Pediatric thoracolumbar spinal injuries: The etiology and clinical spectrum of an uncommon entity in childhood
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/neuroindia.NI_1243_15
Source of Support: None, Conflict of Interest: None
Background: Pediatric thoracolumbar (TL) spinal injuries are uncommon entities with an incidence of 5–34% of all pediatric spinal injuries. There is a scarcity of studies done on the pediatric population in the developing countries like India. This study aims to review our experience with TL spine injuries in children over a 12-year period.
Keywords: Epidemiology, pediatric, spine injury, thoracolumbar, translation injury
Pediatric thoracolumbar (TL) spinal injuries are rare but constitute an important cause of morbidity in this age group. The incidence of pediatric spinal injuries is 1–10% in the published literature.,, Thoracic and lumbar injuries constitute between 5.4% and 34% of all the spine injuries in children.,,, Only a few studies have specifically reported thoracic, lumbar, and sacral injuries in pediatric patients and most of the published studies are from the developed countries.,,, There is scarcity of studies done on pediatric population in the developing countries like India. This study aims to review our experience with TL spine injuries in children and determine the epidemiology, risk factors, mechanisms and patterns of injury, and management in our population.
We retrospectively reviewed all children with TL spine injury who were managed between January 2002 and June 2014 at National Institute of Mental Health and Neurosciences, Bengaluru, India. The data were collected from medical records and we identified all children (age 0–18 years) with spinal trauma (fractures, dislocations, discoligamentous lesions, and/or spinal cord injuries). Data including age, gender, mechanism of injury, pattern and level of injury, radiological assessment of the TL spine, clinical presentation, presence of neurologic deficits, associated injuries, mode of treatment, functional outcome, and mortality rate of the patients were collected.
All patients underwent diagnostic radiographic studies, including plain radiographs (anteroposterior and lateral view). Thin-slice computerized tomography (CT) scans were obtained whenever surgical intervention was planned or when there was a doubtful skeletal abnormality observed on the plain radiograph. Magnetic resonance imaging (MRI) was performed when the patients were planned for surgical intervention or when the plain radiograph was normal in patients with neurological deficits.
Review of the medical records revealed that 196 children with significant injuries of the spine had been managed during the 12-year and 6 month period, of which 90 had TL spine injuries. From January 2002 to June 2014, 4525 patients had been treated at our institute for spine injuries. Spine injuries in children (≤18 years) comprised 4.3% (n = 196), while TL spine injuries in children comprised 2% of all the spine injuries treated in all age groups (n = 90). The 90 children (90/196; 46%) with an injury of the TL spine formed the cohort for this study. The mean age of the studied population was 15.9 ± 3.2 years (range: 2–18 years). Overall, injury was more common in boys than girls in the cohort (3:1) [Table 1].
The most common mode of injury was fall from height. Sixty-four out of 90 (71.1%) children had sustained injury due to fall from height of more than 10 feet, 18/90 (20%) sustained injury following motor vehicle accident, 6/90 (6.7%) due to fall of a heavy object like a sandbag, heavy metal pipes etc., one patient sustained fracture (1.1%) due to a trivial fall while playing and the remaining one patient (1.1%) due to diving into a well from height of more than 30 feet. Thirteen out of 90 (14.4%) patients had associated injuries. The most common associated injury was long bone fracture. Eight out of 90 (8.9%) patients had only long bone fractures, 3/90 (3.3%) had only head injury. Two patients had sustained polytrauma, one of whom had head injury and femur fracture, and the other child had sustained head injury, humerus fracture, and multiple rib fractures with right hemothorax. The injuries were graded according to Frankel's grading [Table 1]. Most of the patients (27.8%) sustained Grade A injury (complete cord injury).
Lumbar spine was the most common spinal level injured [Table 2]. Of the 31 patients who sustained thoracic spine injury, 25 (27.7%) had fractures, 4 (4.4%) had subluxation and the other 2 children had spinal cord injury without radiographic abnormality (SCIWORA). Out of 48 children who had sustained lumbar spine injury, 44 (48.8%) had fractures and another four patients had sustained subluxation. Injury to the dorsolumbar junction was documented in 10 patients, out of which, five had sustained a fracture, one had sustained D12-L1 subluxation, one had sustained translation injury, and three had sustained both fracture and subluxation.
