Is level- and side-specific multifidus asymmetry a marker for lumbar disc pathology?

BACKGROUND CONTEXT: Several reports suggest that level- and side-specific multifidus atrophy or fat infiltration may be related to localized spinal pathology and symptoms. In particular, a study using a porcine model reported rapid level- and side-specific multifidus atrophy and adipocyte enlargement resulting from anterolateral disc or nerve root lesions. PURPOSE: To investigate asymmetry in cross-sectional area (CSA) and fat infiltration in multifidus and other paraspinal muscles in patients with acute or subacute unilateral symptoms of radiculopathy and concordant posterolateral disc herniation. If multifidus asymmetry is indeed related to local pathology, this may serve as a marker for helping to target the search for less clearly identifiable pathology responsible for low back-related symptoms, which currently remains elusive in approximately 85% of those seeking care. STUDY DESIGN: Cross-sectional observational study. PATIENT SAMPLE: Subjects were patients referred to magnetic resonance imaging (MRI) with unilateral leg symptoms of less than 6 weeks onset suggestive of radiculopathy, with a consistent posterolateral lumbar disc herniation verified on imaging. METHODS: Using T2-weighted axial MRI, measurements were obtained for total muscle CSA and signal intensity, functional (fat-free) CSA, and the ratio of functional CSA to total CSA. RESULTS: Forty-three subjects met the inclusion criteria. The ratio of functional CSA to total muscle CSA was smaller on the side of the herniation than on the unaffected side, both below (mean 0.69 vs. 0.72, p=.007) and at the level of herniation (mean 0.78 vs. 0.80, p=.031). Multifidus signal intensity (fat infiltration) was greater on the side of the herniation at the level below the herniation (p=.014). Contrary to expectation, greater total multifidus CSA was found ipsilateral to the pathology at the level of herniation (p=.033). No asymmetries were found at the level above the herniation or in any other paraspinal muscles, with the exception of higher signal in the erector spinae at the level and side of herniation. CONCLUSIONS: Multifidus may be particularly responsive to, or indicative of, localized lumbar disc or nerve root pathology within the first 6 weeks of symptoms as expressed through fat infiltration, but not through CSA asymmetry on MRI. However, such measurements are not reliable markers of lumbar pathology on an individual basis for use in clinical or research settings. (C) 2012 Elsevier Inc. All rights reserved.