Skip to main content
NCBI home page
Clinical Orthopaedics and Related Research logoLink to Clinical Orthopaedics and Related Research
. 2018 Feb 22;476(7):1506–1513. doi: 10.1007/s11999.0000000000000196

Incidence of Neuraxial Abnormalities Is Approximately 8% Among Patients With Adolescent Idiopathic Scoliosis: A Meta-analysis

Michael Faloon 1, Nikhil Sahai 1, Todd P Pierce 1, Conor J Dunn 1, Kumar Sinha 1, Ki Soo Hwang 1, Arash Emami 1,✉
PMCID: PMC6437592  PMID: 29470234

Abstract

Background

Several studies have sought to address the role of routine preoperative MRI in patients with adolescent idiopathic scoliosis (AIS) undergoing deformity correction. Despite similar results regarding the prevalence of neuraxial anomalies detected on MRI, published conclusions conflict and give opposing recommendations. Lack of consensus has led to important variations in use of MRI before spinal surgery for patients with AIS.

Questions/purposes

This systematic review and meta-analysis of studies about patients with AIS evaluated (1) the overall proportion of neuraxial abnormalities; (2) the patient factors and curve characteristics that may be associated with abnormalities; and (3) the proportion of patients who underwent neurosurgical intervention before scoliosis surgery and the kinds of neuraxial lesions that were identified.

Methods

We performed a search of four electronic databases (PubMed, EMBASE, CINAHL Plus, and SCOPUS) utilizing search terms related to routine MRI and AIS, yielding 206 articles. Studies included had at least 20 participants, patients with ages 11 to 21 years, and a Methodological Index for Non-Randomized Studies (MINORS) study quality score of 8 and 16 points for noncomparative and comparative studies, respectively. Non-English manuscripts, animal studies, and those that did not include patients with AIS solely were excluded. Eighteen articles with 4746 patients were included for analysis of the overall proportion of neuraxial abnormalities, 12 articles with 3028 patients for analysis by sex, eight articles with 1603 patients for right main thoracic curve, eight articles with 665 patients for a left main thoracic curve, and 13 articles with 3063 patients and 230 (7.5%) abnormalities for number of neurosurgical interventions before scoliosis correction. The mean MINORS score for studies included was 14 (range, 10-20). Each study was analyzed for the proportion of patients identified with neuraxial abnormalities and associations with specific demographics. We determined the proportion of patients who underwent surgical interventions before scoliosis surgery as well as the types of neuraxial lesions identified. The articles were assessed for heterogeneity and publication bias. Because all groups were determined to be heterogeneous, a random-effects model was used for each group in this meta-analysis; with this analysis, an overlap of 95% confidence intervals suggests no difference at the p < 0.05 level, but this analytic approach does not provide p values.

Results

The pooled proportion of neuraxial abnormalities detected on MRI was 8% (95% confidence interval [CI], 6%-12%). With the numbers available, we found no difference in the proportion of male and female patients with neuraxial abnormalities (18% [95% CI, 11%-29%] versus 9% [95% CI, 6%-12%], respectively). Likewise, there was no difference in the proportion of pooled neuraxial abnormalities in right and left curves (9% [95% CI, 6%-14%] versus 15% [95% CI, 5%-35%], respectively). In the subset of abnormalities analyzed for number of neurosurgical interventions before scoliosis correction, the pooled proportion showed that 33% (95% CI, 24%-43%) underwent neurosurgical intervention before deformity correction. The most common abnormalities of the 367 found on MRI were syringomyelia in 127 patients (35%), Arnold-Chiari Type 1 malformation with syrinx in 103 patients (28%), and isolated Arnold-Chiari Type 1 malformation in 91 patients (25%).

Conclusions

The proportion of patients with AIS who have neuraxial abnormalities is high (8%) and a large number undergo surgical intervention before scoliosis reconstruction. We did not find any particular demographic variables that indicated an increased risk of abnormality. Clinicians should consider advanced imaging before surgical intervention in the treatment of a patient with an idiopathic diagnosis. Preventable variables need to be identified by future studies to establish a better working treatment protocol for these patients.

