SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. REFERENCES

Objective

To analyze the clinical findings, treatment, outcome, and prevalence of cerebral venous thrombosis (CVT) in a large cohort of patients with Behçet's disease (BD) from a single center.

Methods

We reported a series of 64 consecutive patients with CVT who fulfilled the international criteria for BD. Multivariate analysis was performed to define factors that affect prognosis.

Results

Among a cohort of 820 patients with BD, CVT was present in 64 (7.8%). Compared with BD patients without CVT, those with CVT had lower parenchymal central nervous system involvement (4.7% versus 28.7%; P = 0.0001) and higher extraneurologic vascular lesions (62.5% versus 38.8%; P = 0.03). Up to 90% of patients responded to anticoagulation therapy without severe hemorrhagic complications. Neither steroid nor immunosuppressant use provided better outcome. Severe visual loss due to optic atrophy was the main complication of CVT, being found in 15% of patients. In multivariate analysis, papilledema (odds ratio [OR] 7.1, 95% confidence interval [95% CI] 1.6–31.9) and concurrent prothrombotic risk factors (OR 4.6, 95% CI 1.1–20.2) were independently associated with the occurrence of sequelae. Factors associated with relapse of thrombosis were concurrent prothrombotic risk factors (hazard ratio [HR] 4.9, 95% CI 1.5–15.4) and a peripheral venous thrombosis (HR 2.8, 95% CI 0.7–10.5). After a mean ± SD followup of 8.2 ± 6.9 years, 4 deaths unrelated to CVT were noted.

Conclusion

CVT in patients with BD may result in serious neurologic outcomes. Anticoagulation represents a safe and effective therapy. Extensive investigation of prothrombotic disorders should be considered.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. REFERENCES

Behçet's disease (BD) is a vasculitis of unknown etiology characterized by mucocutaneous, ocular, arthritic, and vascular manifestations (1, 2). Its neurologic manifestations (neuro-BD) are relatively rare, but they must be thoroughly investigated due to their severe prognosis (3). The frequency of neurologic manifestations varies, ranging from 5% (4) to 30% (5) of patients. In an autopsy series of 170 patients with Behçet's syndrome, 20% showed pathologic evidence for neurologic involvement (6). Both the central and peripheral nervous systems can be involved. Central nervous system (CNS) manifestations can be divided into 2 main groups: 1) parenchymal involvement, which includes brainstem involvement, hemispheric manifesta- tions, spinal cord lesions, and meningoencephalitic presentations, and 2) nonparenchymal involvement, including dural sinus thrombosis, arterial occlusion, and/or aneurysms (7, 8). Peripheral neuropathy and myopathy are rare (7, 8).

Cerebral venous thrombosis (CVT) is a major manifestation of BD. However, its frequency is difficult to ascertain from the literature, which consists mainly of single case reports or short series (7–13). Nevertheless, CVT represents approximately 30% of all CNS lesions of BD (10, 13). Although the functional outcome from CVT is better than arterial strokes, the outcome of CVT remains unpredictable and may lead to sequelae or even death if not recognized and treated early.

Because of the relatively small number of patients in previous series, the clinical spectrum of CVT, its response to treatment, and its long-term outcome are still uncertain in patients with BD. In this study, we reported 64 consecutive patients with CVT seen in a single university hospital (Pitié-Salpétrière, Paris, France). For all of these patients, the main characteristics, treatment, and long-term outcome of CVT are reported. Multivariate analysis was performed to define factors that affect prognosis.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. REFERENCES

Patients.

Between 1974 and 2006, CVT was diagnosed in 64 (7.8%) of the 820 patients with BD seen in the Department of Internal Medicine and Neurology at the University Hospital La Pitié-Salpétrière, Paris, France. For each patient, the following data were collected: age at diagnosis of CVT; sex; date of meeting criteria for BD; main features of BD including cutaneous, ocular, rheumatologic, neurologic, and/or vascular involvement; contraceptive and tobacco use; and clinical features, outcome, and treatment of CVT. Lumbar puncture, cerebral angiography, brain computed tomography and, since 1985, magnetic resonance imaging (MRI) were performed when clinically indicated. All 64 patients fulfilled the international criteria for classification of BD (14). The diagnosis of CVT was considered in symptomatic patients with either headache, intracranial hypertension, and/or other neurologic manifestations. As previously described (15), the diagnosis of CVT was defined in all patients by angiographic abnormalities: either partial or total lack of filing of at least 1 dural sinus on 2 projections of carotid angiography. Angiography was used in all patients before 1985. Since 1985, in cases of neurologic symptoms, MRI was systematically performed on a 0.5T GE-GGR MR unit (General Electric, Waukesha, WI), except in 6 patients with MRI contraindication who were diagnosed with cerebral angiography. Sagittal T1-weighted (repetition time/echo time/number of echoes 400/21/1 msec) and coronal T2-weighted (1,800/60/3 msec) slices were obtained.

