J Neurosurg (Pediatrics 1) 103:17–23, 2005

Spinal cord injury without radiological abnormality in

preschool-aged children: correlation of magnetic resonance
imaging findings with neurological outcomes

CHENG-CHIH LIAO, M.D., TAI-NGAR LUI, M.D., LI-RONG CHEN, M.D.,

CHI-CHENG CHUANG, M.D., AND YIN-CHENG HUANG, M.D.
Department of Neurosurgery, Chang Gung University and Chang Gung Memorial Hospital; and
Department of Physical Medicine and Rehabilitation, Toyuan General Hospital, Taoyuan, Taiwan

Object. Spinal cord injury without radiological abnormality (SCIWORA) was defined in the era when magnetic
resonance (MR) images were not popularly used as diagnostic tools. Although it is generally accepted that MR
imaging can effectively illustrate the level and severity of spinal cord injury in the acute phase of trauma, only a
few reports of MR imaging studies of SCIWORA have been published. The authors retrospectively reviewed nine
preschool-aged patients with SCIWORA to study the correlation between MR imaging findings and the outcomes
of neurological deficits, with an elimination of the bias for age.
Methods. Clinical manifestations, radiological images, surgical records, and MR imaging studies were reviewed.
The pre- and postoperative neurological statuses of the patients were reappraised using American Spinal Injury
Association scores and Nurick grades. Nonparametric tests were used to analyze the correlations among the vari-
ables of patient characteristics, MR imaging appearances of the injured spinal cord, and neurological outcome.
Conclusions. Among the patients with SCIWORA younger than 8 years old, the different patterns of the injured
spinal cords could be identified using MR imaging as transection, contusive hemorrhage, traumatic edema, and con-
cussion. The MR imaging patterns of SCIWORA had significant prognostic correlations with the neurological out-
comes of these patients; that is, a normal spinal cord appearance was prognostic of a complete recovery of neuro-
logical deficits, and intramedullary lesions correlated with permanent deficits with functional disability.

KEY WORDS • spinal cord injury • spine • SCIWORA •

magnetic resonance imaging • pediatric neurosurgery

cord injuries with different vertebral pathologi-
S
PINAL
cal conditions are frequently encountered in Level 1
trauma centers.23 In young adults (16–45 years old),
SCIs are often associated with vertebral fracture. In patients
older than 45 years of age, however, they are frequently
secondary to spondylotic degeneration, which compromis-
es the spinal canal and predisposes the spinal cord to blunt
injuries during excessive spinal motion, as in hyperexten-
sion and hyperflexion. From infancy to age 16 years, SCI is
rare and it frequently occurs without vertebral trauma.9,10,18
Now familiar to modern neurosurgeons, SCIWORA was
first clearly described as a distinct disease in 1982, when
MR images were not commonly used as a diagnostic tool
for SCI.20 Despite increasing applications in the modern
era, however, MR imaging studies of SCIWORA are
few.4,5,8 Although mainly occurring in the pediatric popula-
tion, it is evident that neurological deficits and outcomes as-
Abbreviations used in this paper: ASIA = American Spinal Cord
Injury Association; CGMH = Chang Gung Memorial Hospital;
CT = computerized tomography; MR = magnetic resonance; SCI =
spinal cord injury; SCIOWRA = SCI without radiological abnor-
mality.
sociated with SCIWORA are much more disabling in chil-
dren younger than 8 years old than in older children.4–6,19 In
the cervical spine, the phenomenon is generally explained
by stating that younger children have weaker paraspinal
muscle control, more horizontal positioning of the spinal
articular facets, a less developed uncinate process, a more
wedge-shaped vertebral body, and a larger-sized head than
do older persons.2,21 In the thoracolumbar spine, the patho-
genesis of SCIWORA may be related to the mismatch in
the elasticity of the tissue of the vertebral column and spinal
cord; that is, the former can endure much longer distraction
than the latter during an acute hyperextension or hyper-
flexion posture in a traumatic episode.12 After the age of 8
years, when Taiwanese children go to elementary school,
the morphology and biomechanics of their vertebrae will
approach those of adults. Because age has such an impor-
tant impact on the severity of SCIWORA, it is reasonable
to view patients with SCIWORA younger than 8 years old
as a distinctive population.6 Therefore, we performed a ret-
rograde study to review preschool-aged children with SCI-
WORA, treated at CGMH after 1994, when our hospital
began using MR imaging for neurological diagnosis. We
emphasized the effect of MR imaging findings in the acute

