NIH CLINICAL CENTER GRAND ROUNDS
Episode 2010-08
Time: 1:06:30
Recorded March 3, 2010
How to Publish Without Perishing: Navigating the Biomedical Publication Process
Cynthia E. Dunbar, MD
Head, Molecular Hematopoiesis Section, Hematology Branch, NHLBI
Pathogen Inactivation of Blood Components
Harvey G. Klein, MD
Chief, Department of Transfusion Medicine, CC
ANNOUNCER: Discussing Outstanding Science of the Past, Present and Future - this is NIH Clinical Center Grand Rounds.
(Music establishes, goes under VO)
ANNOUNCER: Greetings and welcome to NIH Clinical Center Grand Rounds, recorded March 3, 2010. Diverse topics are on the menu for today as we bring you Dr. Cynthia Dunbar, head of the Molecular Hematopoiesis Section in the Hematology Branch of the National Heart, Lung and Blood Institute. Her topic, "How to Publish Without Perishing: Navigating the Biomedical Publication Process." She will be followed by Dr. Harvey Klein, chief of the Department of Transfusion Medicine at the NIH Clinical Center, who will discuss "Pathogen Inactivation of Blood Components."
We take you to the Lipsett Ampitheater at the NIH Clinical Center in Bethesda, Maryland for today's presentation.
DUNBAR: Originally I was asked to speak about ITP a topic I have been involved with for the last two decades, but I didn't have anything particularly new to say about ITP.
I was uncomfortable just presenting other people's data. I was asked instead if I might give a talk on sort of practical issues of getting your papers published in the biomedical literature. This is not directed particularly at publishing your papers in "blood" because I know many of you are not hematologists. I think the process editors think about will help you have insights to apply whether it's clinical research or basic research in any one of areas.
In terms of conflict of interest, one conflict, I am the editor in chief of a journal, owned by the American society of hematology, not for profit. I have no relevant commercial interests to disclose. So in terms of the objectives I'm required to state I want to help you understand how biomedical publication processes occur and hopefully identify and help avoid pit falls that occur in manuscript preparation and submission and get you on to successfully publishing without perishing. Major steps, and the most important steps and the ones I'm not going to talk about a lot except in terms of pit falls that you might want to think about is planning and execution of your science or clinical trial that will be up to you and your mentor or coworkers to figure out.
But more importantly, I'm going to talk about the steps of writing and marketing your paper for publication. There is lots to do. The sociology, the ethics, morals, philosophy of scientific publication. Right now this whole area is changing unbelievably rapidly as is all of journalism. I think Dr. Gallan should invite Neil Young or others who have written reasonably in terms of the conceptual and philosophical issues of whether or not biomedical publication in science is a commodity in a sort of economic sense and whether we should be trying to restrict the marketplace by rejecting papers. I'm not really going to talk about that either.
I want to help you navigate the process as it is, not as it should be. Talk about the nuts and bolts of submission and post submission. This is what we're aspiring to, actually publishing truly important papers. This is Watson and Cricks original description of the structure of DNA. I will point out this was the entire paper. So concise, to the point, telling a nice story is what you should be aspiring to, not necessarily 15 figures and 50 pages to get across what should be relatively hopefully straight forward points and make them in a way that they are clear and important to your audience. And especially as many people come through, especially graduate school when they're writing dissertations, supposed to be writing a 100 page introduction and citing every paper ever written in the field and explaining their methods in incredible detail, that's not what necessarily you're trying to do when you're writing for the biomedical literature in actual journal publications.
So sometimes it's an adjustment as trainees come through from writing their PhD. Dissertation into writing papers for journals in terms of figuring out how to be concise and tell a real story in a format that may be different from what they're used to. So in terms of planning and execution of your experiments or clinical trial, I think that it's important always to think ahead when you're doing your actual data collection and your actual experiments about what the criticisms might be of your work. And when you're actually writing it up for publication or writing it up for a grant, because really, the two processes are quite similar. You need to plan your controls carefully, especially when you're working in translational science where you may have very limited access to patient samples. You have to get it right the first time.
With more and more, unfortunately, cases that it would seem of scientific misconduct or actual fraud, it appears that some of these instances occur because people just didn't have or do the right controls in the first place so they had to make them up when they're actually trying to write things up for publication. Try not to put yourself in that position, at risk of being tempted to doing things the wrong way because you didn't figure out in advance what you would need to tell a complete story and how to tell the story appropriately with all the controls you need.
