Imaging of Spinal Cord Injury: Acute Cervical Spinal Cord Injury, Cervical Spondylotic Myelopathy, and Cord Herniation Kiran Talekar, MD,* Michael Poplawski, MD,† Rahul Hegde, MD,* Mougnyan Cox, MD,† and Adam Flanders, MD* We review the pathophysiology and imaging ﬁndings of acute traumatic spinal cord injury (SCI ), cervical spondylotic myelopathy, and brieﬂy review the much less common cord herniation as a unique cause of myelopathy. Acute traumatic SCI is devastating to the patient and the costs to society are staggering. There are currently no “cures” for SCI and the only accepted pharmacologic treatment regimen for traumatic SCI is currently being questioned. Evaluation and prognostication of SCI is a demanding area with signiﬁcant deﬁciencies, including lack of biomarkers. Accurate classiﬁcation of SCI is heavily dependent on a good clinical examination, the results of which can vary substantially based upon the patient's condition or comorbidities and the skills of the examiner. Moreover, the full extent of a patients' neurologic injury may not become apparent for days after injury; by then, therapeutic response may be limited. Although magnetic resonance imaging (MRI) is the best imaging modality for the evaluation of spinal cord parenchyma, conventional MR techniques do not appear to differentiate edema from axonal injury. Recently, it is proposed that in addition to characterizing the anatomic extent of injury, metrics derived from conventional MRI and diffusion tensor imaging, in conjunction with the neurological examination, can serve as a reliable objective biomarker for determination of the extent of neurologic injury and early identiﬁcation of patients who would beneﬁt from treatment. Cervical spondylosis is a common disorder affecting predominantly the elderly with a potential to narrow the spinal canal and thereby impinge or compress upon the neural elements leading to cervical spondylotic myelopathy and radiculopathy. It is the commonest nontraumatic cause of spinal cord disorder in adults. Imaging plays an important role in grading the severity of spondylosis and detecting cord abnormalities suggesting myelopathy. Semin Ultrasound CT MRI 37:431-447 C 2016 Elsevier Inc. All rights reserved.
Acute Cervical Spinal Cord Injury
pinal cord injury (SCI) is a devastating, life-altering event. Approximately 12,000 new injuries occur annually in the United States,1 and currently there are approximately 227,080-300,938 individuals living in the United States with the sequelae of SCI including permanent paralysis. Not
*Section of Neuroradiology, Sidney Kimmel Medical College at Thomas Jefferson University, Philadelphia, PA. †Department of Radiology, Sidney Kimmel Medical College at Thomas Jefferson University, Philadelphia, PA. Address reprint requests to Kiran Talekar, MD, Section of Neuroradiology, Sidney Kimmel Medical College at Thomas Jefferson University, Philadelphia, PA 19063. E-mail: [email protected]
http://dx.doi.org/10.1053/j.sult.2016.05.007 0887-2171/& 2016 Elsevier Inc. All rights reserved.
surprisingly, the costs to society of SCI are staggering and in 1998 were estimated at $9.7 billion per year.2 The lifetime direct costs of a high tetraplegic injured at age of 25 years can exceed $3 million.1 Males are disproportionately affected with a 4:1 male-to-female ratio, and most of injuries occur between the age of 16 and 30 years. Mirroring the increasing age of the U.S. general population, the average age at injury has increased from 28.7 years of age in the mid-1970s to 39.5 years since 2005. There are currently no “cures” for SCI and the only accepted pharmacologic treatment regimen for traumatic SCI is high dose methylprednisolone (MP), which has been reported to show efﬁcacy in Phase II randomized trials.3 Subsequently, MP administration for acute SCI has become widespread in the United States. Recently the efﬁcacy of this treatment has been 431
K. Talekar et al.
432 questioned and currently is the subject of ongoing debate.4,5 Much of the debate has centered on whether the magnitude of reported improvement with MP is clinically important. The controversy regarding the utilization of MP highlights the critical need for new treatment strategies. To date, the treatment of acute SCI has been characterized, unfortunately, by the paucity of clinical trials. Although the efﬁcacy and safety of MP remains controversial, there is general agreement that any pharmacologic measure should be employed in the ﬁrst few hours after injury. Patient selection for a speciﬁc therapy can be problematic in the acute period because the classiﬁcation system used to grade neurologic impairment is completely dependent upon the accuracy of the neurologic examination. The neurologic examination is accurate and reproducible in ideal conditions. However, the results can vary substantially based upon the level of cooperation, communication, and consciousness of the patient, associated patient comorbidities and the skills of the examiner. Moreover, the full extent of a patient's neurologic injury may not become apparent for days after injury. By then, late implementation of a drug based upon a delayed neurologic assessment is less likely to demonstrate a therapeutic response. In that respect, it is proposed that in addition to characterizing the anatomic extent of injury, metrics derived from conventional magnetic resonance imaging (MRI), and diffusion tensor imaging (DTI), in conjunction with the neurological exam, can serve as a reliable objective biomarker for determination of the extent of neurologic injury and early identiﬁcation of patients who would beneﬁt from treatment.
