NoBlood

Jan B. Wade · #1 (**permalink**) 10-13-2005, 06:18 PM

http://www.medscape.com/viewarticle/412841 (registration required)

Indications for Blood Transfusion in Critical Care Patients Reviewed CME

William J. Sibbald, MD, FRCPC, FCCHSE

Introduction

During the course of a critical illness, the clinician frequently questions the "best" hemoglobin concentration (or hematocrit level) for her patient. In sepsis or systemic inflammatory response syndrome (SIRS), systemic inflammation and widespread tissue injury complicate an insult, infectious (sepsis) or otherwise (SIRS). As the normal process of metabolic autoregulation is often disturbed in these illnesses, more reliance on convective O₂ delivery to support tissue O₂needs ensues. For many years, survival in the sepsis/SIRS continuum has, therefore, been linked to strategies that focus on improving tissue O₂availability, especially by increasing delivery. Blood transfusions to increase O₂ carriage have become a frequently employed means of achieving this therapeutic target, especially since anemia is common in these patients (due to occult blood loss and depressed erythropoiesis.^[1]

Recently, however, concern has been expressed that blood transfusions may also create harm despite their potential for benefit.^[2] The concern of an unfavorable risk:benefit ratio of blood transfusions has prompted research into alternative approaches to improving tissue O₂ delivery -- for example, by the use of blood substitutes and synthetic erythropoietin (EPO). One of the interesting consequences of these research activities has been a need to better understand clinical decision-making in the context of blood transfusions and to "target" hemoglobin levels in critically ill patients. It was in this context that the Society of Critical Care Medicine sponsored an educational session on blood and blood transfusion.

Anemia and the Metabolic Regulation of O₂ Delivery in Critical Illness

Clinical studies show that anemia is common in septic patients.^[3-5] These studies also show that blood transfusions are frequently prescribed to maintain the hemoglobin at more than 100 g/L. In contrast, a clinical trial has demonstrated it may not be necessary to transfuse red blood cells (RBCs) in patients without ischemic heart disease until the hemoglobin concentration falls to less than 60 g/L to 70 g/L.^[4] The ability to tolerate low hemoglobin levels, even in critical illness, is explained by the circulation's exceedingly efficient compensation to anemia.

Tissue oxygenation requires that O₂supply be "matched" with O₂ needs -- and supply can be understood by review of the O₂ delivery equation: O₂delivery = [Ca O₂ Sa O₂ x 1.39] + 0.003 pa O₂

In critical illness, limitations to tissue oxygenation affect all 3 levels of the circulation. In animal models of isovolemic hemodilution, for example, O₂delivery (QO₂) in sepsis is redistributed to the heart and brain because their ability to increase O₂ extraction is limited.^[6,7] This redirected QO₂ comes from "non-vital" organs such as the splanchnic circulation where tissue oxygenation is then maintained because of this organ's capacity to increase O₂ extraction. In addition to such regional circulatory compensation, central and microregional circulations also adapt in anemia.^[6] These changes include an increase in cardiac output and microcirculatory changes that promote improved tissue O₂ extraction, respectively. In summary, blood flow in response to falling hemoglobin concentrations is regulated at the 3 levels of the circulation: (1) the central (ie, increased cardiac output), (2) the regional (ie, a redistribution of blood flow between organs), and (3) the microregional (ie, a redistribution of blood flow within organs).

When the capacity of compensatory mechanisms to anemia is exhausted, hypoxia and, subsequently, tissue injury may then supervene. Because of a generalized circulatory dysfunction, the effectiveness of the circulation's compensation to anemia may be blunted in sepsis/SIRS. For example, animal research shows that the hemoglobin concentration exerts independent and negative effects on the QO₂ in both central and regional circulations.^[7-9] In 1 experiment, myocardial O₂ delivery was sustained after exposure to acute hypoxia only in septic rats that had been transfused to maintain hematocrit levels more than 45%.^[7] This means that sepsis elevated the "optimal" hemoglobin levels in rat hearts ("optimal" hemoglobin is that level required to support tissue oxygenation). In another experiment, sepsis reduced the heart's O₂ extraction reserve (O₂extraction "reserve" is the degree to which O₂ extraction can be increased) -- and transfusing RBCs to normalize hemoglobin levels minimized this lesion.^[9]

There is some biologic plausibility for the notion that the sepsis/SIRS continuum could put demands on the "optimal" hemoglobin levels, which is not clinically apparent. For example, increased RBC transit times in the microcirculation in anemia could contribute to an O₂ extraction defect in sepsis as transit times increase with progressive anemia. It is also possible that normalizing hemoglobin levels by transfusion could lessen the progression of microcirculatory injury in sepsis by reducing the mediator and/or endotoxin load, which contributes to capillary injury and/or carrying and offloading more nitric oxide to the microcirculations (thus providing "cytoprotective" functions in sepsis).

