ABSTRACT
When the CVST was detected in young men patients with headache, BD should come to mind and patients should also be evaluated in this respect. As the risk of extracranial vascular involvement increases in such patients, they should be systematically evaluated in terms of vascular involvement. Although CS treatment could not be stopped completely with anti TNF-α drugs, it could be reduced. Significant radiological improvement was observed with anti-TNF-α treatment.
Headache (n=20, 83.3%) was the most common complaint at admission in patients with BD-associated CVST. In comparison between BD cases with and without CVST, extracranial vascular involvement was more frequent in the BD-associated CVST patients (p=0.03). Twenty-three (95.8%) patients received corticosteroid (CS) treatment. A total of 8 BD-associated CVST patients received anti-tumor necrosis factor-alpha (anti TNF-α) treatment. There was no significant difference in the rate of CS use before and after anti-tumor necrosis factor-α (TNF-α) treatment (p=0.345) but CS dose was significantly reduced after treatment (p=0.018). When patients with BD- associated CVST who receive or did not receive anti-TNF-α treatment compared, on cranial imaging, thrombosis was significantly less in patients received anti TNF-α treatment than who did not (p=0.02).
In this single-center retrospective study, we reported a series of 24 consecutive CVST patients (20 males and 4 females; mean age 34±12 years) were diagnosed with BD according to international study group criteria. The control group included 36 (24 males and 12 females; mean age 32±8 years) consecutive patients with BD without CVST from the same center.
This study aims to analyze the clinical, laboratory findings, treatments and prognosis of Behçet’s disease (BD)-associated cerebral venous sinus thrombosis (CVST) and to compare the clinical features of BD patients with and without CVST.
Introduction
Behçet’s disease (BD) is a systemic vasculitis that can involve many tissues; patients often develop oral and genital ulcers.[1,2] Although many organs and systems can be affected, neurological involvement (neuro-BD) is uncommon. Both the central and peripheral nervous systems may be affected in such patients. The central nervous system (CNS) involvement may be parenchymal or non-parenchymal; the latter includes cerebral venous sinus thrombosis (CVST), arterial occlusion, and/or aneurysms. Although CVST is a major manifestation of non-parenchymal involvement, only a few case reports and few retrospective clinical series have described the treatment and long-term outcomes thereof. Primary treatment of neurological involvement in BD undoubtedly requires immunosuppression with or without corticosteroid (CS). Anti-tumor necrosis factor-α (anti-TNF-α) agents are now used for BD complications but experience on their effect on CVST is limited.[1,3,4]
Since data on the clinical features, disease course and outcome of BD-related CVST is limited, we tried analyzing the clinical, laboratory findings, treatments and outcomes of BD-associated CVST and compare the clinical features of patients with BD with and without CVST who were followed up in a single centre.
Materials and Methods
Patients
Between June 2014 and January 2018, patients diagnosed with BD according to the 1990 guidelines of the International Study Group or the International Criteria for Behçet’s disease were included. CVST was diagnosed based on typical clinical features, magnetic resonance venography (MRV), cranial magnetic resonance imaging (MRI) features, and neurologist opinion. Patients with comorbidities (trauma, infection, malignancy, or oral contraceptive use) were excluded. The control group included thirty-six consecutive patients with BD without CVST from the same period. CVST patients were compared as patients who received and did not receive anti TNF-α treatment.
Methods
Data on 24 patients with BD-associated CVST and the 36 controls were retrieved from our electronic database and retrospectively reviewed. We recorded age at the onset of problems, and at diagnosis, along with disease duration, laboratory parameters, the cranial and extracranial locations of thromboses as revealed by imaging, and the course of cranial imaging over the years. The clinical findings were compared to those of the controls. Immunosuppressant (IS) and anticoagulant treatments prescribed at diagnosis, along with treatment duration and changes therein during follow-up, were recorded. Patients receiving anti-TNF-α treatment were examined in detail in terms of previous and ongoing treatments, treatment duration, and long-term results. The disease activity was evaluated by measuring the levels of clinical and acute phase reactants, imaging results and, patients on anti-TNF treatment, Behçet Disease Current Activity Form (BDCAF) scores. We contacted all patients by telephone.