Level of Evidence

Level III, diagnostic study.

Introduction

The diagnosis of adolescent idiopathic scoliosis (AIS) comes with a certain degree of accepted etiologic uncertainty [9]. Importantly, however, studies have shown that some patients with AIS also have neuraxial abnormalities [8, 10-12]. Patients referred for AIS who show clinical or radiographic evidence of an atypical curve may be further evaluated by MRI of the brain and spine [2, 5]. However, many patients who do not show these signs still have neuraxial abnormalities detected on preoperative MRI [7, 16]. Although not all diagnosed neuraxial abnormalities result in surgical intervention, failure to identify these abnormalities before scoliosis surgery can have catastrophic consequences [13, 24]. The proportion of patients with neuraxial abnormalities found on preoperative MRI in patients with AIS undergoing surgery ranges from 2.1% to 14%, and the proportion of those patients who undergo surgical intervention for an underlying abnormality before scoliosis correction varies from 0% to 55% [1, 4, 6, 18, 22]. Furthermore, reports have varied regarding the proportion of abnormality seen in right and left curves [16, 24].

Recommendations for routine preoperative MRI screening of patients with AIS undergoing reconstructive surgery vary among studies and often directly contradict one another [1, 3,4,5,6, 16, 21, 22, 25]. No meta-analysis or systematic review, to our knowledge, has pooled what seems to be a discrepant evidence base to see whether there are any reliable inferences about the risk of neuraxial abnormalities among patients with AIS. Some studies recommend routine preoperative MRI [1, 16, 22], whereas others recommend selective screening by presence of abnormal curve characteristics or younger age despite similar patient presentations [1, 3,4,5,6, 13, 21, 23]. Therefore, it is important to pool the available data of neuraxial abnormalities as well as describe the particular lesions themselves. Although many lesions may not be clinically relevant, the proportion that may benefit from prior surgical intervention, or change the preoperative treatment plan, is currently unknown.

We therefore performed a systematic review and meta-analysis on the topic of neuraxial abnormalities among patients with AIS to determine (1) the overall proportion of neuraxial abnormalities; (2) the patient factors and curve characteristics that may be associated with abnormalities; and (3) the proportion of patients who underwent neurosurgical intervention before scoliosis surgery and the kinds of neuraxial lesions that were identified.

Patients and Methods

Search Strategy and Criteria

Exemption for institutional review board was received for this study.

A thorough search of four large electronic databases (PubMed, EMBASE, CINAHL Plus, and SCOPUS) was performed to identify any study that evaluated the role of routine preoperative MRI in patients with a diagnosis of AIS. All studies published from January 1976 to September 2015 were evaluated using a combination of the Boolean search strings as follows: routine[title] OR magnetic[title] OR resonance[title] OR imaging[title] OR MRI[title] AND adolescent[title] AND idiopathic[title] AND scoliosis[title].

Two authors (NS, TPP) evaluated each study independently for inclusion in this review. Each study was assessed for quality using the Methodological Index for Non-Randomized Studies (MINORS) criteria [23], a validated system for evaluating the quality of comparative and noncomparative studies. A senior author (AE) served as adjudicator on three studies in which the two authors did not reach the same value.

We applied the following inclusion criteria: (1) at least 20 participants; (2) patients with ages of 11 to 21 years; (3) MINORS study quality score of 8 and 16 points for noncomparative and comparative studies, respectively; and (4) with routine MRI performed for a diagnosis of AIS. Additionally, indications for MRI were routine or per study protocol (curves with atypical features) [23]. Our exclusion criteria were (1) non-English manuscripts; (2) animal studies; and (3) any addressing metabolic disease, genetic abnormality, posttraumatic scoliosis, congenital scoliosis, neurologic injury, or cerebral palsy.

