Clinical History: A 47 year-old man presents with chronic right hip and leg pain. Reflexes are normal. No weakness is found on physical exam. The patient denies bowel or bladder dysfunction. (1a) T2-weighted sagittal and (1b) axial images of the lumbar region are provided. What are the findings? What is your diagnosis?
1a
1b
Figure 1:
(1a) T2-weighted sagittal and (1b) axial images of the lumbar region
Findings
2a
Figure 2:
(2a) The T2-weighted midline sagittal image demonstrates an intramedullary slit-like focus of CSF intensity signal within the distal cord near the conus medullaris (arrow). Grade I spondylolisthesis (short arrow) is present at L5-S1. Disc bulging (arrowhead) without cord compression is also found at T12-L1.
3a
Figure 3:
(3a) The T2-weighted axial image at the most caudal aspect of the lesion demonstrates its location at the central canal (arrow) and confirms its intramedullary location.
Diagnosis
Ventriculus Terminalis (terminal ventricle of the cord).
Introduction
The Ventriculus Terminalis (terminal ventricle) is a developmental variant of the distal cord characterized by an ependymal lined cavity at or near the conus medullaris. Normally the central canal is widest at the conus medullaris, but if prominent or enlarged, it must be differentiated from a cystic neoplasm or syringohydromyelia.
Kernohan in 1924 provided post-mortem light microscopic details of this finding in the developing cord of fetuses, children and adults.1 The terminal ventricle is observable in all fetuses and children much more than in adults, but increases in frequency again in the elderly. In 1995, Coleman reported that MRI revealed a terminal ventricle in 11 of 418 (2.6%) children less than 5 years old.2 Kriss found 5 cases in neonates, all of whom underwent ultrasonography for cutaneous stigmata over two years at a large children’s hospital, and felt it was a normal variant.3
Embryology
The spinal cord develops in two stages. It begins with neurulation manifested as closure of the neural tube, which forms the majority of the cord and is complete by about 27 days of gestation. At about 4.5 weeks, caudally near the tail fold (perhaps as low as S2), neural epithelium and notochord fuse to create a caudal cell mass composed of undifferentiated cells. This cell mass then undergoes vacuolization. The resultant microcysts coalesce and canalize into an ependymal lined tube and unite with the rostral neural tube. Beginning at about 38 days of gestation, retrogressive differentiation occurs via apoptosis (programmed cell death), completing the formation of the conus medullaris, filum terminale and terminal ventricle. Through this process, a major portion of the distal cord involutes. The process is not uniform and microscopic accessory canals and lumina are commonly found in embryos in the conus region, which can lead to a terminal ventricle that is eccentrically located within the central canal.4
4a
Figure 4:
Illustration of canalization and retrogressive differentiation. (1a) Fusion of neural epithelium with the notochord at the caudal end of the embryo results in the caudal cell mass. (1b) By 30 days gestation, multiple cysts are present within the caudal cell mass. (1c) Around 38 days, retrogressive differentiation causes a decrease in size of the caudal mass, and formation of a central lumen. (1d) Ultimately, this lower segment forms the filum terminale, distal conus, and ventriculus terminalis.
Additional MR Examples
5a
Figure 5:
(5a) A T2-weighted sagittal image near the thoracolumbar junction demonstrates a small cystic area (arrow) in the central aspect of the cord in this elderly woman with non-specific back pain and scoliosis.
6a
Figure 6:
(6a) The T2-weighted axial image confirms the finding. Note that the lesion in question is located at the conus medullaris, is non-expansile, and is of short length, typical for a terminal ventricle.
7a
Figure 7:
(7a) A T2-weighted sagittal MR of the thoracic spine demonstrates a small intramedullary conus lesion (arrow) in this 40 year-old male who was scanned after an acute twisting injury led to back pain and left radicular symptoms. The lesion appears eccentric within the cord, which can occur with a terminal ventricle.
8a
Figure 8:
(8a) The T2-weighted axial image confirms the central CSF-signal intensity lesion, lying slightly dorsal to midline, yet still compatible with a terminal ventricle.
9a
Figure 9:
(9a) After gadolinium contrast is given, a fat-suppressed T1-weighted sagittal view reveals no enhancement within the lesion (arrow).
Discussion
The terminal ventricle is usually an incidental finding of little clinical significance. On rare occasions, however, cystic dilatation of the terminal ventricle can occur, resulting in clinical symptoms. Less than 30 such cases have been reported in the literature.5-8 The reason for this enlargement is not known. Trauma and ischemia have been suggested, altering CSF flow within the central canal or leading to disturbance of the Reissner Fiber (a subependymal secretory product) which has a role in CSF regulation.6
Coleman indicated that the average size of the terminal ventricle was 22 x 4.1 x 4.2 mm.2 Reported cases of cystic dilatation of the terminal ventricle have been considerably larger. Signs and symptoms in those affected vary substantially, prompting De Moura Batista to propose classifying the patients into different groups depending on clinical presentation. Those patients’s with bowel or bladder dysfunction, lower extremity weakness or disturbance of reflexes should be more strongly considered for surgical intervention, which might include cyst fenestration or cyst shunting.7 Brisman reported a case of a 57 year-old woman who presented with acute cauda equina syndrome which substantially responded to emergency surgical decompression and cyst drainage.5
Knowing that a small cystic area representing the terminal ventricle can occur as a normal variation can be comforting if the patient has incidental symptoms. However, care must be taken to follow the patient clinically or by imaging at periodic intervals. Only about 2.5% of syrinxes are restricted to the distal cord.2 Therefore, if a small cystic lesion of the conus medullaris is observed it is unknown if it represents a very small localized syrinx or if it is a terminal ventricle that may enlarge over time. Differential diagnoses include cystic neoplasm, syringohydromyelia, cord infarct and artifact, the latter possibly due to pulsation from cord motion or truncation at dark/bright signal interfaces.
