Journal of Clinical Neuroscience xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Journal of Clinical Neuroscience

journal homepage: www.elsevier .com/ locate/ jocn

Review

Minimally invasive spinal surgery for the treatment of traumaticthoracolumbar burst fractures

http://dx.doi.org/10.1016/j.jocn.2014.05.0300967-5868/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +1 312 695 8143; fax: +1 312 695 0225.E-mail address: zsmith@nmff.org (Z.A. Smith).

Please cite this article in press as: Oh T et al. Minimally invasive spinal surgery for the treatment of traumatic thoracolumbar burst fractures. J Clrosci (2014), http://dx.doi.org/10.1016/j.jocn.2014.05.030

Taemin Oh, Justin K. Scheer, Shayan Fakurnejad, Nader S. Dahdaleh, Zachary A. Smith ⇑Department of Neurological Surgery, Northwestern University, Feinberg School of Medicine, 676 N. St Clair Street, Suite 2210, Chicago, IL 60611, USA

a r t i c l e i n f o a b s t r a c t

Article history:Received 7 March 2014Accepted 4 May 2014Available online xxxx

Keywords:Burst fractureEvidence-based medicineManagementMinimally invasiveSpineTraumaTreatment

The optimum management of traumatic thoracolumbar burst fractures is cause for much debate in theliterature. Although minimally invasive surgery (MIS) approaches are increasingly used in the manage-ment of degenerative spinal pathology, their role in treating burst fractures is unknown. Assessing thelevel of evidence (LOE) for use of MIS approaches in vertebral burst fractures may impart betterunderstanding of how to integrate MIS in the treatment schema for these fractures. A comprehensiveliterature review was conducted using MEDLINE for all articles published on traumatic thoracolumbarburst fractures through to July 2013. LOE was assigned according to the standards set forth by the Journalof Clinical Orthopaedics and Related Research and the Oxford Centre for Evidence Based Medicine. Fulltexts were reviewed to select only those articles discussing MIS approaches as a treatment modality. Atotal of 501 articles met both inclusion and exclusion criteria, and 403 of those were published withinthe past two decades. Among those, 35 articles detailed the use of MIS approaches in the managementof traumatic thoracolumbar burst fractures. Only three studies provided high LOE: one level 1 studyand two level 2 studies. Thirteen studies described the use of vertebroplasty or kyphoplasty, but all werelevel 4 or level 5 studies. Currently, the LOE for utilization of MIS approaches to manage traumaticthoracolumbar burst fractures is low. Further work in the form of prospective randomized controlledtrials is needed to ascertain how MIS may be integrated into the treatment scheme for thoracolumbarburst fractures.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Vertebral burst fractures can represent up to 64% of all fracturesof the thoracolumbar spine [1]. Under the Denis classification sys-tem, burst fractures are defined as fractures of the anterior andmiddle spinal columns [2]. Depending on the specific mechanismand location of injury, these fractures were initially subdivided intofive distinct types (A–E) [2]. In 1994, Magerl and colleagues furtherdesignated thoracolumbar burst fractures as A3 compressionfractures under the AO classification system [3]. Often resulting fromhigh-energy axial loads, the etiology for this fracture type is usuallytraumatic, with vertical falls, automobile accidents, and sportinginjuries being the most common inciting factors [4–6]. Certainhigh-risk populations, such as military personnel suffering from bal-listic wounds, can also be prone to burst fracture injuries [7].

Best management practices for thoracolumbar burst fracturesremain controversial, and the debate fundamentally revolves

around the need for conservative versus operative intervention,specifically in patients that are neurologically intact on presenta-tion [6,8–12]. This lack of consensus erects substantial barriers tooptimizing management protocols. In that manner, it is worth con-sidering the potential role of minimally invasive surgery (MIS) inthe treatment algorithm for traumatic thoracolumbar burst frac-tures. Undoubtedly, in the modern health care climate, precedenceis placed on emphasizing patient quality of life while minimizingmorbidity and mortality. Given its minimally invasive nature,MIS is a treatment modality that could potentially bridge the gapbetween operative and non-operative management. Importantly,MIS has emerged over the past decade as an effective method formanaging spinal pathology, particularly degenerative diseases[13–16]. Furthermore, evidence points to improvements in peri-operative factors such as blood loss or operative time with MISapproaches, which translate into reduced hospitalization andlower costs [14,17,18].

