Sensitivity of T1ρ MRI and Pfirrmann Grade to Discogenic Pain

Abstract

Introduction Diagnosis of degenerative disk disease (DDD), a common cause of low back pain (LBP), remains difficult, subjective, and controversial.1 Healthy discs rely on diffusion to transport nutrients and waste products between the surrounding blood vessels and the ordered collagen fibers of the annulus fibrosus (AF), to the central gel-like nucleus pulposus (NP). Age-related degradation is marked by a loss of the gel-like consistency of the NP including decreased proteoglycan content and decreased water concentration,2 while late-stage DDD is characterized by an NP indistinguishable from AF and a collapsed disk space. Due to the lack of a proper gold standard, the presence of pain in each disk is currently determined by provocative discography.3 The technique relies on the patients' subjective perception of pain as a needle is inserted into the disc. With limitations,4 imaging methods have relied on X-ray based methods to detect disk collapse and magnetic resonance imaging-based (MRI) Pfirrmann grading5 method that relies on a semiquantitative and subjective grading of the signal intensity on a midsagittal T2 MRI. An alternative technique is T1ρ MRI, which has been demonstrated to correlate with NP proteoglycan content, and swelling pressure in cadaveric tissue.6–9 In this study, we compare T1ρ measurements and Pfirrmann grading of patients' lumbar discs presurgery to determine the sensitivity of both methods in predicting painful discs (as measured by provocative discography) that eventually required surgery. The objective of this study is to determine the sensitivity T1ρ and Pfirrmann grade in predicting painful discs that were subsequently surgical-treated in back pain patients. Materials and Methods All MRI scans were performed on a 3 Tesla Siemens Tim Trio clinical scanner using the vendor-supplied spine array coil with approval from the IRB and with subjects' consent. T1ρ and T2 MRI were performed on patients a week before fusion surgery ( n = 12, 49 levels, mean age 44 ± 6 years, range 30–53). Discography was performed with the placement of 22 gauge needles into the center of the L2/L3 through L5/S1 discs, using the IntelliSystem with digital pressure display, and the presence of pain at each level was ascertained. Following coregistration and segmentation procedures, average T1ρ (in milli seconds) was recorded from all lumbar discs using algorithms written in Matlab. Pfirrmann grading was performed on T2 MRI of the same discs by a single reader with several years of experience in disk MRI analysis. Statistical descriptives and regressions were performed in SPSS 19.0 to evaluate any relationships between T1ρ and Pfirrmann grade and painful discs that required surgery. Results The figure below shows examples of quantitative T1ρ maps (in color) corresponding to T2-weighted MRIs (grayscale) of a patient with DDD (left panel) and an asymptomatic normal volunteer (right panel). Average T1ρ values are indicated below each disk in ms and the L4-5 disk in A (T1ρ = 42ms) subsequently treated by 360-degree fusion. An ROC analysis of T1ρ and Pfirrmann grade using a continuous rating scale of predictor of painful discs that subsequently underwent fusion surgery demonstrates higher sensitivity of T1ρ. Results of the ROC analysis are shown in Table 1 . [Table: see text] Conclusion T1ρ MRI was more sensitive than Pfirrmann grading in detecting painful discs that required fusion surgery. T1ρ MRI shows a promising ability to detect painful discs and has the potential to measure patient outcomes from back surgery. I confirm having declared any potential conflict of interest for all authors listed on this abstract Yes Disclosure of Interest None declared Adams M, Roughley P. Spine 2006:2151–2161 Buckwalter JA. Spine 1995;20(11):1307–1314 Tehranzadeh J. Radiologic Clinics of North America 1998;36(3):463–495 Sheehan NJ. Annals of Rheumatic Disease 2010;69(1):7–11 Pfirrmann CW, et al. Spine 2001;26(17):1873–1878 Johannessen W, et al. Spine 2006;31(11):1253–1257 Auerbach J, et al. European Spine Journal 2006;15 Suppl 3:S338–344 Blumenkrantz G, et al. Magnatic Resonance Medicine 2010;63(5):1193–1200 Borthakur A, et al. Spine 2011