Management and Outcome Analysis of Pediatric Unstable Thoracolumbar Spine Injury: Large Surgical Series with Literature Review.

Pediatric thoracolumbar spine fractures are considered rare injuries with paucity of publication across the globe. Further, spine injuries in children are comparatively rarer, and pediatric spine differs from adults, both biomechanically and anatomically; so, adult spine management strategy cannot be applied to pediatric cases, and exact guidelines for management of pediatric spinal injury is lacking. The current study is undertaken to study epidemiology, surgical management, and outcome of pediatric dorsolumbar unstable spine injury. A total of 25 pediatric patients were analyzed retrospectively with thoracic, thoracolumbar junction and lumbar spine injuries, who were managed surgically at our institute since June 2008, formed the cohort of the present study. There were 19 males and six females with a mean age 14.8 years. Clinically, complete spinal cord injuries were observed in 11 (44%), and rest 14 had incomplete injury. Most common mode of injury was fall (76%) in contrast to the western countries and the thoracolumbar junction was the most common affected site. Among all patients who underwent surgical intervention, 68% cases had posterior decompression and pedicle screw fixation. Mean duration of hospital stay was 18 ± 31 days. The mean follow-up period was 13.83 ± 5.97 months. In incomplete neurological injury group, a total of 13 patients showed neurological improvement, out of which four cases improved by two Frankel grades, eight patients improved by at least 1 grade, and rest one by Grade 3 (Frankel Grade B to E). Neurological outcome was statistically significant in relation to the level of fracture (P - 0.03) and preoperative Frankel grade, however, other factors, for example, gender, mode of injury, type of fracture, various surgical approaches with instrumentation, and correction of kyphotic deformity were found to be statistically nonsignificant. Surgical management of unstable pediatric dorsolumbar spine is a safe and an effective procedure which can provide good neurological outcome. The current study is one of largest series of cases managed surgically in this part of the world.

INTRODUCTION

Spine injuries in children which varies are relatively rare[12345678] and account for 2%–10% of all spine trauma.[137] Cervical spine injuries constitute most commonly involved region (60%–70%) in comparison to thoracic, thoracolumbar junction (D11-L2), or lumbar spine.[4] The pediatric spine differs biomechanically and anatomically in comparison to adults. The incidence of thoracic and lumbar spine injuries in children varies from 5.4% to 34%.[10111213] The fracture pattern in pediatric age is different comparing with adult owing to difference in anatomy and biomechanics of growing spine.[14] The greater mobility of pediatric spine is primarily contributed by shallow angulations of the facet joints, relative paraspinal immaturity, greater ligamentous flexibility, incomplete ossification of vertebrae, and greater elasticity of nucleus pulposus in the former. These factors collectively make a young child more vulnerable for ligamentous injuries. Further, inability to early detection of ligamentous injuries can lead to spinal dislocation, subluxation, deformity of spine, and spinal cord injury.[4] The pediatric spine injuries carry relatively higher morbidity and mortality.[811141516171819202122232425262728] There is scarcity of literature and only few series reported management of pediatric thoracic, thoracolumbar junction, and lumbar spine injuries.[11293031] A high index of suspicion is necessary to diagnose pediatric spinal injuries, especially in patients with polytrauma. The nature and level of injury are significantly impacted by the mechanism or mode of injury. Noncontrast computed tomographic (CT) imaging is mandatory and provides better delineation of the fracture and also helps to simultaneously allow screening of associated abdominal and chest injuries. Further magnetic resonance imaging is indicated in patients with spinal cord injury and especially to rule out disintegration of posterior osseoligamentous complex.[30313233] Treatment priorities include resuscitation of patient, treatment of the life-threatening associated injuries before mechanical restoration of the injured osteoligamentous column, and preservation or at least restoration of neurological functions.

Aims and objectives

The study was undertaken to evaluate incidence, mode of injuries, fracture pattern in thoracic, thoracolumbar junction and lumbar regions, various prognostic factors, managements, and outcomes of pediatric patients whose age was less than 18 years, who were admitted and managed at Jai Prakash Narayan Apex Trauma Centre between June 2008 and April 2012.