Fifty-one patients were investigated at the time of CVT diagnosis for a prothrombotic state: either protein C, protein S, and antithrombin deficiency; factor V Leiden R506Q mutation, prothrombin G20210A mutation, and methylenetetrahydrofolate reductase (MTHFR) gene C677T mutation; or plasma total homocysteine level (normal range 7.5–15 μmoles/liter), lupus anticoagulant, and anticardiolipin antibodies. Repeated testing was systematically performed to confirm lupus anticoagulant and anticardiolipin antibodies. Earlier patients were rescreened for the determination of new prothrombotic factors.

Statistical analysis.

Data are shown as frequencies and percentages for categorical variables. Quantitative variables are shown as the median and 25th and 75th percentiles or as the mean ± SD. Univariate associations between sequelae and variables were assessed using the Fisher's exact test. All factors with P values less than 0.1 in the univariate analysis were included in a multiple logistic regression model. Variable selection was performed using a backward procedure based on Akaike's criterion. Odds ratios (ORs) with their 95% confidence intervals (95% CIs) are shown as a measure of association.

Factors associated with the occurrence of thrombosis relapse were assessed using a Cox proportional hazards model, using a similar variable selection as described for sequelae. Hazard ratios (HRs) with their 95% CIs are shown as a measure of association. Comparisons of categorical variables were performed using the chi-square test or Fischer's exact test. All goodness-of-fit tests of the regression models were evaluated using the le Cessie and van Houwelingen method for logistic models (16) and the Grambsch and Therneau method for the Cox model (17). All tests were 2-sided at the 0.05 significance level. Analyses were performed using the R statistical package (online at http://www.R-project.org).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. REFERENCES

Characteristics of BD.

The main features of patients with BD are shown in Table 1. Clinical signs of BD included recurrent oral (100%) and genital (71.9%) ulcerations, skin lesions (75%, mainly pseudofolliculitis), and cutaneous hypersensitivity. Eye lesions other than papilledema included retinal vasculitis (n = 11), posterior uveitis (n = 8), and iridocyclitis (n = 5). Other signs of BD such as arthralgias and epididymitis were observed in 37.5% and 3.1% of patients, respectively. Vascular involvement was frequently present (62.5%), with recurrent or multiple venous thrombosis in 38 patients (59.3%, including pulmonary embolism [n = 8], central retinal vein thrombosis [n = 5], jugular vein and inferior vena cava [n = 4], superior vena cava and superior hepatic veins [n = 2], upper extremities [n = 4], and lower extremities [n = 20]). Arterial lesions were observed in 11 patients (17.2%) and included aortic artery aneurysm (n = 5), coronary artery aneurysm (n = 2), and pulmonary and femoral arterial aneurysm (n = 3). The number of BD flares prior to CVT was 1 in 14.1%, 2 in 23.4%, 3 in 29.7%, and ≥4 in 32.8% of patients. The mean ± SD C-reactive protein level (normal value <4 mg/dl) was 36.4 ± 34.4 mg/liter (range 5–132).

Table 1. Characteristics of patients with BD*
ParameterBD with CVT (n = 64)BD without CVT (n = 756)P
  • *

    Values are the number (percentage) unless otherwise indicated. BD = Behçet's disease; CVT = cerebral venous thrombosis; CNS = central nervous system.

Age at BD diagnosis, mean ± SD years26.2 ± 7.527.7 ± 8.20.93
Male sex42 (65.6)499 (66.0)1
Origin   
 North Africa32 (50.0)307 (40.6)0.35
 Africa7 (10.9)109 (14.4)0.57
 Europe25 (39.1)340 (45.0)0.63
HLA–B513/53 (24.5)246/599 (41.0)0.11
Clinical features of BD   
 Oral ulcerations64 (100)733 (96.9)0.92
 Genital ulcerations46 (71.9)524 (69.3)0.91
 Arthralgias24 (37.5)332 (43.9)0.62
 Skin lesions48 (75.0)522 (69.0)0.68
 Extraneurologic vascular lesions40 (62.5)294 (38.8)0.03
 Eye involvement44 (68.8)470 (62.1)0.68
 Parenchymal CNS involvement3 (4.7)217 (28.7)0.0001

Among our 820 patients with BD, when we compared those with and without CVT (Table 1), we found a significantly lower frequency of parenchymal CNS involvement (4.7% versus 28.7%; P = 0.0001) and a higher frequency of extraneurologic vascular lesions (62.5% versus 38.8%; P = 0.03) among patients with CVT.