However, critical and substantial evaluation of MIS techniquesin the setting of spinal trauma is scant. Anecdotally, the benefitsof applying MIS techniques in spinal trauma have been unclear.For example, the senior authors (N.S.D. and Z.A.S.) have employed

in Neu-

2 T. Oh et al. / Journal of Clinical Neuroscience xxx (2014) xxx–xxx

the MIS technique of percutaneous ‘‘internal bracing’’ for thoraco-lumbar trauma patients. However, this has been done in selectpatients with appropriate fracture characteristics and load sharingscores. Other patients during this same period have been treatedwith open pedicle screw fixation and fusion. Lack of available dataon this topic invites exploration of the current available data.Although MIS approaches have been used to manage traumaticthoracolumbar burst fractures [19,20], no study to our knowledgehas analyzed the level of evidence (LOE) advocating for the utiliza-tion of MIS approaches in traumatic thoracolumbar burst fractures.Because studies that offer high LOE (such as randomized controlledtrials [RCT]) can shape the management decision-making process,addressing this gap in the literature is imperative. The primaryobjective of the present study was therefore to conduct a compre-hensive systematic literature review to ascertain the LOE for use ofMIS in the treatment of traumatic thoracolumbar burst fractures,with specific weight and discussion placed on those studies provid-ing high-level evidence.

2. Materials and Methods

2.1. Literature search strategy

This systematic review was performed in accordance withCochrane guidelines [21]. A comprehensive electronic search onPubMed (MEDLINE) was conducted using the following terms:(spine AND burst) or ‘‘burst fracture’’ or (burst and fracture) or(A3 AND fracture AND spine). The most recent search was donein July 2013. Duplicates were discarded and two blinded reviewersindependently screened articles by title and abstract. Inclusion cri-teria were limited to (1) burst fractures of the thoracic or lumbarspine, and (2) caused by traumatic etiology. Once the final dataset was constructed, articles over the past 20 years were includedin the current study (1 January 1992–13 July 2013). This temporalrestriction was placed because anecdotal evidence suggests thatmanagement of burst fractures has evolved tremendously withinthis epoch. Exclusion criteria included (1) cervical fractures, (2)osteoporotic fractures, and (3) pathological burst fractures. Whenarticles did not clearly meet these criteria, full texts were reviewed.All other ambiguities were resolved through consensus or the addi-tion of a third reviewer, as necessary.

Fig. 1. Step-by-step depiction of the article selection process for the literaturereview. A total of 35 articles discussed minimally invasive approaches for thetreatment of traumatic thoracolumbar burst fractures.

2.2. Data extraction and study classification

All articles were categorized based on study type, whichincluded case reports, case series, cadaver studies, comparativestudies, and RCT. Comparative studies were defined as those inves-tigating two or more cohorts without meeting criteria for RCT.Articles focused on anatomical exploration or technical approachesusing cadavers were classified as cadaver studies. Data pertainingto publication year, population size, pre-operative neurologic sta-tus, treatment type (MIS versus open versus conservative), use ofkyphoplasty or vertebroplasty, primary outcome measures, andfollow-up were extracted from each report. MIS was defined asany study directly mentioning an MIS method or the use of any lessinvasive technique. Any discrepancies pertaining to use of MIStechniques were resolved through consensus or the addition of athird reviewer. The standardized reporting scheme provided bythe Journal of Clinical Orthopaedics and Related Research [22],which is an adaptation of the Oxford Centre for Evidence BasedMedicine LOE [23–25], was used to assign the appropriate LOE toeach study. Studies were ranked on a scale of 1 (highest) to 5 (low-est). Again, ambiguities or discrepancies were resolved throughconsensus or the addition of a third reviewer, as necessary. Follow-ing the construction of the final dataset over the past 20 years, all

Please cite this article in press as: Oh T et al. Minimally invasive spinal surgeryrosci (2014), http://dx.doi.org/10.1016/j.jocn.2014.05.030

studies identified as using an MIS treatment modality wereselected and included in this review.

3. Results

3.1. Article selection

The results of our search strategy are portrayed in Figure 1.Preliminary literature search resulted in 1054 abstracts, and 501studies involving traumatic thoracolumbar burst fractures metinclusion and exclusion criteria, with 403 over the past 20 years.Out of the 403 studies, 35 articles were identified as using anMIS technique [19,20,26–58]. All studies were technique-based,with most in the format of a case report or series (n = 28, 80%).There were four comparative studies, two cadaver studies, and asingle RCT. Twenty studies were retrospective, 10 were prospec-tive, and five either did not specify or were indeterminate. LOEbreakdown, in decreasing order of importance, was: level 1(n = 1, 3%), level 2 (n = 2, 6%), level 3 (n = 2, 6%), level 4 (n = 28,80%), and level 5 (n = 2, 6%). Of the three level 1 or level 2 studies,two were comparative studies and one was an RCT. These threestudies are summarized in Table 1.