Along the same lines it's important to seek appropriate statistical help ahead of time. With clinical trials it's more likely you'll do this in the first place, because IRBS and your scientific review process within the institute and if you're working with a drug company, the drug company itself because it's so important for patient protections and so important for going to the FDA to get a drug approved, that you have your trial powered to give you interpretable information. I don't think this as being a problem in clinical trials as it is in the laboratory realm or when you're working with animals where you may not automatically have to get a statistician involved, but in many instances it's important to do that and think about do you have a sample size that allows you to answer the question that you're starting out trying to answer. So do that before, not after, because it really isn't so valid to keep doing experiments until you reach AP. Value that's significant. When you're doing data acquisition, think about how you're going to present this data, whether it's a talk like this, a paper, a grant. When you put your lanes together on a gel or you're photographing cells, try to get things in an order that makes sense. Have the controls in the right place on the gel. Have your experimental samples from two different conditions grouped. I can't tell you how often people get into trouble when they splice different gels together or splicing within gels. And we'll talk more about image manipulation later.
Because when they did the original experiment they were trying to be perhaps efficient, not realizing it would make it difficult later to put things together in a reasonable way. Also always perform experiments to test alternative explanations of the data. Don't go in trying to prove your point, not being open to look at alternative interpretations. So let's say you've gotten your clinical trial completed. You've finished your experiments in the laboratory. Now you're thinking about writing up your work for publication. You need to choose the appropriate target. You have to have a realistic assessment of how novel, how important to the scientific community at large verses a very restricted scientific or clinical community that works on the same things you do. Is your data really broadly important to those outside your direct field? How strong is the data? Do you have many different ways that confirm your conclusions? Do you have statistical power that's really convincing? Those are all things that can help you and hopefully, coworkers, mentors, decide where you should submit the paper? You also have to think about what return you're in. Obviously, if you are not really in a huge rush because your board or site visit or grant application isn't due in 6 months, then you can take the chance of submitting to a very high profile journal that will look great on your CV with very high likelihood that it's going to get rejected immediately or after review. But you can hopefully make that assessment after talking to others, looking at the journal, and seeing what kind of articles they publish.
I sometimes get the impression that people who have submitted to blood have never looked at the journal. I get papers that are on diabetes. Not transplants for diabetes, they're just on diabetes, on -- I don't know, on cardiac surgery, because the heart pumps the blood. I can't figure it out. It's wasting my time and wasting their time and money to submit to a completely the wrong target. So try to think about does this fit. If I were reading the table of contents of blood or circulation or whatever would I be surprised to see this article in the journal.
If I would, no matter how high the impact factor is, don't submit there. So does you topic fit the substance and style of the journal? Think about some journals have very strict length requirements, only allow one or 2 figures no matter what. If you need more space to tell the story, then it's not the right place. Even based on something as mundane as how many figures you're allowed to have. It's important if you're not sure, send a presubmission inquiry. Most reputable journals are happy to receive presubmission inquiries to not waste our time and your time. Write a brief cover letter. Attach the abstract so we can tell what was actually done. Most of the time the editor or the designate will get back to you in a couple of days, and say it sounds interesting, submit, or it doesn't fit our journal at all. If you submit and get an instant rejection, don't get too upset. When you see the entire paper you may see clinical trial had five participants, it wasn't going to make it. It is something you should consider it if you are not entirely sure.
Think about what audience you're hoping to reach. I think I've mentioned that in terms of the table of contents and looking at what's in the journal. Impact factor is not the only factor. I would say at NIH, and in the United States in general, overreliance or overworship of the impact factor is not as big a problem as it may be in some other parts of the world where people actually use formulas based on the impact factor, a number of papers to determine whether people are promoted. But, you know, it should be realized that had this is a very controversial and very flawed in many ways. But it's unfortunately the major metric we have. You need to understand how it is constructed. It's basically the number of times that an article published in 2007 and 2008 is cited in articles in 2009 divided by the total number of citable articles. I'll talk about that in a minute. Published in those 2 years. It's an average number of times that a paper is cited in the 2 years previous to the calculation of the impact factor for each year.
So how can you manipulate the impact factor if you're a journal editor or commercial publisher? There is all kinds of ways. If you only publish review articles, you'll do great, or if you have a review to primary research article that's very high, something like 60% of the top 50 journals are solely review journals. They should be considered completely separately and shouldn't calculate impact factor at all based on review articles. Thompson scientific, the company that does it, are a very controversial issue that so far it's for profit company, the journals pay them. It's very complicated. I don't have time to talk about this today, but it's true. Only publish your journal in the first half of the calendar year. Articles that are published at the beginning of the calendar year have much more impact, they have a longer time to get quoted. In the 2 years previously than an article published in December.
Don't publish negative studies. Don't publish any articles on rare diseases which is a constant struggle for probably those of us at the NIH who like to publish in high profile journals, and think rare diseases and things you learn are very important. But you may not know how important they are until later. Anemia took 10 or 15 years to understand these proteins were very important in breast cancer. If that work had not been done, nobody would have realized that. Rare disease articles don't tend to be cited as much.
Publish lots of short commentaries and letters to the editor that are primary research letters. Letters to the editor don't count in the nominator, only in the numerator. Same with commentaries. And publish the fire sequence of any genome. 80% of the impact factor for nature one year was completely subsumed by the human genome. You can imagine that.
