive to Japanese encephalitis, St. Louis encephalitis, Dengue, less so to yellow fever. This also reflects the antigenic relationship of these viruses. West Nile, JE and SLE are all in the same antigenic complex. So, based on the IgM test alone, it is difficult for us to interpret that this is a recent West Nile infection. [Slide] This is the kind of data that we like to see. This is a single patient. You can see we have two acute specimens, CSF and serum, and then a convalescent serum. If you look at the P to N in the ELISA, strong in CSF. These are all three positive. What really tells us that this is a recent West Nile infection--this is the neutralization test here across the top with all the related flaviviruses--the highest titer in all cases is to West Nile. We also see a greater than four-fold increase in the neutralization titer between the acute and convalescent. So, this is what we would call a complete picture and in this case we can determine that it is a recent West Nile infection. [Slide] This is a typical West Nile serological case picked from the thousands we looked at in 2002. What we do is we run the West Nile IgM ELISA and the St. Louis encephalitis ELISA along with the neutralization, again, just West Nile and SLE. If there is travel history we will include other flaviviruses but we just include the two domestic flaviviruses at this point. One of the points that I would like to make is that this is a trend that we see, a trend that has been backed up by a statistical analysis. Greater than 99 percent of the time the signal in West Nile is going to be three to five times greater than the P to N to SLE. So, in this case we have 12.75 as opposed to 4. This is a very common trend that we see and we can make a presumptive diagnosis of recent West Nile infection but, again, the trend we always see here is a four-fold or greater difference between the SLE and West Nile neutralization titer, as well as a four-fold increase in titer from acute to convalescent serum. [Slide] I just want to talk about a couple of unusual features from our perspective of West Nile, the antibody response to West Nile infection. One of the surprises we determined following the 1999 outbreak in New York is if you look at these numbers here, basically after 200 days almost 80 percent of the West Nile cases still had IgM present and detectable in serum. Remarkably, when you get out to over a year, a year past initial infection, we have over 50 percent of the West Nile cases that still have detectable IgM in their serum. I think this has implications for discussion in terms of the value of an IgM-based test in donor screening. If we know that over 50 percent of the people are going to have IgM a year later, I think that is worthy of a lot of discussion. The numbers were small in that study but there is still I think good statistical support for over 50 percent. [Slide] From looking at several thousand West Nile cases, I will just summarize in these bullets, and the first point is that we find IgM nearly all the time by day of onset. This year, in 2002, going back and looking at the data, I only found two exceptions in the first 500 that we looked at. In other words, how many times do we find negative IgM on day zero or day one of onset, and then on a follow-up bleed we find out that this person actually had West Nile? It is very, very rare. We do find IgM nearly all the time by the time of onset. We find IgG in every case by day seven. I mentioned the trend of a three to five times higher signal in West Nile cases to West Nile antigen and to SLE. I also mentioned that we know that IgM persists for greater than a year. This is a topic that I won't really go into in any detail, but suffice it to say that if it is a secondary flavivirus infection it is very, very difficult to determine the most recently infecting virus, and we have a lot of data where we know that this is very problematic. It is going to be very difficult even with the neutralization test to determine the recently infecting virus. [Slide] I will just mention that since '99 the CDC has been involved in a lot of training of the state public health labs and other agencies. When this all began in '99 there was only a very few number of laboratories that were doing testing for West Nile. So, it was obviously to our advantage to get as many labs trained as possible. We have offered training courses every year since '99. We have trained over 60 public health labs. We do a proficiency panel where we send out blind coded specimens to all these labs and we are getting very good feedback. As you can see, 100 percent agreement on the IgM ELISA this past year and 92 percent agreement on the IgG ELISA. The problem in those cases where there was not complete agreement was that they missed some of our low positive serum samples. [Slide] What we would like to do in the future, wrapping up this serological assay part of what I am going to talk about, we would like to go in the direction of automation. You can see here a Qiagen robot, called a robot twister in a rapid plate. We are trying to automate as much of the IgM and IgG ELISAs as we can. We would like to look at reagent stability. All the things that Dr. Wong mentioned are very much a concern to us. The fact that you need to coat these plates before you do a run. The