Stroke Study Symptomatic Neonatal Arterial Ischemic Stroke: The International Pediatric

Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

Symptomatic Neonatal Arterial Ischemic Stroke: The International Pediatric

Stroke Study
Adam Kirton, Jennifer Armstrong-Wells, Taeun Chang, Gabrielle deVeber, Michael J.
Rivkin, Marta Hernandez, Jessica Carpenter, Jerome Y. Yager, John K. Lynch, Donna
M. Ferriero and for the International Pediatric Stroke Study Investigators
Pediatrics 2011;128;e1402; originally published online November 28, 2011;
DOI: 10.1542/peds.2011-1148

The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://pediatrics.aappublications.org/content/128/6/e1402.full.html

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly


publication, it has been published continuously since 1948. PEDIATRICS is owned,
published, and trademarked by the American Academy of Pediatrics, 141 Northwest Point
Boulevard, Elk Grove Village, Illinois, 60007. Copyright © 2011 by the American Academy
of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.

Downloaded from pediatrics.aappublications.org at University of Edinburgh on June 8, 2012


Symptomatic Neonatal Arterial Ischemic Stroke: The
International Pediatric Stroke Study
WHAT’S KNOWN ON THIS SUBJECT: Neonatal arterial ischemic AUTHORS: Adam Kirton, MD,a Jennifer Armstrong-Wells,
stroke is a common cause of cerebral palsy and other lifelong MD,b Taeun Chang, MD,c,d Gabrielle deVeber, MD,e
neurologic disabilities. Studies to date have been limited by Michael J. Rivkin, MD,f Marta Hernandez, MD,g Jessica
modest sample sizes of regional and heterogeneous populations, Carpenter, MD,d Jerome Y. Yager, MD,h John K. Lynch, DO,i
and Donna M. Ferriero, MD,j for the International
limiting advances in understanding and treatment.
Pediatric Stroke Study Investigators
aDepartments of Pediatrics and Clinical Neuroscience,
WHAT THIS STUDY ADDS: Results from this sample of nearly 250
University of Calgary and Alberta Children’s Hospital, Calgary,
newborns with neonatal AIS provide a novel global view of clinical Alberta, Canada; bSection of Neurology, Department of
presentations, potential risk factors, current investigational and Pediatrics, University of Colorado and Children’s Hospital
treatment practices, and early outcomes. Important directions Colorado, Denver, Colorado; cNeurophysiology and Epilepsy,
dDepartment of Neurology, Children’s National Medical Center,
for future research were also identified.
Washington, DC; eDivision of Neurology, Department of
Pediatrics, Hospital for Sick Children, Toronto, Ontario, Canada;
Departments of fNeurology, Psychiatry, and Radiology,
Children’s Hospital Boston, Harvard Medical School, Boston,
Massachusetts; gDepartamento de Pediatrica, Pontificia
abstract Universidad Catolica de Chile, Santiago, Chile; hPediatric
Neurosciences, Department of Pediatrics, Faculty of Medicine
BACKGROUND: Neonatal arterial ischemic stroke (AIS) has emerged and Dentistry, University of Alberta, Edmonton, Alberta, Canada;
as a leading cause of perinatal brain injury, cerebral palsy, and lifelong iNational Institute of Neurological Disorders and Stroke,

disability. The pathogenesis is poorly understood, which limits the de- National Institutes of Health, Bethesda, Maryland; and jNewborn
Brain Research Institute, University of California San Francisco,
velopment of treatment and prevention strategies. Multicenter studies
UCSF Benioff Children’s Hospital, San Francisco, California
must define epidemiology, risk factors, treatment practices, and out-
comes to advance clinical trials and improve the adverse outcomes KEY WORDS
stroke, newborn, perinatal stroke
suffered by most survivors.
ABBREVIATIONS
METHODS: The International Pediatric Stroke Study is a global re- AIS—arterial ischemic stroke
search initiative of 149 coinvestigators (30 centers in 10 countries). IPSS—International Pediatric Stroke Study
Patients with clinical and neuroimaging confirmation of symptomatic OR—odds ratio
neonatal AIS were enrolled (2003–2007). Standardized, Web-based CI—confidence interval
data entry collected clinical presentations, risk factors, investigations, CT—computed tomography
treatments, and early outcomes. We examined predictors of infarct All listed authors made substantive intellectual contributions to
characteristics and discharge outcome by using multivariate logistic this study, including conception and design, acquisition of data,
and/or analysis and interpretation of data; drafting and/or
regression. revising of the article for important intellectual content; and
RESULTS: Two hundred forty-eight neonates were studied (57% male, final approval of the version to be published. All listed authors
10% premature). Most of them presented with seizure (72%) and non- qualify for authorship and have participated sufficiently in the
focal neurologic signs (63%). MRI was completed for 92% of the infants, work to take public responsibility for appropriate portions of
the content.
although ⬍50% had vascular imaging. Infarcts preferentially involved
the anterior circulation and left hemisphere and were multifocal in www.pediatrics.org/cgi/doi/10.1542/peds.2011-1148
30%. Maternal health and pregnancies were usually normal. Neonates doi:10.1542/peds.2011-1148
often required resuscitation (30%) and had systemic illnesses (23%). Accepted for publication Aug 15, 2011
Cardiac and prothrombotic abnormalities were identified in ⬍20% of Address correspondence to Adam Kirton, MD, Alberta Children’s
the infants. Antithrombotic treatment was uncommon (21%) and var- Hospital, 2888 Shaganappi Trail NW, Calgary, Alberta, Canada T3B
ied internationally. Half (49%) of the infants had deficits at discharge, 6A8. E-mail: adam.kirton@albertahealthservices.ca
and data on their long-term outcomes are pending. PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).
CONCLUSIONS: Newborns with AIS are often systemically sick, Copyright © 2011 by the American Academy of Pediatrics
whereas their mothers are usually healthy. Definitive causes for most FINANCIAL DISCLOSURE: The authors have indicated they have
neonatal AISs have not been established, and large-scale case-control no financial relationships relevant to this article to disclose.
studies are required to understand pathogenesis if outcomes are to be
improved. Pediatrics 2011;128:e1402–e1410

e1402 KIRTON et al
Downloaded from pediatrics.aappublications.org at University of Edinburgh on June 8, 2012
ARTICLES

