Click for next page ( 82


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 81
IV Risk Tolerance

OCR for page 81

OCR for page 81
Beneficial Aspects of Surgical Transfusion Richard K. Spence Where do we find proof of benefit of surgical transfusion? That is a tough call. It is a routine procedure for donors to give blood and just as routine for physicians to obtain blood from a blood bank. Today's physician writes an order and the blood arrives thanks to the blood bankers. The practicing surgeon hardly gives it a thought. Allogeneic blood is available for both elective and emergency surgery, which truly is an advantage of the use of allogeneic blood, because there is no burden on the patient. Use of autologous blood is not always an option. Patients must have enough time before surgery to donate the blood needed for the surgery. This becomes quite an issue in discussions about autologous predonation by patients, for example, with coronary artery disease. There is increasing evidence that heart surgeons, cardiologists, and others are moving patients to the operating rooms faster than they have before. Thus, we are seeing less and less autologous predonation in patients with coronary artery disease. Although we surgeons take it for granted, where is the proof of benefit of transfusion? If we look through the literature for the prospective randomized controlled trial, we will not find it because we will not find an institutional review board ORB) anywhere that will approve a trial of transfusion versus no transfusion in surgical patients. Therefore, analysis has to be based on retrospective data and anecdotal information. Another way to look for benefit is to see what harm might come from lack of a transfusion. These are the two approaches I took in trying to demonstrate benefit. Neither is proof positive, but both provide considerable evidence of benefit. Before transfusion existed, preoperative preparation was prayer. Physicians wrote about the use of transfusion during the Middle Ages, but the use of transfusion really did not come into play until the technology emerged in the 1660s. The story of the history of transfusion follows on the heels of the description of the circulatory system by William Harvey. A friend of Harvey's, Sir Christopher Wren, described how to get access into the veins. He used a sharpened goose quill and an animal bladder and injected a variety 83

OCR for page 81
84 BLOOD AND BLOOD PRODUCTS: SAFETY AND RISK of things into the vessels. The learned societies in England, Prance, and throughout the continent heard about this and said why not give blood to patients? The first recorded transfusion was to Antoine Mauroy, a 34-year-old imbecile who had "escaped Dom his wife's control." He was transfused with several liters of calms blood over 6 days, until his death.55 Richard Lower did a transfusion at a similar time in England, but the French physicians Denis and Emmerez are the ones who get the credit for the first transfusion. Unfortunately, the issue of transfusion risk came into play with this first transfusion because Mauroy's widow sued for malpractice, blaming transfusion for her poor husband's death. As it turns out, she had poisoned the man with strychnine. This experience, plus experience in England, was enough to make the national academies decide that transfusion was not such a good thing because the risks were so great. Thus, transfusion fell out of favor because it really did not seem to have a place in curing disease. Bear in mind that 17th century physicians thought the benefit of transfusion at the time was to restore mental capacity, to cure evil humors, and so forth. Transfusion was not looked at in a scientific way, such as we look at it today, until the early 1800s and James Blundell. James Blundell was a physician-obstetrician in London who got tired of seeing his patients die of postpartum hemorrhage and wanted to do something about this. He was a very learned man who surely knew of the experiments conducted in the 1660s. He set about investigating this whole phenomenon in a scientific manner in the laboratory by instructing two postdoctoral students from Trinidad and Tobago to start transfusing animals. He began by giving lamb's blood to sheep, calves' blood to dogs, and so forth. He discovered quickly that you must stay within species in order to get some benefit Mom transfusion. With his London instrument makers he devised instruments to collect and administer blood with a syringe and a three-way stopcock. He used these devices in the clinical arena, salvaging blood from his bleeding patients and collecting blood from donors. There are some interesting treatises from this period concerning how to def~brinate blood to deal with clotting, including using birch twigs to stir the blood in order to impede clotting. Blundell eventually began transfusing blood to his patients, and had about a 60 percent improvement in survival. In Blundell's work there was a change in the scientific approach to the use of transfusion. For the first time, Blundell was transfusing patients to treat blood volume loss and red cell loss, not to correct mental illness. His devices presaged modem blood equipment, and he 55Diamond, ~ (1980). A history of blood transfusion. In Weintrobe, MM (ed.), Bloodt, Pure and! Eloquent. New York: Leonard & Co.