The initial search yielded 206 articles. After application of the inclusion and exclusion criteria, 18 articles with 4746 patients were included in this meta-analysis (Fig. 1) [1, 3,4,5,6,7,8, 10, 12, 14,15,16,17, 19, 21, 22, 25, 26]. Each study was analyzed for the proportion of neuraxial abnormalities identified. In addition, we evaluated for a potential association between the proportion of neuraxial abnormalities found and curve characteristics and demographics. Finally, we determined the proportion of change in surgical plan for deformity correction if an abnormality was present. All 18 studies and 4746 patients were included for examination of the proportion of neuraxial abnormalities present. The mean age of all patients was 13.8 years (range, 11.2-15.5 years); the mean age was not documented in one study. Twelve studies were included for sex (3028 patients). Eight studies were used for right and left main thoracic curves (1603 and 665 patients, respectively). Thirteen studies were used to pool the proportions of abnormalities undergoing neurosurgical intervention before reconstructive scoliosis surgery (3063 patients).

Fig. 1.

Fig. 1

Open in a new tab

The flowchart demonstrates how 18 reports ultimately were included in this meta-analysis.

The mean MINORS score for all 18 studies was 14 (range, 10-20). For the 12 studies included for sex, the mean MINORS score was 15 (range, 11-20). For the eight studies used for right and left main thoracic curves, the mean MINORS score was 16 and 15 (range, 10-19 and range, 11-18), respectively. Lastly, the mean MINORS score for the 13 studies included for analysis of the proportion of abnormalities undergoing neurosurgical intervention before reconstructive scoliosis surgery was 15 (range, 10-20). Cochran’s Q test, a chi-square based test, was used to assess the heterogeneity within all included studies (between-study variability) in the meta-analysis. A significance level of < 0.05 was used for the statistical test of heterogeneity. Although the Q statistic provided the qualitative information regarding the existence of heterogeneity, I2, calculated as the percentage of the total variability in a set of effect sizes attributable to true heterogeneity, was used to quantify the degree of heterogeneity for each study population. High-resolution forest plots were derived to estimate the pooled proportions of abnormality or odds ratio (OR) corresponding to the 95% confidence interval (CI).

For the 4746 patients included for analysis in study question 1 (overall proportion of neuraxial abnormalities found), the Cochran’s Q was 200.34, which was statistically significant (p < 0.001). The 18 included studies in study question 1 are highly heterogeneous (I2 = 91.5%; 95% CI, 88%-94%; p < 0.01) (Fig. 2). The studies utilized for right and left curve populations were highly heterogeneous as well (I2 = 86% [95% CI, 74%-92%], p < 0.01; I2 = 93.4% [95% CI, 89%-96%], p < 0.01) (Figs. 3, 4), respectively. Likewise, the 12 studies used in analysis for sex are highly heterogeneous (I2 = 84% [95% CI, 73%-90%], p < 0.01). Lastly, the 13 studies in study question 3 are highly heterogeneous (I2 = 44%; 0%-81%; p < 0.01). A random-effects model of meta-analysis was used for the data of the current review, because each study population was heterogeneous. Publication bias was assessed utilizing funnel plots for each population and generally demonstrated greater asymmetry with studies of smaller sample sizes, consistent with publication bias in these studies (Figs. 5-9).

Fig. 2.

Fig. 2

Open in a new tab

The pooled proportion of neuraxial abnormalities utilizing a random-effects model is 8% (95% CI, 6%-12%).

Fig. 3.

Fig. 3

Open in a new tab

The pooled proportion of neuraxial abnormalities in right thoracic curves utilizing a random-effects model is 9% (95% CI, 6%-14%).

Fig. 4.

Fig. 4

Open in a new tab

The pooled proportion of neuraxial abnormality in right thoracic curves utilizing a random-effects model is 15% (95% CI, 5%-35%).

Fig. 5.

Fig. 5

Open in a new tab

Funnel plot for patients with an abnormality demonstrating greater asymmetry with studies of smaller sample sizes, consistent with publication bias in these studies.

Fig. 9.

Fig. 9

Open in a new tab

Funnel plot for patients with surgical intervention before reconstruction and an abnormality demonstrating greater asymmetry with studies of smaller sample sizes, consistent with publication bias in these studies.

Fig. 6.

Fig. 6

Open in a new tab

Funnel plot for sex and findings of an abnormality demonstrating symmetry in studies with large sample sizes and only one study of smaller size.

Fig. 7.

