Article Text

Download PDFPDF

Paediatric cerebral sinovenous thrombosis: findings of the International Paediatric Stroke Study
  1. R N Ichord1,
  2. S L Benedict2,
  3. A K Chan3,
  4. F J Kirkham4,5,
  5. U Nowak-Göttl6
  6. for the International Paediatric Stroke Study Group
  1. 1Department Neurology & Pediatrics, Children's Hospital of Philadelphia, Perelman School of Medicine of the University of Pennsylvania, Philadelphia, Pennsylvania, USA
  2. 2Department Neurology, Primary Children's Hospital, University of Utah, Salt Lake City, Utah, USA
  3. 3Division of Hematology and Oncology, Department of Pediatrics, McMaster University, Hamilton, Ontario, Canada
  4. 4Neurosciences Unit, University College London Institute of Child Health, London, UK
  5. 5University hospital Southampton NHS Foundation Trust, Southampton, UK
  6. 6Department of Pediatric Hematology/Oncology, Universitätsklinikum Schleswig-Holstein, Kiel, Germany
  1. Correspondence to Dr Rebecca Ichord, Department Neurology CTRB 10th Fl, Children's Hospital of Philadelphia, 3501 Civic Ctr Blvd, Philadelphia, PA 19104, USA; ichord{at}


Objectives We evaluated clinical features, treatment practices and early outcome in a multicentre cohort of children with cerebral sinovenous thrombosis (CSVT).

Methods Children with CSVT from 10 countries were enrolled from January 2003 to July 2007 in the International Paediatric Stroke Study. We analysed clinical symptoms, underlying conditions, antithrombotic treatment and neurological outcome at hospital discharge in 170 children.

Results Of 170 children enrolled, 60% were male; median age 7.2 years (IQR 2.9–12.4). Headache, altered consciousness, focal deficits and seizures were common presenting clinical features. Infarction affected 37% and intracranial haemorrhage 31%. Risk factors included chronic disease in 50%; acute systemic illness or head/neck disorders 41%; prothrombotic state 20% and other haematological abnormality 19%. Discharge neurological status was normal in 48%, abnormal in 43% and unknown in 5%. Antithrombotic therapy was common, most often low molecular weight heparin was common, with significant regional variation in treatment practices. Mortality was low (4%) and was associated with no anticoagulation but not underlying chronic disease, anatomic extent of thrombosis or intracranial haemorrhage. Abnormal neurological status at discharge or death was associated with decreased level of consciousness at presentation and the presence of an identified prothrombotic state.

Conclusions Our study extends the observations of previously published smaller studies in children with CSVT that this is a morbid disease with diverse underlying causes and risk factors. Divergent treatment practices among highly specialised centres as well as limited data on treatment efficacy and safety suggest that further study of this condition is warranted.

  • Neurology
  • Vascular Disease
  • stroke

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What is already known on this topic

  • Cerebral sinovenous thrombosis in childhood is fatal in 3%–12% and causes neurological sequelae in up to 62% of survivors.

  • Previously published cohort studies have described epidemiology, treatment practices and outcomes, but are limited due to small sample size or single-centre enrolment.

  • Uncertainty remains regarding risk factors for death and adverse neurological outcome and optimal treatment.

What this study adds

  • The major strengths of our study are its large sample size and representation of multiple international centres not included in smaller studies.

  • This adds valuable new data regarding a rare condition, improving knowledge of risk factors, current treatment practices and outcomes.

  • Our finding of increased mortality in patients not treated with systemic anticoagulation supports previous observations that have suggested there is probable benefit of anticoagulation.


