Fractures and severe injuries to the spine in children are relatively rare and represent between 1 and 10% of injuries reported by various authors.1,5,11,20,25,27 Ruge et al27 found that only 2.7% of 2598 patients with spinal injuries presenting to their institution were younger than 12. The injuries that are more commonly seen in children include a relatively higher prevalence of occipitoatlantal dissociation, upper cervical spine injuries, atlanto-axial rotary subluxation, multiple thoracic compression fractures, vertebral end plate fractures, and spinal cord injury without radiographic abnormality (SCIWORA).
Anatomic and biomechanical differences between the developing child and the adolescent or adult account for the different patterns of injury occurring in different age groups. The ligaments, discs, and surrounding soft tissue structures are more elastic and the musculature is less well developed in children than in adults. This feature explains the relative resilience of the pediatric spine to injury. This also explains why spinal cord injury without evidence of fracture or dislocation and intrapartum birth injuries occur in children and infants.4,21
Spinal injuries in the very young child are relatively uncommon compared with older children, adolescents, and adults. The increased soft tissue elasticity and anatomical differences in children under age eight make upper cervical spine injuries more likely. The frequency of injury is similar between boys and girls. Spinal cord involvement is more common, but the prognosis is better for partial injuries. Surgery is less likely to be needed in very young children than in adolescents and adults. After age eight, the pattern of injuries becomes more like that in the adult. This paper reviews the literature on frequency of cervical injuries in very young children as well as the evaluation and treatment of these injuries.
The developing spine is characterized by the presence of ossifying cartilage, which forms ossific centers joined by synchondroses. These can be confused with fractures in the immature spine and occasionally can be the site of injury. The atlas forms from three ossification centers. There are two for the lateral masses and a third for the anterior arch. The latter may not be visible radiographically at birth and may not appear until 1 year of age. Fusion of the anterior arch of C-1 with the lateral ossification centers usually is complete by age 7 years. By this time, posterior arches are also fused but may appear bifid in spite of an intact cartilaginous ring.
The axis is more complex and formed from ossification centers for the arches and the centrum or body in addition to two centers for the dens, which are radiographically fused at birth. These centers are separated from the centrum by a cartilaginous physis. This cartilaginous physis, the dentocentral cartilage, is positioned below the level of the atlanto-axial facet joints and closes between 5 and 7 years of age. Injuries to this structure heal readily. The location of the dentocentral cartilage within the body helps to avoid a mistaken diagnosis of a dens fracture that more often occurs at the base of the dens. The ossiculum terminale at the tip of the dens appears at about age 7 and fuses to the dens at about age 12. The blood supply to the dens enters from the apical area of the ossiculum and from vessels near the facet joints. Damage to these vessels is responsible for nonunion in the more mature spine. These upper cervical segments drain into the nasopharyngeal and deep cervical lymph nodes. The nasopharynx also drains into this region and this important fact is why infections and surgically induced inflammation in this area are implicated in C1/C2 rotary subluxation (Grisel’s syndrome). The development of the rest of the cervical, thoracic, and lumbar spine follows a similar pattern with two lateral centers of ossification and a third for the vertebral body. The dimensions of the spinal canal increase until closure of the neurocentral synchondroses. Typically, the neurocentral synchondroses close between 3 and 6 years of age and the posterior arches close between ages 2 to 4. These areas should not be confused with fractures (Fig 1).3
The horizontal orientation of the cervical facet joints as seen in lateral view changes to a more vertical orientation with growth. The angulations of the upper cervical facets change from a relatively horizontal angle of about 30° at birth to 60 to 70° by adolescence. The lower cervical segment changes from 55 to 70°. The horizontal orientation of the facet joints along with the relative ligamentous laxity in young children contributes to the common finding of upper cervical subluxation and the relatively higher frequency of upper cervical injuries in young children (Fig 2). Another contributing factor to this pattern of injury is the cartilaginous uncinate process or joints of Luschka, which do not ossify until about the age of 7 years, after which they may contribute to stability.19 The size and weight of a newborn’s head with respect to the weight of the body is greater and the ratio gradually decreases with age. This fact causes the fulcrum of spinal flexion to be located at C2-C3 in very young children, shifting to C5-C6 by late childhood. Radiographically, the spinal anatomy is similar to the adult by age 8 years, when the pattern of injury begins to change to a more adult pattern. Before that, the relatively increased mass of the head and the more horizontal facet joint orientation contribute the relatively greater frequency of upper cervical injuries in young children.7 Finally, the thick cartilaginous vertebral endplates can be a site of injury, separating from bone as they do in the appendicular skeleton.
