NIH CLINICAL CENTER GRAND ROUNDS
Episode 2009-014
Time: 1:01:15
Recorded April 22, 2009
Dangers of Secret Science: Case Study of Hemoglobin-Based Blood Substitutes
Charles Natanson, MD, Senior Investigator and Chief, Anesthesia Section, Critical Care Medicine Department, CC
Leukocyte Antibodies and Transfusion Reactions
David Stroncek, MD, Chief, Cell Processing Section, 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. We have two speakers on today's presentation, recorded on April 22, 2009. Dr. Charles Natanson, senior investigator and chief of the Anesthesia Section, Critical Care Medicine Department, will address “Dangers of Secret Science: Case Study of Hemoglobin-Based Blood Substitutes.” He'll be followed by Dr. David Stroncek, chief of the Cell Processing Section, Department of Transfusion Medicine, who will speak on “Leukocyte Antibodies and Transfusion Reactions.”
We take you to the Lipsett Ampitheater at the NIH Clinical Center in Bethesda, Maryland, where Dr. John I. Gallin, director of the Clinical Center, will introduce today's first speaker.
GALLIN: We have 2 clinical center investigators who will be speaking today and it's really my pleasure to introduce them today.
So our first talk is going to be from Chuck Natanson who will present Dangers of Secret Science, Case Study of Hemoglobin Based Blood Substitutes, Leukocyte Antibodies And Transfusion Reactions. He is a senior investigator and chief of the Anesthesia Section with the Critical Care Medicine Department. I meet Chuck in different places and most recently I met him with the faculty in China a few weeks ago where he works with us on the clinical practices of research here and internationally he does an excellent job.
Dr. Natanson received his BA at the University f Wisconsin and his M.D. at Columbia University. Following his internship at the university, he trained at the University of California of San Francisco. He's a member of a number of societies reflecting his broad academic interests including The American Society for Anesthesiologists, The American Thoracic Society and the American Society for Clinical Investigation.
His research is focused on the pathophysiology in new and therapeutic approaches to septic shock. And most recently his research inclued safe guarding patients in clinical trials. It's been fun to work with Chuck over these years and it's a pleasure to turn the podium over to him.
NATANSON: Thank you, John. And good morning.
Today I’m going to present a meta analysis we did of hemoglobin based blood substitutes. Our concern in doing this meta analysis was the risk of these agents. But there's really a hidden agenda here. What I really am going to discuss, is as the title says, is secret science. And the point of the the talk is going to be that once somebody has gone through the risk of randomization and clinical trial, it's critically important that that data be promptly available to minimize patient risks and to not deter the advancement of science.
But first let me begin with 2 disclosures. I was paid on a 1 time basis by Hemalink to review a cardiac surgery trial in sort of a nondisclosure disclosure, I am not listed on any patent and have no interest with hemoglobin blood based substitutes. This is a fusionable oxygen carrier that's not associated with red blood cells. Most of them have been hemoglobin based but there are fluorocarbons and the other types. But the ones I’m going to discuss today are the most prominently brought to market or attempted to brought to market, hemoglobin based substitutes.
Now the sources for the cell hemoglobin is that red blood cells are taken and they're listed here and they're left here with these tetramers, 2 alphas and and 2 betas. The sources for the cell for hemoglobin that are commercially developed for expired human blood, slaughtered cows, transgenic animals, and more recently recombinant forms of hemoglobin being made from plants, bacteria and fungi.
Our first question is, can we make a cell safe, cell free hemoglobin based cell substitute? The benefits are large. Self hemoglobin requires no matching, no refrigeration. There's no infection risk, a long shelf life and this would be ideal for trauma victims in rural and military settings. The problem has been risks. The safety has not been established.
The reported toxicities in hemoglobin based substitutes are hypertension, myocardial infarction, pancreatitis, renal failure, neurotoxicity and death. Now this is a slide done by Alan Shekter, showing the potential mechanisms of toxicity and I’m showing you an arterial blood vessel and the reaction between nitric oxide and the cell free hemoglobin.
And on the left here is the normal physiology and on the right is the pathology.
In the formal physiology, the hemoglobin lies within the erythrocyte and here you have cell free hemoglobin.
And if I begin with normal physiology, I’m showing here there's a large amount of nitric oxide released here in blue physiology and the hemoglobin molecules are within the red blood cells and we have smooth muscle cells and endothelial cells. And within those is nitricoxide synthase and argenine is used as a substrate to make nitric oxide insitrolyne and nitric oxide diffuses into the smooth muscle and activates cyclace and keeps the arterial vessel dilated.
