Keywords
spinal cord injury - computed tomography - polytrauma - diagnosis
Palavras-chave
traumatismo raquimedular - tomografia computadorizada - politraumatismo - diagnóstico
Introduction
Between 13% and 30% of the patients who are victims of severe trauma present traumatic
injuries in the spine, and about 1/3 of them will require surgical treatment of these
fractures.[1]
[2]
[3]
The description of the scenario and the mechanism of trauma is fundamental in the
initial care to raise the suspicion of lesions to the spine. Most of these lesions
originate from automobile accidents and falls from great height,[4]
[5]
[6] and they are more commonly located in the thoracolumbar and cervical regions.[5]
[6]
After the initial stabilization of the patient, as recommended by the Advanced Trauma
Life Support (ATLS) training program, pelvis, chest and cervical spine profile radiography
exams are performed.[4]
[7] If there is suspicion of spinal trauma in other segments, radiological complementation
is necessary.[4]
The difficulty in obtaining quality exams in simple radiographies of the cervicothoracic
and craniocervical junction, especially in obese patients, as well as the difficulty
to visualize disc-ligamentous and medullar lesions, resulted in the adoption of the
computed tomography (CT) as the standard examination in the evaluation of spinal cord
injuries (SCIs) in many trauma centers.[4]
[8]
[9]
The CT has a sensitivity higher than 98% for the diagnosis of spinal lesions (bone
or disc-ligamentous), which is much higher than that of conventional radiographs.
Additionally, it adds more information and details when the cuts are thinner (1–2 mm)
and the images are reconstructed tridimensionally.[4] About 40% of fractures detected by CT are not observed with simple X-rays, or appear
in an incomplete manner.[4]
[10]
[11]
On the contrary, magnetic resonance imaging (MRI) is not commonly used for the initial
evaluation, given the fact that it is a longer, more cost-intensive examination, not
available in most minor trauma centers, and with greater technical difficulties for
its performance in polytrauma patients.[12]
In this context, the community of surgeons AOSpine recently published a new classification
system for cervical, thoracolumbar and sacral fractures essentially based on the CT,
which is an exam that is fast and widely available, characteristics that are paramount
for the identification of spine instability and for therapeutic decision-making.[13]
Despite the high sensitivity of the CT to evaluate spinal lesions related to the polytrauma
patient, in some situations, especially in cases of osteoligamentous or medullar lesions,
performing the MRI is indispensable to understand the trauma and for therapeutic decision-making,
as well as to avoid catastrophic neurological damage caused by occult instabilities.[13]
[14]
[15]
The objective of the present study is to report clinical cases in which the CT was
insufficient for the diagnosis and therapeutic decision regarding patients with SCI.
Materials and Methods
We present a series of non-consecutive cases cared for by the main author (AFJ) and
obtained from the spine surgery database of the teaching hospital of Universidade
Estadual de Campinas. The database is approved by the Ethics and Research Committee
of the institution.
The inclusion criteria were: patients who were victims of polytrauma with normal spinal
CT or with apparently stable lesions, but who presented highly unstable lesions, or
with neurological risk at admission or delayed.
Cases
-
1) A male patient, 27 years old, victim of a motorcycle accident, was admitted to
the emergency room with Glasgow Coma Scale (GCS) score of 15, but with loss of sensitivity
at level T6, paraplegic, and with hypotonic sphincter (Asia Impairment Scale [AIS]
A). The total spinal CT at admission did not show evident alterations ([Fig. 1A] – sagittal tomography at admission), and a spinal MRI was performed, which evidenced
a spinal cord signal alteration in the T2 sequence from level T1 to the medullary
cone, with distraction between T3 and T4 ([Figs. 1B] and [1C]). The patient was submitted to posterior arthrodesis ([Figs. 1D] and [1E] – postoperative sagittal CT) between T23 and T45.
-
2) A female patient, 44 years old, victim of polytrauma with closed abdominal trauma,
was admitted to the emergency section with GCS 15, without motor deficits. The lumbar
spine CT at admission showed no evident alterations ([Fig. 2A]). She was submitted to exploratory laparotomy with right hemicolectomy and segmental
enterectomy that evolved with postoperative fistula, with no need for surgical retreatment,
but requiring prolonged hospitalization for three weeks with the general surgery team.
About four months after hospital discharge, in an outpatient return consultation,
the condition evolved with deformity and lumbar pain. An MRI and a new spinal CT showed
listhesis between L3 and L4 ([Figs. 2B] and [2C]). The patient was submitted to posterior fixation through the route between L2 and
L5, with total improvement of pain ([Fig. 2A]).
