MRI-based dural sac index improves detection of tethered cord syndrome
Investigators developed and externally validated a retrospective multicenter MRI diagnostic index using dural sac morphology at the L1 to L2 and L2 to L3 levels to identify tethered cord syndrome, including occult cases. The model, which incorporates age and sex adjustments, achieved an area under the curve of 0.930 with sensitivity of 0.921 and specificity of 0.779 in training, and maintained sensitivity above 0.9 and specificity above 0.7 across internal and external validation cohorts. The study reports that this straightforward morphological scoring system reduces reliance on advanced imaging protocols and subjective radiological interpretation. By providing a standardized imaging biomarker, the index offers a practical tool to streamline early diagnosis and surgical planning for neurosurgical and radiology departments.
The original study
Diagnostic utility of dural sac morphology on MRI in tethered cord syndrome: a multicenter study.
- Authors
- Wang H, Wang K, He K, Zhao T, Lau TS, Wang Y, et al.
- Journal
- Journal of neurosurgery. Spine
- Type
- Journal Article
- PMID
- 42497455
Original abstract
OBJECTIVE: Tethered cord syndrome (TCS) results in neurological dysfunction due to spinal cord tethering. Because MRI sensitivity, particularly for occult TCS (OTCS), is limited by variable image quality and anatomy, the authors aimed to develop and externally validate a simple MRI-based diagnostic index derived from dural sac morphology to improve diagnosis across multicenter datasets. METHODS: This retrospective multicenter study included 280 participants from 5 hospitals, encompassing patients with simple TCS, complex TCS, and OTCS along with healthy controls. Participants were allocated into training, internal validation, external test, and external OTCS test datasets. Dural sac morphological features were manually measured on axial T2-weighted MRI scans from levels L1-2 through L5-S1. Logistic regression identified key morphological features at L1-2 and L2-3 levels, which were integrated into a diagnostic index incorporating demographic adjustments for age and sex. Model performance was assessed using receiver operating characteristic curve analysis across all datasets. RESULTS: The final diagnostic model demonstrated strong and consistent diagnostic efficacy. In the training dataset, the corrected diagnostic index achieved a sensitivity of 0.921 and specificity of 0.779 (area under the curve 0.930). High sensitivity (≥ 0.9) and satisfactory specificity (≥ 0.7) were reliably maintained across the internal validation, external test, and OTCS datasets. CONCLUSIONS: This study establishes a practical MRI-based diagnostic index that leverages dural sac morphological parameters at levels L1-2 and L2-3 to accurately diagnose TCS, including challenging occult cases. The model's reliance on straightforward anatomical measurements and demographic corrections eliminates dependence on advanced imaging protocols or subjective interpretations, presenting a robust imaging biomarker that can significantly improve early detection and clinical management of TCS.