J. Neurosurg: Pediatrics / Volume 103 / July, 2005 17

 C. C. Liao, et al.

stage of SCIWORA on long-term outcome prediction in scores. For all statistical tests, probability values less than
these younger children. 0.05 (two-tailed tests) were considered significant.

Clinical Material and Methods Results

Patient Population
Inpatient records at CGMH from January 1994 (when
MR images began to be used for clinical diagnosis in our
hospital) to July 2004 were collected. Charts and radio-
graphs of preschool-aged children (, 8 years old in Taiwan)
were thoroughly reviewed. For selection criteria we used
CT scans and Pang and Pollack’s criteria19 for SCIWORA:
an objective, posttraumatic myelopathy without vertebral
subluxation or fracture as seen on plain x-ray films.
Neurological Evaluation
The neurological deficits of the patients were reap-
praised using the five-tier ASIA scores at the time of acute
injury and at the censoring time.15 For the purpose of sta-
tistical analysis, the alphabetical ASIA scores were numer-
ically assigned, with ASIA Score A being equal to 1, ASIA
Score B being equal to 2, and so on. The functional status-
es of the patients were evaluated using Nurick grades
(Grade 1; normal walking with possible clinical spinal irri-
tation; Grade 2; slight difficulties in walking with normal
domestic and working life; Grade 3; functional disability
limiting normal work and domestic activities; Grade 4; sig-
nificant weakness making walking impossible without
help; and Grade 5; bedridden or wheelchair bound).17 Pa-
tients were considered to have independent functional re-
covery if they achieved Nurick Grades 1 or 2.
Statistical Analysis
Nonparametric tests and the Fisher exact test were used
to analyze the correlations between the variables of patient
characteristics, types of SCIs seen on MR images, and the
acute and final neurological outcomes in terms of ASIA
Clinical Presentations and Treatments
Fifty-eight pediatric patients (infant–18 years old) in
whom the diagnosis of spine injury or SCIs were made were
treated and followed up at CGMH from January 1994 to
July 2004. Nine preschool-aged children (15.5%) younger
than 8 years old fulfilled the criteria for SCIWORA. The
ages ranged from 2 months to 7.5 years old (mean 4.05 6
2.79 years). The follow-up period ranged from 1 to 12 years
(mean 5.14 6 3.81 years). Table 1 shows the patients’ char-
acteristics, status of the neurological deficits, functional re-
coveries, level(s) of SCI, and associated traumas. All pa-
tients in this study received external immobilization for at
least 3 months, and no patient underwent spinal surgery.
Seven patients received steroid treatment during acute post-
traumatic hospitalization (megadose solumedrol therapy in
three patients and dexamethasone infusion in four). Six pa-
tients (75%) had associated traumas in addition to the SCI.
Three patients underwent laparotomies for associated blunt
abdominal trauma, with one close thoracostomy for associ-
ated pneumothorax.
Neurological Outcomes and MR Images
All patients underwent MR imaging between 3 hours
and 14 days after the traumatic incident. A 1.5-tesla mag-
net was used. On the T2-weighted MR image, two patients
displayed an extraneuronal soft-tissue abnormality: one
displayed a hyperintense signal within the intervertebral
discs and the other a folding of the yellow ligament adja-
cent to the cord contusion. No patient had an extramedulla-
ry compressing lesion, an epidural hematoma, or a ruptured
disc. Four types of neuronal injury in the injured spinal
cords were identified using MR imaging. Two patients had

TABLE 1
Characteristics of patients reviewed in the study*
ASIA Score Nurick Grade
Type of Cord Time From
Case Age (yrs), Cause of Level of Injury on MR Associated Injury to Initial Post- Most Re- Post- Most Re-
No. Sex Injury Injury Imaging Injuries MR Imaging FU (yrs) trauma cent FU trauma cent FU