-
3) A male patient, 36 years old, victim of a motorcycle accident, was admitted to
the emergency department with GCS 15, tetraparesis (proximal grade III and distal
grade II strengths in the right upper limb, proximal grade II and distal I strengths
in the left upper limb, proximal grade II and distal grade I strengths in the lower
right hand, and proximal grade III and distal grade II strengths in the left lower
limb, compatible with centromedullary syndrome), hypoesthesia at level T6, and hypotonic
sphincter (AIS C). The CT at admission did not show significant alterations in the
spine ([Fig. 3A]). The patient was maintained with rigid cervical collar and en bloc mobilization
until the MRI, due to the presence of neurological deficit. After the MRI, an extensive
spinal cord injury was found associated with a narrow cervical canal ([Figs. 3B] and [3C]). Subsequently, the patient underwent cervical arthrodesis through the posterior
route between C3 and C6, with spinal-cord decompression ([Figs. 3D], [3E] and [3F]), with significantly improved symptoms with 2 months of outpatient follow-up (AIS
D).
-
4) A male patient, 25 years old, was admitted to the emergency department after being
found on a public road with a history of motorcycle crash and ejected helmet. He evolved
with cardiorespiratory arrest, requiring three cycles of resuscitation. He arrived
intubated with GCS 3, cervical collar, isomyopic pupils, evident right tibial trauma,
and hypovolemic shock. After the initial stabilization measures, he was submitted
to skull (Marshall 2) and cervical spine CTs ([Figs. 4A] and [4B]).
-
a. The patient was submitted to amputation of the right inferior limb (RIL), with
good evolution due to the extent of the traumatic injury in his tibia. Off sedation,
he did not present motor deficits, and fully regained his consciousness. In ∼ 30 days
of hospitalization, he evolved with a vicious posture in the cervical region, and
was submitted to a new CT of the cervical spine on the 34th day of hospitalization,
which evidenced severe dislocation between C1 and C2 ([Figs. 4C] and [4D]). The patient was submitted to posterior arthrodesis with complete improvement of
posture, and evolved without neurological deficits ([Figs. 4E]).
Fig. 1 (A) Computed tomography (CT) of the spine at admission, sagittal cut, without evident
alterations related to trauma. (B) T2-weighted magnetic resonance imaging (MRI) sequence with fat suppression (short
tau inversion recovery – STIR) showing small disc extrusion between T3 and T4, with
change in medullary signal and ligamentous injury between posterior elements of T3
and T4. (C) The T2 sequence shows an interspinal ligament hypersignal, suggesting rupture, besides
the spinal compressive effect extending from T1 to the medullary cone level. (D) Postoperative control CT of the arthrodesis with posterior instrumentation, fixating
4 levels (T2345). (E) Three-dimensional (3D) reconstruction showing the final aspect of the arthrodesis.
Fig. 2 (A) Sagittal CT of the thoracolumbar spine without evident alterations related to trauma
at admission. (B) CT of thoracolumbar spine performed ∼ 4 months after the initial event, when the
patient evolved with lumbar pain and deformity, showing listhesis of L3 over L4 with
distancing of spinal processes. (C) MRI in a T2-weighted sequence showing listhesis of L3 over L4, posterior ligament
injury, and cauda equina compression. (D) Postoperative control CT showing arthrodesis by the posterior pathway between L2
and L5.
Fig. 3 (A) CT at admission, without evident alterations. (B) MRI in a T2-weighted sequence with fat suppression with extensive hypersignal in
posterior elements of the cervical spine and hypersignal in the cervical medulla,
with evident compression, without listhesis. (C) MRI in a T2-weighted sequence, showing traumatic myelopathy from C4 to C7, bulging
of diffuse disc from C5 to C6, determining vertebral canal stenosis and edema area
in the cervical and dorsal muscle-adipose planes. (D and E) Postoperative sagittal section CT with laminectomy and cervical arthrodesis by posterior
route between C3 and C6. (F) CT with 3D reconstruction showing the final aspect of the arthrodesis.
Fig. 4 (A and B) CT at admission, coronal and sagittal images respectively. Good cervical spine alignment
and fracture of the C7 transverse process are noted. (C and D) CT of the 34the day of hospitalization, when the patient began to present a vicious
cervical posture. There was a dislocation between C1 and C2 and an increase in the
atlanto-dental interval in the sagittal section. (E) 3D CT for postoperative control after the patient underwent posterior cervical arthrodesis
with lateral mass screws in C1 and lamina in C2.
