S. Haneder
The Spine Journal
Available online 25 January 2014
In Press, Accepted Manuscript
Background/Context
1H magnetic resonance imaging of the spine can rule out common causes of low back pain (LBP) such as disc protrusions or nerve root compression; however, no significant causal relation exists between morphology and extent of symptoms. Functional MR techniques, such as 23Na, may provide additional information, allowing indirect assessment of vertebral glycosaminoglycan concentrations, decreases in which are associated with early degenerative changes. PURPOSE: To evaluate 23Na-MR of (1) asymptomatic healthy volunteers and (2) symptomatic patients with LPB and correlate the results to the Pfirrmann classification of MR disk morphology.
Study Design/Setting
Retrospective cohort study at an academic medical center.
Patient Sample
Two groups were studied: (1) 55 healthy volunteers (31 m, 24 w; mean age 28.8 years) and (2) 12 patients (6 m, 6 w; mean age: 35.3 years) with a recent history of LBP.
Methods
This study was performed without grant support; the authors have no conflicts of interest to disclose. Lumbar spines of the above groups were examined on a 3.0 Tesla MR-scanner with morphologic 1H and 23Na imaging. Intervertebral disk (IVD) 23Na at each level was normalized (23Nanorm). Distribution and differences between mean 23Nanorm corresponding to each Pfirrmann classification were evaluated in the two study groups (ANOVA). Linear correlations between 23Nanorm, BMI, and age were assessed (Pearson Correlation Coefficient). Gender-dependent differences were evaluated (paired t-test).
Outcome Measures
Physiologic measure – IVD 23Nanorm as determined by 23Na MRI.
Results
A normal distribution of 23Nanorm was confirmed for both groups (p = 0.072 and p = 0.073). The mean Pfirrmann score statistically significantly differed between them (p < 0.0001). 23Nanorm was statistically significantly reduced in degenerated IVDs (Pfirrmann scores 4+5) (p < 0.0001). No statistically significant differences were seen for the mean 23Nanorm of IVDs with the same Pfirrmann score in healthy volunteers and patients (0.469 < p < 0.967). Age (0.007< R2 < 0.202) and BMI (0.074 < R2 < 0.288) showed either weak or no correlation to 23Nanorm. Mean 23Nanorm was significantly (p = 0.0002) greater in women relative to men.
Conclusion
The results underline the feasibility and robustness of 23Na-MR imaging of human IVDs and affirm in a large cohort decreases in 23Na IVD content seen with disc degeneration.