by the Child Neurology Society Ad Hoc Committee on Stroke in Children:
E. S. Roach, M.D., Co-Chair, Dallas, TX
Gabrielle deVeber, M.D., Co-Chair, Hamilton, Ontario
Anthony R. Riela, M.D., Dallas, TX
Max Wiznitzer, M.D., Cleveland, OH
Table of Contents
Despite growing appreciation by neurologists that cerebrovascular disorders occur more often in children than once suspected, the study of stroke in children and adolescents has remained largely descriptive. Child neurologists often encounter children with a cerebrovascular lesion, yet large scale clinical research is difficult because these disorders are less common than in adults and arise from diverse causes. Three fundamental problems hinder both clinical research and the routine clinical care of children with cerebrovascular disease:
(1) The infrequency of cerebrovascular disorders in children makes it difficult to organize multicenter controlled clinical trials of the sort done in adults in recent years. The relative rarity of stroke in children also contributes to the still remaining reluctance of some clinicians to consider the diagnosis in individual children.
(2) The causes of cerebrovascular disease in children are legion, and no one risk factor predominates. Thus, not only is stroke less common in children, but the diversity of risk factors creates a heterogeneous patient population which hinders clinical research.
(3) Despite improved diagnostic techniques which make rapid, noninvasive diagnosis of cerebrovascular disease possible, many physicians still know very little about cerebrovascular disorders in children. This lack of awareness contributes to delayed diagnosis and in the near future will make it more difficult to use thrombolytic agents or other treatments which require early diagnosis and treatment.
Although cerebrovascular disorders occur less often in children than in adults, recognition of stroke in children has probably increased because of the widespread application of noninvasive diagnostic studies such as magnetic resonance imaging (MRI), magnetic resonance angiography (MRA), computed tomography (CT) and, in the neonate, cranial ultrasound studies.1-3 These studies allow confirmation of a diagnosis that in previous years would not have been suspected or at least not recognized as a vascular lesion. Also, the number of patients with cerebrovascular lesions from certain risk factors may have increased as more effective treatments for some causes of stroke have allowed patients to survive long enough to develop vascular complications. Patients with sickle cell disease or with leukemia, for example, now have a longer life-expectancy, and during this time they may have a stroke.
Most of the pediatric cerebrovascular literature consists of single case reports or small groups of children with a common etiology. These reports offer some insight into the relative frequency of various causes of stroke and draw attention to individual risk factors, but their usefulness is otherwise limited. Larger series of children selected for a common anatomic lesion or a single cause offer additional insight into the unique features of cerebrovascular lesions in children,4 but patients collected from large medical centers may not be representative of all children with stroke. None of these studies can accurately judge the incidence of cerebrovascular disease in children.
Schoenberg and colleagues studied cerebrovascular disease in children of Rochester, Minnesota from 1965 through 1974.5 Excluding strokes related to intracranial infection, trauma or birth, they found three hemorrhagic strokes and one ischemic stroke in an average at risk population of 15,834, for an estimated average annual incidence rate of 1.89/100,000/year and 0.63/100,000/year for hemorrhagic and ischernic strokes respectively. Their overall average annual incidence rate for children through fourteen years of age was 2.52/100,000/year. In this population, hemorrhagic strokes occurred more often than ischemic strokes, while in the Mayo Clinic referral population, ischemic strokes were more common. The risk of childhood cerebrovascular disease in this study is about half the risk for neoplasms of the central nervous system of children, but neonates and children with traumatic lesions are excluded. Despite our impression that cerebrovascular disorders are recognized more often in children than in previous years, Broderick and colleagues6 found an incidence of 2.7 cases/1 00,000/year, similar to the figure reported by Schoenberg and colleagues.5 In the Canadian Pediatric Ischemic Stroke Registry incidence of arterial and venous occlusion is estimatedtobe 1.2/100,000 children/year.
