- Email: [email protected]
Neurologic injury because of trauma after Type II odontoid nonunion
The Spine Journal 14 (2014) 903–908
Clinical Study
Neurologic injury because of trauma after Type II odontoid nonunion Christopher K. Kepler, MD, MBAa,*, Alexander R. Vaccaro, MD, PhDa, Florian Dibra, BSa, D. Greg Anderson, MDa, Jeffrey A. Rihn, MDa, Alan S. Hilibrand, MDa, James S. Harrop, MDb, Todd J. Albert, MDa, Kristen E. Radcliff, MDa a
Department of Orthopaedic Surgery, Rothman Institute, Thomas Jefferson University, 1015 Walnut St, Philadelphia, PA 19107, USA b Department of Neurosurgery, Thomas Jefferson University, Philadelphia, PA, USA Received 3 December 2011; revised 11 May 2013; accepted 13 July 2013
Abstract
BACKGROUND CONTEXT: Treatment of Type II odontoid fractures remains controversial, whereas nonoperative treatment is well accepted for isolated Type III odontoid fractures. Little is known about long-term sequelae of nonoperative management or risk of recurrent injury after nonsurgical treatment. We hypothesize that a substantial proportion of odontoid fractures assumed to be acute are actually chronic injuries and have a high rate of late displacement resulting in neurologic injury. PURPOSE: To identify patients presenting with previously unrecognized odontoid fracture nonunions and to document the incidence of new neurologic injury after secondary trauma in this population. STUDY DESIGN: Retrospective case series. PATIENT SAMPLE: One hundred thirty-three patients with Type II odontoid fractures presenting to a Level I trauma center. OUTCOME MEASURES: Computed tomography (CT) and magnetic resonance imaging (MRI) scans, American Spinal Injury Association Motor Score (AMS), and neurologic examination. METHODS: All patients presenting after traumatic injury to a Level I trauma center from May 2005 to May 2010 with a Type II odontoid fracture on CT scan were included. Patients aged less than 18 years and those with pathologic fractures were excluded. Fractures were classified as chronic or acute based on CT evidence of chronic injury/nonunion including fracture resorption, sclerosis, and cyst formation. Magnetic resonance imaging was then examined for evidence of fracture acuity (increased signal in C2 on T2 images). Patients without evidence of acute fracture on MRI were considered to have chronic injuries. Computed tomography and MRI scans were interpreted independently by two reviewers. Chart review was performed to document demographics, AMS, and new-onset neurologic deficit associated with secondary injury. RESULTS: One hundred thirty-three patients presented with Type II odontoid fractures and no known history of cervical fracture with an average age of 79 years. Based on CT criteria, 31/133 (23%) fractures were chronic injuries. Nine additional fractures appeared acute on CT but were determined to be chronic by MRI findings. The overall number of chronic fractures was therefore 40 (30%). Interobserver reliability analysis for classification of fractures as chronic demonstrated k50.65 representing substantial agreement. Of the 40 chronic fractures, 7 patients (17.5%) had new-onset neurologic deficits after secondary injury including 4 motor deficits, 2 sensory deficits, and 1 combined deficit. Although the chronic injury group as a whole had similar AMS to the acute injury group (89 vs. 84, p5.27), the seven patients with new-onset neurologic deficit had an average AMS of 52.4. CONCLUSIONS: A substantial proportion of patients presenting after cervical trauma with Type II odontoid fractures have evidence of nonacute injury. Of these patients, 17% presented with a new neurologic deficit caused by an ‘‘acute-on-chronic’’ injury. Ó 2014 Elsevier Inc. All rights reserved.
Keywords:
Type II odontoid fracture; Odontoid nonunion; Cervical spine trauma; ASIA score; Dens fracture
FDA device/drug status: Not applicable. Author disclosures: Listed at the end of this article. * Corresponding author. Department of Orthopaedic Surgery, Rothman Institute, Thomas Jefferson University, 1015 Walnut St, Curtis Building 1529-9430/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.spinee.2013.07.443
Room 508, Philadelphia, PA 19107, USA. Tel.: (215) 955-9773; fax: (215) 955-4322. E-mail address: [email protected] (C.K. Kepler)
904
C.K. Kepler et al. / The Spine Journal 14 (2014) 903–908
Context The authors aimed to assess whether some Type II odontoid fractures seen acutely are actually acute-on-chronic injuries representing old nonunions. Contribution Using their defined CT and MRI criteria, the authors found that 30% of acutely diagnosed fractures might actually be acute-on-chronic injuries. Implications Assuming the radiographic criteria are accurate, the findings are quite interesting. It is unclear, however, whether the conclusion can be drawn that surgery should be recommended for all presenting Type II fractures in order to avoid the risk of neurological injury due to secondary trauma. There may be a very high prevalence of individuals with nonunions who may never have a problem after subsequent falls. The study design does not lend itself to answering this question. That said, the findings might be useful during shared decision-making in acute presentations. —The Editors
fracture is uncommon, although the overall complication rate remains high [10]. The incidence of neurologic deficit after odontoid fracture has been reported to range from 13% to 25% [1,2,8,9,11,12], with tetraparesis seen in only 2% to 8.5% [1,2,8,9,11], likely because of the relatively capacious spinal canal in the upper cervical region adjacent to the odontoid process. Management of Type II odontoid fractures is controversial, and high rates of fracture nonunion after conservative treatment using an orthosis are well documented, exceeding 75% in some hard collar series [7,11,13]. Similarly, treatment with halo-vest immobilization results in nonunion rates reported as high as 54% [14]. Although high rates of nonunion occur after conservative treatment and immobilization, the long-term clinical significance of odontoid nonunion is not well characterized and has rarely been investigated in a systematic manner. In this study, we sought to identify previously unrecognized odontoid fracture nonunions to document the rate of nonacute injuries in patients presenting with Type II odontoid fractures and report the incidence of new neurologic injury after secondary trauma in patients with odontoid nonunions. We hypothesize that a substantial proportion of odontoid fractures assumed to be acute are actually chronic injuries and that these chronic injuries have a significant rate of later displacement resulting in neurologic injury.
