Childhood stroke is one of the top ten causes of mortality in childhood. A majority of the survivors have neurological and neuropsychological deficits with significant impact on their participation and health related quality of life. WHO stroke definition includes arterial ischaemic stroke (AIS), intracranial haemorrhage (ICH) and cerebral sinovenous thrombosis (CSVT). It is well recognised that childhood stroke is multifactorial and atherosclerosis, the commonest cause of AIS in adults, is extremely rare in childhood. A number of risk factors are described and at least one factor is found in 80% children. Relatively low numbers of stroke in any one centre means that randomised controlled studies are difficult, however recent observational and registry based studies have added important information on epidemiology, causation and treatment of stroke in childhood.
Epidemiology and risk factors for childhood arterial ischaemic stroke
The incidence of childhood stroke (excluding perinatal stroke) has been described to be between 1.3 to 13/100000/year. In the recently concluded study on childhood stroke (SOCS) 96 new cases of AIS and 43 cases of haemorrhagic stroke (HS) were reported in a geographically defined UK population in the south East of England (Mallick et al., 2014). The crude incidence rate of AIS was calculated to be about 1.6/100000/year. The incidence was highest in the age group under 1year (4.4/100000/year). Black and Asian children had higher risk of stroke than white children, and boys were affected as frequently as girls. Possible explanations for the interracial difference relate to the presence of sickle cell disease in the black children and increased incidence of chronic systemic disorders and iron deficiency anaemia in Asian children.
Data from the SOCS study also showed that focal features, in particular hemiparesis was the most common presenting feature in children similar to adults. Diffuse features and seizures in isolation were rare (less than 5% children). Seizures at presentation were commonly seen in children less that 1 year and global features and headache in children over 6 years of age. Risk factors were classified into following groups:- Arteriopathy, cardiac disease, chronic systemic disorders, prothombotic states, acute systemic disorders, chronic head and neck disorders, acute head and neck disorders, and risk factors for adult atherosclerosis.
No risk factor was identified in 16 (17%) cases. One risk factor category was identified in 35 (36%) cases, two in 33 (34%) cases, and three in 12 (13%) cases. Infection as a risk factor was present in 28% of cases.
Presence of arteriopathy has the highest correlation with stroke recurrence risk. Arteriopathy was found in almost a third of the cases in the SOCS and was the commonest risk factor, therefore leading to a strong recommendation that all children with stroke should have a dedicated vascular imaging study i.e. CTA/MRA or DSA.
Diagnosis of arteriopathy was also investigated in Vascular effects of infection in paediatric stroke (VIPS)study (Wintermark et al., 2014). 355 children with AIS were enrolled prospectively in 37 centres across the world -North America (27), Europe (5)Asia (3) Australia (1)and South America (1). Detailed clinical data and imaging was reviewed centrally by the study neurologists and neuroradiologists. Arterial banding, luminal flap and intramural haematoma, ectasia and aneurysm and pseudoaneurysm were considered as definite arteriopathy however stenosis and irregularity of vessel and vessel occlusion were nonspecific and considered as possible arteriopathy as it was difficult to exclude recanalising thrombus. Follow up imaging was helpful in establishing the diagnosis. Rapid resolution of vascular stenosis was consistent with a resolving thrombus whereas stable or progressive stenosis was consistent with arteriopathy. Definite arteriopathy was found in 36% of cases, and possible arteriopathy in 9% of cases. It was possible to give a secondary diagnosis of type of vasculopathy with high degree of certainty in 30%, with dissection, transient cerebral arteriopathy, primary and secondary moya moya being the top four diagnoses. No cases of definite Primary CNS vasculitis were diagnosed in this cohort.
Preceding infection and trauma have also emerged as important risk factors for stroke in children. A population based case control study in Northern California found that preceding minor infections was seen in 33% of cases of childhood stroke (similar to SOCS) (Hills et al., 2012). The mechanism of increased risk is not clear but may be related to promotion of thrombosis through systemic inflammation or by endothelial damage. A history of preceding head and neck trauma requiring medical attention was found in 12% cases and was strongly associated (adjusted odds ratio of 9.0) with stroke in this study. The mechanism of AIS in this group was likely to be dissection.
A recent population based study based on primary care database in the UK identified a 4 fold risk of stroke in the 6 months following chickenpox (Thomas et al., 2014). The data from these studies suggests that the typical arteriopathy following varicella is unilateral and affects terminal ICA and proximal M1 and A1 segments. Vasculopathy may progress over time in a proportion of patients generally stabilising after 6 months. The mechanism of causation of post varicella angiopathy is believed to be direct invasion of the arterial endothelium with subsequent inflammation and stenosis. A similar pattern, but without history of chickenpox, has been labelled as transient cerebral arteriopathy (TCA) or focal cerebral arteriopathy (FCA). TCA/FCA appears to follow a similar course.
