|Home | About | Journals | Submit | Contact Us | Français|
In a patient taking prophylactic anticoagulants after valve replacement, haemorrhage presents a therapeutic dilemma.
A man aged 77, previously independent, developed a right hemiparesis with expressive dysphasia. He was right-handed. 2 years earlier he had undergone prosthetic aortic valve (bileaflet) replacement together with coronary artery bypass grafting, since when he had been taking warfarin with a target international normalized ratio (INR) of 2.5–3.5. On examination he was in sinus rhythm and normotensive, Glasgow Coma Scale 13/15. His dysphasia, sensory and visual inattention, and right hemiplegia were compatible with a left total anterior circulation stroke syndrome, and CT revealed a large left frontoparietal primary intracerebral haemorrhage (Figure 1). His INR was 3.2.
Clearly, this patient was at risk of further bleeding if the warfarin was continued but at risk of thromboembolism from the heart valve if it was stopped. We reversed the anticoagulation rapidly with Beriplex—a prothrombin complex concentrate. A neurosurgeon recommended conservative management and he was treated on a multidisciplinary stroke unit. Initially he made good progress, with some return of function, but on day 11 he deteriorated suddenly with development of a right total anterior circulation stroke syndrome. On this occasion CT showed a right middle cerebral artery infarct (Figure 2), to which he succumbed.
We judged the initial haemorrhage life-threatening, and in these circumstances the British Society for Haematology recommends reversal of anticoagulation. Conventional treatment was not used since with vitamin K the response is slow and unpredictable and with fresh frozen plasma it is rapid but short-lasting. Beriplex has been available for about 10 years but only recently have its efficacy and safety profiles been published; it can stop bleeding immediately without acute risk of thromboembolism.1,2
Should we have started an alternative antithrombotic/anticoagulant (AT/AC)? A meta-analysis of over 13 000 patients3 indicates that, in patients with prosthetic heart valves, the risk of thromboembolism is 4% per year without AT/AC, 2% per year with AT and 1% per year with AC. Prosthetic mitral valves carry twice the thromboembolism risk of aortic valves (4% v 2%), and ball cage valves are more hazardous than tilting disc valves. In anticoagulated patients with metallic heart valves Cannageiter et al.4 reported an incidence of intracranial bleeding of 0.5% per year and of any significant bleed 2.7% per year.
A review of the published work yields little information on management of a patient with a prosthetic heart valve and a major bleed—the time for which warfarin can safely be withheld, the subsequent risk of thromboembolism, the risk of rebleeding or resumption. Karthik et al.5 suggest that warfarin can be withheld for two or three weeks with little risk of thromboembolism, though in their retrospective series there was some rebleeding (gastrointestinal) when anticoagulation was restarted.
We had planned to reintroduce anticoagulation after 14 days but unfortunately our patient had his thromboembolism on day 11.