Comparison of Outcomes for Unstable Lower Cervical Flexion Teardrop Fractures Managed With Halo Thoracic Vest Versus Anterior Corpectomy and Plating

Transcription

1 Comparison of Outcomes for Unstable Lower Cervical Flexion Teardrop Fractures Managed With Halo Thoracic Vest Versus Anterior Corpectomy and Plating SPINE Volume 27, Number 2, pp , Lippincott Williams & Wilkins, Inc. Charles G. Fisher, MD, MHSc, FRCSC, Marcel F. S. Dvorak, MD, FRCSC, Jordan Leith, MD, FRCSC, and Peter C. Wing, MB, MSc, FRCSC Study Design. A retrospective cohort study with crosssectional outcome analysis was conducted. Objectives. To compare the outcome for two groups of patients with unstable cervical flexion teardrop fractures: those treated with halo thoracic vests and those treated with anterior corpectomy and plating. Summary of Background Data. With the evolution of safe and effective anterior cervical plates, the treatment of unstable cervical flexion teardrop fractures has shifted from halo immobilization to surgical stabilization. Although outcomes for these treatment alternatives have been reported, the literature reflects the inherent bias of retrospective studies without standardized health-related quality-of-life outcomes and without a control or comparative group. Furthermore, study populations have lacked homogeneity with respect to fracture patterns. Methods. For this study, 45 patients with cervical flexion teardrop fractures and at least 2 years of follow-up evaluation were identified. Of these patients, 24 were treated with a halo thoracic vest and 21 with anterior corpectomy and plating. Unstable cervical flexion teardrop fractures were defined as those exhibiting failure of the anterior spine under compression and the posterior spine in tension. The primary outcome was radiographic kyphosis at follow-up assessment. Secondary outcomes included the MOS 36-Item Short-Form Health Survey and the Cervical Spine Research Society Long-Term Follow-Up questionnaire. Results. The halo thoracic vest group and the anterior corpectomy and plating group were comparable for baseline demographic and clinical data, except for neurologic deficit (67% of the halo thoracic vest group and 96% of the anterior corpectomy and plating group had neurologic deficit). Most of the injuries occurred at C5. All 45 patients had radiographic follow-up evaluation, but only 17 of the 24 patients in the halo thoracic vest group and 13 of the 21 patients in the anterior corpectomy and plating group (30 of 45 in all) completed the health-related quality-of-life outcome instruments. The mean kyphosis was 11.4 in the halo thoracic vest group and 3.5 in the anterior corpectomy and plating group (P 0.001). The difference remained significant, with control used for the baseline variables. The halo thoracic vest group had five failures, four of which were subsequently managed operatively. No major intra- or postoperative complications occurred in the anterior corpectomy and plating group. There were From the Department of Orthopaedics, Division of Spine Surgery, Vancouver Hospital and Health Sciences Centre, Vancouver, British Columbia. Acknowledgment date: January 25, First revision date: April 30, Second revision date: June 22, Acceptance date: July 2, Device status category: 11. Conflict of interest category: 12. no significant differences in the MOS 36-Item Short-Form Health Survey mental component score and the Cervical Spine Research Society subscales even after adjustment for neurologic deficit. Conclusions. The results of this study indicate that anterior cervical plating is a safe and effective treatment for cervical teardrop fractures, and that it is superior to the halo thoracic vest for restoring and maintaining sagittal alignment and for minimizing treatment failures. There does not appear to be a relation between residual kyphosis and health-related quality-of-life outcomes. Although this raises questions about the relevance of restoring sagittal alignment in the treatment and outcome of cervical flexion teardrop fractures, the study was underpowered for secondary outcomes. Therefore, the secondary outcome results must be interpreted with caution. [Key words: cervical vertebrae, health-related quality of life, kyphosis, spinal fractures, spinal stabilization] Spine 2002;27: The optimal treatment for unstable cervical flexion teardrop fractures has been controversial. 