stroke in children

Stroke in ChildrenE. Steve Roach, MD

AddressDepartment of Neurology, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75235, USA.

Current Treatment Options in Neurology 2000, 2:295–303Current Science Inc. ISSN 1092-8480Copyright © 2000 by Current Science Inc.

Opinion statement• Identification and treatment of the underlying risk factors for stroke reduce the

potential for additional strokes; therefore, a thorough search for treatable risk fac-tors is justified. Because some risk factors can have a cumulative effect, even chil-dren with known risk factors for stroke sometimes need to be evaluated for other conditions. Cerebral angiography is often helpful; I recommend angiography in any child with an unexplained infarction or hemorrhage. Angiography is especially important in children with intraparenchymal hemorrhage because more than one third of such children will prove to have some type of potentially treatable congen-ital vascular anomaly such as an arteriovenous malformation (AVM) or aneurysm.

• The evidence that periodic blood transfusion effectively prevents cerebral infarction due to sickle cell disease is compelling. Transfusions apparently must be continued indefinitely to maintain the reduction of stroke risk, and without iron chelation, chronic transfusion eventually results in severe iron toxicity and, most likely, death, so the decision to begin transfusion is not an easy one. Measurement of the time-averaged mean flow velocity in the large cerebral vessels with transcra-nial Doppler (TCD) is highly predictive of stroke risk in these children, enough to justify its routine use in screening patients with sickle cell disease for stroke risk. I believe that patients with sickle cell disease should be offered chronic trans-fusion after an initial large-vessel stroke or when the TCD results suggest a high risk of stroke. The family must be made aware of the serious complications of chronic transfusion and the importance of complying with chelation once it is started.

• There are no controlled clinical trials to guide the use of anticoagulants, antiplatelet agents, or thrombolytic agents in children, although these drugs are being used more and more often in pediatric patients. For the most part, our approach has been adapted from our experience with adults. Heparin followed by warfarin is often used for sinovenous thrombosis and for arterial dissection. I also suggest long-term anti-coagulation for children with coagulopathy or a high risk of embolism due to congen-ital or acquired cardiac disease. It is reasonable to use a thrombolytic agent in children with an acute infarction; because few children present soon enough after the onset of symptoms, however, thrombolysis is infrequently used.

• Aspirin is used more than other antiplatelet agents in children, largely because of years of experience with aspirin and the lack of evidence that other agents are more effective. Despite its frequent use, there are no unequivocal indications for the use of aspirin in children. Aspirin is often started empirically in children sus-pected to be at substantial risk for additional ischemic stroke but whose risk is ill defined, an approach not too dissimilar from that often used in adult patients. Although the risk of Reye’s syndrome in a child taking daily aspirin for stroke pre-vention is a common concern, I know of no published examples of children who developed Reye’s syndrome while taking prophylactic aspirin. This apparently low risk must be weighed against the often-considerable risk of ischemic stroke that could be reduced by the use of daily aspirin. In situations such as vasculopathy or infarction of unknown cause, the small risk of Reye’s syndrome seems acceptable.

296 Cerebrovascular Disorders

IntroductionAlthough cerebrovascular disorders occur much lessfrequently in children than in the elderly, stroke inchildren is not rare. Clinicians have become more awareof vascular disorders in children, and the widespreadapplication of noninvasive diagnostic studies such asmagnetic resonance imaging, magnetic resonanceangiography, and CT allow confirmation of a diagnosisthat in previous years would have been missed or atleast not recognized as a vascular lesion. Additionally,the number of children who develop cerebrovascularlesions from certain risk factors may have increasedbecause more effective treatments for some causes ofstroke have allowed patients to survive long enough todevelop vascular complications [1, Class IIIc].

Aside from incidence, the most important distinc-tion between cerebrovascular disorders in children andin adults is the variety of conditions that promote strokein children [2, Class IIIc]. Recognition and treatment ofthe cause are essential if subsequent strokes are to beprevented, and it is sometimes the underlying conditionthat determines the patient's eventual outcome. Unlessthe cause is recognized and can be corrected, a childmay have decades-long exposure to the risk factor,greatly increasing the likelihood of subsequent strokes.If a thorough evaluation is done, a likely cause can bedetected in at least three quarters of the children withischemic infarction and in most children with intracere-bral hemorrhage [3,4, Class IIIa2].

