Imaging in Spinal Trauma

Pramit M. Phal, MBBS, FRANZCR, and James C. Anderson, MD

he standard of care in imaging of the spine in trauma patients is constantly changing with the increasing availability of new technology. Multidetector helical computed tomography (CT) allows the spine to be imaged more accurately and expeditiously than previously. MRI also has an important role in the imaging algorithm. The aim of the following article was to provide a contemporary review of imaging in spinal trauma.

Indications for Imaging

There have been multiple studies investigating the necessity of imaging in trauma of the cervical spine. The general goal of these guidelines is to accurately predict which patients are at risk of cervical spine fractures, avoiding the potentially disastrous consequences of not diagnosing a cervical spine fracture. The secondary benet of such guidelines is to reduce unnecessary examinations. The two largest studies are the NEXUS study and the Canadian C-spine study. NEXUS (National Emergency X-radiography Utilization Study group)1 performed a prospective, observational study investigating the usefulness of a clinically based decision instrument in deciding which patients needed imaging of the cervical spine posttrauma. The study was conducted in 21 centers across the USA and studied 34,069 patients who had cervical spine radiography after blunt trauma. Patients needed to meet ve criteria to be classied as having a low probability of injury, namely, no midline cervical tenderness; no focal neurologic decit; normal alertness; no intoxication; and no painful, distracting injury.1 This decision instrument identied 810 of the 818 patients who had cervical spine injury. Of the eight injuries missed, only two were deemed clinically signicant. The sensitivity of this tool was 99% and negative-predictive value was 99.8% for identifying patients with cervical spine injury. It also found that 4309 of the 34,069 examinations could have been avoided (12.6%). The Canadian C-spine rule for radiography in alert and stable trauma patients2 was derived from data from a multi-

Oregon Health and Science University, Department of Radiology, Division of Neuroradiology, Portland, Oregon. Address reprint requests to James C. Anderson, MD, Oregon Health and Science University, 3181 S.W. Sam Jackson Park Road, Mail Code CR 135, Portland, OR 97239. E-mail: andejame@ohsu.edu

center study that prospectively evaluated 20 predetermined standardized clinical ndings before radiography. Patients with unstable vital signs, patients with reduced Glascow Coma Score (GCS) (below 15), and children were excluded from the study. In total 8924 patients were studied, of which 151 (1.7%) had an important C-spine injury. The nal rule asks the three following questions: (1) Is there a high-risk factor mandating radiography (high risk being dened as age 64, dangerous mechanism of injury, or paresthesia in extremities)? (2) Is there a low-risk factor present that allows safe assessment of motion (specied as a simple rear-end motor vehicle collision, sitting position in Emergency Department (ED), ambulatory since injury, delayed onset of neck pain, or absence of midline tenderness)? (3) Is the patient able to actively rotate their neck 45 to the left and right? The rule species that there is no need for imaging when there are none of the factors in question 1 present; at least one of the factors in question 2 is present; and the patient is able to complete the task in question 3. This rule had 100% sensitivity and 42.5% specicity in identifying the 151 clinically important C-spine injuries and also identied 27 of 28 clinically unimportant cervical spine injuries. The estimated radiography rate was 58.2%, reduced from 68.9% (relative reduction of 15.5%). In 2003 a study was published comparing the clinical performance of the Canadian C-spine rule and the NEXUS lowrisk criteria.3 A prospective cohort study was conducted in nine Canadian hospitals. Both sets of criteria were tested in 7438 patients, of whom 162 had clinically signicant injuries. The Canadian C-spine rule had a sensitivity of 99.4% and specicity of 45.1% compared with a sensitivity of 90.7% and specicity of 36.8% for the NEXUS low-risk criteria. The Canadian C-spine rule would have resulted in lower radiography rates: 55.9% compared with 66.6%, and would have missed less important injuries only 1 compared with 16 with the NEXUS low-risk criteria. The ACR has issued its own appropriateness criteria for imaging4 in cervical spine trauma, drawing on NEXUS and Canadian C-spine data, on its own literature review (data from 13,534 patients), and from the considerable experience of its expert panel members. Its recommendations are that no imaging is required in alert patients who have never lost consciousness, are not under the inuence of drugs or alcohol, have no distracting injuries, have no cervical tenderness,

190

0037-198X/06/$-see front matter 2006 Elsevier Inc. All rights reserved. doi:10.1053/j.ro.2006.05.003