Neurologic features of BD.

The main neurologic features of CVT are shown in Table 2. The mode of onset was acute (<48 hours) in 23 patients (35.9%) and progressive in 41 patients (64.1%). No correlation was found between the mode of onset of CVT and the presence of papilledema (P = 0.59) or blindness (P = 0.41), or with the occurrence of sequelae (P = 0.33). The most frequent signs were headache (96.9%), papilledema (62.5%), fever (28.1%), nausea/vomiting (18.7%), focal deficits (12.5%, including hemiparesis and/or hypoesthesia, cranial palsy, monoparesis, and alternate monoparesis), seizures (7.8%), and confusion (6.2%). Cerebral angiography and MRI were performed in 19 and 45 patients, respectively. Focal T2-weighted hypersignals (lenticular, temporal, diencephalic, mesencephalic, and pontine) were observed in 6 patients. Cerebrospinal fluid (CSF) examination was performed in 38 patients. As a medical center–based tertiary reference department in BD, many patients we received were already receiving oral anticoagulation, which prevented CSF study. Thirty-five CSF examination findings were normal. The 3 remaining patients with associated meningoencephalitis showed pleocytosis with a lymphocytic predominance and/or elevated protein content. The CSF pressure was evaluated in 21 patients and was found elevated (>22 cm H2O) in 13 (61.9%). The mean CSF pressure level was 23 cm H2O (median 25, range 11–49).

Table 2. Neurologic features of the 64 patients with BD and CVT*
ParameterValue
  • *

    Values are the number (percentage) unless otherwise indicated. BD = Behçet's disease; CVT = cerebral venous thrombosis; MTHFR = methylenetetrahydrofolate reductase; CSF = cerebrospinal fluid.

Age at CVT diagnosis, mean ± SD years31.7 ± 14.4
CVT as initial feature of BD14 (21.9)
Clinical features of CVT 
 Persistent headache62 (96.9)
 Papilledema40 (62.5)
 Fever18 (28.1)
 Nausea/vomiting12 (18.7)
 Focal deficit8 (12.5)
 Seizure5 (7.8)
 Confusion4 (6.2)
Site of CVT occlusion 
 Superior sagittal sinus41 (64.1)
 Transverse sinuses39 (60.9)
  Left26 (40.6)
  Right26 (40.6)
 Deep cerebral vein4 (6.3)
 Cavernous sinuses3 (4.7)
Associated prothrombotic state16/51 (31.0)
 MTHFR mutation (homozygous/heterozygous), no.6 (4/2)
 Hyperhomocysteinemia, no.6
 Anticardiolipin antibodies, no.4
 Factor V Leiden mutation, no.2
 Prothrombin mutation, no.1
 Protein C deficiency, no.1
Tobacco use22 (34.4)
Contraceptive use8/22 (36.3)
Elevated CSF pressure13/21 (61.9)

Associated prothrombotic factors are listed in Table 2. Sixteen patients (31.0%) showed at least 1 prothrombotic factor associated with BD (MTHFR mutation, homozygous [n = 4] and heterozygous [n = 2], hyperhomocysteinemia [n = 6], anticardiolipin antibodies [mean ± SD 46 ± 23 IgG phospholipid units] and lupus anticoagulant [n = 4], factor V Leiden mutation [n = 2], prothrombin mutation [n = 1], and protein C deficiency [n = 1]). Tobacco and contraceptive use were noted in 34.4% and 36.3% of patients, respectively. Fourteen patients (27%) had >1 risk factor for thrombophilia.

Treatment and outcome of CVT.

