Minimally Invasive Spine Surgery for Unstable Thoracolumbar Burst Fractures: A Case Series

Introduction  Traumatic thoracolumbar burst fracture is a common pathology without a clear consensus on best treatment approach. Minimally invasive approaches are being investigated due to potential benefits in recovery time and morbidity. We examine long-term resolution of symptoms of traumatic thoracolumbar burst fractures treated with percutaneous posterior pedicle screw fixation. Methods  Retrospective clinical review of seven patients with spinal trauma who presented with thoracolumbar burst fracture from July 2012 to April 2013 and were treated with percutaneous pedicle screw fixation. Electronic patient charts and radiographic imaging were reviewed for initial presentation, fracture characteristics, operative treatment, and postoperative course. Results  The patients had a median age of 29 years (range 18 to 57), and 57% were men. The median Thoracolumbar Injury Classification and Severity Scale score was 4 (range 2 to 9). All patients had proper screw placement and uneventful postoperative courses given the severity of their individual traumas. Five of seven patients were reached for long-term follow-up of greater than 28 months. Six of seven patients had excellent pain control and stability at their last follow-up. One patient required revision surgery for noncatastrophic hardware failure. Conclusion  Percutaneous pedicle screw fixation for the treatment of unstable thoracolumbar burst fracture may provide patients with durable benefits and warrants further investigation.

Introduction Traumatic thoracolumbar burst fracture is a common pathology without a clear consensus on best treatment approach. Minimally invasive approaches are being investigated due to potential benefits in recovery time and morbidity. We examine longterm resolution of symptoms of traumatic thoracolumbar burst fractures treated with percutaneous posterior pedicle screw fixation. Methods Retrospective clinical review of seven patients with spinal trauma who presented with thoracolumbar burst fracture from July 2012 to April 2013 and were treated with percutaneous pedicle screw fixation. Electronic patient charts and radiographic imaging were reviewed for initial presentation, fracture characteristics, operative treatment, and postoperative course. Results The patients had a median age of 29 years (range 18 to 57), and 57% were men. The median Thoracolumbar Injury Classification and Severity Scale score was 4 (range 2 to 9). All patients had proper screw placement and uneventful postoperative courses given the severity of their individual traumas. Five of seven patients were reached for long-term follow-up of greater than 28 months. Six of seven patients had excellent pain control and stability at their last follow-up. One patient required revision surgery for noncatastrophic hardware failure. Conclusion Percutaneous pedicle screw fixation for the treatment of unstable thoracolumbar burst fracture may provide patients with durable benefits and warrants further investigation.

Methods
A retrospective clinical review was conducted of adult patients with spinal trauma who presented with thoracolumbar burst fracture from July 2012 to April 2013 at a single institution and were treated with percutaneous pedicle screw fixation. Electronic patient charts and radiographic imaging were reviewed for initial presentation, fracture characteristics, operative treatment, and postoperative course. Patients were contacted for long-term follow-up to assess resolution of pain and other symptoms from their injury.

Surgical Technique
Patients were taken to surgery the day of or one day following injury. Pedicle screws were placed one level above and one level below the level of injury unless the fixed level would be adjacent to the thoracolumbar junction. In this case, the construct was extended an additional level. All pedicle screws were placed in the same fashion. After induction of general anesthesia, preoperative somatosensory evoked potentials were obtained and the patient was turned prone on a Jackson table. The back was prepped and draped in usual sterile fashion. The fluoroscopy unit was draped, and biplanar fluoroscopy was obtained of the injury to visualize and mark the appropriate pedicles. Paramedian incisions were made at the locations of the pedicles. Meticulous hemostasis was achieved with electrocautery. A Jamshidi needle was placed under fluoroscopy, then followed with a K-needle to confirm the trajectory of the pedicle screws (K2M, Serengeti Minimally Invasive Spine Surgery System, Leesburg, Virginia). The Jamshidi needle was removed, and dilators were placed into the wound. The pedicles were then tapped and the pedicle screws were placed. Bilateral posterolateral fusion grafts of corticocancellous bone chips and demineralized bone matrix were placed adjacent to the construct. Wounds were then closed in layers as appropriate.

