Cerebral Venous Sinus Thrombosis: Clinical Characteristics and Factors Influencing Clinical Outcome

In recent years, cerebral venous sinus thrombosis (CVST) has been diagnosed substantially more frequently than in the past due to the expanded use of noninvasive brain imaging methods. However, the actual incidence of CVST is most likely even higher. Many cases remain unrecognized due to the high variability of their clinical presentation and the possible absence of any symptoms. In the most recent publication, which was based on an analysis of a cohort of 9270 patients examined between January 2008 and December 2010 in the Netherlands, the reported incidence of CVST was 1.32/100 000 individuals/year; this incidence was the highest in females aged 31 to 50 years (2.78/100 000).¹ This report indirectly confirmed the results of the International Study on Cerebral Vein and Dural Sinus Thrombosis (ISCVT) performed in 2004, which examined the largest cohort of patients with cerebral phlebothrombosis to date and which reported that females consisted of 74.5% of those experiencing CVST.²

The finding that the ratio of CVST incidence in females and males is 3:1 at young and middle age can only be explained by gender-specific risk factors (GSRFs), which most often include hormonal contraception (HC), pregnancy, puerperium, and hormone replacement therapy (HRT). These factors are likely responsible for the better outcome of CVST in females than in males, despite the same thrombosis severity and a higher frequency of parenchymal lesions.³ Aside from male gender, impaired consciousness at presentation, deep vein thrombosis, and parenchymal hemorrhage were identified as predictors of unfavorable outcome in the ISCVT study.⁴

The aim of our study was to assess the correlations between the location and the extent of venous sinus impairment, clinical presentation during the acute phase, recanalization, the presence of parenchymal lesions, and the clinical outcome in patients with CVST. The identification of (primarily modifiable) outcome predictors and their subsequent control are essential to achieve good clinical outcome in these patients.

The most valuable source of data concerning CVST remains the ISCVT study published in 2004, which analyzed data collected from 624 patients at 89 centers in 21 countries between 1998 and 2001.² Excluding this study, CVST cohorts have been relatively small and have been recruited from individual centers.^6,8,9

The most frequently affected sinuses in our cohort were the left transverse and sigmoid sinus, possibly as a consequence of their morphological inferiority (often being hypoplastic),⁴ followed by the superior sagittal sinus. The association of transverse and sigmoid sinus thrombosis with local inflammation in the nasal, tonsillar, and middle ear area, which was commonly observed in the past, has recently been rare.¹⁰ We confirmed this finding, reflecting a globally reduced incidence of inflammatory phlebothrombosis, due to the increased use of antibiotic treatment. In the majority of our patients, multiple sinuses were involved, and more extensive thrombosis was observed in males than in females. Previously, only 2 reports thoroughly analyzing the correlation between the extent of CVST and the risk of parenchymal complications were published. Although Bergui et al in 1999¹¹ did not detect this relationship, the Mayo Clinic report published 10 years later⁹ obtained an opposing conclusion. Of the 56 patients with CVST examined, 34% presented with a parenchymal complication (edema/infarction or hemorrhage), and the risk of complications directly correlated with the extent of thrombosis (in contrast to age, gender, and the presence of thrombophilic conditions). In the present study, no significant correlation was observed between the extent of thrombosis and either edema (detected in 43% of the cases) or brain hemorrhage (present in 29% of the cases). Nevertheless, vasogenic edema was significantly more frequently caused by thrombosis of the deep venous system and cortical veins, whereas hemorrhage was a consequence of the thrombosis of cortical veins or the right transverse or sigmoid sinus. The dominant role of deep and cortical veins in the risk of parenchymal complications is the logical based on their characteristics; these veins are “terminal,” lacking sufficient collaterals that are able to protect the brain from the increased venous pressure. Thrombosis of the right transverse and sigmoid sinus was associated with an elevated risk of hemorrhage, as confirmed in the ISCVT study,⁴ likely as a consequence of increased venous congestion due to common hypoplasia of the contralateral sinus. Thrombophilic conditions significantly contributed to more extensive thrombosis but not to the occurrence of parenchymal lesions. Although females with inherited forms of a thrombophilic condition (identified in 24% of the females) exhibited an average thrombosis extent score of 2.8, the average thrombosis extent score in males with identified thrombophilic conditions (present in 43% of the males) was 4.6. Approximately one-third of the patients with CVST had parenchymal hemorrhage, which is typically accompanied by more severe clinical symptoms at disease onset⁶ and, occasionally, by worse outcome.^2,8,12–14 In most cases, small parenchymal hemorrhages surrounded by normal tissue or by a zone of perifocal edema of varying size are observed, whereas subdural hematoma and subarachnoid hemorrhage are rare. In our cohort, 29% of the patients had hemorrhage—in all cases at the region of vasogenic edema. Although both types of brain complications (hemorrhage and edema) were associated with a more severe clinical presentation during the acute phase, accompanied by impaired consciousness in 12% of the cases, their presence did not influence long-term clinical outcome in the present study.

