Objective: To detect risk factors for intracranial hemorrhage (ICH) in patients with long-term oral anticoagulant and to identify clinical or radiological data specific of anticoagulant-related ICH. Methods and Patients: Three groups of patients were included. Group 1 represents patients who were admitted because of anticoagulant-related ICH between January 1984 and February 1996. All patients underwent CT scan. Clinical data, anticoagulation parameters, location and volume of the ICH, treatment and the 30-days in-hospital mortality were analyzed. Group 2 consisted of patients selected at random among all patients with spontaneous ICH admitted to our department during the same period of time. Patients without ICH, but regularly taking oral anticoagulants constituted group 3. Results: Seventy-nine patients with anticoagulant-related ICH were compared to 127 patients with spontaneous ICH. The volume of supratentorial ICH was greater in group 1 of patients and was correlated with a worse prognosis. Comparison of group 1 with group 3 (212 controls) demonstrated that length of anticoagulation, prothrombin time or excessive anticoagulation, prior cerebral infarct and use of acénocoumarol, but not age or indication of anticoagulant, were significant risk factors for ICH in multivariate analysis. Conclusions: The results emphasize that anticoagulant-related ICH are not clinically different from spontaneous ICH except for volume of bleeding, and that frequent and careful coagulation monitoring is needed, especially during the first year in order to decrease the risk of ICH.

MULTIPLE DURAL LESIONS MIMICKING MENINGIOMAS IN PATIENTS WITH CCM3/ PDCD10 MUTATIONSCerebral cavernous malformations (CCM) are in most cases located within the brain parenchyma. Familial forms of CCM (FCCM) are characterized by an autosomal dominant pattern of inheritance and the presence of multiple cerebral lesions. 1 Three CCM genes have been identified: CCM1/KRIT1, CCM2/MGC4607, and CCM3/PDCD10. 1,2 They are involved in 53%, 15%, and 10% of patients with FCCM. 3 All mutations lead to premature stop codons. Herein, we identified, within a series of 197 index patients with CCM mutations, four patients with multiple dural-based lesions mimicking meningiomas. All four index patients carried a mutation in CCM3. Methods and results.A consecutive cohort of 267 CCM index patients with either multiple lesions or an affected relative who gave written informed consent were screened for mutations within the three CCM genes as previously reported. 3 This study was approved by our local ethics committee. Genetic screening identified a deleterious mutation in 197 of these index patients, which constituted our study group. Clinical manifestations and mutations of 128 of these index patients were previously reported. 2 A total of 129 out of the 197 index patients had a CCM1 mutation (65.4%), 39 had a CCM2 mutation (19.7%), and 29 had a CCM3 mutation (14.7%) (data not shown).Multiple intracranial dural-based lesions were detected in four index patients whose main imaging features are shown in the figure. Briefly, these four patients (three women/one man; mean age: 48 years, range: 42-64) show the following: 1) multiple parenchymal cavernous angiomas; and 2) multiple dural-based lesions with similar magnetic resonance patterns, i.e., hypointense on T1, hyperintense on T2 sequences, markedly enhanced by gadolinium injection, and a diffuse dural enhancement. Gradient echo MRI did not show hyposignals suggestive of hemosiderin deposits within those dural-based lesions. The number of dural lesions increased with age in two of them (cases C30 and C161). All four pa-tients carry a deleterious mutation in CCM3/ PDCD10. C030 index patient has an R196X stop codon mutation. C107 showed a c.475-1 G ϾA splice site mutation leading to an aberrant splicing of exon 9. C158 has a c.474 ϩ 5GϾA splice site mutation leading to several abnormal transcripts. C161 has a one bp deletion within exon 9 leading to a frameshift and a stop codon at codon 188.The proportion of patients with CCM3 mutations who had multiple dural lesions in this consecutive series is approximately 14% (4/29). Percentages of patients carrying dural-based lesions within the three CCM1/CCM2/CCM3 groups were clearly different (p Ͻ 0.0001/Fisher test). None of the patients with CCM without an identified mutation showed dural-based lesions.

No abstract

No abstract

No abstract