Rational treatment of stroke in children is hamperedby the lack of controlled treatment trials. Such studiesare difficult to perform in children because of the rela-tive rarity of stroke at this age and because of the diver-sity of the causes of stroke in children. Nevertheless,some of our experience in treating stroke in adults canbe adapted for use in children, and accumulating expe-rience with antithrombotic and anticoagulant therapyin children suggests that these agents can be effective inchildren, though we could still benefit from age-specifictrials to establish the drugs’ efficacy and optimal dose.Likewise, thrombolytic agents should theoretically be aseffective in children as in adults, but the safety data areincomplete for children, and the effective dosage forchildren and adolescents needs to be determined.

CAUSES OF STROKE IN CHILDRENPerhaps the most fundamental difference between cere-brovascular diseases in children and in adults is thevariety of risk factors seen in children. Congenital heartdisease and sickle cell disease (SCD), for example, arecommon causes of stroke in children, whereas athero-sclerosis is rare. No risk factors are identified in aboutone fifth of children with ischemic infarction, but theprobability of identifying the cause depends on the

thoroughness of the evaluation. The recognized riskfactors for stroke in children are too numerous to dis-cuss in detail here, but detailed information is available[1,2, Class IIIc].

The source of intracranial hemorrhage is more oftenidentified [1, Class IIIc; 4, Class IIa2]. Common causesof hemorrhage include structural vascular lesions,hemoglobinopathy, and coagulation or platelet dis-orders. Intracranial hemorrhage makes up a larger pro-portion of strokes in children than in adults. In adults,hemorrhage accounts for only 15% to 20% of strokes.Small AVMs or cavernous malformations may not beevident on magnetic resonance imaging, CT, or evenangiography, particularly if surrounded by acute hemor-rhage and edema. For this reason, it may be necessary torepeat the evaluation after the hemorrhage has startedto resolve; some would advocate delaying the angio-gram until the hemorrhage has passed the acute phase.

Cardiac disease Congenital or acquired heart diseaseprobably remains the most common cause of ischemicstroke in children, although the number of children withstroke due to heart disease seems to have diminished inrecent years owing to the tendency to perform correctivesurgery for congenital lesions at earlier and earlier ages.Even so, about 20% of patients with ischemic infarctionhave heart disease, most often some type of congenitalanomaly. Complex cardiac anomalies involving both thevalves and the chambers are collectively the biggest prob-lem, but virtually any acquired or congenital cardiaclesion can sometimes promote stroke.

Arterial dissection Arterial dissection is probably anunderrecognized cause of stroke in children. A dissec-tion can occur after trauma or spontaneously. The injurypreceding an arterial dissection is not always severe,however, suggesting that the artery undergoing dissec-tion may have been structurally unsound [5, Class IIIb].Spontaneous carotid dissection affects the carotid arte-ries more often than the vertebral arteries and the cervi-cal carotid arteries more than their intracranialbranches. Aside from the absence of trauma, the clinicalfindings are similar regardless of whether the dissectionis traumatic or spontaneous [6,7, Class IIIb]. Infarctioncan develop almost immediately after injury, butdelayed embolism can occur hours, days, or even weeksafter the dissection. Either a fluctuating clinical courseor saltatory deterioration suggests distal embolism fromthe site of dissection.

Sickle cell disease Sickle cell disease is a commoncause of stroke in children, although the stroke risk isnot as high as once estimated. Balkaran et al. [8, Class

Stroke in Children Roach 297

IIIa] reported 17 strokes in a group of 310 Jamaicanchildren with SCD followed from birth to at least 14years of age. Two of the 17 patients had subarachnoidhemorrhage, one from an aneurysm. Thus, about 5% ofthe patients in this series suffered ischemic infarction by14 years of age, compared with previous estimates ofabout 15%. Almost half the patients with SCD whohave had one stroke subsequently have another.Although the functional significance of these lesions isnot always obvious, small ischemic infarctions probablycontribute to reduced cognitive function [9, Class IIIa1].Occasional patients with SCD develop intraparenchy-mal hemorrhage, sinovenous occlusion, or subarach-noid hemorrhage. Adults with SCD are more likely todevelop intraparenchymal hemorrhage than arechildren and adolescents, probably because of theweakened collateral channels distal to the arterial steno-sis [10, Class IIIa2; 11, Class IIIa1].

Vasculitis Intracranial vasculitis in children has multi-ple causes and may present clinically with arterialthrombosis, intraparenchymal or subarachnoid hemor-rhage, or sinovenous occlusion. Bacterial meningitis isprobably the most common cause of intracranial arteri-tis and stroke in children, although it may be hard todifferentiate the effects of the infection itself from thoseof stroke in very sick children. Most of the residual focalneurologic deficits following bacterial meningitis prob-ably result from secondary vascular occlusion andstroke. In one series with 198 children with tuberculousmeningitis, 21% had unilateral and 10% had bilateralbasal ganglia infarction [12, Class IIIa2]. Cerebralinfarction from tuberculous meningitis occurs morefrequently in children than in adults. Other disordersthat are well-known causes of arteritis in adults, such astemporal arteritis and isolated angiitis of the nervoussystem, are rarely reported in children.