OCR for page 81
RISK TOLERANCE 85 should perhaps be considered the father of modern transfusion medicine for his use of blood to treat anemia as a result of blood loss. In 1883 Jennings published his summation of all of the works that had been written to that time concerning transfusion, which really was not very much. He looked at 243 cases of blood transfusion, and he again showed that two-thirds of these were successful in saving lives.56 Most of these were for postpartum hemorrhage and the bulk of this material was Blundell's. In 1883, nothing was known about ABO and Rh blood groups. We now know that because of the distribution of blood types in the European/American populations (45 percent type O. 40 percent type A, 11 percent type B. and 4 percent type AB) the chances of inducing a serious reaction by transfusing without regard to ABO group are only about 35 in 100. Blundell, Jennings, and other transfusers of their time were fortunate. If the proportions of blood types in Europeans was different, we probably would not have seen transfusion advance as far as it did. Bear in mind also that a risk-benefit analysis in this period of time could hardly have faulted Blundell for saving two-thirds more patients than he had before with the use of what was a highly risky procedure. He didn't know he was going to have problems with one-third of his patients, what we consider now a major risk ABO incompatibility but getting two-thirds of his patients to survive postpartum hemorrhage was recognized as an undeniable benefit of transfusion. Early in this century scientific approaches to surgery continued to increase, promoting surgical procedures of a magnitude greater than we had seen before. This was initiated in part by anesthesia, which allowed patients to undergo larger and more complex surgical procedures. The surgical world had also heard from Rudolph Virchow, who had noticed that the ink from tattoos on the arms of his patients stopped in the lymph nodes in the armpit. He deduced that the lymph nodes must have a barrier effect, a function that might apply to cancer cells as well. He suggested that a way to treat patients with some specific forms of cancer might be to take out both the cancer and the lymph nodes. This marked the beginning of radical extirpative procedures: radical mastectomy, radical colon resections, and so forth. These procedures could not be done without the use of transfusion. Transfusion medicine was still a relatively young science. In the early 1900s progress came mainly from the work of two pioneers: Alexis Carrell, who showed us how to do vascular anastomoses,\and George Crile, Sr., who Pennings, C ( 1 883). Transfusion: Its History, Indications, and Modes of Application. New York: Leonard & Co.

OCR for page 81
86 BLOOD AND BLOOD PRODUCTS: SAFETY AND RISK wrote Hemorrhage and Transfusion 57 in 1909 and again emphasized the necessity of replacing lost blood with transfused blood. Transfusion soon became popular, and itinerant transfi~sionists, surgeons, and physicians who knew how to do this would travel around the country. It caught the public's attention. Despite great risks in this type of procedure, the benefits were clear to people at this time: preventing death Mom hemorrhage. As Bernheim said in his 1917 book Blood Transfusion, Hemorrhage, and the Anemias,s~ "Hundreds of people have been saved Dom premature death Tom hemorrhage." The literature contains many statements such as this, without any scientific basis for support such as a prospective or randomized trial. This was the perception of transfusion in general. Bernheim also pointed out an additional, indirect benefit of transfusion: operations could be made less hazardous with transfusion. Not only were they made less hazardous, but many were also made possible for the first time. Before blood banks, transfusion required 2 patients and a surgical procedure. The ability to store blood temporarily in a "blood bank" led to farther advances. Being able to take the blood out of the patient and separate it temporarily and then infuse it made it possible to transfuse more patients more easily. Some of the surgical procedures that Bernheim foresaw as becoming possible are listed below. This list is by no means complete, but these are the kinds of things that would not be done today if it were not for transfusion. Even though for a percentage of patients some of these surgical procedures can be done without transfusion, that is only because of the skill and technology of people who have been working in this field for a number of years with blood close at hand. Transplantation. Renal, cardiac, and particularly liver transplants could not have been done without transfusion. - Burn excision. Blood loss is very high in burn excision, and the techniques we have learned now rely on and depend on transfusion. Radical cancer surgery. The extirpative kinds of procedures such as mastectomies, pancreatoduodenectomies with the Whipple procedure, rectal excisions, hepatic resections, and radical prostatectomies all require transfusion. Other surgery made possible by transfusion: cardiac surgery, valve replacement, coronary artery bypass grafts, and vascular procedures of all sorts. 57Crile, G Sr. (1909). Hemorrhage and Transfusion; An Experimental and Clinical Research. New York and London: D. Appleton and Co. s8Bernheim, BM (1917). Blood Transfusion, Hemorrhage, and the Anemias. Philadelphia and London: Lippincott.

OCR for page 81
RISK TOLERANCE 87 In the late 1960s and 1970s, it was routine to type and cross-match 25 to 26 units of blood in preparation for surgery. This is no longer done, but without the ability to transfuse large amounts of blood we would not have seen cardiac surgery advance to where it is today. Since the mid-1980s we have seen about a quarter of a million coronary artery bypass grafts done each year. The patients who receive blood in this group represent about 10 percent of all blood recipients. At Cooper Hospital- University Medical Center in Camden, New Jersey, about 10 to IS percent of overall red cell recipients in 1994 were coronary artery bypass recipients. This is a 550-bed university tertiary-care hospital where about 300 to 400 coronary bypass procedures are done per year, so there is a significant number of patients who are transfused in conjunction with cardiac surgery. The Sanguis study has looked at 43 hospitals in Europe with more than 7,000 patients since about 1990. Some reports are starting to come out from the information gathered. Of all the patients studied and the number of procedures they looked at, 87.7 percent of coronary bypass patients were transfused.S9 Thus, transfusion is still a major part of coronary bypass surgery. More than 80 percent of vascular surgery patients were transfused. That includes both autologous and allogeneic blood, but even considering only allogeneic or banked volunteer donor blood, the numbers are still in the 35 to 50 percent range. Although we are doing better in terms of using autologous blood, we still need allogeneic blood to make these kinds of procedures possible. The discussion to this point has been primarily about red cell transfusion, but therapy with platelets and plasma derivatives is an issue here as well. In 1991 it was reported that platelets are a major part of treatment of coagulopathies associated with cardiopulmonary bypass surgery, particularly in the pediatric age group. Platelets play a major role in coronary surgery as well. When we look at a listing of the primary users of platelets in hospitals we find that the coronary bypass patients are generally in the top one or two positions. Other areas of medicine that would not be possible today without transfusion include level I trauma centers. Orthopedic surgical cases also would be impossible without transfusion. The Sanguis study has shown that 80 to 81 percent of all total hip patients are transfused. Likewise, surgery on patients with coagulopathies would be impossible. Operations such as simple inguinal hernia repair on a patient with hemophilia would not be done without anti-hemophilic factor derived from plasma. Even without surgery, patients 59Group TSS (1994). Use of blood products for elective surgery in 43 European hospitals. Transfusion Medicine, 4.- 251-268.

