In an article published in 2010,1 Plessier et al. investigated 102 patients with acute thrombosis of the portal vein unrelated to cirrhosis or malignancy. The authors found that the formation of thrombosis could be favored by at least one general risk factor and local factors in 52% and 21% of cases, respectively. Although their investigations were exhaustive, one factor was overlooked and deserves specific comment. We recently found the presence of antiannexin V (aANV) antibodies in a 53-year-old man suffering from portal hypertension unrelated to cirrhosis. Our patient had a history of both right sural deep vein thrombosis following an immobilization period and right saphenous paraphlebitis. In 2010 he experienced an episode of acute esophageal variceal hemorrhage. Magnetic resonance imaging (MRI) revealed a segmental thrombosis of the superior mesenteric vein and a portal cavernoma associated with splenomegaly but without evident cirrhosis. No local risk factors and no acquired or inherited thrombotic disorders were initially found through routine screening. Extending our research, we found an antiphospholipid syndrome (APS) with a high titer of aANV immunoglobulin G (IgG) (89 AU; N <8 AU) and a nonsignificant titer IgM (4.76 AU; N <9 AU) (Elisis, Biomedical Diagnostics, Marne la Vallée, France). A liver biopsy sample containing 20 portal tracts with reticulin stain demonstrated normal architecture. One year later, the patient remained safe under vitamin K antagonists and beta-blockers but aANV IgG were still strongly positive (32 AU).
Annexin V acts as an anticoagulant that competes with prothrombin for phospholipid binding sites, and prothrombic effects of aANV antibodies are related to APS.2 Antibodies to annexin V have been identified in association with various pathological conditions, such as fetal loss, and venous and/or arterial thrombosis in patients with systemic lupus erythematosus, systemic sclerosis, and rheumatoid arthritis.3-5 However, aANV antibodies are not routinely tested, unless specifically requested. In patients with unexplained portal thrombosis this test should be carried out because this condition is generally reported in association with prothrombic and procoagulant states. Further studies investigating the prevalence of aANV antibodies and its role in the pathogenesis of portal vein thrombosis are warranted.