Peer-Review Reports

In-Hospital Neurologic Deterioration Following Fractures of the Ankylosed Spine: A Single-Institution Experience Terry K. Schiefer, Brian D. Milligan, Colten D. Bracken, Jeffrey T. Jacob, William E. Krauss, Mark A. Pichelmann, Michelle J. Clarke

Key words Ankylosing spondylitis - Diffuse idiopathic hyperostosis - Spinal cord injury - Trauma -

Abbreviations and Acronyms AS: Ankylosing spondylitis ASIA: American Spinal Injury Association DISH: Diffuse idiopathic skeletal hyperostosis EDH: Epidural hematoma IQR: Interquartile range SCI: Spinal cord injury SLIC: Subaxial Cervical Spine Injury Classification TLISS: Thoracolumbar Injury Severity Scale Department of Neurologic Surgery, Mayo Clinic, Rochester, Minnesota, USA To whom correspondence should be addressed: Michelle J. Clarke, M.D. [E-mail: [email protected]] Citation: World Neurosurg. (2015). http://dx.doi.org/10.1016/j.wneu.2014.12.041 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2015 Elsevier Inc. All rights reserved.

INTRODUCTION Ankylosing spondylitis (AS) and diffuse idiopathic skeletal hyperostosis (DISH) are clinically distinct entities, but both ultimately result in spinal ankylosis. The ankylosed spine, whether secondary to AS or DISH, is very susceptible to fracture. The incidence of spine fractures in patients with AS is estimated to be 4 times greater than the general population (20). This greater incidence of spine fractures is largely due to the combination of rigidity and osteoporosis that develops in these patients (58). The ankylosed spine often consists of very long fused segments. Under stress, these segments act as lever arms exerting large forces at points in the spine that are already weakened because of poor bone quality (12, 14, 55). Spinal fractures result, which can be quite severe even after minor trauma (26). Because these fractures often follow minor trauma and occur in patients who typically

- OBJECTIVE:

To determine the rate and severity of in-hospital neurologic deterioration following vertebral fractures of spinal hyperostosis.

- METHODS:

A retrospective review of 92 fractures in 81 patients with diffuse idiopathic skeletal hyperostosis (42%) or ankylosing spondylitis (58%) was performed. Data on demographics, comorbidities, and fracture and treatment characteristics were recorded. Neurologic presentation and outcomes were categorized using American Spinal Injury Association grades and the modified Rankin Scale. Univariate and multivariate analyses were used to identify risk factors for neurologic deterioration or poor outcome (modified Rankin Scale 4e6).

- RESULTS:

Most fractures (66%) occurred after falls of standing height or less. Presentation was delayed in 41% of patients (median 7 days), and diagnosis was delayed in 21% (median 8 days). Most fractures were extension (60%) or distraction (78%) injuries involving all 3 spinal columns. Median Subaxial Cervical Spine Injury Classification and Thoracolumbar Injury Severity Scale scores were 6 (interquartile range 5e7) and 7 (interquartile range 6e8), respectively. Of patients, 62% underwent open operative fusion either as initial therapy or after failed conservative treatment, 20% had percutaneous instrumentation, and 27% were treated in an external orthosis (52% required open fusion). Neurologic deterioration after presentation occurred in 7 patients (8.6%); 5 of these patients deteriorated after surgical treatment, constituting a 7.6% surgical risk. The presenting American Spinal Injury Association grade and patient age predicted poor outcome at 1-year outcome (P < 0.001). Death occurred in 17 patients within 1 year of injury (23%).

- CONCLUSIONS:

Neurologic deterioration during the initial hospitalization after spinal fractures in the setting of diffuse idiopathic skeletal hyperostosis or ankylosing spondylitis is common, and 1-year mortality is high.

already have long-standing back pain, the diagnosis is often delayed. To make matters worse, these fractures can be missed on plain films and often require computed tomography for diagnosis (25). This delay in diagnosis can lead to later neurologic deterioration secondary to spinal instability or development of epidural hematoma (EDH). For these reasons, patients with AS and DISH who sustain spine fractures are at high risk of in-hospital neurologic deterioration. A low index of suspicion because of a seemingly minor mechanism and diagnostic imaging challenges may result

WORLD NEUROSURGERY - [-]: ---, MONTH 2015

in dangerously unstable fractures treated without appropriate precautions on admission. However, the degree of danger has been poorly quantified. To date, most of the literature describing patients with AS and DISH with spine fractures has been limited to case reports and smaller case series (26, 42, 58). A large case series and systematic review were published only more recently (9, 57). Our goal was to review our large series of patients with AS or DISH who presented with a spine fracture to investigate further delays in presentation and diagnosis and the rate and cause of secondary neurologic deterioration.

www.WORLDNEUROSURGERY.org

1

PEER-REVIEW REPORTS TERRY K. SCHIEFER ET AL.

MATERIALS AND METHODS This study was performed with the approval of the Mayo Clinic Institutional Review Board (10-006769). We performed a search of our electronic medical record using “ankylosing spondylitis” and “diffuse idiopathic skeletal hyperostosis” or “DISH” as key words to identify patients who presented to the Mayo Clinic between January 1994 and February 2012 with a traumatic vertebral fracture of an ankylosed vertebral segment. All patients either previously carried a diagnosis of or met diagnostic criteria for AS (50) or DISH (39). Data gathered included demographics, gender, age, delay in presentation or diagnosis, traumatic mechanism, fracture level or mechanism, instability, and comorbidities (associated injuries, cardiopulmonary disease, obesity, diabetes). Clinical presentation and initial radiographic studies were reviewed to calculate the Subaxial Cervical Spine Injury Classification (SLIC) system (48) and Thoracolumbar Injury Severity Scale (TLISS) (49). We recorded the details of initial and any subsequent bracing, casting, or surgical treatment. Functional and neurologic data were recorded at presentation, at death or hospital discharge, and at follow-up (3 and 12 months) and included American Spinal Injury Association (ASIA) class (34), bowel and bladder function, ambulatory status, pain, hospital disposition, and modified Rankin Scale score (5). For patients who underwent surgical fusion and for whom imaging studies were available at 1 year, evidence of radiographic arthrodesis was assessed. We performed a univariate analysis to identify risk factors for poor outcome and for neurologic deterioration between presentation and hospital discharge. Statistical analysis was performed using JMP 9.0.2 (SAS Institute Inc., Cary, North Carolina, USA).

