Purpose: The goal of this study was to refine clinical MRS to optimize performance and then determine whether MRS-derived biomarkers reliably identify painful discs, quantify degeneration severity, and forecast surgical outcomes for chronic low back pain (CLBP) patients.
Methods: We performed an observational diagnostic development and accuracy study. Six hundred and twenty-three (623) discs in 139 patients were scanned using MRS, with 275 discs also receiving provocative discography (PD). MRS data were used to quantify spectral features related to disc structure (collagen and proteoglycan) and acidity (lactate, alanine, propionate). Ratios of acidity to structure were used to calculate pain potential. MRS-SCOREs were compared to PD and Pfirrmann grade. Clinical utility was judged by evaluating surgical success for 75 of the subjects who underwent lumbar surgery.
Results: Two hundred and six (206) discs had both a successful MRS and independent pain diagnosis. When comparing to PD, MRS had a total accuracy of 85%, sensitivity of 82%, and specificity of 88%. These increased to 93%, 91%, and 93% respectively, in non-herniated discs. The MRS structure measures differed significantly between Pfirrmann grades, except grade I versus grade II. When all MRS positive discs were treated, surgical success was 97% versus 57% when the treated level was MRS negative, or 54% when the non-treated adjacent level was MRS positive.
Conclusion: MRS correlates with PD and may support improved surgical outcomes for CLBP patients. Noninvasive MRS is a potentially valuable approach to clarifying pain mechanisms and designing CLBP therapies that are customized to the patient. These slides can be retrieved under Electronic Supplementary Material.
Keywords: Clinical outcomes; Diagnosis; Low back pain; Magnetic resonance spectroscopy.