Pathophysiology of SCI Similar to acute traumatic brain injury, acute SCI can be divided into primary and secondary injury models. Compared with the brain, the mechanism of acute SCI is less well understood, and most of the research data currently available is derived from animal trials.6 Although transection injuries of the spinal cord do occur in acute trauma, most of acute SCI in humans is caused by blunt trauma, usually in the setting of motor vehicle accidents.7,8 The primary insult to the cord is initiated by transient or ﬁxed loss of integrity of the surrounding bony and ligamentous structures with resultant blunt impact on the spinal cord. In general, the amount of force transmitted to the cord determines the severity of the underlying cord injury.9 The injury may range from transient neurologic deﬁcits because of abnormal axonal ﬁring to dense neurologic deﬁcits due to axonal disruption.7 Aside from preventive measures like lowering the speed limit and enforcing drunk-driving laws, primary SCI is immutable and current interventions are aimed at mitigating secondary spinal injury. Secondary SCI is characterized by subsequent cellular dysfunction, necrosis, and death of initially intact neurons adjacent to the site of primary affect. Several processes are triggered by the injured or dying neurons at the site of primary impact that spread to nearby normal axons and result in propagation of the initial injury. Immediately following acute injury, spinal cord edema occurs, resulting in decreased
perfusion pressure and ischemia related to microvascular perfusion abnormalities.10 The injured axons also release glutamate in large amounts, a potent excitatory neurotransmitter.11,12 Exposure of uninjured neurons to excessive amounts of glutamate is toxic, resulting in inﬂux of calcium and sodium into the cells and initiating a number of deleterious processes including cell death in some cases.11,12 Vulnerable intact neurons close to the site of initial SCI may undergo cell death in the ensuing hours or days via necrosis or apoptosis.13 Apoptosis is controlled cell death that results in minimal inﬂammation, whereas necrosis is a disorderly process of cell death that causes signiﬁcant inﬂammation.10 Severe cord injury tends to result in more extensive necrosis.7 Another important mechanism of secondary SCI is generation and propagation of free radicals. Free radicals are highly reactive molecules that interact with lipids and proteins in cell membranes to cause cellular dysfunction. Free radicals are especially abundant during the reperfusion phase of SCI, and several therapies are speciﬁcally targeted at halting the production of free radicals.14 The processes in secondary spinal injury are complex and interdependent. Optimal management of acute SCI in the future will likely be directed at inhibiting multiple sites in the secondary spinal injury cascade in hopes of achieving a synergistic therapeutic effect.