The frequency of several individual risk factors for stroke in children is known, but in most instances, the occurrence of secondary cerebrovascular disease is so variable that it is difficult to assess the relative contribution of each risk factor to the problem of cerebrovascular disease as a whole. In one report which included both children and young adults, children were less likely than young adult stroke patients to have identifiable risk factors and more often fall victim to infectious or inflammatory disorders.7 The implication is that children may have additional, as yet unknown, risk factors.
Probably the most fundamental difference between cerebrovascular diseases in children and adults is the wide array of risk factors seen in children versus adults ( Table 1 ).8Congenital heart disease and sickle cell disease, for example, are common causes of stroke in children, while atherosclerosis is rare in children. No cause can be detected in about a fifth of the children with ischemic infarction, yet many of these children seem to do well. The recognized causes of cerebrovascular disorders in children are numerous ( Table 1 ), and the probability of identifying the cause depends on the thoroughness of the evaluation. A probable cause of cerebral infarction was identified in 184 of 228 (79%) children in the Canadian Pediatric Ischemic Stroke Registry ( Figure 1 ). The source of an intracranial hemorrhage is even more likely to be found.8
The most common cause of stroke in children is probably congenital or acquired heart disease. In the Canadian Pediatric Ischemic Stroke Registry, heart disease was found in 40 of 228 (19%) of the children with arterial thrombosis. Many of these children are already known to have heart disease prior to their stroke, but in other instances a less obvious cardiac lesion is discovered only after a stroke. Complex cardiac anomalies involving both the valves and chambers are collectively the biggest problem, but virtually any cardiac lesion can sometimes lead to a stroke. Of particular concern are cyanotic lesions with polycythemia, which increase the risk of both thrombosis and embolism.
Both the frequency and the cause of pediatric stroke may depend somewhat on both the geographic location and the specific hospital setting. The Canadian Pediatric Ischemic Stroke Registry, for example, lists only 5 children (2%) with cerebral infarction due to sickle cell anemia. A large metropolitan hospital in the United States might care for this many patients in a year, but early estimates9 that cerebral infarction occurred in 17% of people with sickle cell disease proved far higher than the 4-5% figure derived from more representative samples in Jamaica and in Africa.10,11
Lack of general awareness of cerebrovascular disorders in children probably delays medical attention for children with cerebrovascular disorders. It is not unusual, for example, for children with a cerebral infarction to be brought to a physician several days after the onset of symptoms. In contrast, family members are usually well aware of the significance of an acute neurological impairment in older individuals, and these patients are typically seen by a physician earlier than children with a similar lesion.
Data from the Canadian Pediatric Ischemic Stroke Registry indicate that 48-72 hours often elapse between the onset of symptoms of arterial occlusion and a child's diagnosis (Table 2).12Venousocclusion was discovered a bit more quickly than arterial occlusion, at least in younger children, perhaps because of the common occurrence of epileptic seizures in children with venous thrombosis. This seems to be fairly typical of the pattern seen in the United States as well. The typical adult with anew onset neurological deficit from cerebrovascular disease undoubtedly sees a physician much sooner. It is likely that this delay in the diagnosis of children reflects a lack of awareness by both physicians and families that cerebrovascular disease occurs in children. To the extent that treatment might be improved by earlier evaluation and treatment, prompt recognition and treatment could improve management.
No randomized controlled treatment trials have been completed in children with stroke; many of the procedures increasingly used in children with cerebrovascular disease have been adapted from studies in adults. Accumulating experience with antithrombotic and anticoagulant treatment in children suggests that these agents can be safely used in children, though their efficacy and proper dose still need to be established by controlled trials. Thrombolytic agents should be as effective in children as in adults, but the safety data are inadequate for children and the timing and dosage need to be determined for children and adolescents.
(1) Background: There are no controlled trials on the use of aspirin or other antiplatelet agents in children with ischemic cerebral infarction. Nevertheless, aspirin is being used more and more in the routine clinical care of children with cerebral ischemic disorders.