Varicella A possible link between varicella zosterinfection and ischemic infarction in children has beenpostulated for several years [13,14, Class IIIb]. Most ofthe reported children have experienced a stroke in theanterior circulation, most often in the basal ganglia,and few children have had more than one episode[15–17, Class IIIb]. As often as chickenpox occurs inchildren, proving that it causes cerebral infarction isdifficult. The most convincing evidence so far is arecent case-control study that found that 7 of 11 chil-dren (64%) with idiopathic cerebral infarction hadvaricella during the previous 9 months, compared withonly 4 of the 44 children (9%) in the control group[18•, Class IIb]. How varicella leads to stroke is stilluncertain, but many of these children have transientcerebral arteriopathy [19, Class IIIa2]. Varicella couldlead to delayed stroke via persistent virus and arteritis,similar in some ways to the ipsilateral cerebral infarc-

tion sometimes seen in adults several weeks or monthsafter herpes zoster ophthalmicus.

Moyamoya Moyamoya syndrome is a vasculopathy ofthe cranial arteries, typically the internal carotid arte-ries, leading to progressive intracranial arterial occlu-sion with distal telangiectatic collateral vessels. Thesecollateral vessels are visible as an angiographic blushdistal to the occluded large artery. Moyamoya was orig-inally described in the Japanese and has the highestincidence in Japan, although it is recognized less oftenin most populations. Moyamoya is probably best clas-sified as a radiographic syndrome rather than a specificdisease entity, but often the term moyamoya disease isapplied to children with no defined risk factors (eg,Down’s syndrome, neurofibromatosis, or cranial irra-diation) for vasculopathy [20,21, Class IIIb]. Mostpatients have no known risk factors, but there is noguarantee that these individuals all have the sameunderlying defect.

Congenital vascular anomalies Congen i ta l vascu la rlesions collectively constitute the greatest reason fornontraumatic intraparenchymal hemorrhage inchildren. In our retrospective series of 68 childrenand adolescents with nontraumatic intraparenchymalbrain hemorrhage, for example, almost one third ofthe patients had an AVM or arteriovenous fistula, andaltogether 29 children (42.6%) had some type ofcongenital vascular anomaly [4, Class IIIa2]. The clin-ical presentation of intraparenchymal hemorrhagedue to AVM is no different from that of hemorrhagedue to other causes; however, some patients have pre-existing epileptic seizures, neonates with an AVMoften present with high-output cardiac failure, andinfants sometimes develop hydrocephalus.

Thrombocytopenia Thrombocytopenia has numerouscauses, but most often it results from either immuno-pathic thrombocytopenic purpura or from cancerchemotherapy. In our series of 68 children and adoles-cents with nontraumatic intraparenchymal brainhemorrhage, thrombocytopenia was the most commonhematologic risk factor for brain hemorrhage (8 of 68patients, or 11.8%). Five of these eight patients devel-oped thrombocytopenia from cancer chemotherapyand two others from isoimmune thrombocytopenia [4,Class IIIa2]. Nontraumatic brain hemorrhage due toreduced platelet count does not usually occur withcounts of more than 20,000/mm3, and even with lowercounts spontaneous hemorrhage is uncommon. Brainhemorrhage later in the course of immunopathicthrombocytopenic purpura frequently coincides with asystemic viral infection, probably because the infectionstimulates the production of antiplatelet antibodies andproduces a further decrease in the number of platelets.


Coagulation defects Hemophilia A (factor VIII defi-ciency) and B (factor IX deficiency) are the two mostcommon hereditary coagulation defects to cause intra-cranial hemorrhage, although intracranial bleeding dueto other deficiencies such as deficiency of factors XIII orV sometimes occurs. It is generally the severity of thebleeding tendency rather than the specific coagulationdefect that defines the risk of hemorrhage [22, ClassIIIa2]. Trauma precedes intracranial hemorrhage in upto three fourths of hemophiliacs, though the head

trauma is not always severe, and symptoms fromhemorrhage may be delayed for several days after thetrauma [23, Class IIIa1]. The routine use of vitamin Kinjections in newborns has all but eliminated vitamin Kdeficiency, once the most common acquired coagula-tion defect, as a cause of hemorrhage in neonates in theUnited States. Infants born to mothers taking anticon-vulsants sometimes bleed excessively, apparently owingto reduction in the number of vitamin K–dependentcoagulation factors.