Clinical Assessment of Acute SCI Initial clinical presentation of patients with acute cervical SCI is the main factor determining triage, deﬁning therapeutic options, and predicting prognosis. As such, the initial neurologic assessment should be accurate, consistent, and reproducible in deﬁning the neurological deﬁcits. In addition, an ideal neurologic assessment scale should have prognostic value in determining patient's potential for recovery. Numerous assessment scales have been employed to evaluate SCI patients, and can be divided into 2 general types. The ﬁrst type focuses on the neurological deﬁcits resulting from SCI and is examination speciﬁc. The International Standards for the Neurological Classiﬁcation of SCI (ISNCSCI) is the most widely used and validated system, having undergone multiple revisions, most recently in 2011.15 The second type of scale focuses on SCI patient's functional skills, including ability to care for oneself, perform personal hygiene, ambulate or transfer. These scales aim to determine patient's ability or inability to functionor live independently. In general, the ﬁrst type of scale is used to acutely assess patients with SCI, while both scales are used to deﬁne the chronically injured patient. Scales for functional outcomes include the Barthel Index, the modiﬁed Barthel Index, the Functional Independence Measure (FIM), the Quadriplegic Index of Function, the Spinal Cord Independence Measure (SCIM), the Walking Index for SCI, the SCI Functional Ambulation Inventory, and more recently proposed SCI Computer Adaptive Test.16 The central aspect of the ISNCSCI is classiﬁcation of SCI patients into American Spinal Injury Association (ASIA) Impairment Scale—the AIS, which is a 5-point ordinal scale that classiﬁes individual's injury from A through E (Table). The neurologic level of injury (NLI) refers to the most caudal
Imaging of spinal cord injury
Table AIS Scale A B
Complete No sensory or motor function is preserved in the sacral segments S4-5 Sensory Incomplete Sensory but not motor function is preserved below the neurological level and includes the sacral segments S4-5 (light touch [LT] or pin prick [PP] at S4-5 or deep anal pressure [DAP]) AND no motor function is preserved more than three levels below the motor level on either side of the body Motor Incomplete. Motor function is preserved at the most caudal sacral segments for voluntary anal contraction (VAC) OR the patient meets the criteria for sensory incomplete status (sensory function preserved at the most caudal sacral segments [S4-S5] by LT, PP, or DAP), and has some sparing of motor function more than three levels below the ipsilateral motor level on either side of the body (This includes key or nonkey muscle functions to determine motor incomplete status.) For AIS C—less than half of key muscle functions below the single NLI have a muscle grade Z 3 Motor Incomplete. Motor incomplete status as deﬁned above, with at least half (half or more) of key muscle functions below the single NLI having a muscle grade Z3 Normal If sensation and motor function as tested with the ISNCSCI are graded as normal in all segments, and the patient had prior deﬁcits, then the AIS grade is E. Someone without an initial spinal cord injury does not receive an AIS grade To document the sensory, motor and NLI levels, the ASIA Impairment Scale grade, and the zone of partial preservation (ZPP) when they are unable to be determined based on the examination results
ND, not determined; NLI, Neurologic level of injury; ISNCSCI, International Standards for Neurological Classiﬁcation of Spinal Cord Injury.
segment of the cord with intact sensation and antigravity muscle function strength, provided that there is normal (intact) sensory and motor function rostrally. Figures 1-4 show AIS (“A” in AIS stands for ASIA) grades A through D, respectively. A required portion of the ISNCSI motor examination is testing of key muscle functions corresponding to 10 paired myotomes (C5-T1 and L2-S1) with each muscle group receiving a score
of 0-5. These scores are then summed across myotomes and sides of body to generate a single motor score each for the upper and for the lower limbs (maximum score of 50 for each the upper and lower extremities), and provides means of numerically documenting changes in motor function. The ASIA motor scores are routinely used and reported in the literature to describe functional impairment following SCI and
Figure 1 Spinal cord injury in a 43-year-old female hit by a train. At presentation patient was categorized as AIS grade A. (A) Sagittal CT image demonstrates traumatic anterolisthesis of C6 on C7, with a ﬂexion teardrop fracture at C7, and spinous process fracture of C6. Also present were bilateral jumped facets at C6 or C7 with bilateral comminuted fractures of superior articular C7 facets (not shown). (B) T1-weighted MRI, (C) T2-weighted, and (D) STIR imaging additionally demonstrates signiﬁcant cord compression at C6 or C7 with cord edema extending from C5 to superior aspect of T1. Heterogeneity of cord T2 signal abnormality with foci of low signal suggest hemorrhagic contusion of the cord (arrow). Note: epidural hemorrhage posterior to C6 vertebral body (arrow head). There is diffuse posterior soft tissue edema extending inferiorly from C1.
K. Talekar et al.
Figure 2 Spinal cord injury in a 58-year-old male following a motor vehicle collision. At presentation patient was categorized as AIS grade B. (A) Sagittal CT demonstrates a perched right facet at C5-C6 (arrow). No fracture was seen. (B) T1-weighted MRI, (C) T2-weighted, and (D) short tau inversion recovery (STIR) imaging further demonstrates traumatic anterolisthesis of C5 on C6 with disruption of the disc (arrow head). There is disruption of interspinous ligaments and anterior longitudinal ligament (arrows). Cord edema extends through C5 and C6 level, without evidence of cord hemorrhage.