3.2. Evidence for MIS interventions in thoracolumbar vertebral bodyfractures

3.2.1. Level 1 evidenceThe single level 1 study was a prospective RCT published in

2012 by Jiang and colleagues [20]. Within a cohort of 61 patients,

for the treatment of traumatic thoracolumbar burst fractures. J Clin Neu-

Table 1Summary of articles presenting high-level evidence for use of minimally invasive approaches found in the literature

Authors LOE Study type Techniques Key findings

Jiang et al. [20] 1 RCT Percutaneous versus paraspinal posterior pedicle screw-rodfixation

Percutaneous fixation:� Shorter operative time (p = 0.008)� Lower EBL (p < 0.001)� Shorter hospitalization (p < 0.001)� Lower VAS score at 3 month follow-up (p < 0.05)� Lower ODI score at 3 month follow-up (p < 0.05)

Paraspinal fixation:� Better immediate postoperative LKA correction

(p < 0.001)� Better immediate postoperative VBH correction

(p < 0.001)Pang et al. [50] 2 Comparative Paravertebral muscle versus posterior midline approach PVM approach:

� Lower EBL (p < 0.005)� Shorter recumbency time (p < 0.005)� Shorter time to exercise (p < 0.005)� Higher postoperative VAS score (p < 0.005)

Schmid et al.[19]

2 Comparative Thoracoscopic versus TLIF for anterior column reconstruction TLIF:� Greater postoperative spinal canal narrowing (p < 0.05)� More extensive surgical decompression (p = 0.02)

EBL = estimated blood loss, LKA = local kyphosis angle, LOE = level of evidence, ODI = Oswestry Disability index, PVM = paravertebral muscle, RCT = randomized controlledtrial, TLIF = transforaminal lumbar interbody fusion, VAS = visual analogue scale, VBH = vertebral body height.

T. Oh et al. / Journal of Clinical Neuroscience xxx (2014) xxx–xxx 3

two techniques of posterior pedicle screw-rod fixation were com-pared: the minimally invasive percutaneous (n = 31) versus theparaspinal approach (n = 30). All patients had suffered single-levelburst fractures from T11 to L2, presented within 7 days of incitinginjury without evidence of pre-operative neurologic deficit, had aload-sharing score of 66, possessed an intact invertebral disc,posterior longitudinal ligament, and facet joint, and receivedMRI-confirmed posterior ligamentous complex injury. Patientswith prior history of spinal surgery and/or osteoporotic/pathologicfractures were omitted. After enrollment, patients were randomlyassigned to receive one of the two muscle-sparing approaches. Ther-apeutic intervention involved pedicle screw placement into the ver-tebrae one level above and below the fracture level, and placementwas confirmed radiologically, the accuracy of which was graded ona scale of I–III as per Youkilis et al. [59]. Post-operatively, patientswere fitted with a brace for 8 weeks and encouraged to ambulatewithin 3 days of the procedure. Follow-up was conducted at the 3,6, 12 month marks and yearly afterwards for at least 3 years. Allimplants were removed 1 year post-operatively. At follow-up, thelocal kyphosis angle (LKA) and vertebral body height (VBH) wereascertained and compared to pre-operative values. Pain and qualityof life were assessed using the visual analogue scale (VAS) and Osw-estry Disability Index (ODI), respectively.

There were no evident demographic differences between treat-ment groups. Peri-operatively, percutaneous surgery resulted inshorter operative times (p = 0.008), lower estimated blood loss(p < 0.001), and shorter duration of hospital stay (p < 0.001). How-ever, there was no statistical difference regarding time fromsurgery to ambulation. Hardware failure was not observed. GradeII screw misplacements manifested in 4.8% and 3.3% of patientsin the percutaneous and paraspinal groups, respectively. Therewas no evidence of Grade III misplacement. With respect to clinicaloutcomes, patients randomized to the paraspinal approachachieved better surgical correction of both LKA (p < 0.001) andVBH (p < 0.001). Furthermore, the percutaneous approach led tolower VAS (3.58 ± 0.27 versus 4.47 ± 0.29; p < 0.05) and ODI scores(13.48 ± 6.13 versus 18.40 ± 6.90; p < 0.05) at 3 month follow-up.This difference, however, did not hold with longer follow-up.

3.2.2. Level 2 evidenceTwo prospective comparative studies qualified as level 2 evi-

dence. In 2009, Pang et al. reported their cohort of 62 patients

Please cite this article in press as: Oh T et al. Minimally invasive spinal surgeryrosci (2014), http://dx.doi.org/10.1016/j.jocn.2014.05.030

who were randomized to undergo an approach through the para-vertebral muscle (PVM) (n = 34) or the traditional posteriorapproach (n = 28) [50]. All patients had thoracolumbar fractures(T10–L2) and no neurologic deficits. However, only 47 of the 62patients had documented burst fractures while the others hadcompression fractures. The authors did not clarify which of the47 patients received PVM or the traditional approach. Peri-opera-tive factors were assessed and the VAS was used as the endpoint.More long-term outcomes were not analyzed, despite a meanfollow-up of longer than a year (18.6 months). Overall, the groupreceiving PVM had improved operative times, lower estimatedblood loss, and lower VAS scores (p < 0.005). Patients were alsobedridden for a shorter period of time (p < 0.005).

