Portal de Periódicos da CAPES

Recombinant activated factor VII to control massive post operative bleeding after septic aortobifemoral grafting

2005; Lippincott Williams & Wilkins; Volume: 22; Issue: 10 Linguagem: Inglês

10.1017/s026502150526132x

1365-2346

Mathieu Raux, L. Chiche, E. Vanhille, Bruno Riou,

Trauma, Hemostasis, Coagulopathy, Resuscitation

EDITOR: Recombinant activated factor VII (rFVIIa, NovoSeven®; Novo Nordisk, Copenhagen, Denmark) is mainly used as a pro-haemostatic agent for the treatment of haemophiliac patients with inhibitors [1]. However, it has also proven its clinical effectiveness in patients with Glanzmann thrombasthaenia and thrombocytopaenia [1]. Moreover, rFVIIa has been recently used in patients without pre-existing coagulopathy to control perioperative bleeding in various clinical situations such as trauma [2], cardiac surgery, vascular surgery, hepatectomy, gynaecological surgery or left ventricular assistance. A 56-yr-old man with a prior history of two aortobifemoral grafts for acute abdominal aortic thrombosis, complicated by postoperative acute severe pancreatitis, developed a prosthetoenteric fistula 7 months after an episode of cholangitis and was scheduled for emergency venous autograft replacement. He was treated receiving aspirin and clopidogrel which could not be withdrawn because of the emergency conditions. He had normal platelet count and coagulation screen at the time of admission. The Dacron graft was exposed through a left thoraco-phreno-laparotomy. Surrounding pus was sampled for bacteriological analysis. The graft was completely removed under supra-coeliac aortic cross-clamping and replaced by a bifurcated aortograft fashioned with both superficial femoral veins. An aortic tear occurred during aortic control because of sudden arterial hypertension, requiring a 4 cm thoracic aortic Dacron graft replacement. Aortic clamp time was 25 min at the supra-coeliac level and 50 min at the supra-renal level. During reperfusion, non-surgical bleeding developed requiring administration of 27 units of red blood cells, 10 units of platelets, 18 units of fresh frozen plasma and aprotinin (2 million kallikrein inhibiting units). The immediate postoperative period in the intensive care unit (ICU) was marked by continuous bleeding (>1800 mL h−1) through two of the drains and acute haemostasis disorders were observed (prothrombin ratio 20%, fibrinogen 0.7g L−1, platelets count 32 × 109 L−1, activated partial thromboplastin ratio 5.3). Despite administration of 18 units of blood, 24 units of fresh frozen plasma, 21 units of platelets, 3 g of fibrinogen, continuation of intravenous (i.v.) aprotinin administration and the absence of hypothermia (35.5°C), bleeding was still active and justified a second surgical look 5 h later. Active bleeding from a diaphragmatic vein, a pulmonary wound and the lower aortic graft anastomosis were surgically corrected. Perioperative blood loss required massive transfusion (13 units of blood, 12 units of fresh frozen plasma and 10 units of platelets). In the second postoperative period, the patient remained unstable justifying continuous haemodynamic support with norepinephrine, continuation of aprotinin infusion and transfusion (a further 8 units of blood, 10 of frech frozen plasma and 10 of platelets). Despite in vitro haemostasis tests being in the normal range, postoperative bleeding through the drains was still >500 mL h−1. The administration of rFVIIa (90 μg kg−1 followed 2 h later by a second dose 45 μg kg−1) was associated to an immediate termination of bleeding. Only two more units of blood were administered 10 h later due to anaemia and severe hypoxaemia. There were no further episodes of bleeding, and no significant side-effects were observed. Further progress in the ICU was marked by the occurrence of a collection near the aortobifemoral graft within a month, associated with sepsis due to Enterococcus gallinarium. Despite targeted antibiotics (vancomycin and gentamycin), the patient remained septic. Surgical excision of the abscess together with an allograft replacement of the thoracic prosthesis was performed at day 34 under hypothermic cardiopulmonary bypass. Preoperative haemostasis was normal. Aprotinin was administrated to prevent fibrinolysis. Thoracic aortic cross-clamping lasted 23 min at 25°C. Diffuse non-surgical bleeding developed after weaning off cardiopulmonary bypass, despite heparin reversal by protamine. It required the administration of 13 units of blood, 5 of fresh frozen plasma and 10 of platelets before the incision could be closed. During the immediate postoperative period, the patient remained unstable with active bleeding through the drains requiring further transfusion (5 units of blood, 4 of fresh frozen plasma, 10 of platelets and 3 g of fibrinogen), administration of aprotinin and continuous haemodynamic support with norepinephrine (up to 12 mg h−1). Despite normal in vitro haemostasis tests and normothermia, bleeding through the drains was >500 mL h−1 justifying the administration of a 90 μg kg−1 of rFVIIa. This was followed by the termination of