(2) Safety: In addition to the potential complications of chronic aspirin use seen in adults, children taking daily aspirin could have an increased risk of developing Reye's syndrome. Evidently the risk of Reye's syndrome is fairly small, due perhaps to the low aspirin dose typically used in children. Despite the increasingly common use of aspirin in children with stroke, we were unable to find in the literature even one child who developed Reye's syndrome while taking prophylactic aspirin. One 65 year-old, however, developed Reye's syndrome while taking aspirin for stroke prophylaxis, but he also took additional aspirin for influenza.13
(3)Efficacy: Adaily aspirin dose of 2-3 mg/Kg/day causes an antiplatelet effect, though it remains to be seen whether this dose of aspirin is clinically effective in children.
(1) Background: A decision to use heparin in a child rests on two questions: What is the likelihood of either extension of an infarction or of a second infarction from an embolus which might be prevented
by treatment, and what is the risk of inducing a hemorrhage because of anticoagulation? Much like the situation in adults, heparin should be used in children thought to have a high risk of recurrence and a low risk of secondary hemorrhage.
(2) Safety: There are no large scale trials of heparin in children with ischemic stroke, but increasing clinical experience suggests that children can be treated along the same lines as adult patients with
reasonable safety.8,14,15 Combined experience with over 100 pediatric patients treated for systemic clots with low molecular weight heparin indicates a good safety profile and dose finding feasibility.16 No significant hemorrhagic complications occurred in these initial 100 children's.18
(3) Efficacy: The value of anticoagulation in children is difficult to assess without more information. Anticoagulation is commonly used in children with arterial dissection, dural sinus thrombosis, coagulation disorders, or a high risk of embolism.8,15 It also seems reasonable to anticoagulate a child with progressive deterioration or during the initial evaluation of a new cerebral infarction.8 The loading dose of heparin is 75 units/Kg intravenously followed by 20 units/Kg/hour for children over one year of age (or 28 units/Kg/hour below one year of age). The target APTT to 60-85 is seconds.14
Adult stroke patients who receive low molecular weight heparin for ten days starting within 48 hours.of diagnosis have a better outcome,17 and it may be possible to adapt this approach for children. Low molecular weight heparin (Lovenex, Rhone-Poulenc) can be given to children subcutaneously in two divided doses of 1 mg/Kg/dose (or in neonates, 1.5 mg/Kg every 12 hours).
(1) Background: Experience in children with long term anticoagulation to prevent cerebral infarction is limited, and there is additional concern about anticoagulating an active child who may be prone to minor injuries through normal activities. Nevertheless, warfarin is the most effective means of prolonged anticoagulation in children.
(2) Safety: Clinical experience suggests that warfarin can be used in children and adolescents with reasonable safety. The concern that active children could have an increased risk of hemorrhage due to trauma seems to be largely unfounded, though it is recommended that they avoid activities which carry an especially high risk of injury (e.g., contact sports).
(3) Efficacy: The rationale for using warfarin in children with cerebrovascular disorders follows closely the approach used in adults. Thus, major uses of warfarin treatment in children include congenital or acquired heart disease, hypercoagulable states, arterial dissection, and dural sinus thrombosis. An INR of 2.0 to 3.0 is appropriate for most children on warfarin; for children with mechanical heart valves the INR should be 2.5 to 3.5.
(1) Background: There is ample reason to seek new treatments for children with ischemic cerebral
infarction, because 75% of the children have serious sequelae including neurologic deficit, epilepsy, or death. While there is little information about the use of thrombolytic agents in children with stroke, enough work has been done with adult patients that the technique could possibly be adapted for selected children.
(2) Safety: Urokinase and streptokinase are used infrequently in children with cerebrovascular disease, but no serious complications occurred in the few children treated for dural sinus thrombosis. Thrombolytic therapy for children with non-cerebral thrombotic complications has recently been evaluated. Pooled literature analysis of 203 children treated with thrombolytic agents (including 39 patients who received tPA) indicated that the thrombus was cleared in 80% of the children, but 54% had minor bleeding (not requiring transfusion) and one child suffered an intracranial hemorrhage. In 29 consecutive children treated with tPA (0.5 rng/`Kg) at Toronto's Hospital for Sick Children, the clot was dissolved in 79%, but almost a fourth of these children had bleeding which required transfusion.19,20 Given this high rate of serious bleeding after systemic tPA and the lack of studies demonstrating improved outcome, we can not recommend tPA except in the setting of a controlled clinical trial.