Perinatal stroke has emerged as a demiologists. The long-term goal of the expressed by postnatal week (ethical
common cause of lifelong neurologic prospective registry was to develop restrictions on birth dates). Neonates
disability. Neonatal arterial ischemic and execute international clinical born at ⱕ36 or ⱖ41 weeks’ gestation
stroke (AIS) is a common variety and is trials (https://app3.ccb.sickkids.ca/ were considered to be premature or
defined as acute symptomatic, focal cstrokestudy/). Prospective enroll- postmature, respectively. Investiga-
cerebral infarction in an arterial terri- ment extended from January 1, 2003, tional sites were grouped according to
tory between birth and 28 days of life to July 1, 2007 (149 coinvestigators, 30 region (Europe, Canada, United States,
that is confirmed by neuroimaging.1,2 centers, 10 countries) and included South America, Asia, or Australia).
Most survivors suffer neurologic mor- symptomatic neonatal AIS but not pre- Birth weight was trichotomized
bidity, and perinatal stroke is the lead- sumed perinatal ischemic strokes. The (⬍2500, 2500 – 4000, or ⬎4000 g). Sea-
ing cause of hemiplegic cerebral IPSS office (Hospital for Sick Children) son of event was adjusted according to
palsy.2 Many infants with AIS incur ad- managed Web-based data entry and hemisphere.
ditional sequelae including impair- the master database. Consensus- Potential risk factors were classified
ments in language, cognition, and be- based definitions for diagnosis, inves- into consensus IPSS categories based
havior and epilepsy.3–5 Case-control tigations, outcomes, and treatment on case-control studies and theoreti-
data are limited, and little is under- were applied. Study identification cal considerations only including
stood regarding pathophysiology.2,6 numbers were assigned at enrollment. cardiac, prothrombotic, acute and
That adverse outcomes last for de- Data were deidentified and entered chronic illnesses, arteriopathy, and
cades amplifies the impact of neonatal into a password-protected, Web-based neonatal (maternal, pregnancy, neo-
AIS on patients, their families, and system. Clinical care was not pre- natal, and obstetric). Neonates with
society. scribed by the IPSS. Methods were ap- systemic illness or significant resusci-
Studies of neonatal AIS have been lim- proved by site research ethics boards tation were classified as having “acute
ited by sample size and inconsistent with informed consent. The complete neonatal illness.” Prothrombotic testing
terminology, data collection, and risk- IPSS methodology is described and investigator interpretation varied
factor evaluations. The International elsewhere.11 across institutions, hence the term “pos-
Pediatric Stroke Study (IPSS) was es- sible” prothrombotic abnormality. Treat-
Population
tablished to standardize approaches ment categories included antithrom-
to childhood stroke research on a Cases were enrolled by site investiga- botic, anticonvulsant, and other.
global scale. Treatment options re- tors using established clinical and Outcomes, discharge destination, and
main to be established, and practice radiographic criteria5 including (1) causes of death were recorded.
patterns seem widely discrepant.7 acute neurologic deficit or seizure and
Consensus-based guidelines for pedi- (2) radiographic confirmation of acute Analysis
atric stroke8–10 offer little direction in focal cerebral infarction(s) within ar-
Numerators were expressed over the
neonatal AIS, and global practices re- terial territories corresponding to
available population unless otherwise
garding identification, investigation, clinical manifestations. Infants with
stated. For each outcome of interest,
and management must be established the following conditions were exclud-
variables predefined on evidence were
ed: neonatal cerebral sinovenous
to facilitate systematic studies and compared (␹2 analysis, dichotomous; t
clinical trials. thrombosis,12 presumed perinatal
tests, continuous) and expressed as
ischemic stroke, intracranial hemor-
We examine here the presentations, odds ratios (ORs) with 95% confidence
rhage, global hypoxic-ischemic injury,
clinical associations, investigations, intervals (CIs). To determine indepen-
periventricular leukomalacia, and
treatments, and early outcomes of a dent predictors, multivariate logistic re-
metabolic injury.
large, global population of neonates gression models incorporated univari-
with AIS. Data Abstraction ate P values of ⱕ .1 with testing for
colinearity. Stata 10.0 (College Station,
METHODS Investigators collected data—
TX) was used for statistical calculations.
demographics (age, gender, race/eth-
International Pediatric Stroke nicity, location), clinical presentations, RESULTS
Study imaging, potential risk factors and
The IPSS was established in 2003 by 11 evaluations, treatments, and dis- Patient Population
coinvestigators including pediatric charge outcomes— by using stan- A total of 1194 patients were enrolled
neurologists, hematologists, and epi- dardized forms. Age at diagnosis was in the IPSS, 347 (29%) of which were

PEDIATRICS Volume 128, Number 6, December 2011 e1403


Downloaded from pediatrics.aappublications.org at University of Edinburgh on June 8, 2012
neonates. Neonates with AIS totaled
248 and represented our population
(71% of all neonates, 21% of all AISs).
Median gestational age was 39.0
weeks (mean: 39.0 ⫾ 1.9; range: 31–
42). Birth weight was considered nor-
mal (86%) or small (10%) or large (4%)
for gestational age, and 55% had birth
weights lower than the median for ges-
tational age. Multiple (twin) gestations
were uncommon (3 [1%]). Nineteen of
the infants were born prematurely
(8%), and another 20 (10%) were born
postmaturely. Gender was discrepant;
140 (57%) were male (detailed analy-
sis described elsewhere).11 Geo-
graphic origin of the patients was
United States (51%), Europe (19%),
Canada (16%), South America (6%),
Australia (6%), or Asia (3%) (Fig 1).
FIGURE 1
Clinical Presentations International distribution of neonatal AIS cases. During the period of prospective enrollment (January
1, 2003, through July 1, 2007), the IPSS included 149 coinvestigators at 30 centers in 10 countries.
Most patients presented in the first According to geographic category, the United States constituted the largest share of patients, and
approximately two-thirds of them were enrolled in North America. LOC indicates level of
week of life (87%). Seizures were consciousness.
common at presentation (178
[72%]), as were diffuse neurologic
signs (157 [63%]); the most common
sign was abnormal tone (38%) or
level of consciousness (39%). A mi-
nority of the patients demonstrated
focal neurologic signs (30%), the ma-
jority of which were lateralizing
hemiparesis (95%). Systemic find-
ings included respiratory (26%) and
feeding (24%) difficulties (Fig 2).

