On neonatal lumbar spine sonography, a simple cyst at the conofilar junction is frequently observed and termed as a filar cyst [1]. These cystic structures are midline in location, inferior to the conus, thin walled, anechoic, and fusiform on longitudinal and round on transverse plane on ultrasound. The majority of filar cyst regress. Because of a lack of histologic correlation, the exact pathology of filar cyst is not clear. Irani et al. [1] found filar cyst in 12% of 644 spine US of neonates and infants. They consider that filar cyst is developmental in nature because the incidence of filar cyst detection was inversely related to age up to 6 months; there is a lack of filar cyst descriptions in adults. Thus, neonatal filar cysts incidentally found by ultrasound are considered as a normal variant. Seo et al. [2] followed 50 cases of filar cysts detected in neonatal ultrasound with subsequent MRI at age 5–12 months. Twenty filar cysts (40%) showed later regression, while 30 persisted and 2 showed progressive enlargement requiring surgery. It is noteworthy that these latter two cases had concurrent fibrolipoma of the filum and communication of the cyst with the central canal of the more rostral cord was observed during surgery.
During the first 5 years of life, incidental observation of a focal dilation of the central canal in the conus at lumbar ultrasound or MRI is frequent and termed as a ventriculus terminalis (VT). In their retrospective review of 418 spine MRIs (age 5 days to 20 years), Coleman et al. [3] found focal dilation of the central canal in the conus in 11 of 180 (6%) patients younger than 5 years but not among patients older than 5 years. Thus, asymptomatic localized terminal ventricle is considered as a normal developmental phenomenon. In adulthood, the persistence of the terminal ventricle, especially when unaccompanied by other central nervous system anomalies, is extremely rare. Symptomatic adult cases of enlarged terminal ventricle have been reported in surgical literature under the name of terminal ventricular cyst. The main presentations are either acute or chronic back pain, sciatica, bladder dysfunction, and/or lower extremity weakness with female predominance [4].
Are the ventriculus terminalis and filar cyst different entities or ends of the spectrum with the same pathophysiology? We believe the latter and these are representing the embryologic variation of the distal spinal cord (Fig. 5). Both the ventriculus terminalis and filar cyst are transitional structures in the equinal cord (the conus medullaris and filum terminale) during the process of retrogressive differentiation, which may be still in progress in early infancy [3, 5]. The process of the canalization and retrogressive differentiation of the equinal cord is extremely variable [5]. The filar cyst may represent residual of minor cystic loculation remaining in the equinal cord, which regress early in infancy. The ventriculus terminalis is a focal enlargement of the ependymal lined canal, in which the connection to the central canal of the more rostral spinal cord is maintained. If this connection is either narrowed, kinked or compressed by a mechanism such as cord tethering, deranged CSF flow dynamics may result in dilation of the ventriculus terminalis cavity with increased pressure. Central canal dilatation of the conus is well documented among patients with “tethered cord” such as with lipomyelocele, myelomeningocele, and diastematomyelia [6]. Alteration of the CSF dynamics in the central canal by the cord tethering was considered as a cause.
During the canalization of the equinal cord, vacuoles form at many sites and coalesce into the canal with great viability. The ependymal-lined cyst/canals in the conus are variable and may communicate with the central canal or remain as an isolated cavity [5, 7]. Accessory lateral and dorsal canals may extend into the filum and may communicate into the central canal. These variations are commonly seen in human embryos and referred as “forking.” Lendon and Emery evaluated 77 autopsy specimens of infants under 1 year of age and found the presence of the major “forking” in 10% and minor “forking” in 31% [7]. When “forking” of the canal is present in the conus, the ventriculus terminalis may be off the midline and may even mimic extramedullary location on the imaging, like in the case presented here.