Risk Factors and Outcomes for Cerebral Palsy With Hypoxic-Ischemic Brain Injury Patterns Without Documented Neonatal Encephalopathy.

BACKGROUND AND OBJECTIVES: Perinatal hypoxic-ischemic brain injury is a leading cause of term-born cerebral palsy, the most common lifelong physical disability. Diagnosis is commonly made in the neonatal period by the combination of neonatal encephalopathy (NE) and typical neuroimaging findings. However, children without a history of neonatal encephalopathy may present later in childhood with motor disability and neuroimaging findings consistent with perinatal hypoxic-ischemic injury. We sought to determine the prevalence of such presentations using the retrospective viewpoint of a large multiregional cerebral palsy registry. METHODS: Patient cases were extracted from the Canadian Cerebral Palsy Registry with gestational age >36 weeks, an MRI pattern consistent with hypoxic-ischemic injury (HII, acute total, partial prolonged, or combined), and an absence of postnatal cause for HII. Documentation of NE was noted. Maternal-fetal risk factors, labor and delivery, neonatal course, and clinical outcome were extracted. Comparisons were performed using χ2 tests and multivariable logistic regression with multiple imputation. Propensity scores were used to assess for bias. RESULTS: Of the 170 children with MRI findings typical for HII, 140 (82.4%, 95% confidence interval [CI] 75.7%-87.7%) had documented NE and 29 (17.0%, 95% CI 11.7%-23.6%) did not. The group without NE had more abnormalities of amniotic fluid volume (odds ratio [OR] 15.8, 95% CI 1.2-835), had fetal growth restriction (OR 4.7, 95% CI 1.0-19.9), had less resuscitation (OR 0.03, 95% CI 0.007-0.08), had higher 5-minute Apgar scores (OR 2.2, 95% CI 1.6-3.0), were less likely to have neonatal seizures (OR 0.004, 95% CI 0.00009-0.03), and did not receive therapeutic hypothermia. MRI was performed at a median 1.1 months (interquartile range [IQR] 0.67-12.8 months) for those with NE and 12.2 months (IQR 6.6-25.9) for those without (p = 0.011). Patterns of injury on MRI were seen in similar proportions. Hemiplegia was more common in those without documented NE (OR 5.1, 95% CI 1.5-16.1); rates of preserved ambulatory function were similar. DISCUSSION: Approximately one-sixth of term-born children with an eventual diagnosis of cerebral palsy and MRI findings consistent with perinatal hypoxic-ischemic brain injury do not have documented neonatal encephalopathy, which was associated with abnormalities of fetal growth and amniotic fluid volume, and a less complex neonatal course. Long-term outcomes seem comparable with their peers with encephalopathy. The absence of documented neonatal encephalopathy does not exclude perinatal hypoxic-ischemic injury, which may have occurred antenatally and must be carefully evaluated with MRI.

Risk Factors for Term-Born Spastic Diplegic Cerebral Palsy: A Case-Control Study.

BACKGROUND: To identify if a predetermined set of potential risk factors are associated with spastic diplegic cerebral palsy (SDCP) in term-born children. METHODS: This is a case-control study with cases (n = 134) extracted from the Canadian Cerebral Palsy Registry (CCPR) and controls (n = 1950) from the Alberta Pregnancy Outcomes and Nutrition (APrON) study. Our primary variable was the SDCP phenotype in term-born children. Possible risk factors were selected a priori and include extreme maternal age (<19 or >35 years), pregnancy complications, maternal disease, substance use, perinatal infection, mode of delivery, perinatal adversity (i.e., neonatal encephalopathy presumably on the basis of intrapartum hypoxia-ischemia), sex, and birth weight. Multivariable analyses were used to calculate odds ratios (ORs) and 95% confidence intervals (CIs). RESULTS: Multivariable analysis revealed associations between term-born SDCP and pregnancy complications (OR = 4.73; 95% CI = 1.91 to 10.56), maternal disease (OR = 2.52; 95% CI = 1.57 to 3.93), substance use (OR = 3.11; 95% CI = 2.10 to 4.55), perinatal infection (OR = 2.72; 95% CI 1.32 to 5.10), Caesarean section (OR = 2.35; 95% CI = 1.62 to 3.40), and perinatal adversity (OR = 2.91; 95% CI = 1.94 to 4.50). Multiple regression analysis revealed associations between SDCP and pregnancy complications (OR = 3.28; 95% CI 1.20 to 8.15), maternal disease (OR = 2.52; 95% CI 1.50 to 4.12), substance use (OR = 3.59; 95% CI 2.37 to 5.40), perinatal infection (OR = 3.78, 95% CI 1.71 to 7.72), Caesarean section (OR = 2.72; 95% CI 1.82 to 4.03), and perinatal adversity (OR = 4.16; 95% CI 2.67 to 6.70). INTERPRETATION: Antenatal (pregnancy complications, maternal disease, substance use) and perinatal (infections, Caesarean section, and perinatal adversity) risk factors are associated with an increased risk of SDCP in term-born children, suggesting variable interactions between risk factors to provide a clinicopathologic framework that is different from SDCP observed in preterm-born children.