Cerebral sinovenous thrombosis (CSVT), defined as blood clot in a vein carrying blood from the brain to the heart, affects 0.34–0.67/100 000 children annually, of which 27%–35% are neonates.1–3 Mortality is 3%–12%, and neurological sequelae affect up to 62% of survivors.4–6 Aetiology in the paediatric population includes a wide spectrum of underlying conditions and symptoms at clinical presentation.1–3 ,5 ,7 Recent studies suggest that haematological abnormalities, including prothrombotic disorders and iron deficiency, contribute to development of CSVT.1 ,5 ,8–12 Previously published cohort studies have described epidemiology, treatment practices and outcomes, but are limited due to small sample size or single-centre enrolment. The International Paediatric Stroke Study (IPSS) established a prospective multicentre registry to characterise epidemiology, treatment and outcome of infants and children with arterial ischaemic stroke (AIS) and CSVT. The objective of this paper is to analyse this large multicentre international data set for clinical presentation, underlying diseases and provoking conditions, treatment and outcome at hospital discharge among children with CSVT and to define risk factors for death and abnormal neurological status at hospital discharge.

Methods and design


This study was performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. Research ethics approval was obtained at each site by the local institutional review boards to include data of consecutively enrolled patients in the IPSS database maintained in Toronto, Canada.


The IPSS is an international paediatric stroke registry established in 2003 with the purpose of collecting data on demographic, clinical and radiological features, risk factors, treatment and outcomes at the time of hospital discharge among children from birth up to the age of 19 years with AIS and CSVT. We used the same methodology for enrolment, data abstraction and database management as previously described.13 This analysis included consecutively enrolled cases submitted to the database between 1 January 2003 and 1 July 2007. Sample size was determined by the number of cases enrolled in that time interval that met inclusion criteria. Local treating physicians determined diagnostic and treatment decisions.


Site investigators identified cases and determined eligibility.14 Patients were included in the current childhood cohort study who met consensus-based, published clinical and radiological criteria for the diagnosis of childhood CSVT,6 as follows: (1) age >28 days up to 19 years at onset of symptoms, (2) acute neurological syndrome with signs and symptoms of CSVT, such as altered mental status, headache, papilloedema, seizures or focal neurological deficits and (3) radiological confirmation of CSVT by one or more modalities (MRI with MR venogram, CT venogram or cerebral catheter venography) showing thrombus or flow interruption within cerebral veins or dural venous sinuses, with or without venous distribution infarction.

Data collection

Site investigators abstracted data from medical records onto IPSS case report forms (CRF). Data were entered via a web-based data entry system to the IPSS database, created and maintained in the IPSS data coordinating centre at the Toronto Hospital for Sick Children. Clinical data abstracted included demographics (age, gender, country of origin); clinical symptoms at presentation (altered mental status, headache, seizures, focal neurological deficits); underlying disease conditions (heart disease, genetic disorder, connective tissue disorder, prothrombotic state (including acquired and genetic thrombophilias), haematological disorder, malignancy, pre-existing cerebrovascular disorder); acute illness or provoking condition (systemic disorders, head or neck infections, trauma, treatment with a prothrombotic drug such as L-asparaginase, cranial surgery); neuroimaging findings and treatment modalities (antithrombotic treatment with unfractionated heparin (UFH) or low molecular weight heparin (LMWH), anticonvulsants, antibiotics, immunomodulatory therapy, neurosurgical interventions, transfusion or exchange transfusion). Initial and secondary antithrombotic treatments were reported, where secondary antithrombotic treatment refers to a treatment change from the initial agent to a different agent during admission. Data on prothrombotic states were reported to the registry database only for abnormal results. Practices varied widely across centres regarding testing for prothrombotic states. Data concerning the testing performed on each patient were not available as this was beyond the scope of the IPSS protocol. Data on neuroimaging included site of thrombosis, presence of infarction and presence and type of intracranial haemorrhage. Site of CSVT was classified as involving superficial or deep venous systems or both. Outcome data were collected on discharge disposition (home, inpatient rehabilitation unit, nursing facility), vital status at hospital discharge (alive or dead) and neurological status at discharge classified as normal or abnormal. Data on standardised functional outcomes at intervals past hospital discharge were beyond the scope of the IPSS registry.