1 0.16, M MVA C7–T1 SCT chest contusion 14 days 4.5 A A 5 5

2 2.5, M MVA C7–T1 SCT liver laceration, 7 days 6.5 A A 5 5
pneumothorax
3 3.3, F MVA T-10 SCC small bowel perfora- 6 hrs 1.0 A C 5 5
tion
4 2.0, F fall from T1–2 SCC none 7 days 2.0 B C 5 5
2m
5 7.5, M MVA T2–3 SCC traumatic SAH, chest 7 hrs 3.0 C C 5 5
contusion, spleen
laceration, femor-
al fracture
6 6.8, M fall T-10 SCE none 36 hrs 12 C D 5 2
7 7.2, M fall C4–5 norm mild head injury 4 hrs 4.5 D E 3 1
8 0.5, F fall T5–6 norm none 6 days 6.3 D E 4 1
9 6.5, M MVA C3–4 norm mild head injury 7 days 10.3 D E 3 1
* FU = follow up; MVA = motor vehicle accident; norm = normal appearance of spinal cord; SAH = subarachnoid hemorrhage; SCC = spinal cord contu-
sion with major hematomyelia; SCE = spinal cord edema; SCT = spinal cord transection.

18 J. Neurosurg: Pediatrics / Volume 103 / July, 2005

Pediatric spinal cord injury without radiological abnormality
FIG. 1. Case 1. Magnetic resonance images T1- (A) and T2- (B) weighted obtained in a 2-month-old boy, demonstrat-
ing a complete transection of the cervical spinal cord at the junction of C-7 and T-1. Because the patient had a concomi-
tant chest contusion and delayed pneumonia, imaging was performed 2 weeks after acute injury. The patient had com-
plete sensorimotor dysfunction 4.5 years after the injury.
a transection of the spinal cord (Fig. 1), and both of them
had permanent, complete sensorimotor dysfunction (ASIA
Score A) below the injury level. Three patients had a cord
contusion with hematomyelia located at the level of the in-
FIG. 2. Magnetic resonance images of a major contusive SCI 6
hours posttrauma. A: A T2-weighted, whole spine image reveal-
ing a heterogeneous signal within the spinal cord at the level of T-
10, with folding of the yellow ligament posterior to the spinal ca-
nal. B: A T1-weighted image revealing a cord enlargement at the
corresponding vertebral segment. C: A T2-weighted gradient
image revealing an intramedullary hypointense signal, which rep-
resents a hematomyelia.
J. Neurosurg: Pediatrics / Volume 103 / July, 2005
jury (Fig. 2). One patient (Case 6) weighting 26 kg was wit-
nessed to be paraplegic immediately after a ground-level
slip and fall while running. The patient was conscious and
afebril after the accident. When he was brought to our
emergency department, physical examination showed both
normal blood pressure and cardiac rhythm. Neurological
examination revealed nociceptive hypesthesia, areflexia,
urine retention, and muscular weakness below the T-10
dermatome. Abdominal and spinal CT scans did not dis-
close injury to the solid organs, great vessels, or thoracic
vertebrae. A thoracolumbar MR image obtained 36 hours
after the accident demonstrated an intramedullary, hyperin-
tense T2-weighted signal in the conus medullaris without
Gd–diethylenetriamine pentaacetic acid enhancement (Fig.
3). Because of the obvious history of the traumatic incident,
the injury was thought to be a spinal cord contusion. Fur-
ther investigation of cerebrospinal fluid was not done. The
patient was treated with intravenous dexamethasone (5 mg
every 6 hours) for 1 week and was discharged to rehabili-
tation in outpatient clinics with a neurological status of
ASIA Score C and Nurick Grade 5. After rehabilitation for
10 years, he recovered his ambulatory function to an score
of D but had a Nurick grade of 4. Because of spasticity and
equina deformity of his feet, he needed a walker. He under-
went Achilles tendon lengthening 10.5 years after the in-
jury and could walk steadily without assistance 6 months
after tendon lengthening; therefore, his final functional
grade was coded as a Grade 2 and Score D. Meanwhile, a
delayed thoracolumbar scoliosis with a Cobb angle of 18˚
developed after the 3rd year posttrauma; with bracing for 3
years it was corrected to 14˚ (Fig. 4A). The patient did not
experience other neurological deficits representing new in-
sults to the brain or cervical spinal cord in the following pe-
riod. The follow-up MR images obtained 12 years post-
trauma of the whole spine revealed a normal-appearing
conus medullaris and no other spinal abnormality (Fig.
4B). The other three patients displayed either a normal ap-
pearance or an equivocal intramedullary, heterogeneous,
19

