Reports

This report is based on medical evidence presented at sanctioned medical congress, from peer reviewed literature or opinion provided by a qualified healthcare practitioner. The consumption of the information contained within this report is intended for qualified Canadian healthcare practitioners only.

12th Annual Update on Cardiopulmonary Bypass

Whistler, British Columbia / March 18-23, 2007

As discussed by Dr. Bruce Spiess, Professor of Anesthesiology and Emergency Medicine, Virginia Commonwealth University, Richmond, cardiac surgery uses approximately 20% of the US blood supply. “It has been estimated that a minimum of 25 to 30% of this blood is unnecessary,” Dr. Spiess noted—and the percentage may well be much higher. Moreover, a significant majority of transfused blood goes to a relatively small proportion (about 20%) of cardiac surgery patients. A concerted effort to reduce blood use in this high-risk group could therefore have a substantial impact on overall blood use, he suggested. Certainly, there are good reasons why physicians might want to avoid transfusion of any blood product wherever possible. According to observational database information, transfusion is associated with multiple negative outcomes and the more units a patient receives, the higher the apparent risk.

In his interpretation of the data from a recently published study, Dr. Spiess suggested that one out of 73 to one out of 193 transfusions given to critically-ill patients in the surgical intensive care unit (ICU) was associated with transfusion-related acute lung injury (TRALI) (Rana et al. Transfusion 2006;46(9):1478-83). In turn, the development of TRALI was associated with a 47% mortality rate. Investigators also found that the incidence of transfusion-associated circulatory overload (TACO) in the same study population occurred in one out of 50 patients; TACO was associated with a 20% mortality risk. If these numbers hold true for every ICU in the US, then TRALI and TACO are responsible for between 25,000 and 50,000 deaths a year, Dr. Spiess observed. Transfusion is also associated with an increased risk of ventricular failure, perioperative atrial fibrillation and septicemia, bacteremia and superficial infection. Researchers also found that over a period of 60 months, transfused patients who had undergone coronary artery bypass grafts (CABGs) had twice the mortality rate at 15% compared with 7% for non-transfused patients (Engoren et al. Ann Thorac Surg 2002; 74(4):1180-6). After adjusting for comorbidities and other confounders, transfusion was still associated with a 70% increase in mortality, according to study authors.

It would also appear to be advantageous for a critically ill patient’s hemoglobin levels to drop <70 g/L before initiating red-cell transfusion compared with a more liberal transfusion strategy. In a published study, overall 30-day mortality rates were similar in patients who were transfused when their hemoglobin levels dropped <70 g/L (maintained at between 70 and 90 g/L thereafter) and in those who were transfused when hemoglobin dropped <100 g/L (maintained at 100 to 120 g/L thereafter) (Hébert et al. N Engl J Med 1999;340(6): 409-17). Mortality rates also did not differ between the two groups in patients with clinically significant cardiac disease. However, they were significantly lower among patients who were less acutely ill if they were randomized to the restrictive transfusion strategy as well as among patients who were <55 years of age. In-hospital mortality rates were also significantly lower in the restrictive-strategy group. “We do not know how many times transfusion saves vs. kills but we do know it is costly,” stated Dr. Spiess, “and if you and I cut blood utilization for heart surgery in half, scarcity and cost [of blood] worldwide should change.”

As demonstrated by Dr. Robert Higgins, Professor and Chair, Cardiovascular and Thoracic Surgery, Rush University Medical Center, Chicago, Illinois, blood conservation strategies can significantly decrease the amount of blood needed in surgical practice. As he discussed, the Rush University Medical Center implemented a comprehensive blood management program in which they mandated that all blood transfusions be limited to appropriate clinical indications rather than transfusing to achieve specific hemoglobin or hematocrit levels. Based on this program, patients are no longer eligible for either packed red-blood cells or autologous blood until their hematocrit reaches 17% (20% if they have a history of carotid disease or transient ischemic attacks). “We also discuss all blood-product use decisions with the staff physician,” Dr. Higgins told listeners, “and we do not rely on a single measure of either hemoglobin or hematocrit as a transfusion trigger.”