Association of Gestational Age at Birth and Changes on MRI With Prevalence and Spectrum of Comorbidities in Children With Cerebral Palsy.

BACKGROUND AND OBJECTIVES: For individuals with cerebral palsy (CP) and caregivers, comorbidities may be a greater challenge than neuromotor impairment. Clinicians may make assumptions regarding risk of comorbidities based simply on term vs preterm birth, but this has not been well examined. To better understand factors affecting comorbidity pattern, we investigated the relationship between gestational age (GA) and imaging pattern on the presence of specific comorbidities. METHODS: This is a cross-sectional study of data extracted from the Canadian Cerebral Palsy Registry of children with CP. Multivariable analysis was used to evaluate the relationship between brain injury, GA, and comorbidities. Comorbidities included in the analysis were communication, cognitive, visual, and auditory impairment, seizures in the past year, and gavage feeding. Each comorbidity was assessed as a separate nonexclusive outcome, with GA, MRI pattern, birth weight, postneonatal insult, 5-minute Apgar score, and male sex considered as potential modifiers. RESULTS: The only comorbidity affected by GA on multivariable analysis was seizures within the past year that were more prevalent in term children (odds ratio [OR] 1.1 95% CI 1.0-1.2) and was also affected by Apgar score (OR 0.9 95% CI 0.85-0.94), but not MRI pattern. MRI pattern appeared important for communication impairment (deep gray OR 4.2 95% CI 1.8-10.0; total brain injury OR 8.5, 95% CI 3.2-22.6; malformation OR 2.7, 95% CI 1.3-5.7) and cognitive impairment (deep gray OR 5.6, 95% CI 2.4-13.2; total brain injury OR 10.1, 95% CI 4.0-25.3; malformation OR 3.3, 95% CI 1.6-6.8; watershed OR 3.6, 95% CI 1.4-8.9). Focal injury compared with normal MRI was associated with reduced odds of visual impairment (OR 0.24, 95% CI 0.12-0.48), auditory impairment (OR 0.2195% CI 0.10-0.46) and communication impairment (OR 0.46, 95% CI 0.26-0.82), and overall number of comorbidities (coefficient -0.73, 95% CI -1.2 to -0.31). The number of comorbidities was increased by total brain injury pattern (coefficient 0.65, 95% CI 0.15-1.13) and reduced by focal brain injury (coefficient -0.73, 95% CI -1.2 to -0.31) and increasing 5-minute Apgar score (coefficient -0.11, 95% CI -0.16 to -0.07). DISCUSSION: In those with brain injuries sufficient to cause CP, development of additional comorbidities is less affected by GA at birth and more related to the underlying cause of CP as reflected by MRI patterns.

Risk Factors for Perinatal Arterial Ischemic Stroke: A Machine Learning Approach.

BACKGROUND AND OBJECTIVES: Perinatal arterial ischemic stroke (PAIS) is a focal vascular brain injury presumed to occur between the fetal period and the first 28 days of life. It is the leading cause of hemiparetic cerebral palsy. Multiple maternal, intrapartum, delivery, and fetal factors have been associated with PAIS, but studies are limited by modest sample sizes and complex interactions between factors. Machine learning approaches use large and complex data sets to enable unbiased identification of clinical predictors but have not yet been applied to PAIS. We combined large PAIS data sets and used machine learning methods to identify clinical PAIS factors and compare this data-driven approach with previously described literature-driven clinical prediction models. METHODS: Common data elements from 3 registries with patients with PAIS, the Alberta Perinatal Stroke Project, Canadian Cerebral Palsy Registry, International Pediatric Stroke Study, and a longitudinal cohort of healthy controls (Alberta Pregnancy Outcomes and Nutrition Study), were used to identify potential predictors of PAIS. Inclusion criteria were term birth and idiopathic PAIS (absence of primary causative medical condition). Data including maternal/pregnancy, intrapartum, and neonatal factors were collected between January 2003 and March 2020. Common data elements were entered into a validated random forest machine learning pipeline to identify the highest predictive features and develop a predictive model. Univariable analyses were completed post hoc to assess the relationship between each predictor and outcome. RESULTS: A machine learning model was developed using data from 2,571 neonates, including 527 cases (20%) and 2,044 controls (80%). With a mean of 21 features selected, the random forest machine learning approach predicted the outcome with approximately 86.5% balanced accuracy. Factors that were selected a priori through literature-driven variable selection that were also identified as most important by the machine learning model were maternal age, recreational substance exposure, tobacco exposure, intrapartum maternal fever, and low Apgar score at 5 minutes. Additional variables identified through machine learning included in utero alcohol exposure, infertility, miscarriage, primigravida, meconium, spontaneous vaginal delivery, neonatal head circumference, and 1-minute Apgar score. Overall, the machine learning model performed better (area under the curve [AUC] 0.93) than the literature-driven model (AUC 0.73). DISCUSSION: Machine learning may be an alternative, unbiased method to identify clinical predictors associated with PAIS. Identification of previously suggested and novel clinical factors requires cautious interpretation but supports the multifactorial nature of PAIS pathophysiology. Our results suggest that identification of neonates at risk of PAIS is possible.