Glycosaminoglycan Chemical Exchange Saturation Transfer of Lumbar Intervertebral Discs in Healthy Volunteers.

STUDY DESIGN Evaluation of a new quantitative imaging technique in a prospective study design. OBJECTIVE To assess glycosaminoglycan (GAG) content of lumbar intervertebral discs (IVDs) in healthy volunteers with chemical exchange saturation transfer (CEST). SUMMARY OF BACKGROUND DATA Biochemical alterations of lumbar discs are present before the appearance of morphological changes. GAG loss plays a central role in these degenerative processes. METHODS Lumbar intervertebral discs of healthy controls (26 women, 22 men; mean age 31 ± 8 years; range: 21-49 years) without lumbar back pain were examined at a 3 Tesla magnetic resonance imaging (MRI) scanner in this prospective study. None of the participants were overweight or had previous surgery of the lumbar spine. The MRI protocol included standard morphological, sagittal and transversal T2-weighted (T2w) images to assess Pfirrmann score and to detect disc disorders according to the Combined Task Force classification of five lumbar IVDs (L1 to S1). A prototype glycosaminoglycan chemical exchange saturation transfer (gagCEST) sequence was applied to measure GAG content of the nucleus pulposus (NP) and annulus fibrosus (AF) by identifying the magnetization transfer asymmetry ratio (MTRasym) in a region-of-interest analysis. Morphological and biochemical imaging analysis were statistically tested for quantitative differences between different grades of IVD degeneration and disc disorders. RESULTS gagCEST values of NP demonstrated a significant negative correlation with morphological Pfirrmann score (r = -0.562; P < 0.0001). The MTRasym values were higher in non-degenerative lumbar IVDs (Pfirrmann 1-2) compared with degenerative lumbar discs (Pfirrmann 3-5; 2.92% ± 1.42% vs. 0.78% ± 1.38%; P < 0.0001). The MTRasym values of NP were significantly higher in normal appearing discs compared with herniated IVDs (2.83% ± 1.52% vs. 1.55% ± 1.61%; P < 0.0001). We found a significant negative correlation between gagCEST values and the graduation of disc herniation (r = -0.372; P < 0.0001). CONCLUSION Biochemical imaging with gagCEST distinguished morphologically degenerative from non-degenerative lumbar IVDs (in NP and AF) of healthy volunteers at a clinical 3T-MRI system. The depletion of GAG content in degenerative lumbar discs correlated significantly with the morphological disc classification. We could demonstrate that disc disorders, such as protrusion and extrusion, were accompanied by lower GAG content. LEVEL OF EVIDENCE 2.