The type and location of injury varies with age of the patient. Osenbach and Menezes20 reported on 179 children from birth to age 16 years treated for vertebral column and/or spinal cord injury during an 18-year period. These patients accounted for only 9% of all spinal trauma patients treated. Sixty-two were younger than 8 years of age and 117 were between 8 and 16 years. The younger children sustained a higher incidence of cervical spine injury (79%) than the older children (54%). The upper cervical spine was injured twice as frequently in the younger children. Lower cervical and thoracic injuries were equally frequent in both groups but thoracolumbar junction injuries were more common in the older children. The incidence of neurological deficit was higher in the younger group than in the older group (62% versus 47%, respectively). Nondisplaced fractures were more common in the older patients, but subluxation without fracture and SCIWORA injuries were more frequent in the younger patients. All injuries that involved subluxation alone involved the cervical spine and 84% of those were in the upper cervical spine. The majority of children with fracture and subluxation of the spine (76%) and SCIWORA (76%) had injuries that involved the cervical spine.20
Hadley et al10 reviewed 122 cases of spinal cord and vertebral column injury. There were 18 patients aged 0 to 9 years, 38 patients aged 10 to 14 years and 66 patients aged 15 and 16 years. There were more girls in the youngest patient group. The most common causes of injury were pedestrian automobile accident and falls in the 0 to 9 years age group, whereas the most common injuries in older children were caused by motor vehicle accidents and sports-related traumas. The youngest group also was more likely to sustain neurologic injury. There was a higher incidence of SCIWORA in that age group. In the 0 to 9 years group, 72% of the injuries were in the cervical spine and 50% were centered between the occiput and C2. The older children were injured at levels and with the patterns observed in the adult population. Sixteen had multiple levels of spinal injuries. Only 16% of the 122 patients needed operative intervention. Eighty-nine percent of 38 patients with incomplete neurologic injury improved and 23 were neurologically intact at a median followup of 44 months. Four of 20 patients with complete spinal cord injury improved. Thirty-three percent of those with neurological injury had SCIWORA. These patients were younger and had a worse prognosis.10
In a study of 71 patients younger than 12 years old, Ruge et al27 found that children younger than 3 years represent a subpopulation. Forty-seven of the patients had a spinal cord injury. Overall, the average patient age was 6.9 years and 43% were girls. Falls (38%) were the most common cause of injury followed by automobile-related injuries (20%), with more than half of them automobile-pedestrian accidents. The remainder of injuries consisted of miscellaneous injuries, including gunshot, sports-related, and birth-related injuries. The three years and younger group represented only 9 of 47 patients. Six (77%) of the nine patients under age three were girls and 5 (55%) had C1-C2 injuries. Five (55%) needed surgical intervention. A statistical comparison between children 0 to 3 years of age with those 4 to 12 years showed a significantly higher incidence of C1-C2 injuries, a greater frequency of girls, and less need for surgical stabilization than for children aged 4 to 12 years.27 Therefore, in children younger than 8 years old, most large series have found a relatively higher ratio of girls to boys and a higher prevalence of injuries caused by falls, pedestrian automobile injuries, child abuse, and birth injuries. Injuries in these patients tend to be centered in the upper cervical spine and are more often ligamentous. Although fractures were less common in the younger than 8 group, they do occur in younger children through the cartilaginous endplates and can produce a disturbance of growth.2 Spinal cord injury is relatively more common in younger than in older children and incomplete spinal cord injuries are more likely to improve in the younger patients.1,6,11,20,25,27 In children older than 8 to 10 years, the epidemiology and patterns of injury increasingly resemble those of the adult.