MATERIALS AND METHODS

Three hundred and twelve patients were managed for dorsal, dorsolumbar junction, and lumbar spine injury from June 2008 to April 2012, out of which 25 patients belonged to the pediatrics age group and formed the cohort of the current study. Inclusion criteria included age <18 years, with traumatic unstable fractures involving dorsal, thoracolumbar junction, or lumbar spine and presence or absence of neurological deficits. Patient's data were obtained from computerized patient's record system, clinical case note, operation notes, progress notes, morbidity/mortality note, Picture Archiving and Communication System (X-rays, CT-scan/magnetic resonance imaging [MRI] scans), and also from follow-up records.

Initial assessment on admission included mechanism of injury, time interval since trauma, detailed physical examination and an accurate assessment of the spinal stability and neurological status and the associated major injuries.

Patients were graded clinically using Frankel grading for neurological deficits; preoperative and postoperative radiological evaluation (CT-scan and/or MRI) were evaluated to characterize level of fracture, fracture pattern, spinal canal compromise, degree of kyphosis, and spinal cord compression.

Surgical managements for selecting type of surgical approach were based on neurological deficits, presence of unstable fracture, single vertebral body compression fracture with associated >50% loss of height of vertebral body, disrupted posterior column or more than minimal middle column failure, fracture dislocation and spondyloptosis, kyphotic angulations >20°.

Radiological evaluation including X-ray-anteroposterior and lateral views, CT scan with bone window, and three-dimensional reconstruction were carried out postoperatively and also on subsequent follow-up to look for deformity progression, implant position, and fusion status.

Follow-up assessment at 3, 6, and 12 months following surgery or whenever patient reported for follow-up or latest follow-up was examined for neurological assessment and evaluated with neuroimagings. Detailed neurological examination (Frankel grade), radiological imaging observed to assess fusion/deformity, and Denis pain scale were subjectively assessed by questionnaires.

The data were analyzed using SPSS version 20 software (IBM Corporation, New York, USA). Univariate analysis was done. Student's t-test/Mann–Whitney/Wilcoxon matched-pairs signed-ranks tests were used depending on the nature of continuous data. Fisher square test was used for qualitative data.

RESULTS

A total 25 patients (M = 19, F = 6) with pediatric dorsal, dorsolumbar junction, and lumbar spine injuries were identified. The youngest patient was 8 years of age; while oldest was 18 years with a mean age of 14.8 ± 2.8 years [Table 1]. Most common mode of injuries were fall from height in 19 (76%) patients, followed by road traffic accident 5 (20%) patients [Table 2].

Table 1

Sex distribution

Table 2

Mode of injury

Most common involved site was dorsolumbar junction (56%), dorsal spine (28%), and lumbar spine (16%). Most common type of fracture was burst fracture (64%), fracture dislocation (16%), flexion distraction and spondyloptosis (8%), and least common was compression fracture (4%).

On neurological evaluation basis, 11 patients sustained complete injury (Frankel Grade A), 5 patients and the remaining 4 belonged to Frankel Grade B and C, and rest 4 belongs to Frankel Grade D. Most common mode of injuries was fall from height in 19 (76%) patients, followed by road traffic accident (20%) patients [Table 3].

Table 3

Frankel grade in relation to spinal level of fracture

All patients were managed with surgical intervention, which included posterior decompression and pedicle screw fixation without corpectomy in 68%, posterior transpedicular corpectomy and graft placements (16%), and anterolateral approach in 2 patients [Figure 1]; long segment fixations were done in 72% cases and rest had short segment fixations [Figure 2]; the mean time interval of surgical intervention after injury was 10.5 ± 6.09 days with range from 2 to 31 days.

Figure 1

Preoperative computed tomographic scan of 18-year-old male, Grade III spondylosis at D11–D12 vetrebral level with associated spinal canal compromise (a) sagittal section image of dorsolumbar spine, (b) axial section image view

Figure 2

Postoperative tomographic scan of dorsolumbar spine, 18-year-old male, after surgical correction of Grade III spondylosis at D11–D12 vertebral level, showing complete reduction of fracture dislocation using posterior spinal approach with pedicle screw and rod fixation using (a) sagittal section, (b) axial section image

Perioperative complication included dural tear in two patients which were managed with primary dural closure and sandwich gel foam and fat graft. Another single patient had transient neurological worsening in postoperative period and postoperative MRI spine revealed spinal cord edema was noted at operative site, managed conservatively with steroid and patient gradually improved. Superficial wound infection was observed in 13% patients, all managed conservatively. Mean duration of hospital stay was 18 ± 31 days.