incubation times are also an issue. Again, this test is a 24-hour test. We would really like to try and shorten the incubation times. We have done some of that. The signal does decrease but we need to look at that some more and see if we can come up with a rapid version of this test. Finally, I won't mention this because, you know, Dr. Wong talked about it, but we would really like to move in the direction of Luminex as well for all the obvious advantages. [Slide] Let me talk about the molecular side of all the testing that we do. This is kind of a busy slide but this kind of gives a history of what we have done at CDC. All of these nucleic acid tests are based on really three steps. You have to extract RNA from your sample; amplification by some method; and then detection. I will have one slide about RNA extraction and kind of how we evolved there. Amplification, as with the industry, we have started with standard RT-PCR, followed by agarose gels. We have really moved away from this; we have never really used this at all for West Nile, just very, very early on. The West Nile epidemic in '99 coincided with the development of TaqMan assays in our laboratory so we never really spent a lot of time doing standard RT-PCR. We have done a lot of TaqMan for West Nile. We have also developed SYBR Green consensus primers to look at multiple flaviviruses and some of the other arthropod-borne viruses. I won't talk much about this but we have done a lot of work with NASBA as well, which is a transcription-based amplification system, looking both with electrochemiluminescence as detection or even molecular beacons. [Slide] Just very quickly, we have evolved in the way we process our samples for RNA extraction. We began with what I would now call some very ancient technologies, using liquid-liquid extraction, guanidine isothiocyanate and all those sorts of things. As many of you probably know, they are time consuming and they are limited in the number of samples you can do per day. We then moved into the silica gel-based kits. We can process more this way. In fact, we still use this technology when we don't have a very large number of samples that we are dealing with, but our method of choice now in the larger studies that we are involved in is Qiagen-9604 BioRobot, and this is a 96-well based system where we can do at least 300 samples a day, and this is really taking into account kind of a government work day of eight hours. I know that for those of you in industry, because I spent some of my time working for a private company, eight hours is not, you know, the limitation there; you could work 24 hours if you want. But in an eight-hour shift you can easily do 300 samples. In fact, we are using this now in a very large study that is going to be discussed here I think on the blood bank viremia survey, as I call it. Everything comes in bar-coded. The samples can go right into this slot, here. The bar codes are read and it is a beautiful system because in the end the RNA is eluted into a 96-well plate. The RNA can then be transferred directly into another microtiter plate and you can do the real-time PCR there. [Slide] I think everyone here is probably familiar with TaqMan chemistry. It is an RT-PCR. We need to do reverse transcription. Then we need to do polymerase chain reaction. The real-time aspect is where we have a probe besides the two amplification primers, and probes labeled with fluorescent dye, and there is a quencher and during the amplification Taq polymerase will chew the probe up so you can directly measure the increase in amplification there. [Slide] This is a typical result. In fact, this is from the blood study that we have looked at. Down here, in the baseline you can see actually the 96-well plate. Most of the samples, in fact all of them, are negative, except for our four standards. We include four positive controls in every one of our runs that we make at the beginning of the year and we quantitate the amount of virus. Our standard way of quantitation is to determine the number of plaque-forming units per milliliter. So, in every run that we perform we have these four standards. How you define positivity in TaqMan is time at which the fluorescence goes above the threshold. This is the Biorad I-Cycler. The instrument determines its own threshold and then it determines the time at which the threshold is crossed. We have settled on a cut-off for determining what is positive at 37 for two reasons. One, we found that among samples that we know are negative we see this occasional situation where something will just barely drift over the line after 37. The other is looking at replicates we find that we cannot reproducibly get a signal. If we get a signal of 38 or 39, you repeat that. It may be negative; it may be 38 or 39 again. So, we are trying to stay within the range where the assay is reproducible. [Slide] This is a study that we do every year when we make our positive controls. We look at the detection limit of the assay in terms of our standard quantitation method, which is the number of plaque-forming units. We take the same serial dilutions. We put them on vero cells and we count plaques. It is pretty straightforward. We take the same samples and extract RNA and perform the TaqMan PCR. There are so many ways you can define these things. I am just