Basically, the point is blockbuster papers are responsible for a huge percentage of many journals impact factors. Clinical trials, they're very, very good and very strong at getting other people to reference their own trials or other people in their company to reference trials, even if it has nothing to do with the drug they're writing about. Premature, novel work is very good. Completely made up fabrications seem to do a good job as well. The human cloning papers in science that turned out to be completely fabricated from Dr. Wang. There were 417 citations to the retracted paper after it had been retracted in 2007 alone. And these papers have continued to be cited since that time. So, you know, it's still counting for the impact factor. So it doesn't matter if it's actually right. You can publish and retract it and it will still help. So this is a highly flawed metric. Ok, substance of the paper. Think about these are -- I've gone through these in terms of figuring out your target. Is it significant, novel, descriptive or mechanistic, is your data presentation okay. A complete story or minimal publishable unit. We try to avoid minimal publishable units. You're not really telling a full story. There is lots of issues with that. Every individual project that is different issue that decide when you're ready to publish. Sometimes I've seen in individuals, especially at NIH because we don't have the grant pressure in the same way, sometimes go too far the other way, and try to have a perfect story, the perfect can be the enemy of the good and can take 20 years to finish the story and that drives away postdocs and trainees quickly.
But somewhere in the middle is good. And things like cofirst authorship which has become quite common. Most journals will allow that without any difficulty and it is one way around the issue of trying to put a complete story together, even if 2 individuals really contributed substantively and received first authorship. Sitting on tenure committees and grant review panels, 2 people is fine. You get about equal credit for a very important cofirst author paper than you do for publishing an incomplete story submitting it up into 2 parts.
In terms of style, I think this is very important to follow the journal instructions very carefully. Look at the instructions before you start to write. The abstract is by far the most important part of the paper. It's the only part of the paper that the editor will probably read at all before they decide about sending it out to, you know, an editorial board member or whether they're going to reject it without review if it doesn't seem to be on the topic or behind a quality for the journal. You hope they will look at the text of the paper. They may not. A lot of people leave the abstract until last. They might waste in an abstract they presented at that meeting and didn't even update, has the wrong number of patients. Don't do that. The abstract is critical. Don't make typos, don't make errors in the abstract. It's also very important because if the paper is accepted, the abstract is unfortunately about the only part of the paper anybody is going to read. So they're not going to cite your paper if they don't understand what you did just from reading the abstract. So the abstract itself should be a concise clear story. You don't have to include numbers. You don't to have include p. values. You don't have to include references. Do include what the entire story is and what your main supportive data is and what your conclusions are in that abstract. Write concisely. We've already told you that. We have very short attention spans and getting shorter. The introductions do not have to be a graduate school dissertation. They need to clearly explain why you're doing this study and what relevant background material there is in the discussion, describe clearly what it means, what might be important to go forward. But don't write too much. It's always a mistake.
Read the papers you cite. Accurate referencing is very important. A, it's good science and good practice and shows you're doing your homework correctly. B, if you don't, nothing pisses off a reviewer more than if you don't. You regularly get comments that this author is a moron, they didn't even read my paper and they cited saying the opposite of what my paper concluded. Read the papers you cite. It's important to accurately reference them. You don't have to reference every study. If it's an important review that goes over a decade's worth, you can say as reviewed recently and give that reference, then you don't have to cite every paper and worry whether you cited the right one. As long as you're fair.
The methodology, the other thing that pisses people off, if you cite a paper for a method and the method is not in that paper, that may not lead to rejection of your original paper but will lead to colleagues being annoyed at you and not look favorably at the next paper.
Beware of duplicate publication. Hopefully you've done this before you started to figure out whether somebody else has done exactly what you're trying to publish. Even if it turns out that a month before you submit your paper your colleague across the country publishes an article that shows more or less the same thing, don't try to fool the editor or the reviewers into thinking that paper doesn't exist. If they're at all good, they will notice that and have done a search or read the literature anyway. So don't blow it off. Reference it, discuss it, and point out why it's important to confirm the study or to -- how your study goes beyond what's already been published.
Write in English. If English is not your first language, get help. It is impossible -- I send back without review papers I can't read and cover letters that I can't understand. Try to avoid describing your own data in the text of the paper as extraordinary, amazing, remarkable. Don't use those words. Tell the story in such a way that the reader decides that it's remarkable or extraordinary and amazing. Go back and read Watson and Creek. They don't use those words, but everyone knew they understood and everybody else understood as soon as they saw the data, how earth shattering it was. Avoid personal attacks on competitors or their theories in the text of your article. They may be your reviewer.
