Prevalence and Clinical Significance of Incidental Vertebral Marrow Signal Abnormality in Thoracolumbar Spine MRI

Study Design.A cross-sectional study.Objective.This study investigates the prevalence of incidental vertebral marrow signal abnormality (VMSA) in thoracolumbar spine magnetic resonance imaging (MRI) ordered for the evaluation of back and/or leg pain and assess the clinical work-up for VMSAs.Summary of Background Data.Patients presenting with back pain are often referred for spine MRI for diagnostic evaluation. VMSA is most frequently found in the lumbar spine and is of clinical concern because it can represent malignancy. Standardized procedures for reporting and managing VMSAs do not exist.Methods.The radiology database at the Oregon Health & Science University health system was queried to identify patients with thoracolumbar spine MRI scans performed between January 2014 and June 2016. Patients 16 years or older with MRIs ordered by providers at a multidisciplinary spine specialty clinic for the diagnostic evaluation of back and/or leg pain were included. Radiology reports were searched for keywords pertaining to VMSAs, such as "malignancy." Medical records of these patients were further reviewed for the clinical work-up and final diagnoses pertaining to the VMSA.Results.The study sample included 1503 individual patients, of whom 65 (4%) had MRI radiology reports that described a VMSA. Thirty-one (48%) of the 65 patients with VMSAs had further evaluation recommended by radiology. Ten (32%) of these 31 patients were followed clinically without further diagnostic testing for the VMSA. Of the 65 patients with VMSAs, only one was diagnosed with malignancy (multiple myeloma).Conclusion.While VMSAs are not frequently found on thoracolumbar MRIs ordered to evaluate back and/or leg pain, there is a large amount of heterogeneity in how these abnormalities are documented and managed. This may indicate the need for clinical guidelines for the reporting and management of VMSAs detected on spine MRI and for improvement in communication between radiologists and ordering providers.Level of Evidence: 3.