More recently in 2012, Schmid et al. performed circumferentialfusion on 35 patients [19]. All patients initially underwent short-segment fixation, reduction, and posterolateral fusion for posteriorstabilization. For the anterior column, patients were non-randomlyassigned to receive either (1) thoracoscopic reconstruction with atricortical strut graft or titanium vertebral body implant (n = 14)or (2) unilateral transforaminal lumbar interbody fusion withmonocortical strut grafts (n = 21). All patients had known thoraco-lumbar burst fractures from T12–L2. Patients with neurologic def-icits or pathologic/osteoporotic fractures were excluded. Surgicaloutcomes were analyzed with respect to correction of kyphosis.Clinical outcomes were analyzed with the VAS, ODI, the McGillPain Questionnaire, the Roland and Morris Disability Question-naire, and the Hannover Functional Questionnaire of FunctionalDisability Caused by Backache. No significant differences in demo-graphics or peri-operative factors such as procedure duration andestimated blood loss were seen. Post-operatively, both treatmentmodalities led to significant reduction of kyphosis, but there wereno significant differences between the approaches (p = 0.31). Fur-thermore, scores for all five questionnaires yielded no meaningfuldifferences.

3.3. Vertebroplasty and kyphoplasty

Thirteen studies described the application of vertebroplasty orkyphoplasty, none of which provided high-level evidence[34,35,37,38,42,45–47,51–55]. Twelve were level 4 studies pub-lished as case reports or series; the last article was a level 5 cadav-eric study. Techniques were tested either as primary intervention

for the treatment of traumatic thoracolumbar burst fractures. J Clin Neu-

4 T. Oh et al. / Journal of Clinical Neuroscience xxx (2014) xxx–xxx

or in combination with other approaches. Hartmann et al., forexample, treated 26 patients with kyphoplasty via bone needleinsertion, balloon expansion, and subsequent cementation withpolymethyl methacrylate (PMMA) as the sole treatment modality.Minor cement leakages were observed but overall complicationrates were marginal. While surgery initially led to post-operativereductions in kyphotic wedge and Daniaux angles, this correctionwas lost at follow-up [52].

Other groups utilized vertebroplasty or kyphoplasty in conjunc-tion with other approaches that were either minimally invasive oropen approaches. Blondel and colleagues combined kyphoplastywith PMMA and percutaneous osteosynthesis (posterior percutane-ous approach) in their treatment of 29 patients with traumatic tho-racolumbar burst fractures. Twenty-two patients receivedkyphoplasty first; seven received osteosynthesis followed by kyp-hoplasty. Post-operatively, both groups achieved reduction inkyphosis with similar loss of correction [54]. Similarly, Fuenteset al. applied double-balloon kyphoplasty and percutaneous osteo-synthesis in 18 patients with traumatic thoracolumbar burst frac-tures. PMMA or phosphocalcic cement was used to fill the gap.Post-operative improvements in local kyphosis, vertebral height,and Beck index were all statistically significant (p < 0.001) using thismethodology [55]. On the other hand, Boswell et al. reported a smallseries of two traumatic burst fractures that were treated with opendecompressive laminectomy followed by vertebroplasty through afluoroscopy-assisted transpedicular approach and PMMA implanta-tion [53]. Rahamimov et al. utilized balloon kyphoplasty in conjunc-tion with short-segment pedicle instrumentation and screwaugmentation with PMMA. Mean loss of correction was 3.05� at12 months [51].

4. Discussion

In the world of spine surgery, the field of MIS has emergedwithin the past two decades as a promising novel method for treat-ing patients. The potential benefits are numerous, including cost-efficacy [60] as well as reductions in intra-operative blood lossand hospital stay [61]. However, the efficacy of MIS when com-pared to open techniques remains debated, especially with regardto long-term outcomes [62–64]. In the current study, we demon-strate that the LOE for use of MIS in the management of traumaticthoracolumbar vertebral fractures remains relatively low. Withinthe vast spine literature, only 35 studies assess MIS approachesin some capacity and, among them, only three met criteria for pro-viding robust LOE.

To further compound the issue, high-level studies did notdirectly contrast open to MIS approaches or MIS to conservativetreatment, thus rendering any head-to-head comparisons difficult.Jiang et al. compared two minimally invasive procedures (parasp-inal versus percutaneous) [20] while Schmid et al. had all patientsundergo preliminary posterolateral fusions before administeringsecondary intervention [19]. While Pang et al. studied the efficacyof a more minimally invasive alteration to the traditional posteriorapproach, both still required open midline incisions on all patients[50]. Another weakness of this particular study was the failure toassess long-term outcomes. And with regard to vertebroplastyand kyphoplasty, which have both traditionally been applied inthe treatment of osteoporotic compression fractures [65,66], theLOE in support of their use was low.