Special content areas. I'm a little bit behind time. I'm not going through this. But read what the guidelines are for publishing microarrays, make sure you're doing it right, microscopy. Include all the important information required to understand the images and on clinical trials before you start, it's now required for most journals that you have registered your trial in advance for patient accrual on clinical trials governor or other open access source of clinical trials, accrual information. IRB approval appropriate consent. If doing a randomized phase 3, that you're follow the consort guidelines.
Figures and tables should be self- explanatory. Plan them on what the final size is going to be on a page or computer screen. If the journal requires only one figure or table, that doesn't mean put 30 panels in one figure. That will be approximately the size of a pin head on the page when you're trying to look at them. So if you need 10 figures, then you need to submit to another journal and make 10 different figures. Don't try to cram it all in. Created your figures and tables first, not last. If your figures and tables can't be constructed in a rational way, you probably have to go back and do more experiments and get the data. Don't wait until the end after you've written beautiful text. In the interest of time I'm going to go through this quickly. Image manipulation is a larger and larger issue. We're using so many sophisticated digital methodologies. You can see here that the Russians figured this out first. And when Trolsky came out of favor he disappeared from pictures. The more recent presidential campaign, somebody decided that James Fonda standing next to John Kerry would be an interesting image but it wasn't true, it was pasted together. There is lots of reasons it's wrong to manipulate images. You -- if you've heard Peter Agee talk about in terms of the hydrogen channel on the red cell membrane, on the original western blots, he thought it was a contaminant, a mistake. If he had just cut those off and put in the journal without the other things he thought were unimportant and mistakes, may have never figured out what eventually led to the discovery of it. Because I believe a reader first asked him about whether this could be related to different explanation than just being a contaminant.
If you look in most journals, there are image presentation guidelines -- I can send these to anybody that wants them. Basically, the rules are don't put together things that didn't actually occur together in real life and don't change specific features in a figure if you don't change the entire image, contrast, gain, et cetera. A lot of ways we find these inappropriate manipulations are based on software that were based on art authentication in terms of art fraud. We're looking for inconsistencies. We find them frequently. Most of them are due to not being educated. But some of it is true data manipulation and fraud. So this is the kind of thing we find all the time.
You can see this even without anything very sophisticated. They paste it in bands here. You can find it very easily. We got the original blot. We found out it was all one blot. There were lanes with bubbles in them. They were spliced out. We asked them to put in lines and explain what was done. Here somebody thought 3 cells would look better than 2. They weren't actually on the same image field through the microscope so they dropped another one in. We found this. That's not okay either. You can just put a line and as long as you make it clear they were taken from a single slide, a single patient, a single image, but put together in this way as long as you disclose it.
How to avoid trouble. Keep your original files. Don't convert them to jpeg. Keep them as tif or other resolution in case you have to remake the figures. Make sure you understand how the data was collected. Apply the same adjustments to control and experimental and disclose the adjustments in you figure legends.
Your cover letter. Be concise. You don't need a 3 page explanation why it's important. You need a couple of sentences. Revise your letter if you're sending to your second choice journals. 10% of the time I get letters about I really wanted to believe you should publish this in nature. I'm actually at blood, so did you mean -- I guess I'm going to reject it because you didn't want to submit it me. I don't usually do that. But it sets your hackles up when you know you're the second choice some of the time but you don't need it graveled into you. It's important to exclude people you used to be married to if you're in court or a court case over a patent or a direct competitor that you had bad interactions with. When you exclude more than 2 or 3 people you seem paranoid. And if you exclude every person in your field who might have the expertise to review your paper, that's a big problem for an editor and decide it's not worth it. They know there is no way that 10 different people can be unfair reviewers. Keep it to a minimum and you'll be better off.
Post submission issues. Respond carefully and completely to reviewer comments. Doesn't mean you have to take them as truth. If they misunderstood your paper, explain in a measured and rational way why you shouldn't have to do this control because they just didn't understand what you were doing. But try to address every one in your response letter. If there are 3 or 4 big things asked for and you can only do one or 2, e-mail the editor back and say I can't do this. It's better to do that before you go through a whole revision process. Hear from the editor what they'll accept. If they say forget it, go to another journal. It's better not to send something back in with major holes just because you – hoping they would ignore it. That doesn't work.
If you get a rejection that you feel is not fair based on substantive misunderstanding or misinterpretation of your data, you can appeal and it may work. I'd say about 5-10% of our rebuttals are eventually accepted. And it helps to be not hysterical. It helps to not make death threats in the rebuttal. Or physical -- I mean I got some that were close. It helps to be rational. And -- but after they say no the second time, that's probably when it's time to give up. When you finally get accepted, don't get them to your secretary or someone in your lab who had nothing to do with is the project.