Treatment of BD manifestations other than CVT consisted of steroids in 54 patients (84.4%), immunosuppressants (mainly cyclophosphamide and azathioprine) in 32 patients (50.0%), and colchicine in 59 patients (92.2%). Treatment of CVT consisted of anticoagulation in 62 (96%) of 64 patients (Table 3). The 2 remaining patients were treated with a combination of aspirin and steroids because of the impossibility of regular monitoring of warfarin therapy. One of the 2 patients underwent a ventricular peritoneal derivation because of intracranial hypertension and experienced relapse of CVT. Among the 62 patients that received anticoagulation, 48 also initially received a combined treatment with steroids (prednisone or equivalent 1 mg/kg/day). Oral anticoagulation was maintained for <1 year in 6 (9.7%) of 62 patients, 1–3 years in 15 (24.2%) of 62 patients, and >3 years in 41 (66.1%) of 62 patients. The duration of anticoagulation was dependent on the timing of clinical improvement and on the presence of an associated prothrombotic state or thrombotic complication. Short-term outcome was good, with regression and/or disappearance of neurologic symptoms within 1 month in 56 (87.5%) of 64 patients. Four patients experienced hemorrhagic complications (psoas hematoma [n = 2], leg hematoma [n = 1], and subdural hematoma [n = 1]) that recovered without sequelae.

Table 3. Treatment and outcome of the 64 patients with BD and CVT*
ParametersValue
  • *

    Values are the number (percentage) unless otherwise indicated. BD = Behçet's disease; CVT = cerebral venous thrombosis.

Treatment 
 Anticoagulation, years62 (96.9)
  <16 (9.7)
  1–315 (24.2)
  >341 (66.1)
 Aspirin29 (45.3)
 Steroids54 (84.4)
 Immunosuppressants32 (50.0)
Event during followup 
 Total sinus recanalization33/57 (57.9)
 Recurrent CVT7 (10.9)
 Other thrombotic events17 (26.5)
  Extremity or pelvic venous thrombosis, no.10
  Vena cava, no.5
  Pulmonary embolism, no.2
 Flare of BD28 (43.7)
 Sequelae23 (35.9)
  Blindness, no.6
  Persistent headache, no.6
  Decrease visual activity, no.4
  Altered cognition, no.3
  Hemiparesis, no.2
  Tinnitus, no.2
 Death4 (6.2)
  Myocardial failure, no.2
  Neuro-BD, no.1
  Unknown, no.1

The clinical long-term followup ranged from 6 months to 27 years (mean ± SD 8.2 ± 6.9 years). Repeat of cerebral imaging (by angiography or MRI) was performed in 57 patients. Total sinus recanalization of CVT was evidenced in 33 (57.9%) of 57 patients. The remaining patients without total recanalization had partial or total lack of filing of the involved sinus or vein. Cerebral MRI disclosed an old thrombosis, as evidenced by hyposignal on T2 sequence.

Sequelae of CVT were noted in 23 patients (35.9%), and included blindness (n = 6), persistent headache (n = 6), decreased visual activity (n = 4), altered cognition (n = 3), hemiparesis (n = 2), and tinnitus (n = 2) (Table 3). Among patients with blindness, 2 had a delayed diagnosis of CVT and 2 others had poor adherence to warfarin therapy.

Following CVT, 20 patients (31.2%) experienced a relapse of thrombosis (Figure 1). Four patients (6.2%) relapsed at 1 year, 11 (17.2%) at 3 years, 15 (23.4%) at 5 years, 19 (29.7%) at 10 years, and 20 (31.2%) at 20 years of followup. Of the 20 patients with thrombosis relapse, 7 had a relapse of CVT (Figure 1). One patient (1.6%) relapsed at 1 year, 3 (4.7%) at 3 years, 6 (9.4%) at 5 years, and 7 (10.9%) at 10 years. Relapse of CVT occurred in all except 1 patient, while warfarin therapy was discontinued. Other treatments at the time of relapse included steroids (n = 6, range 10–30 mg/day) and azathioprine (n = 3). No difference was noted regarding the recurrence rate of thrombosis between earlier BD patients (1974–1990) and others (1991–2006).

thumbnail image

Figure 1. Probability of thrombosis relapse after cerebral venous thrombosis (CVT) in patients with Behçet's disease.

Download figure to PowerPoint

A flare of BD occurred in 28 patients (43.7%) after the diagnosis of CVT. After a mean ± SD followup of 8.2 ± 6.9 years, 4 deaths were noted due to myocardial failure (n = 2), CNS involvement of BD (n = 1), and unknown origin (n = 1) (Table 3).

