190 - Accuracy of Early Cerebral Palsy Diagnosis with Structural MRI, General Movements Assessment (GMA) and Hammersmith Infant Neurological Exam (HINE) in Infants Born Preterm

Friday, April 28, 2023

5:15 PM – 7:15 PM ET

Poster Number: 190
Publication Number: 190.142

Nehal A. Parikh, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, United States; Karen Harpster, Cincinnati Childrens, Cincinnati, OH, United States; Stephanie L. Merhar, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, United States; Beth M. Kline-Fath, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, United States

Presenting Author(s)

Nehal A. Parikh, DO, MS (he/him/his)

Professor of Pediatrics
Cincinnati Children's Hospital Medical Center
Cincinnati, Ohio, United States

Background: An expert consensus panel recommended the use of structural MRI (sMRI) at term corrected (CA) age in combination with GMA or HINE before 5 months CA to enable early diagnosis of cerebral palsy (CP) in high-risk newborns (Novak I, 2017). However, the prognostic accuracy of this combination approach is lacking.

Objective: To assess the accuracy of sMRI, GMA, and HINE individually and in combination to predict different severity levels of CP diagnosed at 2 years CA.

Design/Methods: We assembled a multicenter regional cohort of 395 preterm infants born at or below 32 weeks’ gestation from 5 level III/IV Southwest Ohio NICUs from September 2016 to November 2019. All infants were imaged at term CA using sMRI on the same 3T Philips scanner and scans were read by a neuroradiologist masked to clinical and outcome data. Moderate-severe abnormality on sMRI was determined per Novak et al. (2017). An occupational therapist who was masked to clinical data and trained to reliability administered the GMA and HINE tests at 3 months CA. At 22-26 CA, CP diagnosis and severity were determined using the Amiel-Tison standardized neurological exam and Gross Motor Function Classification System (GMFCS). Early CP diagnosis was made as per Novak 2017: abnormal sMRI and either abnormal GMA (no fidgety movements) or HINE score < 56. We used sensitivity, specificity, likelihood ratios, to evaluate the accuracy of these three tests individually and in combination to predict CP at age 2.

Results:

Of 395 infants, 2 withdrew, 1 died, 381 (97%) returned for testing at 3 months CA and 338 (86%) returned at 22-26 months CA for developmental follow-up. Of these 338 preterm infants, 39 (11.5%) were diagnosed with any CP (GMFCS ≥1) and 11 (3.3%) with moderate-severe CP (GMFCS ≥2) at age 2. Infants that returned at 2 years were similar to those that did not except for gestational age and retinopathy of prematurity (Table 1). Infants diagnosed with moderate-severe CP at age 2 had significantly higher odds of abnormal sMRI, GMA, and/or HINE (Table 2). Each test was accurate individually in predicting moderate-severe CP but failed in predicting mild CP due to poor sensitivity (Table 3). Combining sMRI with GMA or with HINE did not improve prediction accuracy.

Conclusion(s):

Early prediction/diagnosis of moderate-severe CP in preterm infants was most accurate with GMA at 3 months CA. The addition of sMRI at term CA did not meaningfully improve prognostic accuracy and thus has implications for current consensus recommendations. Prediction of mild CP was poor for each test individually or in combination.