Ferro JM, Coutinho JM, Dentali F, et al. JAMA neurology. 2019 Dec 1;76(12):1457-65.
Reviewed by Wazim Mohamed, MD
Read the article.*
*You will need to log in to US National Library of Medicine National Institutes of Health to read this article.
The prevalence of cerebral venous thrombosis (CVT) is low and the primary concern after the acute phase is that of recurrent CVT or venous thrombotic events (VTE). Current recommendations for prevention of such events is through anticoagulation using vitamin K antagonists. Direct oral anticoagulants (DOACS), including dabigatran, are effective and safe when used for recurrent stroke prevention as well as treatment and prevention of recurrent VTEs. There are published case series demonstrating the safety and efficacy of DOACS in patients with CVT. However, these were small, lacked controls and randomization.
The current study was a multicenter PROBE design (prospective, randomized, parallel-group, open-label with blinded evaluation of end points) exploratory trial to evaluate the efficacy and safety of dabigatran compared with warfarin in the prevention of recurrent VTE and CVT. It was conducted in nine countries at tertiary medical centers over a period of 2.5 years.
Inclusion criteria: Patients aged 18-79 with a diagnosis of CVT on MRI plus MR venography, CT plus CT venography or intra-arterial venography.
Exclusion criteria included: Inability to swallow medication, CNS infection or major head trauma, surgical procedure for CVT (eg. decompressive hemicraniectomy), recent major bleeding, the need to continue previous anticoagulants, current/recent malignancy and creatinine clearance < 30 ml/min. Endovascular treatment was allowed. Randomization took place 5-15 days after the initial acute treatment with parenteral therapy once the patient was deemed clinically stable. Patients were randomized 1:1 to receive either dabigatran (150 mg twice daily) or warfarin (to maintain INR 2-3). This was an open-label trial, but the end point adjudicators were blinded.
The 24-week treatment period was followed by a 7-day follow-up period, for a total of 25 weeks. The end of treatment visit included an assessment of the modified Rankin scale (mRS) and recanalization assessment using a trial-specific MRI protocol. Patients that discontinued trial medications were followed up for survival, new thrombotic events and bleeding until the end of 25 weeks.
Primary outcome: Composite of number of patients with major bleeding events or VTE at 25 weeks.
Secondary outcomes: VTEs and recanalization rates.
Secondary safety outcomes: Patients with major bleeding, clinically relevant non-major bleeding, any bleeding, new intracranial hemorrhage or worsening baseline hemorrhage.
Exploratory outcomes included: mRS, VTE associated mortality and all-cause mortality. All analyses were descriptive and analyzed as intention-to-treat.
One hundred twenty patients were randomized, and each group was comprised of 55% women with a mean age of 45.2 (±13.8) years. Each group had approximately 30% of patients with a hemorrhagic lesion. The sinuses involved for the diagnosis of CVT were similar between both groups. None of the patients were comatose and only 5% of patients in each group had a GCS between 9 and 14.
The most common presentation were headaches (~90%) and seizures (21.7% dabigatran group vs. 26.7% warfarin group). The predominant risk factor was the use of oral contraceptives (30% dabigatran group vs. 31.7% in warfarin). Genetic thrombophilia was present in 8.3% of patients in the dabigatran group and 5% of patients in the warfarin group. The NIHSS was less than 4 in 96.7% of patients in each group. Overall, 90.8% of patients completed the treatment period. Seven patients in the dabigatran group (one had enlargement of baseline hemorrhage, one patient had an intestinal hematoma and other adverse events in five patients), and four patients in the warfarin group (two patients withdrew from trial, one did not reach therapeutic INR and the other for unknown reasons) discontinued medication prematurely.
The median adherence to dabigatran was 99.7% and warfarin was therapeutic 66.1% of the time.
Primary outcome: No recurrent VTEs were observed in either group, and there were three new major bleeding events for the entire cohort: one intestinal bleeding in the dabigatran group (1.7%; 95% CI 0.0-8.9) and two subdural hemorrhages in the warfarin group (3.3%; 95% CI 0.4-11.5). One patient in the dabigatran group had worsening of the baseline intracranial hemorrhage.
In the evaluation of secondary outcomes, there were no clinically relevant non-major bleeds in the dabigatran group, and one patient had genitourinary bleeding the warfarin group (1.7%; 95% CI 0.0-8.9). Any bleeding irrespective of severity occurred with the same frequency (20%) in each group (95% CI 10.8-32.3). Venous recanalization was evaluated in 55 patients in the dabigatran group and 52 patients in the warfarin group. In the dabigatran group, there was improvement in 60% (95% CI 45.9-73) and 67.3% (95% CI 52.9-79.7) in the warfarin group. In both groups, more than 90% of patients had a mRS 0-1 at 24 weeks and all remaining patients had a mRS of 2. There were no deaths in either group.
The authors conclude that the risk of recurrent VTEs in patients with CVT receiving either dabigatran or warfarin was low. Anticoagulation with either agent was associated with few bleeding episodes, new intracranial bleed or expanding intracranial hemorrhages. This was the first randomized controlled trial to assess the safety and efficacy of DOACS in CVT. The rates of hemorrhage in this trail were comparable to the dabigatran for DVT trials and less than the stroke trials. This is likely due to the population in the stroke trials being older, subject to small vessel disease and longer exposure to anticoagulation. In the current trial, there were more patients that discontinued dabigatran compared to warfarin mainly due to digestive system adverse effects.
Among its limitations, this trial was an exploratory analysis with a small sample size that was not powered to detect statistical differences. Given the low incidence of CVT and low recurrent rates of VTE, such a trial would require more than 2,000 patients and was not deemed feasible. Open-label studies such as this may influence management decisions however outcomes reporting and assessment bias was curtailed with its PROBE design.
The inclusion/exclusion criteria induced sample size bias by limiting recruitment to only tertiary care centers and excluding the sicker population. There were no comatose patients or patients with decreased mentation in the study and more than 95% of patients had a NIHSS less than 4. The risk of recurrent VTEs were also lower than anticipated and the authors attribute this to excellent medication adherence and good INR control. Other factors that may have played a role are that patients were only followed up for 6 months, and the selection bias that may have excluded patients that are at high risk for VTEs (i.e. patients with higher disability and cancer). Even among the patients recruited, there were a lower proportion of patients that were older, that were male, had thrombophilias or myeloproliferative syndromes — factors that increase the risk of recurrent VTEs.
The criteria for diagnosis of CVT that required anticoagulation was left to the enrolling institution. A good proportion of patients had cortical vein involvement and it is unclear how many of these patients had involvement of the deep veins or other major sinuses. A frequent dilemma among clinicians treating CVT is the timing for clinical stability and switching from parenteral therapy to oral anticoagulants. This study outlines a time frame of 5-15 days, however the results do not include the median time to administration of first oral dose.
Overall, this was a well conducted trial, owing to the scarcity of the disease in question, and elucidates that dabigatran may be a safe alternative to warfarin to prevent recurrent VTEs in patients with CVT.