Thirty-two of the 62 BD patients who were treated with anticoagulation therapy also received immunosuppressants. Similar outcomes were observed with both regimens in terms of sequelae (23.3% versus 50%; P = 0.15) and thrombosis relapse (20% versus 43.7%; P = 0.19).

Factors associated with the occurrence of sequelae after CVT.

In univariate analysis, papilledema (not associated with other eye lesions of BD) and concurrent prothrombotic risk factors were associated with the occurrence of sequelae (Table 4). In multivariate analysis, papilledema (OR 7.1, 95% CI 1.6–31.9) and concurrent prothrombotic risk factors (OR 4.6, 95% CI 1.1–20.2) were independently associated with occurrence of sequelae (Table 4). In a separate analysis with hyperhomocysteinemia as the only prothrombotic factor, the OR was 2.5 (95% 0.3–33.4, P = 0.37).

Table 4. Factors associated with the occurrence of sequelae*
ParametersNo. (%)Univariate analysisMultivariate analysis OR (95% CI)
OR (95% CI)P
  • *

    OR = odds ratio; 95% CI = 95% confidence interval; BD = Behçet's disease; CVT = cerebral venous thrombosis; CRP = C-reactive protein.

  • P = 0.01.

  • P = 0.04.

Sex    
 Female (n = 22)6 (27.2)1  
 Male (n = 42)17 (40.5)1.52 (0.43–5.92)0.57 
Genital ulcerations    
 No (n = 18)6 (33.3)1  
 Yes (n = 46)17 (36.9)1.35 (0.38–5.3)0.77 
Arthralgias    
 No (n = 40)16 (40.0)1  
 Yes (n = 24)7 (29.2)0.58 (0.16–1.94)0.42 
Skin involvement    
 No (n = 16)3 (18.7)1  
 Yes (n = 48)20 (41.7)2.43 (0.53–15.55)0.33 
Papilledema    
 No (n = 24)3 (12.5)1 1
 Yes (n = 40)20 (50.0)4.11 (1.18–16.93)0.0187.18 (1.61–31.91)
Fever    
 No (n = 46)16 (38.1)1  
 Yes (n = 18)7 (38.9)1.03 (0.28–3.66)1 
Arterial involvement    
 No (n = 53)19 (38.0)1  
 Yes (n = 11)4 (40.0)1.09 (0.2–5.29)1 
Peripheral venous thrombosis    
 No (n = 26)12 (46.2)1  
 Yes (n = 38)11 (28.9)0.63 (0.21–1.82)0.46 
BD flares prior to CVT    
 1 (n = 9)3 (37.5)1  
 2 (n = 15)5 (35.7)0.93 (0.15–5.61)  
 3 (n = 19)6 (33.3)0.83 (0.15–4.72)  
 ≥4 (n = 21)9 (45.0)1.36 (0.25–7.32)0.93 
Superior sagittal sinus occlusion    
 No (n = 23)7 (30.4)1  
 Yes (n = 41)16 (39.0)0.93 (0.23–3.93)1 
Transverse sinuses occlusion    
 No (n = 25)7 (28.0)1  
 Yes (n = 39)16 (41.0)0.88 (0.23–3.45)1 
Elevated CRP level (>10 mg/liter)    
 No (n = 24)6 (25.0)1  
 Yes (n = 40)17 (42.5)0.52 (0.1–2.59)0.48 
HLA–B5    
 No (n = 40)16 (42.1)1  
 Yes (n = 13)4 (36.4)0.79 (0.14–3.76)1 
Associated prothrombotic factor    
 No (n = 35)11 (31.4)1 1
 Yes (n = 16)9 (56.2)2.75 (0.7–11.32)0.104.65 (1.07–20.2)
Tobacco use    
 No (n = 42)12 (28.6)1  
 Yes (n = 22)11 (50.0)1.49 (0.43–5.19)0.57 
Oral contraceptive use    
 No (n = 14)9 (40.9)1  
 Yes (n = 8)1 (16.7)0.3 (0.01–3.4)0.37 
Anticoagulation    
 No (n = 2)1 (50.0)1  
 Yes (n = 62)22 (37.9)0.62 (0.01–50.13)1 
Duration of anticoagulation, years    
 <1 (n = 6)2 (33.3)1  
 1–3 (n = 15)4 (26.6)1.33 (0.09–20.11)  
 3 (n = 41)17 (41.5)3.09 (0.32–30.25)0.52 
Aspirin    
 No (n = 35)13 (37.1)1  
 Yes (n = 29)10 (34.5)0.77 (0.23–2.59)0.78 
Immunosuppressants    
 No (n = 32)9 (28.1)1  
 Yes (n = 32)14 (43.7)2.15 (0.66–7.3)0.19 
Steroids    
 No (n = 10)5 (50.0)1  
 Yes (n = 54)18 (33.3)0.57 (0.11–2.83)0.49 
Colchicine    
 No (n = 3)1 (33.3)1  
 Yes (n = 61)22 (37.1)0.62 (0.01–50.13)1 
Complete recanalization of CVT    
 No (n = 24)12 (50.0)1  
 Yes (n = 33)11 (33.3)0.59 (0.13–2.61)0.52 