Introduction Odontoid fractures have been reported to represent between 9% and 19% of cervical spine fractures [1–3] and are most common in the elderly. According to the classification developed by Anderson and D’Alonzo [4] and modified by Grauer et al. [5], Type II fractures are the most common type, representing between 56% and 85% of odontoid fractures [1,4,6–9]. Despite the location of the fracture in the proximal cervical spine, where a spinal cord injury can be devastating, neurologic deficit after odontoid
Methods All patients presenting after traumatic injury to a Level I trauma center from May 2005 to May 2010 with a Type II odontoid fracture as identified on computed tomography (CT) scan were included. Patients younger than 18 years of age and those with pathologic fractures were excluded. Furthermore, patients presenting after traumatic injury but with a previously diagnosed odontoid fracture that had been treated either conservatively or surgically were excluded to
Fig. 1. Sagittal computed tomography images demonstrating (Left) atrophic odontoid nonunion with sclerosis and resorption at the fracture line, (Middle) hypertrophic nonunion, and (Right) nonunion with cyst formation at the fracture line on the proximal fragment.
C.K. Kepler et al. / The Spine Journal 14 (2014) 903–908
905
Fig. 2. Sagittal computed tomography (CT) (Left) and magnetic resonance imaging (MRI) (Right) images demonstrating no stigmata of fracture nonunion on CT images but without high T2 signal on MRI indicating nonacuity of fracture.
allow our analysis to be generalized to patients on first presentation. Fractures were classified as chronic or acute based on evidence of chronic injury or nonunion on CT scan including fracture resorption, callus formation, sclerosis, and cyst formation (Fig. 1). Midline sagittal CT cuts were used to measure fracture displacement in both the anteroposterior direction and fracture distraction. Magnetic resonance imaging (MRI) obtained within 1 day of presentation was then examined for evidence of fracture acuity based on the presence or absence of increased signal in the region of the odontoid process on T2 images. Patients without evidence of fracture acuity on MRI were considered to have chronic injuries (Fig. 2). Magnetic resonance imaging was also evaluated to determine the presence of spinal cord compression, compressive hematoma, and signal changes in the spinal cord at the level of the odontoid as seen on T2 images. Imaging studies were interpreted independently by two observers (CKK and KER). Chart review was performed to document patient demographics, history, American Spinal Injury Association Motor Score (AMS), and incidence of new-onset neurologic deficit associated with secondary injury as documented at presentation by clinical examination of the spinal cord injury consult team. Differences between groups were analyzed using chisquared tests for proportions and t tests for continuous variables. Significance was assumed for p value less than .05. Cohen kappa (k) was used to quantify agreement between observers. Kappa values were graded according to a previously described semiquantitative scale: no agreement for values less than 0, slight for 0 to 0.20, fair for 0.21 to 0.40, moderate for 0.41 to 0.60, substantial for 0.61 to 0.80, and near perfect for 0.81 to 1.0. Statistical analysis was performed using SPSS 16.0 (SPSS, Inc., Chicago, IL, USA).
Results Based on the inclusion criteria, 133 Type II odontoid fractures without known history of cervical fracture were identified. This population included 46 men (35%) with an average age of 79 years. The average AMS for the cohort was 86. Based on CT appearance, 31/133 (23%) fractures were chronic injuries. Overall, 65 of the fractures demonstrated no anteroposterior translation, 58 demonstrated posterior displacement of the proximal dens by an average of 4.8 mm, and 11 demonstrated anterior displacement of the proximal dens by an average of 3.2 mm. The most common feature seen on CT scan indicative of nonunion was sclerosis at the fracture line that was seen in 30/31 patients (97%). Cyst formation at the fracture site was seen in 17/31 (55%) of chronic fractures. Five fractures demonstrated a hypertrophic nonunion (16%), 18 demonstrated an atrophic nonunion (58%), and the remaining 8 fractures were eutrophic (26%). There were nine additional fractures that initially appeared to be acute based on CT appearance but were determined to be chronic by MRI findings. The percentage of fractures thought to be acute on CT but appeared chronic on MRI was 9/10358.7%. The overall number of chronic fractures was therefore 40/133 Type II fractures (30%). Interobserver reliability analysis for classification of fractures as chronic demonstrated k50.65 representing substantial agreement between readers. Of the 40 chronic fractures, 7 patients (17.5%) had newonset neurologic deficits after secondary injury including 4 motor deficits, 2 sensory deficits, and 1 combined deficit. In the acute group, there was a 16% rate of new-onset neurologic injury overall, a proportion that was not significantly different in comparison with the chronic group (p5.89). Although the chronic injury group as a whole had similar AMS to the acute injury group (89 vs. 84, p5.27), the seven
906
C.K. Kepler et al. / The Spine Journal 14 (2014) 903–908
patients with new-onset neurologic deficit had an average AMS of 52.4. Of the six patients with chronic fractures and neurologic deficit for whom mechanism of injury was known, all six (100%) were injured through falls from standing height or less, a rate that is statistically similar to the rate of low-energy mechanisms in patients with neurologic injuries from the acute group (8/12566%, p5.23). The acute and chronic injury groups had similarly low percentages of patients with associated subaxial disc herniations and fractures (7% vs. 3%, p5.28) and patients with subaxial injuries who had new neurologic injuries (2.5% vs. 1%, p5.53).