FIG. 3. Magnetic resonance images. A: A T1-weighted image revealing an intramedullary hypointense signal. B:
A T2-weighted image revealing an intramedullary hyperintense signal representing spinal cord edema. The lesion of the
cord may be secondary to spontaneous spinal cord infarction, transverse myelitis, or vascular insult in a low-energy
falling injury.
T2-weighted MR imaging signal of the spinal cord corres-
ponding to their level of posttraumatic neurological deficit.
These patients were considered to have cord concussion
(Fig. 5).26 Although the different types of cord injuries did
FIG. 4. Case 6. A: A follow-up thoracolumbar spine x-ray film
12 years after a falling injury. A thoracolumbar scoliosis with a
Cobb angle of 14˚ is shown. B: A T2-weighted MR image of the
whole spine at the 12-year follow up revealing a normal appear-
ance of the conus medullaris and no other spinal abnormality.
20
C. C. Liao, et al.
not correlate with the severities of the immediate posttrau-
matic neurological deficits represented by ASIA scores
(Kruskal–Wallis test, p = 0.068), they did correlate with the
various neurological outcomes after 1 year of follow up
(p = 0.046). Both patients with cord transection failed to re-
cover from their complete cord injuries (Score A) after 4.5
and 6.5 years of follow up, respectively. The three patients
who had major contusive spinal cord hemorrhages could
only recover to ASIA Score C and had severe functional
dysfunction below the injury levels. These five patients all
had a functional Nurick grade of 5 at the censoring time.
The other three patients, who had no abnormal signal in
the injured cord, recovered completely (ASIA Score E),
and achieved complete functional independence (Nurick
Grade 1) within 72 hours. These three patients were con-
sidered to have a spinal cord concussion. The patients
with positive intramedullary lesions identified using MR
imaging all failed to achieve independent functional re-
covery, and they had significantly worse Nurick grade
than those without abnormal intramedullary signals on
MR imaging (Fisher exact test, p = 0.048). In the present
study, neither recurrent SCIWORA nor death occurred un-
til the last follow-up investigation.
Discussion
Recognized in the pre–MR imaging era, SCIWORA
was a diagnosis of SCI mainly based on radiography, such
as plain x-ray films and CT scans. Usually vertebral frac-
ture, degenerative osteophyte, or segmental subluxation
would not be found in a spine harboring SCIWORA. Not
until the introduction of MR imaging could pathological
abnormalities in the injured spinal cord be clearly defined;
on MR imaging, they manifest as an alteration of the in-
tramedullary MR imaging signal, which can barely be
identified on a CT scan. It is well accepted that MR imag-
ing can display the specific relationship between the in-
jury level of the spinal cord and the clinical neurological
deficits.7 The procedure can also disclose associated para-
J. Neurosurg: Pediatrics / Volume 103 / July, 2005
Pediatric spinal cord injury without radiological abnormality
FIG. 5. A and B: Magnetic resonance images from the current study revealing equivocal swelling appearances of
spinal cord on both T1- and T2-weighted MR images. All three patients with such images presented with partial and tem-
porary neurological deficits that resolved in 72 hours. They were considered to have suffered a spinal cord concussion.
spinal soft-tissue injuries as well as extramedullary com-
pressing lesions such as a ruptured disc or an epidural he-
matoma.3,24,25 In SCIWORA, an MR imaging investigation
may be the only option that can define the actual patholo-