The remission criteria were an absence of BD-related symptoms, no new vascular involvement or progression of existing vascular involvement, normal levels of acute phase reactants, and a methylprednisolone dose ≤10 mg/day for 3 months. Recurrence was defined as the development of new symptoms and/or sufficient progression of a pre-existing BD-related symptom to necessitate an increase or change in treatment. Infliximab was intravenously administered at an induction dose of 5 mg/kg during weeks 0, 2, and 6, and every 8 weeks thereafter. Adalimumab was given subcutaneously (40 mg every 2 weeks) along with etanercept (50 mg weekly).
The approval of the institutional ethical committee was obtained (University of Health Sciences Turkey, Gazi Yaşargil Training and Research Hospital, approval number: 541, date: 11.09.2020).
Statistical Analysis
Histograms and probability plots were generated. We used the Kolmogorov-Smirnov or Shapiro-Wilk test to determine whether variables were normally distributed. Continuous variables with normal distributions are expressed as mean ± standard deviation (SD); all other variables are expressed as median values (range). Categorical variables were compared using the chi-squared or Fisher’s exact test. Continuous variables were compared using Student’s t-test. Non-normally distributed continuous variables were compared using the Mann-Whitney U test and the Wilcoxon test. A p-value <0.05 was considered statistically significant. All tests were performed using SPSS for Windows software (ver. 22.0; IBM Corp., Armonk, NY, USA).
Results
Demographic Characteristics and Clinical Features
In total, 24 (20 males and 4 females) of 571 patients with BD (4.2%) were diagnosed with BD-associated CVST; 54 exhibited neurological involvement (9.4%), and headache [n=20 (83.3%) patients] was the most common complaint at admission. Three (12.5%) patients presented with oral aphthae and 1 (4.2%) patient with hemoptysis due to pulmonary artery aneurysm at admission.
Of the patients with CVST, the most common BD-related symptoms were oral aphthae in 24 (100%) and genital ulcers in 21 (87.5%) (Table 1).
Neurological Imaging
CVST was diagnosed using MRV and cranial MRI. A neuroradiologist (K.A.; 8 years of experience) reviewed all brain MRI scans while blinded to the clinical and laboratory data, and treatments. Multiple sinus involvement (19, 79.2%) was more common than single sinus involvement (5, 20.8%). No patient had a cerebral infarction, bleeding, or permanent neurological damage. The parenchymal involvement was observed in 32 (5.6%) patients and 2 (0.3%) exhibited both parenchymal involvement and CVST. All patients had control imaging including brain MRI and MR venography. In 10 of 23 patients at their final visit, CVST was continuing on MR venography and brain MRI.
Extracranial Thrombosis
In total, 16 (66.7%) patients exhibited extracranial vascular involvement; lower extremity deep vein thrombosis was seen in 8 (33.3%) patients, jugular vein thrombi in 7 (29.2%), and pulmonary thromboembolism in 6 (25.0%) (Table 1).
Laboratory Findings
The erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) level were elevated in 20 (83%) patients at admission. The mean ESR at presentation was 54.08 (0-20) mm/h, and the mean CRP level 66.7 (0-5) mg/dL. After 1 year, the ESR was higher in nine (39%) patients and the CRP level was higher in six (26%); the mean ESR and CRP level were 19.04 mm/h and 9.26 mg/dL, respectively. Compared to baseline, the values decreased significantly after 1 year (both p<0.01). Antinuclear antibody status was negative in 15 patients and weakly positive (≤1:160) in 2 patients. Anti-extracted nuclear antigen antibody status was negative in all patients. Of the 16 patients evaluated, all were negative for antiphospholipid antibodies (anticardiolipin antibody immunoglobulin G (IgG)/IgM, lupus anticoagulant, and anti-beta-2 glycoprotein (anti-ß2GPI) IgG/IgM]. HLA B51 analysis was performed in six (25%) patients; four (16.7%) were positive.