A history of considerable trauma such as a motor vehicle accident, motor vehicle-pedestrian accident, lack of vehicle restraint, fall from a height, penetrating injury, sports accident, or suspicion of child abuse should alert the clinician. Loss of consciousness or clinical symptoms of neck pain, muscle spasm, and pain with vertebral palpation often are present. Signs of trauma to the head, face, or deformity such as torticollis may be present. Bruising and abrasions caused by seat and shoulder belt restraints are common. In newborns and very young children, such signs may be absent. Small children may complain of occipital headache with upper cervical injury and may cradle their head with their hands. Attention should be focused especially on the upper cervical spine in children younger than 8 years. A careful neurologic examination must be done. Noncontiguous spinal injury may be present.13 A diligent search for other injuries such as hemopneumothoracic, aortic, cardiac, abdominal, renal, and bladder injuries must be done.
Children with suspected spinal injuries should be immobilized on a spine board. The neck should be protected against unwanted motion with a cervical collar and sandbags or blankets. Because of the relatively large size of the head compared with the torso in children younger than 8 years, immobilization on a standard spine board produces an undesirable flexion of the cervical spine. This can be avoided by elevating the torso relative to the head on additional supports such as a thin mattress or using a spine board with a recess that allows the head to be lowered relative to the trunk.14 During the remainder of the physical examination and during transportation for other studies or to another institution, the patient’s head and neck are kept in alignment with the rest of the spine and trunk. Flexion of the cervical spine is avoided (Fig 3). Traction in young children must be avoided when upper cervical or atlanto-occipital injuries are possible. Findings from the neurologic examination should be documented with care on a neurologic injury sheet, and the examination should be repeated when there are positive findings or if spinal injury is suspected. If the patient is alert, cooperative, and does not complain of neck pain; if there is no pain with palpation; and if there is a painless full active range of motion, then immobilization can be discontinued.
The standard trauma series includes AP, lateral, and open-mouth radiographs of the cervical spine, which must include C7 through T1. These are done if there is pain, unconsciousness, or positive findings on physical examination. Open-mouth views are difficult to interpret in children younger than 5 years and are comparatively of little value. Active flexion-extension lateral cervical spine radiographs may be done to look for instability in awake, cooperative and neurologically intact patients. They are more likely to be useful if there are suspicious findings on plain AP and lateral radiographs such as segmental kyphosis, equivocal subluxation, or anterior soft tissue swelling.24 The flexion-extension radiographs may not be conclusive if a full range of motion is not possible because of pain or muscle spasm and may need to be delayed. Anteroposterior and lateral radiographs of the thoracic and lumbar spine are done before a patient is removed from the spine board. Anteroposterior, open mouth, and sometimes, lateral radiographs of the upper cervical spine may be difficult to interpret in small infants because of overlapping of the scull or mandible. A lateral radiograph of the scull will aid in the evaluation of the upper cervical and occipital region in these cases. Certain radiographic findings in the immature spine can be confused with injury. Because of the ligamentous laxity and horizontal orientation of the facet joints in young children, there can be a step-off seen on the lateral cervical spine radiograph at C2-C3 and less commonly at C3-C4. This should be smaller than 4 mm. The lateral view should show four smooth contiguous lines along the posterior spinous processes, the posterior laminar line, and the posterior and anterior border of the vertebral bodies (Fig 4). The facets should be well aligned. The soft tissue retropharangeal space at C2 should be no more than 7 mm and the retrotracheal space at C6 should be smaller than 14 mm. These areas may appear abnormally wide in a child who is crying. Widening of the disc space, small avulsion fractures, and spinous process fractures may be associated with apophyseal separation. Occasionally, subtle abnormalities are seen in a child who has a healthy active range of motion and no further pain, leading to the conclusion that there is a normal variant. If not, further investigation is needed. The atlanto-dens interval should be smaller than 4 mm in children younger than 8 years and smaller than 3 mm in older children. Rupture of the transverse atlantal ligament will allow more than 4 mm of anterior displacement of the atlas. In very young children, up to 5 mm may be considered healthy.33 The alar ligaments prevent further displacement. When there is 10 to 12 mm of displacement, these ligaments are ruptured and cord compression results.31 Evaluation of the occipital-cervical junction should begin with plain lateral radiographs. A line should be drawn between the basion B and the posterior arch of the atlas C. Another line is drawn between the opisthion O and the anterior arch of the atlas A. Dividing the distance BC by OA is known as the ratio of Powers et al and averages 0.77. A value of greater than 1.0 is diagnostic of atlanto occipital dislocation23 (Fig 5). Another measurement useful in trauma is measuring the distance between the tip of the odontoid and the basion. This distance is 5 mm on the lateral radiograph in an adult, but it may be up to 10 mm in a child younger than 8 years.36 Harris et al12 described the basion-axis distance, which is the distance between the basion and a vertical line projected cephalad from the posterior border of the odontoid process. This should measure fewer than 12 mm in children younger than 13 years (Fig 6). Computed tomography (CT) scans may be more accurate for detecting high cervical fractures and occipital condyle fracture. Soft tissue swelling seen on lateral radiographs should prompt further investigation.