Now, when you come to the pathologic system, now we have cell free hemoglobin which can be an oxygen carrier during hemolysis. And it's during this state this can quickly react with nitric oxide and use up the nitric oxide and form the hemogate and with this used up, this will cause constriction. Now, not only does it cause basal constriction from this reaction but cell free hemoglobin binds to nitric oxide much more strongly than oxygen. Binding of nitric oxide leads to basal constriction, platelet aggregation and attachment and release procoagatives and release of inflammatory cytokines.
These are ideal conditions for vascular thrombosis of the heart and other organs.
Now the different sponsors of these trials are manufactures of cell free hemoglobin have tried to alter or have altered the cell free hemoglobin in an attempt to minimize these toxicities.
Cell free hemoglobin is cross linked and that is basically to prevent renal failure for the molecule to not break apart, so the alpha alpha chain is bound. Then in addition, some manufactures have polymerized cell free hemoglobin.
Idea behind polymerization is to make a larger molecule that will stay away from the endothelial cell and the potential binding nitric oxide. In addition some manufactures have regulated cell free hemoglobin with the same idea, make a larger molecule so it stays away from the endothelial surface and therefore won't bind nitric oxide.
And finally, the amino acid sequence is being altered by certain manufactures with the idea of venting and combining nitric oxide.
The concern we had when we began this meta analysis was that randomized control trials as early as the mid1990s raised questions about the safety of these hemoglobin based substitutes and none of them demonstrated a meaningful beneficial effect.
And 1 cell free hemoglobin based substitute was approved for use outside the united states and there were recent clinicals trials ongoing in UK., Poland, Soviet Union, Sweden and South Africa and the Czech Republic.
So the meta analysis was to review the the association between hemoglobin based substitutes, myocardial infarction and death in trials and clinical settings. In these methods we used 3 different searches, and we included trials with death or myocardial infarction as outcomes analyzed. We found 70 articles, 41 excluded, because they did not include a blood substitute trial, 16 more were excluded because they were either healthy volunteers or duplicates which left us with 13 trials in the literature. We found 4 trials in press releases, 2 were excluded because there was no quantitative data which left 2 eligible trials and we found 1 trial which was a composite of a number of trials from 1 manufacture from an FDA meeting which gave us a total of 16 randomized control trials included in the meta analysis. We used mortality in myocardial infarction as our outcomes because based on our initial review of the data, these were commonly reported.
12 of the 16 trials reported death and 10 myocardial infarctions. So this is the data from the meta analysis.
So what I have plotted here or shown here is mortality and myocardial infarction, I’m showing the relative risk of myocardial infarction and the relative risk of mortality.
Data on this side favors controls in that you were 10 times, a hundred times, a thousand times more likely to die in the control group and here you were more likely to die if you were I’m sorry.
Here you're more likely to die if you're in the control group with hemoglobin blood based substitutes and here you're more likely to die with the use these cell induced substitutes, here is the mortality and here were the 5 products being produced and Hemopure, Polyheme, Hematour, and Hemaspan.
I’m going to show you the products 1 at a time and I’m going to show them until I complete the meta analysis so let me begin with Hemacyst, there were 7 trials which reported mortality data and 5 trials which reported data on myocardial infarction.
Hemalink there were 2 trials that I’m showing the Hemalink data in red and I’ve turned the previously shown Hemacyst data to white and there were 2 trials which reported mortality data of Hemalink and those same 2 trials reported on myocardial infarction data.
Hemapure, the summary trial from the FDA reported mortality and myocardial infarction data and polyheme, 2 trials reported 2 in mortality in data and Hemaspan which had 1 clinical trial reporting mortality and myocardial infarction data.
When you put this together and analyze it together, what we found is that overall the relative risk of death is increased by 30% and this is statistically significant at the 016 level and there's almost a 3 fold increase at the risk of myocardial infarction and this is highly significant.
Well, of course you do an analysis to see if these can be combined in I squared and as you see here the I squared and 0. I squared that are very high, means that the treatment effect it is differ between the clinical trials that it's a given an I squared 60 or 70%, an I square is the lowest 0 and is saying that the treatments over all are not statistically different and the this give you the same information.
These trials had the similar effects and at least on a statistical bases are okay to combine and what you find here is overall there was a significant increase in mortality in myocardial infarction.
Now, you can't look at only risk, you have to look at potential benefit because always you should look at a therapy in view of risk verses benefit.
And what I’m showing you here is 7 of the hemoglobin based blood substitute trials investigated if the hemoglobin blood based substitute decreased transfusion requirements.