Treatments and Outcomes
Treatments on CVST Diagnosis
Previous treatments included ISs and anticoagulants. In total, 16 (66.7%) patients with BD-associated CVST were prescribed colchicine and 14 (58.3%) were prescribed azathioprine. Twenty-three (95.8%) patients received CS (methylprednisolone or an equivalent) treatment. Eighteen (75%) patients received pulsed CSs and five (20.8%) received CS at a starting dose of 1 mg/kg/day, which that was gradually tapered and stopped. The pulsed protocol was methylprednisolone or an equivalent at 1 g/day for 3-5 days (induction), followed by 1 mg/kg/day (maintenance); the CS was then tapered or stopped. Eleven patients (45.8%) received pulsed cyclophosphamide (CYC) treatment (1 g per month, intravenously for 6 months, or 500 mg every 15 days). Anti-TNF-α (adalimumab) therapy was prescribed for one (4.2%) patient (40 mg subcutaneously every 2 weeks). Low-molecular-weight heparin (enoxaparin, 100 IU/kg x 2) was given during the acute period and then replaced with warfarin (used only for maintenance treatment). Before anticoagulant treatment, all patients were evaluated in terms of pulmonary artery aneurysms. Anticoagulant treatment was prescribed for 19 (79.2%) patients. All patients received enoxaparin (Table 1).
Treatment During Follow-up
Patients who received current treatments were evaluated at outpatient clinic visits every 1-3 months according to their clinical status after discharge. Ten of 16 (66.7%) patients who received colchicine treatment was switched to azathioprine and/or anti TNF-α treatment after one year because they were colchicine-resistant.
Initially, 18 (75%) patients received pulsed CS treatment. Two (8%) patients received repeat-pulsed treatments because of recurrence during follow-up. Of 23 (95.8%) patients prescribed CSs, the doses were reduced and ultimately discontinued in 16 (69.5%). At the last visit, 7 of 23 patients were receiving low dose (<10 mg/day) CS treatment.
Initially, 19 (79.2%) patients were prescribed anticoagulants, of whom 16 (66.7%) were followed up with warfarin only; 10 (43.5%) continued on both anticoagulants and warfarin. None of the 11 (45.8%) CYC-treated patients received a second CYC treatment after 6 months; all were switched to maintenance azathioprine. At the beginning, 14 (58.3%) patients received azathioprine treatment. Number of patients who were treated with azathioprine increased to 16 (66%) after one year. No patient who received colchicine, azathioprine, CYC and CS treatment developed any drug-related complications. Seven (29%) patients were switched to anti-TNF-α treatment because they were resistant to other treatments.
Analysis of CVST Patients Receiving Anti TNF-α Treatment
The mean age of eight patients (five males and three females) who received anti-TNF-α treatments was 30.6±8.3 years; six of these patients were prescribed infliximab; one each received adalimumab and etanercept. The mean disease duration, which was 8.5 [minimum-maximum (min-max: 1-108)] months before anti-TNF-α treatment, increased to 51 (min-max: 10-70) months after anti-TNF-α treatment. The mean duration of CS use before anti-TNF-α treatment was 4 (min-max: 0-12) months and that after TNF-α treatment 2.5 (min-max: 0-45) months. There was no significant difference in the rate of CS use before and after anti-TNF-α treatment (p=0.345). The mean dose of CS before anti-TNF-α treatment was 34.1±22.2 mg/day, and that after anti-TNF-α treatment was 2.7±1.7 mg/day; the CS dose was significantly reduced after treatment (p=0.018). Although seven patients had received anticoagulants for an average of 3 months before anti-TNF-α treatment, this was then discontinued in six; only one patient received anticoagulants for 60 months (Table 2).
Seven of 8 CVST patients received anti TNF-α treatment had received CS treatment before anti TNF-α. At the last visit, 2 patients were continuing to receive CS treatment. When the patients received and did not receive anti TNF treatment were compared in terms of CS and anticoagulant use, there was no significant difference (p=0.679 and p=0.101, respectively) (Table 3).