Conclusions In this series of consecutive patients presenting to a Level I trauma center with a Type II odontoid fracture as identified by CT scan, 30% of patients had a previously undiagnosed odontoid fracture nonunion. These patients likely sustained odontoid fractures from previous trauma that were not previously diagnosed and went on to nonunion. The rate of neurologic injury in patients presenting with chronic odontoid nonunion was 17.5%, similar to that of patients presenting with acute fracture. The etiology of neurologic deficit is likely multifactorial. We identified four patients who had hypertrophic nonunions with secondary spinal cord compression in the setting of preexisting myelopathy (Fig. 1). Another patient had significant displacement of an established nonunion (Fig. 3). The remaining two patients had no active spinal
Fig. 3. Sagittal computed tomography image demonstrating established nonunion with fragment sclerosis and cyst formation with significant displacement of proximal fragment posteriorly.
cord compression but neurologic deficits nevertheless. These patients may have had a transient displacement at the time of injury either at the fracture site or through the C1–C2 articulation that resulted in spinal cord injury before the fracture subsequently reduced, resulting in normal appearing alignment. Most patients presenting with previously unrecognized nonunion were neurologically intact, despite their recent trauma. This is similar to primary odontoid fractures and is consistent with the data from previous investigations that analyzed patients who developed nonunion after Type II odontoid fracture. Hart et al. [13] described a series of patients who were not diagnosed with Type II odontoid fracture immediately after injury and went on to nonunion before radiographs taken for other reasons (neck pain in four patients, minor trauma in one patient) led to the diagnosis of odontoid nonunion. Although at a follow-up of 4.7 years, only two patients had chronic neck pain and all patients remained neurologically intact, it should be noted that this series included only five patients and only one who presented after trauma. Similarly, Ryan and Taylor [7] described a series of patients with odontoid fractures of whom nine patients with Type II nonunion were neurologically intact at a follow-up of 21 months. Our data support the generally benign postinjury course of patients with odontoid nonunion—82.5% of patients with odontoid nonunions had no neurologic deficit in this series of patients presenting after recent trauma. Because patients were included in our series only after a traumatic injury prompted imaging of their cervical spine at a regional spinal cord injury center, it is not surprising that we identified patients with neurologic injury after nonunion even if previous series have not. More than half of the patients with nonunion were transferred from the outside institutions for evaluation and definitive treatment, a characteristic of our cohort that is likely to only increase the bias toward more severely injured patients. This selection bias limits our ability to make general statements about patients with odontoid nonunion who do not present with trauma but also presented an opportunity to study the potential risks of odontoid nonunion. Patients in this study did not know they had a fracture; so by definition, they did not have a preventable injury. Patients, however, with known fractures that are either treated conservatively or those with previously undiagnosed fractures that are incidentally identified on imaging may be similarly prone to morbidity after secondary trauma. Acuteon-chronic injury in these patients may potentially be preventable through fracture stabilization and healing in the acute postinjury period or after delayed presentation. We believe the compromised biomechanics of a fibrous or mobile nonunion may confer a susceptibility to relative fracture instability and resultant neurologic injury in this subpopulation. This suggests that physicians should counsel their patients regarding the possibility of subsequent late displacement and neurologic injury of odontoid nonunions. This is not an affirmation of a particular stabilization
C.K. Kepler et al. / The Spine Journal 14 (2014) 903–908
technique after the initial injury, a topic outside of the scope of this article. Based on these data, we suggest that definitive fracture stabilization at the time of initial injury would have likely prevented severe sudden neurologic injury from secondary trauma in 17% of patients. We acknowledge that surgery may have unintended consequences such as adjacent-level degeneration or subsequent propensity to subaxial injury. Despite these risks, surgical stabilization of Type II odontoid fractures has recently been shown in a prospective observational study to be associated with improved functional outcome at 1 year in an elderly population [15]. When ‘‘benign neglect’’ or conservative treatment of odontoid nonunions is considered, the risks of stabilization should be weighed against the risk of late displacement and neurologic deficit from a remote trauma. There are several limitations to our study. We did not obtain dynamic imaging to distinguish those patients with fibrous nonunion from mobile nonunion, a distinction that may be important in determining susceptibility to neurologic injury from secondary trauma. The use of MRI to identify nonacute fractures is likely not 100% specific for the presence of a fracture; so, it is possible that patients assumed to have a nonunion may, in fact, have been acute presentations. All MRIs were obtained within a day of initial presentation, a factor we feel mitigates this limitation as early imaging should capture bone edema and other stigmata of acute fracture on T2 images. No previous studies have been done to establish the sensitivity and specificity of CT and MRI for identifying chronic odontoid fractures, and we have no other gold standard method of diagnosis to rely on; we did, however, demonstrate good interobserver reliability based on the features we considered to represent chronic injury. We cannot be sure that neurologic deficits did not arise from undiagnosed subaxial injuries. We have no baseline preinjury neurologic examinations because of the undiagnosed nature of the odontoid nonunions and therefore had to rely on patient- and family-provided history to identify when physical examination findings represented a new-onset deficit. Because patients identified in this study as having chronic fractures were not aware of their index injury, most patients likely had no postinjury immobilization that may have contributed to a more mobile nonunion than if a cervical orthosis was used. Therefore, it is unknown if our results are fully generalizable to patients treated conservatively using a cervical orthosis who then develop nonunion. Finally, our study cannot definitively comment on the protective effect of surgical stabilization against neurologic injury from secondary trauma—biomechanical studies are needed to explore the relative force needed to cause spinal cord compression with surgically stabilized versus mobile odontoid processes. Author disclosures: CKK: Nothing to disclose. ARV: Royalties: DePuy Spine (E), Biomet Spine (E), Globus (F), Medtronic (H), K2M (D),