OCR for page 81
88 BLOOD AND BLOOD PRODUCTS: SAFETY AND RISK with hemophilia and the inherent coagulopathies would all die at young ages because we would not have the plasma derivatives with which to treat them. Another way to look at this is to ask what kind of surgery would we do without transfusion? It comes down to surgery that does not invade a body cavity. We can generally do laparoscopic surgery without transfusion, laparoscopic cholecystectomies for example. But surprisingly, the Sanguis Study found that about 15 to 20 percent of those patients are transfused as well. Without the backup of transfusion we would not be doing surgery as we know it today. It cannot be quantified in the same way that risk can with a number like l in 250,000, but the benefit is still clear. In summary we can see the progression in surgical red cell transfusion: Treat shock. After early attempts to cure anY and all diseases with transfusion, Blundell, and later Jennings, shifted attention to treating hemorrhage. In the 1900s Crile talked about hemorrhage and transfusion in shock. Blood was used for a long time to restore volume. The experience with World War I, and on into World War II, when blood became more available, supports this. Promote well being. Once blood became more and more available and the risks were still fairly minimal, transfusion was done for a lot of reasons that perhaps are not physiologically appropriate, such as perking up patients and putting a little pink in their cheeks. Fortunately, we have gotten away from a lot of that. Allow surgery. First and foremost, transfusion's benefit for surgery has been to save lives, by allowing interventions that would otherwise not be possible. Correct coagulopathy. The primary reason we should use red cell transfusion now Is to improve oxygen delivery. All of the foregoing has been focused on life-saving interventions. Another way to look at benefit is, as mentioned at the outset, to look at what kind of harm we see or what kind of damage is done if patients are not transfused when appropriate. To find that information we must look at the studies of patients who refuse transfusions, primarily the Jehovah's Witnesses. There are only a few studies, but there are survivors with hemoglobin (Hgb) levels as low as 3 to 5 g/dl. Some hemodilution studies also provide us information about the harm thta can occur if we do not transfuse. Jeff Carson and I did some work in early 1988 with a series of 125 surgical Jehovah's Witness patients.60 This work showed that there were multiple comorbid 60Carson, JL, RK Spence, RM Poses, G Bonavita (1988). Severity of anemia and operative mortality and morbidity. Lancet, ii: 727-729.

OCR for page 81
RISK TOLERANCE 89 factors that increased mortality in the Jehovah's Witness as hemoglobin level declined (e.g., cardiopulmonary disease), but mortality clearly was inversely related to preoperative hemoglobin level. We also looked at a subset of these patients who underwent elective surgical procedures. The difference in mortality associated with low preoperative hemoglobin is small but significant: 5 percent for patients with preop Hgb between 6 and 10 g/dl versus 3.2 percent for those with pre-operative Hgb > 10 g/dl. A difference of only 1.8 percent may not seem like a lot, but if one considers the 2 to 3 percent mortality now viewed as acceptable in a quarter million coronary bypass patients, then this small difference represents both a large number of patients and a doubling of the operative risk. Jeffrey Carson and I have now studied 528 patients: 206 Jehovah's Witnesses, and 322 who are not, looking at mortality and hemoglobin levels. Although analysis of this information is just beginning, it appears that in the 19 percent of these patients who were transfused. we may be able to demonstrate for the first time that transfusion is a positive benefit in terms of increasing survival or preventing mortality. In brief, we begin to see a significant difference in mortality at a Hgb level of 7 g/dl and below. That is' patients who can be transfused non-Jehovah's Witnesses who got down to 7 g/dl and were transfused had lower mortalities than the comparable Jehovah's Witness patients who were not transfused. We have also learned a lot from the laboratory and from the intensive care units about specific patient needs and some things that can happen if we do not transfi~se. We have a much better idea about how the body adapts to anemia and the multiple factors that influence this: severity of anemia, competency of the cardiovascular and respiratory systems, oxygen requirements of the individual, and duration of the anemia. The competency of the cardiovascular system seems to be the main determinant of how patients respond and compensate for surgical anemia. Patients maintain oxygen delivery when they are anemic by increasing cardiac output, simplistically, by increasing either heart rate or stroke volume. The first requirement for this is a healthy heart. The heart provides these compensatory changes by dilating coronary arteries. This response clearly is limited in the presence of coronary artery disease. Two recent studies have shown an increased incidence of cardiac abnormalities in vascular surgery patients with postoperative Hgb levels below 6] Spence, RK, JA Carson, R Poses, S McCoy, M Pello, J Alexander, J Popovich, E Norcross, RC Camishion (1990). Elective surgery without transfusion: influence of preoperative hemoglobin level and blood loss on mortality. American Journal of Surgery, 15963): 32~324.