RESULTS Study Population Between 1994 and 2012, 81 patients presented to our institution with 92 vertebral fractures involving an area of spinal ankylosis. Fractures occurred in the setting of AS (57.6%, n ¼ 53) and DISH (42.4%, n ¼ 39). Patients were predominantly male (81.5%). Baseline characteristics are provided in Table 1. Patients with fractures associated

2

www.SCIENCEDIRECT.com

NEUROLOGIC DETERIORATION AFTER FRACTURES OF ANKYLOSED SPINE

with DISH were older (P < 0.01) and more frequently had preexisting cardiac disease (P < 0.01), diabetes mellitus (P < 0.005), and other traumatic injuries (P < 0.05) compared with patients with AS (Table 2). Rates of closed head injury, underlying pulmonary disease (generally restrictive), and obesity were similar between the 2 groups. Although 3 patients each presented with 2 distinct spinal fractures and another presented with 3 fractures, for the purposes of the analysis, each fracture was treated as an individual entity. After being successfully treated for a spinal fracture earlier in the study period, 4 patients later presented with a second, unrelated traumatic spinal fracture. Between hospital discharge and 3 months after injury, 10 patients (10.9%) were lost to follow-up, and an additional 8 patients were lost to follow-up before the 1-year mark from injury (19.6%). Presentation and Diagnosis Of patient with fractures (ie individual patients since some patients have more than one fracture) (n ¼ 38), >40% presented in a delayed fashion after the index event thought to be the cause of the injury. For patients who presented in a delayed fashion, the median delay was 7 days (interquartile range [IQR] 2e15 days) after the initial injury. Not only was presentation to medical attention delayed, but diagnosis also was delayed in 20% of the patient cohort (n ¼ 19). After coming to medical attention, whether at our institution or elsewhere, the diagnosis was delayed for a median of 8 days (IQR 3e23 days) in these 19 patients. Trauma Mechanism Most fractures occurred in the setting of relatively minor trauma. Of patientfractures, 66% (n ¼ 56) occurred as a result of falls from either a standing height or less or the patient could not identify any recent traumatic injury, which occurred in 2 patients. Patient-fractures occurred after a fall from greater than a standing height in 9% (n ¼ 8). Patient-fractures resulted from motor vehicle collisions in 25% (n ¼ 21); approximately half of collisions occurred at less than highway speeds (n ¼ 10), and half were high-speed accidents (n ¼ 11). Fracture Characteristics Fractures in this study were identified throughout the spinal axis, although fractures associated with DISH tended to be

Table 1. Baseline Demographics for 92 Fractures in 81 Patients with Spinal Ankylosis Characteristic Male gender

75 (81.5%)

Median age, years (range)

75 (59e82)

Ankylosis syndrome DISH

39 (42.4%)

Ankylosing spondylitis

53 (57.6%)

Presentation delay Median delay, days (IQR) Diagnostic delay Median delay, days (IQR)

38 (41.3%) 7 (2e15) 19 (20.7%) 8 (3e23)

Trauma mechanism (n ¼ 89) Fall of standing height or less, or unknown

56 (65.9%)

Fall of more than standing height

8 (9.4%)

Low-speed MVC

10 (11.8%)

High-speed MVC

11 (12.9%)

Fracture level involved Occipitocervical junction Cervical spine Cervicothoracic junction

1 (1.1%) 30 (32.7%) 5 (5.4%)

Thoracic spine

45 (48.9%)

Thoracolumbar junction

14 (15.2%)

Lumbosacral spine

20 (21.7%)

Columns involved (n ¼ 90) Anterior

86 (95.6%)

Middle

84 (93.3%)

Posterior

80 (88.9%)

Injury involving 2 columns

84 (93.3%)

Injury involving 3 columns

76 (84.4%)

Fracture mechanism (n ¼ 89) Axial load

8 (9%)

Flexion

30 (33.7%)

Distraction

69 (77.5%)

Extension

53 (59.6%)

Rotation

3 (3.4%)

DISH, diffuse idiopathic skeletal hyperostosis; IQR, interquartile range; MVC, motor vehicle collision.

more common in the thoracic spine or thoracolumbar junction (87.2%), whereas fractures associated with AS were better

WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2014.12.041

PEER-REVIEW REPORTS TERRY K. SCHIEFER ET AL.

NEUROLOGIC DETERIORATION AFTER FRACTURES OF ANKYLOSED SPINE

Table 2. Principal Differences Between Patients with Ankylosing Spondylitis and Diffuse Idiopathic Skeletal Hyperostosis AS (n [ 52)

DISH (n [ 39)

70 (55e81)

77 (68e84)

9 (17%)

17 (43.6%)

Diabetes

7 (13.2%)

16 (41.0%)

Polytrauma*

7 (13.2%)

12 (30.8%)

Closed head injury

2 (3.8%)

3 (7.7%)

Variable Age, years (IQR)* Comorbidity Cardiac diseasey z

Lung disease Overweight/obese

7 (13.2%)

4 (10.3%)

18 (38.3%)

20 (58.8%)

27.6 (24.2e33.1) (n ¼ 47)

30.8 (29e39) (n ¼ 34)

SLIC score, median (IQR)z

5 (4e6) (n ¼ 23)

9 (6e9) (n ¼ 7)

TLISS score, median (IQR)

7 (6e8) (n ¼ 27)

7 (6e9) (n ¼ 31)

Median BMI (IQR)

Disposition from hospital Home (with or without assistance)* Acute rehabilitation

y

Skilled facility Deceased

17 (32%)

5 (12.8%)

4 (7.5%)

10 (25.6%)

25 (47.1%)

22 (56.4%)

6 (11.3%)

1 (2.6%)

AS, ankylosing spondylitis; DISH, diffuse idiopathic skeletal hyperostosis; IQR, interquartile range; BMI, body mass index; SLIC, Subaxial Cervical Spine Injury Classification; TLISS, Thoracolumbar Injury Severity Scale. *P < 0.05. yP < 0.01. zP < 0.005.