Discussion
In the present study, we discussed a series of cases of patients with severe polytrauma
and SCI not adequately documented in the spinal CT at hospital admission. The cases
in question reflect the importance of the need for constant neurological surveillance,
with repetition of the physical examination and the performance of new imaging exams,
such as the MRI, in cases of suspicion of instability or neurological injury not explained
by the CT findings.
In a series with 188 polytrauma patients with cervical spine injury, in 37% of the
cases the lateral incidence failed to provide the correct diagnosis, and the CT failed
in 3 cases in which there were disco-ligamentous lesions.[4]
[11]
Nuñes et al,[16] who studied a group of 88 victims of severe polytrauma, comparing the helical tomography
with simple radiographs for the diagnosis of cervical spine lesions, detected that
32 patients (36.4%) with cervical fracture were not diagnosed by X-ray, only by CT.
The authors broadly advocate the use of CT to detect lesions in victims of severe
polytrauma.[16]
Although the CT presents high sensitivity for the diagnosis of fractures, especially
when compared with radiographs, it fails in cases of disco-ligamentous lesions, a
situation in which the MRI has higher sensitivity.[15]
[16]
[17]
Pizones et al[18] prospectively evaluated 33 patients using simple radiography, and classified the
lesions according to the system proposed by the AOSpine. Subsequently, these patients
were submitted to MRIs. From 41 fractures diagnosed using CT and radiography, there
was an increment of 9 additional lesions not revealed by the first 2 exams, totaling
50 fractures. In addition, the MRI detected 18 occult (disco-ligamentous) lesions,
causing the classification of the lesions to change from A to B in 24% of the patients,
and the therapeutic approach in 16% of the cases. The authors concluded that the MRI
is fundamental for the treatment of thoracolumbar traumas.[18]
Based on this premise, Rihn et al[19] conducted a prospective study to evaluate MRI accuracy in the diagnosis of lesions
of the posterior ligament complex (PLC) in victims of thoracolumbar trauma, compared
with the intraoperative findings. The authors concluded that the MRI findings showed
a high negative predictive value, and sensitivity close to 100%; however, they had
relatively low specificity, varying from 51.5% to 80.5%. Such relatively low sensitivity
values and positive predictive value may result in greater surgical interventions
in lesions that could once be treated conservatively. The authors conclude, therefore,
that MRI findings cannot be used alone for the therapeutic decision.[19] Such evidence was corroborated by other studies.[14]
To establish a system of reproducible, simple and efficient classification, the AOSpine
proposed a new model that, in addition to the criteria of morphology and lesion of
ligament complex, also included facet lesion and the neurological status of the patient,
which are fundamental conditions for the therapeutic decision-making regarding thoracolumbar
and cervical fractures. Such conditions are evaluated by CT, because it is a fast
and easy-to-access exam. However, many times, as we show in our sequence of cases,
an MRI or a new CT is necessary to detect SCIs, disco-ligamentous lesions, and delayed
instabilities.[13]
[14]
[15]
Regarding cervical spine trauma specifically, patients with GCS 15, without drug or
narcotic effect, without neurological deficits, and who do not have pain on palpation
of the cervical spine, may have the cervical collar removed without the need for complementary
exams, a recommendation that has been proven by more than 20 clinical trials, which
was also incorporated by the ATLS.[12]
[20] However, awake patients with pain or tenderness in the neck and normal CT may have
the cervical collar removed as long as the radiograph in extension-flexion is adequate,
or the MRI of the cervical spine does not show alterations.[21]
For patients who are symptomatic or who do not have a trustworthy exam, the primary
scanning modality is the CT from the occiput up to T1, with sagittal and coronal reconstruction.
Simple radiographs do not add information, and, therefore, are unnecessary.[20] In the case of patients with normal cervical CT and coarse movements in the four
limbs, and intubated patients, the cervical collar should be maintained until it is
possible to perform a trustworthy clinical examination or an MRI.[21] However, some authors argue that the cervical collar of patients intubated but with
cervical CT without alterations can be removed, since the incidence of ligament injuries
in these cases is lower than 5%, with the incidence of significant lesions from the
clinical point of view in less than 1% of the cases.[20]
Conclusion
Based on the cases reported and the literature review, we conclude that SCI is a multifaceted
and complex disease. The serial neurological evaluation of the individual is necessary,
as it complements the imaging exams.
The CT, which is an essential exam in the first care to the polytrauma patient, is
not always sufficient for the management of SCIs at any level. The use of MRI is fundamental
in some cases, especially when there is doubt between the conservative and surgical
treatments. The combination of the clinical information, the tomographic classification
and, eventually, the MRI data, are fundamental to choose the final conduct and to
improve prognosis.