10 With the evolution of sophisticated spine instrumentation and biomechanically sound surgical principles, these fractures are being treated surgically with greater frequency, although immobilization in a halo thoracic vest (HTV) still is commonly used. Lack of consensus about the ideal treatment for this fracture pattern prompted this review. The unstable cervical flexion teardrop fracture, often referred to as quadrangular fracture, is characterized by failure of the anterior column under compression and by tensile failure of the posterior column. This fracture pattern consists of a coronal split through the ventral part of the vertebral body, with dorsal displacement of the remaining vertebral body that leads to narrowing of the spinal canal. The posterior column fails in tension, which leads to distraction and subluxation of facet joints and/or increased interspinous distance often associated with fractures through the lamina or lateral mass (Figure 1A and 1B). In the mechanistic classification system of Allen et al, 2 this fracture is described as a compressive flexion Stage IV or V injury. Radiographically, the fracture can sometimes have a benign appearance, thereby necessitating vigilant clinical and radiographic assessment for effective diagnosis and treatment. Historically, the results of treatment for this fracture have been plagued by late instability, deformity, malunion, and neurologic deterioration. 5,17,18 Conservative treatment characteristically involves HTV immobilization, which results in union rates reportedly as high as 160

2 Unstable Lower Cervical Flexion Teardrop Fractures Fisher et al 161 Figure 1. Plain radiograph (A) and computed tomography scan (B) of C5 flexion teardrop fracture demonstrating coronal body split, posterior displacement of the vertebral body, and widening of facet joints. 78%. 5,16 18 Botte et al 4 reported local complications of HTV immobilization including pin-site loosening (36 60%), infection (20%), ring migration (13%), pain (18%), and loss of reduction in up to 40% of patients. Among cervical burst and flexion teardrop fractures treated in an HTV, 9% required late surgical intervention, and residual kyphosis was Surgical treatment alternatives include anterior and posterior stabilization procedures, either alone or combined. Posterior fusion with bone graft and wiring reportedly has a 45% nonunion rate and a residual kyphosis of With posterior wiring, it appears that neck pain is related to residual kyphosis. 7 Anterior interbody fusion without fixation followed by traction, an HTV, or both had a 95% union rate, but continued to be plagued with significant complications including graft extrusion, residual instability, and kyphosis. 7,23 Studies on posterior plating for trauma involved mixed cohorts with few teardrop fractures. 6,8 One study described a posterior reduction and plating technique for the cervical flexion teardrop fracture, but the cohort was too small for any inferences to be made. 19 The optimal reported results after treatment of unstable cervical spine fractures appear to result from anterior cervical plating (ACP) with autogenous bone graft (Figure 2). 1,9,11,13,16 Although biomechanical limitations have been described, 22 clinical studies show ACP to be a safe technique with a union rate approaching 100%, no residual kyphosis, and minimal symptomatic neck pain. 1,9,11,13,16 The outcomes in these studies reflect the inherent biases of retrospective studies including inadequate follow-up evaluation and lack of a control or comparative group. The other impediment that disallows firm conclusions regarding treatment efficacy is related to the heterogeneous study cohorts in terms of fracture pattern. To the authors knowledge, no assessments have included only unstable cervical flexion teardrop fractures. Furthermore, the studies have not used psychometrically sound health-related quality of life (HRQOL) questionnaires evaluated by an independent observer over at least a 2-year follow-up period. The objective outcomes used historically to evaluate the management of this fracture pattern have included residual kyphosis, osseous union, early and late instability, and neurologic recovery. Fusion remains difficult to assess because no valid and reliable diagnostic test is available. Neurologic recovery, a very germane clinical Figure 2. Plain radiograph after anterior C5 vertebrectomy, strut graft, and fusion with anterior cervical locking plate.