distributed across the cervical, thoracic, and lumbar spine (Table 1). Most could be classified as 3-column (84.4%) injuries (13). Fractures that were not 3-column injuries involved areas of the spine that were not completely ankylosed. For fractures with available multiplanar imaging (n ¼ 84), the fracture morphology most commonly involved distraction (77.5%), extension (59.%), or flexion (33.7%) moments. The most common fracture observed involved extension and distraction and occurred in the cervical, thoracic, and lumbar spine (Figure 1). Among patients with subaxial cervical spine fractures (n ¼ 30), the median SLIC score was 6 (IQR 5e7). SLIC scores were significantly higher in patients with DISH than in patients with AS (P < 0.005) (Table 2). Among patients with thoracolumbar spine fractures, the median TLISS score was 7 (IQR 6e9).

included percutaneous instrumentation (21.7%), posterior instrumented fusion (51.1%) with or without decompression, circumferential decompression and fusion (4.4%), or anterior decompression and fusion (6.6%). Patients were instrumented over a median of 7 levels (IQR 6e9). More than one-quarter of the cohort underwent external immobilization as the primary treatment (27.1%, n ¼ 25). However, 13 of those patients (52%) required a subsequent surgical procedure; the other 12 healed successfully. Of the 13 patients who failed conservative treatment alone, no significant difference in the SLIC or TLISS scores were noted. Radiographic arthrodesis was confirmed in 56 of 59 patients who completed 1-year postinjury radiographic follow-up, although 10 of those patients healed as a result of an operation performed for a failed course of bracing.

Treatment Most patients underwent a surgical procedure for stabilization of the fracture that

Neurologic Status Most patients were neurologically well at the time of presentation (Table 3), with

WORLD NEUROSURGERY - [-]: ---, MONTH 2015

79% (n ¼ 73) presenting neurologically intact as an ASIA grade E. Although more than one-third of patients were nonambulatory at the time of presentation, likely owing to pain, only 5.4% (n ¼ 5) presented with a complete spinal cord injury (SCI) (ASIA A), and 15.3% (n ¼ 14) presented with an incomplete SCI (ASIA B, C, or D). After presentation, 7 patients demonstrated neurologic deterioration. One patient experienced neurologic deterioration because the fracture was not diagnosed on presentation, and he deteriorated after being discharged from the hospital. One patient experienced neurologic deterioration after traction and closed reduction. There were 5 patients who deteriorated after surgical treatment, constituting a 7.6% surgical risk. Of these 7 patients, 6 patients progressed to a complete SCI (ASIA A), from ASIA B (n ¼ 1), ASIA D (n ¼ 1), or ASIA E (n ¼ 4) status at presentation. The remaining patient worsened to ASIA C from ASIA D after percutaneous instrumentation. From the rest of our cohort, 3 patients improved 1 ASIA grade, and a fourth patient improved to ASIA E after he presented with a severe central cord syndrome. ASIA status remained relatively constant throughout the remainder of the follow-up period (Table 3). Mortality and Outcome In-hospital mortality was 7.6% (n ¼ 7), increasing to 17.1% (n ¼ 14) at 3 months and 23% (n ¼ 17) at 1 year after injury. About 40% of the cohort were discharged from the hospital to either home or acute rehabilitation and then home, with the remainder of patients requiring a skilled nursing facility. The fraction of patients at each follow-up point with a good functional outcome status, defined as a modified Rankin Scale score of 0e3, increased during the follow-up period from 24% at the time of hospital discharge to 52.4% at 3 months and 58.1% at 1 year. DISCUSSION Patients with ankylosed spines such as seen in AS and DISH are at high risk of severe fractures and delayed neurologic deficit after injury. AS is an inflammatory seronegative spondyloarthropathy (1, 3) most commonly affecting men in the third and fourth decades of life (31); it has an

www.WORLDNEUROSURGERY.org

3

PEER-REVIEW REPORTS TERRY K. SCHIEFER ET AL.

NEUROLOGIC DETERIORATION AFTER FRACTURES OF ANKYLOSED SPINE

Figure 1. Sagittal reconstructions of computed tomography scans demonstrating extension distraction fractures of the cervical (A), thoracic (B), and lumbar (C) spines of 3 different patients with ankylosed spines.

estimated prevalence of 0.4%e1.4% of the population (7). Based on the modified New York criteria for AS, a definite diagnosis can be made if the radiologic criterion is associated with at least 1 clinical criterion (50). Clinical criteria include low back pain and stiffness improved with exercise but not with rest, restricted motion of the lumbar spine, and restriction of chest expansion. The radiologic criterion is grade 2 or greater sacroiliitis bilaterally or grade 3e4 unilaterally. Other radiographic features include progressive ankylosis of the spine with involvement of the disc space and the formation of bridging syndesmophytes leading to the classic “bamboo spine” appearance (Figure 2A). Despite the propensity to produce ectopic bone, 60% of these patients may have osteoporosis (4). Approximately 90% of patients with AS are HLA-B27 positive (19). DISH is a noninflammatory systemic disease characterized by calcification and ossification of soft tissues, primarily ligaments and entheses (36). The condition was originally described by Forestier and Rotes-Querol (21) more than 60 years ago as senile ankylosing hyperostosis and classically involves the axial skeleton, particularly the thoracic spine. Utsinger et al. (47) coined the term DISH in 1976 because of the diffuse peripheral ossification characteristic of the disease. Prevalence of the disease is estimated to be 27.3% in elderly white men (29). The

4

www.SCIENCEDIRECT.com

diagnosis is based on radiographic findings in the thoracic spine established by Resnick and Niwayama (39). These include flowing ossification extending over at least 4 contiguous vertebrae, relative preservation of intervertebral disc height in relation to age, and absence of apophyseal joint ankylosis or sacroiliac changes (Figure 2B). Although no clear etiology has been established, there is an association with diabetes mellitus, obesity, and advanced age (30). DISH is not associated with HLA-B27 (36). Owing to the nature of the ankylosed spine, patients with AS or DISH are more susceptible to spinal fractures than the general population. In a large epidemiologic study conducted in Canada over 2 decades, Pirouzmand (38) showed that of 12,192 trauma admissions in the general population, spine fractures were present in 23.2%. Although such large-scale studies have not been conducted in patients with ankylosing spinal disorders, Vosse et al. (52) performed a nested case control study that demonstrated increased risk of vertebral fractures in patients with AS with an odds ratio of 3.26 (95% confidence interval ¼ 1.51e7.02). This study suggests that spinal fractures could be found in 75% of traumas involving patients with an ankylosed spine. In patients with AS or DISH, these fractures tend to occur after minor trauma more frequently than in the general population. In our series, ground-level falls accounted