A comparison of blood transfusion rates and resource utilization for all adult cardiac surgery patients prior to the introduction of their comprehensive blood conservation program vs. several years thereafter showed that costs associated with reduced blood loss in cardiac surgery dropped from approximately $120,000 to approximately $50,000 of total costs. Multiple regression analysis also revealed that blood transfusion was the only statistically significant variable affecting 30-day mortality in all cardiac surgery cases in the program database from 1997 to 2000. “Implementation of a clinical practice guideline which challenges the conventional 10 g/dL transfusion trigger can decrease transfusion rates and resource utilization,” Dr. Higgins concluded, “and blood conservation strategies may be employed in modern cardiothoracic surgical practice with significant benefits to patients.”

OnTrac Ontario Initiative

In the province of Ontario, an initiative has been underway since January 2002 whose aim is to reduce red cell use by 5 to 10%. The OnTrac initiative involves 23 hospitals in Ontario, all of whom have taken a systematic approach towards identifying patients at risk for transfusion prior to surgery and then correcting any possible underlying contributing factors to reduce this risk. Regarding transfusion requirements for three different surgeries (knee arthroplasty, abdominal aortic aneurysm and primary CABG), there was approximately a 35% reduction in transfusion requirements for CABG patients alone 18 months after introducing the program. “Reduction in allogeneic transfusion markedly exceeded the anticipated 5 to 10% for each of these procedures,” reported Dr. C. David Mazer, Department of Anesthesia, St. Michael’s Hospital, Toronto, Ontario.

There was also a net fall in red cell use between 2005 and 2006 among hospitals involved in the OnTrac program compared to those who were not, among whom there was a net increase over the same interval. The average length of stay was also lower among patients who did not receive allogeneic transfusion and the risk of post-operative infection was also significantly higher among transfusion recipients, especially among those who received at least six units of blood. Given the cost of $400/unit of blood, the reduced length of stay in non-transfused patients and consequently reduced work in laboratories and nursing units, the program has also saved the province almost $15 million a year, “so the goal is to keep funding this program and to expand it to other surgical procedures,” Dr. Mazer noted.

He also discussed the risks vs. the benefits of antifibrinolytic therapy, notably aprotinin. Studies have clearly demonstrated that antifibrinolytics reduce the need for transfusion and they have become a standard of care for many cardiac procedures. Yet there has been some controversy about the renal safety of aprotinin. A recent study matched 449 aprotinin-treated patients to 449 patients treated with tranexamic acid from a single-centre database of over 10,000 patients (Karkouti et al. Transfusion 2006;46(3):327-38). Although there was no difference in transfusion rates between the two groups, investigators found that there may be an association between aprotinin use and renal dysfunction. A recent review of the clinical data on aprotinin by the Food and Drug Administration (FDA) in the US suggests that the agent has an “acceptable safety profile.” However, the FDA still recommends that physicians carefully monitor patients for the occurrence of toxicity, particularly to the kidney, heart and central nervous system, and limit its use to those situations where the benefit of reduced blood loss is essential to medical management of the patient.

More definitive data on the safety of antifibrinolytics will be provided by the BART (Blood Conservation Using Antifibrinolytics: Randomized Trial in High-risk Cardiac Surgery), a Canadian multicentre study in which approximately 3000 high-risk patients will be randomized to receive either aprotinin, tranexamic acid or epsilon-aminocaproic acid. Pooled data from a second interim analysis of the BART study indicate that so far, 7.5% of randomized patients experienced massive bleeding, 2.5% required massive transfusions and there was a 0.4% mortality rate due to hemorrhage. Approximately 7% of the BART cohort had to undergo re-operation for massive post-operative bleeding. To date, the stroke and myocardial infarction (MI) rates in BART are each slightly over 3% while the overall mortality rate is approximately 4%.