Those gallies are really important because they are the last chance you have to really fix something. It's embarrassing to have to publish corrections. So look at those carefully yourself if you're the primary author on the paper. Authorship should be negotiated far in advance. It's not the journal's problem. If you're fighting with your coauthors about who is first and one of the postdocs e-mails me to say I should have been first author, I'm supposed to do something about it? I say paper is on hold. As soon as there is problems with that, we just stop and send it back to the corresponding author and say fix it. It's much better now, because every author is now notified at the time of submission. We need the e-mail addresses. We have less problems with people not finding out until the paper is in print, and trying to bring a lawsuit. Also realize what the requirements are for sharing a publication related materials. If you don't follow that, many journals will ban future submissions from your group.
There is lots of issue of conflict of interest and industry. I don't have time to go through that today. It's a whole other topic. Definition of a ghost is an author that contributed to the writing of the paper but was not every acknowledged or listed as an author. There has been a lot of publicity on this in clinical trials where basically the authors listed had never seen or written the paper, and in court said I know I was first author. It's not my fault that the review article said to use a drug I did not believe in because I never saw the paper. This has been a major issue. Jama survey indicates that like in journals like the New England journal, Jama and others had unacknowledged ghost writers. Review articles and lower peer journals, some surveys say 75% were written by somebody else and in some cases never read carefully by the listed author. You can yell at drug companies and try to sue them, but authors have to stop agreeing to this sort of thing and prostituting themselves.
I'll finish now; hopefully you'll be Stockholm to get a Nobel Prize because you published so many fine papers. I'm sorry for going over. I guess I can take a question or 2 at the most.
[APPLAUSE]
QUESTION: I'll start with one question. A friend of mine is an editor, associate editor in the New England journal of medicine told me they are requiring to seek protocols for all clinical trials in the reviewing protocols in the context of the paper. Is this becoming more commonly used and is blood using it? What do you think of it?
DUNBAR: This is becoming a major issue in journal publication. I went to a meeting in Vancouver that involves a lot of high tier clinical journals. We have not done that. Some authors do submit the protocol and supplemental material. Some reviewers ask to see it. If they ask for it, we, do, ask for it to be submitted. The consort guidelines, although they don't provide the clinical trials provide a lot of important information in a checklist that's more easy to digest anyway. I think personally right now we're not going to require this. Most of what we want is in consort or clinical trials.gov.
In terms of eligibility criteria, et cetera. I'm not really sure knowing what the dose is -- I'm not sure a lot of what's in clinical trials is going to be relevant. But it's an ongoing issue. I think it's an interesting one.
Maybe at the end Dr. Dunbar will be available.
KLEIN: Thank you, John. It's a pleasure to be here today. To give full disclosure, I was coauthor of Mollison's 11th Edition. I hope you'll buy a copy of the book. I received no royalties from this book. Dr. David Anstee, my coauthor received some, but Blackwell Publishing makes a lot of money from this book. I have no other disclosures to make.
I will be discussing some non licensed technologies, all of the technologies to inactivate pathogens, are either not licensed or not used in the United States. The objective is to try to convince you we do need effective pathogen inactivation for blood components, to give you a review of the currently available systems, none of which are licensed in the U.S. and to provide an overview of the trends of pathogen inactivation of blood components worldwide.
Now, there are a lot of agents that are transmitted by transfused blood. Over the years, we've developed a number of systems for making blood safe. You all know about the donor history which is more or less an inquisition these days. We examine the donors and the cornerstone is our testing of blood specimens from the donor for infectious disease. But we also divert the first few cc's of blood so that you don't get bacteria from the skin. We leuko reduce blood, so that some of the agents that are cell associated are removed such as cytomegla virus. We inspect the components. We get donation information, after donation, and if a donor is sick, we quarantine or get rid of the unit of blood. We also have deferal registries for individuals who have been infectious for prior donations. We limit exposures to blood. We look at the hemavigilance.
This slide is Dr. Harry Alter's slide who is in the audience. It shows the experience here at the National Institutes of Health. Where if you came here in the 1960s, for open heart surgery, you not only received a lot of blood. But you had about a 30% chance of leaving this hospital with hepatitis. Over the years, the introduction of a variety of measures including going to an all volunteer blood donation system, screening for hepatitis B, screening for hepatitis C, screening for HIV, have reduced the risk to virtually 0. It's obviously not 0 but it's so low that one can no longer do these kinds of studies that Dr. Alter carried out over the many years.
In fact, it's not only hepatitis that's been reduced but it's been HIV as well, and they have been reduced to the extent that we have to calculate infectious transmissions these days. We can no longer measure them directly. And, in fact, the risk of HIV is somewhere about one in every 2 million units transfused, hepatitis C, about one in every million and a half units transfused. Hepatitis B, maybe 1 in every 200,000 units transfused. I won't see one of these in the clinical center for the rest of my career if I'm lucky.
It's not quite the same with bacteria especially in blood stored at room temperature such as platelets. About 1 in every 3,000 units is probably contaminated when we culture them. About 1 in every 75,000 results in clinical sepsis from platelets. And about 1 in every 750,000 results in a fatality.