Based on these findings, it is currently difficult to advocate forutilizing MIS methodologies to treat traumatic thoracolumbarburst fractures. The importance of practicing evidence-based med-icine cannot be understated, as studies providing high LOE are crit-ical to devising optimum guidelines for patient management. Theglobal effects of such studies can be far-reaching, as evidencedby the way the Framingham Heart Study has revolutionized our

Please cite this article in press as: Oh T et al. Minimally invasive spinal surgeryrosci (2014), http://dx.doi.org/10.1016/j.jocn.2014.05.030

understanding of the treatment of cardiovascular disease [67].Similarly, instituting RCT comparing open versus MIS approachescan provide invaluable knowledge on the role that MIS can playin the treatment algorithm for vertebral burst fractures

The primary surgical benefits of utilizing MIS approaches arethe dramatic improvements seen in peri-operative factors suchas blood loss, hospitalization time, and operative times, whichtranslate to reduced chances of incurring surgical morbidity[20,50]. Furthermore, MIS techniques (for example internal brac-ing) circumvent the need for fusion, which could potentiallyreduce the chances for developing fusion-related complicationssuch as proximal junctional kyphosis [68], as well as increasingthe degree of mobility and subsequent hardware removal. On theother hand, it is unclear whether bracing without fusion wouldunilaterally suffice as treatment. Additionally, without fusion,there may also be a concern for screw or rod failure due toincreased loads through the instrumentation leading to constructfatigue and screw loosening. Long-term prospective studies wouldthus go a long way towards addressing these concerns.

The present systematic review highlights the need for improvedstudy design as well as the need to conduct further studies on thistopic matter. Systematic reviews are well-known for providing apowerful tool for the interpretation and evaluation of the collectiveliterature [21]. Within the traumatic thoracolumbar burst fractureliterature, there are a total of six systematic reviews assessing var-ious aspects of management [69–74]. Although most studies aredescriptive, Gnanenthiran et al. statistically compared the benefitsof operative versus conservative management and found that whileoperative intervention improves kyphosis, functional and pain out-comes were roughly equivalent [69]. Based on their own review,Alpantaki et al. devised a treatment algorithm for management[71]. Thus, while this review highlights the dearth of high-qualityevidence for MIS, concerted efforts at conducting level 1 and level2 trials studying MIS interventions in burst fractures may furtherdevelop our understanding of the role that MIS can play and poten-tially lead to changes in management paradigms in the future. Inparticular, future studies should aim towards assessing the impactof MIS on functional outcomes and disability using patient-reported outcome surveys such as the ODI, Euroqol, and the ShortForm (36) Health Survey. Secondary outcomes such as radio-graphic findings, post-operative complications, and neurologicalscores remain important considerations as well, particularly ifstudied in relation to health-related quality of life outcomes.

5. Conclusion

The importance of MIS in the management of traumatic thoraco-lumbar burst fractures is currently unknown. Based on the presentsystematic review, the level of evidence for the use of MISapproaches is low, and distinct benefits in patient quality of life orpatient outcomes have not been clearly delineated. However, futureprospective RCT dedicated to illustrating the benefits or disadvan-tages of MIS compared with open and conservative managementtechniques may provide valuable information on the utility of MIS.

Conflicts of Interest/Disclosures

The authors declare that they have no financial or other con-flicts of interest in relation to this research and its publication.

References

[1] Gertzbein SD. Neurologic deterioration in patients with thoracic and lumbarfractures after admission to the hospital. Spine (Phila Pa 1976)1994;19:1723–5.

[2] Denis F. Spinal instability as defined by the three-column spine concept inacute spinal trauma. Clin Orthop Relat Res 1984:65–76.

for the treatment of traumatic thoracolumbar burst fractures. J Clin Neu-

T. Oh et al. / Journal of Clinical Neuroscience xxx (2014) xxx–xxx 5

[3] Magerl F, Aebi M, Gertzbein SD, et al. A comprehensive classification ofthoracic and lumbar injuries. Eur Spine J 1994;3:184–201.

[4] Bensch FV, Koivikko MP, Kiuru MJ, et al. The incidence and distribution of burstfractures. Emerg Radiol 2006;12:124–9.

[5] Vander Have KL, Caird MS, Gross S, et al. Burst fractures of the thoracic andlumbar spine in children and adolescents. J Pediatr Orthop 2009;29:713–9.

[6] Dai LY, Jiang LS, Jiang SD. Conservative treatment of thoracolumbar burstfractures: a long-term follow-up results with special reference to the loadsharing classification. Spine (Phila Pa 2008) 2008;33:2536–44.

[7] Lehman RA Jr., Paik H, Eckel TT, et al. Low lumbar burst fractures: a uniquefracture mechanism sustained in our current overseas conflicts. Spine J2012;12:784–90.

[8] Kraemer WJ, Schemitsch EH, Lever J, et al. Functional outcome ofthoracolumbar burst fractures without neurological deficit. J Orthop Trauma1996;10:541–4.

[9] Siebenga J, Leferink VJ, Segers MJ, et al. Treatment of traumatic thoracolumbarspine fractures: a multicenter prospective randomized study of operativeversus nonsurgical treatment. Spine (Phila Pa 1976) 2006;31:2881–90.