“Antifibrinolytic therapy is not a good substitute for good surgical technique,” Dr. Mazer cautioned, “but the study will provide important new information on the safety and efficacy of [these] drugs.”

Treatment and Prevention of Massive Blood Loss

Another potentially important new addition to the treatment and, in certain settings, prevention of massive blood loss is the use of recombinant factor VIIa (rFVIIa), a hemostatic agent currently indicated for the treatment or prevention of bleeding in hemophilia patients with inhibitors. “Most practicing cardiac anesthesiologists have encountered the excessively bleeding patient who has literally drained the blood bank’s resources over a 24-hour period,” noted Dr. Patricia Murphy, Associate Professor of Anesthesiology, Toronto General Hospital, Ontario. Standard treatment of excessive surgical bleeding after cardiopulmonary bypass (CPB) includes measurement of coagulation parameters, surgical intervention, the use of systemic hemostatic agents and transfusion of blood and blood products but even with optimal treatment, a small proportion of patients who bleed excessively will continue to bleed, and with dire consequences, she added.

Numerous case reports have described the use of rFVIIa in the cardiac surgical setting, among them one of the larger reported series from Karkouti et al. (Transfusion 2005;45(1):26-34). Among the first 51 cardiac surgery patients who received rFVIIa for intractable blood loss in this series, blood loss and blood product usage were both significantly reduced after patients received 2.4 to 4.8 mg of rFVIIa. When adverse events (AEs) among rFVIIa patients were compared with 51 matched control patients who also experienced massive blood loss, event rates were similar in the two groups, except there was a higher incidence of acute renal dysfunction in the rFVIIa-treated group.

A later study also examined the unadjusted and risk-adjusted AE rate in 114 consecutive cardiac surgical patients who received rFVIIa for refractory excessive blood loss to that of 541 patients who also experienced excessive blood loss but who did not receive the therapy (Karkouti et al. Can J Anaesth 2006;53(8):802-9). The AEs measured in this analysis was a composite of death, stroke, renal failure, MI and major vein thrombosis. Prior to receiving rFVIIa, the median number of blood units required was eight. AE rates in those who were treated early (defined as those who received eight or fewer units of blood) and late (those who received more than eight units of blood) were 30% and 60%, respectively. The AE rate in those who did not receive rFVIIa was 24%. Using a risk-adjustment model that included total red blood cell units, pump time, weaning difficulty, gender, weight and age, the unadjusted AE odds ratio (OR) in the treated group was 2.41 vs. the untreated group, but once adjusted for these confounding variables, the adjusted AE OR was 1.04 for the treated vs. the untreated group, respectively. In the rFVIIa cohort, the adjusted AE OR was lower in the early treated group (OR 0.41, CI 0.18-0.92, P=0.03). Moreover, as Dr. Murphy pointed out, the adjusted AE OR was approximately 60% lower among patients who received rFVIIa early in their post-operative course compared to those who received it after significant blood loss had occurred.

These findings suggest that if rFVIIa is given as a last resort, it may not reduce the morbidity and mortality associated with massive blood loss, Dr. Murphy suggested. Conversely, “if you can predict ahead of time who is going to bleed massively, and if you give this drug early, then you may have a beneficial effect. This is one of the ways I think the drug will have good clinical application.” Another potential role for rFVIIa may be in patients who present for urgent cardiac surgery but who have high international normalized ratios (INRs) that need to be rapidly normalized. In a small series of 13 patients who had an INR in excess of 10, who had clinical bleeding or who required an interventional or surgical procedure, a small dose of between 15 and 20 µg/kg of rFVIIa was found to rapidly reverse the effects of warfarin and return INRs quickly to normal. Patients were then re-anticoagulated and in this series, the use of rFVIIa was not associated with an increased thrombotic risk.