What Do Clinical Studies Tell Us About What We Are Doing?

Transfusing blood is commonly used to improve convective O₂ delivery in critically ill patients.^[3] Despite the publication and dissemination of guidelines describing "consensus" and "best practices" concerning RBC transfusion in hospitalized patients,^[10] there remains considerable variability in individual transfusion practices.^[3] An example of this was also found in the European Transfusion Practices study, a systematic collection of data on all adult patients admitted to 145 nonspecialized intensive care units (ICUs) and collected over a 2-week period in Western Europe.

Jean-Louis Vincent, MD, PhD,^[11] from the Universite Libre de Bruxelles in Brussels, Belgium, reporting on behalf of the study team, indicated that the occurrence of anemia as a component of critical illness was common -- by the time of ICU discharge, the hemoglobin concentration was grouped around a mean of 100 g/L to 110 g/L. With a transfusion trigger (transfusion "trigger" is the hemoglobin concentration used by the clinician to prescribe blood transfusions) that seemed to be around 85 g/L, more than 35% of patients were transfused during their ICU stay. A predictor of blood transfusions seemed to be the length of stay. For patients admitted for at least 48 hours, the frequency of transfusion increased to 56%; for patients admitted for more than a week, 73% were transfused. While patients were more commonly transfused after emergency surgery or trauma, transfusion was also common in patients following elective surgery and in patients with primarily a medical diagnosis. Of note, the use of blood transfusion was not only related to the presence of overt bleeding, but "inadequate hemoglobin without evidence of bleeding" was reported as a common indication for blood transfusion.

Mitchell M. Levy, MD,^[12] of Brown University School of Medicine in Providence, Rhode Island, also reported on the CRIT trial, a North American equivalent of the European Transfusion Practices study. This is a prospective, multicenter, observational cohort study of ICU patients whose goal is to define transfusion practices in American ICUs. The target enrollment to study is 5000 patients. Patients are enrolled within 48 hours of ICU admission and (planned) follow-up is 30 days. Data from a very early and interim analysis suggests the transfusion trigger is similar to what the European study found -- approximately 80 g/L. It is probably higher in patients with cardiovascular disease (approximately 10g/L.). Similar to the European study, exposure to blood transfusion seems to increase with the length of time spent in the ICU.

In a study of RBC transfusion practices in a sample of Canadian ICUs, Hebert and colleagues^[3] have previously reported that heterogeneity characterizes transfusion practices. This work confirmed the findings of the European ICU study; namely, patients who die in ICUs have lower hemoglobin values and are transfused RBCs more frequently. Additionally, higher hemoglobin values in anemic patients with coexistent cardiac disease seemed to be associated with improved outcomes, measured as survival. The latter finding needs to be confirmed with appropriate clinical studies.

What Do Clinical Studies Tell Us About What We Should Be Doing?

Subsequently, Paul C. Hebert, MD,^[13] of the University of Ottawa in Ottawa, Canada, reviewed the findings of the randomized, controlled trial undertaken in Canadian ICUs and reported last year.^[4] The objective of that study was to determine if a restrictive strategy of blood transfusion (transfuse only if the hemoglobin concentration was less than 70 g/L) was equivalent to a more liberal approach (transfuse if the hemoglobin concentration was less than 100 g/L).

As an "equivalency" study, Dr. Hebert noted that the study was designed to exclude differences, using an "effectiveness" approach. What that means is that the study design did not control any other treatment practice for the enrolled patients except the hemoglobin level at which blood transfusion would be prescribed. In the restrictive group, the hemoglobin concentration was maintained between 70 g/L and 90 g/L. In the liberal group, the range was kept between 100 g/L and 120 g/L. Initially, about 2000 patients were screened for inclusion. Subsequently, 420 were randomized to the restrictive group and 418 to the liberal group. Once randomized, the restrictive group received about 2.5 units of blood and the liberal group received twice that amount.