[10] Dai LY, Jiang SD, Wang XY, et al. A review of the management ofthoracolumbar burst fractures. Surg Neurol 2007;67:221–31 [discussion 231].

[11] Domenicucci M, Preite R, Ramieri A, et al. Thoracolumbar fractures withoutneurosurgical involvement: surgical or conservative treatment? J NeurosurgSci 1996;40:1–10.

[12] Denis F, Armstrong GW, Searls K, et al. Acute thoracolumbar burst fractures inthe absence of neurologic deficit. A comparison between operative andnonoperative treatment. Clin Orthop Relat Res 1984:142–9.

[13] Archavlis E, Carvi y Nievas M. Comparison of minimally invasive fusion andinstrumentation versus open surgery for severe stenotic spondylolisthesiswith high-grade facet joint osteoarthritis. Eur Spine J 2013;22:1731–40.

[14] Parker SL, Mendenhall SK, Shau DN, et al. Minimally invasive versus opentransforaminal lumbar interbody fusion for degenerative spondylolisthesis:comparative effectiveness and cost-utility analysis. World Neurosurg 2013[Epub ahead of print].

[15] Mobbs RJ, Sivabalan P, Li J. Minimally invasive surgery compared to openspinal fusion for the treatment of degenerative lumbar spine pathologies. J ClinNeurosci 2012;19:829–35.

[16] Cheng JS, Park P, Le H, et al. Short-term and long-term outcomes of minimallyinvasive and open transforaminal lumbar interbody fusions: is there adifference? Neurosurg Focus 2013;35:E6.

[17] Singh K, Nandyala SV, Marquez-Lara A, et al. A perioperative cost analysiscomparing single-level minimally invasive and open transforaminal lumbarinterbody fusion. Spine J 2013 [Epub ahead of print].

[18] Lucio JC, Vanconia RB, Deluzio KJ, et al. Economics of less invasive spinalsurgery: an analysis of hospital cost differences between open and minimallyinvasive instrumented spinal fusion procedures during the perioperativeperiod. Risk Manag Healthc Policy 2012;5:65–74.

[19] Schmid R, Lindtner RA, Lill M, et al. Combined posteroanterior fusion versustransforaminal lumbar interbody fusion (TLIF) in thoracolumbar burstfractures. Injury 2012;43:475–9.

[20] Jiang XZ, Tian W, Liu B, et al. Comparison of a paraspinal approach with apercutaneous approach in the treatment of thoracolumbar burst fractureswith posterior ligamentous complex injury: a prospective randomizedcontrolled trial. J Int Med Res 2012;40:1343–56.

[21] Cochrane Handbook for Systematic Reviews of Interventions 2011. Availablefrom: www.cochrane-handbook.org.

[22] Levels of Evidence 2013. Available from: http://www.clinorthop.org/author_and_reviewer_information.html [accessed 7/13/2013].

[23] Howick J, Chalmers I, Glasziou P, et al. The 2011 Oxford CEBM levels ofEvidence (Introductory Document) 2011. Available from: Oxford Centre ofEvidence Based Medicine [serial online, accessed 7/13/13].

[24] Howick J, Chalmers I, Glasziou P, et al. Explanation of the 2011 Oxford Centrefor Evidence-Based Medicine (OCEBM) Levels of Evidence (BackgroundDocument) 2011. Available from: Oxford Centre for Evidence BasedMedicine; Oxford UK [serial online, accessed 7/13/2013].

[25] Howick J, Chalmers I, Glasziou P, et al. The Oxford Levels of Evidence 2. 9/16/2013, 2011. Available from: http://www.cebm.net/mod_product/design/files/CEBM-Levels-of-Evidence-2.1.pdf [accessed 7/13/2013].

[26] Keel MJ, Lustenberger T, Puippe G, et al. Table-mounted ring retractor forconsistent visualization in endoscopy-assisted anterior reconstruction of burstfractures of the thoracolumbar junction. Acta Orthop Belg 2013;79:90–6.

[27] Yang WE, Ng ZX, Koh KM, et al. Percutaneous pedicle screw fixation forthoracolumbar burst fracture: a Singapore experience. Singapore Med J 2012;53:577–81.

[28] Spiegl U, Hauck S, Merkel P, et al. 6-Year follow-up of ventral monosegmentalspondylodesis of incomplete burst fractures of the thoracolumbar spine usingthree cortical iliac crest bone grafts. Arch Orthop Trauma Surg 2012;132:1473–80.

[29] Lefranc M, Peltier J, Fichten A, et al. Dual, minimally invasive fixation in acute,double, thoracic spine fracture. Minim Invasive Neurosurg 2011;54:253–6.

[30] Eck JC. Minimally invasive corpectomy and posterior stabilization for lumbarburst fracture. Spine J 2011;11:904–8.

