Anticoagulation delay does not affect the functional outcome of cerebral venous thrombosis

Introduction

Cerebral venous thrombosis (CVT) is a unique type of cerebrovascular disease that varies greatly in its clinical course and prognosis, from complete recovery to disability or death [1, 2]. CVT is usually treated with anticoagulant therapy. Timely and effective anticoagulant therapy can prevent thrombus growth, promote thrombus regression, and improve patient outcome. Conversely, a delay in starting anticoagulation may allow progressive thrombosis, a worsening of clinical status, a more complicated condition, and/or an increased thrombus burden, all of which are detrimental to the outcomes of CVT patients. Unfortunately, the diverse clinical manifestations and neuroimaging features of CVT often lead to delays in admission and diagnosis, which in turn lead to delays in anticoagulation.

Delay in admission and diagnosis in CVT has been reported in previous studies [3, 4], but each of them is not equivalent to anticoagulation delay. First, it may take several days from admission to confirmed diagnosis or anticoagulation, during which time clinical worsening may occur [5, 6]. Second, even after diagnosis, some patients are not treated with anticoagulant therapy in the above studies. Thus, clinical outcomes may be confounded by a lack of uniform therapy. Additionally, the effect of anticoagulation delay on residual headache, as one of the most common sequelae specific to CVT, has not been evaluated.

In the present study, therefore, we sought to identify the potential variables that influence anticoagulation delay and to analyze the effects of this delay on outcomes in a large single-center cohort of patients with CVT. Of particular interest was the effect of anticoagulation delay on residual headache.

Results

Between January 2014 and July 2018, a total of 179 CVT patients were enrolled in the registry. Among them, 5 patients were excluded because of our inability to determine the day of onset due to an insidious clinical course and a history of chronic headache, while 10 patients less than 14 years old were also excluded. Therefore, the remaining 164 eligible patients who were treated initially with heparin and had valid information on anticoagulation delay were included in the present study. The median age of these patients was 35 years (interquartile range [IQR], 25-49), and 53% were female. The initial anticoagulation treatment was subcutaneous injection of low-molecular-weight heparin (LMWH) in 134 patients and intravenous unfractionated heparin (UFH) in 30 patients.

Discussion

In the present study, we systematically investigated the impact of anticoagulation delay on outcomes of patients with CVT. Cerebral hemorrhage on admission neuroimaging was identified as predictive of earlier anticoagulation. Anticoagulation delay had no significant effect on long-term functional outcomes in CVT patients. However, the anticoagulation delay increased the risk of persistent residual headache in both the entire cohort and the subgroup with isolated intracranial hypertension.

Cerebral hemorrhage was the baseline variable identified in our study that was associated with earlier anticoagulation. This result was consistent with the earlier observation that patients with more severe clinical presentation tended to be admitted to the hospital sooner and treated more actively, resulting in earlier diagnosis [4]. According to current guidelines, acute CVT should be treated immediately with initial heparin therapy, regardless of pre-treatment intracranial hemorrhage [1, 7]. It can therefore be reasonably assumed that admission or diagnosis delay is the direct cause of anticoagulation delay. Isolated intracranial hypertension syndrome, which was associated with admission or diagnosis delay in previous studies [3, 4], was not identified as predictive of anticoagulation delay in our study. This discrepancy may reflect the different distribution of diagnoses or anticoagulation delays among our cohort (the median diagnosis delay was 9 days [data not shown]). In our sample, nearly 50% of patients with seizure or any parenchymal lesion were diagnosed later (over 9 days), leading to a longer median delay in the entire cohort, whereas the median diagnosis delay in the abovementioned earlier study was only 7 days [4].

Anticoagulation delay had no significant impact on long-term functional outcomes of CVT patients because those with severe clinical manifestations tended to be admitted earlier and thus diagnosed and treated earlier. By contrast, patients with a relatively mild presentation (i.e., isolated intracranial hypertension syndrome or isolated headache) were always admitted later [3, 8]. Consequently, the benefits of earlier anticoagulation, if any, were difficult to assess across the entire CVT cohort. However, our results show that anticoagulation delay significantly increased the incidence of residual headache after CVT. The incidence of residual headache during the follow-up period varied in previous studies [9–13]. In a report from the Cerebral Venous Thrombosis Portuguese Collaborative Study Group, the incidences of any residual headache and severe headache after CVT were 47% and 8%, respectively [9]. In the International Study on Cerebral Venous and Dural Sinuses Thrombosis cohort, 14% of patients with CVT had severe headache at the end of the follow-up [10]. In a recent analysis of 161 CVT patients from a single center, 20% of survivors reported remaining frequent headaches [12]. In our cohort, 28% of CVT survivors experienced frequent or severe headaches. These incidences are difficult to compare due to the different definitions of remaining headaches. The cause of residual headaches after CVT is unclear, though remaining sinus occlusion, post-thrombotic sinus stenosis, or persistent intracranial hypertension could be the reason for residual headaches in some cases [14].

One limitation of the present study is that valuable information on the cause of residual headaches after CVT, such as persistent raised intracranial pressure or occluded sinuses, was not analyzed and warrants further study. We therefore cannot draw a conclusion about the mechanisms by which anticoagulation delay may lead to more frequent or severe headaches. Another limitation of the study is the difficulty to ascertain the exact day of onset in some of cases with a gradual clinical course or a history of chronic headache. Additionally, given the nature of our single-center design, the generalizability of our results may be hindered. Nonetheless, the strength of our study is the large sample and uniform treatment approach, which minimized outcome confusion from various treatments.

Our findings have several implications for practice. It may be appropriate to pay more attention to remaining headaches in CVT patients who received anticoagulation later. Routine follow-up imaging of the cerebral venous system to re-evaluate potential causes of the headaches may be a reasonable approach for patients with remaining frequent or severe headaches.

In conclusion, anticoagulation delay was found to be common among patients with CVT. Anticoagulation delay was not associated with poor functional outcome, but may have led to an increased risk of residual headaches across our entire cohort and in the subgroup with isolated intracranial hypertension.

Materials and Methods

Supplementary Materials

Supplementary Table