Generalized and focal epileptic symptoms during the acute phase occur in up to 40% of patients (in 20% of the patients in the present study), and residual epileptic syndrome persists in 10% to 30% of these patients (in only 6% of these patients in the present study).² The prophylactic use of anticonvulsants during the acute phase of CVST is controversial and has yet to be supported by any evidence. The observation that the probability of the development of a secondary epileptic syndrome is higher in patients with a supratentorial lesion, as observed in our patients, supports their use in this subgroup of patients.¹⁵

Partial or complete recanalization after 3 to 4 months was detected in 92% of our patients. In the majority of reported patients, partial or complete recanalization is achieved within the first 3 months,^16,17 after which the recanalization rate does not further increase.¹⁸ Therefore, in routine clinical practice, it may not be necessary to perform MRI imaging at long-term follow-up if a noncomplicated course of CVST is observed clinically. Successful recanalization in our cohort did not depend on the presence of a thrombophilic condition but significantly influenced clinical outcome.

The prognosis of CVST is markedly better than brain infarction due to arterial occlusion. The predictors of poor outcome (mRS > 3) observed in 13% of the patients in the ISCVT study (with a 30-day mortality rate of 3.4%) and in the National Inpatient Sample database from the 2000 to 2007 (with a mortality rate of 4.4%) include older age, brain hemorrhage, and concomitant diseases, for example, hematological pathologies, systemic malignancy, and central nervous system infection.^2,19 However, other studies reported even lower mortality—approximately 1%.¹ In the present study, as previously stated, good clinical outcome after 3 to 4 months significantly negatively correlated with a thrombosis location in the left transverse, left sigmoid, or superior sagittal sinus and positively correlated with the achievement of recanalization. However, good clinical outcome did not correlate with other factors, such as the extent of thrombosis, impairment of the deep venous system and cortical veins, impairment of consciousness, edema, or hemorrhage. Regarding the thrombosis location, our finding did not confirm those from the ISCVT study, in which poor clinical outcome among the patients having CVST with intracerebral hemorrhage was associated with thrombosis of the right lateral sinus. This discrepancy might be attributable to the larger size of this sinus, explaining that the drainage cannot be compensated by the contralateral hypoplastic side.⁴

None of the patients in our cohort died. The worst outcome (mRS grade 3) was observed in an 18-year-old male who had a pronounced worsening of visual acuity. An mRS grade of 2 was observed in 10% of the patients; this clinical outcome was due to a residual epileptic syndrome in 3 patients, memory impairment in 1 patient, and persisting alexia in 1 patient.

