The Call for Multicenter Studies of Pediatric Stroke
To the Editor:
We were pleased to see the call for large, collaborative studies of pediatric stroke by Dr Zahuranec et al.1 The authors present population-based incidence data and power calculations detailing the overwhelming number of pediatric stroke cases needed to identify risk factors in case-control studies. As members of the on-going International Pediatric Stroke Study (IPSS), a multicenter prospective childhood ischemic stroke registry, we would like to thank the authors and the editors for highlighting the need for large multicenter studies in childhood stroke, and wish to elaborate on both the role for, and difficulties associated with, such studies.
We believe the most pressing issue in childhood stroke at this time is the identification of effective secondary stroke prevention measures. Outside of certain subgroups of high-risk children (eg, sickle cell disease or congenital heart disease), stroke is rare, and the development of strategies for primary stroke prevention that can be widely implemented is probably unrealistic. Recurrent stroke, however, occurs with alarming frequency (in up to 25% of children within 2 years of an initial stroke, even with best medical management),2 and proven strategies for secondary stroke prevention are lacking. Pediatric neurologists must therefore manage children with stroke by extrapolating from adult studies, but given the vastly different etiologies in children versus adults, this strategy is clearly not adequate.
We cannot design rational randomized controlled trials (RCTs) for secondary stroke prevention in children until data regarding rates and predictors of recurrent stroke are available. The goals, then, of collaborative studies of childhood stroke must also include identifying these rates and predictors. However, as Dr Zahuranec et al highlight, the challenges facing such studies are great.
The first challenge is simply identifying ischemic stroke in children. The poor awareness of pediatric stroke among pediatricians, poor sensitivity of early CT scanning, and wider differential diagnosis for focal deficits in children frequently delay this diagnosis. In the authors’ population-based study, ICD-9 code searches of hospital discharge and emergency department diagnoses identified 6 hemorrhagic strokes, but only 1 ischemic stroke. This ratio of hemorrhagic to ischemic stroke is much higher than the 1:1 ratio previously reported for children, and their incidence of ischemic stroke (0.6 per 100 000) is lower than prior reports.3,4 Although this may reflect the uniqueness oftheir population, it could also reflect difficulty identifying cases by ICD-9 codes. In our own experience with on-going population-based studies, we have been surprised by the low yield of stroke-related ICD-9 code searches compared with text-string searches of radiology databases. Common reasons for missed ICD-9 diagnosis include: (1) stroke diagnosis not being coded at discharge in a child with severe medical illnesses (eg, stroke in the setting of overwhelming sepsis, leukemia, a lupus flare, or cardiac surgery), (2) stroke diagnosed only after discharge when either additional neuroimaging is performed or an in-patient imaging study is reinterpreted, or (3) children with stroke diagnosed and managed only as out-patients (eg, occipital stroke presenting with minor visual deficits).
The second major challenge, as identified by the authors, is identifying sufficient pediatric stroke patients to achieve adequate power in a study. Depending on the prevalence of the risk factor and its strength of association with stroke, they estimated that hundreds to thousands of cases are needed. We similarly estimated that—depending on stroke recurrence rates, the treatment agents used, and the specific population targeted—RCTs for secondary stroke prevention in children would require between 215 and 1600 subjects.5 Only large, well-coordinated multicenter efforts will ever be able to produce such numbers.
Although a population-based cohort study is the superior study design for analyzing risk factors for a first stroke, prospective pediatric stroke registries that are not population-based, such as the IPSS, should provide meaningful data on rates and predictors of stroke recurrence at those centers that will conduct the RCTs. Although limited in terms of generalizability (all subjects have been enrolled at academic tertiary-care centers), the IPSS offers the advantage of rapid enrollment through its web-based registry. Since its initiation in 2002, the IPSS has expanded from 9 to more than 60 centers spanning 17 countries with 76 co-investigators. Over 500 patients from 18 centers have been enrolled into the web-database (https://www.sickkids.on.ca/cstrokestudy). The IPSS has also facilitated the development of a network of investigators interested in childhood stroke, and willing to participate in future RCTs.
In conclusion, we agree that the time is ripe for large multicenter studies of childhood stroke with the ultimate goal of obtaining the evidence desperately needed to guide the management of these children.
Zahuranec DB, Brown DL, Lisabeth LD, Morgenstern LB. Is it time for a large, collaborative study of pediatric stroke? Stroke. 2005; 36: 1825–1829.
Lanthier S, Kirkham FJ, Mitchell LG, Laxer RM, Atenafu E, Male C, Prengler M, Domi T, Chan AK, Liesner R, deVeber G. Increased anticardiolipin antibody IgG titers do not predict recurrent stroke or TIA in children. Neurology. 2004; 62: 194–200.
Broderick J, Talbot GT, Prenger E, Leach A, Brott T. Stroke in children within a major metropolitan area: the surprising importance of intracerebral hemorrhage. J Child Neurol. 1993; 8: 250–255.
Lynch JK, Hirtz DG, deVeber G, Nelson KB. Report of the National Institute of Neurological Disorders and Stroke workshop on perinatal and childhood stroke. Pediatrics. 2002; 109: 116–123.