Article Text

Download PDFPDF

Acquired factor V inhibitor associated with valproic acid use in a cirrhotic patient
  1. B Godart1,
  2. C Boinot2,
  3. C Remblier3,
  4. A Hajjar4,
  5. M Beauchant5
  1. 1Service d’Hépato-Gastroentérologie, CHU la Milétrie, Poitiers, France
  2. 2Laboratoire d’Hématologie, CHU la Milétrie, Poitiers, France
  3. 3Service de Pharmacologie Clinique, Pavillon René le Blaye Nord, Poitiers, France
  4. 4Service de Médecine, Centre Hospitalier, Le Blanc, France
  5. 5Service d’Hépato-Gastroentérologie, CHU la Milétrie, Poitiers, France
  1. Correspondence to:
    Dr B Godart
    Service d’Hépato-Gastroentérologie, CHU La Milétrie, 86021 Poitiers cedex, France; b.godart{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Acquisition of factor V inhibitor is a rare event. The inhibitor most frequently encountered in clinical practice is directed against factor VIII. In a recent review of the literature, Streiff and Ness1 found 126 published cases of factor V inhibitor. The inhibitor emerged after major surgery, haemostatic therapy with bovine thrombin, malignancies, autoimmune disorders, blood transfusion, antibiotic therapy, or for unknown reasons. We report the emergence of factor V inhibitor in a cirrhotic patient receiving valproic acid for seizure control.

A 50 year old man treated for alcoholic cirrhosis was admitted for epistaxis. He had no history of autoimmune disorders or blood transfusion. For three years he had been taking valproic acid 1 g/day orally for seizures, and propranolol 60 mg/day. On admission, prothrombin level was 5% of control, factor V was 1%, and factor II was 49%. Two months previously prothrombin level had been 83% of control on two occasions one month apart. Physical examination showed compensated cirrhosis. Epistaxis was linked to telangiectasia and was controlled by meshing followed by local cauterisation with silver nitrate. Laboratory tests showed no signs of clotting factor consumption (platelet count 68×109/l, D-dimers <500 ng/ml, fibrinogen 3 g/l). Valproic acid was replaced by phenobarbital while propranolol was maintained.

Three months later prothrombin level was 33%, factor V 14%, factor II 72%, factor VII 70%, factor IX 92%, and factor X 65% of control. Factor VIII, XI, and XII levels were normal, as was thrombin time. He was screened for factor V inhibitor by measuring residual factor V after one hour of incubation at 37°C of equal parts of pooled normal plasma and patient plasma. Antibody titre was determined with the Bethesda method used to titre factor VIII inhibitor. Factor V inhibitor titre was 1.0 Bethesda units. Protein immunoelectrophoresis was normal, and tests for antinuclear, antinative DNA, anti-beta2-glycoprotein 1, anticardiolipin, antithyroperoxidase, antimitochondria, and antismooth muscle antibodies were negative. Aspartate aminotransferase was 73 IU/l (normal <40 IU/l), albuminaemia 35.2 g/l, and α fetoprotein 9.6 ng/ml. Abdominal sonography and colonoscopy were normal. Gastroduodenal endoscopy showed grade 1 oesophageal varices. Eighteen months after admission, the patient was asymptomatic and epistaxis had not recurred. Factor V was 41% and weak factor V inhibitor activity persisted (0.6 Bethesda units).

Hypocoagulability due to factor V inhibitor is rare and can be difficult to diagnose in a patient with cirrhosis. Eighty seven of the 126 cases described by Streiff and Ness1 occurred during the last decade, and two thirds of cases followed bovine thrombin exposure. Antibovine factor V and antihuman factor V antibodies can interact, potentially inactivating human factor V in vivo. Other noteworthy causes are blood transfusion, cancer, treatment with betalactam antibiotics or streptomycin, major surgery (usually in patients having received transfusions or betalactam agents), and autoimmune disorders (coeliac disease, bullous pemphigoid, Sjögren’s syndrome, Hashimoto thyroiditis) associated with congenital factor V deficiency. No cause was found in nearly 20% of cases.2 To date, no cases have been linked to dental extraction or other minor surgeries. Anti-factor V antibodies can appear at all ages but most reported cases occurred after age 65 years.2 The inhibitor was discovered fortuitously in nearly 40% of cases following an isolated increase in prothrombin time. Bleeding was the main presenting sign in 60% of cases, and was life threatening in 22%.2

To our knowledge, this is the first reported case of factor V inhibitor associated with valproic acid therapy. It is noteworthy that valproic acid inhibits fatty acid beta oxidation, potentially leading to life threatening microvesicular steatosis.3 However, our patient had no clinical or biological signs of hepatitis. Rare cases of cutaneous vasculitis or lupus-like syndrome have been linked to valproic acid or its prodrug valpromide. Factor V inhibitors have occasionally been detected in patients with such syndromes,4 but our patient had no clinical or biological signs of an autoimmune process. Factor V inhibitor appeared after three years of treatment with valproic acid, and prothrombin level improved partially after drug withdrawal. In previously reported cases, the inhibitor disappeared in 88% of patients overall, after a mean of 10 weeks.2 In patients with no identified cause, the inhibitor only disappeared in 62% of cases after a mean of 23 weeks,2 although this did not affect outcome. Bleeding is difficult to treat in patients with factor V inhibitor. Various approaches have been tried, such as infusion of fresh frozen plasma or, better, platelet concentrates.2 Plasmapheresis has been used to lower antibody titre and high dose immunoglobulin to neutralise the antibodies. Steroids and immunosuppressants (azathioprine, cyclophosphamide), alone or in combination, have been used for long term inhibition of factor V inhibitor synthesis. However, the results are difficult to interpret as the series were small and included patients with heterogeneous manifestations. There is no consensus treatment.

In conclusion, the onset of hypocoagulability linked to a decline in factor V level in a cirrhotic patient should not be systematically attributed to hepatocellular insufficiency; in the absence of marked cytolysis, the presence an acquired factor V inhibitor and a possible drug related cause should be sought.



  • Conflict of interest: None declared.