Diagnosis and Neuroimaging


A brain MRI was obtained from nearly FIGURE 2
all of the infants (92%). Computed to- Clinical presentations of neonatal AIS. Most neonates presented with seizures (dark), although diffuse
mography (CT) scans were obtained neurologic signs (altered level of consciousness or tone) and systemic abnormalities (respiratory or
feeding difficulties) were frequent. Fewer than 30% of the newborns were described as having focal
for 119 (48%) infants. AIS lesions were neurologic deficits. AUS indicates Australia; CAN, Canada; SA, South America.
single in 70% and isolated to the ante-
rior circulation (71%), posterior circu-
lation (9%), or both (20%). Strokes Multiple vascular territory AIS was multivariate analysis, only neonatal re-
were more commonly left-sided (left common (30%). In univariate analysis, suscitation was independently associ-
alone in 51%, right alone in 25%), associations with multiple-territory in- ated with multiple-territory infarction
whereas bilateral lesions were ob- farcts included neonatal resuscitation (OR: 5.5; P ⬍ .0001) (Table 1).
served in 24%. In those with only ante- (OR: 4.58; P ⬍ .0001), maternal hyper-
rior circulation strokes, a strong left- tension (OR: 4.29; P ⫽ .006), and acute Intracranial hemorrhage in 56 neo-
sided preference was seen (73%). systemic illness (OR: 2.13; P ⫽ .03). In nates (23%) was reported. Most of

e1404 KIRTON et al
Downloaded from pediatrics.aappublications.org at University of Edinburgh on June 8, 2012
ARTICLES

TABLE 1 Predictors of Infarct Characteristics


Univariate Multivariate
OR P 95% CI OR P 95% CI
Predictors of multiple infarcts
Significant neonatal resuscitation 4.58 ⬍.0001 20.9–10.02 5.5a ⬍.0001a 2.30–13.17a
Maternal hypertension 4.29 .006 1.52–12.07 2.91 .09 0.84–10.10
Neonatal acute systemic illness 2.13 .03 1.09–4.13 0.8 .66 0.29–2.18
Predictors of hemorrhagic infarct
Neonatal acute systemic illness 18.78 .008 2.13–165.74 8.72 .07 0.81–93.76
Neonatal sepsis 21.5 .002 3.01–153.5 6.69 .22 0.32–141.07
Neonatal acidosis 17.1 .004 2.52–116.2 0.63 .77 0.03–14.58
Bilateral lesions 20.29 .007 2.29–179.21 15.54a .03a 1.28–189.09a
The variables tested were week of event, region, gender, presentation (focal deficit[s], hemiparesis, diffuse neurologic signs, abnormal tone, altered level of consciousness, seizure, feeding
difficulties, respiratory difficulties), imaging completed (CT, MRI, computed tomography angiography, magnetic resonance angiography), lesion features (anterior circulation, posterior
circulation, both anterior and posterior circulations, multiple vascular territories, left any/only, right any/only, bilateral, hemorrhage [any/inside lesion/outside lesion]), cardiac risk
(congenital heart disease, other heart disease, related to surgery, related to catheterization [diagnostic or interventional]), arteriopathy, prothrombotic state, neonatal acute systemic
illness, neonatal sepsis, neonatal acidosis, birth weight, gestational age, prematurity, postmaturity, maternal age, natural conception, low Apgar score (at 1 or 5 minutes), maternal
hypertensive disorder, maternal diabetes, prolonged rupture of membranes, vaginal birth, any assisted delivery, urgent cesarean delivery, significant neonatal resuscitation, treatment (any
antithrombotic, any anticoagulant, unfractionated heparin, low molecular weight heparin, acetylsalicylic acid, anticonvulsants).
a Factor was independently associated with outcome on multivariate analysis (see text).

TABLE 2 Predictors of Neurologic Deficit at Discharge mained an independent predictor (OR:


Predictor Univariate 15.54; P ⫽ 0. 03) (Table 1).
OR P 95% CI Arterial imaging was performed on
Delivery characteristics 53% of the infants (magnetic reso-
Vaginal delivery 1.87 .04 1.05–3.36
Urgent Cesarean delivery 0.41 .01 0.21–0.83
nance angiography in 49%). CT and
Presentation conventional angiography were rare
Hemiparesisa 1.29 ⬍.0001 4.16–17.68 (4% and 3%, respectively). Abnormali-
Tone/reflex abnormalitya 4.56 ⬍.0001 2.26–9.21
ties of arterial imaging were reported
Any focal deficit 8.67 ⬍.0001 4.29–17.50
Seizure 0.37 .001 0.20–0.67 in 29 infants (12% overall, 22% of those
Feeding difficulties 2.31 .01 1.19–4.48 imaged). Abnormal vascular imaging
Newborn risk factors results included descriptions of arte-
Underlying chronic disease 2.32 .01 1.21–4.50
Prothrombotic state 2.24 .02 1.12–4.48 rial occlusion (9 [7%]) and possible ar-
Vasculopathy 9.45 .036 1.16–76.91 teriopathy (stenosis, dissection, or
Fever 3.77 .02 1.19–11.95 vasculitis) (10 [8%]), but further de-
Acidosis 6.65 .08 0.79–56.21
Acute head/neck disorder 2.82 .09 0.86–9.29
tails were not available. No associa-
Infarct characteristics tions with abnormal arterial imaging
Left-sided 0.39 .001 0.22–0.67 were identified.
Right-sided 2.36 .008 1.24–4.47
Hemorrhage outside 0.32 .04 0.11–0.93
Potential Risk factors
The variables tested were week of event, region, gender, presentation (focal deficit[s], hemiparesis, diffuse neurologic
signs, abnormal tone, altered level of consciousness, seizure, feeding difficulties, respiratory difficulties), imaging com- Median maternal age was 29 years
pleted (CT, MRI, computed tomography angiography, magnetic resonance angiography), lesion features (anterior circula-
tion, posterior circulation, both anterior and posterior circulations, multiple vascular territories, left any/only, right (mean: 29.8 ⫾ 5.6; range: 16 – 43). Fifty-
any/only, bilateral, hemorrhage [any/inside lesion/outside lesion]), cardiac risk (congenital heart disease, other heart two percent of women were primipa-
disease, related to surgery, related to catheterization [diagnostic or interventional]), arteriopathy, prothrombotic state,
neonatal acute systemic illness, neonatal sepsis, neonatal acidosis, birth weight, gestational age, prematurity, postmatu- rous, and 15% were of advanced ma-
rity, maternal age, natural conception, low Apgar score (at 1 or 5 minutes), maternal hypertensive disorder, maternal ternal age (ⱖ35 years). History of
diabetes, prolonged rupture of membranes, vaginal birth, any assisted delivery, urgent cesarean delivery, significant
neonatal resuscitation, treatment (any antithrombotic, any anticoagulant, unfractionated heparin, low molecular weight
previous miscarriage was recorded
heparin, acetylsalicylic acid, anticonvulsants). for 27 of the mothers (11%). Diagnosis
a Factor was independently associated with outcome on multivariate analysis (see text).
was not associated with month or sea-
son of birth. The majority of pregnan-
them (62%) were classified as hemor- temic illness (OR: 18.78; P ⫽ .008), aci- cies were conceived naturally (144 of
rhage within the infarct (hemorrhagic dosis (OR: 17.7; P ⫽ .004), sepsis (OR: 159 [91%]). Family-history data were
transformation). In univariate analy- 21.5; P ⫽ .002), and bilateral lesions incomplete, but not a single case of
sis, associations with hemorrhagic (OR: 20.29; P ⫽ .007). In multivariate perinatal stroke in a sibling was re-
transformation included acute sys- analysis, only bilateral lesions re- ported. Maternal medical conditions