Fig. 7

Open in a new tab

Funnel plot for patients with a right curve and abnormality demonstrating greater asymmetry with studies of smaller sample sizes, consistent with publication bias in these studies.

Fig. 8.

Fig. 8

Open in a new tab

Funnel plot for patients with a left curve and abnormality demonstrating greater asymmetry with studies of smaller sample sizes, consistent with publication bias in these studies.

Statistical Analysis

We entered all outcome metrics into an Excel spreadsheet (Excel 2011; Microsoft Corporation, Redmond, WA, USA). All analyses were done using statistical software R version 3.2.0 (The R Development Core Team, The R Foundation, in Vienna, Austria). For this review, events proportion with 95% CI was chosen. With this analysis in the setting of a random-effects model, an overlap of 95% CIs suggests no difference at the p < 0.05 level, but this analytic approach does not provide p values.

Results

The pooled proportion of abnormality detected on preoperative MRI in 18 studies and 4746 patients was 8% (95% CI, 6%-12%) (Fig. 2).

With the numbers available, there were no differences in the proportion of male and female patients with neuraxial abnormalities (18% [95% CI, 11%-29%] versus 9% [95% CI, 6%-12%], respectively). Additionally, there was no difference in the proportion of right and left curves in patients with neuraxial abnormalities (9% [95% CI, 6%-14%] [Fig. 3] versus 15% [95% CI, 5%-35%] [Fig. 4], respectively).

The pooled proportion of patients with neuraxial abnormalities who underwent neurosurgery before scoliosis surgery was 33% (95% CI, 24%-43%). These patients comprise 3% (95% CI, 2%-4%) of all 3063 patients included in this study population. The most common abnormalities of the 367 found on MRI were syringomyelia in 127 patients (35%), Arnold-Chiari Type 1 malformation with syrinx in 103 patients (28%), isolated Arnold-Chiari Type 1 malformation in 91 patients (25%), and tethered cord with or without syrinx in 17 patients (5%). Consequently, almost all 75 patients who underwent neurosurgical decompression had one or a combination of these diagnoses. One patient underwent neurosurgery before curve correction for diastematomyelia, one for syringomyelia with tonsillar ectopia, and one for a spinal cord tumor.

Discussion

Idiopathic scoliosis implies a certain level of uncertainty as a diagnosis. Even so, a considerable number of patients with idiopathic scoliosis who undergo surgery have diagnosable neuraxial abnormalities. Some studies have suggested that certain “atypical findings” in adolescents with idiopathic scoliosis have an increased risk of having an underlying neuraxial abnormality [1, 8, 26], whereas other studies fail to corroborate these findings [10, 17, 22]. The difference may result from insufficient power, which might benefit from meta-analysis to help resolve [1, 9,]. Our study shows an 8% pooled proportion of abnormalities across 18 studies including 4746 patients. We were unable to demonstrate that specific demographics were associated with the proportion of abnormality. Our findings indicate that previously asserted risk factors for neurologic abnormalities in AIS did not consistently demonstrate greater risk than those without them.

Our study has several limitations. We were unable to examine whether typical or atypical curves are associated with neuraxial abnormality, because not enough studies classified their patients in this way. Furthermore, we were not able to examine whether patients with a neurologic deficit on preoperative physical examination were more likely to have a diagnosable neuraxial abnormality, because this was not clearly stated in most of the studies that we analyzed. This limitation does not diminish our findings in any way, however, and only limits the scope of our analysis. The 20-year time range of included studies (1996-2015) raises the possibility of several other limitations including changes in MRI sensitivity for detecting abnormalities and variations in interpretations of the test. We believe that the radiologic limitations of earlier studies do not discredit their findings or reduce the value of including them in this meta-analysis, because neuraxial abnormalities were well documented and described in similar proportion to more recent studies. Routine preoperative MRI was used in 12 studies and atypical features detected on clinical examination or radiographs were used in six, per institutional protocol. It is possible that these six studies that used MRI only for specific indications may have resulted in an inflated estimate in our study of the risk of a neuraxial abnormality (which we calculated to be 8%), and thus limited our ability to generalize well. However, we believe this was not the case because all studies included only examined the subset of patients with AIS with operative indications. Additionally, five of the six studies had proportions between 4.2% and 9.9% with the last being a study by Morcuende et al. [14] in which 11 of 72 patients with AIS (15.3%) underwent MRI for the indication of atypical clinical or radiographic features. Thus, despite heterogeneity in studies, these studies with “atypical” indications are consistent with the overall findings of our pooled population. Furthermore, the consistency of findings in these studies with those that utilized routine MRI makes it less likely that our findings were substantially influenced by publication bias (positive outcome bias). Lastly, it is possible that despite pooling a large number of patients’ data, our analysis may have been underpowered. For example, we found a pooled proportion of 18% abnormalities in males and 9% in females, although with the numbers available, these point estimates were not different.

