Association between diffuse idiopathic skeletal hyperostosis and thoracic kyphosis in patients with cervical myelopathy: a retrospective observational study

Diffuse idiopathic skeletal hyperostosis (DISH) is a non-inflammatory osteophytic disease of the ligaments and entheses that can affect the entire body. The unique pathology of DISH has been recognized to increase the risk of spinal disorders in affected and adjacent regions [1‐3]. When a fracture occurs in the area of a DISH lesion, the structure of the spine destabilizes, possibly leading to a spinal cord injury. A predisposition to ossification has been known to influence not only activities of daily living but also quality of life [4, 5]. Notably, the spinal ankylosis caused by DISH may lead to impaired posture [6], also called sagittal imbalance, throughout the entire spine. However, very few studies [7, 8] have evaluated spinal sagittal balance according to whether or not DISH is present. Therefore, to clarify the effect of DISH on radiologic parameters related to sagittal alignment, we retrospectively investigated the prevalence of DISH in patients with cervical myelopathy caused by cervical OPLL or cervical spondylosis and compared the sagittal alignment of the spine between those with and without DISH, using propensity score matching.

The prevalence of musculoskeletal disorders has been steadily increasing in parallel with our aging society [13, 14]. DISH is a pathologic state identified by heterotopic ossified bridging in the thoracic spine [15] and is also a systemic condition characterized by ossification of entheses throughout the body [16‐18]. In a study of 504 healthy volunteers, Banno et al. [19] demonstrated on lateral radiographs that DISH had a prevalence of 14.3% in men and 4.3% in women. Similarly, Hirasawa et al. [20] reviewed 558 patients with an average age of 66.7 years who underwent CT chest to pelvis and found that the prevalence was 38.7% in men and 13.9% in women. Therefore, spine surgeons and physicians should recognize the high prevalence of DISH in elderly patients.

Few groups [7, 8] have investigated the association between sagittal alignment and the presence of DISH. Yamada et al. [7] investigated a patient population in which lumbar spine surgery was required and demonstrated that patients with DISH had significantly more thoracic kyphosis and less lumbar lordosis compared with patients without DISH. Uehara et al. [8] investigated sagittal spinal parameters and the occurrence of DISH in 411 patients from a registry that included data on a population of 5352 and showed that those with DISH in the thoracic spine had a higher sagittal vertical axis, a higher T1-slope, and more thoracic kyphosis compared with those without DISH. In the present study, we found marked differences in thoracic kyphosis and in the cervical vertical axis and C7 slope. Interestingly, we did not find a significant difference in the C2-7 lordotic angle or in any spinopelvic parameter, including SS, PT, PI. and LL, between those with DISH and without DISH. These findings indicate that patients with cervical myelopathy mostly posture cervical straight alignment with reciprocal lumbar lordosis to prevent the spinal cord compression in the canal of the cervical spine.

DISH often coexists with ossification of the spinal ligaments, including the posterior longitudinal ligament, ligamentum flavum, and interspinous and supraspinous ligaments [21‐25]. Nishimura et al. [25] showed that the prevalence of DISH was as high as 48.7% in patients with cervical OPLL. They also speculated that bony bridging may develop at the thoracic level as a result of DISH and progress to the cervical and/or lumbar spine with aging. Similarly, 22 (37.3%) of the 59 patients with DISH in the present study also had cervical OPLL. Katzman et al. [26] collected data for 1500 men from the Osteoporotic Fractures in Men study and for 1267 women from the Study of Osteoporotic Fractures and assessed the prevalence of DISH and follow-up radiographs for approximately 4 years to evaluate the change in sagittal alignment in these populations. They identified DISH in 15% of men and 12% of women and found no significant difference in kyphotic change between those with and without DISH during follow-up. Interestingly, they also demonstrated that women with DISH had less kyphotic change in the thoracic spine over 15 years than those without DISH. Hirai et al. [27] prospectively collected data for 239 patients with cervical OPLL and categorized them into three grades according to the number of levels with OPLL using the cervical OP index. They found that the OP index grade was significantly correlated with extension of DISH. Similarly, Yoshii et al. also investigated a relation between the prevalence of ossification of anterior longitudinal ligament (OALL) and ossification of nuchal ligament (ONL) in cervical OPLL patient dataset, and demonstrated that the existence of ONL is significantly associated with age, male, cervical and thoracic OALL [24]. The consistent findings regarding coexistence of ossified lesions throughout the spine in patients with DISH suggest a need to consider confounding factors when assessing the association between clinical findings and radiologic findings. Although longitudinal investigation will be needed in the future, we assume that DISH may extend from the thoracic spine proximally and/or caudally with advancing age.

Patients with severe kyphosis often complain of a variety of symptoms, including severe low back pain, depression, gastroesophageal reflux caused by excessive pressure on the abdomen, and difficulty gazing straight ahead. Thoracic kyphosis has generally been considered to be a spinal deformity resulting from vertebral fractures, which are prevalent in the elderly because of osteoporosis. Katzman et al. [28] demonstrated that thoracic hyperkyphosis in the elderly is correlated with loss of muscle mass at the lower lumbar levels. It has also been shown that the range of motion at the lumbar spine is more limited in patients with DISH than in those without DISH [29]. Although the mechanisms that cause excessive kyphosis in patients with DISH are yet to be investigated, it is likely that bony bridging of the anterior longitudinal ligament gradually progresses with worsening of sagittal imbalance. Furthermore, high C-SVA associated with thoracic hyperkyphosis often influences postoperative clinical neurologic outcomes as well as cervical alignment change after cervical spine surgery. Sakai et al. reported that neurological recovery after cervical laminoplasty was poor in cervical myelopathic patients with preoperative high C-SVA [9, 30]. In addition, we previously demonstrated that graft dislodgement likely occur after anterior cervical surgery in cervical OPLL patients with high C-SVA caused by ankylosing thoracic spine [31]. Therefore, attentions to sagittal alignment in not only the thoracolumbar spine but also global spine including the cervical spine should be paid in DISH patients with spine related disorders.

DISH has been recognized to be not only a systemic skeletal abnormality affecting the entire body but also a result of metabolic syndrome. Okada et al. [32] reviewed 327 patients who attended for a routine wellness check and found that the prevalence of metabolic syndrome was significantly higher in patients with DISH than in those without DISH (28.9% vs 16.0%). Furthermore, in a study by Lantsman et al. [33] the cross-sectional area of visceral fat on abdominal CT scans was greater in patients with DISH than in healthy control subjects. Fassio et al. [34] reported that 20.8% of 96 postmenopausal women with type 2 diabetes mellitus had DISH and that the serum sclerostin level was lower in the women with DISH than in those without DISH. In the present study, although not a statistically significant finding, the prevalence of diabetes mellitus was higher (46.4%) in the DISH group. Furthermore, it has been reported that patients with DISH have an increased risk of coronary artery calcification [35, 36]. These findings suggest that adipokines might be associated with not only obesity but also ectopic bone formation in patients with diabetes mellitus, raising the possibility of a link between the onset and extent of DISH and a systemic metabolic disorder.

This study has several limitations. First, it had a cross-sectional rather than a longitudinal design, so no conclusions could be reached concerning causality. Second, it included a patient population with a specific disease; therefore, further investigations are needed in the general population to accumulate more evidence. Nevertheless, despite these limitations, we believe that our findings provide important information on the radiologic features of DISH.

This study is the first to compare sagittal alignment in patients with cervical myelopathy according to whether or not they have DISH. We found that patients with DISH were more likely to have excessive kyphosis in the thoracic spine, a high C7 slope, and high C-SVA.