3 162 Spine Volume 27 Number outcome significantly influenced by numerous biologic, biomechanical, and clinical variables, is probably an inappropriate outcome in small cohorts composed of neurologically normal, incomplete, and complete patients. Clinically, failure of treatment often is defined, although arbitrarily, as a loss of alignment. Therefore, kyphosis seems to be the most appropriate of these variables. Kyphosis is accurately measured and thought to be related to patient outcome. 7,12,21 Maintaining overall sagittal balance in the spine has been postulated as critical in avoiding an escalating cycle of abnormal forces that may lead to progressive kyphotic deformity, compromise of the adjacent spinal motion segments, and potentially even neurologic symptoms. 21 Cervical lordosis recently has been measured in the normal population, with mean cervical lordosis reported at 40. Most cervical lordosis has occurred between C1 and C2, with only 6 of lordosis between C4 and C7. 12 The clinical relevance of cervical kyphosis in varying degrees is not fully understood. Standardized HRQOL questionnaires were developed to measure subjective data in a reliable and valid manner. The accurate collection of qualitative data is essential because it reflects what is relevant to the patient and society. Furthermore, it allows for evaluation of associations between so-called hard objective data and soft subjective data. In the Acute Spinal Cord Injury Unit (ASCIU) at Vancouver General Hospital two surgeons tended to favor HTV treatment for cervical flexion teardrop fractures on the basis of successful anecdotal experience and the perception that risks of surgery were being avoided. Two other surgeons favored ACP. This diversity of opinion regarding treatment is mirrored in the recent report by Glaser et al, 10 who demonstrated significant variability in surgeons preferred treatment of cervical fractures. It was this divergence of opinion that prompted the authors to review the outcome for these two forms of treatment in a group of patients with a single fracture pattern. The purpose of this study was twofold. The primary objective was to determine whether any significant difference existed in sagittal plane alignment after clinical and radiographic union in patients with unstable cervical burst fractures treated with a HTV and those treated with ACP. The secondary objective was to determine, using a generic and a disease-specific questionnaire, whether there was any difference in HRQOL. The authors designed a retrospective cohort study with crosssectional outcome assessment to answer these questions. Methods Study Protocol. A spine database generated retrospective cohort analysis was conducted for all subaxial cervical spine injuries treated by the ASCIU at Vancouver General Hospital from 1992 to Inclusion criteria required the presence of cervical flexion teardrop fractures with failure anteriorly in compression and posteriorly in tension with or without neurologic deficit. A 2-year follow-up period also was required for the patients included in the study. The exclusion criteria disallowed prohibitive medical comorbidity, a second contiguous cervical spine fracture, multiple trauma, or a seronegative spondyloarthropathy such as ankylosing spondylitis. Treatment Protocols. The decision to treat the patient with either HTV or ACP was made at the discretion of the treating surgeon. Two surgeons performed ACP, and two others used HTV management. Deviation from this practice pattern occurred only if there was neurologic deterioration or loss of function in patients treated by the physicians favoring HTV. In the surgeons preferring ACP, only patient preference resulted in HTV management. Medical management of these patients was provided on the ASCIU by four attending surgeons and the house staff. Once the patients were medically stabilized, and imaging (radiographs, computed tomography [CT] scan, magnetic resonance imaging [MRI]) was completed, those in the HTV treatment group were placed in halo traction for a variable period (1 to 28 days) to accomplish reduction with subsequent application of an HTV. Patients were then mobilized in the HTV, they underwent routine radiographic follow-up assessment. Adjustments were made in the HTV as necessary. Duration of HTV stabilization uniformly approximated 12 weeks. Surgically treated patients underwent identical medical management and imaging. They frequently were stabilized acutely with skeletal tong traction. Anterior cervical plating was performed by a standard anterior cervical approach after awake fiberoptic intubation. The affected vertebral body and adjacent discs were removed, and a complete decompression, including the posterior longitudinal ligament, was performed. The anterior column of the spine was