You can see here, the 7 trials, 1 did show a significant decrease. Two trials of hemoglobin based substitute showed non-significant decreases, 2 trials, the decreases that were significant early were completely offset by increased blood requirements later.
And in trials it's unknown because they were stopped early because of safety reasons.
Some of these same hemoglobin based trials investigated neurologic outcome from stroke or preventing organ failure and death in different clinical settings, but no trial reported a statistically significant meaningful long term beneficial outcome.
Now the next thing you would like to see is subgroup analysis, looking at the different types of trials and seeing if the effect is consistent. And what I’m showing you here now is dividing the data by the type of patient that was in the clinical trial, whether it was the trauma elective surgery, cardiac surgery or stroke patients.
You see here there were 5 trials more than a thousand patients that were done in trauma victims and you can see here it's very consistent, there's an adverse effect on mortality and an adverse effect on myocardial infarction. You look at the elective surgery there were about 2000 trials, 7 patients very consistent that there's an adverse effect in mortality and myocardial infarction and you look at elective surgery with 560 patients and here the relative risk is on the other side, but if you look at the myocardial infarction dates, it's again very consistent. There's increased risks and in terms of stroke, the 1 trial again, there was an increase in mortality.
So in terms of cross types of patients there seem to be a consistent effect that the risk was there regardless of what type of surgery was done or what type of patient was studied.
You also want to see, did it make a difference? What type of control was used in these various studies. And 9 of these studies had a non-blood product control, almost a thousand patients and you can see the result is very consistent with the result of an increase in mortality of myocardial infarction and if the blood product was used, there's this trend here where it may be near the no effects line but myocardial infarction seems to have the same effect size. And 3 of these studies reported both trials or patients received blood products and non-blood products and that was very consist went the overall effect.
So it didn't seem to matter what type of control was used in these clinical trials.
In the last analysis, what I’m going to do is remove 1 trial sequentially, 1 trial sequentially and show you, the effect doesn't change that no 1 single trial is responsible for this harmful effect.
And here I have removed the Hemacyst trial which leaves 7 trials in 2000 patients and the effect is very consistent. If I remove the Hemapure, it's consistent, Hemalink, very consistent, polyheme or the Hemaspan and if we did sensitivity analysis between published and unpublished data whether we looked at a high or low tetramer content or what kind of biophysical characteristic it was altered by, none of this data showed any difference. So we found a significant consistent effect on mortality and myocardial infarction and no beneficial effects.
So what we concluded is based on the analysis of the available data and clinical trials, hemoglobin based blood substitutes are associated with a significant increase of death to myocardial infarction.
Now there are some limitations. Because much of this data was not available in published form, we had to rely on press releases and committee presentations for some real results and 1 of these products was based on a full food and drug administration analysis and based as a single trial representing 39% of the data in the meta analysis.
And these were not standardized across the data we gave the report if the clinicians thought there was a myocardial infarction.
Well, now turns more to the point of the talk. I’m now going to show you a cumulative mortality and myocardial infarction analysis which shows you when the data became available, what the or when the clinical trials were completed, what the mortality was over time as it accumulated. So now I’m showing you the hemoglobin blood based substitute trials, the myocardial infarction, here I have the year the study was completed, the number of studies, cumulative number of patients, again, cumulative mortality and myocardial infarction data the all important I squared is when was the treatment effect consistent across these studies such that had a statistical basis that said they're not different and you can combine them.
And let me begin the first trial, 1996, the data became available and it showed an increased trend for an increased myocardial infarction in mortality. By 1998, the data were available from 3 trials and again you could see, there's now a significant effect on mortality and we're beginning to see myocardial infarction data. Looking at the cumulative data by 1999, we have 7 trials 700 patients and again the mortality looks pretty much the same and the myocardial infarction we have more data but the I squared are not very high, bullpen still more than you would want but still more than you would want. But by the year 2000 when the trials were done representing 2500 patients you can see a significant effect on mortality and the myocardial infarction and the I squared are a consistent effect.
So now have you cumulative data suggesting you to you by the year 2000 there is a significant effect on mortality and myocardial infarctions. In 2002, another trial was done representing another hundred patients and it didn't add any information. The effect on mortality and myocardial infarction is the same. You then look at 2004, now another 2 trials have been done since 2000 again, it's representing another 400 patients. There's no difference in the effect on mortality and myocardial infarction. By 2005, we now have 15 trials, again, 4 more trials have been done. The year 2000 again we added no new information. There is the same cumulative mortality and myocardial infarction data and the I squared are 0. By 2007, we have 16 trials, close to 4000 patients and again it's added no new information. The myocardial infarction and mortality data the I squares are 0 and the significant.