Treatment

• Large lesions can generate increased intracranial pressure, but smaller lesions in the cerebral hemispheres generally do not produce enough mass effect to require treatment. We usually do not treat the edema unless it is extensive or unless the patient is deteriorating.

• Large hemispheric infarcts or hemorrhages and smaller lesions in the cere-bellum are another matter. As with increased intracranial pressure from other causes, rapid reduction of the PaCO2 by hyperventilation is the quickest way to lower the pressure. Patients with a large cerebral hemor-rhage and those with a cerebellar infarction or hemorrhage may benefit from surgical evacuation of the lesion. Steroids are not effective for cyto-toxic edema, and we do not ordinarily use them in stroke patients.

• Epileptic seizures are relatively common during the acute stage of both intracranial hemorrhage and infarction, but most patients do not have seizures and there is no evidence that the prophylactic use of antiepileptic agents is beneficial. We do not advocate prophylactic treatment except for children with subarachnoid or intraparenchymal hemorrhage due to a suspected aneurysm or an AVM, in whom a seizure might increase the risk of rebleeding. We try to use a nonsedating drug such as phenytoin rather than phenobarbital.

Heparin and low-molecular-weight heparin

Standard dosage Heparin: The loading dose of heparin is 75 U/kg intravenously followed by 20 U/kg/h for children older than 1 year of age (or 28 U/kg/h for those younger than 1 year of age) [24, Class IIIc]. We use a target-activated partial thromboplastin time of 45 to 60 seconds, although others aim for 60 to 80 seconds [25, Class IIIc].

Low-molecular-weight heparin (Lovenox; Rhone-Poulenc Rorer Pharmaceuti-cals, Collegeville, PA): This drug can be given to children subcutaneously in two divided doses of 1 mg/kg per dose (or in neonates, 1.5 mg/kg every 12 hours) [25, Class IIIc].

Contraindications Active bleeding or high risk of hemorrhage, thrombocytopenia with antiplatelet antibodies, evidence of drug hypersensitivity.

Main drug interactions There is no specific information for use in children, but other drugs that could cause or exacerbate bleeding, such as warfarin and antiplatelet agents, should be used with caution.

Main side effects Hemorrhagic transformation of infarction or exacerbation of other hemorrhage, spinal epidural hematoma following lumbar puncture, thrombocytopenia.

Special points As with adult patients, the use of anticoagulation in children with acute cerebral infarction rests on two questions: 1) What is the likelihood of either extension of the infarction or a second infarction from an embolus that might be prevented by

Pharmacologic treatment


treatment? 2) What is the risk of inducing a hemorrhage because of anticoagula-tion? We anticoagulate children with a high recurrence risk and a minimal risk of secondary hemorrhage. Most of the time, heparin is used initially while the war-farin dose is adjusted for long-term use. How long to continue anticoagulation is somewhat arbitrarily based on whether complications have occurred and on our estimation of the degree of ongoing risk.

Using this approach, anticoagulation is often used in children with arterial dissection, dural sinus thrombosis, coagulation disorders, or a high risk of embo-lism. The use of anticoagulation in children with arterial dissection or sinovenous thrombosis is based on the available data from adult series. Although intraparen-chymal hemorrhage is common behind the thrombosed sinus, there is little indica-tion that anticoagulation worsens the bleeding and increasing evidence of an improved outcome. Some children receive heparin during the initial assessment but do not continue long-term treatment unless a specific reason is discovered.

Experience with more than 100 children treated for systemic clots with low-molecular-weight heparin indicates reasonable safety; none of these children developed hemorrhagic complications [26, Class IIa]. Eventually, low-molecular-weight heparin may replace unfractionated heparin because of its more predictable pharmacologic action and more favorable side-effect profile.

Cost/cost effectiveness The usefulness of the medication has not been established in children.

Warfarin

Standard dosage Loading dose: 0.2 mg/kg on day one; for patients with liver dysfunction or a prior Fontan procedure, consider starting at 0.1 mg/kg/d. On days two through four, administer 25% to 50% of the initial dose, based on the International Normalized Ratio (INR). An INR of 2.0 to 3.0 is appropriate for most children receiving war-farin; for children with mechanical heart valves, the INR should be 2.5 to 3.5 [25, Class IIIc]. An INR of greater than 3.0 has also been suggested for patients with the antiphospholipid antibody syndrome [27, Class IIIa1].

Contraindications Active bleeding, hemorrhagic arterial infarction, recent or pending surgery of the eye or nervous system, high risk of significant trauma, and pregnancy. Patients with protein C deficiency should receive heparin before warfarin therapy is initiated.