going to settle in on the number of plaque-forming units per milliliter. We have two primer probes that we use in our West Nile assay. We actually have three but we spend most of our energies on these two. One is in the envelope gene; one is in the 3' non-coding region. You can see that there is a difference in sensitivity, about a three-fold difference. We have transferred all this technology. Basically everybody I have talked to has noticed the same difference between these two primer probes. For comparison, we also have synthesized a set that has been published by Dr. Lipken's laboratory. You can see that those primers are really somewhere in the middle between these two. [Slide] If we talk about copy number, which in some cases is more relevant to what many people in the industry look at, I have looked at two things in our lab. Both are plasmid and just a double-stranded DNA as a target. These are quantitated by optical density and also running an agarose gel with standards. You can see in this case that we can detect five copies of a plasmid containing the envelope gene. This is with the envelope set which is the most sensitive set. This is not a concentration; this is actually the number of copies in our test volume of 5 mcl. So, five copies here. When we do the same kind of experiment with a double-stranded DNA molecule, not a plasmid, we are at 12 copies. Dr. Kramer is going to talk, I am sure, but in her laboratory's publication it looks as if they are talking about 37 copies of single-stranded RNA. So, I think this is fairly good agreement because we are really talking about two different kinds of amplification. Here, this is just PCR. Here we are talking about including the RT step since RNA is the starting point. [Slide] The other thing we do in our lab is we have an internal positive control. This, again, is the same figure I showed you a minute ago, looking at a whole 96-well plate of samples. At the same time that we have the West Nile primer probe set, we have added an artificial RNA molecule and we have a primer probe set that is labeled with another fluorescent dye called JOE. You can see that the only thing here positive are the positive controls. Down here everything comes up as we would expect. This is to ensure that there was efficient amplification in all wells. [Slide] This is just another way of looking at the data, comparing TaqMan. I haven't gotten into this much, but we have developed nucleic acid sequence-based amplification for West Nile, and this is typical of what we see. The sensitivity is about the same. On one day NASBA might be better, on another day TaqMan. [Slide] Specificity, just to point out that the TaqMan and NASBA assays are specific not only for all West Nile strains here, but in the case of our envelope it is actually specific for a particular phylogenetic clade of West Nile viruses, and no reactivity with other arthropod-borne viruses. [Slide] Dr. Petersen mentioned this. We have looked at the complete genome of West Nile strains in the U.S. since '99. We have nearly completed a 2002 strain and we are finding the same thing from year to year, just very minimal, 15-20 nucleotide changes from year to year and most of those silent. In fact, with the 2002 strain we have only seen one amino acid difference when you compare it back to the '99 strain. So, they are very genetically stable viruses. [Slide] I will just summarize what all this is showing us about viremia and, again, Dr. Petersen talked about this but we know from everything we have done so far that viremia in humans is very low. If you take all of the positive TaqMan, positive human sera from this year, the range is between 100 to 150 PFU/ml with an average of 18 PFU/ml. I will just point out this should initiate a lot of discussion about NAT testing in pools. People ask me about that all the time. I think that if we diluted 1:16, of all the ones we looked at this year we would lose about half. Half of them would no longer be positive. Viremia is short lived, and this is another point that we have a considerable amount of data on where we have looked at a lot of West Nile cases from this year from which we have day zero or day one onset of symptoms, and in that situation it is less than five percent of the time that we can detect West Nile virus. In fact, this year we only detected 2 out of 500 or 600 that we looked at. So, by the time there is onset of symptoms there is not really detectable viremia. The other point is that there is a very, very short window when you are going to detect both IgM and West Nile virus by these nucleic acid tests. Again, this year we only found four that fall into that category where they are IgM positive and they are TaqMan positive as well. There are other studies out there, studies that we were involved with this year in Louisiana where we went looking for viremia. We went to clinics. My colleagues at CDC did this study, looking for viremic people and what we found was a fair number of IgM positives but not a single viremic case. [Slide] This is how I envision things based on the data. We have viremia that precedes onset of illness, and then the appearance of IgM at onset, and then this very, very short time period where you can detect both IgM and virus. [Slide] The same thing applies to TaqMan. We have transferred this