Abstract

*Beth Vettiyil M.D., Michael Aguilar M.D. and Oganes Ashikyan M.D.

ABSTRACT

Purpose: To compare STIR with T2-/T1-weighted (T2W/T1W) images in detecting abnormal endplate signal in the lumbar spine (L. spine) in patients presenting with back pain. Subjects and methods: After obtaining IRB exemption, we retrospectively reviewed 50 routine clinical lumbar MR studies of patients with back pain. All patients underwent MRI on 1.5T or 3T scanners. All scans consisted of sagittal T1W, T2W, and STIR sequences as well as axial T2W sequences. Patients with a history of lumbar spine surgery, infection, or neoplasm were excluded from this study. We evaluated a total of twelve endplates in each patient from T12-S1 levels. We recorded whether abnormal signal was present on T2W/T1W sequences, STIR sequences or both sequences. Modic type change was recorded for each endplate based on the appearance on T2W and T1W images. Results: There were 24 male and 26 female patients with ages ranging from 25 to 85 years. Abnormal STIR signal in the endplates was seen in 22% of all endplates in the lumbar spine. Abnormal T2W/T1W signal in the endplates was seen in 20.8% of all endplates. 4.7% of all endplates demonstrated abnormal signal only on STIR images. 5.2% of all endplates demonstrated abnormal signal only on T1W/T2W images. Overall, 8% of patients had abnormal findings only on STIR images, and 8% of patients had abnormal findings only on T1W/T2W images. Abnormal STIR signal was present in 83.3%, 72.3% and 91.7% of L. spine endplates that demonstrated Modic type I, II and III changes, respectively. Conclusion: STIR and T1W/T2W images provide complimentary, but sometimes different information about endplate integrity in the L.spine. Spine MRI protocols may benefit from the addition of STIR or other fat saturated fluid sensitive sequences to conventional T1W/T2W images to maximize MR sensitivity for detection of endplate pathology.