|Home | About | Journals | Submit | Contact Us | Français|
Recombinant factor VIIa has been licensed in the United Kingdom since 1996 for the control of bleeding in hemophilic patients who are actively bleeding or are about to undergo surgery. Medical practitioners are also applying recombinant factor VIIa toward the control of bleeding in patients without hemophilia. Although the efficacy of recombinant factor VIIa has been shown in many patients, concerns have arisen about the risk of thrombotic adverse events.
Herein, we report the case of a 73-year-old woman who underwent major coronary surgery and developed fatal systemic venous thrombosis after recombinant factor VIIa was used in an attempt to control her severe postoperative bleeding. We review the medical literature and discuss the risks of using recombinant factor VIIa to control severe bleeding after cardiac surgery.
Recombinant human coagulation factor VIIa (rFVIIa) has been licensed in the United Kingdom since 1996 to control bleeding in patients who have hemophilia and are actively bleeding or are about to undergo surgery. Medical practitioners have expanded the application of rFVIIa toward the control of postoperative bleeding. In such off-label use, the efficacy of rFVIIa has been shown in many patients. However, safety concerns have arisen, particularly concerning the risk of thrombotic adverse events, when rFVIIa has been used in this manner.
We report the case of a 73-year-old woman who underwent elective coronary artery bypass grafting (CABG) and was given rFVIIa in an attempt to control severe postoperative bleeding. We review the medical literature and discuss the off-label application of rFVIIa to the control of severe bleeding that may occur after major cardiac surgery.
In January 2006, a 73-year-old woman (body weight, 64 kg) was admitted to our hospital for elective CABG. Her medical history included hypertension, hypercholesterolemia, and a transient ischemic attack 10 years earlier. Preoperative carotid duplex ultrasonography revealed bilateral calcified plaques in the bulbs and bifurcations of the common carotid arteries, without significant stenosis. All preoperative blood results, including the clotting profile (Table I), were normal.
After inducing anesthesia, we cannulated the right internal jugular vein with an 8.5F quadruple-lumen cannula and an 8.5F catheter introducer in the pulmonary artery. The patient received 3 grafts, as planned; the duration of cardiopulmonary bypass (CPB) was 86 min. As the patient was being weaned from CPB, the proximal anastomosis of one of the vein grafts was disrupted, which led to a dissection of the ascending aorta. Therefore, CPB was reinstituted with deep hypothermia (18 °C), followed by circulatory arrest for 14 minutes, in order to replace the ascending aorta with a Dacron graft. The duration of the 2nd run of CPB was 180 min.
After CPB was discontinued, severe bleeding from all suture lines of the Dacron graft began, and it continued despite all attempts at surgical hemostasis. The coagulopathy was initially treated with 6 units of fresh frozen plasma, 3 units of platelets, and then with 10 units of cryoprecipitate. Nevertheless, the bleeding continued. Approximately 2 hours after the discontinuation of CPB, 90 μg/kg of rFVIIa was administered, divided into 2 doses, 20 minutes apart. Table I shows the patient's clot-ting profile before and after rFVIIa administration. After another 90 minutes, the bleeding was controlled; the patient received an intra-aortic balloon pump and was transferred to the intensive therapy unit. Her chest was left open after surgery by stenting with a retractor. The total chest-tube drainage was 450 mL over the next 16 hours, at which time the chest was closed.
The patient remained intubated and on mechanical ventilation for 9 days, during which time she required substantial cardiovascular and respiratory support. Shortly after extubation, she developed pyrexia and was started empirically on broad-spectrum antibiotics; an attempt was then made to change the central line.
Three attempts were made to cannulate the left internal jugular vein under ultrasonographic guidance. In each instance, the vein and needle were clearly visible, but the wire failed to move into the venous lumen. Therefore, the femoral vein was cannulated instead. The next day, the right subclavian vein was cannulated under ultrasonographic guidance, and ultrasonographic examination of the neck vessels was performed.
Extensive thrombi were found in the left internal jugular vein (Fig. 1), the right internal jugular vein (Fig. 2), and the left subclavian veins. Although some flow was seen in the right subclavian vein, a nonocclusive thrombus could not be ruled out, due to the presence of some flow beyond the thrombus. The patient was immediately started on intravenous heparin. Her res-piratory function had deteriorated by this time, and she was reventilated. Shortly afterwards, she developed a distended and tender abdomen that necessitated an exploratory laparotomy. She was found to have an ischemic perforation in the transverse colon, with necrotic areas in the caecum and ascending colon. The vessels were thrombosed, providing morphologic evidence of mesenteric thrombosis. The patient underwent a subtotal colectomy with an end ileostomy. During the next 2 weeks, she underwent 2 more laparotomies with further small-bowel resection. Despite these efforts, the patient's condition continued to deteriorate. She developed multiple-organ failure and died after 30 days in the intensive therapy unit.