PEDIATRICS Volume 128, Number 6, December 2011 e1405


Downloaded from pediatrics.aappublications.org at University of Edinburgh on June 8, 2012
were uncommon with the exception of (2%) strokes were associated with ex- Anticonvulsant treatment was com-
gestational hypertension and/or pre- tracorporeal membrane oxygenation. mon (169 [67%]). Early discontinuation
eclampsia, reported in 10%. Factors Results of prothrombotic testing were of anticonvulsants is suggested, be-
such as gestational diabetes, oligohy- inconsistent. The presence of a possi- cause they were present at discharge
dramnios/polyhydramnios, and prena- ble thrombophilia (see “Methods”) in only 2% of those treated. The only
tal trauma, infections, or bleeding was reported in 47 (19%) of the cases, other common medications were
were each recorded in ⬍1%. including an increased lipoprotein(a) antibiotics (23%).
Potential perinatal factors included level,13 methylene tetrahydrofolate re-
maternal fever (9%), prolonged rup- ductase (MTHFR) mutations (11 [7 Outcomes
ture of membranes (6%), and pres- hetero, 4 homozygous]), elevated ␤-2 Durations of hospital stay could not be
ence of meconium (23%). A prolonged glycoprotein level,8 factor V Leiden,6 accurately determined because of eth-
second stage of labor was reported for prothrombin gene 20210A,5 low anti- ical restrictions on the use of birth
22%. Mode of delivery included emer- thrombin III level,3 antiphospholipid dates. A majority of the infants were
gent cesarean (26%), elective cesar- antibodies,2 plasminogen activator in- discharged from the hospital (95%).
ean (15%), assisted vaginal (14%), and hibitor 1,2 and low protein S level.1 Six Neurologic deficits at discharge were
spontaneous vaginal (45%) delivery. children had multiple prothrombotic documented in 49% of them. Data re-
Apgar scores were usually normal abnormalities. garding type and severity of deficits
(score ⱕ 5 at 1 minute [25%] and 5 were incomplete. Only 5 deaths were
minutes [6%]). Significant early neona- Treatment reported (2%). Long-term outcome
tal resuscitation (assisted ventilation, Only fifty-two (21%) of the infants re- data are not yet available. In univariate
chest compressions, intubation, medi- ceived antithrombotic medication. analysis, only vaginal delivery (OR:
cations) was documented in 30% of the Most of them were anticoagulated (42 1.87; P ⫽ .04) predicted discharge def-
cases. The term “birth asphyxia” was [17% overall]) with low molecular icit, whereas urgent cesarean delivery
used in only 21 cases (8%). Data on weight heparin (28 [11%]), unfraction- seemed protective (OR: 0.41; P ⫽ .01).
placental abnormalities was recorded ated heparin (12 [5%]), or warfarin (2 Neonatal predictors of discharge defi-
in only 20 cases (8%), but of these [⬍1%]). Antiplatelet therapy (acetyl- cit in univariate analysis included focal
cases, 16 (80%) were classified as “ab- salicylic acid) was given to 14 (6%). deficits (OR: 8.67; P ⬍ .0001), hemipa-
normal.” Details of placental patho- Multiple antithrombotic medications resis (OR: 1.29; P ⬍ .0001), tone abnor-
logic findings were not recorded. Co- were given to 10 (19% treated, 4% mality (OR: 4.56; P ⬍ .0001), feeding
occurrence with acute systemic overall). Treatment patterns for anti- difficulties (OR: 2.31; P ⫽ .01),
illnesses was common (23%), includ- thrombotic medications differed geo- vasculopathy (OR: 9.45; P ⫽ .04), pro-
ing dehydration (22%), fever (7%), gas- graphically (P ⫽ .01): Europe, 44%; thrombotic state (OR: 2.24; P ⫽ .02),
troenteritis (9%), shock (8%), sepsis Canada, 27%; Australia, 17%; South and fever (OR: 3.77; P ⫽ .02). Newborn
(3%), acidosis (3%), and meningitis America, 16%; Asia, 14%; and United presentation with seizure decreased
(3%). As already defined, 40% had evi- States, 14%. the probability of discharge deficit (OR:
dence of an acute neonatal illness. Univariate analysis revealed an associ- 0.37; P ⫽ .001). Infarct characteristics
Associated cardiac factors were iden- ation between antithrombotic treat- associated with deficits included right
tified in 43 (18%) of the infants. Com- ment and cardiac disease (OR: 2.8; P ⫽ hemispheric lesions (OR: 2.36; P ⫽
plex congenital heart disease was .005) or presence of thrombophilia .008), whereas left hemispheric le-
most common (38 of 43 [88%]). Echo- (OR: 2.7; P ⫽ .005). Several markers of sions (OR: 0.39; P ⫽ .001) and hemor-
cardiography details were insufficient disease severity were associated with rhage outside the infarct (OR: 0.32; P ⫽
for further analysis. Acquired cardiac antithrombotic treatment, including .04) were inversely associated.
conditions were rare (2 [⬍1%]). Diag- presentation with diffuse signs (OR: Because presentation with focal defi-
nosis within 72 hours of a cardiac pro- 2.25; P ⫽ .02) or altered level of con- cits and hemiparesis were strongly
cedure was reported in 12 cases (28% sciousness (OR: 2.1; P ⫽ .02). Congeni- collinear, the former was not included
of cardiac cases, 5% overall), including tal or acquired heart disease was an in the multivariate model. Only the
cardiac surgery6 and catheterization independent predictor of receiving un- presence of hemiparesis (OR: 7.09
(6 [5 diagnostic]). No cases of clinically fractionated heparin (OR: 13.77; P ⬍ [95% CI: 1.54 –32.47]; P ⫽ .01) or tone
asymptomatic heart disease present- 0.001 and OR: 59.67; P ⫽ .008, abnormality (OR: 4.43 [1.30 –15.10];
ing with stroke were described. Five respectively). P ⫽ .02) at presentation were indepen-

