DOES A MODERN MRI PROTOCOL IMPROVE DIAGNOSIS OF LUMBAR DISC DISEASE COMPARED TO AN ORIGINAL MRI PROTOCOL FROM THE 1980’S?

Abstract

As part of a 10 year follow-up study investigating the relationship between MRI-diagnosed disc disease and low back pain (LBP), a comparison of MRI image acquisition protocols was conducted. The aim was to establish whether the modern protocol produced improved diagnoses of lumbar disc disease. This is of significance when attempting to determine links between lumbar disc disease and LBP. The proposed hypothesis was that little difference in the pathology reported of MRI lumbar spines between the surface coil acquired images (Coil-MRI) and phased-array acquired images (Phased-MRI) would be found.

Methods: Local ethics committee approval was granted for this study. 31 male subjects (aged 35–71 years) were recruited and underwent two subsequent scans. For both Coil-MRI and Phased-MRI scans sagittal dual echo, T1, axial T1 and T2 images were acquired. A Consultant Radiologist blindly reviewed the 62 scans continuously and reported on the pathology. Disease pathology assessment consisted of disc degeneration, disc herniation (based on 5-grade classification systems), facet hypertrophy (FH) and nerve root compression (NRC).

Results: A wide range of pathology was reported at all disc levels, particularly with regard to disc degeneration and herniation. Kappa agreement statistics were computed for each pathological feature at all disc levels. Disc degeneration and herniation reports were statistically consistent for all disc levels (kappa range: 0.6–0.8, p< 0.05 for degeneration & 0.5–0.7, p< 0.05 for herniation). The results show that at the L4/L5 disc level, 1 in 10 discs were reported as ‘moderately degenerate’ (an increase of 1 grade) in Phased-MRI scans. At the same disc level, 1 in 6 discs were reported as ‘moderately herniated’ in Phased-MRI scans compared to ‘bulging’ in Coil-MRI scans, indicating that Phased-MRI coil scans may improve clarity in particular for herniation diagnosis. Pathology for FH and NRC were limited, with the majority of subjects (over 91% for FH and NRC irrespective of protocol) presenting with normal features.

Conclusion: The statistical results indicate that few differences in pathological diagnosis of lumbar disc disease occurred, however Phased-MRI appears to increase confidence in diagnosing more severe features at some disc levels.