Analysis showed that patients with an acute physiology and chronic health evaluation (APACHE) II score less than 20 had a better survival when allocated to the restrictive group. Regarding pulmonary edema and acute respiratory distress syndrome, an overwhelming effect was seen in the restrictive group -- that is, there was less pulmonary edema, both cardiac and noncardiac, in the group associated with restrictive transfusion practices (however, the absence of pulmonary edema did not seem to explain the survival benefit). Dr. Hebert emphasized that this part of the study was too underpowered to determine if patients transfused liberally failed to wean from mechanical ventilation any faster than patients allocated to the restrictive group.

From the Hebert study, we learn that blood transfusions are probably not to be recommended in the critically ill patient (without heart disease and with an APACHE II score < 20) until the hemoglobin concentration is less than 70 g/L. The question then becomes the following: "Is the survival advantage in patients with a hemoglobin less than 100 g/L because the critically ill are really better off with a lower hemoglobin concentration?" (ie, the optimal hemoglobin concentration is depressed in a critical illness) or "is there harm associated with the use of packed, stored blood as a means of keeping the hemoglobin over 70 g/L to 80 g/L?" There may be some evidence favoring the latter explanation.

What Might the Problem Be?

William J. Sibbald, MD,^[14] of the University of Toronto, Toronto, Ontario, Canada, reviewed animal and clinical studies that may provide evidence for "harm" of transfusing blood into patients with a damaged microcirculation, as occurs in sepsis. In a study reported in the early 1993, blood transfused to increase the hemoglobin concentration in septic ICU patients improved calculated O₂ delivery but concurrently depressed intramucosal oxygenation (measured by gastrointestinal tonometry) when blood administered had been stored for more than 12 to 15 days.^[15] The biologic rationale for this result was then explored over the subsequent years. In an animal model developed to assess the efficacy of transfused blood, a consistent finding was a lack of efficacy of transfused "old" stored rat blood in improving tissue O₂ consumption compared with fresh stored blood (and with blood substitutes).^[16] Potential reasons for a lack of efficacy of "old" stored RBCs in acutely restoring tissue oxygenation are at least 3 in number: (1) 2,3-diphosphoglycerate (DPG) depletion; (2) altered erythrocyte-endothelial interactions; and (3) altered RBC/nitric oxide homeostasis.

In both humans and in nonhuman primates, systemic DPG levels, as well as p50 values (a measure of oxyhemoglobin affinity indicated by the O₂ tension at 50% hemoglobin saturation), fall after infusing DPG-depleted RBCs (while 2,3-DPG undergoes regeneration following transfusion, the rate at which levels return to normal in man is variable and can take from 24 hours to several days). A second possible cause of failure of "old" transfused blood to increase tissue oxygenation is a loss of membrane deformability as occurs in stored RBC and the possibility of increased adherence of RBCs to activated endothelium (as may occur in the sepsis/SIRS continuum.^[17] In vitro studies demonstrate that RBCs treated with endotoxin have an increased adhesiveness to pulmonary artery endothelial cells, dependent upon the concentration of lipopolysaccharide and the presence of divalent cations calcium and magnesium.^[17] Finally, if blood storage depleted the ability of the red cell to form nitrosylated hemoglobin, then a critical imbalance could occur favoring excess nitric oxide binding to heme, which would potentially cause vasoconstriction and thereby paradoxically reduce O₂ delivery.

More Alternatives?

Howard Corwin, MD,^[18] of the Department of Anesthesiology at Dartmouth-Hitchcock Medical Center in Lebanon, New Hampshire, reported on alternatives to blood transfusions to increase hemoglobin concentrations in critically ill patients. Because there is an anticipated shortage of blood and concern there may be an adverse effect of using stored blood in critically ill patients, stimulation of erythropoiesis with EPO has been explored for its efficacy. Dr. Corwin discussed the randomized, controlled trial of patients allocated to EPO on ICU day 3. In an intent-to-treat analysis, patients were followed for 42 days following randomization. EPO 300 units or placebo were administered subcutaneously, daily for 5 days and then every other day for 2 weeks. Patients who remained in the ICU for greater than 2 weeks received EPO until they left the ICU. Study drugs were held for hematocrits greater than 38 and all patients received supplemental iron. "Physician practice" was allowed regarding the administration of blood transfusions. Blood transfusions were reduced by almost 50% in the EPO group. The total number of units transfused in the EPO group was 166 and total units transfused in the placebo group was 305.

