SEVERAL investigators have reported the astonishing effect of recombinant activated factor VII (rFVIIa) in trauma patients with diffuse bleeding. Currently, rFVIIa is approved for the treatment of patients with hemophilia with inhibitors to factors VIII and IX. Conditions with increased thromboembolic risk, including trauma, extensive tissue damage, sepsis, arteriosclerosis, and disseminated intravascular coagulation, may be considered contraindications for the drug. Thrombotic complications in trauma patients are rarely observed. To our knowledge, this is the first report of a patient who experienced a cerebral sinus thrombosis in the posttraumatic period after rFVIIa administration.
A 19-yr-old man was involved in a frontal motorbike accident. He had an open shaft fracture of the femur, pneumothorax, and lung contusions. A small frontal brain contusion was only visible on a follow-up computed tomographic (CT) scan, without any initial neurologic impairment.
After the pneumothorax was sufficiently drained, the patient underwent surgical reposition of his thigh. Stabilization was achieved with external skeletal fixation. The intense bleeding was under control after reposition and intermittent tamponade. No damage of major vessels was found. An extensive supracondylar hematoma was drained.
Twelve hours after stabilization of the fracture, the patient showed signs of bleeding. Before and during surgical revision of the thigh, he received 16 units of erythrocytes, 11 fresh frozen plasmas, 8 single-donor platelet concentrates, and a single bolus dose of 240 KU (60 μg/kg body weight) rFVIIa. Again, no damage of major vessels was found. Bleeding stopped promptly after administration of rFVIIa. During the operation, an intramuscular hematoma was drained, and fasciotomy was performed on the thigh and lower leg. After this, the patient was stable, with a hemoglobin concentration at 7.5–8 mg/dl.
Because of the patient’s pulmonary impairment, it was difficult to wean him off the respirator. After return to spontaneous ventilation, he showed changing levels of consciousness. A follow-up CT scan of the brain was performed at day 5, which showed a small frontal contusion without any space-occupying character (fig. 1A).
In the second week, the patient started to present signs of infection of the fracture site. His body temperature peaked on day 13 up to 39.7°C and decreased immediately after surgical revision and débridement. The leukocyte count at that time was 15.1 × 109/l, and C-reactive protein, with a maximum of 236 mg/l on day 9, had decreased to 100 mg/l on day 15. Reactive thrombocytosis occurred, with a maximum of 1,080 × 109/l on day 19.
On day 15, the patient presented with a unilateral dilated pupil with remaining prompt reaction to light, spontaneous inward rotation, extensions of the upper and lower limbs, and vomiting despite drained gastric tube. Native and contrast-enhanced CT scans were performed to rule out septic dissemination and did not show any pathologic findings (fig. 1B). Bacterial meningoencephalitis was ruled out by lumbar puncture. A possible systemic herpes simplex infection after the labial lesions was treated prophylactically with acyclovir until viral genome testing results were found to be negative by polymerase chain reaction.
After another 12 h, the patient again showed signs of impaired brainstem function. A second CT scan showed slightly dilated lateral ventricles. After the insertion of ventricular drainage, intracranial pressure did not increase.
Because of the persistent impaired consciousness, another follow-up CT scan was performed on day 21, and hypodense areas in the anterior parts of both thalami were found (fig. 1C). Bilateral thalamic edema after thrombosis of internal cerebral veins was considered to be the most likely diagnosis, but to visualize the thrombus and to rule out other causes of thalamic hypodensities, such as ischemic infarction, a magnetic resonance imaging study was performed on the same day. The corresponding areas of both thalami showed signs of edema (figs. 2A–C) but no contrast enhancement. Signal intensity in diffusion-weighted imaging was slightly increased, probably because of T2 sensitivity of the sequence, but not to the extent typically seen in ischemic infarction. There were no signs of intracerebral hemorrhage.
The thrombus itself was found in the proximal part of the straight sinus, obstructing the confluence of the inferior sagittal sinus and the great cerebral vein of Galen. It was surrounded by a narrow stripe of contrast media (fig. 3A). Phase contrast magnetic resonance angiography showed a remaining flow signal in the area of the thrombus (fig. 3B).
To achieve differentiation of the thrombus against a venous anomaly, conventional angiography was performed on the following day. In late venous images, sustained contrast was detected in the great cerebral vein of Galen, the internal cerebral veins, and the thalamostriate veins (fig. 3C), making venous congestion the most probable cause for thalamic edema.
Intravenous anticoagulant therapy was initiated, and within 3 weeks, the neurologic situation improved gradually (figs. 2B and C).
In several case reports 1,2and small studies, 3antihemorrhagic effects of recombinant activated factor VII (rFVIIa) has been described in diffuse hemorrhage of trauma and surgical patients. Massive transfusion dilutes coagulation factors and impairs platelet number and function. Excessive treatment with fluids such as hydroxyethyl starch preparations might directly compromise coagulation. Concomitant hypothermia causes slowing of enzymatic activities of the coagulation cascade and dysfunction of platelets. Release of procoagulant substances from ruptured tissues leads to a complex consumptive coagulopathy with enhanced fibrinolysis. Metabolic abnormalities, such as acidosis and hypocalcemia, further deteriorate coagulation.
Under these circumstances, diffuse bleeding often persists, despite apparently adequate surgical procedures and treatment with blood products and conventional hemostatic agents. In single cases 1and small study groups, 2rFVIIa was successfully given to achieve hemostasis in severe bleeding, without previous coagulopathy. Thromboembolic complications have been described, such as venous thrombosis, 4myocardial infarction, 5and disseminated intravascular coagulation. 6,7The described thrombotic events often have occurred in the context of preexisting risk factors, e.g. , crush injury, tissue necrosis, septicemia, atherosclerosis, or previous administration of activated prothrombin complex, making it difficult to pinpoint the contribution of rFVIIa to these adverse events.
In our patient, obvious clinical signs of his cerebral sinus thrombosis became evident 14 days after administration of rFVIIa, when the patient presented with a unilateral dilated pupil and spontaneous inward rotation. With a half-life of rFVIIa of approximately 6 h, a direct relation seems rather unlikely. However, because cerebral sinus thrombosis in the age group of our patient and with the presented trauma pattern is rare and extensive clinical experience with rFVIIa is lacking, we are considering a relation to the administration of rFVIIa, in contribution with an endothelial lesion of the cerebral sinus caused by the mild brain injury. With the exposure of subendothelium after the small blood vessel injury, cell-bound tissue factor may be exposed and may cause formation of microthrombus. Without enhancement of the coagulation by rFVIIa, the microthrombus might have resolved quickly, without any clinical recognition. With the administration of rFVIIa and the generation of a high thrombin burst, a more stable clot was probably created, not resolving until day 14 when, under septic conditions, the coagulation became activated. When toward day 14 the patient’s septic situation aggravated, the thrombus may have increased in volume, or smaller thrombi may have occluded the remaining drainage beside the initial thrombus, compromising the venous drainage of the thalamic region.
A similar delay between rFVIIa administration and thrombosis formation was seen by Van der Planken et al. 4in a hemophilia A patient with inhibitors and severe infectious disease. This patient experienced a distal deep venous thrombosis 18 days after rFVIIa transfusion. Also, d’Oiron et al. 6described a pulmonary embolism of the lung 5 days after discontinuation of rFVIIa infusion.
Despite the promising beneficial effect of rFVIIa for hemostasis in massively bleeding trauma patients, the thrombotic risk of the drug must be kept in mind. Therefore, close monitoring is necessary for early identification of thrombotic complications.