Diffuse idiopathic skeletal hyperostosis (DISH) is a common skeletal disorder in which an abundance of bone is formed along the anterior longitudinal ligament, leading to segmental fusion.¹ This condition was first described by Forestier and Rotes-Querol in 1950 and further defined by Resnick et al. in 1975.^2,3 This condition leads to progressive segmental fusion of the spine, resulting in pain, stiffness, and dysphagia in the cervical spine. Despite inappropriate bone formation, quantitative CT studies have shown no increase in bone mineral density in patients with DISH.⁴ As a result, patients with DISH and low bone mineral density are at a higher risk for incurring vertebral fractures with long lever arms on either side of the fracture, resulting in highly unstable lesions.

Given the long lever arms in either direction, the standard recommendation has been for long-segment fusion to minimize the risk of neurological compromise.⁵ Despite this, unstable fractures do not occur in a vacuum and individual clinical circumstances must be considered. Here, we present the largest case series to date of patients with extension-type DISH fractures treated with external orthosis alone. This series allows a critical evaluation of the natural history of the disease, as well as a critical consideration of external orthosis as a treatment option in patients for whom surgical intervention may not be a safe or attainable option.

Methods

A retrospective review was conducted from 2015 to 2022 at a large level 1 trauma center. Following this, departmental records were queried to identify patients presenting with DISH fractures over the past 7 years, including all patients admitted to the hospital and for whom a neurosurgeon was consulted. Individual charts were then reviewed and patients who underwent conservative therapy with external orthosis alone were identified. The use of orthosis, as opposed to surgical intervention, was decided on the basis of a combination of factors that may make selected patients poor surgical candidates, e.g., age, body mass index, fragility, medical comorbidities, and concurrent traumatic injuries. Individual radiographic and clinical information was reviewed by two separate authors (P.S.P. and G.P.G.).

Patients were considered to have complete follow-up if they were cleared from their orthosis after being seen in clinic by the managing spine surgeon and had supporting radiographic evidence indicating fusion, including a combination of upright radiographs, dynamic radiographs, and CT imaging based on the location of the fracture and clinical suspicion. Following clearance, patients were gradually weaned out of their respective orthosis over the course of several weeks. Only extension-type fractures without evidence of gross displacement consistent with AO Spine type B3 fractures with an associated M2 modifier were included. Fractures at all levels were included and subcategorized accordingly. DISH was defined by anterior osteophytes spanning at least 4 vertebral levels with normal disc spaces and no normal sacroiliac joints. Patients with neurological deficits were excluded from analysis. An orthopedist was consulted on all patients to ensure adequate fitting and maximal tolerance and compliance of bracing. The bracing schedule included wearing the orthosis at all times, with the exception of those patients with skin breakdown, who could remove the orthosis to prevent skin injury and for hygiene purposes, at which times strict spinal precautions were maintained. Additionally, strict spinal precautions were maintained prior to brace placement, and postbracing radiographs were obtained and reviewed to ensure that correct alignment was maintained in the chosen brace. Fracture morphology was subcategorized as either transosseous or transdiscal (Fig. 1). Statistical analysis was conducted in Excel version 2204 (Microsoft Corp.). Chi-square analysis was conducted for categorial variables, and the Student t-test was used for continuous variables. A p value < 0.05 was chosen for statistical significance.

CT images obtained in a 71-year-old male presenting with a T10–11 extension fracture in the setting of DISH after falling down stairs. A: Sagittal image demonstrating a transdiscal extension fracture with widening of the disc space. B: Sagittal image demonstrating a transosseous component extending to the left T10 pedicle through the vertebral body.

Results

Twenty-seven patients were identified as presenting with extension-type DISH fractures during the inclusion time frame. Patients had an mean age (± SD) of 78.92 ± 9.90 years and were most commonly White males (Table 1). Most fractures occurred at the level of the thoracolumbar junction, with 22/27 (81.5%) occurring between T5 and L2. Three fractures occurred between C4 and C7, and 1 fracture occurred at L5. Five patients were excluded from analysis as they either were lost to follow-up or died prior to the scheduled follow-up.

Twenty-two patients had complete follow-up on final evaluation (Table 2). Of these 22 patients, 21 (95.5%) were treated successfully with external orthosis and able to have their orthosis cleared at follow-up without the development of symptoms. Regarding fracture morphology, transosseous morphology was most common, occurring in 10/22 (45.5%) cases, followed by transdiscal morphology in 7/22 (31.8%) cases. Combined transdiscal and transosseous morphology occurred least often, in 5/22 (22.7%) cases. One patient developed acute spinal cord injury at the T8–9 level 1 month after presentation during a prolonged hospitalization, secondary to dysphagia and placement issues (Fig. 2). In this patient, spinal precautions and bracing were not being maintained and he was no longer wearing his brace at all. Unfortunately, he was found to have acute-onset paraplegia, at which time MRI was obtained demonstrating cord compression, requiring posterior decompression and instrumented fusion (Fig. 3). No cases of complications related to bracing were reported.