gy of the spinal cord and indicate whether surgical inter-
vention is necessary.
According to the existence and extension of an intrame-
dullary hematoma and edema, Armstrong classified the T2-
weighted MR imaging appearances of acute posttraumatic
spinal cords into three types.1 The larger area of a low T2-
weighted MR imaging signal within the cord represents the
larger intramedullary hematoma and the poorer neurologi-
cal recovery. Some MR imaging appearances of SCIWO-
RA do not follow the Armstrong classifications1 of SCI. In
early reports, a normal appearance or minimal swelling of
the spinal cord was frequently seen on MR images, espe-
cially those depicting spines of children older than 8
years.18,22 In this group, the function of the long-track axons
is temporarily disturbed by the transmission of externally
applied kinetic energy, rather than by the direct compres-
sion or distraction of the spinal cord. This level constitutes
the mildest form of SCIWORA, termed spinal cord con-
cussion, in which the neurological deficits usually recover
within 72 hours without neurological sequela.26 Matsumu-
ra, et al.,14 reported the earliest single case report to corre-
late the abnormal MR imaging appearance of SCIWORA
with neurological deficits: a 3-year-old patient with SCI-
WORA whose spinal cord had a hyperintense T2-weighted
signal in the acute traumatic phase and who consequently
suffered from permanent paraparesis. Grabb and Pang8 sub-
divided the MR imaging appearance of seven patients with
SCIWORA into four categories: major hemorrhage, with
hemorrhage in at least 50% of the transverse area of the
cord; minor hemorrhage, with evidence of hemorrhage in
J. Neurosurg: Pediatrics / Volume 103 / July, 2005
less than 50% of the transverse area of the cord; edema
only, with evidence of a hyperintense T2-weighted signal in
the cord; and a normal cord appearance. Permanent and
complete SCI will follow cord transection and major hem-
orrhage, but partial recovery of neurological function can
be expected in the types involving minor hemorrhage or
edema only. Complete recovery to Score E can be expect-
ed only in patients with a normal-appearing spinal cord on
MR imaging. Also noted is the influence of age. In Grabb
and Pang’s series,8 three patients younger than 8 years old
suffered from transection or hemorrhagic injury of their
spinal cord with severe permanent neurological deficits.
Comparatively, only one of four patients older than 8 years
of age had hemicordal edema and suffered from minor neu-
rological deficits. With regard to analyzing the MR imag-
ing data of patients younger than 8 years old separately,
however, Grabb and Pang’s patient population was too
small to address the significant correlation between abnor-
mal MR images and poor neurological deficits. Bosch, et
al.,4 recently reported results of a large series of 60 MR
imaging studies of SCIWORA with nine abnormal and 51
normal MR images. They found that abnormal MR imag-
ing studies were three times more likely to occur in chil-
dren younger than 8 years old compared with older chil-
dren. As in previous reports, abnormal intramedullary MR
images were also correlated with a failure of neurological
recovery. To eliminate the bias of age, our study focused on
children younger than 8 years of age, whose spines are
much more flexible and resistant to mechanical stress than
are their spinal cords.11 We did not observe the minor hem-
orrhagic cord injuries according to Grabb and Pang’s clas-
sification. The pattern of cord transection in the acute stage
of SCIWORA represents the complete disruption of the
neuronal anatomical continuity, and it certainly implies the
21