for 66% of spinal fractures. Similarly, Whang et al. (58) reported falls as the most common mechanism in their series at 76.7%. Smaller AS and DISH series report similar findings (24, 37, 46, 51). In contrast, estimates of spine fractures occurring from ground-level falls in the general population range from 12.9% (38) to 20% (32). The disproportionate number of spinal fractures resulting from minimal trauma in patients with an ankylosed spine compared with the general population further underscores how fragile and brittle the spine is in these patients. Spine fractures in patients with AS or DISH also tend to be more biomechanically severe and cause more frequent neurologic injury than spine fractures that occur in the general population. In the general population, simple compression fractures represent the most common traumatic fracture (32). However, in patients with an ankylosed spine, the spine is typically fused at the site of the fracture. Fractures tend to cross all 3 spinal columns and result in a highly unstable injury. Caron et al. (9) found 3-column fractures in all 112 patients in their series with hyperextension as the most common mechanism (89%). In our series, 3-column extension distraction fractures were the most commonly observed fracture pattern, found in 66% of patients, which was lower than expected. Emphasizing the severity of these fractures, we found the median SLIC score was 6 (IQR 5e7), and the TLISS score was 7 (IQR 6e8). In another series comprising patients with AS or DISH, the average SLIC or TLISS score was 5.1 (58). The median SLIC score in the patients with DISH in our series (9 [IQR 6e9]) was significantly higher than the median score in the patients with AS (5 [IQR 4e6]). For both classification systems, scores >4 are considered severe, and operative treatment is recommended (48, 49). Not surprisingly given the increased frequency of severe unstable fractures, SCIs are also more common in patients with AS or DISH compared with the general population. Although our study showed SCIs to be present in only 19% of patients, Caron et al. (9) reported SCIs in 58% of patients. Similarly, Whang et al. (58) reported that 60% of the patients in their series had SCI. In a

WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2014.12.041

PEER-REVIEW REPORTS TERRY K. SCHIEFER ET AL.

NEUROLOGIC DETERIORATION AFTER FRACTURES OF ANKYLOSED SPINE

Table 3. Neurologic Status, Follow-Up, and Functional Outcomes Presentation

Discharge or Death (n [ 92)

3 Months (n [ 69)

1 Year (n [ 57)

A

5 (5.4%)

9 (9.8%)

7 (10.1%)

5 (8.8%)

B

3 (3.3%)

1 (1.1%)

0 (0%)

0 (0%)

C

2 (2.2%)

2 (2.2%)

1 (1.4%)

1 (1.8%)

D

9 (9.8%)

8 (8.7%)

8 (11.6%)

4 (7.0%)

Variable ASIA class

73 (79.3%)

72 (78.3%)

53 (76.8%)

47 (82.5%)

Bowel dysfunction

E

12 (13%)

15 (16.3%)

11 (15.9%)

8 (14%)

Bladder dysfunction

13 (14.1%)

15 (16.3%)

12 (17.4%)

9 (15.8%)

Independent

36 (39.1%)

13 (14.1%)

31 (44.9%)

32 (56.1%)

Assistance required

24 (26.1%)

52 (56.5%)

26 (37.7%)

18 (31.6%)

Nonambulatory

32 (34.8%)

27 (29.3%)

12 (17.4%)

7 (12.2%)

Ambulatory status

Pain level None

2 (2.2%)

8 (9.6%)

32 (46.4%)

36 (63.2%)

Nonlimiting pain

26 (28.3%)

57 (68.6%)

24 (34.8%)

15 (26.3%)

Limiting pain

64 (69.6%)

18 (21.8%)

13 (18.8%)

6 (10.5%)

Follow-up available

e

92 (100%)

82 (89.1%)

74 (80.4%)

0 (0%)

10 (10.9%)

18 (19.6%)

1 (1.1%)

6 (7.3%)

15 (20.3%)

1

3 (3.3%)

21 (25.6%)

24 (32.4%)

2

5 (5.4%)

10 (12.2%)

3

13 (14.1%)

7 (8.5%)

4

49 (53.3%)

18 (22%)

5

14 (15.2%)

7 (8.5%)

7 (7.6%)

14 (17.1%)

Lost to follow-up mRS 0

e

6 (deceased) Median follow-up (IQR), days

e

13 (7e28)

99 (74e120)

Poor outcome (mRS 4e6)

e

70 (76%)

39 (47.6%)

0 (0%) 4 (5.4%) 9 (12.2%) 5 (6.8%) 17 (23%) 368 (236e426) 31 (41.9%)

All variables were similar between the patients with ankylosing spondylitis and diffuse idiopathic skeletal hyperostosis. Median time to death or last follow-up was 216 days (IQR 44e380 days). ASIA, American Spinal Injury Association; IQR, interquartile range; mRS, modified Rankin Scale.

retrospective cohort series, Westerveld et al. (56) found 57.1% of patients with AS and 30% of patients with DISH were admitted with neurologic deficit compared with 12.6% of control subjects. In large trauma series of the general population, SCIs were found in 23.3%e 50.2% of patients who had an associated traumatic spinal fracture (28, 32, 38). Many patients with AS or DISH have a significant delay in presentation to medical

attention and the time to diagnosis, which may result in delayed neurologic injury. In their systematic review of the literature, Westerveld et al. (57) also found that delays in presentation and diagnosis were common. Of the patients in our series, 40% had a delayed presentation after trauma, with a median time of 7 days (IQR 2e15). There are 2 possible reasons for this delay. First, many patients with AS or DISH have chronic back pain, and new