OCR for page 81
90 BLOOD AND BLOOD PRODUCTS: SAFER AND RISK 10 g/dl.62 These two studies sounded the warning bell about undertransfiasion. If we look for more of this evidence, we would find that morbidity and mortality does increase in patients whose Hgb level is below 10 g/dl, particularly if they have coronary or pulmonary disease. Jefferey Carson has been looking at a database of 1,400 or 1,500 nontransfi~sed Jehovah's Witness patients. He has developed some preliminary information that seems to show that if we look at preoperative hemoglobin level, going Mom 14 g/dl down to 2 g/dl and look separately at patients with and without cardiopulmonary disease, mortality clearly increases in the presence of cardiopulmonary disease when preoperative hemoglobin is less than 10 g/dl. One can assume from this that transfusion should have a beneficial effect in these patients. Gerson Greenburg has done some seminal work in this area and is beginning to move us toward what you might term a physiologic transfusion trigger of 12 ml of oxygen per minute per kg.63 He has shown in a number of settings that this seems to be the lower limit for the heart. Below this level the heart muscle starts to deteriorate or convert to anaerobic metabolism. Thus, we are beginning to see the possibility of a more exact determination of when to employ transfusion. There are clear limits to cardiac response in anemia, but we may be able to override these with transfusion. Let me digress for just a minute to talk about trauma. We have known for some time with a number of animal studies that there is a finite limit to the amount of volume that can be replaced with asanguinous fluids. We can quantify that to some extent in trauma. It is now well known that one of the things that we need to do in trauma is to move quickly; the golden hour concept that Donald Trunkey has popularized has led to the establishment of level I trauma centers around the country. The trauma center in Camden is a fairly busy one, with about 1,800 patients a year. That means admitting a seriously injured patient every 5 hours throughout the year. In 1994, 26 percent of all red cells and components went to trauma patients; thus, the trauma patients are prime users of blood in the hospital. It is that transfused blood which gets these patients into the hospital, into the operating room, and treated within that golden hour. Hemorrhage may be classified into four levels by amount of blood lost. When we reach class three hemorrhage, the standard for fluid replacement is crystalloid and blood. 62Nelson, AH, LA Fleisher, SH Rosenbaum (1993). Relationship between postoperative anemia and cardiac morbidity in high-risk vascular patients in the intensive care unit. Critical Care Medicine, 21(6):860-866. Christopherson, R. S Frank, E Norris, et al. (1991). Low postoperative hematocrit is associated with cardiac ischemia in high-risk patients (abstract). Anesthesiology, 75 (3A): A100. 63Greenburg, A (1988). Indications for transfusion. Scientific American Medicine, 3: 1-16.

OCR for page 81
RISK TOLERANCE 91 If blood is not replaced at this point we start to see anaerobic metabolism, tissue loss, and patient death. The benefit is pretty clear in trauma, as transfusions allow us to move patients forward faster and also allow us to do the surgery needed to save lives. One last area of medicine where transfusion is prominent is the anemia thatresults from chronic diseases, chronic infections, inflammations, malignancies, and so forth. It is facile to say that one should just correct the underlying disease that causes the anemia, but in fact this is not always that easy to do. Transfusion thus becomes an important part of treatment in these patients when they require surgery. In conclusion, let me summarize the evidence for the beneficial effects of transfusion. There is some evidence that there would be limitations to surgery without transfusion. Many of the things we consider commonplace in surgery now would not be done without transfusion. There is some evidence in studies of Jehovah's Witnesses associating increased morbidity and mortality with lower hemoglobin levels, and we are beginning to get evidence that transfusion can turn that around. There is evidence for red cell-related limits of cardiac compensation, and we are beginning to acquire more solid physiologic findings about those limits that will help us to determine in the future just what we can do and what the transfusion intervention will provide. Finally, there are known tolerable limits to red cell loss from hemorrhage. I would close by urging you to remember, as you contemplate the risks and benefits of transfusion, that all solutions create new problems; good solutions are those that introduce problems more manageable than those they replace.