e1406 KIRTON et al
Downloaded from pediatrics.aappublications.org at University of Edinburgh on June 8, 2012
ARTICLES

dently associated with discharge radiation to the neonatal brain.17 Our ing of the head and neck is required in
deficit. results have helped better character- neonatal stroke protocols to help re-
ize hemorrhagic changes associated solve this issue.
DISCUSSION with neonatal AIS. Associations with Disordered coagulation has long been
To our knowledge, this is the first re- markers of acute disease suggest that suspected of contributing to perinatal
port of a large multinational cohort of systemic processes such as altered stroke. Previous case-control data for
newborns with symptomatic AIS. Our hemostatic mechanisms might in- neonatal AIS are sparse, and estimates
data include several clinically relevant crease the likelihood of intracranial have varied widely (20%– 68%).18–22 Ev-
findings. Our population was generally bleeding. Although not evaluated in de- idence does support a role for throm-
“sicker” than those in previous stud- tail, severe hemorrhages that required a bophilia in neonatal AIS, particularly
ies. Neonates with AIS are typically de- change in management were not re- protein C deficiency, elevated lipopro-
scribed as nonencephalopathic, other- ported. Therefore, repeat imaging for tein(a) level, and factor V Leiden.23 A
wise well children who present at 1 to hemorrhagic changes might not be re- 2010 meta-analysis of thrombophilia in
2 days of life with seizures.3,13–15 How- quired, particularly with CT scanning. Ad- pediatric stroke found only 22 of 185
ever, at least 25% of our neonates dition of blood-sensitive magnetic reso- studies eligible, and only 6 focused on
showed signs of acute illness, includ- nance sequences (gradient echo, “perinatal stroke.”23 The authors con-
ing emergent cesarean delivery, re- susceptibility-weighted images) is a via- cluded that studies have been “contra-
suscitation, low Apgar scores, and/or ble alternative, and incorporation of dictory or inconclusive due to lack of
systemic illness. Despite this, the term such sequences to neonatal neuroim- statistical power.”23 Additional limiting
“birth asphyxia” or suggestion of aging protocols might enhance our un- factors of these studies and ours in-
global hypoxic-ischemic encephalopa- derstanding of disease mechanisms. clude inconsistent laboratory methods
thy was rare. Although they likely Noninvasive vascular imaging is now with a lack of controls matched for
share risk factors and AIS can co- routinely available and recommended perinatal factors including age, in
occur with hypoxic-ischemic encepha- by current pediatric stroke consensus which developmental changes in he-
lopathy,16 most cases seemed to guidelines.8 Despite this, fewer than mostasis are paramount and norma-
clearly distinguish the two. Such in- half of the neonates with AIS had cere- tive values are not well established.24
creased accuracy in diagnosis might bral angiographic studies performed. Our results do little to resolve the
reflect a combination of better neuro- Furthermore, dedicated angiography poorly understood role of thrombo-
imaging (diffusion MRI), enhanced ex- of the cervical vasculature was rarely philia in neonatal AIS, and fully pow-
perience of the IPSS investigators, and reported. Arterial dissection in neo- ered, carefully controlled studies are
better awareness of disease patterns. nates has been described and might required.
Such distinctions are essential for represent an underrecognized entity Consistent with the controversial role
choosing investigations, treatment op- with cervical imaging so underper- of antithrombotic treatment for neo-
tions, outcome prediction, family coun- formed. The documentation of arterial natal AIS, treatment was uncommon
seling, and research progress. occlusion in ⬍10% of the cases with and varied according to region. Recent
Neonates were unlikely to present with angiographic studies suggests multi- neonatal cerebral sinovenous throm-
focal deficits, which emphasizes the ple possibilities. This finding is per- bosis studies within the IPSS12 and
need for a high index of clinical suspi- haps most consistent with systemic elsewhere25 found similar interna-
cion and prompt neuroimaging to di- embolic events, although obvious risks tional discrepancies in treatment. Cur-
agnose stroke. Modern neuroimaging for embolism are present in a small rent consensus-based guidelines8–10
has greatly improved the detection minority of patients (eg, congenital vary in their recommendations re-
and understanding of neonatal AIS. heart disease). In addition, recanaliza- garding neonatal AIS treatment, al-
Our findings support the use of MRI tion of cerebral arteries might occur though they consistently support anti-
with diffusion-weighted sequences as quickly, and with most imaging per- coagulation therapy in congenital
the first-line imaging modality in most formed days after birth, occlusions heart disease, and such an association
cases of neonatal encephalopathy, might often not be evident. It is inter- was observed here. Our study could
which facilitates the accurate diagno- esting to note that despite a very low not examine the safety or efficacy of
sis of neonatal AIS. The role of CT scan- number of cases, “arteriopathy” was such interventions, although serious
ning is limited and has little advantage associated with deficits at discharge complications were not reported, and
over MRI but poses additional risk of (8 of 9 cases). Routine vascular imag- no association with intracranial hem-

