Patients with ischemic heart disease have always been of special concern for transfusion therapy. There are a number of observational studies showing that moderate to severe anemia in patients with cardiac disease is associated with adverse outcomes, including higher mortality rates. This observation correlates nicely with what we know about the unique aspects of the physiology of our beating hearts. While most tissues have the ability to extract more oxygen from the blood when demand increases, the heart muscle (myocardium) extracts near-maximal concentrations of oxygen under resting conditions.
For that reason, the primary adaptive response to increased oxygen demand or decreased oxygen supply is to increase coronary artery blood flow. However, this adaptive response is impaired in patients with abnormal coronary circulation, such as atherosclerosis, leading to critical oxygen delivery levels at lower hemoglobin levels than in patients with “normal” coronary circulation. For these reasons, patients with ischemic heart disease have traditionally been transfused at higher hemoglobin levels than other patients with the assumption that blood transfusions improve outcomes.
I discussed the Transfusion Requirements in Critical Care (TRICC) and its implications in a previous article. The TRICC trial compared a liberal transfusion strategy (transfusion trigger of 10 gm/dL) to a restrictive strategy (transfusion trigger of 7 gm/dL) in 838 critically ill patients.1 The conclusion was that no patients benefited from a more liberal strategy, and most had better outcomes being transfused more conservatively (less is more).However, patients with severe cardiac disease were specifically excluded from the recommendation that “a restrictive strategy of red cell transfusions is at least as effective as and possibly superior to a liberal strategy in critically ill patients, with the possible exception of patients with acute myocardial infarction or unstable angina.” Furthering the concern about patients with ischemic heart disease was the criticism that some physicians had refused to enroll high risk cardiac patients in the TRICC trial, lowering the number of cardiac patients that were studied. This issue was addressed in a post-hoc analysis of cardiac patients in the TRICC trial, which still concluded that “most hemodynamically stable, critically ill patients with cardiovascular disease may receive a transfusion safely when hemoglobin concentrations decrease below 7.0 g/dL and may be maintained at hemoglobin concentrations between 7.0- 9.0 g/dL.”2 However, the possible exception to conservative therapy remained patients with acute coronary syndromes (ACS). In the same year this secondary analysis was published, a large retrospective review of patients having myocardial infarctions was published in the New England Journal of Medicine which also cautioned against restrictive transfusion therapy.3 While the study was impressive in the number of patients included (79,000), it was of overall low quality because the patient population was from the early 1990’s, a number of patients were excluded from review, and the primary emphasis was admission hematocrit without a variety of other confounding factors being considered. In spite of the limitations, the conclusion was not terribly surprising: anemic heart attack victims do worse than non-anemic heart attack victims, and the more anemic they are, the worse they do. What was surprising was the accompanying editorial that failed to emphasize the multiple limitations of the study, and instead endorsed a transfusion trigger as high as 11 gm/dL in patients with myocardial ischemia to avoid adverse outcomes. What we now know is that this editorial recommendation has led to many more deaths for heart attack victims, not less.
The first investigator to scientifically question the logic of anemia and transfusions in patients with ischemic cardiac disease was Dr. Sunil Rao from Duke. In a 2004 study published in JAMA, Rao reviewed 24,000 patients enrolled in FDA trials for acute coronary syndromes.4 Using this comprehensive data set and a variety of sophisticated statistical tools, including multivariate analysis and propensity scoring, he came to a shocking conclusion: patients with acute coronary syndromes get worse with blood transfusions, not better. In fact, his conclusion was that heart attack patients transfused at a hematocrit above 25% (or a hemoglobin of 8 gm/dL) were more likely to have a second heart attack and were four times more likely to die within a month. Understandably, this study was met initially with skepticism due to the counter- intuitive findings, but there are now at least two other published studies with similar conclusions. A study of 2400 myocardial infarction patients in Israel demonstrated remarkably similar findings to the Rao study,5 and an analysis of over 4100 patients with ST- elevation MI (STEMI) using propensity scoring concluded that mortality rates in transfused patients were three to four times higher at 30 days, six months and one year.6 The conclusion of the latter study was “blood transfusion is associated with increased short- and long- term mortality in the setting of STEMI.”