failure of neurological recovery. Although major hemor-
rhagic cord injuries involving more than 50% of the trans-
verse area of the cord were prognostic of permanent com-
plete deficits in previous reports,4,8 this prediction was not
universally true in our three patients. Two of them (Cases 4
and 5) retained partial sensorimotor function with Scores B
and C, respectively, after the initial trauma. Despite com-
plete paraplegia immediately after injury, the patient in
Case 3 recovered to Score C with a muscle power grade of
1 to 2 in the key muscles of the lower limbs 2 days later. All
three children recovered to ASIA Score C only, however,
remaining functionally disabled at the most recent follow
up. Mottled increased T2-weighted MR imaging signals
were identified in an inpatient (Case 6) who was injured in
a low-energy fall when running. The edematous spinal cord
could have resulted from vascular compromise during the
injury, but the presence of transverse myelitis, multiple
sclerosis, spontaneous spinal infarction, or an intramedul-
lary tumor could produce a similar MR image. After 12
years of follow up, the patient regained functional ambula-
tion without recurrent neurological deficits. The progres-
sive improvement of neurological function without a recur-
ring event13 and lack of initial enhancement of the lesion on
MR imaging11 made the diagnosis of multiple sclerosis,
transverse myelitis, or intramedullary tumor less likely. Be-
cause the patient’s abdominal CT scan did not disclose aor-
tic injury or aneurysm, spontaneous spinal cord infarction
(which shares a presentation and disease course similar to
that seen in Case 6) should be considered as a possible
cause of his sudden paraplegia.16 Our determination of spi-
nal contusion was preferred, however, because of the obvi-
ous prior traumatic incident and the low probability of risk
factors such as hypotension and artherosclerosis.
With elimination of the bias for age and anatomical in-
fluence, the MR imaging findings of SCIWORA in our stu-
dy were prognostic of neurological outcomes, just as in the
adult population. Magnetic resonance imaging may not be
able to lead to better surgical or medical treatment to alter
the outcomes of children with SCIWORA, but it does pro-
vide reliable prognostic information to predict the long-
term outcome in the early posttraumatic stage. A normal
MR image of the spinal cord in children with SCIWORA
was prognostic of complete recovery in our study (p =
0.048). Predictably, the worst recovery, with a complete
and permanent sensorimotor loss, was correlated with cord
transection. Although variable recoveries occurred with the
other transitional patterns of cord injury in which the anat-
omical continuity of the spinal cord was maintained, per-
manent neurological deficits and functional disability could
be expected. We recommend that MR imaging be per-
formed in any child who is thought to have suffered SCI, as
soon as the patient becomes clinically stable.
Conclusions
Spinal cord injury without radiological abnormality is a
distinct SCI of the pediatric population. Children younger
than 8 years of age usually have more severe neurological
deficits and outcome compared with older counterparts.
Among patients with SCIWORA younger than 8 years of
age, different cord injury patterns can be identified using
MR imaging, such as transection, contusive hemorrhage,
22
C. C. Liao, et al.
traumatic edema, and concussion. The MR imaging pat-
terns of SCIWORA have significant prognostic correla-
tions with the neurological outcomes of these patients.
Spinal cord transection is prognostic of complete and per-
manent neurological dysfunction, and spinal cord concus-
sion is prognostic of a complete neurological recovery.
Patients with the other two patterns of cord injury have
partial recoveries but remain functionally disabled. Mag-
netic resonance imaging provides a practical interpreta-
tion of the severity and a proper prediction of the progno-
sis of SCIWORA in these very young, preschool-aged
children.
References
1. Armstrong DC: Pediatric spine, in Stark DD, Bradlty WG Jr
(eds): Magnetic Resonance Imaging, ed 3. St. Louis: Mosby,
1999, Vol 3, pp 1917–1936
2. Bailey DK: The normal cervical spine in infants and children.