OCR for page 81
98 BLOOD AND BLOOD PRODUCTS: SAFETY AND RISK DISCUSSION Harold Sox: Why did it take so long for hepatitis C to become recognized as something to be avoided rather than something to be put up with? Elaine Eyster: Those of us who were caring for persons with hemophilia recognized it. For a majority of them transaminase levels were measured from time to time. Sometimes they would be abnormal. Sometimes they would be normal. Some persisted in having abnormal transaminase levels, but we noticed this year after year, and the patients were well. They had no clinical illness that anyone could detect, and the hemophilia-related bleeding problems were so overwhelming that it didn't seem appropriate to focus an undue amount of attention on these laboratory abnormalities. At the same time research was going on, but we really couldn't do much about the problem until we had a diagnostic test to identify the agent that was causing the elevated transaminase levels. By 1990, a huge amount had been learned about this, and we are just now beginning to learn about the natural history of hepatitis C. If you talk to many persons with hemophilia who received clotting factor concentrates during that era, you will find that it would have been intolerable to them to think about not having these products to provide them with the kind of life that was worthwhile for them. Everyone relied on these concentrates at the time in order to make it possible for these persons to lead as normal a life as possible and to prevent risks from bleeding, particularly CNS hemorrhage, which was then and is still the major cause of death in HIV-negative persons with hemophilia. Paul Russell: Why is it necessary to have such large pools? I know the risk is very much dependent on the size of the pool. Is there some scientific reason for it, or is it a practical matter of production? Elaine Eyster: All I know is that the pools can range anywhere from a few thousand, say, perhaps 5,000 (the average was about 20,000), and subsequently I have heard reports of as many as 50,000 donors. Comment from the audience: The Food and Drug Administration requires a minimum of at least 1000 different donors in pools from which gamma globulin is prepared, so as to include a wide variety of antibodies in the product. FDA standards for lot-to-lot consistency are also easier to meet with increasing numbers of donors in each lot, since the effects of anomalous donations are diluted. More to the point, pool size reductions do not effectively reduce risk for patients who require lifelong treatment.

OCR for page 81
RISK TOLERANCE 99 Bernard Horowitz: If the issue is non-A, non-B hepatitis or hepatitis C, you have to take into account the fact that the underlying risk at that time for just blood from a single donor was something close to 1 percent. Why didn't you as a treater or any of the treaters do something to avoid exposure to hepatitis C? It is very easy to calculate that even going to a single-donor product would not have helped. With respect to the pool size, it certainly would not have helped if it is not going to even help at a single-donor level. It is just not relevant to hepatitis C at that time. Did you see any difference in hepatitis C transmission rates among the users of cryoprecipitate who would use 100, 300, or 1,000 units, compared with those among the users of concentrates? Elaine Eyster: No. Once a person with severe hemophilia took the amount of cryoprecipitate that it took to treat his disease, there was virtually no difference in the incidence of hepatitis C, because he would soon be taking 1,000 units of cryoprecipitate derived from 1,000 different donors. Henrik Bendixen: I can understand why hepatitis was tolerated in the hemophilia population because the trade-off was so outstanding. It is more difficult to understand why it was tolerated for so long in patients who received transfusions for, for example, surgical operations, in whom the incidence of hepatitis was considerable. And although mortality was not known to be a great problem, morbidity certainly was. That morbidity alone in the general population should have made us a little bit more intolerant than we were. Harvey Klein: It was not widely appreciated until Harvey Alter's longitudinal study over a 20-year period that there was significant morbidity, let alone mortality. There was no excess liver-related deaths in transfused individuals, and the pathologists kept saying, "Where is all the cryptogenic cirrhosis that we are supposed to be seeing at autopsy in recipients of blood?" The answer was that it wasn't being seen. It took important, expensive longitudinal studies, primarily by an individual who stayed at the same institution for about 30 years and followed both the patients and the recipients, to demonstrate scientifically that not only was it present but that it was of clinical import. Lew Barker: I find this word "tolerate" a little bit strange. We tolerate graft-versus-host disease in people who have transplants. People who work in hepatitis were always looking for tests so that they could do something about it. When there were tests that were nonspecific, they were extensively evaluated. There were meetings on the question of the extent of pathogenesis. Until specific tests were available, the situation was very murky, and the solutions were not at hand.

OCR for page 81

OCR for page 81
Examples of Risks That We Tolerate Harvey G. Klein What risks do we tolerate and what risks don't we tolerate? The Leishmania tropica paradigm is illustrative. During Operation Desert Storm we sent one-half million men and women to the Persian Gulf and from that half million 31 service personnel later developed symptoms that were associated with infection by L. tropica.7~ L. tropica was not transmitted by blood transfusion. However, a related parasite, Leishmania donovani, has a parasitemic phase. While there is a case in the literature of a transmission by a blood transfusion, clearly this is not a major public health problem. However, in the aftermath of AIDS and out of concern for the potential transmission of leishmaniasis, the Armed Forces made the decision to defer one-half million potentially eligible blood donors for a year. There was great discussion about this decision at the time. The outcome tells us something about the general tolerance for infectious diseases or potentially infectious diseases transmitted by the blood supply. We have known for many years that transmission of bacterial infections by whole blood and by platelets occurs. Perhaps 1 in 1,000,000 infected red blood cell units transfused causes significant clinical outcomes because of bacterial contamination. As many as 1 in 2,500 platelet concentrates may transmit bacteria that are clinically important. But this doesn't seem to have raised the public ire. We tolerate this because there is no secondary transmission, because these are generally treatable infections if discovered in time, and because the fatality rate isn't appreciable or appreciated except for what we report to the Food and Drug Administration (FDA). "Tolerate" does not, however, imply acceptance or approval. There was considerable excitement a couple of years ago about one of these organisms, Yersinza enterocolitia. Y. enterocolitia is a gram-negative aerobic rod that produces endotoxin, requires iron and carbohydrate for growth, Persian Gulf Veterans Coordinating Board (1995). Unexplained illnesses among Desert Storm veterans. Archives of Internal Medicine, 155: 262-268. 101