CT images obtained in an 80-year-old male presenting with a T7–8 extension fracture in the setting of DISH after a fall from a ladder. Sagittal (A) and coronal (B) images demonstrating a transdiscal extension fracture with widening of the disc space (arrows).

MR images obtained in an 80-year-old male presenting with a T7–8 extension fracture with acute neurological decline. Sagittal (A) and axial (B) images demonstrating significant cord compression at the T7–8 level.

Discussion

The radiographic appearance of senile ankylosis hyperostosis was first described by Forestier and Rotes-Querol in 1950.³ According to Resnick et al., these manifestations include osteophytes flowing anterolaterally along at least 4 contiguous vertebral bodies, the absence of degenerative disease, and the absence of ankylosis of the facet joints, or sacroiliac bony fusion.² The treatment of fractures in the setting of DISH represents a complicated pathology because of the presence of long lever arms above and below the fracture level, resulting in increased instability. Given these biomechanical considerations, these lesions are most appropriately treated with surgical stabilization for the prevention of neurological injury.^5–7 In addition to mechanical forces across the index fracture, the poor bone quality associated with DISH also makes these fractures unideal for treatment with conservative therapy.⁴

Despite biomechanical considerations, the presence of DISH increases significantly with advanced age and has been associated with obesity, diabetes mellitus, hyperlipidemia, etc.^8–10 Specifically, although the ideal treatment of these fractures is surgical intervention, in circumstances where surgical intervention is of exceptionally high risk or prohibitive, the surgeon should be knowledgeable regarding treatment of these fracture types with external orthosis and the associated outcomes. In a series conducted by Cirillo et al. in 2022, 22 patients presenting with AO Spine type B3 fractures were evaluated, with 13 of these patients having either DISH or ankylosing spondylitis.¹⁰ In their series, postoperative complications were present in 63.1% of patients, including surgical site infection rates of 22.7% and mortality rates of 13.6%. In another series conducted by Balling and Weckbach evaluating 23 patients undergoing surgical intervention for hyperextension injuries in the setting of DISH, the surgical complication rate was found to be 13.0%, with the medical complication rate being much higher at 65.2% and a 3-month mortality rate of 17.4%.⁵ In the most comprehensive study to date, by Caron et al. in 2010,¹¹ 122 patients with ankylosing spinal disorders presenting with fracture were evaluated, and of these patients, 75 were treated surgically. In their series, as many as 84% of patients had at least one medical complication and 26% had some type of surgical complication, with wound infection being the most common at 16%. The abovementioned studies highlight that while most of these fractures should be treated surgically, the thoughtful spine surgeon should consider the risks and benefits of surgical intervention compared with orthosis.

Regarding conservative treatment with orthosis, few studies have been reported to date. Although small case reports and series have demonstrated that fracture healing may occur with external orthosis, no large series has been reported to date quantifying the likelihood of healing and the underlying risk of developing neurological injury.^12–14 In the previously mentioned study by Caron et al. in 2010,¹¹ bracing was chosen as the definitive treatment in 20 patients; however, it was chosen for palliative reasons in 17 patients. Therefore, formal follow-up data were limited regarding union or neurological status. Of the follow-up data that were available, there was no change in angulation, translation, or displacement in those who were braced. Our results demonstrated that although a good radiographic and clinical outcome can be achieved with external orthosis in cases of high surgical risk, there remains a small risk of delayed neurological injury.

The present study included several important limitations. Although this study included the largest series to date for evaluation of this patient population, conclusions were difficult to draw given the relatively small sample size. Additionally, in our series 5 patients were lost to follow-up, making conclusions regarding these patients difficult. Although it can be determined that bracing may result in good clinical outcomes, further study is required to understand the rate for potential spinal cord injury. Additionally, our study was largely represented by fractures at the thoracolumbar junction, which were most common. Interestingly, however, we also reported good outcomes in the treatment of cervical fractures that would presumably be more unstable comparatively. Further research comparing transdiscal injuries and transosseous injuries may also improve future understanding of these fracture patterns.

Conclusions

Here, we demonstrated that bracing of extension fractures in the setting of DISH may result in good clinical and radiographic outcomes in high-risk surgical candidates; however, a risk of neurological injury secondary to delayed instability remains. This risk should be balanced against the risks of operative intervention, which are very high in this patient population.

Disclosures

Dr. Resnick: advisory board of NIDUS.

Author Contributions

Conception and design: Page, Greeneway, Lake, Hanna, Resnick. Acquisition of data: Page, Greeneway. Analysis and interpretation of data: Page, Greeneway, Resnick. Drafting the article: Page. Critically revising the article: Page, Josiah, Hanna. Reviewed submitted version of manuscript: Page, Lake, Brooks, Hanna, Resnick. Approved the final version of the manuscript on behalf of all authors: Page. Statistical analysis: Page. Administrative/technical/material support: Page. Study supervision: Page.