WORLD NEUROSURGERY - [-]: ---, MONTH 2015

posttraumatic pain may not be recognized as pathologic. Second, as mentioned earlier, the traumatic events that lead to these fractures are often minor and could be thought to be trivial. In our series, 2 patients did not even report a preceding traumatic event. In either case, presentation may be delayed until pain intensifies or there is a neurologic decline. Even after presenting to medical attention, many of these patients experience a significant delay in diagnosis. In our series, 20% of patients experienced a delay in diagnosis after presentation for a median of 8 days (IQR 3e23). There are likely several causes for these delays in diagnosis. First, clinicians may have a very low suspicion of serious injury given the minor trauma associated with most of these fractures and delay obtaining imaging studies. Second, patients may not yet have a diagnosis of AS or DISH alerting clinicians to the fragility of their spine and the need for imaging studies even after trivial trauma. Third, x-ray imaging of these patients can be very difficult to interpret given the extensive changes associated with AS and DISH, and fractures can be missed. Multiple studies of patients with fractures in the setting of spinal ankylosis cite many examples of fractures that were initially missed on plain x-rays and seen only later on computed tomography or magnetic resonance imaging (10, 22, 23, 27, 41, 53, 54). Because of the difficulty in interpreting plain films in these patients, advanced imaging techniques (computed tomography or magnetic resonance imaging) have been recommended as the screening tool of choice (20). It is recommended that the entire spine be imaged in these patients because many may have noncontiguous fractures that are asymptomatic (40). In our series, 7% of patients had noncontiguous fractures, including 1 patient in whom a second fracture was overlooked at admission. Both the delay in presentation and the delay in diagnosis can be very dangerous given the severity of most of these fractures and the subsequent risk of neurologic deterioration. In their series, Caron et al. (9) reported a 20% delay in diagnosis and an 81% likelihood of decline in patients in whom the diagnosis was delayed. In our series, of the 19 patients with diagnostic delays, 1 patient experienced a severe delayed neurologic

www.WORLDNEUROSURGERY.org

5

PEER-REVIEW REPORTS TERRY K. SCHIEFER ET AL.

NEUROLOGIC DETERIORATION AFTER FRACTURES OF ANKYLOSED SPINE

Figure 2. Sagittal reconstructions of computed tomography scans performed in patients with ankylosing spondylitis (A) and diffuse idiopathic skeletal hyperostosis (B).

deficit. He initially presented neurologically intact after being rearended in a low-speed motor vehicle accident. No imaging studies were performed, and he was discharged from the emergency department. He presented 6 days after the initial trauma with a complete SCI, and an unstable 3-column T4 fracture was diagnosed. Even after presentation and diagnosis, neurologic deterioration in patients with AS or DISH after spine trauma is common. Colterjohn and Bednar (11) performed a retrospective review of 281 patients with cervical spine fractures, of which 15 had secondary neurologic deterioration. Of these 15 patients, 40% had an ankylosing spinal disorder compared with 0.5% of the remainder of their cohort. They concluded that chronic multilevel spinal arthritis with ankylosis is a primary risk factor for secondary neurologic deterioration in patients with cervical spine injury. Other studies showed similar findings (18, 33, 52). In our series, we observed a secondary neurologic decline to a complete SCI (ASIA A) in 7 patients after presentation (8.6%), including the man described in the previous paragraph.

6

www.SCIENCEDIRECT.com

One patient demonstrated neurologic deterioration after traction and closed reduction of a C5-6 fracture and was found to have a protruding disc and spinal EDH. He was immediately taken to the operating room for decompression and fusion, but his neurologic status did not significantly improve. Deterioration occurred in 5 patients after surgical fixation of fractures, which constitutes 7.6% of patients who underwent surgical treatment. In 2 patients, deterioration occurred secondary to subluxation after operative fixation and EDH compression. One of these patients was morbidly obese and woke up without lower extremity function immediately after surgery despite preserved intraoperative monitoring. The other patient deteriorated 3 days after surgery. Imaging studies showed subluxation and compression of the neural elements. Both patients returned to the operating room for emergent decompression, but no neurologic improvement occurred thereafter in either patient. One patient fell after percutaneous fixation of a fracture resulting in complete SCI and was moved to comfort cares. Of the 2 remaining patients, 1 patient had intraoperative loss of somatosensory evoked

potentials that was thought to be due to an intraoperative vascular injury, and no obvious cause was found for deterioration in the other patient. Other series report rates of secondary neurologic deterioration between 0% and 43% (9, 45, 46, 58). Delay in diagnosis has been cited as a primary risk for secondary neurologic deterioration (9). The formation of a spinal EDH has also been identified as a significant risk factor in patients with AS and patients with DISH (2, 8, 9, 16, 17, 59). Several strategies may help prevent secondary neurologic decline. First, removal of the spine board in patients with a severe fixed kyphotic deformity has been advocated because a forced flat position may further compromise the spinal cord in these patients (35, 45). Second, the entire spine should be screened so that noncontiguous fractures are not missed. Third, intraoperative monitoring with motor evoked potentials and somatosensory evoked potentials should be considered because of the highly unstable nature of these fractures and risk for EDH formation. Finally, spinal cord decompression should be considered at the time of internal fixation when neurologic deficit is present because of the probable presence of an EDH (9, 41). The inherent instability of the fractured ankylosed spine should be considered when choosing operative versus nonoperative treatment for these fractures. In our series, bracing was chosen as the primary method of treatment in 25 patients (27.1%). Of these, 13 patients (52%) failed nonoperative treatment and required subsequent surgery. Most of the cases of failed nonoperative treatment were due to progressive deformity, persistent pain, and fracture nonunion. One patient developed a new radiculopathy. No significant differences in presenting SLIC or TLISS scores of patients were found compared with patients who did not fail conservative treatment alone. Other groups reported similar disappointing results with conservative management in this patient population. In their series of 7 patients with AS, Taggard and Traynelis (44) employed trials of bracing in 2 patients, and both trials failed after 3e4 months. Paley et al. (37) published their series of 8 patients with DISH in which 5 patients were placed in halo immobilization for primary treatment; 80% of these patients failed the

WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2014.12.041

PEER-REVIEW REPORTS TERRY K. SCHIEFER ET AL.

