Ensured by highly qualified UTE images, we for the first investigated the feasibility of UTE imaging in assessing the damage of CEPs for patients with chronic low back pain. As shown in the results, the image quality was confirmed by high SNR and CNR (mean SNR: 31.48±2.73; mean CNR: 9.1±1.97). Excellent inter-observer agreement of CEP and VEP damaged was confirmed by high value of 0.839 (95%CI:0.925~1.05; P<0.001) in Cohen's kappa test.
Subtracted UTE imaging has been shown an effective clinical method to display CEP changes noninvasively, because it maintained bound water signal with ultrashort T2* and eliminated the effect of nucleus pulposus and annulus fibrosus by removing the signal contribution from free water protons (1, 14). Based on this advantage, bound water weighted CEP presented hyperintense signal while the signal of VEP was hypointense. The structure of CEP could thus be well visualized and the image contrast between CEP and VEP was strong.
A modified CEP grading system, based on six-level Rajasekaran grading method (8), was developed in this study, defining additional criteria on CEP and VEP damage. Unlike displaying only the whole endplate on conventional T1 weighted imaging, UTE imaging with well distinguished CEP and VEP allows for accurate diagnosis of CEP and VEP damage. The original six-level Rajasekaran grading method without accurate grading for CEP and VEP is thus not sufficient, and a new grading system with extra new grading is highly requested in this study.
Chen. KC et al. considered that the defect of CEP might cause weakened or lost VEP bone mineral density, and then led to VEP damage (10). In other words, the integrity of CEP ensured that VEP was not damaged. UTE imaging provided a possibility for observing pathological processes that transit from CEP damage to VEP damage in a noninvasive way. Damaged VEP appears mostly schmorl's nodule sign, a classic sign of endplate defect. When VEP defects, nucleus pulposus herniation presents as the characteristic schmorl's nodule and may be along with vertebral bone marrow involvement (e.g., edema), a so-called Modic change, which is also the boundary between grade III and IV. Grade III was defined as CEP with local depression but complete continuity, VEP was not involved and without Modic change. The corresponding cases in our scheme showed that the integrity of CEP ensured that VEP was not involved and vertebral bone marrow was not damaged on subtracted UTE images (Fig. 2c). Grade IV was defined as CEP begin with defects, the damaged area of CEP less than 25% of total area, VEP involved and along with Modic change, while the corresponding cases showed that both CEP and VEP were damaged with vertebral bone marrow involved on subtracted UTE images (Fig. 2d), so Modic change occurred. UTE imaging could display obviously the process grade III transforms to IV, which the damage extends from CEP to VEP. Subtracted UTE images could be observed that whether CEP and VEP damage, bone marrow involve or not so that the corresponding TEPS could be calculated accurately for each IVD and used for further clinical application.
In addition, using Kendall's TAU-B analysis, significant relationship between the TEPS and IVD degeneration was revealed by high coefficient of 0.864. It indicates that the increase of CEP degree might induce more serious disc degeneration, although pathologic results are requested for further validation. IVD degeneration was a major cause of chronic low back pain (2, 3).
Many factors contribute to disc degeneration including age and Body Mass Index (BMI). Rade. M et al. reported that when TEPS was included as predictor in multivariable model, the covariates such as age and BMI became insignificant that mean endplate defect was the prevalent factor in disc degeneration (16). A minor interruption of endplate may be sufficient to cause significant changes in the mechanical environment of IVD and effect diffusion paces for maintaining IVD nutrition and nucleus pulpoid hydration (17, 18). In addition, herniating annulus fibrosus may led to CEP away from the subchondral bone and promoting endplate defect (19, 20). In another way, anaerobic bacteria may also enter the IVD through the defect of endplate and cause IVD degeneration (21).
Rajasekaran. S et al. reported that the cut-off value of TEPS was 5, over which the prevalence of IVD degeneration was significantly higher. When TEPS is above 5, mechanical environment and nutritional pathways of IVD are blocked, IVD degeneration is impossible to reverse through natural healing or adaptive remodeling (4). The following treatments of biological treatment, stem cell therapy, geneticmanipulation, anabolic stimuli, even growth factors that are identified as possible treatments in the future were not suitable for these IVDs (8). Therefore, it is necessary to evaluate TEPS accurately in order to help for the clinical treatment and evaluate disc degeneration grade. Subtracted UTE images could show well structural CEP images and the corresponding TEPS could be calculated accurately on UTE images.
This study also has some limitations. First, histologic references of patients corresponding to the new grades are lacked. Second, CEP minor structural damage (e.g., fissure or crack) may not appear in the UTE images, but these defects represent microscopic compositional changes, such as acidophilic degeneration, focal proliferation, calcification and cell death, which are known to occur in CEP (22). Third, although Rade. M et al. mentioned that endplate defect was the prevalent factor in disc degeneration, the multi-aetiological nature and multifactorial condition (e.g., BMI, age, sex, genetics) of IVD degeneration may influence the assessment of CEP defects and their roles in IVD changes. Nonetheless, our study raises awareness that UTE imaging provides the opportunity to identify CEP defects which would be missed and reveals the diagnosis of CEP defects to be clinically beneficial in predicting the development and severity of IVD degeneration, which may have effects on clinical treatment options and patient outcomes.