It may reassure the caring physician to think that rFVIIa can be used with efficacy to control postsurgical bleeding in the absence of hemophilia. However, it is prudent to consider that no good data establish the safety of such an off-label application.
Recombinant factor VIIa is thought to act locally at the site of tissue injury and vascular-wall disruption by binding to exposed tissue factor and generating small amounts of thrombin that are sufficient to activate platelets. The activated platelet surface can then form a template on which rFVIIa can directly or indirectly mediate further activation of coagulation, resulting in the generation of much more thrombin, and, ultimately, in fibrinogen-to-fibrin conversion.1,2
Data that support the efficacy of rFVIIa as a hemostatic agent in bleeding non-hemophiliac patients have been available since 1999, when the 1st descriptions appeared in the medical literature. Initially, these were case reports of coagulopathic trauma patients3 and patients who had developed coagulopathy after cardiac surgery.4 The authors of these reports expressed concerns about the possibility of thrombotic complications but did not highlight venous thrombosis as a problem.
The efficacy of rFVIIa as a prophylactic agent to decrease bleeding in the non-coagulopathic patient has been tested in 3 randomized, controlled surgical trials.5–7
When rFVIIa was used to decrease bleeding after majorliver resection, there was no significant reduction in either the number of patients transfused or the volume of blood products administered. The authors reported an approximate 5% incidence of thromboembolic complications in both the placebo and treatment groups.5 When rFVIIa was used to decrease bleeding after retropubic prostatectomy, there did appear to bea significant beneficial effect: perioperative blood loss in the treatment group was approximately half that of the 2,500 mL in the placebo group.6 Conversely, in delayed pelvic reconstruction after traumatic fracture, rFVIIa was not shown to decrease bleeding or reduce the need for blood transfusion.7 In both of the last 2 studies,6,7 there was no detectable thrombosis.
In other studies, rFVIIa was shown to prevent the extension of intracerebral hemorrhage and to improve short-term functional outcome, despite slightly more frequent thromboembolic adverse events.8 The use of rFVIIa also enabled better control of variceal bleeding in cirrhotic patients.9
Several researchers have speculated on the reasons for the lack of efficacy of rFVIIa in cases of postoperative bleeding. Suggestions included the timing of the drug's administration, and the possibility that rFVIIa cannot improve local blood coagulation unless coagulation is already deranged. It is not recommended that rFVIIa be used as a prophylactic agent in patients who are undergoing surgery, when there is a high risk of bleeding.10
Reports from 2 recent randomized, controlled studies11,12 discuss the use of rFVIIa in patients who had already developed coagulopathy. In the 1st study,11 the use of rFVIIa after complex cardiac surgery was found to reduce the need for red-cell and coagulation-product transfusion. However, 1 stroke and 1 myocardial infarction occurred in the treatment group—an incidence that was not significantly different from adverse events noted in the control group. Although the researchers acknowledged the small size and insufficient power of their study, they concluded that their use of rFVIIa reduced the need for transfusion and led to no significant increase in the number of adverse events. The 2nd study12 showed that, after blunt trauma, the use of rFVIIa significantly reduced the need for red-cell transfusion and for massive transfusion. However, the authors detected a 3% incidence of thromboembolic complications in their treatment group.
In a case-matched study, Karkouti and colleagues13 used propensity scoring to determine the efficacy and safety of rFVIIa as rescue therapy after cardiac surgery. Despite the cautiously optimistic conclusions suggested by those authors, we believe that the efficacy of rFVIIa in that study remains questionable, in view of the significantly greater initial rate of bleeding in the treatment group.
The authors of several case series4,14,15 have reported varying details of thromboembolic complications without highlighting the complications as a problem. In contrast, a recently published 16-patient case series16 revealed a 25% incidence of thromboembolic complications. Most recently, O'Connell and colleagues17 studied the incidence of thromboembolic adverse events in 168 reports, noted that the mortality rate in those reports was high (30%), and found that in 36 of 50 reported deaths (72%), the thromboembolic event was the probable cause of death. In concluding that most of the reports of thromboembolic adverse events were due to the off-label use of rFVIIa, with its concomitant frequent risk of death or severe morbidity, the researchers called for randomized controlled trials of rFVIIa in this application.
The case of our patient shows that the use of rFVIIa in an attempt to control excessive bleeding after coronary surgery—particularly in the absence of established indications and dosing regimens—can lead to serious morbidity and to death. Randomized clinical trials are warranted in order to establish the safety and efficacy of rFVIIa as prophylactic treatment and as therapy when severe bleeding occurs in association with major cardiac surgery.
Address for reprints: Sharif Al-Ruzzeh, PhD, Leeds General Infirmary, 18 Fielding Way, Leeds LS27 9AB, United Kingdom. E-mail: ku.oc.oohay@hezzurlafirahs