Among AS patients, SCI is a major complication regardless of spinal fracture or not. Our results showed that the SCI rate after spinal trauma in AS patients was 57.1% (60/105), while the SCI rate in cases also involving spinal fractures was 52.1% (49/94), which was lower than the average 67.2% rate reported in the largest meta-analysis,4 but higher than the 32.1% rate reported in the largest single-institution series.23 Lukasiewicz et al. reported a 21.2% rate of SCI in patients with AS and spinal fracture.16 Another two single-institution studies reported 58% and 19.7% rates.11,13 These inconsistent rates of SCI after spinal fractures in AS patients may be due to differences in medical referral standards, different severities of trauma, and different severities of AS.
The diagnoses of fractures in AS patients are frequently delayed, ranging from 17.1–65.4%4,11,23 as patients are known to have chronic pain even in the absence of trauma. Therefore, aggravating pain following a minor trauma may be overlooked. Due to alterations in bone density, radiologic assessment for fracture in AS patients could be difficult and easily misinterpreted, especially in cases involving fractures at the thoracic spine and thoracolumbar junction. In our study, the rate of delayed diagnosis for spinal fracture was 31.4% (33/105), with 69.7% (23/33) of those cases involving fractures between the T8 and L1 vertebrae. Six of the 33 cases were attributed to a doctor’s oversight, while the remaining cases were due to the patients’ delayed visits. Overseeing may expose these patients to high risk of delayed SCI. In our study, one-third of these 33 patients with a delayed diagnosis of spinal fracture developed delayed SCI. In the initial post-trauma period, 90.9% of these patients had axial pain such as neck or back pain, whereas 33.3% had limb numbness. The attending physician should be reminded of the highly possible devastating consequences of AS patients, even after a low-energy trauma. We suggest routine radiographic examination for all AS patients after trauma and additional computed tomography (CT) if axial pain progresses. Magnetic resonance imaging is a reasonable option for spinal cord injuries and to rule out occult fractures.12,24−26
Bamboo spine is a radiographic feature seen in AS that occurs as a result of vertebral body fusion by marginal syndesmophytes. The consequential radiographic appearance is radiopaque spicules that completely bridge the adjoining vertebral bodies. In our study, 74.3% (78/105) patients exhibited this feature. We also observed that these patients with existing bamboo spine had a higher spinal fracture rate than those without bamboo spine (p = 0.006, Table 4). However, the results showed no significant correlation between bamboo spine and SCI (p = 0.367, Table 4). This contradicted the finding that AS with spinal fracture was significantly related to SCI (p < 0.001). To clarify this, we analyzed the relationship between bamboo spine and spinal fracture in the 60 patients with SCI. Four of the 11 patients who had SCI without spinal fracture presented with bamboo spine, while 43 of the 49 patients with both SCI and spinal fracture had bamboo spine. A comparison of these two groups revealed a significant difference (p = 0.001, Table 2). Patients with non-complex compression fractures with the posterior ligamentous complex (PLC) intact (Orthopedic Trauma Association classification type A) had a lower rate of SCI than those with complex fractures, such as flexion/distraction type (B.1 and B.2, PLC disrupted), hyperextension type (B.3), or shear/rotation type (type C) fractures (p < 0.001). Subluxation or dislocation was also found to be a risk factor for the AS patients in developing SCI (p < 0.001). The severity of spinal structure disruptions is thus predictive of SCI in AS patients. SCI in cases involving a complex fracture is not only caused by the destruction resulting from the direct impact, but also by further compression from bone fragments, hematoma, or disk material.27 Thus, we hypothesize that AS patients with existing bamboo spine have a high probability of experiencing spinal fracture, but such a mild fracture will not necessarily be severe enough to cause SCI.
There are no universal guidelines for the management of spinal trauma in AS patients.26 Non-operative treatment has long been recommended for fractures of ankylosed spines.26,28 The management options include bed rest, skeletal traction, bracing, or immobilization with a halo-vest.29 Conservative methods are reasonable for non-displaced or minimally displaced fractures.28 However, under the inherent instability of these fractures and their high potential of acute displacement, catastrophic situation may occur.30 Therefore, surgical fixation with long segmental instrumentation combined with fusion is highly recommended.30 Furthermore, the compression of neurologic elements often requires surgical evacuation. Recent studies have demonstrated a trend for higher complication rates in non-operative patients. Patients with bed rest are associated with higher pulmonary complications29 and present a risk of neurological deterioration.3,4,29 Surgical stabilization usually includes anterior, posterior, or combined fixation, often accompanied by decompression with laminectomy and various osteotomy techniques for deformity correction.31,32 In our study, 3 patients with initial AIS A received close reduction and halo-jacket fixation and all 3 patients (100%) had complications: one experienced screw loosening, two developed pneumonia. In contrast, seven of 9 patients (77.8%) with initial AIS A had complication after surgery, suggesting a trend of lower complication rates in severe SCI.
All 14 patients with SEH developed SCI, thus making SEH an important predictive factor of SCI (p = 0.003). SEH occurred in 13.3% of the AS patients, which is much higher than in the rates ranged from 0.5–7.5% of the general population.33,34 As opposed to severe fractures that caused SCI immediately, SEH may bring about subacute SCI hours after the trauma. The mechanisms underpinning cervical SEH formation are not fully understood, but disruption of the posterior longitudinal ligament and spinal epidural vessel rupture may play an important role.34,35 Symptomatic SEH is thought to be a neurosurgical emergency. In the general population, better long-term neurological recovery was noted after early surgical intervention. However, no major case study has reported in AS patients. In our study, the investigated SCI patients received surgical treatment at 1.8 ± 3.2 days after trauma. Some of these patients had delayed surgical treatment due to delayed diagnosis, old age, higher co-morbidity, or poly-trauma requiring other treatments.
None of the patients who had complete SCI at admission showed improvement in AIS grade after 2 year, as compared to the 58.3% (35/60) of patients in the incomplete SCI group who showed improvement. Overall, AS patients with incomplete SCI had better long-term neurological recovery. Complication rates were significantly higher (p = 0.001) in patients with SCI. The 4 patients who died within 1 year were significantly older (with an average age of 69.2 years). This echoes the findings of higher mortality in older AS patients after spinal fracture in previous studies.11,13 In our study, although these AS patients with the different complexity of fracture with different degree of SCI, more than 53% patients get the benefit of long-term neurological recovery after surgical treatment.