Classification of AIS in children
Standardised classification systems of stroke in adults e.g. TOAST classification has promoted detailed understanding of adult stroke and facilitated research into interventions, natural history and outcomes. This system is not useful in children. A number of attempts had been made to classify childhood stroke in a similar way but this has been difficult as a number of different terms and classifications have been used by different workers. A consensus based classification system for childhood AIS has been proposed. A panel of collaborators of the IPSS (international paediatric stroke study) used a modified Delphi method to propose The CASCADE (Childhood AIS Classification and Diagnostic Evaluation) classification based on the predominant anatomic site of disease in blood vessel to improve standardisation of terms used in childhood stroke and to facilitate research (Bernard et al., 2012). Following primary categories have been proposed
- Small vessel arteriopathy of childhood
- Unilateral focal cerebral arteriopathy of childhood
- Bilateral cerebral arteriopathy of childhood
- Aortic Cervical arteriopathy
- Cardio emboli
- Other (Undetermined)
- Multifactorial (more than 1 primary site of disease)
The disease is further classified into acute and chronic and when chronic can be further classified into reversible, stable progressive or indeterminate. Inclusion of further important secondary causes e.g. haemoglobinopathy, inflammation, genetic syndromes and thrombophilia has been proposed as the next stage of development. There are some drawbacks to the system as it has been developed on the basis of expert consensus rather than evidence from studies. The primary classification includes subcategories that are not strictly anatomic e.g. aorto-cervical arteriopathy. However it is hoped that CASCADE would provide a framework to add on to as more information about childhood AIS is uncovered.
Thrombolysis in AIS
Treatment of adult AIS has been revolutionised by the advent of thrombolysis and mechanical clot retrieval. The current recommendations for intravenous thrombolysis in adults recommend treatment within 4.5 hours. Logistic and diagnostic difficulties have hindered attempts to obtain robust data on safety and efficacy of thrombolysis in paediatric stroke. The aetiology of childhood AIS is diverse and very different from adult AIS where the main cause is atherosclerotic vascular disease. There is a high proportion of stoke mimics in children presenting with acute neurological symptoms. There is often a long delay between the presentation and diagnosis of AIS in children. Data from the cohort in the SOCS study showed that the median delay to first neuroimaging was 3.9 hours in patients with AIS (Mallick et al., 2014). This was shorter if the patients had reduced LOC. However the first neuroimaging was diagnostic in only 66% cases of AIS and in the patients with nondiagnostic early imaging the median delay to diagnosis was 44 hours.
National institute of neurological disorders had funded a phase I trial of thrombolysis in paediatric stroke (TIPS) to decide the safety and efficacy of 3 doses of intravenous tissue type plasminogen activator (TTPA) in children between the ages of 2-17 years. This trial commenced in 2010 and involved 17 centres (Rivkin et al., 2015; Bernard et al., 2014). TIPS was closed in December 2013 because of low recruitment however did result in increasing the preparedness of the centres involved in the study to deliver thrombolysis in selected patients. All centres reported increased ability to triage, diagnose and treat stroke rapidly. The investigators concluded that a lot more effort needs to be directed at early diagnosis of paediatric stroke before thrombolysis trials could take place. They also proposed guidelines for creation of a primary paediatric stoke units.
Thus, despite a number of case reports describing successful treatment, mechanical or intravenous thrombolysis is not standard practice in paediatric AIS at present.
IPSS data from 262 cases of CSVT cases (92 neonates) enrolled from 2003 to 2007 has been published (Ichord et al., 2015). Out of 170 paediatric cases analysed 8 had a combination of AIS and CSVT. There was male predominance and the median age was 7.2 years. 25% patients were under 3 years of age. Headache, altered level of consciousness, seizures and focal neurological deficits were common presenting features. Venous infarction was seen in 37% and intracranial haemorrhage in 31%. 6 children (4%) died and 43% were neurologically abnormal at time of discharge. Risk factors included chronic health disorder in 50% and acute systemic illness or acute head and neck infections in 41%. Prothombotic states were seen in 20%. Haematological malignancy and iron deficiency anaemia were the commonest chronic health conditions seen. Other conditions included congenital heart disease, inflammatory bowel disease and solid tumours.
Antithrombotic treatment was given in 141 out of 170 children (83%), of whom 33 (28%) had intracranial haemorrhage at the time of start of treatment. Antithrombotic treatment was withheld in 29 children out of whom 10 had intracranial haemorrhage at the time of presentation. There was considerable variation in the anticoagulation protocol in different countries. LMWH (low molecular weight heparin) was the commonest agent used for initial treatment in Europe and USA whereas UFH (Unfractionated heparin) use was more common in Australia and South America. Anticoagulation had been withheld in 3 of the 6 children who died. A strong association of death and nontreatment with anticoagulation (OR 5.2) was shown.
Reduced level of consciousness at presentation but not anticoagulation treatment was associated with abnormal neurological status. There was also no association of extent of thrombosis, presence of haemorrhage or underlying chronic health condition with abnormal neurological status at discharge.
This study further supports the safety of anticoagulation and reduction of mortality with anticoagulation demonstrated in previous smaller studies, however a controlled clinical trial would still be necessary to assess the effect of confounding factors such as underlying disease, extent of haemorrhage and infarction to clarify the benefits of treatment.