technology to all the state public health labs. We got pretty good agreement, not quite as good as with the ELISA. There are issues in the state labs, false positives, as you would expect with any amplification assay. Some labs had problems with false positives. The other extreme is failure of some of these labs to detect our lowest positive. We sent a very, very low positive and some labs didn't get that. [Slide] This is my last slide, just to acknowledge that I have a great group of people that I work with at the CDC, and this work represents a collaborative effort between all of us. Sometimes people are surprised; they think the CDC has hundreds of people working on this. This is it. [Laughter] Two of the main people that do the TaqMan testing, and I have done a lot of it this year because one of the main people got pregnant and had a baby. Anyway, I want to acknowledge them for their hard work and support at CDC, and thank you for having me this morning. DR. RIOS: Thank you, Rob. We would like to break now for ten minutes only, if you don't mind. We will be back at 10:35. Thank you. [Brief recess] DR. LANCIOTTI: We need to begin to make some attempt to get back on schedule. Our next speaker is Dr. Robert Myers, from the Maryland Department of Health. He is going to talk about assays that they are using in their lab. Dr. Myers has been a great support and colleague to us at CDC. He has been one of the labs that we have relied on for second opinions and for follow-up testing, and so forth. Dr. Myers? Utilizing Nucleic Acid Amplification Technologies in Arboviral Testing Programs: Practical Experiences from the Maryland Public Health DR. MYERS: Thank you. [Slide] Basically, as Rob mentioned, the CDC has rapidly transferred their developmental technologies to public health laboratories and we basically validated and implemented them. We are very grateful to them. Our technological capacity has increased remarkably over the last three years. [Slide] Basically I am going to give you a very brief overview of RT-PCR technology which is the cornerstone technology for us. We are going to talk about the applications to our public health-related arbovirus testing activities. I am going to show you relative levels of West Nile virus RNA in various sample matrices--birds, mosquitoes, horses, humans. Because this is basically a blood transfusion committee, I am going to relate our experiences with handling West Nile RNA positive human specimens. [Slide] Over the last four years since West Nile first appeared in the country, arboviral testing programs in public health laboratories, which have languished for decades, have really taken off. I can say from personal experience that over the last three summers we have devoted a great deal of time and resources toward these testing activities. [Slide] These include surveillance--dead bird surveillance, mosquito pools, equine and mammal surveillance and, of course, human surveillance for encephalitis and meningitis patients, as well as doing some diagnoses. [Slide] Real-time PCR technology is really the cornerstone of our laboratory as far as nucleic acid amplification. It allows for the simultaneous detection of agent-specific amplification, detection of specific agent nucleic acid sequences; of course, as Rob mentioned, the specific hybridization by complementary probes and, of course, there are several reaction strategies and instruments that are used in real-time PCR assays. [Slide] Advantages are high throughput. We have really done tens of thousands of these assays in the last few summers. Without this technology we couldn't have. Our results turn around quickly enough so that public health intervention, so to speak, mosquito control and public education campaigns for arbovirology could really be useful. It does reduce but doesn't totally eliminate cross-contamination of PCR products. You will see that the levels of viral RNA are very high in birds, for example, and mosquito pools and you can get cross-contamination even at the level of unamplified RNA. Sensitivity, as you saw before, low copy numbers can be detected; multiplex capabilities, which we do utilize in our laboratory; and, of course, you can get a fairly accurate quantification over a fairly wide dynamic range. [Slide] TaqMan, very quickly--you see the hybridization probe, the quencher molecule, the reporter molecule. It cannot be extended by Taq. Of course, the biophysical principle is the light energy coming in to excite the reporter molecule. The wave of the energy that is transferred is given off to the quencher molecule as long as they are closely together and, therefore, no fluorescence is given off. [Slide] After the RT step, when you are actually doing the DNA PCR, of course, the probe binds. You get an elongation from the Taq. The probe starts to be displaced. It is chewed up and broken up, separating the quencher molecule from the reporter molecule and now the reporter molecule is in solution and is excited and gives off fluorescence that is picked up by the instrument. Of course, the more product you have the more fluorescence you have. [Slide] This is some of our bird data, just giving you an idea of what a routine one looks like. This is our control. This is our lysate control to make sure the RNA extraction works properly in each batch. These