In uncooperative, intubated, or obtunded patients, it is advantageous to clear the spine within 72 hours of admission. This can be done with magnetic resonance imaging (MRI) or CT scanning at the same time if these studies are needed for evaluation of brain or abdominal injuries. Frank et al9 established that their protocol of doing MRI to clear the cervical spine within 72 hours of admission to the intensive care unit if it could not be cleared by other methods, which allowed earlier removal of the collar, improved mobilization and pulmonary care, and fewer skin problems. This resulted in an average decrease in hospitalization of 4.6 days and a savings of $7,700 per patient. They found that the T2-weighted images on MRI were more sensitive in diagnosing soft tissue ligamentous injuries.9 Although CT scanning gives better bony detail, MRI can delineate bone and soft tissue injury and permits predictive classification of spinal cord injuries.29
Although plain tomography scans show excellent bone detail, they use more radiation than CT scans and are increasingly unavailable in hospitals. Myelography still is useful for studying spinal cord compression, but is invasive.
Injuries to the atlanto-occipital articulation and the upper cervical spine are a common form of injury in very young children. The spectrum of injuries includes fractures, subdural hematoma, spinal cord injury, techtorial membrane injury, facial injury, and muscular-ligamentous injury. Atlanto-occipital dislocations historically have been considered to be rare and often fatal. Improved resuscitation and transportation of patients sustaining these injuries has increased the rate of survival. Diagnosis of these rare injuries is difficult, particularly in young children. They may be associated with brain, head, neck, facial, and spinal cord injury, but are not necessarily associated with these. Most are caused by automobile-pedestrian injuries. In one large series,34 the average age of the patients was 3.5 years. Neck pain and signs or symptoms of spinal cord injury should prompt investigation when there is a history of severe trauma. It is essential to diagnose these injuries as soon as possible in order to avoid causing neurologic damage because of undetected instability.16 Magnetic resonance imaging provides detailed anatomic imaging of the neural and ligamentous soft tissues in this region.34 Fusion from the occiput to C2 usually is recommended for patients who survive and who are unstable. For the patient without neurologic deficit (which is unusual), occiput to C1 fusion has been recommended in order to preserve head rotation.30
Odontoid physeal fractures are rare and unique to young children, occurring through the neurocentral synchondroses of C2, which may not fuse until age 7. Odontoid fractures would be best diagnosed on the lateral cervical spine radiograph. The dens may be tipped forward; less frequently the dens may be oriented posteriorly. The majority of these injuries heal with external immobilization, most often with a halo vest or cast.28 If needed, closed reduction can be done with anesthesia after halo placement with fluoroscopic control. Surgery often is not needed unless alignment cannot be maintained.18
Atlanto-axial rotary subluxation can occur in young children. This may be caused by trauma in approximately 20 to 45% of patients. Other causes include recent upper respiratory tract infection or inflammation from recent surgery (Grisel’s Syndrome). These children present with the so-called cock robin position, with the chin rotated to one side and the head flexed to the opposite side. In infants and very young children, this must be distinguished from congenital muscular torticollis. These children have tightness in the sternocleidomastoid muscle on the side opposite to which the chin is rotated. In infants, there is a palpable mass. Acquired torticollis should lead to investigation of atlantoaxial rotary subluxation. Other nontraumatic causes can include congenital anomalies, tumor in the upper cervical region, or posterior fossa, and Arnold-Chiari malformation, syringomyelia. Interpretation of plain radiographs of the upper cervical spine can be difficult because of head tilting and rotation. Open-mouth oblique radiographs can be done in cooperative patients, but in younger children dynamic CT scans yield the best information. The first line of treatment involves the use of a cervical collar, anti-inflammatory medication, and occasionally, physical therapy. Cervical traction and muscle relaxants may be needed when treatment is delayed. Surgery should be needed in a minority of cases. The most important factor influencing outcome is a delay in diagnosis and treatment.22,32
Birth-related injuries to the spine are rare; however, one study35 found that 10% of stillborn infants had evidence of spinal cord and brainstem injury at autopsy. They most often involve the cervical spine, followed by the cervical-thoracic junction, and thoracolumbar junction. There is an association with forceps delivery and presentation with apnea and initial flaccid paralysis and a high mortality rate.17 In one study of five neonates with birth-related spinal injuries, four were given an incorrect diagnosis including Werdnig-Hoffman syndrome, occult myelodysplasia, and birth asphyxia. Only one had an abnormal plain xray.26 Thoracic and thoracolumbar injuries have been seen with a transverse position.15 Diagnosis of spinal injury may be difficult with plain xrays in the largely cartilaginous neonatal spine. Ultrasound has been described for detecting spinal cord disruption at the bedside and has been compared with MRI.8 Immobilization is needed if instability is present and surgical stabilization for displaced injuries may be warranted.