Follow-up

The longest follow-up period was 41 months and shortest being 3 months with a mean follow-up of 13.83 ± 5.97 months. However, two patients were lost to follow-up after 3 months following surgery, and hence, 23 patients were evaluated for neurological status, postoperative pain by questioner's methods, radiographic X-ray, and CT scan, where possible for evaluation of fusion status [Bar Diagram 1]. The neurological improvement in complete injury group (Frankel Grade A) was observed in 5 patients; however, 4 patients did not show any improvement. In incomplete neurological injury group at admission, 13 patients out of 14 improved but 1 patient failed to improve. Out of improving group, four patients improved by two Frankel Grade from C to E, 8 patients improved by 1 grade, Frankel Grade B to C, C to D and D to E, and 1 patient improved by Grade 3 (Frankel Grade B to E).

Bar Diagram 1

Neurological outcome in pediatric traumatic dorsolumbar spine injury in relation to location

DISCUSSION

Traumatic spinal injury in pediatric population is rare; few case series reported incidence of 1%–10%.[3478141718192023] The cervical spine is most commonly injured site in the pediatric spine injury while thoracolumbar spine injury is less common.[714] The fracture pattern and prediction of level of fracture in pediatric spine injury are different compared to adult population, which can be attributed to flexibility of spine and supporting structure, growing ossification center, mode of injury disproportional had size related to body, elasticity, and compressibility of bone.[20]

Dogan et al.[13] analyzed 89 pediatric patients with thoracic, thoracolumbar junction, lumbar and sacral spine injury and observed 46 were male and 43 were female with a mean age of 13.2 years. In our studies, the mean age was 14.8 ± 2.8 years with 19 male and 6 female.

In 2004, Carreon et al. analyzed mode of injury and noted motor vehicle accident accounted for 58% pediatric spine injury while Dogan et al.[13] reported motor vehicle accidents accounted for 57%, while sports-related accidents for 21%, but in the current study, most common cause of injury was fall from height 76%, followed by road traffic accidents 20% (n = 5).[8] Orenstein et al.[31] observed 67% of the injuries were caused by motor traffic accidents and 30% involved sports, while in series of Givens et al.,[32] traffic accidents accounted for in 68% of the injuries while the sports-related accidents contribute 28%.

In a study carried out by Dogan et al.,[13] out of 89 spinal injury cases, only 13 had neurological deficit, who had injury involving seven thoracic, five thoraco-lumbar, and 1 with lumbar spine injuries, respectively, and rest 76 patients were neurologically intact. In our study, on admission, complete spinal cord injury (Frankel Grade A) was observed in 11 patients, while incomplete spinal cord injury in 14 patients (Frankel Grade B = 5, C = 5, D = 4, E = 0). Among 11 patients with complete spinal cord injury group, six patients had dorsal lumbar junction fracture; rest five patients had dorsal fracture.

In 2003, Reddy et al.[27] analyzed the distribution of spine fractures in the pediatric population in relation to gender and mechanism of injury after reviewing 2614 pediatric cases, out of which 84 patients had vertebral fracture and additional 50 had neurological injury without radiological abnormality. Out of the total of 164 patients, the incidence of thoracic spine injury was 28.7%, lumbar spine injury was 23.2%, thoracolumbar junction in 14.6%, and least common was craniocervical junction injury in 6.7%. In a study by Dogan et al.,[13] thoracic spine fracture accounted for 46%, thoracolumbar region for 19.2%, lumbar region for 29.8%, and sacral region in 5%. In our study, most common site of fractures was dorsolumbar junction (56%) followed by dorsal spine fracture (28%) and least involved was lumbar spine 16% (n = 4).

In analysis carried by McAfee et al.,[2122] most common type of fracture was compression fracture, but in the current study, most common type of fracture was burst fracture. Most common site of fracture was thoracolumbar junction. Patients with fracture dislocation and spondyloptosis had higher incidence of associated neurological injury compared to simple fracture.