Statistical analysis

Characteristics of the patient population were summarised with descriptive statistics, expressed as medians and IQRs. All statistical analyses were performed using the StatView 5 software package (SAS Institute, Cary, North Carolina, USA) and MedCalc (V.11.1.1., Mariakerke, Belgium). Continuous data are presented as median values and IQR. For categorical data, frequency distributions were compared between groups using χ2 test or Fisher's exact test (F), as appropriate. The criterion for statistical significance was set at p=0.05 based on two-sided test. In order to evaluate (i) clinical symptoms (no clinical symptoms vs. any clinical symptoms), (ii) underlying medical conditions (no underlying medical condition vs. any underlying medical condition), (iii) thrombophilic risk factors (no thrombophilia vs. any thrombophilia) and (iv) initial anticoagulation (UFH or LMWH vs. no anticoagulation), analysis was performed using logistic regression. Adjustment variables for anticoagulation included recruitment areas (Australia, Canada, Europe, South America, the USA). Risk ratios and adjusted risks were expressed as ORs together with 95% CIs. Clustering by centre was not accounted for in the analysis because of small numbers contributed by each centre.


Patient demographics

Investigators at 30 sites in 10 countries enrolled 1187 evaluable cases of AIS and CSVT between January 2003 and July 2007 ( Figure 1 shows enrolment characteristics by age group (neonate vs. non-neonate) and diagnostic subtype. A total of 170 patients aged 1 month to 19 years with a diagnosis of childhood CSVT were enrolled, representing 65% of patients in the IPSS cohort with CSVT and 14% of the combined IPSS cohort with either AIS or CSVT of all ages. Of these, 162 had CSVT alone and 8 had combined AIS and CSVT (figure 1). There was a male predominance, with 102 boys (60%) (table 1). The median age was 7.2 (IQR 2.9–12.4) years, with 25% of children under the age of 3 years.

Table 1

Demographic, clinical and radiological characteristics and discharge status

Figure 1

Enrolment characteristics of International Paediatric Stroke Study (IPSS) registry. DNQ, did not qualify or excluded for insufficient data.

Clinical and radiological features, risk factors, treatment and outcome

Table 1 summarises clinical and radiological characteristics, risk factors and discharge status. Common presenting signs and symptoms in children with CSVT included headache, altered consciousness, focal deficits and seizures. The superficial dural sinuses alone were most commonly involved (56%). Cerebral venous infarction affected 37% and intracranial haemorrhage 31% of cases. Clinical neurological status at discharge was normal in 48%, abnormal in 43% and unknown or not reported in 5%. Six children (4%) died before hospital discharge. In total, 84% of children were discharged to home, while 7% went to inpatient rehabilitation and 6% to another hospital or nursing facility.

Chronic health disorders were reported in 50% of cases; acute systemic illness or acute head/neck disorders in 41% and neither or unknown in 9%. A prothrombotic state was identified in 34/170 (20%) of cases, 28 with isolated prothrombotic states without other chronic underlying conditions and the remainder associated with other chronic underlying disorders. As diagnostic practices varied widely across centres, all subjects were not tested for all prothrombotic factors. Data collection regarding which patients were tested for each prothrombotic factor was not available as this was beyond the scope of the IPSS protocol. Specific prothrombotic abnormalities were reported by annotations on the CRFs, detailed in table 1. Haematological malignancy and iron deficiency were each reported in around 10% of cases, but only one patient had a haemoglobinopathy (sickle cell anaemia). The anatomical extent of the thrombosis was not associated with presence of prothrombotic factors.

Table 2 summarises initial and secondary antithrombotic treatment. Antithrombotic treatment was given in 141 of the 170 (83%) children, of whom 33 (28%) had intracranial haemorrhage present at the time of treatment. Antithrombotic treatment was withheld in 29/170 children, of whom 10 had intracranial haemorrhage. Among the 6 children who died, 3 received antithrombotic treatment.

Table 2

Antithrombotic treatment N (%)

Clinical–radiological associations and outcome predictors

Table 3 shows the associations of clinical and radiological features with outcome. There was an association of death and non-treatment with anticoagulation, with an OR of 5.2. Abnormal neurological status at discharge was associated with decreased level of consciousness at presentation, but was not associated with anticoagulation treatment. There was no association of death or abnormal neurological status with presence of an underlying chronic disease, extent of thrombosis or presence of intracranial haemorrhage at initial presentation.

