Treatment and outcome of childhood cerebral sinovenous thrombosis

Abstract

Objective: To test our hypothesis that anticoagulation is associated with better neurologic outcomes in childhood cerebral sinovenous thrombosis (CSVT), we analyzed treatment and outcomes in a population of 410 children from the International Pediatric Stroke Study (IPSS).

Methods: We included patients enrolled in the IPSS registry with a diagnosis of CSVT at age >28 days with radiologic confirmation, in isolation or with concomitant arterial ischemic stroke. The primary outcome was the neurologic status at discharge. We defined unfavorable outcome as severe neurologic impairment or death at discharge. The Pediatric Stroke Outcome Measure was used for long-term outcome in those with follow-up. Predictors of anticoagulation use and outcome were analyzed by logistic regression.

Results: Most children (95%) had identifiable risk factors, and 82% received anticoagulation. Shift analysis demonstrated better outcomes at discharge in children who were anticoagulated, and this persisted with longer-term outcomes. In multivariable analysis, anticoagulation was significantly associated with favorable outcomes (adjusted odds ratio [aOR] unfavorable 0.32, p = 0.007) whereas infarct was associated with unfavorable outcome (aOR unfavorable 6.71, p < 0.001). The trauma/intracranial surgery was associated with a lower odds of anticoagulation use (aOR 0.14, p < 0.001).

Conclusions: Within the IPSS registry, children with risk factors of trauma or intracranial surgery were less likely to receive anticoagulation for CSVT. Anticoagulation was associated with a lower odds of severe neurologic impairment or death at hospital discharge, but this finding is limited and needs further confirmation in randomized, controlled, prospective studies.

Figure 1. Outcome distributions in the patients with and without anticoagulation

(A) Neurologic outcomes at the time of hospital discharge. Stacked bars represent the percent of patients by anticoagulation status. *p < 0.05 for the shift analysis by ordinal regression adjusted for age and sex. (B) Boxplot of PSOM at long-term follow-up. **p < 0.001 by the Mann-Whitney U test. PSOM = Pediatric Stroke Outcome Measure.

Figure 2. Subgroup analyses of univariate predictors of unfavorable outcome

Ordinal regression analysis was used to analyze each variable, and the odds ratio presented represents the odds of each outcome severity relative to all levels below it. LOC = level of consciousness.

Figure 3. Neurologic outcomes in children with cerebral sinovenous thrombosis and trauma/intracranial surgery

Stacked bars represent the percent of patients by anticoagulation status. *p < 0.05 for the shift analysis by ordinal regression adjusted for age and sex.

Figure 4. Treatment of childhood cerebral sinovenous thrombosis in the International Pediatric Stroke Study registry

(A) Patients were categorized as receiving no antithrombotic treatment, antiplatelets alone, anticoagulation alone, or both anticoagulation and antiplatelets. (B) Percent of patients who received anticoagulation within each risk factor category. *p < 0.0001.