So the question you ask is why did the trial continue after the year 2000? By the year 2000 we established that there was no substantial or meaningful beneficial effect and a significant increase over all in these trials of myocardial infarction and death. If an analysis had been completed in the year 2000, it would have shown significant risks.
A moratorium of trials would have prevented subsequent deaths (…) however, in the year 2000 that data were not available to scientists or the public. Why did the trials continue? Why wasn't the data available to scientists or the public?
Data on over 2200 patients from the Hemapure trials were completed by the year 2000 but they only became publicly known in 2007 during an open food and drug administration advisory panel meeting, a public citizen activist group had to sue the food and drug association to get this meeting open and the company decided to make the data public.
A large polyheme trial was only made public when the manufacture issued a press release in December 2006 and this press release was after a wall street journal article that was very critical that had obtained some of the data from this trial and this was 6 years after the trial was completed. Seven Hemasis trials were completed by 1998, however these trials were published by 1993 and 2003 because of the time to get things to the editor and get things published. Data from at least 2 additional hemoglobin based blood substitute trials were terminated for safety reasons and they still haven't been published that is the quantitative data are not available. One hemoglobin based substitute was associate wide gastrointestinal side effects, hypertension and increase in peripheral resistance and another was reported to produce an increase in myocardial infarctions as had been described in the 2 previously published cardiac surgery trials. And it is possible that there are still more clinical trials that have not been made public.
So in terms of hidden science, I think that this meta analysis and cumulative meta analysis underscores the inefficiency and risks of failing to promptly report the results of clinical trials. Scientists are unable to build on success or failures of other researchers, testing similar products, and patients with report, can be exposed to risks unnecessarily.
You can ask the question, how does it occur? How can data be kept from scientists and the public?
Well sponsors of investigation drugs are required by law to report the results of clinical trials in a timely fashion to the FDA, even if they never publish the data. But the FDA, does not make the data public unless the product is approved or the product is discussed at an advisory meeting so it stays hidden. The FDA does have policies that treat the FDA policies treat safety and effectiveness data from investigational drugs as trade secrets. There's an FDA reauthorization bill that provides for a database registering ongoing trials as well as eventual posting of study results and adverse events.
But there's no requirement for this of unimproved products. So they stay hidden.
How can we prevent hidden science? Well the first thing we might do is amend the U.S. Food and Drug Cosmetic Act to make it clear that safety and effectiveness data from investigational drugs are shared and not trade secrets. And pass a bill to require that trials of unapproved products need to be registered and results promptly posted.
So in conclusion, in order to protect trial participants and to advance science, safety and efficacy data for investigational products should be shared and not treated as trade secrets. All trials approved and unapproved drugs should be registered so the fail tower complete or publish trial results can be assessed.
Thank you.
[ applause ]
QUESTION: First I would like to say that was really terrific and I wonder if in your analysis you looked at post year 2000, the number of excess deaths that could be attributed to these products?
NATANSON: I haven't done it, I have not done that analysis, but I haven't done that analysis. I’ll say the second half which I wasn't going to say.
I don't think because you understand it's a it's a small percentage, it's a 30% increase in mortality but a very small mortality rate, so I don't think the numbers would be impressive in terms of how many people listen 1 person is too many.
But I don't think it would be a large number.
QUESTION: So you commented on the new law of not only registering clinical trials phase 2 4 and reporting outcomes for products that licensed either for the indication that you studied them or for another indication and we've discussed this in this room last week at our town meeting and had Deborah Zahrn, from the Library of Medicine and talk about how to do that and the difficulties and we're sort of in a transition period where we haven't each as an institution done a very good job of complying with the current law because there's no rules that go with the law. But the current law does state that you have to report the outcomes as they occur on a clinical trial within the year after the completion of the data from the last patient. So what you're suggesting is that we expand that law to include all trials whether or not, it's for a licensed drug. Correct?
NATANSON: Yes.
QUESTION: You focused on the clinical trials. There's also the issue of the preclinical data that was used as a basis for progressing to clinical trials. And I think your case study brings up the question that there might be opportunities to assess the power of preclinical modeling with respect to translational efforts like this. Because my somewhat limited memory of the preclinical studies were that they weren't very strong in terms of particularly issues related to toxicity.
NATANSON: It would be relatively easy to argue that the preclinical data was showed toxicity. I mean, John has 10 years ago, closed down the Army operation because of the toxicities found in rabbits and none of the toxicity was born out then in the human clinical trial. So in this case I think what that meant is you have to be more careful and you had to be even more circumspect as (…) coming in because there wasn't a clear case that the animal was showing benefit and there were increased risks in the preclinical data.