During the follow-up period, all 23 patients were evaluated for CVST by MR venography and brain MRI. It was found that thrombosis continued in 10 patients. At the last visit of anti-TNF-α treatment, the cerebral venous thrombosis evident on cranial MRV disappeared in seven of the eight patients and did not recur (Figures 1 and 2). The thrombosis decreased in the other patient (Figure 3). On cranial imaging, thrombosis was significantly less in patients received anti TNF-α treatment than who did not (p=0.02) (Table 3). No patient who received anti-TNF-α treatment developed any permanent neurological sequelae or drug-related complications.
Comparison Between BD Cases with and without CVST
The control group included thirty-six consecutive patients with BD without CVST from the same period; they all had oral aphthae. Skin involvement was more common in the controls, while the extracranial vascular involvement was more frequent in the patients with BD with CVST (p=0.03 and p=0.03, respectively) (Table 4).
Discussion
Of all patients with BD, 9.4% showed neurological involvement. The incidence rates of CVST were 4.2% in BD and 44% in neuro-BD patients. Anti-TNF-α treatment was relatively safe and effective in patients with BD-associated CVST; most patients exhibited no recurrence. The frequency of neuro-BD ranges from 5% to 35%.[5] In an autopsy series, 20% of 170 patients with BD exhibited neurological involvement.[6] Bolek et al.[7] reported a neuro-BD rate of 18.4%. Among a cohort of 820 patients with BD, 64 (7.8%) had CVST. The prevalence of CVST in neuro-BD patients varies from 10% to 20%.[8,9] In this study, the rate of neurological involvement among all patients with BD was 9.4%. The incidence of CVST in patients with BD was 4.2%, compared to 44% in neuro-BD patients. Although some studies found that CVST was more common in males (65.6% and 68.5% of all patients), others reported no gender difference.[4,8,10] We found that the disease was significantly more common in young males (83.3%). The CVST disease duration was approximately 6.5 years. The most common initial complaint was headache; most patients presented to the neurology clinic. BD is the most common cause of CVST in some Middle Eastern countries.[11,12] The most common findings of BD-associated CVST are headache, focal neurological deficits, and changes in consciousness, which are mainly attributed to intracranial hypertension. Saadoun et al.[8] found that the initial complaint was often a severe headache that had developed a few days earlier, followed by fever and focal deficit. In an Italian study, the most common presentation of neuro-BD was an acute attack with motor symptoms and cognitive changes.[13] In contrast, most of our patients complained of sub-acute chronic and progressive headache, although some also showed behavioral changes. The parenchymal involvement was seen in only two CVST patients. No patient exhibited permanent neurological deficits, either at admission or during long-term follow-up. The headache improved in most patients. CVST patients may develop visual problems caused by neurological involvement or uveitis.[14] Eight of our CVST patients had uveitis, but none suffered permanent vision loss. As the gold standard imaging modalities, MRV and cranial MRI, which were used in this study, play critical roles in CVST diagnosis. Approximately 80% of our patients exhibited involvement of more than one sinus. The transverse sinus was the most commonly involved, followed by the superior sagittal sinus, consistent with the literature.[4,15] Two-thirds of patients exhibited extracranial vascular involvement: arterial involvement (pulmonary artery aneurysms and thrombi) was seen in eight patients, while venous and cardiac thromboses were observed in all other patients (mostly deep vein thromboses in the lower extremities). Compared to the patients with BD without CVST, extracranial vascular involvement was more common in CVST patients. As reported by Shi et al.,[4] among others, when BD is complicated by thrombosis, any vascular thrombus increases the risk of other thromboses.[16] Tunc et al.[17] found that, in BD patients with CVST, the risk of extracranial vascular involvement was high. Therefore, after the diagnosis of BD-associated CVST, all patients should be thoroughly evaluated for possible vascular involvement of other organs (including the heart). Although CSs are the cornerstone of treatment for neuro-BD, colchicine, azathioprine, and CYC serve as steroid-sparing agents that prevent relapse.[18] High-dose CS and CYC treatments are recommended for patients with BD exhibiting major organ involvement, but long-term use is limited by toxicity. In our study, five (20%) patients received high-dose CSs and 2 (8%) were given CYC before anti-TNF-α treatment.