So blood is pretty safe but what are the sources of the residual risk? Well, there is what we call the window period, that is, a donor is infected, say, by hepatitis agent, but comes in before our test is sufficiently sensitive to pick it up. That's the zero negative window. There may be viral variants. We've seen strains of hepatitis agents and HIV that wouldn't be detected by current tests unless they're modified.
You can have individuals who are infectious but negative for an antibody screen. There are errors in testing, those are rare. And for testing for bacteria, if you don't sample where the bacteria are, a few bugs can get through and grow in your units of blood. The big risk is new agents.
I'm going to give you one example of that which is West Nile virus. In 1999, probably as a result of mosquitoes coming into Kennedy international airport, West Nile virus was introduced into the United States. Over the next several years you can see by the year 2000, West Nile virus in mosquitoes and in people spread down the east coast. By 2001 it reached the Mississippi River. By 2002, it was already to the west coast. By 2007, West Nile virus had much reported in every state in the United States except Alaska and Hawaii. It's now endemic in the United States. West Nile virus is transmitted by blood. We have a test for West Nile virus. It took several years to get it available for screening blood donors but there was already a test when West Nile virus came to the United States. It was investigational test. It didn't take all that long to get the test available for screening blood. Nevertheless, there were probably thousands of people infected by blood transfusions with West Nile, and hundreds that we knew of who had substantial morbidity and some mortality. And West Nile virus is a relatively mild infectious agent.
Let's go back to the HIV era, it entered the blood supply sometime in the 70s, we think. The first cases of aids which were not known as AIDS at the time, occurred in 1980s. By the time the first real transfusion transmitted a case of what we thought was aids was reported, the risk in San Francisco was already about 1% of being infected from a blood transfusion. So there was a long, silent period. Once we recognized that and put in appropriate screening tests and screening questions, the risk was reduced drastically but I would propose to you that if a new agent entered the blood supply today, with a long silent period, before the first clinical case was described we would be in no better position today than we were back in the 1980s or at least in not much better position than we were then.
So the strategy of developing a test after you survey and find a disease is a reactive strategy. And that is the strategy that we've used in the United States since 1938 when we started testing blood for syphilis, there was a long period before we developed another test to test the hepatitis. Since then a number of tests we put on to screening blood has risen drastically. Yet there are a lot of agents we don't test for yet because there isn't a test or because tests aren't sufficiently sensitive to screening.
And this again is a slide that Dr. Alter put together that we have published in blood in a review that we hoped would increase the impact factor of blood. [LAUGHTER] showing the interval between the time a disease is discovered and the time that you actually have a screening test and you can see the hepatitis b was 30 years, for hepatitis c, it was 15 years. Live, 3 years or more -- HIV. 3 years or more. West Nile, probably about 4 years from the time it entered the United States. For the latest test for Chagan's disease, 4 years from the time that a screening test was recommended until we started screening blood. During that period of time, people who are infected with these agents. If we were to screen out all contaminated units we wouldn't have a blood supply. If we did risk assessments on all the agents that might infect blood, I think this is unachievable.
So I hesitated to show this based on the Olympic outcome. Wayne Gretzky, when asked how you score so many goals, he said you have to skate to where the puck is going not to where the puck has been, which seems obvious. That's what the pharmaceutical did with blood plasma after the aids epidemic. Factor 8 affected so many hemopheliacs, even there were many factor concentrates available. They went to a pathogen reduction strategy. So since 1987 there hasn't been a single transportation mission of HIV, HPV, or HCV by a placema traction of any kind. Factor 8, factor 9. No West Nile virus was transmitted ever by a plasma fraction.
The problem, of course is how to selectively damage or remove a wide range of intra and extra cellular pathogens without affecting the blood components. The solution, of course, is to identify a target that is present in the pathogens but not present in blood components. And one of these might be DNA since most pathogens that we know of are nucleated in the blood components, red cells, platelets, and individual units of plasma do not have nucleated cells.
We learned a number of lessons from the pharmaceutical industry in plasma activation. We found out the components can be well maintained if you add inactivating agents. They didn't encounter any toxicity in 20 years of enacting pathogens. Immunogentcity was not encountered. Viral safety was achieved. You do need methods that inactivate logs of different viruses. You need methods that inactivate more if you're not testing for the individual virus as well.
But for individual components, there are other considerations. You have more virus in a single unit of blood, perhaps, than you do in a pool of 30,000 units where an infected unit might be really diluted. There are more proteins to consider in units of fresh frozen plasma than a factor 8 concentrate. You have a limited ability to purify individual opponents than fractionated units of plasma. Cells are more fragile and many proteins and inactivating in bags is not like inactivating in tanks which is used for plasma fractions.
The ideal technology would be non toxic, nonmutagenic. It would cover a broad spectrum of acts have minimal impact on the blood components, simple, removable, and compatible with good manufacturing practices. There isn't anything that really meets all of those today.