Main drug interactions Numerous drugs either increase or diminish the anticoagulant effect of warfarin. A partial list of drugs that increase anticoagulation include allopurinol, amio-darone, chloramphenicol, cimetidine, methyldopa, phenothiazines, sulfonamides, and tamoxifen. Drugs that decrease the anticoagulation effect of warfarin include barbiturates, carbamazepine, griseofulvin, haloperidol, oral contracep-tives, and rifampin.

Main side effects Gastrointestinal or genitourinary bleeding or excessive bleeding following trauma are the primary concerns with warfarin. Changes in diet or the addition of other drugs can either increase or diminish the anticoagulant effect of warfarin, either of which could put the patient at increased risk.

Special points There are no controlled clinical trials that specifically demonstrate the efficacy of warfarin in children with a risk of stroke, but warfarin treatment in children with cerebrovascular disorders parallels the approach used in adults. Indications for warfarin treatment in children include congenital or acquired heart disease, hyper-coagulable states, arterial dissection, and dural sinus thrombosis. Children already taking heparin usually begin warfarin on the first or second day of heparin treat-ment. Some authors recommend starting warfarin after 5 days of heparin in patients with deep vein thrombosis or pulmonary embolism [25, Class IIIc].

The fear that active children could have an increased risk of hemorrhage due to trauma incurred during normal daily activities is logical but unsubstantiated; they should avoid activities that carry an especially high risk of injury, such as contact sports. Clinical experience suggests that warfarin can be used in most children and adolescents with reasonable safety.

Cost/cost effectiveness Warfarin is a relatively inexpensive medication; the dose and monthly cost vary with the age and size of the child.


Antiplatelet agentsAspirin is used more often than other antiplatelet agents in children because of years of experience with aspirin and the lack of evidence that other agents are more effective.

Standard dosage A daily dose of 2 to 3 mg/kg/d induces an antiplatelet effect, but it is not yet certain how clinically effective this dose is in children.

Contraindications Active bleeding, hypersensitivity to aspirin, urticaria, bronchospasm.Main drug interactions Combination with nonsteroidal anti-inflammatory agents could potentiate gastric

irritation. Other drugs that could cause or exacerbate bleeding should be used with caution in conjunction with aspirin.

Main side effects Daily aspirin at this dose seems to be reasonably well tolerated in most children. In addition to the potential complications of chronic aspirin use seen in adults (such as excessive bleeding), children taking daily aspirin might have an increased risk of developing Reye’s syndrome. I am not aware of any children who have devel-oped Reye’s syndrome or had severe bleeding that was attributed to aspirin. We have used daily aspirin in about three dozen children without complications. One 65-year-old man developed Reye’s syndrome while taking aspirin for stroke pro-phylaxis, but he also took additional aspirin for influenza [28, Class IIIb].

Special points There are no controlled clinical trials that demonstrate the efficacy of aspirin or other antiplatelet agents in children with a risk of stroke. Just how well aspirin works in children and the most effective dose of aspirin are unanswered questions. Nevertheless, aspirin is often used empirically in children suspected to be at risk of additional ischemic stroke but whose risk factors are ill defined. Reasonable uses of aspirin include idiopathic stroke or transient ischemic attacks, recurrent infarc-tion or transient ischemic attacks in children not previously treated with aspirin or who are not candidates for anticoagulants, and cerebral infarction and patent fora-men ovale [25, Class IIIc].

Controlled clinical trials with aspirin or other antiplatelet agents have not been done 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.

Cost/cost effectiveness Aspirin is inexpensive; the dose and the monthly cost vary with the age and size of the child.

Thrombolytic agents

Standard dosage Tissue plasminogen activator: Based its use in adults, tissue plasminogen activator (tPA) can be given at 0.9 mg/kg, with 10% given as an intravenous bolus and the remainder infused over the following hour [29, Class I]. No guidelines have been published for infusion directly into a thrombosed vessel in children, but 2-mg boluses into the thrombosed vessel, to a maximum cumulative dose of 0.2 mg/kg, seem to be reasonable.

Urokinase: This drug is not currently available, but its use may eventually be resumed. The approach is to administer a loading dose of 4000 U/kg over 10 minutes, followed by 4000 U/kg/h for 6 hours [25, Class IIIc].

Contraindications Active bleeding, hemorrhagic stroke, delayed presentation after ischemic stroke, recent neurosurgical procedure.