[31] Kakarla UK, Little AS, Chang SW, et al. Placement of percutaneous thoracicpedicle screws using neuronavigation. World Neurosurg 2010;74:606–10.

[32] Smith WD, Dakwar E, Le TV, et al. Minimally invasive surgery for traumaticspinal pathologies: a mini-open, lateral approach in the thoracic and lumbarspine. Spine (Phila Pa 1976) 2010;35:S338–46.

Please cite this article in press as: Oh T et al. Minimally invasive spinal surgeryrosci (2014), http://dx.doi.org/10.1016/j.jocn.2014.05.030

[33] Ni WF, Huang YX, Chi YL, et al. Percutaneous pedicle screw fixation forneurologic intact thoracolumbar burst fractures. J Spinal Disord Tech 2010;23:530–7.

[34] Doody O, Czarnecki C, Given MF, et al. Vertebroplasty in the management oftraumatic burst fractures: a case series. J Med Imaging Radiat Oncol 2009;53:489–92.

[35] Carbognin G, Sandri A, Girardi V, et al. Treatment of type-A3 amyelicthoracolumbar fractures (burst fractures) with kyphoplasty: initialexperience. Radiol Med 2009;114:133–40.

[36] Wang MY, Ludwig SC, Anderson DG, et al. Percutaneous iliac screw placement:description of a new minimally invasive technique. Neurosurg Focus 2008;25:E17.

[37] Korovessis P, Hadjipavlou A, Repantis T. Minimal invasive short posteriorinstrumentation plus balloon kyphoplasty with calcium phosphate for burstand severe compression lumbar fractures. Spine (Phila Pa 1976) 2008;33:658–67.

[38] Afzal S, Akbar S, Dhar SA. Short segment pedicle screw instrumentation andaugmentation vertebroplasty in lumbar burst fractures: an experience. EurSpine J 2008;17:336–41.

[39] Matsubara H, Tsuchiya H, Kawahara N, et al. Ilizarov external fixator for burstfracture of the lumbar spine: a case report. J Orthop Surg (Hong Kong)2007;15:380–3.

[40] Maciejczak A, Barnas P, Dudziak P, et al. Posterior keyhole corpectomy withpercutaneous pedicle screw stabilization in the surgical management oflumbar burst fractures. Neurosurgery 2007;60:232–41 [discussion 241–2].

[41] Watanabe K, Yabuki S, Konno S, et al. Complications of endoscopic spinalsurgery: a retrospective study of thoracoscopy and retroperitoneoscopy. JOrthop Sci 2007;12:42–8.

[42] Oner FC, Verlaan JJ, Verbout AJ, et al. Cement augmentation techniques intraumatic thoracolumbar spine fractures. Spine (Phila Pa 1976)2006;31:S89–95 [discussion S104].

[43] Sahin S, Resnick DK. Minimally incisional stabilization of unstable L5 burstfracture. J Spinal Disord Tech 2005;18:455–7.

[44] Verlaan JJ, van de Kraats EB, Oner FC, et al. The reduction of endplate fracturesduring balloon vertebroplasty: a detailed radiological analysis of the treatmentof burst fractures using pedicle screws, balloon vertebroplasty, and calciumphosphate cement. Spine (Phila Pa 1976) 2005;30:1840–5.

[45] Verlaan JJ, van de Kraats EB, Oner FC, et al. Bone displacement and the role oflongitudinal ligaments during balloon vertebroplasty in traumaticthoracolumbar fractures. Spine (Phila Pa 1976) 2005;30:1832–9.

[46] Chen JF, Lee ST. Percutaneous vertebroplasty for treatment of thoracolumbarspine bursting fracture. Surg Neurol 2004;62:494–500 [discussion 500].

[47] Chen JF, Wu CT, Lee ST. Percutaneous vertebroplasty for the treatment of burstfractures. Case report. J Neurosurg Spine 2004;1:228–31.

[48] Horn EM, Henn JS, Lemole Jr GM, et al. Thoracoscopic placement of dual-rodinstrumentation in thoracic spinal trauma. Neurosurgery 2004;54:1150–3[discussion 1153–4].

[49] El Saghir H. Extracoelomic mini approach for anterior reconstructive surgeryof the thoracolumbar area. Neurosurgery 2002;51:S118–22.

[50] Pang W, Zhang GL, Tian W, et al. Surgical treatment of thoracolumbar fracturethrough an approach via the paravertebral muscle. Orthop Surg 2009;1:184–8.

[51] Rahamimov N, Mulla H, Shani A, et al. Percutaneous augmentedinstrumentation of unstable thoracolumbar burst fractures. Eur Spine J2012;21:850–4.

[52] Hartmann F, Gercek E, Leiner L, et al. Kyphoplasty as an alternative treatmentof traumatic thoracolumbar burst fractures Magerl type A3. Injury2012;43:409–15.