There were 14 cases of multilevel injury, out of which 11 were contiguous with involvement adjacent vertebra, and three were noncontiguous with at least one normal vertebra present between the injured vertebrae. Of the three patients with multilevel injury, one patient had sustained an injury to the craniovertebral junction and dorsal spine with type II C2 fracture and D6, D7, and D8 body fracture. Radiological instability was present in 62 of 90 (68.9%) patients.
All the patients were closely monitored and the vitals were stabilized. Surgical fixation was performed in 18 out of 90 patients. Though radiological instability was present in 62 (68.9%) patients, patients with complete injury (Frankel Grade A), patients with polytrauma and the patients who refused consent were not operated. Nine out of 18 patients (50%) underwent only fusion and another nine patients (50%) underwent laminectomy/corpectomy of the fractured vertebra in addition to fusion [Figure 1] and [Figure 2].
Follow-up was available in 21 patients (23.3%) with the median follow-up period of 5 months (range: 1–72 months). Out of these, 15 patients had undergone surgical fixation. Ten out of 21 patients (47.6%) had improvement of at least one grade in Frankel's grading and 13/21 (62%) patients were ambulant at follow-up, while there was no improvement in the other 5 patients (23.8%) [Table 3]. The outcome was compared with the level of injury, type of injury, Frankel's grade, and surgical procedure, but none of the analyzed variables had a statistically significant influence on the outcome.
Though children are often considered as smaller adults, the anatomical and physiological characteristics of almost every system in the body are significantly different in a child, based on the important tenet of development and growth. The spinal column of children is different from that of the adult by having increased cartilage/bone ratio, increased soft tissue hyperelasticity, and the presence of secondary ossification centers. As the children grow, the ossification center enlarges leading to reversal of cartilage/bone ratio.,, These differences influence the biomechanical properties of spine, especially the presence of secondary ossification centers. Physeal injury in children leads to physeal growth arrest which in turn leads to increased deformity compared to those without physeal arrest.
In the present study, pediatric spine injuries accounted for 4.3% of all spine injuries managed in our institute, while TL spine injuries in children comprised 2% (n = 90). The cervical spine injuries were more common than TL injuries in children. Other investigators also have reported similar increased incidence of pediatric cervical spine injury compared to TL injury in various studies.,, Dogan et al., however, reported 52.9% of TL injuries out of all pediatric spinal injuries. The higher incidence of cervical spine injury can be attributed to the increased range of motion in the cervical spine compared to the TL spine. The lower incidence of TL spine injury can be attributed to the protection offered by the rib cage. The injury often occurs at the TL junction where the relative more mobile lumbar segment meets the relatively fixed thoracic spine.
In our study, 81% of the children who sustained TL injury were in the age group of 15–18 years. Only 4.4% (4/90) of our cohort were <10 years of age. Most of the published studies reported increased incidence of TL injuries in adolescents,,, though some authors had reported higher frequencies of 17–21%, in children <10 years. The nucleus pulposus of intervertebral disc in younger children is more hydrophilic and the vertebral body is more cancellous due to which the vertebral body and the disc can withstand more compression before it fractures. Furthermore, as mentioned before, the rib cage adds to the protection of the thoracic spine. These factors can explain the lower incidence of TL injuries in the younger age group.
The etiologies of injury in children reflect the sociocultural differences between the developed and developing countries. While the majority of the studies from the West report motor vehicle accident as the most common cause of injury,,, our series noted that the most common cause of injury was fall from height (71.1%). This probably indicates that the children are less exposed to high-velocity motor vehicle injuries in our population. This also underscores the probable inadequacy of awareness and safety measures in living and work/play areas to prevent accidental falls and injuries to children in our population.