Main drug interactions Other drugs that could cause or exacerbate bleeding should be used with caution.Main side effects Reperfusion of infarcted brain carries a risk of hemorrhage. None of the throm-

bolytic agents has seen extensive use in children with cerebrovascular disease. Thrombolysis for children with noncerebral thrombotic complications has recently been evaluated. Of 203 children pulled from the literature who were treated with thrombolytic agents (including 39 patients who received tPA), the thrombus was cleared in 80% of the children, but 54% had minor bleeding (not requiring trans-fusion) and one child suffered an intracranial hemorrhage. In 29 consecutive chil-dren treated with tPA in one series, the clot was dissolved in 79%, but almost one fourth had bleeding that required transfusion [30, Class IIIc]. Some of these children evidently received a higher tPA dose (0.5 mg/kg/h for 6 hours) [25, Class IIIc] than used in the adult trials, which might increase the com-


plication rate in this series. Although experience with tPA is limited in children, it is probably reasonable to use it cautiously in children, using the same approach as in adults.

Special points It is not unusual for children with a cerebral infarction to be brought to a physician several hours or even days after the onset of symptoms. Until a greater sense of urgency develops, thrombolytic agents are unlikely to play an important role in the treatment of stroke in children. There is little information about the effectiveness and the risks of thrombolytic agents in children, although some of our experience with adults can probably be adapted for selected children. Children with arterial occlusion who are seen within 3 hours would be suitable candidates for thromboly-sis, and longer intervals after venous thrombosis may be acceptable.

Urokinase is presently not available owing to production concerns. Direct administration of urokinase has been used with apparent success in a few children with dural sinus thrombosis [31, Class IIIa2; 32, Class IIIb], but there is little pub-lished experience with these agents in children with arterial thrombosis. Streptoki-nase is not usually recommended.

The available data are insufficient to recommend tPA strongly in children with ischemic stroke outside the setting of a clinical trial, but there is no reason to believe that it would be less effective in children than in adults provided it is used in a similar fashion.

Cost/cost effectiveness Usefulness of these medications has not been established in children.

Blood transfusion for sickle cell disease

Standard procedure Red cell transfusions are given approximately every 3 to 4 weeks, sufficiently often to maintain the percentage of sickle hemoglobin below 30%. Iron chelation is started after the serum ferritin level increases.

Contraindications Lack of compliance with iron chelation is at least a relative contraindication to chronic transfusion.

Complications Chronic transfusion programs are justified despite concern about eventual iron toxicity because about half the patients with one stroke due to SCD will have another stroke.

Special points The Stroke Prevention in Sickle Cell Study Project (STOP) investigated the effec-tiveness of periodic transfusions to prevent the first stroke in children at high risk (ie, those whose mean average blood-flow velocity consistently exceeded 200 cm/sec) [33••, Class I]. Children were randomly assigned to receive either periodic blood transfusions or standard clinical care (no transfusion). The study was termi-nated early because the stroke frequency in the group not receiving transfusion was 10 times higher than in the group receiving transfusion. Thus, the stroke risk from SCD can be predicted by TCD, and chronic transfusion can prevent even the first stroke due to SCD in children at high risk [33••, Class I].

Stopping the transfusions after a stroke-free interval reintroduces the higher stroke risk, but just how much the sickle hemoglobin has to be suppressed to maintain the lower stroke risk is debatable. In one report, the frequency of the transfusion schedule was reduced in 15 patients with SCD, none of whom had another stroke during the 4-year follow-up period. Maintaining the hemoglobin S near 50% (instead of the usual target of 30%) required an average of 31% less transfused blood [34•, Class IIIa2]. Complete cessation of transfusion after a stroke-free interval reintro-duces the higher risk of stroke [35, Class IIIa2], but, if effective, reduction of the amount of transfused blood would reduce the risk of iron toxicity [34•, Class IIIa2]. The optimal transfusion frequency needs further study.

Cost/cost effectiveness The cost of monthly transfusions varies in different institutions, but the trans-fusion cost is probably less than the treatment cost incurred after repeated cerebral infarctions.

Interventional procedures


Surgery for moyamoyaVarious surgical procedures have been touted for moyamoya: cervical sympathec-tomy, intracranial grafts of omentum or temporalis muscle, and direct or indirect bypass of the superficial temporal artery to the cerebral circulation. Without controlled studies, it is hard to gauge the relative value of these procedures.

Standard procedure A trial of medical management with aspirin is acceptable before committing the patient to surgery [25, Class IIIc]. I have also used warfarin in two children in whom aspirin failed and who were not candidates for surgery, but considering the increasing risk of hemorrhagic complications in adults with moyamoya, I would prefer not to use warfarin in adult patients.

Adults and larger children typically undergo direct anastomosis of the super-ficial temporal artery to the middle cerebral artery. Because of the smaller vessel caliber in younger children, many surgeons prefer to do an indirect bypass (the encephaloduroarteriosynangeiosis procedure) [36, Class IIIa1].