907
Aesculap (B), BI Medical (C), Book royalties (C), Stryker Spine (G), Osteotech (B), Alphatec (B); Stock Ownership: K2M (F), Cytonics (percent of entity unknown), Disc Motion Technology (D), Location-Based Intelligence (~20% of entity), Progressive Spinal Technology (percent of entity unknown), Computational Biodynamics (percent of entity unknown), Stout Medical (B), Bonovo Orthopedics (100,000 Shares), Electrocure (D), Vertiflex (30,000 Shares), Flagship Surgical (D), In Vivo (percent of entity unknown), Pearl Diver (C), Small Bone Innovations (30,000 Shares), Neucore (22,000 Shares), Crosscurrent (125,000 Shares), Onset Medical (1,535 Shares), Orthovita (10,000 Shares), Syndicom (2,750 Shares), Biomerix (25,000 Shares), Paradigm Spine (146,875 Shares), Spinology (28,750 Shares), Gamma Spine (~15% of entity), Replication Medica (15,250 Shares), Breakaway Imaging (100,000 Shares), Crosstree Medical (41,667 Shares), Globus (816,123 Shares), FlowPharma (percent ownership of stock options unknown), Advanced Spinal Intellectual Properties (30% of entity), SpineMedica (number of shares unknown), RIS (!5,000 options); Consulting: Gerson Lehrman Group (B), Guidepoint Global (B), Mediacorp (B); Board of Directors: AO Spine (None, North America Board of Directors travel expenses stipend, B), ASIA (None, Past President Board member), NASS (None, Past Program Committee Co-chairman); Research Support—Staff and/or Materials: AO Spine (E), Cerapeutics (C); Fellowship Support: Stryker Spine (E); Other: NuVasive (C), Applied Spinal Intellectual Property (C), Confluent Surgical (B). FD: Nothing to disclose. DGA: Royalties: Medtronic (F), DePuy Spine (H); Consulting: DePuy Spine (D), Synthes (B); Speaking and/or Teaching Arrangements: DePuy Spine (D, consulting); Trips/Travel: DePuy Spine (D, consulting); Other Office: Society for Minimally Invasive Spinal Surgery (None, First Vice President). JAR: Other Office: Federation of Spine Associations (None, President). ASH: Royalties: Biomet Spine (F), Alphatec Spine (F), Stryker (C), Zimmer (C), Amedica (C), Aesculap (B); Stock Ownership: Amedica (20,000 Shares, !1%), Lifespine (20,000 Shares, !1%), Spinal Ventures (initial investment was 3% but has been diluted), Syndicom (20,000 Shares); Private Investments: Benvenue (B), Nexgen (B), Paradigm Spine (B), Pioneer (10,000 Shares, C), PSD (B), Vertiflex (B); Scientific Advisory Board: Amedica (20,000 shares of options). JSH: Stock Ownership: Axiomed (option to buy 15,000 shares); Consulting: Depuy Spine (C, Paid directly to institution/employer); Speaking and/or Teaching Arrangements: Medtronic (None, no longer a consultant for this company); Trips/Travel: Stryker (None); Board of Directors: Jefferson Medical College Physician Board (None); Scientific Advisory Board: Axiomed (None, Medical Advisory Board), Geron (None), Depuy (C, Paid directly to institution/employer), CNS (None, Executive Board, Chairman of Publication Committee, Editor of CNSQ, Chairman for Neurostimulation Project); Other Office: Pennsylvania Neurologic Society (None, Board of Pennsylvania Neurosurgical Society);
908
C.K. Kepler et al. / The Spine Journal 14 (2014) 903–908
Research Support—Staff and/or Materials: NACTN (E, Paid directly to institution/employer). TJA: Relationships Outside the 1-Year Requirement: CSRS (12/2008, Royalties, A). KER: Nothing to disclose. The disclosure key can be found on the Table of Contents and at www.TheSpineJournalOnline.com.
References [1] Hadley MN, Browner C, Sonntag VK. Axis fractures: a comprehensive review of management and treatment in 107 cases. Neurosurgery 1985;17:281–90. [2] Greene KA, Dickman CA, Marciano FF, et al. Acute axis fractures. Analysis of management and outcome in 340 consecutive cases. Spine 1997;22:1843–52. [3] Goldberg W, Mueller C, Panacek E, et al. Distribution and patterns of blunt traumatic cervical spine injury. Ann Emerg Med 2001;38:17–21. [4] Anderson LD, D’Alonzo RT. Fractures of the odontoid process of the axis. 1974. J Bone Joint Surg Am 2004;86-A:2081. [5] Grauer JN, Shafi B, Hilibrand AS, et al. Proposal of a modified, treatment-oriented classification of odontoid fractures. Spine J 2005;5:123–9. [6] Hanigan WC, Powell FC, Elwood PW, Henderson JP. Odontoid fractures in elderly patients. J Neurosurg 1993;78:32–5.
[7] Ryan MD, Taylor TK. Odontoid fractures in the elderly. J Spinal Disord 1993;6:397–401. [8] Polin RS, Szabo T, Bogaev CA, et al. Nonoperative management of types II and III odontoid fractures: the Philadelphia collar versus the halo vest. Neurosurgery 1996;38:450–6; discussion 456–7. [9] Kuntz C IV, Mirza SK, Jarell AD, et al. Type II odontoid fractures in the elderly: early failure of nonsurgical treatment. Neurosurg Focus 2000;8:e7. [10] Smith HE, Kerr SM, Maltenfort M, et al. Early complications of surgical versus conservative treatment of isolated type II odontoid fractures in octogenarians: a retrospective cohort study. J Spinal Disord Tech 2008;21:535–9. [11] Clark CR, White AA III. Fractures of the dens. A multicenter study. J Bone Joint Surg Am 1985;67:1340–8. [12] Harrop JS, Sharan AD, Przybylski GJ. Epidemiology of spinal cord injury after acute odontoid fractures. Neurosurg Focus 2000; 8:e4. [13] Hart R, Saterbak A, Rapp T, Clark C. Nonoperative management of dens fracture nonunion in elderly patients without myelopathy. Spine 2000;25:1339–43. [14] Koivikko MP, Kiuru MJ, Koskinen SK, et al. Factors associated with nonunion in conservatively-treated type-II fractures of the odontoid process. J Bone Joint Surg Br 2004;86:1146–51. [15] Vaccaro AR, Kepler CK, Kopjar B, et al. Functional and quality-oflife outcomes in geriatric patients with type-II dens fracture. J Bone Joint Surg Am 2013;95:729–35.