reconstructed with a structural iliac crest autograft. The alignment of the spine with the structural graft was adjusted to ensure lordotic positioning and to avoid persistent distraction of the posterior elements, which could potentially have distracted the neurologic structures over the involved levels. This was accomplished most frequently by extending the neck to reposition the head. This repositioning was performed only after decompression. Fixation was obtained by applying an anterior cervical locking plate (CSLP; Synthes, Paoli, PA). The patients did not undergo posterior instrumentation or other procedures. After surgery, they were mobilized in a Philadelphia collar (Philadelphia Cervical Collar, Westville, NJ). Rehabilitation protocols were identical for both groups of patients. Radiographic Evaluation. The primary outcome was the kyphosis across the two injured motion segments at the time of clinical and radiographic union. Radiographs obtained for measurement were standard upright lateral cervical spine radiographs. If nonunion or instability was suspected clinically and radiographically, then lateral flexion extension radiographs were taken and reported by the surgeon. The radiographic follow-up period was not 2 years because follow-up evaluation reflected the standard clinical practice of routine radiographs until clinical radiographic follow-up assessment was achieved. If a patient in the HTV group failed halo management secondary to neurologic deterioration or, unable to achieve or maintain adequate alignment, was converted to surgical treatment, then the kyphosis in the postoperative radiograph was measured, with the patient remaining in the HTV group (intent-to-treat protocol). The radiographs were measured by two

4 Unstable Lower Cervical Flexion Teardrop Fractures Fisher et al 163 independent reviewers. The Cobb angle technique was used over the two injured motion segments. If there was disagreement, the radiographs were remeasured. If there still was no consensus, the average value was taken. Outcome Assessments. Institutional ethics approval to review the medical charts and contact the patients for administration of the HRQOL questionnaires was obtained from the University of British Columbia and Vancouver General Hospital. An independent research staff member contacted patients by telephone and administered the questionnaires. The secondary outcomes measurement included the MOS 36-Item Short- Form Health Survey (SF-36) and the Cervical Spine Research Society (CSRS) questionnaire. Short Form-36. The SF-36 is a generic HRQOL questionnaire that consists of 36 items and assesses eight health dimensions. A physical and mental component score can be derived. The physical component score ranges from 73 (high level of functioning) to 8 (low level of functioning), and the mental component score ranges from 74 (excellent state of mental health) to 10 (poor state of mental health). This questionnaire has been found reliable, valid, and responsive when administered over the telephone or by mail. 15,24 Cervical Spine Research Society Questionnaire. The CSRS, a recently developed disease-specific questionnaire, is available in baseline, short-term follow-up, and long-term follow-up versions. In this retrospective study, only the CSRS long-term follow-up evaluation questionnaire was administered because it had been at least 2 years since injury for all the patients at the time of the follow-up evaluation. Five subscale scores, involving those for neck pain, arm pain, symptoms, functional status, and psychological outlook, were derived, ranging from 0 (excellent state of health) to 100 (poor state of health) (personal communication, Dr. Bendebba, 1998). In addition, satisfaction with treatment was assessed by a single CSRS question: If your neck condition was to remain the same as it is now, how satisfied would you be? This question was scored using a 5-point scale ranging from 0 (extremely satisfied) to 4 (extremely dissatisfied), with a lower score indicating a higher level of satisfaction. At the time of this study, work was in progress to assess the reliability and validity of this question. Sample Size and Statistical Analysis. Sample size was calculated on the basis of a 5 effect change for a mean normal cervical sagittal angulation at C4 C7 of 1. It was thought that 5 was outside measurement error and probably clinically relevant, although to the authors knowledge, nothing in the literature identifies a degree of cervical kyphosis that would be recognized as clinically relevant. The study was appropriately powered for the primary outcome using an alpha of 0.05 and a beta of 0.1. In each group, 21 patients were needed. Statistical analysis for a comparison between the HTV and ACP groups was conducted using Fisher s exact test for categorical