To me, the most surprising study in the cardiology literature regarding transfusion risks and benefits examined the impact of bleeding and transfusion during cardiac catheterization, also known as percutaneous coronary intervention (PCI).7 A major source of conflict between cardiac surgery teams and cardiologists has been a differing perception of the relative risks of ischemia and bleeding. The cardiac surgery and critical care literature for the past decade has recognized the significant risks of transfusions, but the cardiology literature has emphasized the risks of ischemia, at times seeming to treat bleeding and transfusion as more of a nuisance complication. The result of this emphasis on ischemia has been a push by cardiologists to take cardiac patients to surgery shortly after loading doses of platelet inhibitors, resulting in significantly higher bleeding and reoperation rates. There have also been concerns about unrecognized blood loss during PCI, resulting in more anemia upon presentation to the cardiac OR and ultimately to more blood transfusions. Published in the Journal of the American College of Cardiology, data from 90,000 patients undergoing PCI were examined to determine the independent effects of bleeding and blood transfusion. Similar to the previously mentioned studies, the conclusion was that bleeding and transfusion increased the risk of in hospital mortality by 350% (3.5X). Of note, this hazard was greater than that observed between in-hospital myocardial infarction (MI) and 1-year mortality in PCI patients. To state this more simply, if you are forced to choose between a heart attack and a unit of blood while having a cardiac catheterization, you should opt for the heart attack because your risk of dying is actually lower. The authors summed up by stating: “Collectively, these studies demonstrate a robust association among major bleeding, blood transfusion, and increased mortality in patients undergoing PCI and are in keeping with reports of decreased survival among patients with acute coronary syndromes who experience major bleeding or require blood transfusion.”7
I consider this to be a “must read” paper for cardiologists not only because of the clinical findings, but also because of an in-depth discussion of the possible mechanisms in play. Cardiologists particularly dislike inflammation and coagulation in their cardiac patients, but when they transfuse blood products they are in fact giving a proinflammatory procoagulant due to the effects of the “storage lesion.” Compounding the issue, many physicians still adhere to the adage “if you’re going to give one unit, give two”, so volume overload may also be a contributory factor. Undoubtedly the increase in mortality for PCI patients with bleeding complications is a series of events, beginning with recognized or unrecognized bleeding in a major artery (femoral artery), which then leads to tachycardia, hypotension and activation of the sympathetic and coagulation systems, which then leads to blood transfusions which help to fuel inflammation and coagulation. In any event, the best course of events is primary prevention, that is, to prevent the bleeding complication from occurring in the first place. According to Dr. Rao, bleeding in PCI can occur in up to 10% of cases, resulting in a tremendous increase in costs and complications. Recommended strategies to reduce bleeding complications in PCI include vascular closure devices and/ or more attention to groin site pressure and hemostasis. My recommendation is to use ultrasound guidance for catheterization sheath placement, rather than performing a blind “stick” which can result in vascular injury. Vascular surgeons and anesthesiologists have successfully used ultrasound guidance for central line placement and for nerve blocks, so cardiologists could easily master this technique. Finally, Dr. Rao strongly recommends a radial artery approach to PCI which dramatically reduces complications because the location in the wrist makes bleeding detection and control much easier. The argument against radial PCI is that the smaller artery makes the approach technically more challenging and perhaps not as widely applicable. Dr. Rao tends to disagree and points to the fact that less than 5% of PCIs are currently done using the radial artery, so perhaps there is some prejudgment. I would only offer that many general surgeons initially doubted that a laparoscopic approach to gallbladder surgery was worth the time and effort to learn given their years of experience with an open approach; it is now the standard of care.
- Hébert PC, Wells G, Blajchman MA, et al. A multicenter, randomized, controlled trial of transfusion requirements in critical care. NEJM 1999;340(6):409-68.
- Hebert PC, Yetisir E, Martin C, et al. Is a low transfusion threshold safe in critically ill patients with cardiovascular diseases? Crit Care Med 2001;29(2):227-34.
- Wu WC, Rathore SS, Wang Y, et al. Blood transfusion in elderly patients with acute myocardial infarction. N Engl J Med 2001;345(17):1230-36.
- Rao SV, Jollis JG, Harrington RA, et al. Relationship of blood transfusion and clinical outcomes in patients with acute coronary syndromes. JAMA 2004;292(13):1555-62.
- Aronson D, Dann EJ, Bonstein L, et al. Impact of red blood cell transfusion on clinical outcomes in patients with acute myocardial infarction. Am J Cardiol 2008;102:115-19.
- Shishebohr MH, Madhwal S, Rajagopal V, et al. Impact of blood transfusions on short-term and long-term mortality in patients with ST-elevation myocardial infarction. J Am Coll Cardiol Intv 2009;2(1):46-53.
- Doyle BJ, Rihal CS, Gastineau DA, et al. Bleeding, blood transfusion, and increased mortality after percutaneous coronary intervention: implications for contemporary practice. J Am Coll Cardiol 2009;53(22):2019-27.