GALLIN: It's now my privilege to introduce our next speaker, Dr. David Stroncek who's going to discuss leukocyte antibodies and transfusion reactions. Dr. Stroncek has been chief of the Cell Processing Section in the Clinical Centers of Transfusion Medicine since 2007. His section manufactures a variety of cellular and cell therapy products for Clinical Center patients and to enable many of the studies that are being done here.
Before his current position, he was chief of transfusion medicines laboratory services section for 11 years. Earlier in his career, he served as the first medical director for the National Marrow Donor Program.
Dr. Stroncek earned a bachelor's degree from the University of Minnesota and an M. D. Degree from the university's medical school. He completed an internship in residency and internal medicine, also at the University of Minnesota affiliated hospitals and completed a hemaatology oncology fellowship there as well. His research focused on the use of gene and micro-RNA expression analysis for the assessment of cellular therapies.
He has done pioneering work in the fields of unrelighted marrow donor transplantation and the mobilization and collection of peripheral stem cells for allogeneic transplantation.
He's also an expert on neutrophil antigens and transplants related to acute lung injury.
David, welcome.
STRONCEK: Thank you, John.
My lecture will focus more on the hazards of transfusion medicine.
Another title for this talk could be how technical aspects of transfusion medicine can impact the public and that's the real interest in this subject.
This story starts about 2 years ago when the American Association of Blood Banks released the guidance document suggesting that all blood centers and transfusion services prevent the transfusion of plasma containing blood components from donors that were leukocyte aloe immunized. This policy was supposed to take effect gradually and be completed by the fall, by the end of last year but they have loosened it so centers have more time.
What do we mean by leukocyte antibody? Well, leukocyte antibodies are 2 types of antigens, classic HLA, class 1 and HLA class 2 and then there's a lesser known group of leukocyte antigens, human neutrophil antigens and they're broken down into 3 well described systems human neutrophil antigen, 1, 2, 3 and 3 and there are couple others but these are the most common and clinically relevant systems.
In this lecture I’ll review the types of transfusion reactions caused by leukocyte antibodies, I’ll discuss the role of neutrophils and HLA antibodies and transfusion reactions and then suggest strategies to discuss mediated transfusion reactions.
I have no disclosures.
I’ll discuss the roles of thesis different types of antibodies in transfusion reaction. I have directed both a neutrophil antibody reference lab and an HLA neutrophil bias lab so I have no bias as to which bodies are important.
So the 3 types of transfusion reactions are febrile transfusion reactions, granulocyte transfusion reactions and pulmonary transfusion reactions or transfusion related to lung injury. Transfusion reactions are the ones we've all seen very commonly when a cellular blood product is transfused. These are the straight forward fever and chills types of reactions, patients often experience. They're really not trivial because some of these patients can be very uncomfortable and it's really not a good type of thing to happen when blood's transfused. These occur following the transfusion of a cellular blood product and they've been well known to be caught by the reaction of leukocytes in the blood and with leukocyte antibodies in the transfusion recipient.
It's also been well known for many years that just removing some of the leukocytes from the blood product prevents these reactions, simply removing a buffing will prevent these reactions but now we have good filters available and a number of institutions including our own now, remove all leukocytes from blood, remove leukocytes from all cellular blood products and this prevents this reaction very nicely.
The other type of reaction is similar but involves granulocyte transfusion but these are a bit more severe. And in these cases, they not only experience, fevers, chills but they develop dyspnea and frank hypoxia. And again these are due to leukocyte antibodies in the transfusion recipient and in case, reactions are more severe because granulocyte concentrates contain probably a hundred fold more leukocyte than a typical blood product and that probably contributes to trapping of leukocytes in the lung.
These reactions could be prevented by collecting granulocytes from donors that are compatible with the transfusion recipient but that's difficult to do because the number of people who want to donate granulocyte and limited so we handle these by stopping the granulocyte transfusion. Usually most patients only need granulocyte transfusion for a matter of a week or 2 so that hasn't been too problematic.
The interesting types of leukocyte mediated transfusion reactions are the pulmonary reaction and these again are fevers, chills, dyspnea, hypoxia, patients may have transient leukopenia, clinically these occur within 6 hours of the transfusion and what's interesting about these are they're caused by the transfusion of antibodies in the blood component donor, so it's the opposite situation, the other reactions. The most severe form of these reaction system known as TRALI and that's been the subject to fairly intense investigation recently.