(3) Efficacy: The delayed diagnosis which so often occurs in children with ischemic stroke reduces the likelihood that a child with an ischemic stroke will be seen early enough to benefit from thrombolytic agents. Intravascular urokinase or streptokinase have been used with apparent success in a few children with dural sinus thrombosis, 8,21-23 but there is even less experience with these agents in children with arterial thrombosis. The available data are insufficient to comment on the effectiveness of any of the thrombolytic agents in children with ischemic stroke. Certainly they would be expected to produce unacceptable roles of bleeding as seen in adults if given more than 4-6 hours after onset of stroke.
(1) Background: About half of the patients with a stroke due to sickle cell disease will have another stroke,11 and this increased risk can be reduced by repeated transfusions to suppress the level of circulating sickle hemoglobin to 30% or less. The risk of stroke increases again if the transfusions are discontinued even after a prolonged stroke-free interval, so most patients who begin transfusions must continue them.
Safety: Although the risk can be reduced by iron chelation, iron toxicity from repeated blood transfusions remains a major problem. Cohen and colleagues 24 proposed a less aggressive transfusion program to maintain the hemoglobin S near 50%; this regimen required an average of 31% less transfused blood and still no infarctions occurred. Miller and colleague had similar results, although their follow-up period was shorter. This new approach needs to be studied further.
Efficacy: Although no randomized clinical trials were ever done, years of clinical experience have produced general agreement that periodic transfusion greatly reduces the risk of ischemic cerebral infarction due to sickle cell disease. A patient who has had one stroke has about a 90% risk of having additional infarctions. The Stroke Prevention Trial in Sickle Cell Anemia (STOP) is now investigating the use of transcranial Doppler (TCD) to identify children at greatest risk for their first cerebral infarction due to sickle cell disease. This study could prove that periodic transfusions reduce the risk of ischemic infarction in children with sickle cell disease and that TCD can be used to identify those at greatest risk.
Given the paucity of information about many aspects of childhood stroke, what is the best approach to the diagnosis and management of stroke in children? How should our methods in children differ from those used in adults? Until more information on childhood stroke is available, we must of necessity continue to adapt the knowledge obtained from adult stroke patients. Itshould not be necessary to repeat in children all the work already done in adults, but we do need to identify areas which are age specific.
In some respects, our study of stroke in children recapitulates some of the early work in adult stroke patients. Databases such as the Canadian Pediatric Ischemic Stroke Registry will continue to provide data on the causes of childhood stroke as well as the patients' treatment and outcome. Under the best of circumstances, such databases are limited by the fact that the correct diagnosis may not be recognized or reported to the registry. Larger case series which concentrate on one cause of stroke or one anatomic lesion need to be published. Epidemiologic studies need to be reassessed to reflect better diagnostic techniques and the increased recognition of stroke in children by physicians.
Several specific causes of cerebrovascular disease are relatively common in children and have a high enough risk of stroke to make collaborative trials feasible. There are several potentially productive areas to study. Research should initially focus on the more common disorders or on children with risk factors which are usually identified before a stroke occurs:
(a) The Stroke Prevention Trial in Sickle Cell Anemia (STOP) trial now underway could serve as a model for studies of childhood stroke from other causes. Sickle cell disease is common in some medical centers, and cerebral infarction frequent enough to make a study feasible. Early diagnosis and treatment probably improve the patient's outlook. Additional multicenter trials for patients with sickle cell disease could also address the use of hydroxyurea or other drugs in stroke prevention.
(b) Sinovenous thrombosis seems to occur relatively more often in children than other cerebrovascular lesions and can now be identified quickly and noninvasively with MRI/MRA. Collaborative studies to evaluate systemic anticoagulation and/or thrombolysis should be feasible, particularly if similar trials in adults continue to show promise.
