Unusual redo mitral valve replacement for bleeding in Glanzmann thrombasthenia

Introduction

Glanzmann thrombasthenia (GT) is an autosomal recessive congenital or acquired hemorrhagic disorder described initially by Edward Glanzmann in 1918. ¹ The defect is characterized by either quantitative and/or qualitative abnormalities of the platelet membrane glycoprotein IIb/IIIa receptor for fibrinogen, resulting in abnormal platelet aggregation. In GT, inter-platelet interaction is defective and final thrombus production is abnormal. Due to the bleeding tendency, it is quite unexpected that a thrombotic event would develop in patients with GT. However, there are several reports showing that a potent thrombotic process can evolve.^2,3 Recommendations for long-term anticoagulation in GT are still lacking. ⁴ It has been suggested that warfarin therapy maintaining the international normalized ratio (INR) between 1.5 and 2.0 can be safely used to prevent thrombosis in patients with GT. ³ We describe the case of a patient with mitral valve (MV) regurgitation and GT who underwent mechanical MV replacement. Warfarin therapy caused devastating bleeding. Consequently, we had to replace the mechanical valve with a biological prosthesis to solve the problem of long-term anticoagulant therapy. To our knowledge, this is the first report of MV replacement in GT, highlighting the danger of oral anticoagulation in this pathology.

Case report

A 23-year-old man was admitted to our hospital with a diagnosis of MV regurgitation secondary to subacute bacterial infective endocarditis caused by Streptococcus viridans. The diagnosis of GT was previously known in this patient with past history of petechiae, epistaxis, and gingival bleeding in early childhood. A complete blood count including platelets (235,000/µL) and morphology were normal. Routine coagulation screening tests and measurements of coagulation factors and von Willebrand factor were within the reference range. Bleeding time with the Ivy method was 48 min (normal range 4–11 min). A platelet aggregation test showed abnormal responses to adenosine diphosphate, collagen, and epinephrine, but normal agglutination with ristocetin. Unfortunately, we do not have flow cytometry and specific markers to determine the type of GT. Because MV repair was not feasible, a 29-mm St. Jude Medical mechanical prosthesis (St. Jude Medical, Inc., St. Paul, MN, USA) was inserted in the mitral position. Multiple platelet transfusions were required in the operating room and continued for a week. Despite 2000 mL of postoperative bleeding in the first 24 h, the postoperative course in the intensive care unit was uneventful. Oral anticoagulant therapy with 5 mg of warfarin per day was initiated on the 7th postoperative day; the INR was 1.8. Immediate devastating complications including bilateral hemothorax, pericardial effusion, epistaxis, persistent gastrointestinal bleeding, and hematuria were simultaneously observed. A lower dose of warfarin with 2.5 mg per day was attempted with an INR of 1.6, but gastrointestinal bleeding and hematuria were extremely persistent. A decision was made to replace the mechanical valve with a biological prosthesis. The patient underwent a second MV replacement 2 weeks after the first operation. A 29-mm St. Jude Medical Epic biological mitral prosthesis was installed. Screening for human leukocyte antigen and platelet antibodies was not available. Recombinant activated factor VII was administered in a single dose of 60 µg kg⁻¹ after a limited transfusion of platelets, fresh frozen plasma, cryoprecipitate, and red cell concentrates immediately after weaning from cardiopulmonary bypass. The second postoperative course was uneventful. Postoperative bleeding was 500 mL in the first 24 h. No thrombi were found inside the heart on an echocardiographic study prior to hospital discharge on the 10th day after repeat MV replacement. No need for oral anticoagulation has been observed in this case.

Discussion

The primary objective of this report is to show how MV replacement must be faced in patients with GT. This is a hematological disorder due to the quantity or quality of glycoprotein IIb/IIIa integrin on the platelet surface, resulting in poor platelet aggregation. Glycoprotein IIb/IIIa inhibitors inducing a transient GT-like situation used in percutaneous coronary interventions have been developed from observations of GT. ⁵ Since GT represents a bleeding tendency, it is quite unexpected that a thrombotic event would develop in these patients. However, there are several reports showing that a potent thrombotic process can evolve.^2,3 There is also evidence that prosthetic materials can directly activate coagulation factor XII and induce thrombus formation independent of platelets. Moreover, it has been suggested that warfarin therapy maintaining the INR between 1.5 and 2.0 can be safely used to prevent thrombosis in patients with GT. ²

Our case exhibited life-threatening major bleeding with an INR of 1.6–1.8. Normally, the risk of major bleeding is 0.6–1.4 per 100 patient-years on anticoagulant therapy with INR of 1.5 to 3.0.^6,7 Although the interaction between GT and warfarin is not yet clear, the risk of oral anticoagulant therapy in patients with GT remains unrecognized. These anticoagulants block the hepatic enzyme vitamin K epoxide reductase, reducing the liver’s ability to recycle vitamin K. Vitamin K serves as a cofactor in the addition of hydroxyl groups to specific glutamic acid residues. The addition of hydroxyl groups allows these specific residues to bind calcium, which is critical for the formation of salt bridges. These bridges allow the formation of multi-molecular complexes that facilitate the rapid activation of factor X and prothrombin. When warfarin is administered, clotting factors II, VII, IX, and X lack the modified glutamic acid residues and consequently, thrombin generation is delayed and reduced. Thrombin is the most powerful agonist for platelet aggregation. ⁸ In view of the patient’s age, our first choice was MV repair. Unfortunately, this was not possible because of severe valvular destruction. So, in accordance with our hospital policies for this age group, we decided to install a mechanical prosthesis. Complications arising from warfarin therapy were multiple and hard to treat. It is highly likely that warfarin played a large part in the major bleeding because thrombin is implicated in both the coagulation cascade as well as poor platelet aggregation in GT. Two weeks later, the patient had to return to the operating room and the mechanical valve was replaced with a biological prosthesis to avoid any anticoagulant therapy.

The binomial condition of MV surgery with GT remains a challenging situation in cardiac surgery, particularly when a mitral prosthesis needs to be implanted in a young patient. MV reconstruction must be considered the first choice; if that is not possible, the second option should be a biological prosthesis to avoid any long-term anticoagulation. We strongly recommend avoiding the use of warfarin in patients with GT.