Spinal injuries in the very young child are relatively uncommon compared with older children, adolescents, and adults. The increased soft tissue elasticity and anatomic differences in children younger than 8 years make upper cervical spine injuries more likely. The frequency of injury is similar among boys and girls in contrast to a higher ratio of boys to girls in older children and adolescents. Spinal cord involvement is more common, but the prognosis is better for partial injuries. Spinal cord injuries more frequently occur in the absence of apparent bony injury. Surgery is less likely to be needed in very young children than in adolescents and adults. Ruge et al,27 in their study limited to children younger than 12 years, found that these trends are more pronounced in children younger than 3 years. The unique anatomy of the growing upper cervical spine, more horizontal facets, increased soft tissue elasticity, and increased mass and size of the head compared to the trunk explain these findings. After age 8 years, the pattern of injuries becomes more like that in the adult. Injury to the spine of the young child is comparatively rare. Therefore, it is hoped that this review of the anatomy, epidemiology and radiographic evaluation will be useful to those who will occasionally have a need to evaluate a child with a spinal injury in early childhood.
1. Anderson JM, Schutt AH: Spinal injury in children: A review of 156 cases seen from 1950 through 1978. Mayo Clin Proc 55:499-504, 1980.
2. Aufdermaur M: Spinal injuries in juveniles. Necropsy findings in twelve cases. J Bone Joint Surg 56B:513-519, 1974.
3. Bailey DK: The normal cervical spine in infants and children. Radiology 59:712-719, 1952.
4. Byers RK: Spinal-cord injuries during birth. Dev Med Child Neurol 17:103-110, 1975.
5. Dickman CA, Zabramski JM, Hadley MN, Rekate HL, Sonntag VK: Pediatric spinal cord injury without radiographic abnormalities: Report of 26 cases and review of the literature. J Spinal Disord 4:296-305, 1991.
6. Eleraky MA, Theodore N, Adams M, Rekate HL, Sonntag VK: Pediatric cervical spine injuries: Report of 102 cases and review of the literature. J Neurosurg 92:12-17, 2000.
7. Farley FA, Hensinger RN, Herzenberg JE: Cervical spinal cord injury in children. J Spinal Disord 5:410-416, 1992.
8. Fotter R, Sorantin E, Schnieder U, et al: Ultrasound diagnosis of birth related spinal cord trauma: Neonatal diagnosis and follow up and correlation with MRI. Pediatr Radiol 24:241-244, 1994.
9. Frank JB, Lim CK, Flynn JM, Dormans JP: The Efficacy of MRI in Pediatric Cervical Spine Clearance. Spine 27:1176-1179, 2004.
10. Hadley MN, Zabramski JM, Browner CM, Rekate H, Sonntag VK: Pediatric spinal trauma: Review of 122 cases of spinal cord and vertebral column injuries. J Neurosurg 68:18-24, 1988.
11. Hamilton MG, Myles ST: Pediatric spinal injury: Review of 174 hospital admissions. J Neurosurg 77:700-704, 1992.
12. Harris JH, Carson CG, Wagner LK: Radiologic diagnosis of traumatic occiptocervical dissociation: 1. Normal occipitocervical relationships on lateral radiographs on supine subjects. AJR Am J Roentgenol 162:887-892, 1994.
