[Cerebral infarction in full-term newborns: MR imaging features].

PURPOSE

Perinatal stroke is the most frequent cause of congenital hemiplegia. The MR imaging features of cerebral infarction in full-term newborns will be reviewed and the underlying etiologies discussed.

PATIENTS AND METHODS

Retrospective study (1999-2005) including 15 newborn infants without history of fetal distress with early seizures (before day 7) with ischemic or hemorrhagic infarct on MR. All MR examinations were reviewed by consensus (MBD, CHA) using a checklist and results were correlated with findings at clinical follow-up (mean follow-up of 1 year).

RESULTS

MR showed ischemic strokes in 10 infants (5 left, 4 right, 1 bilateral) in the middle (n=9) or anterior (n=1) cerebral artery distribution or with extensive bilateral distribution. The cortex and subcortical regions were involved in all cases with ipsilateral basal ganglia/internal capsule extension in 4 cases. The imaging features were related to the timing of the MR examination: within 5 days (n=4): cortical effacement on T1W and T2W images, increased signal on DWI with reduced signal on ADC; between 7-14 days (n=6): T1W hyperintense and T2W hypointense cortex. Five infants had unifocal (3/5) or multifocal (2/5) hemorrhagic infacrtion. Of 13 patients followed-up from 3 months to 3 years (1 lost to follow-up, 1 deceased), 11 had no clinical deficits, 1 had hemiparesis, and 1 had asymmetrical muscle tone. The latter 2 infants had involvement of the posterior limb internal capsule, with basal ganglia and hemispheric involvement. In the literature, this association of lesions is considered predictive of hemiplegia and a high rate of sequelae is reported following neonatal stroke: 50-75% with motor deficit and/or seizure disorder. The 4 infacts with hemorrhagic infarction followed-up for 14 months to 5 years had a globally favorable neurological outcome.

CONCLUSION

Cerebral infacrtion in full-term newborns without associated risk factor is variable and should be recognized. Early MR, before day 2, with diffusion-weighted sequences, allows diagnosis but follow-up MR after day 7 is necessary to better depict the extent of permanent lesions. The presence of hemorrhage is not a predictive factor of poor neurological outcome.