MRI-derived three-dimensional modelling reveals cervicothoracic subarachnoid space narrowing in syringomyelia-affected cavalier king charles spaniels

Syringomyelia (SM) is a disorder of cerebrospinal fluid (CSF) circulation strongly associated with Chiari-like malformation (CM) in Cavalier King Charles Spaniels (CKCS). Although CM is nearly ubiquitous in the breed, not all affected dogs develop signs of CM-related pain (CM-P) or Syringomyelia (SM), raising questions about underlying mechanisms. CM-P without SM has been linked to brachycephaly, whereas SM involves additional craniocervical conformational changes. This study tested whether CM-P affected CKCS with and without SM were more likely to have reduced cisterna magna volume and cervicothoracic subarachnoid space (SAS) narrowing compared to each other and with CKCS with CM but no clinical signs (CM-N). A retrospective dataset of 168 CKCS was reviewed. After applying exclusion criteria and matching for body size, three groups were analysed: CM-N, CM-P without SM, and CM-P with SM (n = 10 per group). Three-dimensional SAS models from the foramen magnum to T3/T4 were generated by segmenting MRI data in 3D Slicer and lofting contours in ANSYS SpaceClaim. Whole SAS volume was lower in SM cases than in both comparison groups (ANOVA p = 0.0033, η² = 0.344). Cisterna magna volume and surface area did not differ, but the volume-to-surface area ratio was reduced in SM affected dogs (p = 0.011, η² = 0.274), consistent with a relatively constricted SAS. Slice-by-slice analysis revealed focal reductions in annular area and hydraulic diameter at the occiput, C2–C6, and T1–T3, with hydraulic diameter more sensitive to local change. Corresponding decreases in Womersley number indicated a viscous-dominated CSF flow regime and loss of pulsatile compliance. Principal Component Analysis confirmed distinct clustering of SM cases, while CM-P and CM-N dogs overlapped; displacement along the principal axis correlated with syrinx diameter, linking morphology to clinical severity. CKCS with CM-P and SM have global SAS volume loss with focal constrictions, whereas CKCS with CM-P without SM does not differ from CM-N. These findings support a model in which SM develops from limited cranial and spinal SAS capacity to accommodate and redistribute pulsatile CSF flow, altering pressure transmission along the neuraxis. The 3D reconstruction method offers a reproducible framework for quantifying these changes and linking structure to flow dynamics. Limitations include incomplete neuroaxis coverage, post-SM analysis (cause versus effect), small sample size, and single-operator measurements.