The individualized prediction of the disease course is extremely difficult. Nevertheless, in contrast to arterial brain infarction, the achievement of a very good recovery is possible even in severe cases. The prognosis of CVST is better during pregnancy and puerperium, as its common spontaneous regression and high recanalization rate are much more favorable than those of other conditions. In these cases, the probability of recurrence is very low (≤ 1%)^18,20; therefore, it is not necessary to discourage these patients from further pregnancies.²¹

In 2009, a subgroup analysis of the data from the ISCVT study dedicated to gender differences in CSVT was published.³ Females accounted for 75% of the cases and were significantly younger, and those (65%) with at least 1 GSRF exhibited a more favorable prognosis. Similar results were observed in the present study. Although the females with GSRFs in our cohort insignificantly more frequently exhibited a severe clinical presentation during the acute phase of the disease and experienced edema and hemorrhage than the males, they exhibited a significantly better clinical outcome than the males. Complete functional recovery (mRS 0-1) was achieved in 97% of the females with GSRFs but only in 79% of the males. Compared with females, males exhibited a higher incidence of an inherited thrombophilic condition (43% vs 24%), significantly more extensive thrombosis, and insignificantly more frequent impairment of the deep and cortical veins.

The better outcome observed in the females with GSRFs could be explained by their younger age (which is typically associated with reduced comorbidities). However, multifactorial analysis performed on the females in the ISCVT study did not identify age as a significant outcome predictor. Alternatively, the absence of GSRFs was a strong independent predictor of unfavorable outcome in females. Another explanation could be the generally better physical condition of females using HC because they are medically supervised, as HC would not be prescribed to females known to exhibit RFs for a thrombotic complication. Similarly, pregnancy is less likely or is postponed in females with health problems. In accordance with this explanation is the finding of the ISCVT study that females without GSRFs share the same risk profile as males.²² However, one may argue that less frequent systemic illness among younger females could also play a role in their more favorable outcome.

In recent years, several studies have revealed possible mechanisms by which the estrogens contained in HC increase the risk of thrombosis.^23–27 In response to the use of HC, the levels of procoagulant factors such as II, VII, VIII, X, and vWf are increased, the concentrations of protein S and antithrombin are decreased, and a resistance to activated protein C develops. The progestin component of HC also enhances the risk of venous thrombosis. However, the thrombogenic potential of HC does not exclusively depend on its composition. For the initiation of cerebral venous and sinus thrombosis, the presence of a multiple causative factors is typically required. The highest risk represents a combination of RFs at several levels—at the HC level (impact of different generations of progestins, the form of the drug, and the length of exposure), at the patient level (age, smoking status, arterial hypertension, obesity, immobility, etc), and at the inherited thrombophilic condition level.²⁷ A meta-analysis published in 2006 reported an odds ratio of 5.59 for CVST risk of HC users relative to those who do not use an HC (95% confidence interval [CI]: 3.95-7.91); in combination with a factor V Leiden or prothrombin mutation, the odds ratio increased to 30 and 79.3, respectively.²⁸ One may assume that to prevent the synergic effect of RFs, it would be beneficial, in addition to assessing for inherited thrombophilic conditions, to avoid prescribing an HC to females with other identified RFs. Although the long-term outcome of CVST was highly favorable in the vast majority of our females, it is unclear whether this result would be the same if 14% (5 of 37) of them had not undergone invasive and potentially dangerous recanalization treatment.

Several limitations of the present study should be mentioned. Its retrospective observational nature represents the first limitation. Second, a relatively small sample size and the very limited comparison of females without GSRFs make it difficult to draw conclusions regarding the impact that gender on CVST. Third, this sample size did not enable the use of logistic regression analysis for the statistical evaluation of the results. Fourth, the clinical outcomes were confounded by a lack of uniform therapy.

In conclusion, CVST represents a relatively benign disease that primarily affects young individuals. However, CVST may result in unfavorable outcome, particularly in females with GSRFs, specifically HC users, in whom CVST develops more frequently and exhibits a more severe course during the acute phase. Nevertheless, in the present study, the clinical outcome after 3 to 4 months was significantly better in the females with GSRFs than in the males. Moreover, good clinical outcome significantly negatively correlated with a thrombosis location in the left transverse, left sigmoid, or superior sagittal sinus and positively correlated with successful recanalization.

The authors thank Josef Bukač, MSc, PhD, Charles University Faculty of Medicine in Hradec Králové, for testing the statistical significance of the results of this study.