Our analysis demonstrated that approximately 8% of 4746 patients across 18 studies had neuraxial abnormalities (Fig. 2). Given such a high proportion of abnormality present in this population, the clinician treating patients with this supposed idiopathic diagnosis may wish to consider ordering an MRI routinely. Future studies should be conducted with multicenter cooperation at hospitals with large pediatric patient populations and aim to establish identifiable variables for scoliosis progression before patients become operative candidates.

It generally is accepted that MRI is warranted in the workup of patients with scoliosis with atypical features, including a left thoracic curve [2, 20]. Although we did find a greater pooled proportion of neuraxial abnormality in left thoracic curves (Fig. 4) than right thoracic curves (Fig. 3), our analysis could not identify a difference (15% [95% CI, 5%-35%] versus 9% [95% CI, 6%-14%], respectively). Additionally, we did not find a difference in the proportion of males or females with an underlying neurologic abnormality (18% [95% CI, 11%-29%] versus 9% [95% CI, 6%-12%], respectively). It is possible that this analysis may have been underpowered and, thus, unable to demonstrate a difference. These findings suggest, however, that atypical features may not indicate as dramatic of an increased risk of abnormality by some studies. It is also possible that patients without these features may have a higher risk of abnormality than previously acknowledged [2, 20]. Either way, our analysis suggests that atypical characteristics do not demonstrate a difference that warrants a distinct approach to the diagnostic or treatment protocol.

In addition to the proportion of neuraxial anomalies, studies also disagree on whether preoperative MRI changes the surgical plan for deformity correction [1, 3, 4, 6]. Of the 13 studies with data on change of surgical plan, eight reported a change in surgical plan and five reported none with a preoperative neurosurgical intervention. Types of abnormality identified were similar between both groups (change in plan versus no change in plan), but no indication of disease severity was given. Therefore, it was difficult to objectively determine which patients needed preoperative neurosurgery or a change in surgical planning. Furthermore, neurosurgical intervention indications may vary by institution, geographic location, and time. Nonetheless, our analysis showed that a very high proportion of the abnormality cohort (33%, in our pooled analysis) experienced a change in surgical plan through preoperative decompression or intraoperative management may warrant reconsideration of adding MRI to the standard preoperative workup. Interestingly, two of the studies with the highest proportion of neurosurgical intervention in abnormalities detected (eight of 18 [55%] and 13 of 26 [50%]) gave opposing recommendations as to whether to routinely order an MRI for preoperative patients with AIS [4, 22]. Singhal et al. [22] advocated for universal preoperative MRI, whereas Davids et al. [4] advocated selective screening. This finding highlights important discrepancies in best practice, because researchers with nearly identical findings (both studies on this point found that approximately half of their patients underwent neurosurgical intervention before elective scoliosis surgery) do not agree about key elements of screening. Our study suggests that currently there are insufficient data to recommend selective screening. Therefore, we recommend screening all preoperative patients with AIS.

In this meta-analysis, we found that the overall proportion of neuraxial abnormalities found on MRI in patients with AIS is high (8% in a pooled analysis of > 4000 patients) and we were unable to identify particular patient or curve characteristics that were associated with underlying neuraxial abnormalities. Clinicians should consider advanced imaging before moving forward with surgical intervention in the treatment of a patient with an idiopathic diagnosis. In determining an accepted level of uncertainty while moving forward with surgical intervention in the treatment of AIS, it is imperative to establish a zero tolerance for potentially preventable catastrophic neuraxial complications of this elective procedure. Future studies conducted at centers with large pediatric patient populations should aim to establish risk factors before patients become operative candidates.