Factors associated with relapse of thrombosis after CVT.

In univariate analysis, male sex, peripheral venous thrombosis, concurrent prothrombotic risk factors, and the use of immunosuppressants were associated with relapse of thrombosis (Table 5). In multivariate analysis, concurrent prothrombotic risk factors (HR 4.9, 95% CI 1.5–15.4) and peripheral venous thrombosis (HR 2.8, 95% CI 0.7–10.5) were independently associated with a relapse of thrombosis (Table 5). In a separate analysis with hyperhomocysteinemia as the only prothrombotic factor, the HR was 5.7 (95% CI 1.7–5.7, P = 0.005).

Table 5. Factors associated with relapse of thrombosis*
ParametersNo. (%)Univariate analysisMultivariate analysis HR (95% CI)
HR (95% CI)P
  • *

    HR = hazard ratio; 95% CI = 95% confidence interval; BD = Behçet's disease; CVT = cerebral venous thrombosis; CRP = C-reactive protein.

  • P = 0.12.

  • P = 0.007.

Sex    
 Female (n = 22)4 (18.2)1  
 Male (n = 42)16 (38.1)2.61 (0.87–7.83)0.086 
Genital ulcerations    
 No (n = 18)7 (38.9)1  
 Yes (n = 46)13 (28.3)0.62 (0.24–1.57)0.31 
Arthralgias    
 No (n = 40)13 (32.5)1  
 Yes (n = 24)7 (29.2)0.84 (0.33–2.11)0.71 
Skin involvement    
 No (n = 16)2 (12.5)1  
 Yes (n = 48)18 (37.5)2.73 (0.63–11.79)0.18 
Papilledema    
 No (n = 24)5 (20.8)1  
 Yes (n = 40)15 (37.5)2.28 (0.83–6.3)0.11 
Fever    
 No (n = 46)15 (33.3)1  
 Yes (n = 18)5 (26.3)0.78 (0.28–2.16)0.64 
Arterial involvement    
 No (n = 53)14 (26.4)1  
 Yes (n = 11)6 (54.5)1.84 (0.7–4.79)0.21 
Peripheral venous thrombosis    
 No (n = 26)4 (15.4)1 1
 Yes (n = 38)16 (42.1)2.94 (0.98–8.79)0.0542.83 (0.76–10.5)
BD flares prior to CVT    
 1 (n = 9)1 (11.1)1  
 2 (n = 15)4 (26.7)3.06 (0.34–27.5)  
 3 (n = 19)6 (31.6)2.99 (0.36–24.9)  
 ≥4 (n = 21)9 (42.9)3.33 (0.42–26.6)0.61 
Superior sagittal sinus occlusion    
 No (n = 23)5 (21.2)1  
 Yes (n = 41)15 (36.6)1.27 (0.42–3.83)0.67 
Transverse sinuses occlusion    
 No (n = 25)7 (41.2)1  
 Yes (n = 39)12 (30.8)0.73 (0.29–1.86)0.51 
Elevated CRP level (>10 mg/liter)    
 No (n = 24)7 (29.1)1  
 Yes (n = 40)13 (32.5)1.13 (0.3–4.17)0.86 
HLA–B5    
 No (n = 40)15 (37.5)1  
 Yes (n = 13)3 (23.1)0.66 (0.19–2.29)0.51 
Associated prothrombotic factor    
 No (n = 35)6 (17.1)1 1
 Yes (n = 16)9 (56.2)4.5 (1.49–13.58)0.0084.87 (1.54–15.4)
Tobacco use    
 No (n = 42)11 (26.2)1  
 Yes (n = 22)9 (40.9)2.06 (0.81–5.21)0.13 
Oral contraceptive use    
 No (n = 14)8 (36.4)1  
 Yes (n = 8)2 (25)0.66 (0.14–3.09)0.59 
Anticoagulation    
 No (n = 2)1 (50)1  
 Yes (n = 62)19 (30.6)0.73 (0.1–5.49)0.76 
Duration of anticoagulation, years    
 <1 (n = 6)2 (33.3)1  
 1–3 (n = 15)4 (26.7)0.28 (0.03–3.11)  
 >3 (n = 41)14 (34.1)0.66 (0.15–2.92)0.53 
Aspirin    
 No (n = 35)9 (30)1  
 Yes (n = 29)9 (31)0.82 (0.32–2.09)0.67 
Immunosuppressants    
 No (n = 32)6 (19.4)1  
 Yes (n = 32)14 (43.8)2.42 (0.93–6.3)0.07 
Steroids    
 No (n = 10)3 (30)1  
 Yes (n = 54)17 (31.5)1.11 (0.32–3.8)0.87 
Colchicine    
 No (n = 3)1 (33.3)1  
 Yes (n = 61)19 (31.1)0.49 (0.06–3.81)0.49 
Complete recanalization of CVT    
 No (n = 24)11 (45.8)1  
 Yes (n = 33)9 (27.3)0.58 (0.24–1.4)0.23 