variables and Wilcoxon s rank-sum test for numerical variables. Multivariate linear regression analysis was conducted to assess the relation between kyphosis measure and treatment group (i.e., HTV or ACP) while adjusting for motor score. All P values were two-sided. Table 1. Baseline Variables HTV ACP Male (n) Female (n) 2 4 Age (yr) Mean SD (yr) Smoke (n) 9 10 Mechanism of injury (n) Sports 4 3 Fall from height 10 7 Motorcycle accident 2 3 Motor vehicle accident 8 8 Level of injury (n) C3 1 0 C4 3 0 C C6 3 7 C7 4 6 Motor score Mean SD Length of stay (days) Mean SD (days) Results Patient Demographics Review of the Spine Program database showed that 45 patients were eligible for the study on the basis of inclusion and exclusion criteria. There were 24 in the HTV group and 21 in the ACP group. The HRQOL questionnaires (67%) were completed by 30 patients, and 13 patients could not be located. One patient refused to participate, and one patient had died. The mean age was years in the HTV group and years in the ACP group. The two groups were comparable for baseline and demographic data, except for the American Spinal Injury Association (ASIA) motor score, 3 which represented a major neurologic deficit caused by a complete or incomplete spinal cord injury (Table 1). In the HTV group, 16 of the 24 patients (67%) had a neurologic deficit, as compared with 20 of the 21 patients (95%) in the ACP group. Complete tetraplegia (ASIA A and B) was present in 6 of the 24 patients (25%) in the HTV group and 15 of the 24 of the patients (63%) in the ACP group. This was associated with mean initial ASIA motor scores of in the HTV group and in the ACP group from a total possible score of 100 (neurologic normality). Most of the neurologic injuries occurred at C5. Five patients in the HTV group were classified as failures, four of which went on to require surgical treatment. All five failed when mobilizing in the HTV. Two of these patients experienced neurologic deterioration that required urgent surgical intervention. They both recovered neurologically after surgery. In three of the patients, the HTV failed to maintain satisfactory alignment as defined by the treating surgeon. Two of these patients required surgery. The third was treated with bed rest and halo traction for 6 weeks. The four patients who converted to surgical intervention were analyzed in the HTV group on the basis of an intent-to-treat philosophy. Their radio-

5 164 Spine Volume 27 Number Table 2. Secondary Outcomes HTV ACP P Value Short Form-36 Physical Component Score* Mean SD Mental Component Score* Mean CSRS Neck Pain Score Mean SD Arm Pain Score Mean SD Symptom Score Mean SD Functional Status Score Mean SD Psychological Outlook Score Mean SD Satisfaction [number (%)] Extremely satisfied 7 (41) 2 (15) Moderately satisfied 2 (12) 4 (31) Neither satisfied/dissatisfied 5 (29) 2 (15) Moderately dissatisfied 0 (0) 2 (15) Extremely dissatisfied 3 (18) 3 (23) * Norm-based scoring. PCS and MCS have a mean of 50 and a standard deviation of 10. PCS scores range from 73 (high level of functioning) to 8 (low level of functioning). MCS scores range from 74 (excellent state of mental health) to 10 (poor state of mental health). Scores range from 0 (excellent state of health) to 100 (poor state of health). Scores range from 0 (extremely satisfied) to 4 (extremely dissatisfied). graphic assessment was performed using postoperative films. Outcomes were collected after ACP, but they were analyzed in the HTV group. Radiographic Evaluation The entire sample of 45 patients had follow-up radiographs. The average radiographic follow-up period was months for the HTV group and months for the ACP group. The range of follow-up time was 2 months to 88 months. The mean kyphosis was 11.4 (range, 0 35 ) for the HTV group and 3.5 (0 14 ) for the ACP group. The difference was statistically significant (P 0.001). In the multivariate model, the mean kyphosis score was significantly lower in the ACP group than in the HTV group (P 0.016) after adjustment for the motor score. No significant interaction was detected between treatment group and motor score (P 0.350). Outcome Assessments Only 17 of the 24 patients in the HTV group and 13 of the 21 patients in the ACP group completed the HRQOL instruments, giving a response rate of 67%. The average HRQOL follow-up period was months for the HTV group and months for the ACP group. The range of follow-up time was 29 to 80 months. The mean SF-36 physical and mental component scores and the CSRS subscale scores, including the patient satisfaction score, were calculated for each of the two treatment groups (Table 2). There was no significant difference between the two groups in terms of either generic or disease-specific outcomes, with the exception of the SF-36 physical component score. Discussion Probably the most common treatment protocol for unstable cervical flexion teardrop fractures with compressive failure of the anterior column and distraction through the posterior elements involves vertebrectomy, anterior strut graft, and anterior cervical plating. The results of this study suggest that compared with HTV immobilization, ACP is the optimal treatment for restoring and maintaining sagittal alignment, and for minimizing treatment failures secondary to neurologic deterioration or loss of alignment. This surgical finding is consistent with that of noncomparative surgical studies in the literature. 