Impact of cervical stenosis on multiple sclerosis lesion distribution in the spinal cord

Objective: To determine whether demyelinating lesions attributable to multiple sclerosis (MS) occur more commonly in regions of pre-existing cervical stenosis (CS). Design/Methods: One hundred comorbid MS/CS patients and 100 MS-only controls were identified via ICD codes and radiology reports from a retrospective chart review of the records of the University of Pennsylvania Hospital System (UPHS) from January 1st, 2009 to December 31st, 2018. For each patient, axial and sagittal T2 sequences of cervical MRI scans were examined. The cervical cord was split into 7 equal segments comprising the disc space and half of each adjacent vertebral body. Each segment was assessed for the presence of MS lesions and grade 2 CS or higher by previously published criteria. Lesions which were concerning for spondylotic-related signal change based on imaging characteristics were excluded (n = 6, 3.2%). Clinical data was extracted from the electronic medical record. Results: Average age at the time of MRI was 57.0 +/- 10.5 years and average time with MS diagnosis was 15.3 +/- 9.2 years. The majority of patients had a diagnosis of relapse-remitting MS (81.0%) and the F:M ratio was 3.5. Eighty-five percent of patients were on treatment at the time of MRI, most often glatiramer acetate (35.0%). Spinal segments with at least grade 2 stenosis were significantly associated with the presence of an MS lesion in the same segment (chi(2) = 19.0, p < 0.001, OR = 2.6, 95% CI 1.8-3.7). Conclusions: Our data suggest there is a significant association between segments of spinal cord with at least moderate CS and segments with MS lesions. Further analysis is required to assess if cervical stenosis is a causative or aggravating factor in multiple sclerosis.