Cardiac disease remains the most common cause of ischemic cerebral infarction in children. Most of these children have congenital heart lesions which are identified well before an infarction occurs, and ischemic infarction may occur frequently enough to warrant controlled trials of prophylactic agents or of neuro-protective agents during surgery when the risk of stroke is higher. Thrombolytic agents could play a greater role in children with heart disease because their families could be taught to recognize the significance of an acute neurologic deficit.
Moyamoya is an uncommon condition but it could be studied via a collaborative approach. Most patients in the recent literature have had various surgical procedures designed to increase blood flow to the brain. But no controlled trials have ever been done to assess these operations, and there is some evidence that the natural history of untreated moyamoya may be less devastating than sometimes suggested. In one group of 27 children, 5 patients had no sequelae, 9 had only headache or transient ischemic symptoms, and 7 had mild intellectual or motor impairment. Only 6 of the 27 had a poor outcome: 1 death, 2 who required continuous care, and 3 who required special schooling or institutionalization. Only 11 of these 27 patients had surgery.26 The fact that so many patients do well without intervention makes it difficult to evaluate treatment in the absence of controlled trials.
Several pediatric hospitals offer extracorporeal membrane oxygenation (ECMO), a technique which requires ligation of the right carotid artery. In some centers, the carotid artery is eventually reconstructed once ECMO is no longer needed.27 These children provide an opportunity to study the long term effects of altered cerebral circulation and, for the children whose carotid is reopened, to explore the effects of carotid artery trauma on the development of atherosclerosis.
Increased awareness of these disorders by the public, and by medical personnel will potentially improve accessibility of pediatric stroke patients to newer forms of thrombolytic and neuroprotective agents. Increased awareness by research teams and research funding agencies will provide the means for the intervention trials critically necessary to realize that potential.
|Table 1: Risk Factors for Pediatric Cerebrovasculardisease*|
Congenital Heart Disease
Ventricular septal defect
Atrial septal defect
Patent ductus arteriosus
Complex congenital heart defects
Acquired Heart Disease
Rheumatic heart disease
Prosthetic heart valve
Systemic Vascular Disease
Volume depletion or systemic hypotension
Superior vena cava syndrome
Systemic lupus erythematosus
Inflammatory bowel disease
Drug abuse (cocaine, amphetamines)
Malignant atrophic papulosis
NADH-CoQ reductase deficiency
Vasospasm with subarachnoid hemorrhage
Hematologic Disorders and Coagulopathies
Hemoglobinopathies (sickle cell anemia, sickle
Immune thrombocytopenic purpura
Thrombotic thrombocytopenic purpura
Disseminated intravascular coagulation (DIC)
Leukemia or other neoplasm
Congenital coagulation defects
Oral contraceptive use
Pregnancy and the postpartum period
Antithrombin IR deficiency
Protein S deficiency
Protein C deficiency
Congenital serum C2 deficiency
Liver dysfunction with coagulation defect
Vitamin K deficiency
Structural Anomalies of the Cerebrovascular System
Arterial fibromuscular dysplasia
Agenesis or hypoplasia of the internal carotid or
Hereditary hemorrhagic telangiectasia
Fat or air embolism
Foreign body embolism
Carotid ligation (e.g., ECMO)
Vertebral occlusion following abrupt cervical rotation
Posttraumatic arterial dissection
Blunt cervical arterial trauma
Posttraumatic carotid cavernous fistula
Coagulation defect with minor trauma
Amniotic fluid/placental embolism
Penetrating intracranial trauma
* Modified from Roach and Riela. Pediatric Cerebrovascular Disorders, New York. Futura Publishing Co., 1995.
Figure 1: Effect of Age At Event On Mean Diagnosis Time* (O-18 yrs.; N=80)
*data from the Canadian Pediatric Ischemic Stroke Registry
Return to Table of Contents
Last updated May 17, 2011