1,9,11,16 In the current study, 100% clinical and radiographic union was obtained in the patients treated operatively. There were no major early or late local postoperative complications, which can include neurologic deterioration, vascular injury, esophageal or tracheal injury, instrumentation failure, or nonunion. In the only retrospective comparative study for flexion teardrop and cervical burst fractures, 2 of 35 patients required a second operation for screw complications. 13 The same study found final alignment to be 2.2 in the operative group and 12.6 in the HTV group, similar to the findings in the current study, but not comparable because the investigators mixed different fracture types. 13 Patients treated with an HTV had an average residual kyphosis of Two patients failed HTV treatment because of neurologic deterioration. Two patients, because of HTV inability to obtain and maintain reduction, were converted to surgical intervention. The neurologic deteriorations were a motor loss of one grade in a central cord patient and episodic weakness and numbness in a previously neurologically normal patient. The neurologic changes resolved with surgical intervention. The radiographic and outcome assessment, performed after surgical treatment in the four patients converted to surgical intervention, was included in the HTV group analysis in an intent-to-treat type philosophy, which tended to bias the results toward the null hypothesis. According to the null hypothesis, there would be no significant difference in the sagittal alignment between the two treatment groups. The average kyphosis of the four patients converted to surgery was 4, supporting this bias. Despite this, a significant difference remained between the two treatment groups with respect to kyphosis. The sagittal weight-bearing axis lies posterior to the bodies of C2 and C7, promoting cervical lordosis. Shifting of this axis anterior puts increased biomechanical stress on the posterior dynamic neck stabilizers keeping the head upright and sagittally balanced. As kyphosis increases, so does the flexion movement, thereby increasing the likelihood of progressive kyphosis. Compensatory changes in the thoracic and lumbar spine occur to maintain overall sagittal alignment. 12 Theoretically, the clinical sequelae of these compensatory changes are pain and neurologic deficit.

6 Unstable Lower Cervical Flexion Teardrop Fractures Fisher et al 165 The relation between residual kyphosis and functional outcome has yet to be determined in the literature. White and Punjabi 27 have used kyphosis in developing guidelines for instability in a cadaver model. Sears and Fazl 18 have shown kyphosis to be a predictor for potential failure of HTV treatment. Hardacker et al 12 clearly defined normal overall and segmental cervical alignment, looking at associations with age and symptoms. Despite this work, the effect of sagittal plane deformity on overall patient function and subjective outcome remains to be determined. In the current study, the significant difference in residual kyphosis between the two groups did not seem to have a negative influence on the HRQOL outcome. The statistically significant difference in the SF-36 physical component score probably was related to the large number of patients in the ACP group with severe neurologic deficits. The fact that four patients in the HTV group crossed over to the ACP group but were included in the HTV group for data analysis also may have influenced the results of the generic and disease-specific HRQOL outcomes. In other words, if these four patients had been analyzed in the ACP group, then the HRQOL outcomes possibly would have been different between the two groups. Because HRQOL was not the primary outcome and thus was poorly powered, subgroup analysis was not performed. The SF-36 mental component scores were essentially the same in the two groups, despite a significantly higher proportion of serious neurologic impairment in the surgically treated group. This finding appears to be consistent with the literature. 