Results
Seven patients with thoracolumbar burst fracture were treated with percutaneous screw fixation. The cases are summarized in ►Tables 1 through 4. The median age of patients was 29 years (range 18 to 57), and 57% were men. The mechanism of injury was motor vehicle accident in four patients and fall from height in three patients. Only two patients presented with thoracolumbar spine monotrauma. All of the patients were neurologically intact on presentation except for one patient who had cauda equine syndrome. The median Thoracolumbar Injury Classification and Severity Scale score was 4 (range 2 to 9). One patient presented with multiple thoracolumbar fractures; the rest of the patients only had a single fractured vertebra. Four patients had an injury at L1, one at L2, and one at T3. The seventh patient had multiple fractures with injuries at T10, T11, and L4. The preoperative loss of height of the injured vertebrae had a median of 17.5% (range 5 to 60%). The median preoperative Cobb angle was 8 degrees (range 4 to 21 degrees). Four patients had no canal compromise, two had 5 to 10% compromise, and two had 75% compromise.
All patients underwent surgery using the procedure described previously. Patients were instrumented either one or two levels above and below the level of the fracture at the surgeon's discretion. Only one intraoperative complication occurred: in the patient with multiple fractures, an anterior breach of the left L3 pedicle led to a hemorrhage. No intervention was required for the hemorrhage. Immediate postoperative imaging revealed no misplaced screws. The loss of height was reduced to a median of 5% (range 5 to 40%), and the Cobb angle was corrected to a median of 5 degrees (range 2 to 11 degrees). All four patients with preoperative canal compromise had significant resolution of the canal compromise. The median length of admission for our patients was 6 days (range 4 to 22). All radiographic factors remained stable at the last imaging follow-up (median 9.5 months, range 6 to 18 months).
Five of the seven patients were reached via telephone to assess long-term resolution of pain and other symptoms. On average, the follow-up was 32.8 months (range 29 to 38) for these five patients. The last follow-up of the two patients who could not be reached via telephone was at the time of their last imaging follow-up at 10 and 6 months. Six of the seven patients had near-complete to complete resolution of their lower back pain. ►Figs. 1 to 4 demonstrate radiographic findings for patient number 1, who had an excellent outcome.
The patient with cauda equina syndrome had complete recovery of neurologic function without pain at last clinical follow-up. The one patient who required an open revision is still in significant pain and has been working closely with a pain specialist to control her symptoms.

Discussion
Currently, there is no consensus on the optimal treatment of thoracolumbar burst fractures. Due to improvements in spinal instrumentation technology and improved operator skill, multiple groups have argued for the use of percutaneous pedicle screw placement as part of the treatment for this pathology. As noted by Dhall et al in 2014, percutaneous pedicle screw fixation for unstable thoracolumbar fusion has not been examined in the literature with sufficiently long follow-up. 11 Many studies have examined the use of percutaneous pedicle screw placement for thoracolumbar burst fractures. Shen et al compared posterior fixation versus nonoperative treatment and found no difference in functional outcome at 2 years, but short-segment fixation showed advantages in pain relief and kyphosis correction. 12 Blondel et al treated 29 patient with A3 fractures using a combination of percutaneous pedicle screw placement and balloon kyphoplasty and achieved 11 degrees of local kyphosis correction with a 2degree angle loss at last follow-up of 24 months. 2 No cases of screw migration were noted; however, two patients had lateral cement leakage. All patients had significant improvement in pain. This findings are similar to the series by Fuentes et al. 13  Our series includes longer-term follow-up than most of the aforementioned studies. Our patients generally reported good pain and functional outcomes following their surgeries. One patient required a second surgery after T12-L2 pedicle screw fixation due to increased back pain resulting from a left L2 screw fracture, increased anterior wedging at L1, and focal kyphosis at T12-L1. The length of stay for our patients is longer than some in the literature, but this is likely due to the other traumatic injuries sustained by our patients in addition to their thoracolumbar injury.
Several studies have investigated the use of balloon kyphoplasty for management of thoracolumbar burst fractures. One benefit of posterior fixation is the avoidance of risks    15 Five patients suffered from radiculopathy secondary to cement extrusion. Meta-analyses have reported rates of symptomatic cement extrusion to be between 0.2 and 1.5%. [16][17][18][19][20][21] In comparison, one intraoperative complication was noted within our series, yielding a rate of 14.3%. Further study with a larger sample size is necessary to ensure validity. Postoperatively, six of the seven patients within our series had good outcomes during their follow-up. However, one patient with multiyear follow-up experienced hardware failure after 8 months of significant pain relief. Revision surgery provided stabilization for 38 months (with ongoing follow-up at this time), although she developed lower back and bilateral thigh pain after several months of pain relief when she became pregnant. This pain was attributed by a pain specialist to bilateral sacroiliac joint dysfunction, which was found in a prospective study to have an incidence rate of 6.3% in pregnant women. 22 The pain was treated with joint injections and oral medications.
It is unclear what role her hardware failure and subsequent revision surgery played in the development of sacroiliac joint dysfunction. However, this case highlights the need for long-term follow-up with these patients given that hardware failure can occur many months after surgery. Currently, the literature reports pedicle screw fracture to be a not uncommon event. In a report of 38 patient with pedicle screw instrumentation for thoracolumbar burst fractures, Carl et al described two cases of broken pedicle Table 3   screws and seven cases of bent pedicle screws during an average follow-up of 22 months. 23 In a study examining the effect of transpedicular grafting on short-segment pedicle screw fixation for thoracolumbar burst fractures, Alanay et al reported 2 of 20 patients with pedicle screw fracture. 24 Further investigation is required to determine if the risk of pedicle screw fracture, or other hardware failure, is significantly different in our procedure from accepted values in the literature due to a seven-patient series. Ultimately, treatment of unstable thoracolumbar burst fractures with percutaneous pedicle screw fixation may be a viable strategy that requires further investigation to evaluate long-term outcomes and adverse effects in large cohorts. This investigation provides some evidence that such studies will be warranted due to the resolution of pain seen in the majority of our patients.

Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of Interest
The authors have no personal or institutional interest with regards to the authorship and/or publication of this manuscript.
Contributorship Statement P.A.C., N.A., and R.F.S. were involved in the design and conception of this manuscript. P.A.C. and R.F.S. performed the literature search. P.A.C. compiled the primary manuscript. N.A. and R.F.S. critically revised the manuscript. All authors have approved the manuscript as it is written.