PEDIATRICS Volume 128, Number 6, December 2011 e1407


Downloaded from pediatrics.aappublications.org at University of Edinburgh on June 8, 2012
orrhage was observed. Neonatal AIS important new knowledge but carries scribed here should be considered
carries an extremely low recurrence important limitations. As an interna- hypothesis-generating only. Fortu-
rate,2 so withholding anticoagulation tional registry, our samples are not nately, ongoing follow-up of this IPSS
therapy would seem reasonable out- population-based, and comparative cohort will facilitate future studies to
side the circumstance of congenital “normal” populations are not avail- more accurately characterize neuro-
heart disease. Antithrombotic clini- able. Unavoidable bias is present with logic outcomes.
cal trials will not likely occur for neo- IPSS investigators diagnosing, investi-
natal AIS; neuroprotective interven- gating, and enrolling patients; their ex- ACKNOWLEDGMENTS
tions seem a more likely future pertise differs from that of general Original start-up funding for the IPSS
focus. neonatologists or pediatric neurolo- was provided by the Child Neurology
gists. Accurate and complete data col- Foundation. The funding source played
Because most of the infants presented
lection across dozens of international no role in the design, collection, analy-
with seizures, the frequent use of anti-
centers and investigators is challeng- sis, or interpretation of data, writing
convulsants is expected. However,
ing, even with modern Web-based of the report, or the decision to submit
treatment seemed to be discontinued
data-entry systems and established the article.
by discharge for most patients. This
might parallel practices in other forms IPSS protocols. The detailed data and The original IPSS investigators were
of neonatal encephalopathy (eg, enormous sample size required to Steve Ashwal, MD (Loma Linda Univer-
hypoxic-ischemic encephalopathy), for answer the most challenging ques- sity School of Medicine, Loma Linda,
which the natural history is better un- tions of causation were not possible. CA), Gabrielle deVeber, MD, MHSc (Hos-
derstood and early cessation of sei- Our neurologic outcome data were pital for Sick Children, Toronto, On-
zure medications might be reasonable. only available at discharge, whereas tario, Canada), Donna Ferriero, MD
Stroke-induced neonatal seizures have much longer follow-up intervals (University of California, San Fran-
been studied less, and the possibilities (years) are required for meaningful cisco, CA), Heather Fullerton, MD, MAS
of both ongoing subclinical seizures interpretation. Also, selection of (University of California, San Fran-
and medication adverse effects need symptomatic cases only excludes im- cisco), Rebecca Ichord, MD (Children’s
to be considered. Prospective studies portant populations, including those Hospital of Philadelphia, Philadelphia,
of neonatal AIS seizure treatment in- with ultrasound-detected subcorti- PA), Fenella Kirkham, MB, BChir (Uni-
cluding cerebral monitoring are cal “perforator” lesions26 and later- versity College London Institute of
presenting presumed perinatal isch- Child Health, London, United Kingdom),
required.
emic stroke.27 John K. Lynch, DO, MPH (National Insti-
Our sample size provided a unique op- tute of Neurological Disorders and
portunity to perform multivariate lo- CONCLUSIONS Stroke, National Institutes of Health,
gistic regression and generate specu- Interpretation of our neurologic out- Bethesda, MD), Finbar O’Callaghan,
lative hypotheses regarding outcomes come data are extremely limited. Be- MBChB (Bristol Royal Hospital for Chil-
of interest. That presentation with cause most neonates do not manifest dren, Bristol, United Kingdom), Steve
acute focal deficits predicts the same obvious focal deficits and admission Pavlakis, MD (Maimonides Medical
at discharge is hardly surprising. How- times are generally short, it is not sur- Center, Brooklyn, NY), Guillaume
ever, could the association of neonatal prising that those who do will still have Sebire, MD, PhD (Université de Sher-
resuscitation with multiple infarcts re- discharge deficits. However, it is now brooke Fleurimont, Sherbrooke, Qué-
flect underlying diseased placenta? well established that children will bec, Canada), and Andrew Willan, BA,
Potential reasons for the independent “grow into their deficits” after perina- MSc, PhD (Hospital for Sick Children);
association observed between hemor- tal brain injury. Even simple motor out- the IPSS institutions that enrolled at
rhage and bilateral lesions are less ev- comes (cerebral palsy) are not rea- least 20 patients were (numbers in pa-
ident, and a wide CI should be inter- sonably characterized until 18 to 24 rentheses indicate the number of pa-
preted with caution. Currently minimal months, whereas more complex defi- tients enrolled) Hospital for Sick Chil-
understanding of neonatal stroke cits (eg, learning, behavioral) are often dren (147) (Gabrielle deVeber, MD,
pathophysiology suggests value in the not appreciated until school age.4,28,29 MHSc, Andrew Willan, BA, MSc, PhD,
consideration of such previously unre- Therefore, our results should not be Adam Kirton, MD, Mahendra Moharir,
ported associations. used to try to predict long-term neuro- MD, Rand Askalan, MD, PhD, and Mari-
Our analysis of a large, international logic outcomes in neonates with AIS, anne Sofronas, MA), Münster Univer-
sample of neonates with AIS provides and the potential associations de- sity Pediatric Hospital, Münster, Ger-

e1408 KIRTON et al
Downloaded from pediatrics.aappublications.org at University of Edinburgh on June 8, 2012
ARTICLES