There are currents methods being used today in Europe and elsewhere for plasma, platelets, red cells, whole blood. The first of these is solvent detergent added to plasma to inactivate viruses that are encapsuled. This is the first method ever been used performed on small pools of plasma, less than 2500 units. You put the agent in, you incubate for 4 hours, at 30 degrees. You remove the solvent and detergent by oil and absoprtion chromatography and you aliquot the plasma and freeze it. Solvent and detergent disrupt the lipid envelopes. It can be applied only to acellular components such as plasma because it disrupts cells as well.
Now, I put this slide up to show that it's important that you have reserve capacity because our tests for measuring the log kill are not particularly sensitive so you really don't want a method that inactivates viruses completely at the end of the incubation period. You'd like them to be inactivated as best we can measure very early so if there is residual virus that we can't measure, that will be inactivated as well. Solvent detergent is licensed in Europe. It's not available in the U.S. primarily because inactivated plasma decreases protein S and alpha 2 antiplasmin. And there were some coagulation problems in liver transplant patients and a black box warning was placed on this agent in the United States. However, it's widely available in Europe.
More than 8 million units have been transfused to more than 3 million patients over 18 years. There have been no reported adverse events. It also likely eliminates the risk of transfusion related acute lung injury in plasma, an interesting side effect. There is another method available in Europe, methylene blue, a die you're very familiar with is added to blood and blood is exposed to visible light. This is the methylene blue, a tricyclic aeromatic photo sensitizer. Methylene blue intercalates into the acid of the pathogens, cross links DNA and single strands. It has reactive oxygen species when exposed to light and oxidizes guanosine. It uses visible light. This is the system used in Europe, where methylene blue is added to plasma. It's filtered to remove cells exposed to life. The methylene blue is removed by a depletion filter and the plasma is stored frozen.
Here is the system in use in many parts of Europe. In fact, it's the only plasma currently being used in the United Kingdom for patients born after January 1, 1996. It's being used widely in France where it began to be marketed in 2007 as well as in Germany and had marketing authority in Switzerland and Austria. There have been concerns about the carcinogenicity of methylene blue. These are likely overstated. It's been used for areas other than blood transfusion for 35 years without any evidence of this in human beings.
So for envelope viruses, there have been no transmissions of the big 3 agents. For non envelope viruses, these agents don't work particularly well. If you combine them with screening, you have virtually sterile plasma. So these plasma inactivating agents are here to stay.
Cell associated agents aren't a problem. We filter them out before we inactivate the plasma. They have been investigationed and have an excellent safety profile. Excellent viracidal capability. There is a good chance if a new virus emerges it will probably be lipid ensapulated as the hepatitis viruses and HIV was. Not entirely sure, but this could eliminate that risk of a new agent in plasma.
I'm going to move to platelets now. They're more of a problem than plasma because they're a cellular component. There are 2 methods used for inactivating pathogens in plasma. The first is psoralen, amotosalen, so-called S59. This is what it looks like. Psoralens are available in many foods. It intercalates between bases of acids you then treat with ultra violate light. There is some loss of number and activity of platelets, and this is a cartoon of how that works. This is the amotosalen molecule. It intercalates into the DNA, you then expose it to ultra violate light. This is a very effective way of inactivating pathogens. The ultra violate light is relatively low energy. The system is very similar to that of methylene blue. You add the psoralen, you incubate an illuminator, then remove psoralen, and platelets are ready to go. Platelets treated this way are generally spended in a combination of plasma and a suspension medium. This technology will inactivate 5-6 logs of the major viruses.
There have been phase 3 clinical trials, both in Europe and the United States. And in both of those trials hemostasis using platelets inactivated with psoralen were equivalent to those untreated . This is licensed and used in much of Europe, not yet licensed in the United States. However, in a Dutch study, a randomized control study, the only study that was not conducted under the auspices of the company that makes this inactivation technology, the trial was terminated because of excess bleeding in the treatment arm. That study has not been published. We don't know why that occurred. That is something to bear in mind. This system is also used for plasma. So it's conceivable we could inactivate 2 different blood components using this technology. Again, it's in clinical trials and being used in Europe, not licensed in the United States.
The second technology for platelets is riboflavin. Vitamin B2. An essential nutrient. Everyone is familiar with it. When added to blood intercalates into nucleic acids. You can expose them to a high energy light. It results in breaking nucleic acid. So if this is your virus, and you add riboflavin, again, it cross links and then when exposed to light, produces reactive oxygen species which break nucleic acid. This technology which is called miraosl by the manufacturer uses a higher energy of light and therefore may provide more damage to the blood components. Light in UVA and UVB range. The technology is almost identical to what I showed you previously with the psoralens. This has been used for plasma. It's CE marked, being used in Europe. Also used for platelets and used in Europe and it's thought to be successful for whole blood but no data really in clinical studies have been published. The real problem, is red cells. How do we inactivate pathogens in read cells without damaging the cells, seeing it's difficult to use technologies that require light activation in a red component.