Imaging in spinal trauma

191 comes the preferred initial screening test for moderate- and high-risk groups. A study by Hanson11 and coworkers validated a clinical rule that identied patients at high risk for cervical spine injury, as it is these patients that benet from going straight to CT. The factors described included three mechanistic parameters: high-speed motor vehicle accident (35 mph/56 kmph), a crash with a death at the scene, and a fall from a height (10 ft/3 m). There were also three clinical parameters based on the primary patient survey: signicant closed head injury or intracranial hemorrhage on CT; neurologic symptoms or signs referable to the cervical spine; and pelvic or multiple extremity fractures. If any of these factors were present, helical CT was used as the initial screening modality. The study included 4285 patients. When the rule was applied to patients directly presenting to their trauma center, 40 of 462 high-risk patients (9%) were found to have a cervical spine fracture. In addition to reduced sensitivity of fracture detection, another potential disadvantage of plain X-rays compared with CT is the increased examination time. In multitrauma patients, timely radiographic clearance of the cervical spine is an important element of management. Daffner12,13 found that the average time for a six-view cervical spine series was 22

Figure 1 Sagittal reformatted image reveals small fracture at the anterior-inferior aspect of the C6 vertebral body. This fracture was in plane on the transverse images and more easily seen on the reformatted images.

and have no neurologic ndings. Patients not fullling these criteria should have three-view radiography followed by helical CT or proceed straight to CT if cranial CT is also to be performed.

Plain Films versus CT

Plain lms have been shown to be inferior to CT with respect to fracture detection in a number of studies,5-8 although there are no randomized controlled trials comparing the two modalities; as such, a trial would be ethically unacceptable. A recent meta-analysis by Holmes and Akkinepalli9 showed that pooled sensitivities for plain radiography was 52% and for CT was 98% for identifying patients with cervical spine injury. The question arises as to when it is appropriate to utilize CT as the rst-line investigation of cervical spine injury. In this era of economic rationalization, cost benet analysis becomes an integral part of any discussion. While the initial cost of CT may be greater than plain radiography, both Blackmore an coworkers7 and Grogan and coworkers10 make the point that when institutional costs associated with a missed cervical spine fracture are taken into account, helical CT be-

Figure 2 Sagittal reformatted image through the facets demonstrates widening of the facet joint, indicating ligamentous injury in this patient.

192

P.M. Phal and J.C. Anderson

Figure 3 Transverse image (A) and coronal reformatted image (B) of a fracture of the occipital condyle show the complementary nature of multiplanar imaging.

minutes. This compared with around 12 minutes when a cervical spine CT was added to a cranial CT. In around 79% of patients one or more of the cervical spine radiographs had to be repeated.13 The current role of cervical spine plain lms in the evaluation of cervical spine trauma is difcult to dene and somewhat dependent on the availability of CT. If the patient is at moderate to high risk of injury, it seems reasonable to proceed to CT, particularly if a cranial CT is to be performed. Plain radiography of the cervical spine is indicated if the patient is at low risk of cervical spine injury but imaging is deemed necessary according to NEXUS/Canadian C-spine rule. Lateral cervical spine X-rays also nd utilization in the ED as a quick screening test where it is often performed in conjunction with supine chest and pelvic X-rays. In this setting the lateral cervical spine X-ray may expeditiously identify abnormal levels and conditions such as atlanto-occipital dissociation that require specic management. Daffner14 makes the point that plain lms are useful in motion degraded CT studies and in identifying fractures in the axial plane (such as dens fractures), although the utilization of sagittal and coronal reformations makes the second point less relevant. The use of multiplanar reformations in the routine interpretation of CT studies of the cervical spine allows easier

detection of fractures and misalignment that were difcult to visualize on transverse plane imaging alone (Figs. 1 to 3).