many (122) (Ulrike Nowak- Göttl, MD, University School of Medicine (54) ter, Bangkok, Thailand (Montri Saeng-
Christine Düring, MD, and Anne Krüm- (Steve Ashwal, MD, and Chalmer Mc- pattrachai, MD), British Columbia Chil-
pel, MD), University of Texas South- Clure, MD, PhD), Schneider Children’s dren’s Hospital, Vancouver, British
western Medical Center, Dallas, Texas Hospital, New Hyde Park, New York (46) Columbia, Canada (Bruce Bjornson,
(94) (Michael M. Dowling, MD, PhD, Pa- (Li Kan, MD, MS, Robin Smith, MD, Jo- MD), Children’s Central Hospital,
tricia Plumb, RN, MSN, Janna Journey- seph Maytal, MD, and Rosemarie Sy- Tbilisi, Georgia (Nana Tatishvili, MD),
cake, MD, and Katrina van de Bruin- Kho, MD), Children’s National Medical Children’s Hospital of Buffalo, Buffalo,
horst, MA), Ohio Stroke Registry (94) Center, Washington, DC (39) (Jessica New York (E. Ann Yeh, MD), Children’s
Akron Children’s Hospital, Akron, Ohio Carpenter, MD, Taeun Chang, MD, and Hospital of Eastern Ontario, Ottawa,
(Abdalla Abdalla, MD), Cincinnati Chil- Steven Weinstein, MD), University of Ontario, Canada (Peter Humpherys,
dren’s Hospital Medical Center, Cincin- California, San Francisco (37) (Donna MD), Children’s Hospital of Wisconsin,
nati, Ohio (Tonya Phillips, MD), Cleve- Ferriero, MD, and Heather Fullerton, Milwaukee, Wisconsin (Catherine
land Clinic, Cleveland, Ohio (Neil MD, MAS), Maimonides Medical Center Amlie-Lefond, MD, and Harry T. Whelan,
Friedman, MD), MetroHealth Medical (26) (Steve Pavlakis, MD, Sharon Good-
MD), Denver Children’s Hospital, Den-
Center, Cleveland (Elie Rizkallah, MD), man, PNP, and Kim Levinson, PNP), Ri-
ver, Colorado (Timothy Bernard, MD,
Nationwide Children’s Hospital, Colum- ley Hospital, Indianapolis, Indiana (26)
and Neil A. Goldenberg, MD, PhD), Hos-
bus, Ohio (Warren Lo, MD, and Khaled (Meredith Golomb, MD, MSc), Winnipeg
pital Dr Sotero del Rio, Puente Alto,
Zamel, MD), Rainbow Babies and Chil- Children’s Hospital, Winnipeg, Mani-
Chile (Manuel Arriaza Ortiz, MD), Mc-
dren’s Hospital, Cleveland (Max Wiz- toba, Canada (24) (Mubeen Rafay,
Master University Medical Centre,
nitzer, MD, and Karen Lidsky, MD), Pon- MBBS, MSc, Frances Booth, MD, Mi-
tificia Universidad Catolica de Chile, chael Salman, MD, Charuta Joshi, MD, Hamilton, Ontario, Canada (Anthony
Santiago, Chile (78) (Marta Isabel Her- Namrata Shah, MD, and Monica Nash, Chan, MBBS), Miami Children’s Hospi-
nandez Chavez, MD), Royal Children’s RN), Children’s Hospital of New York, tal, Miami, Florida (Marcel Deray, MD,
Hospital, Melbourne, Victoria, Austra- New York, New York (22) (Geoffrey and Zaid Khatib, MD), Queen Mary Hos-
lia (75) (Professor Paul Monagle, Mark Heyer, MD), Great Ormond Street Hos- pital, Hong Kong, China (Virginia Wong,
MacKay, MD, Chris Barnes, MD, Janine pital, London, United Kingdom (21) (Vi- MD), Université de Sherbrooke Fleuri-
Furmedge, RN, BSc, and Anne Gordon, jeya Ganesan, MBChB, MD), Stollery mont (Guillaume Sebire, MD, PhD), Uni-
MSc, BAppSc), University of Utah and Children’s Hospital, Edmonton, Al- versity of Rochester Medical Center,
Primary Children’s Medical Center, berta, Canada (21) (Jerome Y. Yager, Rochester, New York (Jill M. Cholette,
Salt Lake City, Utah (70) (Susan L. Bene- MD), and Pediatric Institute Hospital, MD, Shalu Narang, MD, and Norma B.
dict, MD, and James F. Bale Jr, MD), Kuala Lumpur, Malaysia (20) (Hussain Lerner, MD, MPH), and University of
Children’s Hospital of Philadelphia (63) Imam, MBBS, FRCP, DCH); and the IPSS Texas, San Antonio, Texas (Shannon
(Rebecca Ichord, MD, Daniel Licht, MD, institutions that enrolled ⬍20 patients Carpenter, MD, and Kurt Bischoff,
and Sabrina Smith, MD), Loma Linda were Bangkok Hospital Medical Cen- MSc).
REFERENCES
1. Raju TN, Nelson KB, Ferriero D, Lynch JK; cerebral artery territory infarction. Stroke. group of the American Heart Association
NICHD-NINDS Perinatal Stroke Workshop 2008;39(2):403– 410 Stroke Council and the Council on Cardio-
Participants. Ischemic perinatal stroke: 5. Lee J, Croen LA, Lindan C, et al. Predictors of vascular Disease in the Young. Stroke. 2008;
summary of a workshop sponsored by the outcome in perinatal arterial stroke: a 39(9):2644 –2691
National Institute of Child Health and Hu- population-based study. Ann Neurol. 2005; 9. Monagle P, Chalmers E, Chan A, et al. Anti-
man Development and the National Institute 58(2):303–308 thrombotic therapy in neonates and
of Neurological Disorders and Stroke. Pedi- children: American College of Chest Physi-
6. Kirton A, deVeber G. Advances in perinatal
atrics. 2007;120(3):609 – 616 cians evidence-based clinical practice
ischemic stroke. Pediatr Neurol. 2009;40(3):
2. Nelson KB. Perinatal ischemic stroke. 205–214 guidelines (8th edition). Chest. 2008 ;133(6
Stroke. 2007;38(2 suppl):742–745 7. Fullerton H, deVeber G; IPSS Investigators. suppl):887S–968S
3. Sreenan C, Bhargava R, Robertson CM. Ce- Practice variability in the treatment of 10. Paediatric Stroke Working Group. Stroke in
rebral infarction in the term newborn: clin- childhood acute ischemic stroke: results of Childhood: Clinical Guidelines for Diagno-
ical presentation and long-term outcome. J the International Pediatric Stroke Study sis, Management and Rehabilitation. Lon-
Pediatr. 2000;137(3):351–355 [abstr]. Stroke. 2007;38(2):485 don, United Kingdom: Royal College of
4. Ricci D, Mercuri E, Barnett A, et al. Cognitive 8. Roach ES, Golomb MR, Adams R, et al. Man- Physicians: 2004
outcome at early school age in term-born agement of stroke in infants and children: a 11. Golomb MR, Fullerton HJ, Nowak-Gottl U, de-
children with perinatally acquired middle scientific statement from a special writing Veber G. Male predominance in childhood