as predictors of outcome.16 The ASIA standards cannot be accurately employed in SCI patients who cannot be accurately examined because of confounding factors,17 and are not applicable to adolescents and children.18 Despite these limitations, the ASIA standards provide the most validated, consistent, and reliable system for neurological assessment, and is recommended (Level II evidence) as the preferred tool in the care of acute SCI patients.16
Although the AIS does not assess functional ability, several functional recovery measures have been developed as outcome measures for persons living with SCI. Anderson et al19 reported the consensus analysis of a multinational work group in 2008. Overall, they recommend SCIM III as the one best measure of global disability and functional recovery speciﬁc for SCI. The SCIM aims to quantify the ability to perform everyday tasks and captures the economic burden of disability, as well as the
Figure 3 Spinal cord injury in a 47-year-old male who was toppled by a wave. At presentation patient was categorized as AIS grade C. (A) Initial outside hospital T1-weighted MRI, (B) T2-weighted, and (C) STIR imaging demonstrates prevertebral edema from C1 to C4-C5 and mild widening of anterior C3-C4 disc space with focal disc T2 signal abnormality (arrow head). There is spinal cord edema spanning the C3 level. There is interspinous edema extending from C2 to C5. Patient was transferred to our facility because of worsening neurological function and a 32-hour follow-up (D) T2-weighted, and (E) short tau inversion recovery (STIR) MRI demonstrated progression of cord edema with a new focus of T2 hypointensity (arrow) within the cord at C3-C4 indicating progression to hemorrhagic contusion.
Imaging of spinal cord injury
Figure 4 Spinal cord injury in a 25-year-old following a hyperextension injury while playing rugby. At presentation patient was categorized as AIS grade D. (A) T1-weighted MRI and (B) T2-weighted imaging demonstrates acute transmembraneous disc herniation with discontinuity of the posterior longitudinal ligament at C3-C4 (arrow) causing signiﬁcant spinal cord compression, and mild spinal cord edema. Additional disc bulge is seen at C4-C5. No fracture was identiﬁed on computed tomography.
effect of their disability on the patient's overall medical condition and comfort. SCIM consists of 3 subscales that cover the related but distinct subsets of self-care (6 items; score range: 0-20), respiration and sphincter management (4 items; score range: 0-40), and mobility (9 items; score range: 0-40). The total score ranges from 0-100. The mobility subset is further subdivided into 2 subscales: room and toilet, and indoors and outdoors. Individual item scores range from 2-9 points. SCIM scores a task higher in patients who accomplish it with less assistance, aids, or medical compromise than other patients.20 With level I evidence, the SCIM III is recommended as the preferred functional outcome assessment tool for care and follow-up of SCI patients.16
Neurologic Recovery in SCI In general, spontaneous neurologic recovery from SCI is inversely related to the severity of the initial neurologic injury. However, there is reported variability in recovery within similar AIS grades. Most of neurological recovery in SCI patients occurs during the ﬁrst 6-9 months.21-23 Afterward, the rate of improvement rapidly drops off with a plateau being reached 12-18 months postinjury with little additional improvement after that time. Early improvement in neurological status is associated with greater recovery than slow improvement.24 Late recovery following complete SCI, deﬁned as motor recovery greater
than one year after injury, is rare but can occur. Recovery of motor function distal to the zone of injury in patients with complete SCI is relatively rare, and when it does occur it tends to be minimal and nonfunctional. The national SCI database indicates that about 15% of all AIS grade A patients admitted within 1 week of injury convert to incomplete status by 1 year. However, only 2.3% of initially complete patients regain signiﬁcant motor function below the injury level, that is, to AIS grade D.25 Other studies have reported complete to incomplete conversion rates ranging from 4%-34%.21,22,26-28 Recovery in motor complete, sensory incomplete injuries (AIS grade B) is mixed, with about 50% attaining ambulatory status.29,30 By deﬁnition, individuals with an AIS grade B injury have some initial preservation of distal sensation, including the S4-5 dermatomes, but no accompanying motor function. It is known that this type of sensory sparing inﬂuences prognosis—those with sacral or lower extremity pin prick sensation—have a better chance of walking than those with only light touch sensation.30-34 While recovery in motor complete injuries is generally poor, recovery from incomplete injuries (AIS grades C and D) is generally good, although this is inﬂuenced by the degree of motor deﬁcit and the age of the patient. For AIS grade C patients younger than 50 years of age at the time of injury, the chance of walking exceeds 90%, while those over age 50 have only a 42% chance of walking.35 Up to 95% of individuals with AIS grade D injuries will recover the ability to ambulate.29