bleeding. Thereafter, norepinephrine requirements markedly decreased (3 mg h−1), and no further transfusion was required after 2 h. Clinical and biological follow up did not highlight any side-effect. The patient's condition gradually improved and he was later discharged for rehabilitation. Analysis of these events shows that rFVIIa was effective in controlling massive postoperative bleeding in a surgical patient with impaired haemostasis due to both sepsis and to uninterrupted preoperative antiplatelet therapies. Several factors may explain the coagulopathy observed and thus the dramatic blood loss and massive transfusion required in this patient. First, active perioperative bleeding and coagulation activation due to tissue factor release at the site of vascular injury is markedly increased in cases of repeat surgery and/or sepsis [3]. Second, despite normal preoperative haemostasis tests, aspirin and clopidogrel induced important platelet dysfunction, while these drugs could not be withdrawn due to the emergency conditions. Both aspirin and clopidogrel therapies were routinely prescribed to the patient by his general practitioner because of coronary disease. Third, supra-coeliac aortic cross-clamping is associated with increased bleeding due to increased fibrinolysis secondary to splanchnic ischaemia, and due to heparin administration [4]. Our patient experienced massive postoperative diffuse bleeding inaccessible for surgical haemostasis, despite warming, transfusions (according to blood loss, in vitro haemostasis results and French recommendations), antifibrinolytic (aprotinin) therapy and appropriate calcium replacement. Particular attention must be paid to routine preoperative antiplatelet therapy. First, they justified platelet transfusion (12 units) prior to the first surgery. Second, serum clearance of these therapies was dramatically increased due to massive bleeding. The administration of rFVIIa was associated twice with rapid termination of postoperative bleeding. At these therapeutic doses, plasma level of rFVIIa is about 100 times higher than endogenous FVIIa level [5]. rFVIIa binds with tissue factor at the site of injury and with activated platelets, leading ultimately to a burst in thrombin formation [1]. This case confirms the previous reports of the efficacy of rFVIIa as a rescue therapy in several situations of severe postoperative bleeding. No thromboembolic complication was observed in this patient. However, we think that these events illustrate some important recommendations that have been recently emphasized concerning the use of rFVIIa in these off-label conditions [6]. First, rFVIIa is only an adjunctive haemostatic therapy that should not replace surgery and or embolization when they are required. In our patient, surgical bleeding required a second look and it is not likely that rFVIIa could have solved this problem. Second, rFVIIa should not replace the correction of the various factors that are known to interfere with haemostasis: platelet, coagulation factors, fibrinogen, red cells, calcium and temperature [1,6]. Nevertheless one can wonder whether an earlier administration of rFVIIa should have been able to decrease the amount of transfusion required in this patient. rFVIIa is a costly drug but the cost of transfusion was greater in our patient. We administered a first dose of 90 μg kg−1 of rFVIIa which is the recommended dose in France for off-label use in severe bleeding. In other reports of perioperative bleeding or trauma, doses between 60 and 144 μg kg−1 have been used [2,7]. However, in severe trauma patients, a higher dose has been recently examined in a randomized study. In our patient, the dose used appeared to be sufficient since bleeding stopped but it should be emphasized that the rate of bleeding was not very high. In patients with higher rate of bleeding, one can suspect that the clearance of rFVIIa is artificially increased by blood loss, thus requiring administration of higher doses. This important point warrant further studies. In conclusion, our patient experienced massive bleedings after repeat septic vascular abdominal aortic surgeries. The bleeding was twice successfully controlled using i.v. rFVIIa. Anaesthesiologists should be aware that rFVIIa may represent an efficient therapy if conventional therapy failed to control bleeding in such clinical situations. M. Raux L. Chiche E. Vanhille B. Riou 1Department of Anesthesiology and Critical Care, Centre Hospitalo-Universitaire Pitié-Salpêtrière, Assistance Publique-Hôpitaux de Paris, Université Pierre et Marie Curie, Paris, France 2Department of Vascular Surgery, Centre Hospitalo-Universitaire Pitié-Salpêtrière, Assistance Publique-Hôpitaux de Paris, Université Pierre et Marie Curie, Paris, France 3Department of Anesthesiology and Critical Care, Centre Hospitalo-Universitaire Pitié-Salpêtrière, Assistance Publique-Hôpitaux de Paris, Université Pierre et Marie Curie, Paris, France 4Department of Emergency Medicine and Surgery, Centre Hospitalo-Universitaire Pitié-Salpêtrière, Assistance Publique-Hôpitaux de Paris, Université Pierre et Marie Curie, Paris, France