These early observations suggest that there is “perhaps a role for rFVIIa in cardiac surgical patients who present for urgent or emergency surgery and who are anticoagulated with warfarin,” Dr. Murphy observed. Some of her colleagues have also given rFVIIa prophylactically to Jehovah’s Witness patients upon separation from CPB if they are deemed to be at increased risk of bleeding by virtue of undergoing a “redo” procedure, the presence of anticoagulation or a low platelet count prior to the operation, or who were on CPB for a long time.

“Having looked at the literature, I do not think rFVIIa should be routinely used for every excessively bleeding patient. You have to think about the risk factors a patient might have for thrombotic events including cerebrovascular disease, increased risk of pulmonary embolism or deep vein thrombosis,” Dr. Murphy advised. “So I think its use should be individualized to patients… but the future of rFVIIa as a universally accepted hemostatic agent is promising.”

As discussed by Dr. Linda Shore-Lesserson, Chief, Division of Cardiothoracic Anesthesiology, Montefiore Medical Center, Bronx, New York, early exploration of the use of novel antiplatelet glycoprotein IIb/IIIa antagonists suggests that platelets may be better protected after cardiac surgery when they are “paralyzed” during the surgery with the use of these agents, after which they return to normal function.

Summary

Questions and Answers

This question-and-answer session was conducted with Dr. Keyvan Karkouti, Assistant Professor of Anesthesia and Clinical Epidemiology, University of Toronto, Ontario.

Q: There is a theoretical risk of patients having a thrombotic event when given rFVIIa but you did not observe an excess of events in the rFVIIa-treated cardiac patients compared with matched controls in your series. How real do you feel the risk of thrombotic events is with rFVIIa in the cardiac setting?

A: The drug is a hemostatic agent so like any other drug that improves hemostasis, it has some thrombotic risk. The best estimates are in other patient groups where it does seem to have some thrombotic risk but it is not major. In extrapolating from other data, that risk is pretty reasonable in patients who are massively bleeding and whose complications from massive bleeding are several-fold higher than the thrombotic risk of the drug. In our studies where we adjusted for everything to see if there was an increased risk, we did not see any increased risk, but the samples were not that large, so there is still a potential for the agent to increase thrombotic risk. Nevertheless, all the signals are that it is a pretty reasonable risk to take, especially because the risk of massive blood loss is pretty high.

Q: At what point do you feel surgeons should intervene with rFVIIa when faced with massive blood loss?

A: The earlier you control bleeding, the better it is because the more blood lost, the more blood products patients require and the worse the outcomes. We have shown that by the time a patient loses one to two blood volumes (five to six units of red blood cells), that is the threshold when outcomes get markedly worse, so that would be the point of intervention. As well, with cardiac surgery, systemic tissue-factor expression seems to be time-related so if you wait several hours after surgery, in theory, you are more likely to have thrombotic complications. Our institutional guidelines say that first you go back to surgery, you correct coagulation abnormalities with blood products (and we use antifibrinolytic therapy in every patient); but once all of those things are done and patients are still bleeding, our guidelines indicate that you can consider using rFVIIa, although the risk:benefit profile must be carefully assessed for each individual patient.

Q: How great is the need for better hemostatic agents in the cardiac surgical setting?

A: The need is great. We looked at data from seven Canadian hospitals and about 10 to 15% of patients have enough blood loss to require five units or more of blood and we know that significant blood loss increases risk [of complications] and that these patients do a lot worse than those who do not get blood. And while it is only an observation, it seems that since we started using rFVIIa, we do not get as many patients bleeding massively in the ICU, whereas before, patients either died in the ICU or they ended up in the ICU for a very long time because of complications of blood loss. Things seems to have changed for the better, and at least we can say that the evidence is becoming stronger that rFVIIa is beneficial if properly used, with the caveat that we still do not know its full risk profile.

Note: At the time of printing, in Canada, recombinant factor VIIa is indicated in hemophilia A/B patients with inhibitors to FVIII or FIX for the treatment of bleeding episodes.