Radiology 59:712–719, 1952
3. Beers GJ, Raque GH, Wagner GG, Shields CB, Nichols GR II,
Johnson JR, et al: MR imaging in acute cervical spine trauma.
J Comput Assist Tomogr 12:755–761, 1988
4. Bosch PP, Vogt MT, Ward WT: Pediatric spinal cord injury
without radiographic abnomality (SCIWORA): the absence of
occult instability and lack of indication for bracing. Spine 27:
2788–2800, 2002
5. Dare AO, Dias MS, Li V: Magnetic resonance imaging correla-
tion in pediatric spinal cord injury without radiographic abnor-
mality. J Neurosurg (Spine 1) 97:33–39, 2002
6. Dickman CA, Zabramski JM, Hadley MN, Rekate HL, Sonntag
VK: Pediatric spinal cord injury without radiographic abnor-
malities: report of 26 cases and review of the literature. J Spi-
nal Disord 4:296–305, 1991
7. Flanders AE, Schaefer DM, Doan HT, Mishkin MM, Gonzalez
CF, Northrup BE: Acute cervical spine trauma: correlation of
MR imaging findings with degree of neurologic deficit. Radiol-
ogy 177:25–33, 1990
8. Grabb PA, Pang D: Magnetic resonance imaging in the evalua-
tion of spinal cord injury without radiographic abnormality in
children. Neurosurgery 35:406–414, 1994
9. Hadley MN, Zabramski JM, Browner CM, Rekate H, Sonntag
VKH: Pediatric spinal trauma. Review of 122 cases of spinal
cord and vertebral column injuries. J Neurosurg 68:18–24,
1988
10. Hamilton MG, Myles ST: Pediataric spinal injury: review of
174 hospital admissions. J Neurosurg 77:700–704, 1992
11. Lakhanpal SK, Maravilla KR: Multiple sclerosis, in Stark DD,
Bradlty WG Jr (eds): Magnetic Resonance Imaging, ed 3. St.
Louis: Mosby, 1999, Vol 3, pp 1379–1402
12. Leventhal HR: Birth injuries of the spinal cord. J Pediatr 56:
447–453, 1960
13. Lublin FD, Reingold SC: Defining the clinical course of multi-
ple sclerosis: result of an international survey. National Multi-
ple Sclerosis Society (USA) Advisory Committee on Clinical
Trials of New Agents in Multiple Sclerosis. Neurology 46:
907–911, 1996
14. Matsumura A, Meguro K, Tsurushima H, Kikuchi Y, Wada M,
Nakata Y: Magnetic resonance imaging of spinal cord injury
without radiologic abnormality. Surg Neurol 33:281–283,
1990
15. Maynard FM Jr, Bracken MB, Creasey G, Ditunno JF Jr,
Donovan WH, Ducker TB, et al: International standards for
neurological and functional classification of spinal cord injury.
American Spinal Injury Association. Spinal Cord 35:266–274,
1997
16. Monteiro L, Leite I, Pinto A, Stocker A: Spontaneous thora-
J. Neurosurg: Pediatrics / Volume 103 / July, 2005
Pediatric spinal cord injury without radiological abnormality
columbar spinal cord infarction: report of six cases. Acta Neu-
rol Scand 86:563–566, 1992
17. Nurick S: The pathogenesis of the spinal cord disorder associ-
ated with cervical spondylosis. Brain 95:87–100, 1972
18. Osenbach RK, Menezes AH: Pediatric spinal cord and vertebral
column injury. Neurosurgery 30:385–390, 1992
19. Pang D, Pollack IF: Spinal cord injury without radiographic
abnormality in children—the SCIWORA syndrome. J Trauma
29:654–664, 1989
20. Pang D, Wilberger JE Jr: Spinal cord injury without radiographic
abnormalities in children. J Neurosurg 57:114–129, 1982
21. Pennecot GF, Gouraud D, Hardy JR, Pouliquen JC: Roentgeno-
graphical study of the stability of the cervical spine in children.
J Pediatr Orthop 4:346–352, 1984
22. Pollack IF, Pang D, Sclabassi R: Recurrent spinal cord injury
without radiographic abnormalities in children. J Neurosurg
69:177–182, 1988
23. Proctor MR: Spinal cord injury. Crit Care Med 30 (Suppl 11):
S489–S499, 2002
J. Neurosurg: Pediatrics / Volume 103 / July, 2005
24. Selden NR, Quint DJ, Patel N, d’Arcy HS, Papadopoulos SM:
Emergency magnetic resonance imaging of cervical spinal cord
injuries: clinical correlation and prognosis. Neurosurgery 44:
785–793, 1999
25. Wittenberg RH, Boetel U, Beyer HK: Magnetic imaging and
computer tomography of acute spinal cord trauma. Clin Or-
thop Rel Res 260:176–185, 1990
26. Zwimpfer TJ, Bernstein M: Spinal cord concussion. J Neu-
rosurg 72:894–900, 1990
Manuscript received November 8, 2004.
Accepted in final form February 14, 2005.
Address reprint requests to: Cheng-Chih Liao, M.D., Depart-
ment of Neurosurgery, Chang Gung Memorial Hospital, 5, Fu-shing
Street, 333, Kweishan, Taoyuan, Taiwan. email: liao2901@adm.
cgmh.org.tw.
23