Clinical Study
Neurologic injury because of trauma after Type II odontoid nonunion Christopher K. Kepler, MD, MBAa,*, Alexander R. Vaccaro, MD, PhDa, Florian Dibra, BSa, D. Greg Anderson, MDa, Jeffrey A. Rihn, MDa, Alan S. Hilibrand, MDa, James S. Harrop, MDb, Todd J. Albert, MDa, Kristen E. Radcliff, MDa a
Department of Orthopaedic Surgery, Rothman Institute, Thomas Jefferson University, 1015 Walnut St, Philadelphia, PA 19107, USA b Department of Neurosurgery, Thomas Jefferson University, Philadelphia, PA, USA Received 3 December 2011; revised 11 May 2013; accepted 13 July 2013
Abstract
BACKGROUND CONTEXT: Treatment of Type II odontoid fractures remains controversial, whereas nonoperative treatment is well accepted for isolated Type III odontoid fractures. Little is known about long-term sequelae of nonoperative management or risk of recurrent injury after nonsurgical treatment. We hypothesize that a substantial proportion of odontoid fractures assumed to be acute are actually chronic injuries and have a high rate of late displacement resulting in neurologic injury. PURPOSE: To identify patients presenting with previously unrecognized odontoid fracture nonunions and to document the incidence of new neurologic injury after secondary trauma in this population. STUDY DESIGN: Retrospective case series. PATIENT SAMPLE: One hundred thirty-three patients with Type II odontoid fractures presenting to a Level I trauma center. OUTCOME MEASURES: Computed tomography (CT) and magnetic resonance imaging (MRI) scans, American Spinal Injury Association Motor Score (AMS), and neurologic examination. METHODS: All patients presenting after traumatic injury to a Level I trauma center from May 2005 to May 2010 with a Type II odontoid fracture on CT scan were included. Patients aged less than 18 years and those with pathologic fractures were excluded. Fractures were classified as chronic or acute based on CT evidence of chronic injury/nonunion including fracture resorption, sclerosis, and cyst formation. Magnetic resonance imaging was then examined for evidence of fracture acuity (increased signal in C2 on T2 images). Patients without evidence of acute fracture on MRI were considered to have chronic injuries. Computed tomography and MRI scans were interpreted independently by two reviewers. Chart review was performed to document demographics, AMS, and new-onset neurologic deficit associated with secondary injury. RESULTS: One hundred thirty-three patients presented with Type II odontoid fractures and no known history of cervical fracture with an average age of 79 years. Based on CT criteria, 31/133 (23%) fractures were chronic injuries. Nine additional fractures appeared acute on CT but were determined to be chronic by MRI findings. The overall number of chronic fractures was therefore 40 (30%). Interobserver reliability analysis for classification of fractures as chronic demonstrated k50.65 representing substantial agreement. Of the 40 chronic fractures, 7 patients (17.5%) had new-onset neurologic deficits after secondary injury including 4 motor deficits, 2 sensory deficits, and 1 combined deficit. Although the chronic injury group as a whole had similar AMS to the acute injury group (89 vs. 84, p5.27), the seven patients with new-onset neurologic deficit had an average AMS of 52.4. CONCLUSIONS: A substantial proportion of patients presenting after cervical trauma with Type II odontoid fractures have evidence of nonacute injury. Of these patients, 17% presented with a new neurologic deficit caused by an ‘‘acute-on-chronic’’ injury. Ó 2014 Elsevier Inc. All rights reserved.
Keywords:
Type II odontoid fracture; Odontoid nonunion; Cervical spine trauma; ASIA score; Dens fracture
FDA device/drug status: Not applicable. Author disclosures: Listed at the end of this article. * Corresponding author. Department of Orthopaedic Surgery, Rothman Institute, Thomas Jefferson University, 1015 Walnut St, Curtis Building 1529-9430/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.spinee.2013.07.443
Room 508, Philadelphia, PA 19107, USA. Tel.: (215) 955-9773; fax: (215) 955-4322. E-mail address: [email protected] (C.K. Kepler)
904
C.K. Kepler et al. / The Spine Journal 14 (2014) 903–908
Context The authors aimed to assess whether some Type II odontoid fractures seen acutely are actually acute-on-chronic injuries representing old nonunions. Contribution Using their defined CT and MRI criteria, the authors found that 30% of acutely diagnosed fractures might actually be acute-on-chronic injuries. Implications Assuming the radiographic criteria are accurate, the findings are quite interesting. It is unclear, however, whether the conclusion can be drawn that surgery should be recommended for all presenting Type II fractures in order to avoid the risk of neurological injury due to secondary trauma. There may be a very high prevalence of individuals with nonunions who may never have a problem after subsequent falls. The study design does not lend itself to answering this question. That said, the findings might be useful during shared decision-making in acute presentations. —The Editors
fracture is uncommon, although the overall complication rate remains high [10]. The incidence of neurologic deficit after odontoid fracture has been reported to range from 13% to 25% [1,2,8,9,11,12], with tetraparesis seen in only 2% to 8.5% [1,2,8,9,11], likely because of the relatively capacious spinal canal in the upper cervical region adjacent to the odontoid process. Management of Type II odontoid fractures is controversial, and high rates of fracture nonunion after conservative treatment using an orthosis are well documented, exceeding 75% in some hard collar series [7,11,13]. Similarly, treatment with halo-vest immobilization results in nonunion rates reported as high as 54% [14]. Although high rates of nonunion occur after conservative treatment and immobilization, the long-term clinical significance of odontoid nonunion is not well characterized and has rarely been investigated in a systematic manner. In this study, we sought to identify previously unrecognized odontoid fracture nonunions to document the rate of nonacute injuries in patients presenting with Type II odontoid fractures and report the incidence of new neurologic injury after secondary trauma in patients with odontoid nonunions. We hypothesize that a substantial proportion of odontoid fractures assumed to be acute are actually chronic injuries and that these chronic injuries have a significant rate of later displacement resulting in neurologic injury.