conservative treatment. Bracing is also associated with higher reported rates of mortality (9); this may be due to the fact that patients with significant medical comorbidities may not tolerate open surgery, and bracing is used as an alternative or palliative approach. When bracing of these fractures is considered, it should be in patients who maintain good spinal alignment, have little or no compression of the neural elements, and are reliable with regard to close clinical follow-up. In addition to the poor results of bracing, high SLIC and TLISS scores and the risk of neurologic deterioration without decompression argue for the surgical treatment of most spinal fractures in patients with AS or DISH. In their systematic review of the literature, Westerveld et al. (57) suggested that surgical treatment is favorable over conservative management because of lower complication and mortality rates and higher rates of neurologic improvement. In our series, most patients underwent a surgical procedure for fracture stabilization (84%), including patients who failed conservative treatment with a brace. Surgical treatment included percutaneous instrumentation (21.7%), open posterior instrumented fusion with or without decompressive laminectomy (51.1%), circumferential decompression and fusion (4.4%), and anterior decompression and fusion (6.6%). Most authors tend to recommend posterior approaches with instrumentation extending 2e4 segments above and below the fracture because of the inherent fragility of the ankylosed spine (9, 41, 43, 58). Reported fusion rates with this approach have been consistently high, reaching 100% in some series (9, 41, 58). In patients with AS and cervical spine fractures, Einsiedel et al. (15) reported a 50% failure rate with anterior fusion only and cautioned against an anterior standalone approach. In our patient cohort, 59 patients who had undergone surgical fixation were available for radiographic follow-up at 1 year. Among these patients, there was a 95% rate of arthrodesis. Morbidity and mortality also tend to be worse in patients with AS or DISH compared with the general population. In the general population, Pirouzmand (38) reported a 16.7% case-fatality rate after traumatic spinal fracture. In contrast, in a series of patients with ankylosed spines,

NEUROLOGIC DETERIORATION AFTER FRACTURES OF ANKYLOSED SPINE

Caron et al. (9) reported that 84% of patients experienced at least 1 medical or surgical adverse event, and 32% were dead at 1 year. In our series, mortality at 1 year was 23%, although this may be an underestimate secondary to losses to follow-up. Most of the fatalities in the general population series are due to highvelocity trauma, whereas most mortalities in patients with AS or DISH occur after minor trauma. The cause of this trend is likely multifactorial. First, patients with AS or DISH present at a significantly older age than the general population. The mean age in our series was 75 years compared with 36e50 years in the general population (28, 32, 38). Second, patients with AS or DISH have higher rates of neurologic injury. As discussed previously, the rates of associated SCIs in patients with an ankylosed spine are higher than the rates found in large epidemiologic studies (9, 58). We performed a univariate analysis to look for predictors of poor outcome (modified Rankin Scale score 0e3) in our series. Of age, polytrauma, obesity, presenting ASIA grade, and bracing versus fusion, both presenting ASIA grade (P ¼ 0.0002) and age (P ¼ 0.011) were significant predictors of a poor outcome. In multivariate analysis, age remained predictive of poor outcome (odds ratio ¼ 1.76 per decade of age, 95% confidence interval ¼ 1.08e2.5) as did presenting ASIA grade (odds ratio ¼ 2.4 per ASIA grade, 95% confidence interval ¼ 1.4e5.7). Caron et al. (9) also found that age was the primary predictor of mortality in their series. Finally, patients with AS or DISH have high rates of comorbidities (6, 9, 57). In our series, 28% of patients had cardiac disease, 25% had diabetes, and 47% were overweight or obese. Although comorbidities were prevalent in patients with AS and patients with DISH in our series, there were significantly more patients with cardiac disease, diabetes, and other injuries in the DISH group (Table 2). Westerveld et al. (57) postulated that this trend is due to the fact that DISH is a rapidly increasing condition in affluent societies. The overall goal of this article was not only to add our experience of spine fractures in the setting of AS or DISH to the literature but also to help clinicians counsel patients and their families. We think it is important for clinicians to emphasize that compared with the general

WORLD NEUROSURGERY - [-]: ---, MONTH 2015

population, patients with an ankylosing spine disorder are more susceptible to spinal fractures even after minor trauma, have more severe spinal fractures, and are more likely to have delays in presentation and diagnosis. After presentation, these patients continue to be at higher risk for secondary neurologic deterioration even after treatment, are more likely to fail conservative treatment, and have higher rates of morbidity and mortality. In our experience, 25% of patients had a neurologic injury (whether due to the initial trauma or secondary neurologic decline), there was an 8% in-hospital mortality rate, and mortality at 1 year was 23%. It is hoped that this information will prepare clinicians better to manage these complex patients and inform patients and their families of what can be expected. CONCLUSIONS Patients who present with a traumatic spinal fracture in the setting of an ankylosed spine as a result of either AS or DISH represent a challenge to the spine surgeon. Because of the fragile nature of the ankylosed spine and the long lever arms the fused segments create, even low-energy mechanisms can result in very severe 3-column unstable spine fractures. Spine surgeons must have a high suspicion and low threshold to obtain multiplanar imaging studies when evaluating these patients to avoid delays in diagnosis that could potentially be devastating. These patients often fail nonoperative treatment and require surgical intervention with large fusion constructs to treat the fracture definitively. Neurologic deterioration after presentation is common. When counseling patients and their families, it is important to communicate that mortality and poor outcomes are more common in patients with an ankylosed spine than in the general population. REFERENCES 1. Akkoc N, Khan MA: Overestimation of the prevalence of ankylosing spondylitis in the Berlin study: comment on the article by Braun et al. Arthritis Rheum 52:4048-4049 [author reply 40494050], 2005. 2. Aoki Y, Yamagata M, Ikeda Y, Nakajima F, Nakajima A, Ohtori S, Inaoka T, Takahashi K: Failure of conservative treatment for thoracic spine fractures in ankylosing spondylitis: delayed neurological defecit due to spinal epidural hematoma. Mod Rheumatol 23:1008-1012, 2013.

www.WORLDNEUROSURGERY.org

7

PEER-REVIEW REPORTS TERRY K. SCHIEFER ET AL.

3. Bakland G, Nossent HC, Gran JT: Incidence and prevalence of ankylosing spondylitis in Northern Norway. Arthritis Rheum 53:850-855, 2005. 4. Bessant R, Keat A: How should clinicians manage osteoporosis in ankylosing spondylitis? J Rheumatol 29:1511-1519, 2002. 5. Bonita R, Beaglehole R: Recovery of motor function after stroke. Stroke 19:1497-1500, 1988. 6. Bransford R, Koller H, Caron T, Zenner J, Hitzl W, Tomasino A, Mayer M: Cervical spine trauma in diffuse idiopathic skeletal hyperostosis: injury characteristics and outcome with surgical treatment. Spine (Phila Pa 1976) 37:1923-1932, 2012. 7. Braun J, Sieper J: Ankylosing spondylitis. Lancet 369:1379-1390, 2007. 8. Campagna R, Pessis E, Feydy A, Guerini H, Thevenin F, Chevrot A, Drape JL: Fractures of the ankylosed spine: MDCT and MRI with emphasis on individual anatomic spinal structures. AJR Am J Roentgenol 192:987-995, 2009.

