Radiological Signs in Spinal Cord Infarct (SCI) and Association with Anatomic Location.
Ahya Ali1, Trevor Glenn2, Philippe-Antoine Bilodeau3, Shamik Bhattacharyya4
1Westchester Medical Center, 2Mass General Brigham, 3Massachusetts General Hospital, 4Brigham and Women's Hospital
Objective:
To describe radiological signs of SCI on MRI and assess association with location, cause, and severity.
Background:
MRI lesion characteristics increase specificity of etiologic diagnoses for spinal cord lesions.
Design/Methods:
A single tertiary health system retrospective review of patients with definite, probable and possible SCI. Spinal cord MRIs were reviewed by three investigators (AA, TG, SB). Numerical or ordinal data were analyzed with Mann-Whitney tests, while binary data were analyzed with Fisher exact test. Univariate analysis identified associations between radiological signs and location, cause, and severity of SCI. Statistical significance was p<0.05.
Results:
32 patients were included; 56% male with mean age 60.6 years (SD 18.0). SCI was classified as 40.6% definite periprocedural infarcts, 40.6% definite spontaneous, 6.3% probable spontaneous, 12.5% possible spontaneous. Median time from symptom onset to MRI was 1 day (range 0-27). Median number of vertebral segments affected was 4 (range 1-12); 68.8% longitudinally extensive (≥3 segments), 6.3% with non-contiguous lesion. Of those with available images, 87% had diffusion restriction, 9.5% had T1 weighted post-contrast enhancement. On T2 weighted imaging, patterns seen were hologrey (n=11), owl eye (n=8), holocord (n=7), anteromedial spot sign (n=3), H sign (n=3), bagel sign (n=2), anterior pencil like hyperintensity (n=2), bright spotty sign (n=1), bilateral posterior segment (n=1). More than one radiological sign seen in 34% of patients. The spinal segments affected were thoracic-to-conus (n=13), cervical (n=8), thoracic (n=6), cervical-to-thoracic (n=4), and conus (n=1).
Cervical cord location was associated with spontaneous SCI (p=0.04). Thoracic cord location was associated with Definite periprocedural SCI (p=0.02). No significant correlations between T2 imaging pattern and location or severity of infarct.
Conclusions:
Radiological signs in SCI are often overlapping and include those seen with inflammatory etiologies. Cervical lesions correlate with spontaneous SCI. Large scale studies regarding radiological signs and their association with spinal cord level may add to etiology specific diagnosis.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.