We next analyzed factors associated with relapse of CVT. A trend toward a positive association was found between relapse of CVT and associated prothrombotic risk factors (HR 2.8, 95% CI 0.42–21.86; P = 0.21), whereas a trend toward a negative association was found between relapse of CVT and total sinus recanalization of CVT (HR 0.29, 95% CI 0.02–1.99; P = 0.22).

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. REFERENCES

Since the first description of neurologic involvement in BD in 1941 (18), only single cases and small series of CVT in patients with BD have been reported (7–13). Certain features of the disease have been hinted at in most of these reports. Here we present the largest series in an effort to provide a definitive clinical picture and outcome of CVT in patients with BD and to define factors that affect prognosis.

The current study provides information about the prevalence of CVT in patients with BD. Among a cohort of 820 patients with BD, CVT was present in 64 (7.8%). Previous studies found similar results, with a prevalence of CVT from 10–12% among patients with neuro-BD (8, 19). Conflicting results have been reported in other studies (20, 21). In a prospective study of 323 consecutive patients with BD, only 2 cases of CVT were found among 17 patients with neurologic symptoms (4). In a review of Japanese autopsy registry data consisting of 170 cases of BD, CVT was present in only 4 of 58 cases with pathologic alterations of the CNS (6). Kidd et al reported CVT in 2 of 50 cases with neuro-BD, and 4 patients presented with isolated intracranial hypertension (7).

Our data shows that CVT related to BD has a characteristic clinical picture. The patient was frequently a young male with an approximately 5-year evolution of BD. However, CVT might be the initial feature of BD in up to 20% of patients. Onset was usually progressive, with persistent headache evolving over a few days, and less frequently, fever or focal deficit. Strikingly, only 3 cases of concurrent meningoencephalitis were noted, stressing the exceptional association between CVT and parenchymal CNS involvement in BD (8, 19). Among our patients with BD, when we compared those with and without CVT, we found a 5 times lower frequency of parenchymal CNS involvement in those with CVT. Superior sagittal sinus was the main site of CVT occlusion, accounting for up to 60% of the CVT.

To our knowledge, there has not been a prospective placebo-controlled trial of treatment in patients with BD with neurologic complications. Treatment of CVT in patients with BD is not standardized either. Long-term followup has not been available in most reports. Treatment for dural venous sinus thrombosis involves anticoagulation, and some authors advocate the concurrent use of corticosteroids (3, 13, 22) and/or immunosuppressants (23) in patients with BD. In the present study, most of the patients rapidly improved following anticoagulation without severe hemorrhagic complications, underscoring the beneficial effect of such therapy. These results challenge the recently published recommendations for the treatment of CVT in patients with BD, advocating the use of corticosteroids and stating that anticoagulation should be avoided due to the possibility of fatal bleeding (24). Neither steroid use nor immunosuppressive regimens provided better outcome in terms of sequelae or to prevent thrombotic relapse. Up to 30% of patients treated with steroids and/or immunosuppressants experienced a relapse of thrombosis and/or sequelae. Compared with anticoagulation therapy alone, immunosuppressants did not improve the neurologic outcome or the thrombosis relapse rate in our study. In a recent study, steroids in the acute phase of CVT unrelated to BD were not useful in patients without parenchymal cerebral lesions (25). The frequency of multiple thrombotic events in patients with BD and CVT provides a rationale for the use of a prolonged anticoagulation at least in those with associated prothrombotic risk factors.