Radiation Exposure
A subject that is often given little consideration in deciding how the cervical spine should be imaged is the relative radiation exposure between the different modalities. A study in the UK demonstrated that most doctors underestimate the dose of radiation from most diagnostic tests.15 Estimates from our institution are that the effective radiation dose from a CT of the cervical spine is 2.5 mSv compared with 0.04 mSv from a single radiograph of the cervical spine (standard three view would be 0.12 mSv). The BEIR-VII report16 estimates that 1 in 100 persons exposed to 100 mSv will develop a malignancy related to radiation exposure. While there are no statistically validated estimates of the risk of radiation-induced malignancy below 100 mSV, after reviewing all available data, BEIR considers the linear no threshold the most appropriate model for risk estimation. This model stipulates that the cancer increases linearly with effective dose. Thus if there are 100 cancers in 100,000 exposed to 100 mSV, CT of the cervical spine with patients exposed to 2.5 mSv will cause 2.5 malignancies per 10,000 patients. It is estimated that in

Imaging in spinal trauma

193

Figure 4 Sagittal reformatted image (A) demonstrates widening of the posterior spinous processes posteriorly (black arrow) and anterior displacement of the prevertebral fat plane anteriorly (white arrows). A T2-weighted sagittal MRI (B) on the same patient conrms the edema in the prevertebral space (white arrow) and the associated disc disruption at C4/5. Cord edema (black arrow) is also noted.

the USA 0.9% of cancers could be caused be diagnostic Xrays corresponding to 5695 cases per year.17 Rybicki and coworkers18 postulate that it may be possible to lower milliampere setting to decrease the radiation dose to the thyroid. The tradeoff is a noisier image, with a concomitant effect on fracture detection. As in all cases of radiation exposure in imaging, the justication is that the enhanced diagnostic ability and subsequent patient benets should outweigh the potential risks. The advantages of CT in imaging of cervical spine trauma are readily apparent and the immediate urgency is usually more important than a theoretical risk of radiation-induced malignancy. Clinicians should however be mindful of the risks and perhaps consider these and the alternate imaging modalities in patients who are not at high risk of cervical spine fracture.

The Role of MRI

The role of magnetic resonance imaging (MRI) in acute spinal trauma is to evaluate neurological symptoms and suspected ligamentous disruption.4 MRI can directly visualize the spinal cord, allowing assessment of spinal cord compression, contusion, and hemorrhage. In acute traumatic myelopathy19

lesions such as epidural hematoma, acute disc prolapse, and ligamentous disruption are demonstrated to advantage. These entities are important to recognize as prompt surgical correction has the potential of preserving neurologic function. The strengths of MR imaging in acute spinal trauma are well documented in a study by Holmes20 and coworkers for the NEXUS group, where MRI identied 69/69 (100%) cases of spinal cord injury and 38/38 (100%) cases of ligamentous injury. However the low fracture detection rate of 85/154 fractures (55%) suggests MRI is not an appropriate screening modality for detection of fractures. However Katzberg21 and coworkers comment that MRI and conventional radiography are complementary examinations and that the combination is competitive with CT for the diagnosis of osseous injury. In Holmes study20 MRI detected 37/43 (86%) cases of vertebral subluxation and 14/18 cases of locked facets. By comparison, CT detected 721/740 fractures (97%), 0/30 (0%) cases of spinal cord injury, 9/36 (25%) cases of ligamentous injury, 76/88 (86%) cases of vertebral subluxation, and 34/35 (97%) of locked facets. MRI has further benets in detecting noncontiguous areas of injury. In one study cervical spinal MRI revealed noncon-

194

P.M. Phal and J.C. Anderson

Figure 5 Sagittal reformatted image (A) and transverse images (B) of the thoracolumbar spine demonstrates the utility of the reformatted images in detection of in plane fractures. Fracture of the anterior-superior endplate of L2 (white arrow) and posterior elements of L1 (black arrow).

tiguous cervicothoracic junction/upper thoracic injury in 28%.22 In a study by Green and Saifuddin23 MRI of the entire spine was performed in all admissions and 77% of cases had a noncontiguous injured level. MRI offers prognostic information regarding potential recovery post spinal cord injury.24 Imaging factors associated with poor functional recovery are hemorrhage, long segments of edema, and high cervical location of injury (Fig. 4). In the past CT myelography has been in the assessment of acute spinal injury with good accuracy. Compared with MRI, CT myelography has the disadvantages of increased patient manipulation, having to perform a lumbar puncture, the use of ionizing radiation, and poor evaluation of intrinsic cord pathology. In addition, performing a lumbar puncture and removing CSF below the level of spinal cord injury may exacerbate the damage.19 The main disadvantages in performing MRI in the acute setting are that it is time consuming, it maybe difcult to monitor unstable patients while in the MR scanner, and the images are sensitive to motion artifact. There are also some patients in whom MRI is contraindicated, namely those with pacemakers, certain types of aneurysm clips, and inner ear implants and those patients with metallic ocular foreign bod-

ies. CT myelography may have a role in patients in whom MRI is contraindicated.