OCR for page 81
102 BLOOD fIND BLOOD PRODUCTS: SAFETY AND RISK grows well at 4C, and infects red blood cells stored at that temperature. Prior to 1987, only six cases of transfusion-related transmission had been reported. Then ten cases and seven deaths associated with red cell transfusions were reported in the United States between 1987 and 1993. All of the recipients became symptomatic during the transfusion, but all the donors were well at the time. However, when careful epidemiologic studies were performed by the Centers for Disease Control, five of the nine donors had had gastroenteritis within 4 weeks of donation and all of the infected units had been stored for more than 25 days. The suggestion was made that blood banks could deal with this issue by either using screening histories or shortening the storage interval of blood. When you try to screen a large population of donors for gastroenteritis, you will find a lot of gastroenteritis within the last 2 weeks of donation. Questions about gastroenteritis didn't seem to select out people who transmitted Y. enterocolitia any better than people who did not. Shortening blood storage interval would have created a major supply problem for blood collectors in the United States and clearly would have had a minimal positive impact on the health of the country. We tolerated the risk of Y.enterocolitia. Graft-versus-host disease has been mentioned in the setting of bone marrow transplants. GraD-versus-host disease also occurs in the setting of blood transfusion. We have long known that it is caused by immunocompetent lymphocytes given to an immunocompromised host, but we now know that it is not just immunocompromised hosts who are at risk. A recipient whose HLA tissue type is similar to that of the blood donor, such as a relative, is also at risk. We now know, especially from studies in Japan, that it may not necessarily be a relative but anyone whose tissue type is similar because of population characteristics. We also know that mortality from transfusion- associated graft-versus-host disease approaches 90 percent. Animal studies show that it takes very few transfused cells to cause this disease. We have known this for years. We could eliminate graft-versus-host disease from blood transfusion essentially by irradiating blood at 2,500 centiGray (cGy). Do we want to do this? No. The costs outweigh the potential benefits. We tolerate this particular risk of blood transfusion for most recipients receiving blood Mom unrelated donors. Does allogeneic blood transfusion alter the recipient's immune response? The data go all the way back to the 1960s. We have known for almost 35 years that people who receive blood transfusions develop Fc receptor blocking factors. Their lymphocyte numbers decline, there is a decrease in the helper lymphocytes (CD4), and an increase in the number of the suppressor (CD8) lymphocytes. Those findings were described in hemophiliacs receiving concentrates of protein and clotting factors long before HIV was reported in

OCR for page 81
RISK TOLERANCE 103 the blood supply.72 That clearly muddied the waters a little bit when HIV entered the blood supply. We also see numbers of activated lymphocytes and down-regulation of antigen-presenting cells. These are all laboratory phenomena, and we have known about them for many years in many populations of transfused individuals. The real question was, do these have any kind of clinical significance? The answer to that is probably yes. We know that there is improved organ allograft survival in transfusion recipients. There is a suggestion that people with cancer who receive allogeneic transfusions have decreased survival or increased recurrence of their malignancy postsurgery. There are data to suggest that people who receive allogeneic blood have increased numbers of postoperative infections. There are also data to suggest that by giving allogeneic transfusions we can . prevent recurrent abortions that are immune mediated, suppress immune inflammatory disease such as ulcerative colitis and Crohn's disease, and . reactivate latent viruses, such as cytomegalovirus and HIV. The data go back to the 1970s. The work of Opelz and colleagues73 on graft survival in patients receiving cadaver kidneys demonstrates that those people who received blood transfusions had better organ graft survival, and there was also a dose response. This suggests that there was an immunosuppressive effect of allogeneic blood. There was great controversy about the data initially, but they have subsequently been confirmed. Data from 1989, reported in the New England Journal of Medicine,74 showed that if you match the donor and the recipient of allogeneic blood for one haplotype at the DR locus and then look at either kidney or heart transplant recipients, those who received blood transfusion with a match at the DR locus had a better survival of their allograft than did those who received no transfusion or received unmatched blood. This again suggests a fairly potent immunosuppressive effect of blood. 72Gowperts, E, R deBiasi, R DeVreker (1992). The impact of clotting factor concentrates on the immune system in individuals with hemophilia. Transfusion Medicine Reviews, 6(1): 4~53. 730pelz, G. DPS Sengar, MR Mickey, PI Terasaki (1973). Effect of blood transfusion on subsequent kidney transplants. Transplant Proceedings, 5: 253-259. 74Lagaaji, EL, IPH Hennemann, M Ruigrok, et al. (1989). Effect of one-HLA-DR-antigen- matched and completely HLA-DR-mismatched blood transfusions on suvival of kidney and heart allografts. New England Journal of Medicine, 321: 701-705.