There's been a very loose definition of TRALI until about 4 or 5 years ago when people became more interested in this disease and now, there's been some formal consensus group vs developed consensus definitions, simply TRALI is a New Orleans onset of lung injury associate wide a transfusion, this should be associate wide a transfusion or plasma containing blood product must occur within 6 hours of the transfusion. Specifics of the diagnosis, it has to be new since onset of lung injury, if they have signs or lung injury signs that are going on before the transfusion, then it's really imposition to attribute the acute lung injury to the transfusion. Clinically x ray findings should show bilateral pulmonary edema. The lung injury that's bilateral infiltrate shouldn't be due to fluid overload, so the pulmonary artery occlusion should be less than 18. And also there has to be fairly significant hypoxia, so the o2 saturation should be less than 90 or the pulmonary artery to o2 ratio of less than 300. Most importantly this is a clinical diagnosis, so, it doesn't have to be dependent on presence or absence or other factor in the blood product.
So why have we become more worried about TRALI recently, and it's because the high incidence of mortality associated with TRALI.
This shows a summary of the transfusion related fatalities reported to the FDA between 2004 and 2006. I would be showing you this data from 10 years ago, hepatitis and HIV would be leading this list, but now, because of the better testing of blood components, TRALI is the leading cause of transfusion related mortality and because of that, the FDA and the transfusion committee have transfusion community have put a lot of effort into trying to reduce the incidence of transfusion related acute lung injury.
A couple things about the incidence and the products involved. The reported incidence of TRALI is really highly variable, anywhere from 1 to 10,000 units have been reported to cause TRALI. This is probably somewhat dependent on that these type of reactions vary from different patient populations and also they may be more apparent in an intensive care setting where patient transfusions are monitored more closely because of general intensive care units have reported rates at close to 1 to 1000 transfusions causing TRALI, but it's been more like 1 to 10,000 in the general population.
Of course all blood products, fresh frozen plasma and red platelets have been implicated but those that have more plasma and FTT and platelet vs been more frequently implicated. Luckily, the symptoms associate wide TRALI resolve on their own within 24 48 hours. The care and treatment for TRALI is really just supportive treatment. On some cases the symptoms may resolve in 6 to 8 hours before it's clear, the diagnosis is completely nailed down. Unfortunately the mortality rate is, some people do die from TRALI.
Most today’s report from the mortality rate of about 10%, but a few studies have reported the rate as high as 50%.
So what's the roll that leukocyte antibodies cause these transfusion reactions or TRALI.
This goes back to 1957 and in study, Britain the goal was to see what would happen when leucoglutens will transfused into healthy subjects. He obtained 250 mls of plasma from 3 highly immunized individuals and transfused that plasma into 3 different individuals.
Two only had mild reactions but 50 mills into the transfusion of the third product, the blood was associated with transfusion with vomiting, diarrhea, chills, fever, hypotension, dyspnea, cyanosis, and leukopenia so this looked like related to acute lung injury.
Interestingly, they didn't get a lung x ray that day, but 2 days later they showed bilateral pulmonary infiltrates and 2 days later the chest x ray was normal so this is really gave us the idea that leucogluten can cause pulmonary type transfusion reaction.
At that time, HLA antigens and antibodies weren't well described or were neutrophil antigens so any type of antibody, leukocyte antibody was just called the leucogluten.
What is some of the other data that antibodies that activate granulocytes can cause pulmonary problems like I think 1 of the best clinical examples goes back to dialyze or nutropenia and in the 70s, from Minnesota, they were studying a number of people had noted that in hemodialysis then, when cellophane membranes were used, the patients developed acute nutripenia within 30 minutes of the start of dialysis, even though dialysis went on for 3 to 4 hours they quickly recovered but it was associate wide a significant drop in 02 saturation and increased arterial gradient and increase the fusing capacity.
They went on to show that these cellophane membranes used in dialysis did activate complement and activated complement did cause neutrophils to clump or aggregate and they hypothesize that these neutrophils are being trapped in the lungs and causing pulmonary damage and hypoxia. They tested in this a rabbit model. They incubated plasma and cellophane which activated complement and they infuse that to rabbits and they again they saw a fall in neutrophil counts and in oxygen levels and they looked at the lungs and they found leucoglutens in the vessels and perivascular edema so this again provides evidence that activated granule sites can get trapped in the lungs and cause pulmonary damage.