Introduction Odontoid fractures have been reported to represent between 9% and 19% of cervical spine fractures [1–3] and are most common in the elderly. According to the classification developed by Anderson and D’Alonzo [4] and modified by Grauer et al. [5], Type II fractures are the most common type, representing between 56% and 85% of odontoid fractures [1,4,6–9]. Despite the location of the fracture in the proximal cervical spine, where a spinal cord injury can be devastating, neurologic deficit after odontoid
Methods All patients presenting after traumatic injury to a Level I trauma center from May 2005 to May 2010 with a Type II odontoid fracture as identified on computed tomography (CT) scan were included. Patients younger than 18 years of age and those with pathologic fractures were excluded. Furthermore, patients presenting after traumatic injury but with a previously diagnosed odontoid fracture that had been treated either conservatively or surgically were excluded to
Fig. 1. Sagittal computed tomography images demonstrating (Left) atrophic odontoid nonunion with sclerosis and resorption at the fracture line, (Middle) hypertrophic nonunion, and (Right) nonunion with cyst formation at the fracture line on the proximal fragment.
C.K. Kepler et al. / The Spine Journal 14 (2014) 903–908
905
Fig. 2. Sagittal computed tomography (CT) (Left) and magnetic resonance imaging (MRI) (Right) images demonstrating no stigmata of fracture nonunion on CT images but without high T2 signal on MRI indicating nonacuity of fracture.
allow our analysis to be generalized to patients on first presentation. Fractures were classified as chronic or acute based on evidence of chronic injury or nonunion on CT scan including fracture resorption, callus formation, sclerosis, and cyst formation (Fig. 1). Midline sagittal CT cuts were used to measure fracture displacement in both the anteroposterior direction and fracture distraction. Magnetic resonance imaging (MRI) obtained within 1 day of presentation was then examined for evidence of fracture acuity based on the presence or absence of increased signal in the region of the odontoid process on T2 images. Patients without evidence of fracture acuity on MRI were considered to have chronic injuries (Fig. 2). Magnetic resonance imaging was also evaluated to determine the presence of spinal cord compression, compressive hematoma, and signal changes in the spinal cord at the level of the odontoid as seen on T2 images. Imaging studies were interpreted independently by two observers (CKK and KER). Chart review was performed to document patient demographics, history, American Spinal Injury Association Motor Score (AMS), and incidence of new-onset neurologic deficit associated with secondary injury as documented at presentation by clinical examination of the spinal cord injury consult team. Differences between groups were analyzed using chisquared tests for proportions and t tests for continuous variables. Significance was assumed for p value less than .05. Cohen kappa (k) was used to quantify agreement between observers. Kappa values were graded according to a previously described semiquantitative scale: no agreement for values less than 0, slight for 0 to 0.20, fair for 0.21 to 0.40, moderate for 0.41 to 0.60, substantial for 0.61 to 0.80, and near perfect for 0.81 to 1.0. Statistical analysis was performed using SPSS 16.0 (SPSS, Inc., Chicago, IL, USA).
Results Based on the inclusion criteria, 133 Type II odontoid fractures without known history of cervical fracture were identified. This population included 46 men (35%) with an average age of 79 years. The average AMS for the cohort was 86. Based on CT appearance, 31/133 (23%) fractures were chronic injuries. Overall, 65 of the fractures demonstrated no anteroposterior translation, 58 demonstrated posterior displacement of the proximal dens by an average of 4.8 mm, and 11 demonstrated anterior displacement of the proximal dens by an average of 3.2 mm. The most common feature seen on CT scan indicative of nonunion was sclerosis at the fracture line that was seen in 30/31 patients (97%). Cyst formation at the fracture site was seen in 17/31 (55%) of chronic fractures. Five fractures demonstrated a hypertrophic nonunion (16%), 18 demonstrated an atrophic nonunion (58%), and the remaining 8 fractures were eutrophic (26%). There were nine additional fractures that initially appeared to be acute based on CT appearance but were determined to be chronic by MRI findings. The percentage of fractures thought to be acute on CT but appeared chronic on MRI was 9/10358.7%. The overall number of chronic fractures was therefore 40/133 Type II fractures (30%). Interobserver reliability analysis for classification of fractures as chronic demonstrated k50.65 representing substantial agreement between readers. Of the 40 chronic fractures, 7 patients (17.5%) had newonset neurologic deficits after secondary injury including 4 motor deficits, 2 sensory deficits, and 1 combined deficit. In the acute group, there was a 16% rate of new-onset neurologic injury overall, a proportion that was not significantly different in comparison with the chronic group (p5.89). Although the chronic injury group as a whole had similar AMS to the acute injury group (89 vs. 84, p5.27), the seven
906
C.K. Kepler et al. / The Spine Journal 14 (2014) 903–908
patients with new-onset neurologic deficit had an average AMS of 52.4. Of the six patients with chronic fractures and neurologic deficit for whom mechanism of injury was known, all six (100%) were injured through falls from standing height or less, a rate that is statistically similar to the rate of low-energy mechanisms in patients with neurologic injuries from the acute group (8/12566%, p5.23). The acute and chronic injury groups had similarly low percentages of patients with associated subaxial disc herniations and fractures (7% vs. 3%, p5.28) and patients with subaxial injuries who had new neurologic injuries (2.5% vs. 1%, p5.53).