OCR for page 81
104 BLOOD AND BLOOD PRODUCTS: SAFETY AND RISK At about the same time as these 1989 data, retrospective data were coming in the literature, primarily on colon cancer.75 In patients who were operated on for colon cancer, those who received allogeneic blood transfi~sions had a more rapid recurrence of cancer than those who did not. There are about 40 of these studies in the literature, including a series of prospective randomized double-blind studies. At least one of these, appearing in the American Journal of Clinical Oncolo`~,76 suggested that, using autologous blood as a control, those people who received allogeneic blood had more rapid recurrence of cancer and a higher death rate. Several other studies didn't determine that, but there is still a suggestion that this is so. Is that an immunosuppressive effect that we tolerate? I don't know. There is also the question of postoperative infection. In retrospective reviews of l9 studies, 17 studies found that transfusion was a significant predictor of postoperative infection and in 12 of these 17 studies it was the single best predictor of postoperative infection.77 Other factors that might be associated with transfusions, such as low hematocrit, blood loss during surgery, duration of surgery, and a host of other factors, when looked at in multivariate analysis, were not significant. Also, those who received autologous blood as a control in matched studies had fewer postoperative infections than those who received allogeneic blood.78 Is that also an immunosuppressive effect that we tolerate? Five prospective studies of postoperative infection also looked at allogeneic blood transfusion either without controls or with autologous blood as control, or compared allogeneic blood with allogeneic blood that had the white cells taken out of it. In four of these five studies there are fairly strong suggestions that there is an effect of allogeneic transfusion that was associated with more frequent postoperative infections.79 Some numbers, but not real data, are available from at patients with AIDS who receive blood transfusions. There was a report from Australia in 1989 suggesting that patients who were at high risk for AIDS had a shorter survival if they received transfusions. That was probably because they were sicker 7sBurrows, L, PI Tarrtar (1982). Effect of blood transfusion on colonic malignancy recurrence rate. Lancet, ii: 662. 76Heiss, MM, K-W Jauch, CN Delanoff, et al. (1994) Blood transfusion-modulated tumor recurrence: a randomized study of autologous versus homologous blood transfusion in colorectal cancer surgery. American Journal of Clinical Oncology, 12: 1859-1863. 77Blumberg, N. JM Heal (1988). Transfusion and host defenses against cancer and infection. Transfusion, 29: 236. 78Heisse, MM, W Mempel, K-W Jauch, et al. (1994). Beneficial effect of autologous blood transfusion on infectious complications after colorectal surgery. Lancet, 342: 1328-1333. 79Klein, HG (1996). Immunomodulation caused by blood transfusion. In Betz, CD, SN Swisher, S Kleinman, RK Spence, RG Strauss (ed.), Clinical Practice of Transfusion Medicine. New York: Churchill Livingstone.

OCR for page 81
RISK TOLERANCE 105 patients. John Ward reported the same kind of phenomenon Dom the Centers for Diease Control and Pevention (CDC), looking at more rapid progression of AIDS in patients who received blood transfi~sions.~ Again, one obvious explanation may be that those who were sicker needed the transfusions. Additional retrospective studies show decreased survival and increased wasting and bacterial infections in patients with AIDS who received blood transfusion. Michael Busch at Irwin Memorial Blood Center took cells infected with HIV from HIV-positive patients and cocultured these cells with a variety of allogeneic blood cells from healthy blood donors. It is well-known that if you take cells that are infected with a variety of viruses and stimulate them, you can activate the viruses, and the viruses will propagate in the cells. Busch has shown that if you use relatively small amounts of the lectin phytohemagglutinin, you can stimulate these in vivo-infected lymphocytes, and p24 antigen appears in the supematant of the culture. The next experiment he performed was to take mononuclear cells from healthy donors and coculture them with these HIV-infected lymphocytes. He used different concentrations of mononuclear cells from different healthy donors and found a dose response relationship showing that lymphocytes endogenously infected with HIV were stimulated by coculture with allogeneic mononuclear cells. If you separated out these cells and looked at granulocytes, monocytes, lymphocytes, or unfractionated mononuclear cells, you saw exactly the same kind of phenomenon, but if you took all the white cells out of the allogeneic blood by filtration or washing, or used plasma devoid of lymphocytes, you did not see this phenomenon. It appears that, in vitro, allogeneic lymphocytes or white cells have some role in the reactivation of latent viruses: cytomegalovirus, HIV, and maybe some oncogenic viruses as well. Lots of leukocytes are included in transfusions of red cells, platelets, and granulocytes. We can remove these if we want to. Is this an important clinical phenomenon, and if so should we attack it with this very large and expensive therapeutic maneuver? Or is this something that we can tolerate until further data are available? The phenomenon of reactivation of latent viruses is now being studied in vivo in a multicenter prospective study sponsored by the National Heart, Lung and Blood Institute. Data are critical when we try to determine which risks are "tolerable," although political decisions often trump data. Award, JW, JO Busch, HJ Perkins, et al. (1989). Natural history of transfusion-associated infection with human immunodeficiency virus: Factors influencing the rate of progression of the disease. New England Journal of Medicine, 321: 947-951. Busch, MP, T-H Lee, J Heitman (1992). Allogeneic leukocytes but not therapeutic blood elements induce reactivation of latent HIV-1 infection: Implications for transfusion support of infected patients. Blood, 80: 2128-2135.

