Syringomyelia can be diagnosed in isolation but is more commonly found in the presence of craniocervical junction anomalies or spinal dysraphism. The origin of syringomyelia has been hypothesized to be either congenital or acquired. The purpose of this study was to determine the incidence of syringomyelia within the fetal and postnatal population with neural tube defects (NTDs).
A review was performed of the authors' fetal MRI database of pregnancies with imaging between March 2004 and November 2011 for evaluation of an intrauterine anomaly detected via prenatal ultrasonography. Those cases with an NTD were then selected and a chart review was performed of all prenatal and postnatal imaging as well as available clinical history.
A total of 2362 fetal MRI examinations were performed, and 109 of these were patients with an NTD. Of the 2362 studies reviewed, 2 cases of fetal syringomyelia were identified. Both fetal syrinxes were identified in fetuses with CSF flow disturbances (1 case each of encephalocele and myelomeningocele). Both fetal MRI examinations were performed late in gestation, at 31 and 38 weeks, respectively. The patient with an encephalocele was excluded from the spinal NTD population; therefore a syrinx was identified in 0.08% (2/2362) of the entire population of fetuses who underwent MRI, or 0.9% (1/109) of fetuses with a spinal NTD. Sixty-three of the 109 patients with an NTD had postnatal clinical data available for review. Twenty-nine (46%) of 63 had a syrinx identified during the follow-up period. Of this group, 50 patients had an open NTD and 27 (54%) of 50 developed a syrinx. Among the patients with an open NTD who developed a syrinx, only 7% did not have or develop hydrocephalus, compared with 35% of the patients who did not develop a syrinx (p < 0.05). There were nonsignificantly more frequent shunt revisions among those patients who developed a syrinx, and a syrinx developed in all patients who required surgical Chiari malformation decompression or tethered cord release. The initial identification of a spinal cord syrinx varied greatly between patients, ranging from 38 weeks gestation to greater than 4 years of age.
These data suggest that syringomyelia is not a congenital embryonic condition. A syrinx was not identified in fetuses who underwent imaging for other intrauterine anomalies. In the population of patients with NTDs who are known to be at high risk for developing syringomyelia, the pathology was only identified in 2 third-trimester fetuses or postnatally, typically in the presence of hydrocephalus, shunt placement, Chiari malformation decompression, or tethered cord release. The study supports the authors' hypothesis that a syrinx is an acquired lesion, most likely due to the effects of abnormal CSF flow.