MRI in the Differential Diagnosis
10a
Figure 10:
(10a) Epidermoid tumor. A T2-weighted sagittal image in a 37 year old woman presenting with chronic low back and bilateral leg pain. A small intradural CSF intensity lesion (arrow) which did not enhance (on other images) and did not change over 1 year is present.
11a
Figure 11:
(11a) A T1-weighted sagittal image demonstrates CSF-like signal within the lesion (arrow).
12a
Figure 12:
(12a) The lesion appears extramedullary/intradural on the T2-weighted axial view. The location of the lesion at the distal conus and its fluid signal appearance are typical for epidermoid tumor.
13a
Figure 13:
(13a) Hemangioblastoma. A T2-weighted sagittal image in a 62 year-old female presenting with left radicular pain demonstrates a small cystic, mildly expansile lesion (arrow) in the distal cord, above the conus medullaris.
14a
Figure 14:
(14a) The post-contrast fat-suppressed T1-weighted sagittal sequence demonstrates enhancement of a mural nodule (arrow) typical for hemangioblastoma.
15a
Figure 15:
(15a) Caudal aspect of syrinx. A 36 year-old female presented with low back pain radiating into both legs. The T2-weighted sagittal image reveals subtle increased signal within the distal cord (arrow), well above the conus medullaris. The finding mandates additional imaging to further evaluate the superior extent of the lesion.
16a
Figure 16:
(16a) The additional T2-weighted sagittal image of the thoracic region reveals the caudal extent of the lesion (arrow). The abnormality appears isointense to CSF and did not enhance (not shown), compatible with a thoracic syrinx. The lesion is too long and too cranially located to be considered a terminal ventricle.
17a
Figure 17:
(17a) Cord infarction. A T2-weighted sagittal image is provided in a 73 year old woman who developed acute paraplegia after an aortofemoral angioplasty. A poorly circumscribed lesion (arrow) is seen in the conus medullaris, without expansion.
18a
Figure 18:
(18a) A post-contrast fat-suppressed T1-weighted sagittal image demonstrates enhancement (arrows) within the cauda equina.
19a
Figure 19:
(19a) Smoothly marginated nerve root enhancement is confirmed on the corresponding axial view. Such enhancement may persist for months after cord infarction.
20a
Figure 20:
(20a) Terminal ventricle versus small syrinx. A T2-weighted sagittal image in a 22 year old male with 4 years of low back pain. A small, slightly expansile intramedullary lesion (arrow) is apparent at the conus medullaris. The lesion does not enhance (not shown).
21a
Figure 21:
(21a) T2-weighted axial through the region of interest. Is the lesion really comprised of 2 intramedullary cysts, suggesting a syrinx? The location is as expected for a terminal ventricle, which may be eccentrically located within the cord. Without abnormal reflexes, bowel or bladder dysfunction, or muscle weakness, the follow up is the same for either entity, and indicated to exclude further cystic dilatation of the terminal ventricle or enlargement of a small distal syrinx.
22a
Figure 22:
(22a) Cystic Dilatation of Terminal Ventricle. 41year old female had been followed for years for a "cyst" within the distal cord. Symptoms were back pain only. No myelopathic signs. STIR Sagittal reveals well circumscribed cystic dilation of the distal cord. No enhancement after gadolinium contrast (not shown).
23a
23b
Figure 23:
(23a,23b) Cystic Dilatation of Terminal Ventricle: 49 year old female underwent MR for radiating left hip pain and left foot paresthesia. Orthogonal T2w images demonstrate intramedullary cyst just cranial to the tip of the conus. No enhancement after contrast was given (not shown).
Conclusion
The ventriculus terminalis, or terminal ventricle, is an ependymal lined space within the conus that usually does not persist into adulthood. When present in adults, the terminal ventricle is usually an incidental finding unrelated to current symptoms. Rarely, the terminal ventricle may dilate and cause significant clinical symptoms. MR is useful in the detection and diagnosis of the terminal ventricle, and can be used to differentiate this variant from other lesions of the cord or conus. In cases in which a lesion is indeterminate in appearance, a lack of contrast enhancement indicates a non-aggressive abnormality. Terminal ventricle may be indistinguishable from a distal syrinx, though either lesion can simply be followed based on clinical symptoms, in order to assess for possible expansion.
References
1 Kernohan JW. The Ventriculus Terminalis: Its Growth and Development. J Comp Neurol 1924; 38:107-125
2 Coleman LT, et al. Ventriculus Terminalis of the Conus medullaris: MR Findings in Children. AJNR 1995; 16:1421-1426
3 Kriss VM, et al. The Ventriculus Terminalis of the Spinal Cord in the Neonate: A Normal Variant on Sonography. AJR 1995; 165: 1491-1493
4 Barkovich AJ. Congenital Anomalies of the Spine. In: Pediatric Neuroimaging, 4th ed. Lippincott Williams and Wilkins, 2005
5 Brisman JL, et al. Cystic Dilation of the Conus Ventriculus Terminalis Presenting as an Acute Cauda Equina Syndrome Relieved by Decompression and Cyst Drainage. Neurosurgery 2006; 58:585-586
6 Ciappetta P, et al. Cystic Dilation of the Ventriculus Terminalis in Adults. J Neurosurg Spine 2008; 8: 92-99.
7 De Moura Batista L, et al. Cystic Lesion of the Ventriculus Terminalis: Proposal for a New Clinical Classification. J Neurosurg Spine 2008; 8: 163-168
8 Dullerud R, et al. MR Imaging of Two Operated Patients and Review of the Literature. Acta Radiologica 2003; 44:444-446