Anticardiolipin antibodies in patients with liver disease

OBJECTIVE: Our aim was to test the hypothesis that anticardiolipin antibodi es (aCL) may cause an antiphospholipid syndrome and thrombotic events in pa tients with liver disease. METHODS: aCL were measured in 116 healthy controls and 372 patients with li ver disease of different stage and etiology: 136 cases secondary to hepatit is C virus (HCV) infection, 139 due to hepatitis B virus (HBV) infection, 6 9) with alcoholic liver damage, and 28 cryptogenic in origin. Prior thrombo tic events were recorded. The results were related to age, gender, stage, s everity, and etiology of the liver disease, as well as to the occurrence of organ- and nonorgan-specific autoantibodies. RESULTS: aCL were positive in 4.4% of controls and in 18.8% of patients (p < 0.0002). Patients with aCL were more frequently men with an advanced cirr hosis and simultaneous occurrence of anti-smooth-muscle antibodies (ASMA) i n serum (p < 0.0006); their liver damage was often secondary to HBV (37.3%) or alcohol abuse (18.5%). At conditional logistic regression analysis, onl y the presence of ASMA (odds ratio [OR] = 3.02, 95% confidence interval [CI ] 1.7-5.5, p = 0.0003), HBV (OR = 3.4, 95% CI 1.6-7.2, p = 0.0013), or alco holic liver disease (OR = 5.3, 95% CI 2.3-12.2, p = 0.0001) were independen tly associated with aCL. Thrombosis was encountered in 24 patients (6.4%). At conditional logistic regression analysis, thrombosis was significantly a ssociated with advanced age (OR = 1.07, 95% CI 1.0-1.1, p = 0.0094), develo pment of hepa tocellular carcinoma (OR = 17.8, 95% CI 1.6-196.0, p 0.01), H BV etiology (OR = 6.3, 95% CI, 1.6-24.6, p = 0.0076), or cryptogenic liver disease (OR = 54.8, 95% CI 5-599.9, p = 0.001). Of the five patients with n ewly documented portal thrombosis during the follow-up, only one tested pos itive for aCL. CONCLUSIONS: In patients with nonautoimmune liver disease, aCL production i s an epiphenomenon of the liver damage and is not associated with thromboti c complications. These data do not support the hypothesis that HCV is a cau se of the antiphospholipid syndrome. (C) 1999 by Am. Coll. of Gastroenterol ogy.