Table 3

Risk factors for mortality and abnormal neurological status at discharge in children


The aim of this study was to describe clinical symptoms, underlying medical conditions, anticoagulant treatment and acute neurological outcome in 170 children with CSVT from a multicentre international registry. This is one of the largest cohort studies with the most diverse geographical representation published to date.

Our study extends the observations of previously published smaller studies in children with CSVT that this is a morbid disease with diverse underlying causes and risk factors. The age distribution with median age of 7 years and male predominance of 60% are comparable with previous reports.2–5 As seen in other studies, acute head and neck disorders in otherwise healthy children are dominant causes of childhood CSVT. Certain chronic diseases known to predispose to CSVT are clearly represented in this cohort, including congenital heart disease, inflammatory bowel disease, haematological disorders and malignancy. Presenting signs and symptoms were dominated by non-specific symptoms of headache and altered consciousness, with focal neurological deficits (43%) more common than reported in prior studies. Cerebral venous infarction or haemorrhage or both were more common than in previously published studies, possibly reflecting widespread use of MRI. The mortality rate of 4% is comparable to that reported in recent prospective cohort studies from Europe and Canada.3 ,4 Abnormal neurological status at discharge was documented in 43% of children in this study, further illustrating the serious nature of this disease. While long-term outcome data were not available in our study, these observations are consistent with reported rates of long-term neurological impairments and neurocognitive deficits in 40%–60% of survivors.3 ,5 ,8 ,12 ,15 ,16 Evaluation of risk factors for poor outcome or death identified coma as a risk factor for death and neurological sequelae, although this is only statistically significant for the latter group, probably because of the small number of deaths. Our finding of increased mortality in patients not treated with systemic anticoagulation is in keeping with other cohort studies showing probable benefit of anticoagulation.7 Further study in a larger cohort or in a clinical trial where confounding factors such as underlying disease, presence of infarction or haemorrhage can be measured and controlled for in the analysis would be helpful in clarifying the effect of anticoagulation treatment on mortality and long-term functional outcome.

We found significant regional differences in antithrombotic treatment practices. A significant difference in the mode of initial anticoagulant administration was found with respect to different patient recruitment areas. In Canada, Europe and the USA, children with CSVT onset were treated with LMWH in approximately 60%–69% of cases, while in Australia and South America the use of UFH was more frequent.

Our study has a number of strengths and limitations. Important limitations include wide variations in practices concerning evaluation of prothrombotic risk factors, differences across centres in modality and timing of imaging, choice of treatment and the lack of data on long-term outcome. Central review and adjudication of imaging would be ideal for any cohort study of paediatric CSVT, but it was beyond the scope of our study.

The major strengths of our study are its large sample size and representation of multiple international centres. Notably, our study includes a large number of patients from the USA and other regions of the world (South America), which have previously been under-represented in cohort studies of paediatric CSVT. This adds valuable new data regarding a rare condition, thereby improving knowledge of risk factors, current treatment practices and short-term outcomes, and making the cumulative published data more widely generalisable.

We found that children are more often treated with anticoagulation compared with neonates in the same centres. Jordan et al reported that 52% of neonates with CSVT enrolled in the IPSS database by the same centres as in our study received antithrombotic therapy compared with 83% of children with CSVT reported in our study. The basis for this finding cannot be determined from our study. Differences between neonates and children concerning risk factor profiles and prevalence of intracranial haemorrhage may have led to different judgements by practitioners about the potential risks vs. benefits of antithrombotic treatment. Moreover, the current treatment guidelines for the treatment of CSVT are more conservative for neonates compared with children.17 ,18 While the safety of anticoagulation for CSVT in neonates and children is supported increasingly by observational studies, its efficacy in improving survival, preventing short-term neurological morbidity and improving long-term functional outcome deserves further study. Our finding of diverse treatment practices indicates that much uncertainty exists among treating physicians, even in highly specialised centres. This suggests that treatment trials are needed in neonates and children. Age-related differences in risk factor profiles, clinical signs and symptoms, rates of associated haemorrhage, as well as profound maturational changes in hemostasis, suggest that clinical trials evaluating antithrombotic treatment in neonates would be different from those in children.