20,26 The CSRS satisfaction with treatment was generally better in the HTV group, and in all other CSRS subscales except the arm pain subscale, for which the HTV scores were higher than the ACP scores. These results must be interpreted with caution. They are hypothesis-generating at best. The study is underpowered for an investigation of these variables, and because of the low patient follow-up assessment with the HRQOL instruments (i.e., 67% of the sample), any inferences from the results are possibly not valid. The issue is raised, however, whether kyphosis affects clinical outcome from a HRQOL perspective. An analogy can be easily drawn to the thoracolumbar spine, in which numerous long-term studies have shown no detrimental relation between increased thoracolumbar kyphosis and outcome in magnitudes up to ,25 The average follow-up period in the current study was 55 months, and the potential biomechanically detrimental sequelae of kyphosis may manifest years later. Other issues may revolve around age, arthritic changes, and ability of the remaining cervical spine or thoracic and lumbar spine to compensate to maintain overall sagittal alignment. Additionally, the ACP group may eventually have less mobility with the fusion of two vertebral motion segments. Unfortunately, at the time of the current study, there was no validated disease-specific outcome instrument for the cervical spine, and the CSRS instrument was still in the development stage. It had not yet been rigorously tested with respect to reliability, validity, and responsiveness. Subsequently, the American Academy of Orthopedic Surgeons and the North American Spine Society published a disease-specific outcome measure for the cervical spine. Any future studies probably should use this instrument, with the caveat that it may not yet be validated for a population of trauma patients. Other variables collected, such as level and mechanism of injury, did not seem to influence either the primary or secondary outcomes. Similarly, with multivariate linear regression analysis, there was no influence of any variables other than treatment on kyphosis, including the ASIA motor score. The authors failed to observe the initial kyphosis and the subtleties of this distinct fracture pattern, which may have affected failure and end-point kyphosis. Neurologic outcome, although of considerable interest and patient relevance, was not studied as a primary or secondary outcome in this study. The cohorts were small and heterogeneous with respect to initial neurologic status, with nine patients in the ASIA E category (neurologically normal) and six patients classified as ASIA D (minimal deficit). Valid evaluation would have necessitated subgroup assessment of complete and incomplete patients, further diluting the power of the analysis. Furthermore, there would need to be control for probable confounders such as age, initial motor score, and mechanism of injury, and this is difficult with such small numbers. The appropriately powered primary outcome, a homogeneous cohort from a fracture pattern perspective, and the use of psychometrically solid outcomes are unique strengths of this study relative to the literature. Nevertheless, considerable limitations exist. The study was retrospective, so data collection, treatment, and diagnostic techniques were not standardized and controlled in the manner of a prospective study. The radiographic follow-up period was not 2 years, but this probably is not relevant because clinical and radiographic union usually is apparent 2 to 4 months after intervention. Strong treatment selection bias occurred, with the large preponderance of neurologically impaired patients in the surgical ACP group. There was only a 67% response rate for the HRQOL questionnaires, and this was well below the recognized gold standard of 80%. Considerable doubt exists about whether the HRQOL outcome results are generalizable to the study population. The use of these outcomes, however, appears to be feasible and has contributed potentially to a normative data set. Furthermore, it hopefully will influence more studies to use standardized objective outcome measures that are relevant to the patient. In conclusion, this study has contributed support to anterior cervical strut graft and plating as a safe and effective technique for managing the unstable cervical flexion teardrop fracture, providing excellent radiographic outcome without fixation failures. This study also has generated questions regarding the relation of

