LAUSR

e read with great interest the recent article by Xu W et al, in which the authors compared cervical alignment among three-level total disc replacement (TDR), two prostheses with one cage (2D1C), one prosthesis with two cages (1D2C), and anterior cervical discectomy and fusion (ACDF), then identify the importance of the cervical balance of cervical spondylotic myelopathy. We highly appreciate their contribution to this important topic. However, we think that several details in the study design need consideration, and we wish to discuss some issues with the authors to further our understanding of the matter. First, we believe that the assessment of ASD of C7-T1 was hard to achieve via lateral x-rays because the disk spaces of C7-T1 of some patients were obscured by the underlying structure on the lateral X-rays, which was even reflected in the figures in the article. The radiological evaluation of ASD in this study may lack reliability and accuracy. Schiebler et al reported that magnetic resonance imaging (MRI) could provide a highly accurate picture of the disk’s microarchitecture, and findings of early disk degeneration could be seen in vivo before either the loss of disk height or loss of central disk signal intensity. The study of Pfirrmann et al. showed that degeneration of disks can be detected noninvasively by MRI with sufficient accuracy. Therefore, MRI can be a better alternative. Furthermore, the authors presented the radiological criteria for disk degeneration by plain films. However, the reference provided was from Alhashash et al; it only narrated criteria for degeneration on MRI in detail, which was according to Matsumoto et al that defined disk space narrowing >25% as one of the criteria of ASD. However, in this study, the value is 50%. We believe that authors should state the basis of the radiological criteria in this study. Second, the operated segments in this study included C3C6 and C4-C7. Thus the adjacent segments are also different. Theoretically, there are differences in the incidence of ASD in different segments. Alhashash et al believed that ASD in patients who underwent ACDF was more likely to