Thrombophilia as a common predisposing factor in pseudotumor cerebri

Pseudotumor cerebri (PTC), also known as idiopathic or benign intracranial hypertension, is related in a significant number of cases to cranial venous outflow obstruction, defined as abnormal venous flow on imaging or venography that may or may not have demonstrable cerebral venous thrombosis which represents an extreme end of the spectrum. Indeed, it has been argued that such a mechanism underlies all cases of PTC, although depending on how assiduously such abnormalities are sought, or on the limitations of imaging modalities, this may not be documented.^1,2  Recently, we reviewed a large series of PTC patients, including a subgroup in whom the cranial venous outflow was investigated in detail, and found that 31% had evidence of venous outflow obstruction.² 

Clearly, there is the possibility of an underlying thrombophilia in PTC patients with cranial venous outflow obstruction with or without demonstrable thrombosis. An association of PTC with systemic lupus erythematosis and an increased incidence of anticardiolipin antibodies have been described; however, the importance of other thrombophilias in PTC with cranial venous outflow obstruction has not been examined.³  Martinelli et al recently described the association of thrombophilia, particularly hyperhomocysteinemia, and overt cerebral vein thrombosis.⁴  The importance of Factor V Leiden (FVL) and the PT20210 mutation had previously been recognized in cerebral venous thrombosis.^5,6 

We had the opportunity to examine 25 consecutive patients with PTC who were admitted to the neurosurgical unit between 1998 and 2000 and who previously had been assessed for venous outflow obstruction. Only 4 patients had frank cerebral venous thrombosis. The average age was 30 years (range, 11-49 years) and the majority female (23 of 25). All patients had a normal platelet count and screening coagulation studies (activated partial thromboplastin time and prothrombin time). Protein C (PC), protein S (PS; functional), antithrombin levels (AT; STA, chromogenic, Diagnostica Stago, Asnieres-sur-Seine, France), and activated protein C resistance were measured (GradiLeiden V, Gradipore, Australia). Lupus anticoagulant (LA) was examined using the Dilute Russel Viper Venom Time (DRVVT) and Kaolin Clotting Time (KCT) with confirmation by addition of excess phospholipid. The anticardiolipin antibody (ACA) was measured using an immunoglobulin G enzyme-linked immunosorbent assay and defined as positive if the titer was moderate to high and persisted for more than 6 weeks. Fasting serum homocysteine levels were measured (normal range, < 23 μM) without oral methionine loading, and all patients confirmed to have normal B12 and folate levels. Mutations for FVL and PT20210 were sought (multiplex amplification refractory mutation system polymerase chain reaction [ARMS PCR]). Fisher exact test was used for statistical analysis.

Thrombophilic defects were detected in 68% (Table 1). There were 2 patients with low PS levels (41% and 43%; normal, 80%-130%), 4 patients with APCR and FVL, 2 with PT20210, 6 with positive ACA, 3 with positive LA (none of whom had positive ACA), and 2 with elevated fasting homocysteine. The rate of FVL and PT20210 is in keeping with rates seen in other thrombotic disorders such as deep vein thrombosis.⁷  We confirm the high frequency of positive LA and ACA as previously described in PTC.³  Again, this is in keeping with other thrombophilic groups.⁸  The possibility exists that ACA positivity is related to an infective cause, and 2 of 6 patients had probable infected shunts at time of testing.

Only one patient had a personal history of deep venous thrombosis (no thrombophilia detected), whereas 3 patients had a family history of venous thromboembolism (all with thrombophilia). There was no history of autoimmune disease and all patients had antinuclear antibody less than a titer of 1:80. No patient was receiving estrogen preparations but, given the striking female predominance, one must speculate on the relationship between endogenous estrogen and underlying thrombophilia. The association between users of oral contraceptives and cerebral venous thrombosis has been described.^4,6 

A slight predominance of thrombophilic defects was seen in patients with documented venous outflow obstruction: 77% (7 of 9) versus 62% (10 of 16) (P = .048). As mentioned, such a diagnosis may be difficult due to the limitations of imaging. It is, therefore, possible that some patients with thrombophilia did, in fact, have undetectable cranial venous outflow obstruction.

In conclusion, there is a high incidence of thrombophilic defects in PTC, particularly in patients with cranial venous outflow obstruction. Cranial venous outflow obstruction may exist in the absence of frank cerebral vein thrombosis and may underlie the pathophysiology of this condition. Further investigation is underway.