We thank research assistants Elisa Wilson, BSc, and Jeffrey Templeton, BA, and research manager Ann-Marie Pontigon, BSc, for their work on this study.



  • FJK and UNG shared senior authorship.

  • Collaborators Coinvestigators: Members of the International Paediatric Stroke Study participating in data collection for this manuscript are as follows. Original investigators—Steve Ashwal, MD, Loma Linda University School of Medicine, Loma Linda, California; Gabrielle deVeber, MD, MHSc, The Hospital for Sick Children, Toronto, Ontario, Canada; Donna Ferriero, MD, University of California, San Francisco, California; Heather Fullerton, MD, University of California, San Francisco, California; Rebecca Ichord, MD, Children's Hospital of Philadelphia, Philadelphia, Pennsylvania; Fenella Kirkham, MA, MD, BChir, University College London Institute of Child Health, London, UK; John K. Lynch, DO, MPH, National Institute of Health/National Institute of Neurological Disorders and Stroke, Bethesda, Maryland; Finbar O'Callaghan, MBChB PhD, Bristol Royal Hospital for Children, Bristol, UK; Steve Pavlakis, MD, Maimonides Medical Center, Brooklyn, New York; Guillaume Sebire, MD, PhD, Université de Sherbrooke Fleurimont, Sherbrooke, Quebec, Canada; and Andrew Willan, BA, MSc, PhD, Hospital for Sick Children, Toronto, Ontario, Canada. Institutions enrolling at least 20 patients (numbers in parentheses indicate patients enrolled). The Hospital for Sick Children, Toronto, Ontario, Canada (147): Gabrielle deVeber, MD, MHSc, Andrew Willan, BA, MSc, PhD, Adam Kirton, MD, Mahendra Moharir, MD, and Marianne Sofronas, MA; Münster University Pediatric Hospital, Münster, Germany (122): Ulrike Nowak-Göttl, MD, Christine Düring, MD, and Anne Krümpel, MD; University of Texas Southwestern Medical Center, Dallas, Texas (94): Michael M. Dowling, MD, PhD, Patricia Plumb, RN, MSN, Janna Journeycake, MD, and Katrina van de Bruinhorst, MA; Ohio Stroke Registry (94): Akron Children's Hospital, Akron, Ohio: Abdalla Abdalla, MD; Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio: Tonya Phillips, MD; Cleveland Clinic, Cleveland, Ohio: Neil Friedman, MD; MetroHealth Medical Center, Cleveland, Ohio: Elie Rizkallah, MD; Nationwide Children's Hospital, Columbus, Ohio: Warren Lo, MD, Khaled Zamel, MD; Rainbow Babies and Children's Hospital, Cleveland, Ohio: Max Wiznitzer, MD, and Karen Lidsky, MD; Pontificia Universidad Catolica de Chile, Santiago, Chile (78): Marta Isabel Hernandez Chavez, MD; Royal Children's Hospital, Melbourne, Victoria, Australia (75): Professor Paul Monagle, Mark MacKay, MD, Chris Barnes, MD, Janine Furmedge, RN, BSc, and Anne Gordon, MSc, BAppSc; The University of Utah and Primary Children's Medical Center, Salt Lake City, Utah (70): Susan L. Benedict, MD, and James F. Bale, Jr, MD; Children's Hospital of Philadelphia, Philadelphia, Pennsylvania (63): Rebecca Ichord, MD, Daniel Licht, MD, and Sabrina Smith, MD; Loma Linda University School of Medicine, Loma Linda, California (54): Steve Ashwal, MD, and Chalmer McClure, MD, PhD; Schneider Children's Hospital, New Hyde Park, New York (46): Li Kan, MD, MS, Robin Smith, MD, Joseph Maytal, MD, and Rosemarrie Sy-Kho, MD; Children's National Medical Center, Washington, DC (39): Jessica Carpenter, MD, Taeun Chang, MD, and Steven Weinstein, MD; University of