NEUROLOGIC DETERIORATION AFTER FRACTURES OF ANKYLOSED SPINE

18. Farmer J, Vaccaro A, Albert TJ, Malone S, Balderston RA, Cotler JM: Neurologic deterioration after cervical spinal cord injury. J Spinal Disord 11:192-196, 1998. 19. Feldtkeller E, Khan MA, van der Heijde D, van der Linden S, Braun J: Age at disease onset and diagnosis delay in HLA-B27 negative vs. positive patients with ankylosing spondylitis. Rheumatol Int 23:61-66, 2003. 20. Finkelstein JA, Chapman JR, Mirza S: Occult vertebral fractures in ankylosing spondylitis. Spinal Cord 37:444-447, 1999. 21. Forestier J, Rotes-Querol J: Senile ankylosing hyperostosis of the spine. Ann Rheum Dis 9:321-330, 1950. 22. Harrop JS, Sharan A, Anderson G, Hillibrand AS, Albert TJ, Flanders A, Vaccaro AR: Failure of standard imaging to detect a cervical fracture in a patient with ankylosing spondylitis. Spine 30: E417-E419, 2005.

9. Caron T, Bransford R, Nguyen Q, Agel J, Chapman J, Bellabarba C: Spine fractures in patients with ankylosing spinal disorders. Spine 35:E458-E464, 2010.

23. Hendrix RW, Melany M, Miller F, Rogers LF: Fracture of the spine in patients with ankylosis due to diffuse skeletal hyperostosis: clinical and imaging findings. AJR Am J Roentgenol 162: 899-904, 1994.

10. Charles Y, Buy X, Gangi A, Steib J: Fracture in ankylosing spondylitis after minor trauma: radiological pitfalls and treatment by percutaneous instrumentation. A case report. Orthop Traumatol Surg Res 99:115-119, 2013.

24. Hitchon PW, From AM, Brenton MD, Glaser JA, Torner JC: Fractures of the thoracolumbar spine complicating ankylosing spondylitis. J Neurosurg 97:218-222, 2002.

11. Colterjohn NR, Bednar DA: Identifiable risk factors for secondary neurologic deterioration in the cervical spine-injured patient. Spine 20:2293-2297, 1995. 12. Davey-Ranasinghe N, Deodhar A: Osteoporosis and vertebral fractures in ankylosing spondylitis. Curr Opin Rheumatol 25:509-516, 2013. 13. Denis F: The three column spine and its significance in the classification of acute thoracolumbar spinal injuries. Spine 8:817-831, 1983. 14. Diederichs G, Engelken F, Marshall L, Peteres K, Black D, Issever A, Barrett-Connor E, Orwoll E, Hamm B, Link T; Osteoporotic Fractures in Men Research Group: Diffuse idiopathic skeletal hyperostosis (DISH): relation to vertebral fractures and bone density. Osteoporos Int 22:1789-1797, 2011. 15. Einsiedel T, Schmelz A, Arand M, Wilke H-J, Gebhard F, Hartwig E, Kramer M, Neugebauer R, Kinzl L, Schultheiss M: Injuries of the cervical spine in patients with ankylosing spondylitis: experience at two trauma centers. J Neurosurg Spine 5:33-45, 2006. 16. Elgafy H, Bransford R, Chapman J: Epidural hematoma associated with occult fracture in ankylosing spondylitis patient: a case report and review of the literature. J Spinal Disord Tech 24:469-473, 2011. 17. Farhat SM, Schneider RC, Gray JM: Traumatic spinal extradural hematoma associated with cervical fractures in rheumatoid spondylitis. J Trauma 13:591-599, 1973.

8

www.SCIENCEDIRECT.com

25. Hunter T, Dubo H: Spinal fractures complicating ankylosing spondylitis. Ann Intern Med 88: 546-549, 1978. 26. Hunter T, Forster B, Dvorak M: Ankylosed spines are prone to fracture. Can Fam Physician 41: 1213-1216, 1995. 27. Jacobs W, Fehlings M: Ankylosing sponduylitis and spinal cord injury: origin, incidence, management, and avoidance. Neurosurg Focus 24:E12, 2008. 28. Kattail D, Furlan JC, Fehlings MG: Epidemiology and clinical outcomes of acute spine trauma and spinal cord injury: experience from a specialized spine trauma center in Canada in comparison with a large national registry. J Trauma 67: 936-943, 2009.

33. Mathews M, Bolesta M: Treatment of spinal fractures in ankylosing spondylitis. Orthopedics 36: 1203-1208, 2013. 34. Maynard FM Jr, Bracken MB, Creasey G, Ditunno JF Jr, Donovan WH, Ducker TB, Garber SL, Marino RJ, Stover SL, Tator CH, Waters RL, Wilberger JE, Young W: International Standards for Neurological and Functional Classification of Spinal Cord Injury. American Spinal Injury Association. Spinal Cord 35:266-274, 1997. 35. Moreau AP, Willcox N, Brown MF: Immobilisation of spinal fractures in patients with ankylosing spondylitis. Two case reports. Injury 34:372-373, 2003. 36. Olivieri I, D’Angelo S, Palazzi C, Padula A, Mader R, Khan MA: Diffuse idiopathic skeletal hyperostosis: differentiation from ankylosing spondylitis. Curr Rheumatol Rep 11:321-328, 2009. 37. Paley D, Schwartz M, Cooper P, Harris WR, Levine AM: Fractures of the spine in diffuse idiopathic skeletal hyperostosis. Clin Orthop Relat Res 267:22-32, 1991. 38. Pirouzmand F: Epidemiological trends of spine and spinal cord injuries in the largest Canadian adult trauma center from 1986 to 2006. J Neurosurg Spine 12:131-140, 2010. 39. Resnick D, Niwayama G: Radiographic and pathologic features of spinal involvement in diffuse idiopathic skeletal hyperostosis (DISH). Radiology 119:559-568, 1976. 40. Samartzis D, Anderson DG, Shen FH: Multiple and simultaneous spine fractures in ankylosing spondylitis: case report. Spine 30:E711-E715, 2005. 41. Sapkas G, Kateros K, Papadakis SA, Galanakos S, Brilakis E, Machairas G, Katonis P: Surgical outcome after spinal fractures in patients with ankylosing spondylitis. BMC Musculoskelet Disord 10:96, 2009. 42. Schoenfeld AJ, Harris MB, McGuire KJ, Warholic N, Wood KB, Bono CM: Mortality in elderly patients with hyperostotic disease of the cervical spine after fracture: an age- and sex-matched study. Spine 11:257-264, 2011. 43. Shen FH, Samartzis D: Surgical management of lower cervical spine fracture in ankylosing spondylitis. J Trauma 61:1005-1009, 2006.