OCR for page 81
106 BLOOD AND BLOOD PRODUCTS: SAFER AND RISK DISCUSSION Harold Sox: There is individual decision making, which is trying to decide whether a physician and patient should arrive at a decision to continue the factor concentrate, for instance, or try some alternative approach. There is another type of decision making, which is societal decision making. The individual decision making basically drives the question of whether to treat or not treat, and the societal question drives the question of whether we should clean up the product and get rid of the risk. We need to recognize both of those examples of risk tolerance. When we reach the end of our societal tolerance we begin to devote money to research to get the product cleaned up. Your talk dealt mostly with the question of what risks we do tolerate. At the end you talked about what risks we should tolerate. That last question is probably the one we should be asking. Elaine Eyster: How does this apply to the situation where there are no suitable alternatives? If the decision is treat or don't treat because there are no suitable alternatives of forms of treatment, then what is one to do? Harold Sox: Our committee concluded that when there is a great deal of uncertainty, when the stakes are very high, that is the time to completely inform the patient about those risks or lack of knowledge or certainty about the nature of those risks and to engage the patient fully in making a decision that we probably shouldn't try to make for them. That is the strategy that I would recommend. Elaine Eyster: What if you don't know the extent of the problem though? How do you present that information to the patients so that they can decide? Harold Sox: We describe the problem as best we can, admit our lack of knowledge such as it is, and bring the patients into the decision. The problem is that we cannot make the decision or solve the problem for them. Harvey Klein: The question of societal decision making is a very interesting one because society has already decided in terms of HIV infection what it will tolerate in the American blood supply, and that is zero, even though that is not a really rational concept from what we know about the biology of various viruses. It probably isn't a rational concept in terms of cost effectiveness, but society has clearly made that decision. In terms of individual decision making, we now are going to be asking donors about Creutzfeldt-Jakob disease (CJD). Conceivably, we should be informing patients that this is a potential risk in blood transfusion. I would find that very difficult to explain to a

OCR for page 81
RISK TOLERANCE 107 patient, because there never has been a case transmitted by transfused blood, but because we are going to be screening our blood supply for CJD, it raises the question of what we do at the patient end. Elaine Eyster: We are starting to tell our hemophilia patients about CJD. It is one of the most absurd discussions I have ever had, trying to tell them that there might be this problem but we don't know if there is a problem, and although the disease has never been shown to be transmitted in blood, it could be. They need the product, and even the recombinant version might not be totally free of it for all we know. Lew Barker: I would like to take this a little further into the realm of deep uncertainty. It is becoming evident that we all walk around with a large collection of fragments of retroviral genomes in many of our somatic cells or white cells. These may or may not be dealt with by irradiation or whatever measures we have. They might even be activated, and we obviously transfuse these all the time in people. We have other ways of exchanging lymphocytes, as was learned from infectious mononucleosis some years ago. One of the reasons that I have a problem with "tolerating" the associated but totally unknown risk is that we don't have any idea what to do about it. Other than the attractive alternative of trying to move totally away from allogeneic blood transfusions, I think once we understand jumping genes or whatever they are, there will be something else to concern ourselves with. These may be the problems of the next century associated with allogeneic blood transfusions. Asking people to make these decisions when they cannot really understand these things and we cannot explain them very well is a tough conundrum. James Allen: Certainly science and literacy are becoming real barriers for us now in terms of communication of risk, communication of what the actual threats are, and what can be done about it. How individuals view risk is very interesting. Risks that they choose, such as riding on a motorcycle with or without a helmet, are risks that individuals freely make for themselves. But when it comes to blood, the media has helped us foster in this country the concept that we should not tolerate any risk in the blood supply. If we move away Tom allogeneic transfusions, what are we going to substitute them with, and what are the risks associated with that? Whatever it is, maybe we can reduce the risks, but it may take an long time for us to really establish that, given the level of risk that we have at present. There is also finite risk because we cannot eradicate all risk. In the early 1980s there were two temporally fairly closely linked episodes of bacterial contamination with Pseudomonas pnorescens in red cell units coming from the same blood collection center. FDA and CDC got involved in the investigation

OCR for page 81
108 BLOOD AND BLOOD PRODUCTS: SAFETY AND RISK in that blood center. Subsequently, a third case occurred several months later, and it was from the same source. The investigators went through the whole blood system at that collection center and never could identify a real breach in the process or a potential source of the infection. The problem either went away or never recurred. Whatever that risk is, it is still there. It is an intrinsic risk in the system. We didn't fix anything. We have to acknowledge that there is that baseline level of a potential problem in the system. To the extent that we can identify possible solutions and put in those solutions at a reasonable cost given the resources available, we should do so, but there is alwaYs Doing to be a finite risk of one type or another. , ~ ~ . Joseph Fratantoni: Much as science has evolved in the last 20 or 30 years, I think this policy of bringing the patient into the decision has evolved somewhat compared with the early 1 980s. Harold Sox: That is correct. The style of decision making back in the early 1980s was much more weighted toward the physician taking the lead and the patient basically nodding and saying, "Yes, doctor, yes, doctor." For many decisions video disks are now available to inform the patients. When they come to the conference with the physician it is on a reasonably equal footing. Things have changed a lot. Michael Stoto: The challenge is not only talking honestly about the risk of the uncertainty with a patient but trying to find ways to adapt that in individual circumstances to help the doctor and the patient work together to understand the risks and benefits in a given situation, and what they should do together, given what they know and what they don't know, in order to make a decision about this patient's care. I don't think we are doing it yet, but many are attempting not only to be clear about the uncertainty but really to adapt what is known to individual patients.