Neonates from the IPSS cohort with CSVT were reported in 2010 by Jordan and colleagues (Jordan et al., 2010). This study showed that the proportion of neonates treated with anticoagulation was considerably less (52%) compared to children (83%). There was significant regional difference in treatment. Neonates in US were less likely to be treated with anticoagulation however there was no regional difference in outcome. Thus anticoagulation appeared to be safe but no association could be demonstrated in the neonatal group between treatment and abnormal neurological outcome or death.
The French society for paediatric neurology (SPNF) working with European paediatric neurology expert group (EPNS) reviewed a number studies and published the following recommendations (Lebas et al., 2012):
'In the absence of any contraindication, it is reasonable to initiate anticoagulation during the acute phase of CSVT in children. Prolonging treatment over 3–6 months is justified according to numerous individual factors. In the absence of any contraindication, anticoagulation may be considered individually during the acute phase of CSVT in neonates for duration of 6–12 weeks.'
Similar to AIS there are case reports of mechanical and interventional thrombolysis in CSVT but no systematic studies and this is not standard practice in children at present.
Bernard, T. J., Manco-Johnson, M. J., Lo, W., MacKay, M. T., Ganesan, V., Deveber, G., ... Ichord, R. (2012). Towards a consensus-based classification of childhood arterial ischemic stroke. Stroke, 43(2), 371–377. doi:10.1161/STROKEAHA.111.624585
Bernard, T. J., Rivkin, M. J., Scholz, K., De Veber, G., Kirton, A., Gill, J. C., ... Amlie-Lefond, C. (2014). Emergence of the primary pediatric stroke center: Impact of the thrombolysis in pediatric stroke trial. Stroke, 45(7), 2018–2023. doi:10.1161/STROKEAHA.114.004919
Hills, N. K., Johnston, S. C., Sidney, S., Zielinski, B. A., & Fullerton, H. J. (2012). Recent trauma and acute infection as risk factors for childhood arterial ischemic stroke. Annals of Neurology, 72(6), 850–858. doi:10.1002/ana.23688
Ichord, R. N., Benedict, S. L., Chan, A. K., Kirkham, F. J., Nowak-Göttl, U., & International Paediatric Stroke Study Group. (2015). Paediatric cerebral sinovenous thrombosis: findings of the International Paediatric Stroke Study. Archives of Disease in Childhood, 100(2), 174–9. doi:10.1136/archdischild-2014-306382
Jordan, L. C., Rafay, M. F., Smith, S. E., Askalan, R., Zamel, K. M., DeVeber, G., & Ashwal, S. (2010). Antithrombotic Treatment in Neonatal Cerebral Sinovenous Thrombosis: Results of the International Pediatric Stroke Study. Journal of Pediatrics, 156(5). doi:10.1016/j.jpeds.2009.11.061
Lebas, A., Chabrier, S., Fluss, J., Gordon, K., Kossorotoff, M., Nowak-Göttl, U., ... Tardieu, M. (2012). EPNS/SFNP guideline on the anticoagulant treatment of cerebral sinovenous thrombosis in children and neonates. European Journal of Paediatric Neurology, 16(3), 219–228. doi:10.1016/j.ejpn.2012.02.005
Mallick, A. A., Ganesan, V., Kirkham, F. J., Fallon, P., Hedderly, T., McShane, T., ... O'Callaghan, F. J. (2014). Childhood arterial ischaemic stroke incidence, presenting features, and risk factors: A prospective population-based study. The Lancet Neurology, 13(1), 35–43. doi:10.1016/S1474-4422(13)70290-4
Mallick, A. A., Ganesan, V., Kirkham, F. J., Fallon, P., Hedderly, T., McShane, T., ... O'Callaghan, F. J. (2015). Diagnostic delays in paediatric stroke. Journal of Neurology, Neurosurgery & Psychiatry, 86(8), 917–921. doi:10.1136/jnnp-2014-309188
Rivkin, M. J., DeVeber, G., Ichord, R. N., Kirton, A., Chan, A. K., Hovinga, C. A., ... Amlie-Lefond, C. (2015). Thrombolysis in pediatric stroke study. Stroke; a Journal of Cerebral Circulation, 46(3), 880–5. doi:10.1161/STROKEAHA.114.008210
Thomas, S. L., Minassian, C., Ganesan, V., Langan, S. M., & Smeeth, L. (2014). Chickenpox and risk of stroke: A self-controlled case series analysis. Clinical Infectious Diseases, 58(1), 61–68. doi:10.1093/cid/cit659
Wintermark, M., Hills, N. K., DeVeber, G. A., Barkovich, A. J., Elkind, M. S. V, Sear, K., ... VIPS Investigators. (2014). Arteriopathy diagnosis in childhood arterial ischemic stroke: results of the vascular effects of infection in pediatric stroke study. Stroke; a Journal of Cerebral Circulation, 45(12), 3597–605. doi:10.1161/STROKEAHA.114.007404