Morphometric features of posterior cranial fossa are different between Chiari I malformation with and without syringomyelia

Abstract

To investigate whether the posterior cranial fossa (PCF) morphology in Chiari I malformation without syringomyelia (also calledsyrinx) (CMI-only) is different from that in Chiari I malformation with syrinx (CMI-S). Nineteen CMI patients without syrinx constituted the CMI-only group, whereas 48 CMI patients with syrinx were assigned to the CMI-S group. Another cohort of 40 age-matched asymptomatic adolescents was enrolled to serve as the control group. Six measurements were evaluated and compared between these three groups from T1-weighted magnetic resonance(MR)imaging, including the length of the clivus (AB), the anteroposterior diameter of the foramen magnum (BC), the length of the supraocciput (CD), the anteroposterior diameter of the posterior fossa (DA), the posterior fossa height (BE) and the clivus gradient ([Formula: see text]). The posterior cranial fossa morphology in relation to syrinx severity was also investigated. Compared to the normal controls, the AB, CD, DA, BE and [Formula: see text] were significantly larger in the CMI-S group. Similar changes in AB, CD, DA and BE were also demonstrated in the CMI-only group, while the clivus gradient ([Formula: see text]) was found to be normal when compared with the control group. A significantly decreased clivus gradient was observed in the CMI-S group as compared to CMI-only group. In addition, the clivus was significantly flattened in patients with a distended-syrinx in comparison to those with a non-distended syrinx. Small size of the posterior fossa was detected both in CMI cases with and without syrinx. The clivus gradient served as the only morphologic difference in the PCF between CMI-S and CMI-only patients and was correlated with the severity of the syrinx, may support the theory that the restricted circulation of cerebrospinal fluid at the anterior paramedial subarachnoid space contributes to the formation of a syrinx.