When Is Surgery Indicated for Thoracolumbar Burst Fractures? Kirkham B. Wood, MD,* Andrew K. Simpson, MD,* and Avraam Ploumis, MD* ,† Burst fractures of the thoracolumbar spine are a common injury and can be source of great pain and disability. Fortunately, in most instances, treatment can be nonoperative with excellent results. However, there are certain situations in which a surgical approach may be indicated: a fracture with a significant neurological deficit or an “unstable” burst fracture with disruption of the posterior ligamentous complex. The literature is reviewed and surgical outcomes are presented. Semin Spine Surg 24:235-239 © 2012 Elsevier Inc. All rights reserved. KEYWORDS thoracolumbar, fracture, surgery, stable, outcomes T raumatic burst fractures of the spine are one of the most frequent injuries, and the thoracolumbar junction (T10 to L2) is the most common location; it is seen in approxi- mately 6% of patients experiencing blunt trauma, 1 and L1 is the vertebra most frequently involved (Figs. 1-3). The conservative management of burst fractures has his- torically been short periods of bed rest and/or fitting with a cast or brace. 2,3 Many authors have described the nonopera- tive management of burst fractures and found it to be, in most situations, comparable with operative care. 2,4-9 Wood et al, 4 in a prospective study of neurologically intact patients with stable thoracolumbar junction burst fractures, followed 53 patients randomized into operative and nonoperative groups and found no differences in the clinical outcomes at almost 5 years after injury. However, there are certain aspects of a thoracolumbar burst fracture that may well be best treated with an operative approach. This article will examine some of those circum- stances. Classification Systems Historically, classification systems, such as Denis 10,11 and that of Margerl et al, 12 reflected the morphologic aspects of the fracture, but little else, and in situations of ligamentous injury or neurologic deficits, offered no guidance to treatment. In 2005, Vacarro et al 13 and members of the Spine Trauma Study Group introduced the injury classification and severity score system that includes a spinal injury severity score (Ta- ble 1). It is made up of 3 separate components: the morphol- ogy of the fracture, the integrity of the posterior ligamentous complex, and any neurological involvement. It has been shown to exhibit excellent overall reproducibility and reli- ability. 14-16 Different aspects of the injury within each com- ponent are scored hierarchically and can then be summed to recommend either operative or nonoperative care. As an ex- ample, complete injury to the posterior ligaments as in a flexion-distraction injury is weighted more than an intact situation and thus flexion-distraction injuries are more likely to require surgery. Similarly, an incomplete spinal cord def- icit is weighted more than a single nerve root injury and may similarly guide surgical decision making. Neurological Injury Thus, thoracolumbar burst fractures may be associated with varying degrees of neurological deficits, ranging from iso- lated nerve deficits to conus medullaris injuries to frank pa- ralysis. There are no level I or level II 17 data that directly com- pare the outcomes of surgery versus nonoperative care for thoracolumbar burst fractures with a neurological deficit. However, there are several level III evidence studies that have retrospectively compared the 2 treatment meth- ods. 3,18-25 The Scoliosis Research Society committee on morbidity and mortality reported outcomes on 1000 *Department of Orthopaedic Surgery, Massachusetts General Hospital, Har- vard Medical School, Boston, MA. †Department of Orthopaedic Surgery and Rehabilitation, University of Io- annina, Ioannina, Greece. Address reprint requests to Kirkham B. Wood, MD, Department of Ortho- paedic Surgery, Massachusetts General Hospital, Harvard Medical School, 55 Fruit St Yawkey OCC #3800, Boston, MA 02114. E-mail: KBWOOD@PARTNERS.ORG 235 1040-7383/$-see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1053/j.semss.2012.05.008