Routine Term Cerebral Mri in Predicting Neurodevelopmental Outcomes of Preterm Infants

Abstract

Background: Several cerebral MRI markers have been recently reported as having significant correlations with developmental outcomes in preterm infants. It is currently unclear whether the routine characterization of white matter abnormalities in premature infants at term by MRI leads to improved predictions regarding their prognosis. Objective: To systematically review the evidence in the literature about the clinical usefulness of routine term brain MRI in predicting the neurological outcomes of preterm infants. Design/Methods: Pubmed search in the Clinical Query section using the words brain MRI preterm infants and prognosis. The search yielded 154 articles. Inclusion criteria were: cohort studies in preterm infants gestational age (GA) < 34 weeks and otherwise unselected; cerebral MRIs performed at or near term; standardized neurodevelopmental assessments at 18 months corrected GA and beyond. Results: Of the 154 articles, 13 articles met our inclusion criteria. Positive correlations with abnormal development were found with a wide variety of cerebral MRI markers notably parenchymal lesions and cysts, ventriculomegaly, T1 hyperintensity signals, white matter injury (WMI) scores, apparent diffusion coefficient (ADC), diffuse excessive high signal intensity (DEHSI). The wide variety in the types and locations of markers studied, the timing and type of test used to assess outcomes and the statistical analyses employed for correlations hindered comparison between studies. Seven articles provided sufficient data to permit positive and negative predictive values (PPV and NPV) calculations for abnormalities found on MRI. PPV ranged between 10%-43% for moderate/severe developmental delay, and up to 60% (n=6 with CP) to 82% for CP. The NPV ranged between 74%-100%. One study did not find any advantage of term cerebral MRI over baseline neurological examination. Conclusions: Many abnormalities which have been described have a PPV for an abnormal outcome of less than 50%, casting doubt on their usefulness for families or for orienting follow up or intervention programs. The higher NPV might be useful to reassure parents (and perhaps delay follow-up appointment?) but once again, without certainty. The relatively higher PPV for CP may be of value but is based on very small numbers of affected babies. Cerebral MRI at term age is an interesting tool for research on neurodevelopmental consequences of prematurity. However, for routine clinical practice its usefulness is unproven. We found no research indicating this test helped neither parents nor preterm infants at the present moment, this should be indicated at the moment of the exam, which is not always benign in fragile babies prone to apneas with sedation.