13. Heilman CB, Riesenburger RI: Simultaneous noncontiguous cervical spine injuries in a pediatric patient: Case report. Neurosurgery 49:1017-1020, 2001.
14. Herzenberg JE, Hensinger RN, Dedrick DK, Phillips WA: Emergency transport and positioning of young children who have an injury of the cervical spine: The standard backboard may be hazardous. J Bone Joint Surg 71A:15-22, 1989.
15. Journeau P, Bourcheiux LM, Wagner A, Padovani JP, Pouliquen JC: Obstetric Dislocation of the Thoracic Spine: Case report and review of the literature. J Pediatr Orthop 10B:78-80, 2001.
16. Kenter K, Worley G, Griffin T, Fitch RD: Pediatric traumatic atlanto-occipital dislocation. J Pediatr Orthop 21:585-589, 2001.
17. MacKinnon JA, Perlman M, Kirpalani H, et al: Spinal Cord Injury at Birth: Diagnostic and prognostic data at birth in twenty two patients. J Pediatr 122:431-447, 1993.
18. Mandabach M, Ruge JR, Hahn YS, McClone DG: Pediatric Axis Fractures. Pediatr Neurosurg 19:225-232, 1993.
19. Ogden JA: Spine. In Skeletal Injury in the Child. Ed 3. New York, Springer Verlag 708-789, 1999.
20. Osenbach RK, Menezes AH: Pediatric spinal cord and vertebral column injury. Neurosurgery 30:385-390, 1992.
21. Pang D, Wilberger Jr JE: Spinal cord injury without radiographic abnormalities in children. J Neurosurg 57:114-129, 1982.
22. Phillips WA, Hensinger RN: The Management of Rotatory Atlanto-axial Subluxation in Children. J Bone Joint Surg 71A:664-668, 1989.
23. Powers B, Miller MD, Kramer RS, Martinez S, Gehweiller Jr JA: Traumatic anterior atlanto-occipital dislocation. Neurosurgery 4:12-17, 1979.
24. Ralston ME, Chung K, Barnes PD, Emans JB, Schutzman SA: The role of flexion-extension radiographs in blunt pediatric cervical spine trauma. Acad Emerg Med 8:237-245, 2004.
25. Rekate HL, Theodore N, Sonntag VK, Dickman CA: Pediatric spine and spinal cord trauma: State of the art for the third millennium. Childs Nerv Syst 15:743-750, 1999.
26. Rossitch Jr E, Oakes WJ: Perinatal spinal cord injury: clinical, radiographic and pathologic features. Pediatr Neurosurg 18:149-152, 1992.
27. Ruge JR, Sinson GP, McLone DG, Cerullo LJ: Pediatric spinal injury: The very young. J Neurosurg 68:25-30, 1988.
28. Sherk HH, Nicholson TT, Chung SM: Fractures of the odontoid process in young children. J Bone Joint Surg 60A:921-924, 1978.
29. Sledge JB, Allred D, Hyman J: MRI in Evaluating injuries to the pediatric thoracolumbar spine. J Pediatr Orthop 21:288-293, 2001.
30. Sponseller PD, Cass JR: Atlanto-occipital fusion for dislocation in children with neurological preservation: A case report. Spine 22:344-347, 1997.
31. Steel HH: Anatomical and mechanical considerations of the atlanto-axial articulation. In: Proceedings of the American orthopaedic association. J Bone Joint Surg 50A:1481, 1968.
32. Subach BR, McLaughlin MR, Albright L, Pollack IF: Current management of pediatric atlantoaxial rotatory subluxation. Spine 23:2174-2179, 1998.
33. Sullivan JA: Fractures of the spine in children. In Green NE, Swiontkowski MF (eds). Skeletal Trauma in Children. Ed 2. Philadelphia, WB Saunders 343-368, 1998.
34. Sun PP, Poffenbarger GJ, Durham S, Zimmerman RA: Spectrum of occipital injury in young children. J Neurosurg (Spine 1) 93:28-39, 2000.
35. Towbin A: Central nervous system damage in human fetus and newborn infants. Am J Dis Child 119:521-542, 1970.
36. Wholey MH, Bruwer AJ, Baker Jr HL: The lateral roentgenogram of the neck; with comments on the atlanto-odontoid-basion relationship. Radiology 71:350-356, 1958.