7 166 Spine Volume 27 Number residual kyphosis in the cervical spine to patient outcome. Acknowledgments The authors acknowledge the assistance of Vanessa Noonan, MSc (Spine Program Research Coordinator), in the conduct of this study. Key Points This retrospective study compared two treatment groups for one fracture type. Multivariate regression was used as a control for variability within treatment groups, negating the need for matching. Cross-sectional psychometrically sound outcomes with at least a 2-year follow-up period were available. An appropriately powered primary outcome was observed. Conclusions pointed to anterior cervical plating as effective treatment with respect to restoring normal alignment and reducing failures. The findings promote hypothesis-generation with respect to healed sagittal alignment and HRQOL outcomes. References 1. Aebi M, Zuber K, Marchesi D. Treatment of cervical spine injuries with anterior plating: indications, techniques, and results. Spine 1991;16(Suppl): S Allen BL, Ferguson RL, Lehmann TR, et al. A mechanistic classification of closed, indirect fractures and dislocations of the lower cervical spine. Spine 1982;7: American Spinal Injury Association. International Standards for Neurological and Functional Classification of Spinal Cord Injury. Revised Chicago: American Spinal Injury Association, Botte MJ, Byrne TP, Abrams RA, et al. The halo skeletal fixator: current concepts of application and maintenance. Orthopedics 1995;18: Bucci MN, Dauser RC, Maynard FA, et al. Management of posttraumatic cervical spine instability: operative fusion versus halo vest immobilization: analysis of 49 cases. J Trauma 1988;28: Ebraheim NA, Rupp RE, Savolaine ER, et al. Posterior plating of the cervical spine. J Spinal Disord 1995;8: Favero KJ, Van Peteghem PK. The quadrangular fragment fracture: roentgenographic features and treatment protocol. Clin Orthop Rel Res 1989;239: Fehlings MG, Cooper PR, Errico TJ. Posterior plates in the management of cervical instability: long-term results in 44 patients. J Neurosurg 1994;81: Garvey TA, Eismont FJ, Roberti LJ. Anterior decompression, structural bone grafting, and Caspar plate stabilization for unstable cervical spine fractures and/or dislocations. Spine 1992;17(Suppl):S Glaser JA, Jaworski BA, Cuddy BG, et al. Variation in surgical opinion regarding management of selected cervical spine injuries: a preliminary study. Spine 1998;23: Goffin J, van Loon J, Van Calenbergh F, et al. Long-term results after anterior cervical fusion and osteosynthetic stabilization for fractures and/or dislocations of the cervical spine. J Spinal Disord 1995;8: Hardacker JW, Shuford RF, Capicotto PN, et al. Radiographic standing cervical segmental alignment in adult volunteers without neck symptoms. Spine 1997;22: Koivikko MP, Myllynen P, Karjalainen M, et al. Conservative and operative treatment in cervical burst fractures. Arch Orthop Trauma Surg 2000;120: Magerl F, Aebi M, Gertzbein SD, et al. A comprehensive classification of thoracic and lumbar injuries. Eur Spine J 1994;3: McHorney CA, Ware JE, Rachel JF, et al. The MOS 36-Item Short-Form Health Survey (SF-36): III. Tests of data quality, scaling, assumptions, and reliability across diverse patient groups. Med Care 1994;32: Ripa DR, Kowall MG, Meyer PR Jr, et al. Series of ninety-two traumatic cervical spine injuries stabilized with anterior ASIF plate fusion technique. Spine 1991;16(Suppl):S Rockswold GL, Bergman TA, Ford SE. Halo immobilization and surgical fusion: relative indications and effectiveness in the treatment of 140 cervical spine injuries. J Trauma 1990;30: Sears W, Fazl M. Prediction of stability of cervical spine fracture managed in the halo vest and indications for surgical intervention. J Neurosurg 1990;72: Signoret F, Jacquot FP, Feron JM. Reducing the cervical flexion teardrop fracture with a posterior approach and plating technique: an original method. Eur Spine J 1999;8: Smith DE, Dvorak MF, Fisher CG, et al. Neurological and Functional Outcomes in Complete Spinal Cord Injury: A Follow-up Study. Presented at the North American Spine Society Meeting, Chicago, Illinois, October 23, Spivak JM, Glordana CP. Cervical kyphosis. In: Bridwell K, Ducked R, eds. The Textbook of Spinal Surgery. 2nd ed. Philadelphia: Lippincott-Raven, 1997: Sutterlin CE, Mc Affee PC, Warden KE, et al. A biomechanical evaluation of cervical spinal stabilization methods in a bovine model: static and cyclical loading. Spine 1988;13: van Peteghem PK, Schweigel JF. The fractured cervical spine rendered unstable by anterior cervical fusion. J Trauma 1979;19: Ware JE, Sherbourne CD. The MOS 36-Item Short-Form Health Survey (SF-36). I. Conceptual framework and item selection. Med Care 1992;30: Weinstein JN, Collalto P, Lehmann TR. Thoracolumbar burst fractures treated conservatively: a long-term follow-up. Spine 1988;13: Westgren N, Levi R. Quality of life and traumatic spinal cord injury. Arch Phys Med Rehabil 1998;79: White AA, Punjabi MM. Clinical Biomechanics of the Spine. Chapter 5. Part 2. Philadelphia: JB Lippincott, 1978: Address reprint requests to Charles G. Fisher, MD, MHSc, FRCSC D605 Heather Pavilion 2733 Heather Street Vancouver, British Columbia V5Z 3J5 cfisher@vanhosp.bc.ca