Fetal gastrointestinal MRI: all that glitters in T1 is not necessarily colon.

BACKGROUND: It has been described that both the colon and distal ileum present with a physiological hypersignal on T1-weighted sequences during the second and third trimesters of pregnancy because of their protein-rich meconium content, it was unclear whether the normal characteristics that have been described on fetal MRI can be applied to gastrointestinal (GI) obstructions. OBJECTIVE: To analyse the localisation value of T1 hypersignal within dilated bowel loops in fetuses with gastrointestinal tract obstruction. MATERIALS AND METHODS: A retrospective 4-year multicentre study analysing cases of fetal GI obstruction in which MRI demonstrated T1 hypersignal content in the dilated loops. Data collected included gestational age (GA) at diagnosis, bowel appearance on US, CFTR gene mutations and amniotic levels of gastrointestinal enzymes. The suggested prenatal diagnosis was eventually compared to postnatal imaging and surgery. RESULTS: Eleven patients were included. The median GA at US diagnosis was 23 weeks (range 13-32). In eight cases there was a single dilated loop, while several segments were affected in three. The median GA at MRI was 29 weeks (range 23-35). One case presented with cystic fibrosis mutations. Final prenatally suspected diagnoses were distal ileal atresia or colon in nine cases and proximal atresia in two. Postnatal findings were proximal jejunal atresia in nine cases and meconium ileus in two. In five cases the surgical findings demonstrated short bowel syndrome. CONCLUSION: In cases of fetal occlusion, T1 hypersignal should not be considered as a sign of distal ileal or colonic occlusion. The obstruction may be proximal, implying a risk of small bowel syndrome, which requires adequate parental counselling.

Rhombencephalosynapsis and related anomalies: a neuropathological study of 40 fetal cases.

Rhombencephalosynapsis is an uncommon cerebellar malformation defined by vermian agenesis with fusion of the hemispheres and of the dentate nuclei. Embryologic and genetic mechanisms are still unknown, and to date, no animal models are available. Ultrasound diagnosis is generally suspected after 22 weeks of gestation, and usually the abnormality is suggested by ventriculomegaly. Morphological analysis of 40 fetuses after medical termination of pregnancy allowed us to confirm that rhombencephalosynapsis was always associated with other brain abnormalities or malformations: Purkinje cell heterotopias, fusion of colliculi, forking and/or atresia of the aqueduct and of the third ventricle resulting in a fusion of the thalami, agenesis of the corpus callosum, lobar holoprosencephaly and neural tube defects. Pons and medulla were very infrequently abnormal. Furthermore, complete autopsy made it possible to separate either pure neurologic phenotypes, or associated with extraneural anomalies from syndromic forms: Gomez-Lopez-Hernandez syndrome (1 case) and VACTERL-H syndrome (6 cases). The number of our fetal cases strongly suggests that VACTERL-H association related with rhombencephalosynapsis emerges as a non-random association. Furthermore, recurrence and consanguinity were noted in two different families, which argue for a sporadic or inherited cause. From our results, it could be suggested that rhombencephalosynapsis may be due to defective genes regulating formation of the roof plate and the development of midline cerebellar primordium at the junction of the mesencephalon and of the first rhombomere.