Summary

There are many reasons for the observed variability around blood transfusion practices in the ICU. There is clearly clinical uncertainty about the optimal hemoglobin level -- that concentration of hemoglobin that, after appropriate volume resuscitation, maximizes tissue oxygenation and facilitates survival. Clinicians believe that improving tissue oxygen availability is an important component of their care for critically ill patients. However, clinical studies are beginning to suggest that the frequency of blood transfusions used in ICUs in the past may not be necessary to optimize patient survival in the future. Indeed, there is some intriguing evidence that excessive use of blood transfusions may be harmful. Because there is some concern for the use of blood transfusions as a means to improve tissue oxygenation, and there is clearly concern about supply when blood transfusions are appropriate, other strategies, such as the use of synthetic EPO, may become an option. What is certain is something we have taken for granted -- the use of blood transfusions to improve tissue oxygenation in critically ill patients is undergoing intense re-examination both at the bench and at the bedside. This is clearly a "story in progress," with studies both planned and underway to identify how best to use blood, blood products, blood substitutes, and synthetic EPO in the ICU.

References

Krafte-Jacobs B, Levetown ML, Bray GL, Ruttimann UE, Pollack MM. Erythropoietin response to critical illness. Crit Care Med. 1994;22:821-826.
Purdy FR, Tweeddale MG, Merrick PM. Association of mortality with age of blood transfused in septic ICU patients. Can J Anaesth. 1997;44:1256-1261.
Hebert PC, Wells G, Martin C, et al. Variation in red cell transfusion practice in the intensive care unit: a multicentre cohort study. Crit Care. 1999;3:57-63.
Hebert PC, Wells G, Blajchman MA, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. Transfusion Requirements in Critical Care Investigators, Canadian Critical Care Trials Group. N Engl J Med. 1999;340:409-417.
Van Iperen CE, Gaillard CA, Kraaijenhagen RJ, Braam BG, Marx JJ, van de Wiel A. Response of erythropoiesis and iron metabolism to recombinant human erythropoietin in intensive care unit patients. Crit Care Med. 2000;28:2773-2778.
Chapler CK, Cain SM. The physiologic reserve in oxygen carrying capacity: studies in experimental hemodilution. Can J Physiol Pharm. 1986;64:7-12.
Morisaki H, Sibbald WJ, Martin CM, Doig G, Inman K. Hyperdynamic sepsis depresses circulatory compensation to normovolemic anemia in conscious rats. J Appl Physiol. 1996;80:656-664.
Fox G, Bersten A, Lam C, et al. Hematocrit modifies the circulatory control of systemic and myocardial oxygen utilization in septic sheep. Crit Care Med. 1994;22:470-479.
Bloos FM, Morisaki HM, Neal AM, Martin CM, Ellis CG, Sibbald WJ. Sepsis depresses the metabolic oxygen reserve of the coronary circulation in mature sheep. Am J Respir Crit Care Med. 1996;153:1577-1584.
Perioperative red cell transfusion. NIH Consens Statement. 1988;7:1-6.
Vincent JL. Transfusion in critical care. Program and abstracts of the 30th International Educational and Scientific Symposium of the Society of Critical Care Medicine; February 10-14, 2001; San Francisco, California.
Levy MM. Transfusion in critical care. Program and abstracts of the 30th International Educational and Scientific Symposium of the Society of Critical Care Medicine; February 10-14, 2001; San Francisco, California.
Hebert P. Transfusion in critical care. Program and abstracts of the 30th International Educational and Scientific Symposium of the Society of Critical Care Medicine; February 10-14, 2001; San Francisco, California.
Sibbald WJ. Transfusion in critical care. Program and abstracts of the 30th International Educational and Scientific Symposium of the Society of Critical Care Medicine; February 10-14, 2001; San Francisco, California.
Marik PE, Sibbald WJ. Effect of stored-blood transfusion on oxygen delivery in patients with sepsis. JAMA. 1993;269:3024-3029.
Fitzgerald RD, Martin CM, Dietz GE, Doig GS, Potter RF, Sibbald WJ. Transfusing red blood cells stored in citrate phosphate dextrose adenine-1 for 28 days fails to improve tissue oxygenation in rats. Crit Care Med. 1997;25:726-732.
Eichelbronner O, Sielenkamper A, Cepinskas G, Sibbald WJ, Chin-Yee IH. Endotoxin promotes adhesion of human erythrocytes to human vascular endothelial cells under conditions of flow. Crit Care Med. 2000;28:1865-1870.
Corwin HI. Transfusion in critical care. Program and abstracts of the 30th International Educational and Scientific Symposium of the Society of Critical Care Medicine; February 10-14, 2001; San Francisco, California.