There's been animal models using that more closely mimic the transfusion situation that have demonstrated this too, there's 2 well, well published models, both use neutrophils and neutrophil antibodies and profuse them into isolated animal lungs, 1 involve the use of rabbit lung and antibodies to 3a antigen and complement. The other involved rat lungs and antibodies to any to antigen 2a and didn't require complement, but both models did show that the weight of the lungs increased with the neutrophils and antibodies were both transfused but neither 1 alone, that the increased vascular pressure occurred and capillary leukostasis occurred so this study again provided evidence that neutrophil antibodies can cause pulmonary injury. Really, the kind of the what do we know for clinical data for the leukocyte antibodies, causing TRALI, in 1983 a group of the mayo clinic reported 5 cases of pulmonary injury associate wide transfusion and they called this coin, they coined it transfusion related acute lung injury, in these 5 cases there were 19 implicated donors and they tested all these donors for HLA antibodies and found at least 1 antibody each of the 1 donor with the HLA antibody and each of the 5 cases so their hypothesis was that HLA antibodies were important cause of transfusion related to acute lung injury.
They also published a larger series in 1985 and the idea took hold that HLA antibodies can cause TRALI.
Now this was a reasonable assumption in the 1980s but really this is a series of case reports so they don't hold up to current standards for clinical investigation. And the problem with these studies is that, when you're looking at TRALI, particularly, years ago when it gave platelet components for multiple donors instead of aphoresis platelet, most patients don't get just 1 blood product, they get transfuse wide 2 products such as platelet, cells, and FPP product and so if you go back to look at leukocyte implicate indeed TRALI, you have to look at the antibodies in the general population and it's well known in blood donors, 4 to 7% of blood donors have an HLA antibody approximate if you look at women, particularly women who had 3 or 4 pregnancies, up to 3% of them will have hrl,a antibodies, so if you transfuse 3, 4, 5, 6 of these products 1 of them will have HLA antibodies. So it's impossible to study these by themselves. You need a case control group.
It's not the same with neutrophil antibodies, neutrophil are very common and at most will .4% of female donors have neutrophil antibodies and most studies show far less than that.
So it was a long time though before a prospective control study was done to look at the roll of these various factors and transfusion related acute lung injury, this was done in an intensive care set being at the mayo clinic where they monitor their patients closely and they look at 960 intensive care unit patients receive a blood transfusion, they identified 368 of these who had worsening respiratory status within 24 hours of the transfusion and of those some had fluid overload and some had the acute lung injuries after 6 hours and some had other reasons for that change in respiratory status and they identified 74 people that met the criteria for TRALI or acute lung injury which would be TRALI and then they looked at the patients that got transfusions and didn't have any worsening of the pulmonary status and they identified 74 match controls and then they compared what blood products each group got and what they found is that, the risk factor, 1 of the risk factors for developing TRALI was the volume of plasma patients got that had were from female donors so they look to see, they made an estimate based on the number of units of fresh frozen plasma and people were transfused with the number of platelets and they estimated the volume of plasma was that infused. They knew the gender of the donors and there was a 5 fold risk of getting TRALI, more female donor, plasma was transfused.
The transfusion of neutrophil blood from a donor of the neutrophil antibody was also a significant risk factor as was the presence of an HLA antibody and HLA class 2 antigen in the donor.
Interestingly, I’m going into this but other people have reported that during storage blood releases leased factors that could activate granulocytes 1 was lineup o tc18 and the presence of that in the blood product was also associated with an increased risk for TRALI but what they didn't find was that HLA class 1 antibody, no increased risk of TRALI when a donor had an antibody to HLA class 1 antigen.
Again, study, doesn't mean HLA antibodies aren't problematic but it does suggest that antibodies to neutrophils and class 2 antigens are much more of a problem: another issue you would expect if leukocyte antibodies do cause transfusion reactions is if a person donates repeatedly if you go back and look at previous donations that those are transfusion reactions.
We're lucky at the Clinical Center, we collect all the we transfuse. We had 1 donor that donated every month for over 2 years, she donated 27 aphoresis products and we found 1 of those was associate wide the transfusion reaction, we went and tested her with neutrophil antibodies she had an antibody with the nn2a and we went back and looked at all 39 of her blood products that were transfused. 12 of them caused reaction, there was a wide variety of reactions some of them fairly severe and none of them meeting the criteria for TRALI.
The point is these neutrophil antigens seem to cause reactions in a high frequency of transfusion recipients.
There's been other studies like this so called look back studies that have evolved donors with antibodies to neutrophil antigens and HLA class 1 antigens and typically it's 1 donor in each study, and multiple transfusions and to summarize this quickly, if you look at all of the (…) studies involving antibodies to neutrophil antigens there were 111 transfusions and 28 or 25% caused reactions. Or if you look at the reports of look back studies involving antibodies to HLA class 1 and class 2 antigens there was 127 transfusions reported and only 2 cause transfusion reactions or 2%. So again, it looks like neutrophil antibodies are much more problematic than HLA antibodies.