Imaging of the Thoracolumbar Spine

There are no validated criteria for imaging the thoracolumbar spine. AP and lateral radiographs have been the traditional method of screening for injury in the thoracolumbar spine in trauma patients. As with imaging of the cervical spine, CT is being used with increasing frequency and there is increasing evidence that it is more accurate than plain radiography. Trauma patients often undergo CT scanning of the chest, abdomen, and pelvis and the data obtained from this scan can be used to assess the thoracolumbar spine. A study by Wintermark and coworkers25 addressed the issue of whether images from reconstructed data from CT examinations of the chest, abdomen, and pelvis are adequate screening for injury of the thoracic and lumbar spine. In this study 100 patients with severe trauma were reviewed. Plain radiographs were compared with 2.5-mm-thick multidetector CT images reconstructed at an interval of 2 mm. The images were assessed for the presence, location, and stability of fractures as well as

Imaging in spinal trauma

image quality. Mean sensitivity and interobserver agreement for detection of unstable fractures 97.2% and 0.95 for CT and 33.3% and 0.368 for plain radiographs. Overall detection was 78.1% with CT and 32% with plain X-rays. A similar study was performed by Roos and coworkers26 who found that 4 2.5 mm multidetector CT images reconstructed from CT of the chest and abdomen allowed accurate fracture detection and classication with sensitivities of 98 and 97% and specicities of 97 and 97% in their two readers. All major fractures were identied, with only a single anterior wedge fracture missed by both readers. The readers also rated the image quality as excellent in approximately two-thirds of the axial and multiplanar reformatted images (Fig. 5). Brown and coworkers,27 Hauser and coworkers,28 and Sheridan and coworkers29 have performed similar analyses comparing plain radiography and CT in trauma of the thoracolumbar spine. Combined sensitivities for detection of injury are 67% for plain lm compared with 98% for spiral CT. Numerous authors have also commented that utilizing CT is more time efcient compared with plain lms.27 Interestingly, in Wintermarks25 analysis the median CT time was 7 minutes longer than plain lms (40 compared with 33 minutes). There is however less patient manipulation with CT. As in the cervical spine there is more exposure to ionizing radiation with CT compared with plain lm. Wintermark25 estimates the average effective dose with CT to be 19.42 mSv compared with 6.36 mSv with plain lms. However if the patient is to undergo CT of the chest, abdomen, and pelvis anyway, it may actually be a more efcient use of ionizing radiation. MRI of the thoracic and lumbar spine offers the same advantages as it does in the cervical spine. It is good for detection of non-displaced vertebral body fractures. Scanning in the sagittal plane allows assessment of injury at noncontiguous levels. MRI also allows assessment of ligaments, disc, and soft tissues as well as the spinal cord, conus, and cauda equina. Also, there is no exposure to ionizing radiation.