Children with multiple injuries pose a significant diagnostic and management difficulty. In a series by Dai et al., 19% of the children with multiple injuries had a delay in diagnosis of TL injury. While previous studies had shown head injury as the most common associated injury in TL spine injury, in our series 11.1% (10/90) of the patients had long bone fractures and only 5.6% (5/90) had head injury.,, This may be possibly due to the differences in the etiology and mechanisms of injury. While motor vehicle accidents may be prone to result in head injury, fall from height resulting in spinal fractures can have a higher probability of associated long bone fractures in limbs.
In this study, lumbar spine (53.3%) was the most common level injured. In contrast, thoracic spine has been reported as the most common level injured in some of the previous studies.,, In our series, the vertebral injuries ranged from minor injuries like posterior element fracture to major injuries such as burst fracture and translation. Most of these injuries were burst fractures (60%) followed by compression fractures (24.4%). Dogan et al., and Santiago et al., had reported a higher incidence of compression fractures compared to our study.,
SCIWORA was first described by Pang and Wilberger in 1982 and was defined as objective signs of acute traumatic myelopathy in the absence of spinal column findings on plain radiographs, flexion-extension radiographs, and/or a CT scan image. However, since the advent of MRI, many cases described as SCIWORA in the literature actually have demonstrable injury to the spinal cord, soft tissue components of the spinal column, or vertebral body endplate., The incidence of SCIWORA in various published series varies between 3.3% and 34%,, and around 13% of cases with SCIWORA involve the thoracic cord. The incidence of SCIWORA in the thoracic cord is lower than that seen in the cervical cord.,, In the present series, the incidence of SCIWORA of the thoracic cord was 6.4% (2/34) which is similar to that seen in the other studies.
In our study, the proportion of patients who underwent surgical intervention was 20% (18/90), which reflects our selection criteria for surgical intervention that patients with complete cord injury were not considered for surgery in the acute period. Most series from the West have reported similar proportion of surgical intervention in their cohort (16–33%).,, Lee et al., have described the TL injury classification and severity score. The classification system is based on three major categories namely the morphology of the injury, the integrity of the posterior ligamentous complex and the neurologic status of the patient. This score can be utilized as a guideline for the surgical management of TL spine injuries. Various other classifications of instability are also available., The factors influencing decision-making in our institute were management by multiple surgeons, presence of complete cord injury, as well as willingness of the patients for undergoing surgery.
Most of the children (16/18) in our series underwent posterior transpedicular fusion involving one level above and one level below the fractured vertebra. Two patients underwent a combined anterior and posterior approach and 360° fusion. Dogan et al., have also described an anterior approach and a combined anterior and posterior approach for the treatment of spinal fractures.
Various unique considerations need to be evaluated in the surgical management of pediatric spinal trauma. The smaller size of the vertebral body and pedicles have to be carefully inspected preoperatively and the correct size of the screws has to be selected. The margins of error are smaller with smaller pedicles and smaller spinal canals. Correction of the deformity should also take into consideration the future physiological growth of the patient.
Neurological recovery in children with SCI is thought to be better than that in the adult population. Several studies have shown a good neurological recovery following SCI, with incomplete lesions having the best prognosis, although severe complete injuries can also improve over time. In the present study, 16/21 patients (76.2%) had improvement of at least one grade in Frankel's grading and 13/21 (62%) patients were ambulatory at follow-up. However, the follow-up is very limited in our cohort and, therefore, there could be a bias in the determination of outcome parameters, with patients who may not have improved being lost to follow-up.
Progressive kyphosis at the involved vertebra is common and the incidence is more after a laminectomy and in patients who are treated conservatively for unstable spinal fractures., Nearly half of the patients in our series who underwent posterior fusion also underwent laminectomy of the involved vertebra. However, there was no difference in the neurological outcome and delayed deformity between patients who underwent a laminectomy and those who did not. There were no cases of delayed deformity in both medically and surgically managed patients. However, the follow-up in our cohort was not sufficient to make definite observations about the development of delayed kyphosis.
The incidence of TL spine injuries in pediatric patients is low, as had been reported in literature. The etiology of the injury is unique in the context of a developing country. TL injuries are most common in children older than 10 years and mainly involve the lumbar region. A significant proportion of the patients sustain associated long bone fracture or head injury. The mortality associated with TL injuries is low. When indicated, surgical fusion of the involved vertebra is safe and effective.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3]