Footnotes

Each author certifies that neither he, nor any member of his immediate family, has funding or commercial associations (consultancies, stock ownership, equity interest, patent/licensing arrangements, etc) that might pose a conflict of interest in connection with the submitted article.

All ICMJE Conflict of Interest Forms for authors and Clinical Orthopaedics and Related Research® editors and board members are on file with the publication and can be viewed on request.

Each author certifies that his institution waived the human protocol for this investigation and that all investigations were conducted in conformity with ethical principles of research.

References

  • 1.Ameri E, Andalib A, Tari HV, Ghandhari H. The role of routine preoperative magnetic resonance imaging in idiopathic scoliosis: a ten years review. Asian Spine J. 2015;9:511–516. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Barnes PD, Brody JD, Jaramillo D, Akbar JU, Emans JB. Atypical idiopathic scoliosis: MR imaging evaluation. Radiology. 1993;186:247–253. [DOI] [PubMed] [Google Scholar]
  • 3.Benli IT, Uzumcugil O, Aydin E, Ates B, Gurses L, Hekimoglu B. Magnetic resonance imaging abnormalities of neural axis in Lenke type 1 idiopathic scoliosis. Spine (Phila Pa 1976). 2006;31:1828–1833. [DOI] [PubMed] [Google Scholar]
  • 4.Davids JR, Chamberlin E, Blackhurst DW. Indications for magnetic resonance imaging in presumed adolescent idiopathic scoliosis. J Bone Joint Surg Am. 2004;86:2187–2195. [DOI] [PubMed] [Google Scholar]
  • 5.Diab M, Landman Z, Lubicky J, Dormans J, Erickson M, Richards BS, members of the Spinal Deformity Study Group. Use and outcome of MRI in the surgical treatment of adolescent idiopathic scoliosis. Spine (Phila Pa 1976). 2011;36:667–671. [DOI] [PubMed] [Google Scholar]
  • 6.Do T, Fras C, Burke S, Widmann RF, Rawlins B, Boachie-Adjei O. Clinical value of routine preoperative magnetic resonance imaging in adolescent idiopathic scoliosis: a prospective study of three hundred and twenty-seven patients. J Bone Joint Surg Am. 2001;83:577–579. [DOI] [PubMed] [Google Scholar]
  • 7.Inoue M, Minami S, Nakata Y, Otsuka Y, Takaso M, Kitahara H, Tokunaga M, Isobe K, Moriya H. Preoperative MRI analysis of patients with idiopathic scoliosis: a prospective study. Spine (Phila Pa 1976). 2005;30:108–114. [DOI] [PubMed] [Google Scholar]
  • 8.Karami M, Sagheb S, Mazda K. Evaluation of coronal shift as an indicator of neuroaxial abnormalities in adolescent idiopathic scoliosis: a prospective study. Scoliosis. 2014;9:9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Konieczny MR, Senyurt H, Krauspe R. Epidemiology of adolescent idiopathic scoliosis. J Child Orthop. 2013;7:3–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Lee RS, Reed DW, Saifuddin A. The correlation between coronal balance and neuroaxial abnormalities detected on MRI in adolescent idiopathic scoliosis. Eur Spine J. 2012;21:1106–1110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Maiocco B, Deeney VF, Coulon R, Parks PF., Jr Adolescent idiopathic scoliosis and the presence of spinal cord abnormalities: preoperative magnetic resonance imaging analysis. Spine (Phila Pa 1976). 1997;22:2537–2541. [DOI] [PubMed] [Google Scholar]
  • 12.Mejia EA, Hennrikus WL, Schwend RM, Emans JB. A prospective evaluation of idiopathic left thoracic scoliosis with magnetic resonance imaging. J Pediatr Orthop. 1996;16:354–358. [DOI] [PubMed] [Google Scholar]
  • 13.Mineiro J, Weinstein SL. Delayed postoperative paraparesis in scoliosis surgery. Spine (Phila Pa 1976). 1997;22:1668–1672. [DOI] [PubMed] [Google Scholar]