As X-ray spine may miss the spine injury. CT study of spine is mandatory in pediatric spinal injuries which provides better delineation of the fracture and also adds in the screening for associated chest and abdominal trauma. MRI imaging may be indicated in patients with suspected spinal cord injury and to rule out posterior osseoligamentous complex injury especially in patients showing severe tenderness during examination. Spinal cord injury without radiological abnormality is an entity that is unique to the pediatric population due to the cartilaginous nature of bone and elasticity of the soft tissue; the traumatic deforming force does not result in a bony fracture; however, the neural structures may be injured resulting in neurological deficit.[2930313233]

Management of spine injury becomes difficult especially in those having associated polytrauma.[10] Spinal injury may be associated with systemic injury including head injury, chest injury, intraabdominal injury, and long bone fracture.[51925] In analysis carried by Dogan et al., associated injury included head injury (28%), fracture of long bones (13%), chest and abdominal injuries (11.2%), and cervical spine injury (4.7%).[13] In the present study, chest injury and hemothorax/pneumothorax were noted in 4 patients and abdominal injury in 4 patients. However, there was no statistical significant association between mode of injury and associated injury.

Surgical management for thoracolumbar fractures is usually advocated for unstable spinal fractures or those presenting with neurological deficit. The surgical approach depends on the preference and experience of the surgical team; however, there is an increasing trend recently for the posterior approach. Further neurological injuries associated with thoracolumbar fractures need to be managed aggressively, especially in those cases with incomplete neurological injury. Even in cases with complete neurological injury, early surgical stabilization definitely helps to initiate early rehabilitation program usually within 2–3 weeks of the injury and can help to avoid respiratory infection, pressure sores, and urinary tract infections as well as psychological support.[33]

Management of thoracic and lumbar spine injury is guided by level of fracture, pattern of fracture, neurological deficits, kyphotic deformity, and integrity of osseoligamentous structure and presence of associated injury. In Dogan et al.'s series, 74% patients were managed conservatively and only 26% patients needed surgical intervention, which included posterior fixation in 16 patients, anterior corpectomy and fusion done in six cases, and combined approach in 1 patient.[13]

Surgical options for operative stabilization include posterior decompression and pedicle screw fixation without corpectomy, posterior transpedicular corpectomy and graft placements, anterolateral approach, combine anterior and posterior approaches (360°), and minimally invasive spine surgery.

Roy and Gibson[26] reported posterior fusion in pediatric patients is suitable and provides solid arthrodesis and excellent range of motion. However, Gambardella et al. and Glass et al. advocated equally efficacious of anterior or posterior approaches as both surgical approaches, provide good alignment, solid fusion, and successful arthrodesis.[1516] In our study, all (n = 25) patients underwent surgical intervention because of unstable spine fracture.

Moller et al. in 2006 reported good long-term outcome in thoracic and lumbar vertebral fracture and no or minor neurological deficit, in conservatively managed patients.[24] Dogan et al. observed complete neurological recovery in 75% having incomplete neurological and nine patients who had complete neurological deficit, 1 patient improved by 2 Frenkel Grade, 1 patient improved by 1 Frenkel Grade, and one recovered completely, however, rest 6 fail to show neurological recovery.[13] In the current series, patients available for follow-up in 23 patients; the neurological improvement was observed in 5 patients with complete injury (Frankel Grade A group) and rest 4 patients did not show any improvement. In incomplete injury group (Frankel Grade C, D, and E), 13 patients improved and one case failed to improve. The high recovery rate in pediatrics believed to be related to rapid healing property of bone and ligament in children.[18] There is statistical significant association with preoperative neurological status and good neurological outcome (P - 0.029).

Neurological outcome was also statistically significant in relation to the level of fracture (P - 0.03). Dorsal fracture has poor outcome compared to thoracolumbar and lumbar fracture. In dorsal spine injury, only 50% cases showed improvement whereas in dorsolumbar junction and lumbar spine, 84% and 91% were improved respectively in neurological status.

On comparison with the outcome of a spinal cord injury between pediatrics and adult population, spinal cord injury has a better prognosis in children.[30] Similarly, in other study, neurological improvements in children were better when compared to adult population. Neurological outcome was not statistically significant in relation to various other factors.

CONCLUSION

Pediatric spine injuries are rare. Fall from height was the major cause of spinal injuries in Indian pediatric patients in contrast to the western countries. Surgical interventions including instrumentation and fusions are effective. The neurological outcome depends on Frankel grade at time of admission as well as site of fracture along the spine. Mechanisms of injury, initial management, diagnostic examinations, and definitive treatment of such lesions are different from those of the adult, and an individualized approach to each case, looking for specific injury patterns, avoids miss diagnosis.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.