RESUMO
The aim of this study was to investigate to what extent annular fissures, vertebral and endplate changes, and Modic changes (MCs), coexist in low back pain (LBP) patients by using multiple imaging modalities. Sixty-two LBP patients (mean age 45 years, range 24-63, 53% men) were examined with same-day CT-discography and MRI. Intervertebral discs punctured for discography (n = 204) were evaluated on MRI [Pfirrmann grade, High-Intensity Zone (HIZ)] and on CT-discograms [Modified Dallas Discogram Score (DDS)]. DDS≥ 1, i.e., disc fissures involving the outer annulus were further digitomized into delimitable fissuring (<50% of annulus affected) or non-delimitable annular fissuring. Using both MRI and CT, adjacent vertebrae and endplates were assessed for MC, vertebral sclerosis, and a modified endplate defect score (EPS). In 194 discs the contrast agent was adequately injected during discography, of which 160 (83%) displayed outer annular fissures, with 91 (47%) of the latter being delimitable fissures. Most discs with delimitable fissures were moderately degenerated; 68% Pfirrmann grade ≤3, 71% EPS ≤ 2, and 12% displayed MC. The majority (76%) of MCs were associated with advanced adjacent disc degeneration; 84% Pfirrmann grade ≥4, 76% with non-delimitable annular fissuring, 59% EPS≥ 4, and 34% EPS of 3. A total 95 HIZ (47%) were found, of which 54 had delimitable fissuring, while the remainder displayed non-delimitable fissuring. Vertebral sclerosis was commonly observed (26%), both with MCs (73%) and without MCs (27%), and not specifically linked to MC type 3. A total of 97% of segments with vertebral sclerosis displayed outer annular fissures. These findings were significant (0.046 > p > 0.0001), except between HIZ and adjacent sclerosis (p = 0.303). To conclude, the present study confirmed a close interplay between the disc and adjacent vertebra and endplates. The fact that a majority of discs with delimitable annular fissures did not coexist with pronounced endplate changes and/or MCs, however, supports the theory that disc fissuring is an early event in the degenerative cascade. This was further supported by the fact that MCs were strongly linked to extensive disc fissuring and to advanced endplate damage. Further, vertebral sclerosis was common also in vertebra without MCs and strongly associated to annular fissuring, indicating that sclerosis is a previously underestimated feature of a general degenerative process.
RESUMO
Modic changes (MCs) are gaining increased interest as potential generators of low back pain (LBP). The current aim was to investigate possible spinal loading effects on the MRI signal in MCs in patients with LBP. Supine lumbar MRIs were performed and immediately repeated with axial loading in 100 LBP patients. A total of 43 patients (23 male, mean age 45.7 years) had MCs. Each Modic was outlined on all sagittal T2-weighted images (>25% affected vertebrae). For reference, regions of interest were placed in both vertebrae without Modic and in Modic-free tissue in vertebrae with Modic. The Modic signal intensity, normalized to cerebrospinal fluid, and Modic volume were compared between MRIs with and without spinal loading. Of the 94 MCs, 36.2% (n = 34) were type I, 58.5% (n = 55) were type II, and 5.3% (n = 5) were type III. No differences in Modic volume (mean 0.046 cm3; p = 0.25) between the MRIs with and without spinal loading were found. In addition, no significant changes in Modic signal were induced by loading (mean 1.5% difference; p = 0.308). Loading increased the signal in the reference regions of interest in vertebrae both with Modic (mean 5.5%; p = 0.002) and without (mean 3.5%; SD 0.09; p = 0.02). To conclude, MRIs performed with and without spinal loading showed no change in either volume or signal of MCs, suggesting that most MCs are not instantaneously influenced by biomechanical load.