So, how are people going about preventing these reactions? And first of all, it's easy.
As I mentioned for fibro reactions we transfuse leukocyte reduced blood components.
For granulocyte transfusions we stop the reactions but avoiding aloe immunized donors is more problematic. For whole blood, for plasma for transfusion, it hasn't been problematic as we thought. But ends up most of our all of our plasma for transfusion comes from whole blood donors, so collecting whole blood, making red blood cells and in the past we would take it's typical in the U.S. for about half the plasma to go to transfusion and the other half to go off to the fractionate or to make AMIN or IGG, or clotting factors.
So what the blood center started doing was saving identifying male and female donors and using the male donors for transfusion and the female donor plasma for fractionation, it's been difficult meeting the needs but we've been able to meet the needs for transfusions and this worked quite well, and it's transparent to the donor so it's not with women who have the blood and we do use all the blood products, so this is as l. Problem is aphoresis platelets and the transfusion now come from aphoresis platelets and it's easy, if the donor's mail, we transfuse the products but the alternative for females aren't as good. We could decide not to collect platelets from women but then we wouldn't have enough platelets to meet the clinical needs.
The other issue is, it would be difficult to tell women we don't want to donate platelets.
We could not collect them from donors which some places are doing so we could test the female donors for leukocyte antibodies and exclude women with leukocyte antibodies.
That's a strategy many centers are taking.
What you think now is that the best way to do is to (…) and he will test for neutrophil antibodies.
As I mentioned few females have antibodies to neutrophils and these often cause reaction, unfortunately the test suggest difficult and kits that can test the important aspect gens aren't really available. In contrast HLA antibody nobody's really testing for neutrophil antibodies instead what centers have started to do is test female donors for HLA antibodies, unfortunately, many females do have antibodies to HLA antigens and these rarely cause reactions but the reason people are testing for them is that the testing is easy. We've been testing for HLA antibodies for years, it's important for organ transplant. There's a number of kids available so this test can be done quick and easy. This is problematic because we're probably if people take this route, they're probable we're losing excluding many women from donating needlessly which is problematic.
So in conclusion what is beginning to happen is that many donors are testing women for HLA antibodies and excluding those with HLA antibodies to class 1 antigen, this will result in exclusion of some donors unnecessarily and we do fear that this will alienate a number of females and prevent them and keep them from donating blood products even though they don't have HLA antibodies, at the clinical center, we do monitor our transfusion very closely. If some transfusion is associated with the reaction, we quickly we immediately test the donor for neutrophil antibodies and HLA antibodies and we do exclude donors so we're not prospectively testing any donors for leukocyte antibodies.
Hopefully the manufactures are working quickly to improve assays for tests for neutrophil antigens and hopefully good assays will be available in a year and we could possibly start testing our donors.
Another thing is that 2 or 33 years down the line is they're using platelets, so they're treat indeed a way to enactivate viruses and bacteria and those procedures involve washes platelets at the end and resusc pending them in solutions, so there are no plasma available and so then you get around the issue of transfusioning leukocyte antibodies but that concludes my presentation, thank you for your attention.
[ applause ]
GALLIN: we have time for a few questions. I’ll start by asking you if people think the pathophysiology of the pulmonary complication system due to the physical clumps of neutrophils that get lodged inside the capillaries or to the release of neutrophil products as a consequence of that event?
STRONCEK: Well, I think for sure it has to be the release of either enzymes or toxic oxygen species or other consequences of granulocytes because it doesn't appear that having neutrophils trapped in the lungs causes a problem. We do know that when you transfuse any granulocyte product, a number of those do end up in the lung and don't leave. So it's probably not just the trapping itself.
GALLIN: Okay, well, we're at the hour and we thank both of our speakers for wonderful presentations.
[ applause ]
(Music fades in, under VO)
ANNOUNCER: We were pleased to bring you two speakers on today's episode of NIH Clinical Center Grand Rounds, recorded on April 22, 2009. Dr. Charles Natanson, senior investigator and chief of the Anesthesia Section, Critical Care Medicine Department, addressed “Dangers of Secret Science: Case Study of Hemoglobin-Based Blood Substitutes.” He was followed by Dr. David Stroncek, chief of the Cell Processing Section, Department of Transfusion Medicine, who spoke on “Leukocyte Antibodies and Transfusion Reactions.” 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 for today's podcast at the Grand Rounds podcast page at www.cc.nih.gov/podcast/grandroundpodcasts.html. 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.