29. Kiss C, O’Neill TW, Mituszova M, Szilagyi M, Poor G: The prevalence of diffuse idiopathic skeletal hyperostosis in a population-based study in Hungary. Scand J Rheumatol 31:226-229, 2002.

44. Taggard DA, Traynelis VC: Management of cervical spinal fractures in ankylosing spondylitis with posterior fixation. Spine 25:2035-2039, 2000.

30. Kiss C, Szilagyi M, Paksy A, Poor G: Risk factors for diffuse idiopathic skeletal hyperostosis: a casecontrol study. Rheumatology (Oxford) 41:27-30, 2002.

45. Thumbikat P, Hariharan RP, Ravichandran G, McClelland MR, Mathew KM: Spinal cord injury in patients with ankylosing spondylitis: a 10-year review. Spine 32:2989-2995, 2007.

31. Kubiak EN, Moskovich R, Errico TJ, Di Cesare PE: Orthopaedic management of ankylosing spondylitis. J Am Acad Orthop Surg 13:267-278, 2005.

46. Trent G, Armstrong GW, O’Neil J: Thoracolumbar fractures in ankylosing spondylitis. High-risk injuries. Clin Orthop Relat Res 227:61-66, 1988.

32. Leucht P, Fischer K, Muhr G, Mueller EJ: Epidemiology of traumatic spine fractures. Injury 40: 166-172, 2009.

47. Utsinger PD, Resnick D, Shapiro R: Diffuse skeletal abnormalities in Forestier disease. Arch Intern Med 136:763-768, 1976.

WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2014.12.041

PEER-REVIEW REPORTS TERRY K. SCHIEFER ET AL.

48. Vaccaro AR, Hulbert RJ, Patel AA, Fisher C, Dvorak M, Lehman RA, Anderson P, Harrop J, Oner FC, Arnold P, Fehlings M, Hedlund R, Madrazo I, Rechtine G, Aarabi B, Shainline M; Spine Trauma Study Group: The subaxial cervical spine injury classification system: a novel approach to recognize the importance of morphology, neurology, and integrity of the discoligamentous complex. Spine (Phila Pa 1976) 32: 2365-2374, 2007. 49. Vaccaro AR, Zeiller SC, Hulbert RJ, Anderson PA, Harris M, Hedlund R, Harrop J, Dvorak M, Wood K, Fehlings MG, Fisher C, Lehman RA, Anderson DG, Bono CM, Kuklo T, Oner FC: The thoracolumbar injury severity score: a proposed treatment algorithm. J Spinal Disord Tech 18: 209-215, 2005. 50. van der Linden S, Valkenburg HA, Cats A: Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria. Arthritis Rheum 27:361-368, 1984. 51. Vosse D, Feldtkeller E, Erlendsson J, Geusens P, van der Linden S: Clinical vertebral fractures in patients with ankylosing spondylitis. J Rheumatol 31:1981-1985, 2004. 52. Vosse D, Landewe R, van der Heijde D, van der Linden S, van Staa T-P, Geusens P: Ankylosing

NEUROLOGIC DETERIORATION AFTER FRACTURES OF ANKYLOSED SPINE

spondylitis and the risk of fracture: results from a large primary care-based nested case-control study. Ann Rheum Dis 68:1839-1842, 2009. 53. Waldman S, Brown C, Lopez de Heredia L, Hughes R: Diagnosis and managing spinal injury in patients with ankylosing spondylitis. J Emerg Med 44:e315-319, 2013.

58. Whang PG, Goldberg G, Lawrence JP, Hong J, Harrop JS, Anderson DG, Albert TJ, Vaccaro AR: The management of spinal injuries in patients with ankylosing spondylitis or diffuse idiopathic skeletal hyperostosis: a comparison of treatment methods and clinical outcomes. J Spinal Disord Tech 22:77-85, 2009.

54. Wang Y-F, Teng MM-H, Chang C-Y, Wu H-T, Wang S-T: Imaging manifestations of spinal fractures in ankylosing spondylitis. AJNR Am J Neuroradiol 26:2067-2076, 2005.

59. Wu CT, Lee ST: Spinal epidural hematoma and ankylosing spondylitis: case report and review of the literature. J Trauma 44:558-561, 1998.

55. van der Weijden MA, Claushuis T, Nazari T, Lems W, Dijkmans B, van der Horst-Bruinsma IE: High prevalence of low bone mineral density in patients within 10 years of onset of ankylosing spondylitis: a systematic review. Clin Rheumatol 31:1529-1535, 2012.

Conflict of interest statement: This article was made possible by Clinical Translational Science Award Grant Number UL1 TR000135 from the National Center for Advancing Translational Sciences, a component of the National Institutes of Health. The contents of this article are solely the responsibility of the authors and do not necessarily represent the official view of the National Institutes of Health.

56. Westerveld L, van Bemmel JC, Dhert W, Oner F, Verlaan J: Clinical outcome after traumatic spinal fractures in patients with ankylosing spinal disorders compared with control patients. Spine J 14: 729-740, 2014. 57. Westerveld LA, Verlaan JJ, Oner FC: Spinal fractures in patients with ankylosing spinal disorders: a systematic review of the literature on treatment, neurological status and complications. Eur Spine J 18:145-156, 2008.

WORLD NEUROSURGERY - [-]: ---, MONTH 2015

Received 10 April 2014; accepted 15 December 2014 Citation: World Neurosurg. (2015). http://dx.doi.org/10.1016/j.wneu.2014.12.041 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2015 Elsevier Inc. All rights reserved.

www.WORLDNEUROSURGERY.org

9