PEDIATRICS Volume 128, Number 6, December 2011 e1409


Downloaded from pediatrics.aappublications.org at University of Edinburgh on June 8, 2012
ischemic stroke: findings from the interna- an increasing source of radiation exposure. atic review and meta-analysis of
tional pediatric stroke study. Stroke. 2009; N Engl J Med. 2007;357(22):2277–2284 observational studies. Circulation. 2010;
40(1):52–57 18. Golomb MR, MacGregor DL, Domi T, et al. 121(16):1838 –1847
12. Jordan LC, Rafay MF, Smith SE, et al. Anti- Presumed pre- or perinatal arterial isch- 24. Andrew M, Vegh P, Johnston M, Bowker J,
thrombotic treatment in neonatal cerebral emic stroke: risk factors and outcomes. Ann Ofosu F, Mitchell L. Maturation of the hemo-
sinovenous thrombosis: results of the Inter- Neurol. 2001;50(2):163–168 static system during childhood. Blood. 1992;
national Pediatric Stroke Study. J Pediatr. 19. deVeber G, Monagle P, Chan A, et al. Pro- 80(8):1998 –2005
2010;156(5):704 –710 thrombotic disorders in infants and chil- 25. Berfelo FJ, Kersbergen KJ, van Ommen CH,
13. Rafay MF, Cortez MA, de Veber GA, et al. Pre- dren with cerebral thromboembolism. Arch et al. Neonatal cerebral sinovenous throm-
dictive value of clinical and EEG features in Neurol. 1998;55(12):1539 –1543 bosis from symptom to outcome. Stroke.
the diagnosis of stroke and hypoxic isch- 20. Günther G, Junker R, Strater R, et al. Symp- 2010;41(7):1382–1388
emic encephalopathy in neonates with sei- tomatic ischemic stroke in full-term
26. Abels L, Lequin M, Govaert P. Sonographic
zures. Stroke. 2009;40(7):2402–2407 neonates: role of acquired and genetic pro-
templates of newborn perforator stroke.
thrombotic risk factors. Stroke. 2000;
14. Fujimoto S, Yokochi K, Togari H, et al. Neona- Pediatr Radiol. 2006;36(7):663– 669
31(10):2437–2441
tal cerebral infarction: symptoms, CT find- 27. Kirton A, deVeber G, Pontigon AM, MacGregor D,
ings and prognosis. Brain Dev. 1992;14(1): 21. Lynch JK, Nelson KB, Curry CJ, Grether JK.
Shroff M. Presumed perinatal ischemic stroke:
Cerebrovascular disorders in children with
48 –52 vascular classification predicts outcomes. Ann
the factor V Leiden mutation. J Child Neurol.
15. Govaert P, Ramenghi L, Taal R, de Vries L, 2001;16(10):735–744 Neurol. 2008;63(4):436–443
Deveber G. Diagnosis of perinatal stroke I: 28. Westmacott R, Barry V, MacGregor D, deVe-
22. Curry CJ, Bhullar S, Holmes J, Delozier CD,
definitions, differential diagnosis and regis- ber G. Intellectual function in preschool and
Roeder ER, Hutchison HT. Risk factors for
tration. Acta Paediatr. 2009;98(10): school-aged children with a history of acute
perinatal arterial stroke: a study of 60
1556 –1567 mother-child pairs. Pediatr Neurol. 2007; neonatal stroke [abstr]. Stroke. 2007;38(2):
16. Ramaswamy V, Miller SP, Barkovich AJ, Par- 37(2):99 –107 581
tridge JC, Ferriero DM. Perinatal stroke in 23. Kenet G, Lutkhoff LK, Albisetti M, et al. Impact 29. Westmacott R, MacGregor D, Askalan R, de-
term infants with neonatal encephalopathy. of thrombophilia on risk of arterial isch- Veber G. Late emergence of cognitive defi-
Neurology. 2004;62(11):2088 –2091 emic stroke or cerebral sinovenous throm- cits after unilateral neonatal stroke. Stroke.
17. Brenner DJ, Hall EJ. Computed tomography: bosis in neonates and children: a system- 2009;40(6):2012–2019

e1410 KIRTON et al
Downloaded from pediatrics.aappublications.org at University of Edinburgh on June 8, 2012
Symptomatic Neonatal Arterial Ischemic Stroke: The International Pediatric
Stroke Study
Adam Kirton, Jennifer Armstrong-Wells, Taeun Chang, Gabrielle deVeber, Michael J.
Rivkin, Marta Hernandez, Jessica Carpenter, Jerome Y. Yager, John K. Lynch, Donna
M. Ferriero and for the International Pediatric Stroke Study Investigators
Pediatrics 2011;128;e1402; originally published online November 28, 2011;
DOI: 10.1542/peds.2011-1148
Updated Information & including high resolution figures, can be found at:
Services http://pediatrics.aappublications.org/content/128/6/e1402.full.
html
References This article cites 38 articles, 2 of which can be accessed free
at:
http://pediatrics.aappublications.org/content/128/6/e1402.full.
html#ref-list-1
Subspecialty Collections This article, along with others on similar topics, appears in
the following collection(s):
Premature & Newborn
http://pediatrics.aappublications.org/cgi/collection/premature
_and_newborn
Permissions & Licensing Information about reproducing this article in parts (figures,
tables) or in its entirety can be found online at:
http://pediatrics.aappublications.org/site/misc/Permissions.xht
ml
Reprints Information about ordering reprints can be found online:
http://pediatrics.aappublications.org/site/misc/reprints.xhtml

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly


publication, it has been published continuously since 1948. PEDIATRICS is owned, published,
and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk
Grove Village, Illinois, 60007. Copyright © 2011 by the American Academy of Pediatrics. All
rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.

Downloaded from pediatrics.aappublications.org at University of Edinburgh on June 8, 2012

You might also like