The present study allows us to find factors associated with a worse prognosis of CVT (i.e., occurrence of sequelae and relapse of thrombosis). By multivariate analysis, patients with papilledema related to CVT had an approximately 7 times higher rate of sequelae. Papilledema was observed in 62.5% of our patients. Severe visual loss due to optic atrophy was the main complication of CVT, accounting for 43.5% of the whole sequelae and being found in 15% of the entire cohort. This is in sharp contrast with the low rate (0.5–1%) of severe visual loss usually found in large series of CVT (26). Clinical and neuroimaging features of CVT in patients with BD are similar to those of patients with CVT of other origins (15, 27). However, an atypical presentation of CVT is not unusual in BD, raising the possibility of delayed or misdiagnosed CVT. Fever was found in one-third of our patients at diagnosis of CVT. Severe visual impairment attributed to bilateral optic atrophy has been described in patients with BD and intracranial hypertension (28, 29).

Thirty-one percent of BD patients experienced a relapse of thrombosis, of which 11% had a relapse of CVT. This was much higher than the 3–4% of recurrent thrombotic events found in large studies of CVT unrelated to BD (26, 30). More than half of the thrombosis relapses occurred in the first 5 years following diagnosis of CVT. Complete recanalization of CVT was observed in two-thirds of our patients with BD and might prevent relapse of CVT. Logistic regression analysis shows that patients with peripheral venous thrombosis or with concurrent prothrombotic risk factors had 2.8 and 4.9 times higher odds of thrombosis relapse, respectively.

Previous studies (31–33) reported on the association between major vessel involvement and CVT in BD, raising the hypothesis of a vasculo-Behçet's subtype of the disease (34). Along this line, we found a significantly higher frequency of extraneurologic vascular lesions among patients with CVT compared with patients without CVT. In the present study, 56% of patients who experienced a relapse of thrombosis following CVT had an associated prothrombotic factor. We found a combination of prothrombotic disorders in one-third of our BD patients with CVT. Hyperhomocysteinemia was found in up to 12% of patients and can be treated directly with vitamin supplementation. Interestingly, concurrent prothrombotic risk factors were also independently associated with the occurrence of sequelae. Siligardi et al (35) found a higher frequency of prothrombin gene mutation in BD patients with ocular disease, and particularly in those with severe eye involvement. Our results are also reminiscent of those found by Carhuapoma et al (36) who reported, in a series of patients with CVT and anticardiolipin antibodies, concurrence of prothrombotic disorders in 75% of cases. A similar coexistence of risk factors was previously described in a large series of CVT unrelated to BD (26).

Since the earliest reports, it has been known that neuro-BD has a high mortality rate, and in the 1960s this was estimated to be ∼25% within the first year after neurologic onset (37). In more recent series of neuro-BD, the mortality rate was nearly 10% (7, 8). This was much higher than the 10-year mortality rate of a general BD series from Turkey, which was ∼5% (38). It is well established that CNS disease in the form of CVT has a much better prognosis than parenchymal CNS involvement (8, 19). Our series confirms that the prognosis of CVT in patients with BD is good following satisfactory treatment. After a mean followup of 8 years, only 4 deaths (6%) of 64 patients were noted that were unrelated to CVT but by an underlying condition (i.e., myocardial failure and CNS involvement of BD). These results are in agreement with those found in large series of CVT unrelated to BD, where death occurred in 5–10% of patients (26, 30).

In conclusion, CVTs are found in nearly 10% of BD patients. Anticoagulation represents a safe and effective therapy. Although the functional outcome of CVT is better than arterial strokes, CVT may result in serious neurologic outcomes. Early recognition and treatment may reduce poor outcomes.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. REFERENCES

Dr. Saadoun had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Saadoun, Wechsler, Resche-Rigon.

Acquisition of data. Saadoun, Wechsler, Trad, Le Thi Huong, Sbai, Dormont, Amoura, Cacoub, Piette.

Analysis and interpretation of data. Saadoun, Wechsler, Resche-Rigon, Piette.

Manuscript preparation. Saadoun, Wechsler, Resche-Rigon, Le Thi Huong, Cacoub, Piette.

Statistical analysis. Resche-Rigon.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. REFERENCES