[53] Boswell S, Sather M, Kebriaei M, et al. Combined open decompressivelaminectomy and vertebroplasty for treatment of thoracolumbar fracturesretrospective review of 41 cases. Clin Neurol Neurosurg 2012;114:902–6.

[54] Blondel B, Fuentes S, Pech-Gourg G, et al. Percutaneous management ofthoracolumbar burst fractures: evolution of techniques and strategy. OrthopTraumatol Surg Res 2011;97:527–32.

[55] Fuentes S, Blondel B, Metellus P, et al. Percutaneous kyphoplasty and pediclescrew fixation for the management of thoraco-lumbar burst fractures. EurSpine J 2010;19:1281–7.

[56] Huang TJ, Hsu RW, Liu HP, et al. Analysis of techniques for video-assistedthoracoscopic internal fixation of the spine. Arch Orthop Trauma Surg1998;117:92–5.

[57] Huang TJ, Hsu RW, Sum CW, et al. Complications in thoracoscopic spinalsurgery: a study of 90 consecutive patients. Surg Endosc 1999;13:346–50.

[58] Merom L, Raz N, Hamud C, et al. Minimally invasive burst fracture fixation inthe thoracolumbar region. Orthopedics 2009;32.

[59] Youkilis AS, Quint DJ, McGillicuddy JE, et al. Stereotactic navigation forplacement of pedicle screws in the thoracic spine. Neurosurgery2001;48:771–8 [discussion 778–9].

[60] Allen RT, Garfin SR. The economics of minimally invasive spine surgery: thevalue perspective. Spine (Phila Pa 1976) 2010;35:S375–82.

[61] Khoo LT, Fessler RG. Microendoscopic decompressive laminotomy for thetreatment of lumbar stenosis. Neurosurgery 2002;51:S146–54.

[62] Ang CL, Phak-Boon Tow B, Fook S, et al. Minimally invasive compared withopen lumbar laminotomy: no functional benefits at 6 or 24 months aftersurgery. Spine J 2013 [Epub ahead of print].

[63] Fourney DR, Dettori JR, Norvell DC, et al. Does minimal access tubular assistedspine surgery increase or decrease complications in spinal decompression orfusion? Spine (Phila Pa 1976) 2010;35:S57–65.

for the treatment of traumatic thoracolumbar burst fractures. J Clin Neu-

6 T. Oh et al. / Journal of Clinical Neuroscience xxx (2014) xxx–xxx

[64] Payer M. ‘‘Minimally invasive’’ lumbar spine surgery: a critical review. ActaNeurochir (Wien) 2011;153:1455–9.

[65] Gan M, Zou J, Song D, et al. Is balloon kyphoplasty better than percutaneousvertebroplasty for osteoporotic vertebral biconcave-shaped fractures? ActaRadiol 2013 [Epub ahead of print].

[66] Yimin Y, Zhiwei R, Wei M, et al. Current status of percutaneous vertebroplastyand percutaneous kyphoplasty–a review. Med Sci Monit 2013;19:826–36.

[67] Mahmood SS, Levy D, Vasan RS, et al. The Framingham Heart Study and theepidemiology of cardiovascular disease: a historical perspective. Lancet2014;383:999–1008.

[68] Kim YJ, Bridwell KH, Lenke LG, et al. Proximal junctional kyphosis in adultspinal deformity after segmental posterior spinal instrumentation and fusion:minimum five-year follow-up. Spine (Phila Pa 1976) 2008;33:2179–84.

[69] Gnanenthiran SR, Adie S, Harris IA. Nonoperative versus operative treatmentfor thoracolumbar burst fractures without neurologic deficit: a meta-analysis.Clin Orthop Relat Res 2012;470:567–77.

Please cite this article in press as: Oh T et al. Minimally invasive spinal surgeryrosci (2014), http://dx.doi.org/10.1016/j.jocn.2014.05.030

[70] Boerger TO, Limb D, Dickson RA. Does ‘canal clearance’ affect neurologicaloutcome after thoracolumbar burst fractures? J Bone Joint Surg Br 2000;82:629–35.

[71] Alpantaki K, Bano A, Pasku D, et al. Thoracolumbar burst fractures: asystematic review of management. Orthopedics 2010;33:422–9.

[72] Oner FC, Wood KB, Smith JS, et al. Therapeutic decision making inthoracolumbar spine trauma. Spine (Phila Pa 1976) 2010;35:S235–44.

[73] Thomas KC, Bailey CS, Dvorak MF, et al. Comparison of operative andnonoperative treatment for thoracolumbar burst fractures in patientswithout neurological deficit: a systematic review. J Neurosurg Spine 2006;4:351–8.

[74] Yi L, Jingping B, Gele J, et al. Operative versus non-operative treatment forthoracolumbar burst fractures without neurological deficit. CochraneDatabase Syst Rev 2006:CD005079.

for the treatment of traumatic thoracolumbar burst fractures. J Clin Neu-