  • 14.Morcuende JA, Dolan LA, Vazquez JD, Jirasirakul A, Weinstein SL. A prognostic model for the presence of neurogenic lesions in atypical idiopathic scoliosis. Spine (Phila Pa 1976). 2004;29:51–58. [DOI] [PubMed] [Google Scholar]
  • 15.Nakahara D, Yonezawa I, Kobanawa K, Sakoda J, Nojiri H, Kamano S, Okuda T, Kurosawa H. Magnetic resonance imaging evaluation of patients with idiopathic scoliosis: a prospective study of four hundred seventy-two outpatients. Spine (Phila Pa 1976). 2011;36:E482–485. [DOI] [PubMed] [Google Scholar]
  • 16.Ozturk C, Karadereler S, Ornek I, Enercan M, Ganiyusufoglu K, Hamzaoglu A. The role of routine magnetic resonance imaging in the preoperative evaluation of adolescent idiopathic scoliosis. Int Orthop. 2010;34:543–546. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Qiao J, Zhu Z, Zhu F, Wu T, Qian B, Xu L, Qiu Y. Indication for preoperative MRI of neural axis abnormalities in patients with presumed thoracolumbar/lumbar idiopathic scoliosis. Eur Spine J. 2013;22:360–366. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Rajasekaran S, Kamath V, Kiran R, Shetty AP. Intraspinal anomalies in scoliosis: an MRI analysis of 177 consecutive scoliosis patients. Indian J Orthop. 2010;44:57–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Richards BS, Sucato DJ, Johnston CE, Diab M, Sarwark JF, Lenke LG, Parent S; Spinal Deformity Study Group. Right thoracic curves in presumed adolescent idiopathic scoliosis: which clinical and radiographic findings correlate with a preoperative abnormal magnetic resonance image? Spine (Phila Pa 1976). 2010;35:1855–1860. [DOI] [PubMed] [Google Scholar]
  • 20.Samuelsson L, Lindell D, Kogler1 H. Spinal cord and brain stem anomalies in scoliosis: MR screening of 26 cases. Acta Orthop Scand. 1991;62:403–406. [DOI] [PubMed] [Google Scholar]
  • 21.Shen WJ, McDowell GS, Burke SW, Levine DB, Chutorian AM. Routine preoperative MRI and SEP studies in adolescent idiopathic scoliosis. J Pediatr Orthop. 1996;16:350–353. [DOI] [PubMed] [Google Scholar]
  • 22.Singhal R, Perry DC, Prasad S, Davidson NT, Bruce CE. The use of routine preoperative magnetic resonance imaging in identifying intraspinal anomalies in patients with idiopathic scoliosis: a 10-year review. Eur Spine J. 2013;22:355–359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Slim K, Nini E, Forestier D, Kwiatkowski F, Panis Y, Chipponi J. Methodological index for non-randomized studies (MINORS): development and validation of a new instrument. ANZ J Surg. 2003;73:712–716. [DOI] [PubMed] [Google Scholar]
  • 24.Stockl B, Wimmer C, Innherhofer P, Kofler M, Behensky H. Delayed anterior spinal artery syndrome following posterior scholiosis correction. Eur Spine J. 2005;14:906–909. [DOI] [PubMed] [Google Scholar]
  • 25.Winter RB, Lonstein JE, Heithoff KB, Kirkham JA. Magnetic resonance imaging evaluation of the adolescent patient with idiopathic scoliosis before spinal instrumentation and fusion: a prospective, double-blinded study of 140 patients. Spine (Phila Pa 1976). 1997;22:855–858. [DOI] [PubMed] [Google Scholar]
  • 26.Wu L, Qiu Y, Wang B, Zhu ZZ, Ma WW. The left thoracic curve pattern: a strong predictor for neural axis abnormalities in patients with ‘idiopathic' scoliosis. Spine (Phila Pa 1976). 2010;35:182–185. [DOI] [PubMed] [Google Scholar]

Articles from Clinical Orthopaedics and Related Research are provided here courtesy of The Association of Bone and Joint Surgeons

RESOURCES