Contraindications Patient too unstable for surgery, coagulopathy.Complications Anesthetic complications, hemorrhage, intracranial infection, ischemic stroke.Special points No randomized controlled clinical trials have been done for either medical or surgical

treatments of moyamoya, but patency of the grafted vessel can often be demon-strated as well as postoperative improvement of the cerebral blood flow. Yamada et al. [37, Class IIIa1] compared preoperative and postoperative angiograms after encephaloduroarteriosynangeiosis and found excellent collateral vasculature in 30%, good collateral development in 46%, and poor collateral development in 24%. The prognosis for patients with moyamoya is not uniformly grim without surgery, however, and the fact that many patients do relatively well makes it difficult to evaluate treatments. Fukui [38•, Class IIIa], for example, presented data from 821 patients in the Japanese moyamoya registry. A little more than two thirds of these patients were treated with various surgical procedures, whereas the remain-ing 23% were treated medically. There was no significant difference between the medical and the surgical treatment groups, although as the author points out, there could have been a tendency to operate on the more severely affected patients. Indications for surgery include symptomatic cerebrovascular insufficiency due to carotid vasculopathy, especially if medical management has been ineffective, as well as inadequate perfusion reserve and cerebral blood flow [39, Class IIIa1].

Cost/cost effectiveness The cost varies with the specific procedure used and the institution where the surgery is done, but the one-time cost of successful surgery is cheaper than the long-term costs of caring for a severely disabled individual.

References and Recommended ReadingPapers of particular interest, published recently, have been highlighted as:• Of special interest•• Of outstanding interest

1. Roach ES, Riela AR: Pediatric Cerebrovascular Disorders, edn 2. New York: Futura; 1995.

2. Riela AR, Roach ES: Etiology of stroke in children. J Child Neurol 1993, 8:201–220.

3. Brower MC, Rollins N, Roach ES: Basal ganglia and thalamic infarction in children. Etiology and clinical features. Arch Neurol 1996, 53:1252–1256.

4. Al-Jarallah M, Al-Rifai T, Riela AR, Roach ES: Spon-taneous intraparenchymal hemorrhage in children: a study of 68 patients. J Child Neurol 2000, 15:284–289.

5. Adelman LS, Doe FD, Sarnat HB: Bilateral dissecting aneurysms of the internal carotid arteries. Acta Neuro-pathol 1974, 29:93–97.

6. Lannuzel A, Moulin T, Amsallem D, et al.: Vertebral artery dissection following a judo session: a case report. Neuropediatrics 1994, 25:106–108.

7. Schievink WI, Mokri B, Piepgras DG: Spontaneous dissections of the cervicocephalic arteries in child-hood and adolescence. Neurology 1994, 44:1607–1612.

8. Balkaran B, Char G, Morris JS, et al.: Stroke in a cohort of patients with homozygous sickle cell disease. J Pediatr 1992, 120:360–366.

9. Kinney TR, Sleeper LA, Wang WC, et al.: Silent cerebral infarctions in sickle cell anemia: a risk factor analysis. Pediatrics 1999, 103:640–645.

Surgery


10. Van Hoff J, Ritchey AK, Shaywitz BA: Intracranial hem-orrhage in children with sickle cell disease. Am J Dis Child 1985, 139:1120–1123.

11. Ohene-Frempong K, Weiner SJ, Sleeper LA, et al.: Cerebrovascular accidents in sickle cell disease: rates and risk factors. Blood 1998, 91:288–294.

12. Schoeman JF, Van Zyl LE, Laubscher JA, Donald PR: Serial CT scanning in childhood tuberculous menin-gitis: prognostic features in 198 cases. J Child Neurol 1995, 10:320–329.

13. Kamholz J, Tremblay G: Chickenpox with delayed contralateral hemiparesis caused by cerebral angiitis. Ann Neurol 1985, 18:358–360.

14. Bodensteiner JB, Hille MR, Riggs JE: Clinical features of vascular thrombosis following varicella. Am J Dis Child 1992, 146:100–102.

15. Silverstein FS, Brunberg JA: Postvaricella basal ganglia infarction in children. Am J Neuroradiol 1995, 16:448–452.

16. Caekebeke JFV, Peters ACB, Vandvik B, et al.: Cerebral vasculopathy associated with primary varicella infec-tion. Arch Neurol 1990, 47:1033–1035.

17. Kovacs SO, Kuban K, Strand R: Lateral medullary syndrome following varicella infection. Am J Dis Child 1993, 147:823–825.