Conclusions In this series of consecutive patients presenting to a Level I trauma center with a Type II odontoid fracture as identified by CT scan, 30% of patients had a previously undiagnosed odontoid fracture nonunion. These patients likely sustained odontoid fractures from previous trauma that were not previously diagnosed and went on to nonunion. The rate of neurologic injury in patients presenting with chronic odontoid nonunion was 17.5%, similar to that of patients presenting with acute fracture. The etiology of neurologic deficit is likely multifactorial. We identified four patients who had hypertrophic nonunions with secondary spinal cord compression in the setting of preexisting myelopathy (Fig. 1). Another patient had significant displacement of an established nonunion (Fig. 3). The remaining two patients had no active spinal
Fig. 3. Sagittal computed tomography image demonstrating established nonunion with fragment sclerosis and cyst formation with significant displacement of proximal fragment posteriorly.
cord compression but neurologic deficits nevertheless. These patients may have had a transient displacement at the time of injury either at the fracture site or through the C1–C2 articulation that resulted in spinal cord injury before the fracture subsequently reduced, resulting in normal appearing alignment. Most patients presenting with previously unrecognized nonunion were neurologically intact, despite their recent trauma. This is similar to primary odontoid fractures and is consistent with the data from previous investigations that analyzed patients who developed nonunion after Type II odontoid fracture. Hart et al. [13] described a series of patients who were not diagnosed with Type II odontoid fracture immediately after injury and went on to nonunion before radiographs taken for other reasons (neck pain in four patients, minor trauma in one patient) led to the diagnosis of odontoid nonunion. Although at a follow-up of 4.7 years, only two patients had chronic neck pain and all patients remained neurologically intact, it should be noted that this series included only five patients and only one who presented after trauma. Similarly, Ryan and Taylor [7] described a series of patients with odontoid fractures of whom nine patients with Type II nonunion were neurologically intact at a follow-up of 21 months. Our data support the generally benign postinjury course of patients with odontoid nonunion—82.5% of patients with odontoid nonunions had no neurologic deficit in this series of patients presenting after recent trauma. Because patients were included in our series only after a traumatic injury prompted imaging of their cervical spine at a regional spinal cord injury center, it is not surprising that we identified patients with neurologic injury after nonunion even if previous series have not. More than half of the patients with nonunion were transferred from the outside institutions for evaluation and definitive treatment, a characteristic of our cohort that is likely to only increase the bias toward more severely injured patients. This selection bias limits our ability to make general statements about patients with odontoid nonunion who do not present with trauma but also presented an opportunity to study the potential risks of odontoid nonunion. Patients in this study did not know they had a fracture; so by definition, they did not have a preventable injury. Patients, however, with known fractures that are either treated conservatively or those with previously undiagnosed fractures that are incidentally identified on imaging may be similarly prone to morbidity after secondary trauma. Acuteon-chronic injury in these patients may potentially be preventable through fracture stabilization and healing in the acute postinjury period or after delayed presentation. We believe the compromised biomechanics of a fibrous or mobile nonunion may confer a susceptibility to relative fracture instability and resultant neurologic injury in this subpopulation. This suggests that physicians should counsel their patients regarding the possibility of subsequent late displacement and neurologic injury of odontoid nonunions. This is not an affirmation of a particular stabilization
C.K. Kepler et al. / The Spine Journal 14 (2014) 903–908
technique after the initial injury, a topic outside of the scope of this article. Based on these data, we suggest that definitive fracture stabilization at the time of initial injury would have likely prevented severe sudden neurologic injury from secondary trauma in 17% of patients. We acknowledge that surgery may have unintended consequences such as adjacent-level degeneration or subsequent propensity to subaxial injury. Despite these risks, surgical stabilization of Type II odontoid fractures has recently been shown in a prospective observational study to be associated with improved functional outcome at 1 year in an elderly population [15]. When ‘‘benign neglect’’ or conservative treatment of odontoid nonunions is considered, the risks of stabilization should be weighed against the risk of late displacement and neurologic deficit from a remote trauma. There are several limitations to our study. We did not obtain dynamic imaging to distinguish those patients with fibrous nonunion from mobile nonunion, a distinction that may be important in determining susceptibility to neurologic injury from secondary trauma. The use of MRI to identify nonacute fractures is likely not 100% specific for the presence of a fracture; so, it is possible that patients assumed to have a nonunion may, in fact, have been acute presentations. All MRIs were obtained within a day of initial presentation, a factor we feel mitigates this limitation as early imaging should capture bone edema and other stigmata of acute fracture on T2 images. No previous studies have been done to establish the sensitivity and specificity of CT and MRI for identifying chronic odontoid fractures, and we have no other gold standard method of diagnosis to rely on; we did, however, demonstrate good interobserver reliability based on the features we considered to represent chronic injury. We cannot be sure that neurologic deficits did not arise from undiagnosed subaxial injuries. We have no baseline preinjury neurologic examinations because of the undiagnosed nature of the odontoid nonunions and therefore had to rely on patient- and family-provided history to identify when physical examination findings represented a new-onset deficit. Because patients identified in this study as having chronic fractures were not aware of their index injury, most patients likely had no postinjury immobilization that may have contributed to a more mobile nonunion than if a cervical orthosis was used. Therefore, it is unknown if our results are fully generalizable to patients treated conservatively using a cervical orthosis who then develop nonunion. Finally, our study cannot definitively comment on the protective effect of surgical stabilization against neurologic injury from secondary trauma—biomechanical studies are needed to explore the relative force needed to cause spinal cord compression with surgically stabilized versus mobile odontoid processes. Author disclosures: CKK: Nothing to disclose. ARV: Royalties: DePuy Spine (E), Biomet Spine (E), Globus (F), Medtronic (H), K2M (D),
