Ankylosing spinal disorder (ASD) patients are at a greater risk for spinal fractures due to osteoporosis and rigidity of the spinal column. These fractures are associated with a high risk of neurologic compromise resulting from delayed or missed diagnoses. Although computed tomography (CT) is usually the initial imaging modality, magnetic resonance imaging (MRI) has been proposed as mandatory to help identify spinal injuries in ASD patients with unexplained neck or back pain or known injuries to help identify noncontiguous fractures. However, some studies have also shown that neurological injury can result from the required patient transfer and positioning for an MRI. The purpose of our study was to assess the frequency with which an MRI identified an injury not previously identified with CT, and whether this affected the treatment and outcome of the patient. Secondarily, we attempted to identify clinical or CT findings that may render an MRI particularly useful. Retrospective review. Patients with ASD who sustained acute spine fractures from 2005 to 2015. Acute fractures identified by CT scan and MRI upon admission; neurologic status upon admission and discharge, mode of injury, type of fracture, and final intervention before and after MRI assessment. A total of 124 patients with a diagnosis of diffuse idiopathic skeletal hyperostosis (DISH) or ankylosing spondylitis (AS) were identified by searching the radiology database of a level I trauma center with diagnosis keywords. Final radiology reports were assessed to determine presence and type of fracture(s) from CT. MRI report was then reviewed to assess if additional fractures or injuries were identified beyond that already known from the CT. Neurologic status upon admission and discharge, mode of injury, type of fracture, and final intervention were determined by inpatient notes and/or operative reports. No source funding or conflict of interest was present pertaining to this study. In the designated time frame, 124 ASD patients with injuries of the spine were identified who had obtained both a baseline CT and MRI. Six patients (4.8%) had additional injuries on MRI that had not been identified with CT. Four of these six patients had a change in treatment plan (three operative and one nonoperative) based on subsequent MRI findings. These included a (1) C4-5 hyperextension injury, (2) C6-7 hyperextension injury, (3) C7 bony fracture with C5-T4 epidural hematoma, and (4) C5-C6 hyperextension injury treated in a brace. Two of the six patients that had additional injuries identified on MRI had no change in their treatment plan. One patient had an additional lumbar extension injury identified above a previously diagnosed injury on CT, which was managed with a Thoracolumbosacral Orthosis (TLSO) according to the original plan. One patient died who had a known odontoid fracture and a suspected C6-7 hyperextension injury, and was identified on MRI as also having a C3-C4 hyperextension injury and a C2 spinal cord transection. In this study, 3.2% (4/124) of patients with ASD who presented to a level I trauma center with an acute spine injury identified with CT required a change in their treatment plan based on subsequent MRI findings. Only one fracture was missed on CT imaging, with the other missed injuries all being either disco-ligamentous hyperextension injuries through mobile discs or intracanal pathology. Our recommendation is that the routine use of MRI be limited to patients with nonankylosed levels in which a disco-ligamentous injury may have occurred, and in patients with neurological deficits that require investigation of the spinal canal to assess for causes of neurological injury.
Copyright remains with the authors and publisher.
Ankylosing spinal disorder (ASD) patients are at a greater risk for spinal fractures due to osteoporosis and rigidity of the spinal column. These fractures are associated with a high risk of neurologic compromise resulting from delayed or missed diagnoses. Although computed tomography (CT) is usually the initial imaging modality, magnetic resonance imaging (MRI) has been proposed as mandatory to help identify spinal injuries in ASD patients with unexplained neck or back pain or known injuries to help identify noncontiguous fractures. However, some studies have also shown that neurological injury can result from the required patient transfer and positioning for an MRI. The purpose of our study was to assess the frequency with which an MRI identified an injury not previously identified with CT, and whether this affected the treatment and outcome of the patient. Secondarily, we attempted to identify clinical or CT findings that may render an MRI particularly useful. Retrospective review. Patients with ASD who sustained acute spine fractures from 2005 to 2015. Acute fractures identified by CT scan and MRI upon admission; neurologic status upon admission and discharge, mode of injury, type of fracture, and final intervention before and after MRI assessment. A total of 124 patients with a diagnosis of diffuse idiopathic skeletal hyperostosis (DISH) or ankylosing spondylitis (AS) were identified by searching the radiology database of a level I trauma center with diagnosis keywords. Final radiology reports were assessed to determine presence and type of fracture(s) from CT. MRI report was then reviewed to assess if additional fractures or injuries were identified beyond that already known from the CT. Neurologic status upon admission and discharge, mode of injury, type of fracture, and final intervention were determined by inpatient notes and/or operative reports. No source funding or conflict of interest was present pertaining to this study. In the designated time frame, 124 ASD patients with injuries of the spine were identified who had obtained both a baseline CT and MRI. Six patients (4.8%) had additional injuries on MRI that had not been identified with CT. Four of these six patients had a change in treatment plan (three operative and one nonoperative) based on subsequent MRI findings. These included a (1) C4-5 hyperextension injury, (2) C6-7 hyperextension injury, (3) C7 bony fracture with C5-T4 epidural hematoma, and (4) C5-C6 hyperextension injury treated in a brace. Two of the six patients that had additional injuries identified on MRI had no change in their treatment plan. One patient had an additional lumbar extension injury identified above a previously diagnosed injury on CT, which was managed with a Thoracolumbosacral Orthosis (TLSO) according to the original plan. One patient died who had a known odontoid fracture and a suspected C6-7 hyperextension injury, and was identified on MRI as also having a C3-C4 hyperextension injury and a C2 spinal cord transection. In this study, 3.2% (4/124) of patients with ASD who presented to a level I trauma center with an acute spine injury identified with CT required a change in their treatment plan based on subsequent MRI findings. Only one fracture was missed on CT imaging, with the other missed injuries all being either disco-ligamentous hyperextension injuries through mobile discs or intracanal pathology. Our recommendation is that the routine use of MRI be limited to patients with nonankylosed levels in which a disco-ligamentous injury may have occurred, and in patients with neurological deficits that require investigation of the spinal canal to assess for causes of neurological injury.
Copyright remains with the authors and publisher.