18.• Sebire G, Meyer L, Chabrier S: Varicella as a risk factor for cerebral infarction in childhood: a case-control study. Ann Neurol 1999, 45:679–680.

This recent case-control study offers the best evidence so far that varicella is a risk factor for stroke in children. Seven of 11 children (64%) with idiopathic ischemic stroke had varicella during the 9 months before their stroke, compared to only 4 of the 44 children (9%) in the control group.19. Chabrier S, Rodesch G, Lasjaunias P, et al.: Transient

cerebral arteriopathy: a disorder recognized by serial angiograms in children with stroke. J Child Neurol 1998, 13:27–32.

20. Mitchell WG, Fishman LS, Miller JH, et al.: Stroke as a late sequela of cranial irradiation for childhood brain tumors. J Child Neurol 1991, 6:128–133.

21. Aylett SE, Britton JA, De Sousa CMCP: Down syndrome and moyamoya disease: presentation with subarach-noid hemorrhage. Pediatr Neurol 1996, 14:259–261.

22. Eyster ME, Gill FM, Blatt PM, et al.: Central nervous system bleeding in hemophiliacs. Blood 1978, 51:1179–1188.

23. Andes WA, Wulff K, Smith WB: Head trauma in hemo-philia: a prospective study. Arch Intern Med 1984, 144:1981–1983.

24. Michelson AD, Bovill E, Andrew M: Antithrombotic therapy in children. Chest 1995, 108:506S–522S.

25. Andrew M, deVeber G: Pediatric Thromboembolism and Stroke Protocols. Hamilton, Ontario: B.C. Decker Inc.; 1997:1–63.

26. Massicotte P, Adams M, Marzinotta V, et al.: Low molec-ular weight heparin in pediatric patients with throm-botic disease: a dose finding study. J Pediatr 1996, 128:313–318.

27. Khamashta MA: The management of thrombosis in the antiphospholipid-antibody syndrome. N Engl J Med 1995, 332:993–997.

28. Peters LJ, Wiener GJ, Gilliam J, et al.: Reye's syndrome in adults: a case report and review of the literature. Arch Intern Med 1986, 146:2401–2403.

29. National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group: Tissue plasminogen activator for ischemic stroke. N Engl J Med 1995, 333:1581–1587.

30. Leaker M, Massicotte MP, Brooker L, Andrew M: Throm-bolytic therapy in pediatric patients: a comprehensive review of the literature. Thromb Haemost 1996, 76:132–134.

31. Horowitz M, Purdy P, Unwin H, et al.: Treatment of dural sinus thrombosis using selective catheterization and urokinase. Ann Neurol 1995, 38:58–67.

32. Griesemer DA, Theodorou AA, Berg RA, Spera TD: Local fibrinolysis in cerebral venous thrombosis. Pediatr Neurol 1994, 10:78–80.

33.••Adams RJ, McKie VC, Hsu L, et al.: Stroke prevention trial in sickle cell anemia (“STOP”): study results. N Engl J Med 1998, 339:5–11.

This multicenter randomized controlled clinical trial demon-strated conclusively that periodic blood transfusion reduces the risk of ischemic stroke due to sickle cell disease in patients at higher risk (determined by TCD).34.• Cohen AR, Martin MB, Silber JH, et al.: A modified

transfusion program for prevention of stroke in sickle cell disease. Blood 1992, 79:1657–1661.

This study of 15 patients with sickle cell disease suggests that a less intense transfusion program designed to maintain the sickle hemoglobin near 50% could prevent most infarctions, with a lower risk of iron toxicity.35. Wang WC, Kovnar EH, Tonkin IL, et al.: High risk of

recurrent stroke after discontinuance of five to twelve years of transfusion therapy in patients with sickle cell disease. J Pediatr 1991, 118:377–382.

36. Matsushima T, Inoue T, Suzuki SO, et al.: Surgical treatment of moyamoya disease in pediatric patients—comparison between the results of indirect and direct revascularization procedures. Neurosurgery 1992, 31:401–405.

37. Yamada I, Matsushima Y, Suzuki S: Childhood moyamoya disease before and after encepalo-duro-arterio-synangiosis: an angiographic study. Neuro-radiology 1992, 34:318–322.

38.• Fukui M: Current state of study on moyamoya disease in Japan. Surg Neurol 1997, 47:138–143.

Of the 821 patients from the Japanese moyamoya registry summarized in this report, 23% were treated without surgery; no significant differences in outcome were noted between the medically and the surgically treated groups.39. Ikezaki K: A rational approach to treatment of

moyamoya disease in childhood. J Child Neurol 2000, 15:350–356.

stroke in children

Documents