907
Aesculap (B), BI Medical (C), Book royalties (C), Stryker Spine (G), Osteotech (B), Alphatec (B); Stock Ownership: K2M (F), Cytonics (percent of entity unknown), Disc Motion Technology (D), Location-Based Intelligence (~20% of entity), Progressive Spinal Technology (percent of entity unknown), Computational Biodynamics (percent of entity unknown), Stout Medical (B), Bonovo Orthopedics (100,000 Shares), Electrocure (D), Vertiflex (30,000 Shares), Flagship Surgical (D), In Vivo (percent of entity unknown), Pearl Diver (C), Small Bone Innovations (30,000 Shares), Neucore (22,000 Shares), Crosscurrent (125,000 Shares), Onset Medical (1,535 Shares), Orthovita (10,000 Shares), Syndicom (2,750 Shares), Biomerix (25,000 Shares), Paradigm Spine (146,875 Shares), Spinology (28,750 Shares), Gamma Spine (~15% of entity), Replication Medica (15,250 Shares), Breakaway Imaging (100,000 Shares), Crosstree Medical (41,667 Shares), Globus (816,123 Shares), FlowPharma (percent ownership of stock options unknown), Advanced Spinal Intellectual Properties (30% of entity), SpineMedica (number of shares unknown), RIS (!5,000 options); Consulting: Gerson Lehrman Group (B), Guidepoint Global (B), Mediacorp (B); Board of Directors: AO Spine (None, North America Board of Directors travel expenses stipend, B), ASIA (None, Past President Board member), NASS (None, Past Program Committee Co-chairman); Research Support—Staff and/or Materials: AO Spine (E), Cerapeutics (C); Fellowship Support: Stryker Spine (E); Other: NuVasive (C), Applied Spinal Intellectual Property (C), Confluent Surgical (B). FD: Nothing to disclose. DGA: Royalties: Medtronic (F), DePuy Spine (H); Consulting: DePuy Spine (D), Synthes (B); Speaking and/or Teaching Arrangements: DePuy Spine (D, consulting); Trips/Travel: DePuy Spine (D, consulting); Other Office: Society for Minimally Invasive Spinal Surgery (None, First Vice President). JAR: Other Office: Federation of Spine Associations (None, President). ASH: Royalties: Biomet Spine (F), Alphatec Spine (F), Stryker (C), Zimmer (C), Amedica (C), Aesculap (B); Stock Ownership: Amedica (20,000 Shares, !1%), Lifespine (20,000 Shares, !1%), Spinal Ventures (initial investment was 3% but has been diluted), Syndicom (20,000 Shares); Private Investments: Benvenue (B), Nexgen (B), Paradigm Spine (B), Pioneer (10,000 Shares, C), PSD (B), Vertiflex (B); Scientific Advisory Board: Amedica (20,000 shares of options). JSH: Stock Ownership: Axiomed (option to buy 15,000 shares); Consulting: Depuy Spine (C, Paid directly to institution/employer); Speaking and/or Teaching Arrangements: Medtronic (None, no longer a consultant for this company); Trips/Travel: Stryker (None); Board of Directors: Jefferson Medical College Physician Board (None); Scientific Advisory Board: Axiomed (None, Medical Advisory Board), Geron (None), Depuy (C, Paid directly to institution/employer), CNS (None, Executive Board, Chairman of Publication Committee, Editor of CNSQ, Chairman for Neurostimulation Project); Other Office: Pennsylvania Neurologic Society (None, Board of Pennsylvania Neurosurgical Society);
908
C.K. Kepler et al. / The Spine Journal 14 (2014) 903–908
Research Support—Staff and/or Materials: NACTN (E, Paid directly to institution/employer). TJA: Relationships Outside the 1-Year Requirement: CSRS (12/2008, Royalties, A). KER: Nothing to disclose. The disclosure key can be found on the Table of Contents and at www.TheSpineJournalOnline.com.
References [1] Hadley MN, Browner C, Sonntag VK. Axis fractures: a comprehensive review of management and treatment in 107 cases. Neurosurgery 1985;17:281–90. [2] Greene KA, Dickman CA, Marciano FF, et al. Acute axis fractures. Analysis of management and outcome in 340 consecutive cases. Spine 1997;22:1843–52. [3] Goldberg W, Mueller C, Panacek E, et al. Distribution and patterns of blunt traumatic cervical spine injury. Ann Emerg Med 2001;38:17–21. [4] Anderson LD, D’Alonzo RT. Fractures of the odontoid process of the axis. 1974. J Bone Joint Surg Am 2004;86-A:2081. [5] Grauer JN, Shafi B, Hilibrand AS, et al. Proposal of a modified, treatment-oriented classification of odontoid fractures. Spine J 2005;5:123–9. [6] Hanigan WC, Powell FC, Elwood PW, Henderson JP. Odontoid fractures in elderly patients. J Neurosurg 1993;78:32–5.
[7] Ryan MD, Taylor TK. Odontoid fractures in the elderly. J Spinal Disord 1993;6:397–401. [8] Polin RS, Szabo T, Bogaev CA, et al. Nonoperative management of types II and III odontoid fractures: the Philadelphia collar versus the halo vest. Neurosurgery 1996;38:450–6; discussion 456–7. [9] Kuntz C IV, Mirza SK, Jarell AD, et al. Type II odontoid fractures in the elderly: early failure of nonsurgical treatment. Neurosurg Focus 2000;8:e7. [10] Smith HE, Kerr SM, Maltenfort M, et al. Early complications of surgical versus conservative treatment of isolated type II odontoid fractures in octogenarians: a retrospective cohort study. J Spinal Disord Tech 2008;21:535–9. [11] Clark CR, White AA III. Fractures of the dens. A multicenter study. J Bone Joint Surg Am 1985;67:1340–8. [12] Harrop JS, Sharan AD, Przybylski GJ. Epidemiology of spinal cord injury after acute odontoid fractures. Neurosurg Focus 2000; 8:e4. [13] Hart R, Saterbak A, Rapp T, Clark C. Nonoperative management of dens fracture nonunion in elderly patients without myelopathy. Spine 2000;25:1339–43. [14] Koivikko MP, Kiuru MJ, Koskinen SK, et al. Factors associated with nonunion in conservatively-treated type-II fractures of the odontoid process. J Bone Joint Surg Br 2004;86:1146–51. [15] Vaccaro AR, Kepler CK, Kopjar B, et al. Functional and quality-oflife outcomes in geriatric patients with type-II dens fracture. J Bone Joint Surg Am 2013;95:729–35.