are some positive birds. This is the envelope. [Slide] Two of our instruments--we have an ABI 7700 which is a 96-well instrument, and a Roche light cycler which is a glass capillary tube, 36 position. This is more of a work horse. It is an older instrument. It does a data dump. We have to wait three or four hours for results. We truly do get real-time results off the Roche light cycler, which we do for a lot of our confirmation testing. In addition, this summer we have gotten a Biorad I-cycler. [Slide] As mentioned, at the lab at Fort Collins and at our lab too, we do have a robotic extraction system. When we are doing thousands of tests, normally we can do 96 specimens in about an hour and a half. Our human work, our volume is not very high. Because we are handling high titer bird samples and mosquito pool samples, our human work is done manually. [Slide] This is our normal testing logarithm for PCR testing. We use two multiplex TaqMan PCRs. We use the WN-3, the most sensitive set from CDC that was developed, the internal positive RNA control, the IPC control. That has worked for us. The Patuxent Wildlife Center in Maryland was feeding heavy metals to birds and the birds were dying, and they wanted to make sure they weren't dying of West Nile, and it totally inhibited our PCR. It also controls for technicians to make sure they added RNA to the reaction mixture, reaction plate. We also look for eastern equine encephalitis virus which is endemic here, in Maryland, on the eastern shore. We do get a few cases every decade. We also are looking for SLE. We had a LaCrosse case appear, a serological LaCrosse case in Maryland last year. For selected mosquito pool types and also all of our human samples we also run a singleplex PCR, real-time PCR for LaCrosse. [Slide] All of our positives are confirmed with additional primer probe sets for West Nile. This is our screening. We use the Lifkin. There was one developed at Walter Reed, called 156, that we use on the light cycler and, of course, the WN-3 and, of course, the non-structural polymerase gene of West Nile. [Slide] As you know, certain bird species, particularly the corvids, are highly susceptible to West Nile. This is a very important component, our West Nile surveillance plan. Birds tend to become positive first in an area that West Nile enters. In this particular area here, in the Washington, D.C. area, this year the Virginia State Health Department picked up a bird in the Virginia suburbs of Washington in late April. We picked one up in the District of Columbia in early May. I don't think we saw mosquito pools positive in this area until July, and we didn't see human cases probably until late July. So, really they are a very early indicator of initial activity and it is an important program. [Slide] This is just a summary of our testing. We have had 1800 submitted. About 1700 were suitable for testing. They come in all states of decomposition. Roughly one in three are positive, 18 different species, and about 92 percent of the positives are either corvids or raptors. In Maryland we cut off at two per zip code. The District of Columbia stops testing after they basically have ten positives, I believe, in each ward. They stopped probably in early August. Our West Nile surveillance program just stopped last week. [Slide] These are our average CTs or distribution of our CTs, 600-and-some positives that we had. As I mentioned, the brains are in various stages of decomposition. We have had things with maggots crawling through them and road kill, and we still can pick up and confirm levels of West Nile. As you saw before, there are extraordinarily high levels of viruses in birds, 1010, 1012 experimentally. We have less than pristine samples and our estimation is about four logs of RNA PFU equivalents/ml. [Slide] Mosquito testing is by our Department of Agriculture mosquito control. We have pools of 1-40 individual mosquitoes. We test them for West Nile, EEE, SLE and certain species for LaCrosse. They are grounded to a suspension of collagen B-bill mixer, and then we also save some of this homogenate before activation to go into tissue culture. We are concerned about emerging infectious arboviruses being imported into our country through containerized cargo or at airports. I am sure you are familiar with the malarial situation here, in Maryland and Virginia, this year. With PCR you only find what you are looking for. The tissue culture casts a broader net for us and occasionally we find things. So far they have been relatively benign, like Jamestown Canyon. We tested 7400 pools; 46 were positive by PCR for West Nile. Pools represent over 100,000 individual mosquitoes. [Slide] This is our breakdown of CTs, primarily Culex and also Aedes albopictus, the Asian tiger mosquito which is an import. One of our positive pools ended by on the editorial page of Science this summer. There was an editorial comment about globalization of a pool that we found here, in Montgomery County. This is our average CT and it looks like it is about 3.5 logs. When we have done tissue culture probably 85 to 90 percent of the time we can isolate West Nile from our PCR positive specimens. In 2001 we did find EEE on the eastern shore and Culiseta melanura, which is a swamp type mosquito that transmits to birds. We also had a positive bird in 2001