RSNA 2014 

Abstract Archives of the RSNA, 2014


PDS244

3T MRI of Cerebellar Hemorrhage in Preterm Infants: Scoring System and Relationship to Neurodevelopmental Outcomes

Scientific Posters

Presented on December 3, 2014
Presented as part of PDS-WEA: Pediatric Wednesday Poster Discussions

Participants

Mai-Lan Ho MD, Presenter: Nothing to Disclose
Dawn Gano MD, Abstract Co-Author: Nothing to Disclose
Olga Tymofiyeva MD, Abstract Co-Author: Nothing to Disclose
Hannah Glass MD, Abstract Co-Author: Nothing to Disclose
Donna Ferriero MD, MS, Abstract Co-Author: Nothing to Disclose
A. James Barkovich MD, Abstract Co-Author: Research Consultant, General Electric Company

PURPOSE

Cerebellar hemorrhage (CH) is a significant imaging finding in preterm infants, and is routinely evaluated by MRI at 3 Tesla in our practice. We will describe a standardized scoring system for CH on 3T MRI, and correlate with supratentorial brain injury and neurodevelopmental outcomes.

METHOD AND MATERIALS

46 preterm infants (< 37 weeks gestational age) admitted to our neonatal ICU between 2011-2013 underwent 3T MRI with 3-D T1, axial T2, and axial susceptibility-weighted sequences. A pediatric neuroradiologist blinded to clinical data reviewed cases for intraventricular hemorrhage (IVH), supratentorial white matter injury (WMI), ventriculomegaly (VM), and cerebellar hemorrhage (CH). CH was graded based on laterality, number, size, location(s), and involved lobe(s). Neurodevelopmental assessment included neuromotor scores at birth and 6 months, and Bayley-III testing at 12 months.

RESULTS

Mean gestational age of subjects was 28.8 weeks (range 25.3 - 32.3 weeks). 11 subjects (24%) had CH, of which 27% were graded as mild, 18% as moderate, and 55% as severe. Of the patients with CH, 55% had IVH, 27% had WMI, and 18% had VM. For the 35 patients without CH, 9% had IVH, 17% had WMI, and 6% had VM. Based on Fisher’s exact test, CH was significantly associated with IVH (p = 0.02), but not WMI (p = 0.4) or VM (p = 0.3). Neurodevelopmental assessment did not identify functional motor deficits at birth or 6 months. Bayley-III motor, language, and cognitive scores at 12 months were 96 ± 26, 102 ± 23, 96 ± 23 for subjects with CH and 104 ± 10, 110 ± 10, 116 ± 11 in those without CH. Using a 2-sided t-test, CH was significantly associated with cognitive (p = .0003) and motor (p = 0.1) subscores, but not language (p = 0.7).

CONCLUSION

We have implemented a standardized scoring system for preterm cerebellar hemorrhage on 3T MRI. In our cohort, CH was significantly associated with IVH, but not WMI or VM. At 1 year of age, CH is significantly associated with cognitive and motor subscores on the Bayley-III. In conjunction with the literature, we conclude that preterm cerebellar injury likely represents a form of germinal matrix hemorrhage, with long-term implications for cognitive and motor function.

CLINICAL RELEVANCE/APPLICATION

3T MRI is the imaging standard for evaluation of cerebellar hemorrhage in preterm infants, and requires a standardized scoring system to assist in grading of overall brain injury and prediction of neurodevelopmental outcomes.

Cite This Abstract

Ho, M, Gano, D, Tymofiyeva, O, Glass, H, Ferriero, D, Barkovich, A, 3T MRI of Cerebellar Hemorrhage in Preterm Infants: Scoring System and Relationship to Neurodevelopmental Outcomes.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14004332.html