There is one technology currently in trial, a tricyclic quinancrine alkylating agent. This is what it looks like. It has a linking agent to an effecter. This alkalates proteins and some membrane proteins. It too is removed by absorption after being added to blood. The agent is added to blood at a very low PH. It will cross link nucleic acid in pathogens. Going to physiologic PH, the molecule is released and can be removed. This has been in Phase 3 clinical trials. They were halted in Europe, and in the United States, after a antibodies to neo antigens were discovered, probably related to deposits on surface of the red cell. The technology has been manipulated and is back in phase 1 trials and we're hopeful that will no longer be a problem and we'll have a technology for inactivating pathogens in red cells.
Plasma technologies are here to stay. Cellular technologies focus on acid, they are available for platelets in Europe, not the united states. Red cell technologies are in development. Generally, the experience in clinical studies has been good. All of these will be expensive technologies. These slides were simple to show you that all of these technologies are quite effective at removing envelope pathogens, non envelope pathogens, gram positives, gram negatives, parasites, almost every agent that has a nucleus.
There are some cautions regarding these. Each one is different. Each one has a different spectrum of pathogen reduction. The activity of the log reduction is different for each of these technologies. They have different activities and different components. Different metabolites, different profiles of adverse reactions.
Why haven't they been licensed in the United States? Well, first of all, the blood in the United States is very safe. There is not a great reason to do this. There isn't a single method to inactivate all components. You'd have to use 2 or 3 different methods which is a problem.
Our strategy of surveillance and screening has been effective in the past. And we can't inactivate all agents so some of the small non enveloped agents like hepatitis a virus and some spores and viruses that very high titres are problems. None of these inactivate prions.
People are worried about the potential risk of adding chemicals to blood. And, of course, cost is a big issue.
Let me summarize by saying blood is extraordinarily safe in the United States. Pathogen reduction technologies could add an additional layer of safety especially for the window period infection. If we're going to get these licensed, they have to have a tolerable benefit risk, broad inactivation, minimal damage to cells, little potential toxicity as well as a fail safe manufacturing system. None of them will be perfect. There is no guarantee against the next evolving agent. But most of these would have inactivated agents that we know of in the past.
Finally, depending upon geography and blood donor characteristics, that might really alter the benefit risk calculus. While some of these may not be useful in the United States, they might be really game changing when used in the developing world when most blood isn't tested at all.
Thank you very much.
[APPLAUSE]
Questions for Dr. Klein?
QUESTION: [INAUDIBLE]
KLEIN: In general, to the best of my knowledge, the amount of radiation you need really inactivates platelets, if you're trying to use platelets. There is some damage to the red cells as well. So I think in addition to the sort of risk of having radiation as an inactivator in general, the damage to the cells has been the major reason.
QUESTION: Why do you think that inactivation of plasma with the SD reduces the risk of trolley?
KLEIN: Well, it's an interesting question. Some people have postulated that it's strictly because they're small pools so if you have a high titre antibody it's diluted out. I don't think that's the entire story. I think it may well have something to do with possibly cytokines or lipids. There has been at least one animal model suggesting that lipids in the plasma and in other blood components activate white cells and result in trolley. So nobody really knows but there hasn't been a single case either reported or in people that I've talked to around the world, nobody has ever seen a case of trolley with a millions of units of SD plasma that have been used.
I thank both speakers for great presentations.
[APPLAUSE]
ANNOUNCER: You've been listening to NIH Clinical Center Grand Rounds recorded March 3, 2010. Diverse topics were the order of the day as we brought you Dr. Cynthia Dunbar, head of the Molecular Hematopoiesis Section in the Hematology Branch of the National Heart, Lung and Blood Institute. Her topic, "How to Publish Without Perishing: Navigating the Biomedical Publication Process." She was followed by Dr. Harvey Klein, chief of the Department of Transfusion Medicine at the NIH Clinical Center, who discussed "Pathogen Inactivation of Blood Components." You can see a closed-captioned videocast of this lecture by logging onto http://videocast.nih.gov -- click the "Past Events" link -- or by clicking the "View Videocast" link on the podcast homepage at www.cc.nih.gov/podcast. The NIH CLINICAL CENTER GRAND ROUNDS podcast is a presentation of the NIH Clinical Center, Office of Communications, Patient Recruitment and Public Liaison. For more information about clinical research going on every day at the NIH Clinical Center, log on to http://clinicalcenter.nih.gov. From America's Clinical Research Hospital, this has been NIH CLINICAL CENTER GRAND ROUNDS. In Bethesda, Maryland, I'm Bill Schmalfeldt at the National Institutes of Health, an agency of the United States Department of Health and Human Services.