TNF-α is a major proinflammatory cytokine that plays a critical role in the pathogenesis of BD. It is widely accepted (including by the European League Against Rheumatism) that anti-TNF- agents (usually infliximab, adalimumab, and etanercept) are effective for treating BD.[19-21] In our study, six (25%) patients received infliximab, while one (4%) received adalimumab and another (4%) etanercept. There was no recurrence in any patient during 43 months of anti-TNF-α treatment. Significant clinical and laboratory improvements were evident by 3 months. Seven patients had received CSs for an average of 4 months before anti-TNF-α treatment, and initially remained on the CSs. However, these were discontinued after 3 months in four patients, and after 24 months in the remaining patient.
At the final follow-up, only two patients were taking ≤4 mg/day methylprednisolone or equivalent. Although the duration of CS use before and after anti-TNF-α treatment did not differ between the patients (p=0.345), the CS dose decreased significantly after anti-TNF-α treatment (mean ± SD, 34.1±22.2 to 2.7±1.7 mg/day, p=0.018). In patients with BD-associated CVST and vascular BD, it is unclear how long anti-TNF-α treatments should be given. Aksoy et al.[22] reported recurrence in two of three patients with vascular BD who discontinued anti-TNF-α treatment. Our patients were followed up for an average of 43 months; all entered remission during month 3 and none suffered recurrence. The mean BDCAF score was 2.2±1.4 at baseline and decreased significantly to 0.5±0.9 at month 3, and 0.3±0.5 at the last follow-up (both p=0.01). Cranial MRI revealed complete resolution of the thromboses of seven patients, but significant recanalization was seen in one case. No patient had any treatment-related complications. In a series of neuro-BD patients, the mortality rate was reported to be around 10%. Neuro-BD patients with CVST have better prognoses than those with parenchymal involvement.[23,24] Saadoun et al.[8] reported only 4 deaths (6%) among 64 patients followed up for a mean of 8 years; diseases other than BD (e.g. myocardial failure) or parenchymal neurological involvement attributable to BD caused the deaths. The mortality rate was similar to that of patients with CVST unrelated to BD.[25] In our study, one (4%) patient died of myocardial infarction outside the hospital. Our findings are consistent with the literature. Although the use of anti-TNF-α drugs to treat parenchymal neuro-BD has received some attention,[7,26] their use in neuro-BD patients with CVST is limited and the studies were retrospective. Although the number of patients is small, we found that long-term anti-TNF-α agents can be safely used to treat CVST, which is the best-known form of non-parenchymal neuro-BD. Although CS treatment could not be stopped completely with anti TNF-α drugs, it could be reduced.
Study Limitations
Our study was retrospective and no disease activity index was universally applied to evaluate neuro-BD activity. Therefore, we used the BDCAF index adopted in other studies to evaluate anti-TNF-α treatment efficacy. Other patients were evaluated according to the absence of CVST-related disease activity, the laboratory and imaging results, and the CS doses required.
Conclusion
As the risk of extracranial vascular involvement increases in BD-CVST patients, they should be systematically evaluated in terms of vascular involvement. Although the number of patients is small, we found that long-term anti-TNF-α treatment was safe and effective. Although CS treatment could not be stopped completely with anti TNF-α drugs, it could be reduced. Significant radiological improvement was observed with anti-TNF-α treatment.
Ethics
Ethics Committee Approval: The approval of the institutional ethical committee was obtained (University of Health Sciences Turkey, Gazi Yaşargil Training and Research Hospital, approval number: 541, date: 11.09.2020).
Informed Consent: Retrospective study.
Peer-review: Internally peer-reviewed.
Authorship Contributions
Concept: L.A., M.S., Design: L.A., Data Collection or Processing: L.A., K.A., C.G., Analysis or Interpretation: L.A., K.A., M.Ö., M.S., Literature Search: L.A., Writing: L.A.
Conflict of Interest: No conflict of interest was declared by the authors.
Financial Disclosure: The authors declare that they have no relevant financial.
