The prognostic value of quantitative diffusion-weighted MRI after pediatric cardiopulmonary arrest

M Yacoub; B Birchansky; M Mlynash; M Berg; L Knight; K G Hirsch; F Su; Revive Initiative at Stanford Children’s Health

doi:10.1016/j.resuscitation.2018.11.003

The prognostic value of quantitative diffusion-weighted MRI after pediatric cardiopulmonary arrest

Resuscitation. 2019 Feb:135:103-109. doi: 10.1016/j.resuscitation.2018.11.003. Epub 2018 Dec 18.

Authors

M Yacoub¹, B Birchansky², M Mlynash³, M Berg⁴, L Knight⁴, K G Hirsch³, F Su⁴; Revive Initiative at Stanford Children’s Health

Affiliations

¹ Stanford University, Pediatric Critical Care Medicine, 770 Welch Road, Suite 435, Palo Alto, 94304, United States. Electronic address: myacoub@stanford.edu.
² Stanford University, 770 Welch Road, Suite 435, Palo Alto, 94304, United States.
³ Stanford Stroke Center, Department of Neurology and Neurological Sciences, 701 Welch Road, Suite 325, Palo Alto, 94304, United States.
⁴ Stanford University, Pediatric Critical Care Medicine, 770 Welch Road, Suite 435, Palo Alto, 94304, United States.

PMID: 30576784
DOI: 10.1016/j.resuscitation.2018.11.003

Abstract

Objectives: The prognostic value of quantitative diffusion-weighted magnetic resonance imaging (DWI MRI) in predicting neurologic outcomes after pediatric cardiopulmonary arrest (CPA) has not been determined. The aim of this study was to identify a DWI MRI threshold for brain volume percent that correlates with neurologic outcome in children who remain comatose or display significant neurologic deficits immediately after resuscitation from CPA.

Methods: This single-center retrospective study analyzed DWI MRIs of pediatric patients who remained neurologically impaired after CPA. Any MRI obtained within 2 weeks after CPA was analyzed. The apparent diffusion coefficient (ADC) value of each voxel within the brain was determined. Percentage brain volume with voxels below each ADC threshold between 300 and 1200 × 10^-6 mm²/s with a step of 50 were calculated. Area under the receiver operating characteristics curve (AUC) was used to identify optimal DWI MRI thresholds for brain volume percent most predictive of poor neurologic outcome. The primary outcome measure was neurologic outcome 6-months after CPA based on Pediatric Cerebral Performance Category (PCPC) score. Poor neurologic outcome was defined as PCPC score of 3-6, or a worsening from baseline score ≥1 if baseline PCPC score was ≥3.

Results: Twenty-six patients were included in this study. The median age was 8.5 years (2.2-14) and median time from CPA to MRI was 4 days (2-7). Two ADC thresholds for brain volume percent had the largest AUC for predicting poor neurologic outcome. An ADC threshold of <600 × 10^-6 mm²/s in ≥7% of brain volume; and <650 × 10^-6 mm²/s in ≥11% of brain volume both demonstrated a specificity of 1.0 (0.76-1.0, 95% CI) and a sensitivity of 0.8 (0.44-0.96, 95% CI) for poor outcome.

Conclusions: In pediatric patients who remain comatose or have significant neurologic deficits after CPA, quantitative DWI MRI correlates with neurologic outcome. Both an ADC threshold of <600 × 10^-6 mm²/s in ≥7% of brain volume and <650 × 10^-6 mm²/s in ≥11% of brain volume are highly specific for predicting poor neurologic outcome. A prospective trial to validate these thresholds is needed.

Keywords: DWI; Neurologic outcome; Pediatric cardiopulmonary arrest; Prognosis; Quantitative brain MRI.

Published by Elsevier B.V.

MeSH terms

Brain / diagnostic imaging
Brain / pathology
Child
Coma* / diagnosis
Coma* / etiology
Correlation of Data
Diffusion Magnetic Resonance Imaging / methods*
Female
Heart Arrest* / complications
Heart Arrest* / therapy
Humans
Male
Nervous System Diseases* / diagnosis
Nervous System Diseases* / etiology
Neurologic Examination
Organ Size
Outcome Assessment, Health Care / methods
Predictive Value of Tests
Prognosis
Resuscitation* / adverse effects
Resuscitation* / methods
Retrospective Studies
United States