195
evaluation of the cervical spine: an evidence-based approach. Skeletal Radiol 29:632-639, 2000 Woodring JH, Lee C: Limitations of cervical radiography in the evaluation of acute cervical trauma. J Trauma 34:32-39, 1993 Holmes JF, Akkinepalli R: Computed tomography versus plain radiography to screen for cervical spine injury: a meta-analysis. J Trauma 58:902-905, 2005 Grogan EL, Morris JA Jr, Dittus RS, et al: Cervical spine evaluation in urban trauma centers: lowering institutional costs and complications through helical CT scan. J Am Coll Surg 200:160-165, 2005 Hanson JA, Blackmore CC, Mann FA, et al: Cervical spine injury: a clinical decision rule to identify high-risk patients for helical CT screening. AJR Am J Roentgenol 174:713-717, 2000 Daffner RH: Identifying patients at low risk for cervical spine injury: the Canadian C-spine rule for radiography. JAMA 286:1893-1894, 2001 Daffner RH: Cervical radiography for trauma patients: a time-effective technique? AJR Am J Roentgenol 175:1309-1311, 2000 Daffner RH: Controversies in cervical spine imaging in trauma patients. Emerg Radiol 11:2-8, 2004 Shiralkar S, Rennie A, Snow M, et al: Doctors knowledge of radiation exposure: questionnaire study. BMJ 327:371-372, 2003 Monson R, Cleaver JE, Abrams HL, et al: Biological Effects of Ionizing Radiation (BEIR) VII: health risks from exposure to low levels of ionizing radiation, in Health Risks from Exposure to Low Levels of Ionizing Radiation: BEIR VII Phase 2. Washington, DC, The National Academies Press, 2006 Berrington de Gonzalez A, Darby S: Risk of cancer from diagnostic X-rays: estimates for the UK and 14 other countries. Lancet 363:345351, 2004 Rybicki F, Nawfel RD, Judy PF, et al: Skin and thyroid dosimetry in cervical spine screening: two methods for evaluation and a comparison between a helical CT and radiographic trauma series. AJR Am J Roentgenol 179:933-937, 2002 Quint DJ: Indications for emergent MRI of the central nervous system. JAMA 283:853-855, 2000 Holmes JF, Mirvis SE, Panacek EA, et al: Variability in computed tomography and magnetic resonance imaging in patients with cervical spine injuries. J Trauma 53:524-529, discussion 530, 2002 Katzberg RW, Benedetti PF, Drake CM, et al: Acute cervical spine injuries: prospective MR imaging assessment at a level 1 trauma center. Radiology 213:203-212, 1999. Choi SJ, Shin MJ, Kim SM, et al: Non-contiguous spinal injury in cervical spinal trauma: evaluation with cervical spine MRI. Korean J Radiol 5:219-224, 2004 Green RA, Saifuddin A: Whole spine MRI in the assessment of acute vertebral body trauma. Skeletal Radiol 33:129-135, 2004 Flanders AE, Spettell CM, Friedman DP, et al: The relationship between the functional abilities of patients with cervical spinal cord injury and the severity of damage revealed by MR imaging. AJNR Am J Neuroradiol 20:926-934, 1999 Wintermark M, Mouhsine E, Theumann N, et al: Thoracolumbar spine fractures in patients who have sustained severe trauma: depiction with multi-detector row CT. Radiology 227:681-689, 2003 Roos JE, Hilker P, Platz A, et al: MDCT in emergency radiology: is a standardized chest or abdominal protocol sufcient for evaluation of thoracic and lumbar spine trauma? AJR Am J Roentgenol 183:959-968, 2004 Brown CV, Antevil JL, Sise MJ, et al: Spiral computed tomography for the diagnosis of cervical, thoracic, and lumbar spine fractures: its time has come. J Trauma 58:890-895, discussion 895-896, 2005 Hauser CJ, Visvikis G, Hinrichs C, et al: Prospective validation of computed tomographic screening of the thoracolumbar spine in trauma. J Trauma 55:228-234, discussion 234-235, 2003 Sheridan R, Peralta R, Rhea J, et al: Reformatted visceral protocol helical computed tomographic scanning allows conventional radiographs of the thoracic and lumbar spine to be eliminated in the evaluation of blunt trauma patients. J Trauma 55:665-669, 2003

8. 9.

10.

11.

12. 13. 14. 15. 16.

17.

18.

19. 20.

21.

22.

23.

References
1. Hoffman JR, Mower WR, Wolfson AB, et al: Validity of a set of clinical criteria to rule out injury to the cervical spine in patients with blunt trauma. National Emergency-Radiography Utilization Study Group. N Engl J Med 343:94-99, 2000 2. Stiell IG, Wells GA, Vandemheen KL, et al: The Canadian C-spine rule for radiography in alert and stable trauma patients. JAMA 286:18411848, 2001 3. Stiell IG, Clement CM, McKnight RD, et al: The Canadian C-spine rule versus the NEXUS low-risk criteria in patients with trauma. N Engl J Med 349:2510-2518, 2003 4. American College of Radiology Expert Panel on Musculoskeletal Imaging. Appropriateness criteria for suspected cervical spine trauma. ACR appropriateness criteria. Available from: http://www.guideline.gov/ summary/summary.aspx?doc_id8587. Reston, VA, 2002 5. Griffen MM, Frykberg ER, Kerwin AJ, et al: Radiographic clearance of blunt cervical spine injury: plain radiograph or computed tomography scan? J Trauma 55:222-226, discussion 226-227, 2003 6. Nunez DB Jr, Zuluaga A, Fuentes-Bernardo DA, et al: Cervical spine trauma: how much more do we learn by routinely using helical CT? Radiographics 16:1307-1318, discussion 1318-1321, 1996 7. Blackmore CC, Mann FA, Wilson AJ: Helical CT in the primary trauma 24.

25.

26.

27.

28.

29.