California San Francisco, San Francisco, California (37): Donna Ferriero, MD, and Heather Fullerton, MD; Maimonides Medical Center, Brooklyn, New York (26): Steve Pavlakis, MD, Sharon Goodman, PNP, and Kim Levinson, PNP; Riley Hospital, Indianapolis, Indiana (26): Meredith Golomb, MD, MSc; Winnipeg Children's Hospital, Winnipeg, Manitoba, Canada (24): Mubeen Rafay, MBBS, MSc, Frances Booth, MD, Michael Salman, MD, Charuta Joshi, MD, Namrata Shah, MD, and Monica Nash, RN; Children's Hospital of New York, New York, New York (22): Geoffrey Heyer, MD; Great Ormond St Hospital, London, UK (21): Vijeya Ganesan, MBChB, MD; Stollery Children's Hospital, Edmonton, Alberta, Canada (21): Jerome Y. Yager, MD; and Pediatric Institute Hospital, Kuala Lumpur, Malaysia (20): Hussain Imam, MBBS, FRCP, DCH. Institutions enrolling <20 patients—Bangkok Hospital Medical Center, Bangkok, Thailand: Montri Saengpattrachai, MD; British Columbia Children's Hospital, Vancouver, British Columbia, Canada: Bruce Bjornson, MD; Children's Central Hospital, Tbilisi, Georgia: Nana Tatishvili, MD; Children's Hospital of Buffalo, Buffalo, New York: E. Ann Yeh, MD; Children's Hospital of Eastern Ontario, Ottawa, Ontario, Canada: Peter Humpheries, MD; Children's Hospital of Wisconsin, Milwaukee, Wisconsin: Catherine Amlie-Lefond, MD, and Harry T. Whelan, MD; Denver Children's Hospital, Denver, Colorado: Timothy Bernard, MD, and Neil Goldenberg, MD; Hospital Dr Sotero del Rio, Puente Alto, Chile: Manuel Arriaza Ortiz, MD; McMaster University Medical Centre, Hamilton, Ontario, Canada: Anthony Chan, MBBS; Miami Children's Hospital, Miami, Florida: Marcel Deray, MD, and Zaid Khatib, MD; Queen Mary Hospital, Hong Kong, China: Virginia Wong, MD; Université de Sherbrooke Fleurimont, Sherbrooke, Quebec, Canada: Guillaume Sebire, MD, PhD; University of Rochester Medical Center, Rochester, New York: Jill M. Cholette, MD, Shalu Narang, MD, and Norma B. Lerner, MD, MPH; and University of Texas San Antonio, San Antonio, Texas: Shannon Carpenter, MD, and Kurt Bischoff, MSc.

  • Contributors RNI: first author; contributed to study design, patient recruitment and data collection, and data analysis. SLB and AKC: contributed to patient recruitment and data collection; critical review of the manuscript. FJK: contributed to study design, patient recruitment and data collection, and data analysis, performed statistical analysis. UN-G: contributed to study design, patient recruitment and data collection, and data analysis; performed statistical analysis.

  • Funding The International Paediatric Stroke Study was funded by the Child Neurology Foundation.

  • Competing interests Dr Ichord served as a consultant to Berlin Heart as member of clinical event committee for IDE trial of Berlin EXCOR Pediatric ventricular assist device.

  • Ethics approval Approval was provided by the local IRB/EC in every institution participating in the study.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement All data used in this study are maintained in a database under the control of the International Paediatric Stroke Study (IPSS) at the Toronto Hospital for Sick Children. Access to that data is available to any and all registered contributors to the IPSS in keeping with the Publication Policy of the IPSS.

Linked Articles

  • Atoms
    R Mark Beattie