436 In contrast to recovery below the zone of injury, most people with complete tetraplegia have some local recovery immediately cephalad to the zone of injury (ie, the zone of partial preservation). Most of individuals with complete tetraplegia gain a motor level, although there are differences dependent on the initial level. If the initial motor level is C4, 70% will gain C5 motor function; the corresponding rates for C5 to C6 and C6 to C7 are 75% and 85%, respectively.23 Recovery more than 2 levels below the most caudal level with motor function is rare, being seen in only 1% of cases.22 One factor that accounts for this wide range in recovery is the difﬁculty performing an accurate neurological assessment during the early phase of injury. Inﬂuences that can affect reliability of this clinical examination include those affecting cognition (traumatic brain injury, drug effects, and psychological disorders) as well as communication (ventilator dependency and language barrier).17 There is a potential therefore to initially misclassify patients, contributing to ambiguity in our understanding of the relationship of the initial injury to spontaneous recovery. As new novel therapies for SCI become routinely available, it is imperative that we improve our precision in classifying patients as close as possible to the time of injury to maximize beneﬁts of these therapies. Other objective methodologies are needed for identifying appropriate patient populations for treatment delivery; optimal therapy may need to be delivered in the ﬁrst few hours or days, yet neurologic examination may require a week's delay before providing an accurate prognosis.17 The variability in recovery for patients with incomplete SCI has hindered the study of promising treatments in the acute injury period for such patients. As a panel sponsored by the International Campaign for Cures of SCI Paralysis (ICCP) reported “trials involving motor incomplete SCI patients, or trials where an accurate assessment of AIS grade cannot be made before the start of the trial, will require large subject numbers or better objective assessment methods.”36 The substantial variability in recovery within similar injury classiﬁcation suggests that the arbitrary delineation among injury classiﬁcations (AIS: A-E) may not be representative of all of the potential functional categories. Potentially functional subclasses may exist that need to be deﬁned through means other than the neurologic examination, such as MRI. With this additional information, the ability to perform rigorous patient selection before initiation of therapy would be greatly enhanced and measuring efﬁcacy of these therapies would be much more feasible.
Conventional MRI of Acute Cervical SCI Currently, MRI provides the only means to directly inspect the damaged spinal cord, therefore it has the potential to complement the assessment provided by the subjective neurologic examination in gauging the degree of injury in SCI. Moreover, MRI evaluation is not operator dependent and the assessment of the MRI features is reproducible among observers. MRI provides excellent deﬁnition of intramedullary hemorrhage and edema in animal models. On MRI, cord edema is indicated by intramedullary hyperintensity on T2-weighted sagittal
K. Talekar et al. images, and acute cord hemorrhage (deoxyhemoglobin) is indicated by focal decreased signal on T2-weighted and gradient echo images. The combination of MRI lesion length, cord caliber, and degree of preservation of white matter in MRI cross-section has a signiﬁcant relationship to functional status in animals and the pathologic ﬁndings at autopsy. The MRI appearance of experimentally induced SCI has been used to explain the variability in functional deﬁcit among animals subjected to identical injuries. A signiﬁcant shortcoming of MRI is its limited capability in demonstrating preserved white matter tracts at the level of injury; this observation becomes signiﬁcant in estimating preserved functional capacity. With the advent of diffusion techniques and tractography algorithms based upon diffusion parameters, MRI now has the capacity to assess the integrity of spinal white matter. The depiction of parenchymal SCI on MRI not only correlates well with the degree of neurologic deﬁcit, but it also bears signiﬁcant implications regarding prognosis and potential for neurologic recovery.37-43 Many clinical investigations have reported that the MRI patterns of SCI correlate with the neurologic deﬁcit at presentation. Kulkarni et al initially proposed 3 MRI injury patterns for SCI and correlated these with the 5-part AIS and total motor scores. Intramedullary hemorrhage (Type I pattern of injury) equated with a severe neurologic deﬁcit and a poor prognosis. Cord edema alone (Type II pattern of injury) was found in patients with mild to moderate initial neurologic deﬁcits who subsequently showed neurologic improvement.42 Schaefer et al40 reﬁned the MRI patterns of SCI by including the size of the injured segment. Cord edema that extended for more than the span of one vertebral segment was associated with a more severe initial deﬁcit than smaller areas of edema. Cord hemorrhage was associated with the most severe neurologic abnormalities.40 Flanders et al38 demonstrated that spinal cord hemorrhage in the cervical region was a strong predictive ﬁnding for a complete neurologic injury. The location of the hemorrhage corresponded anatomically to the level of neurologic injury. Although the location of spinal cord edema related imprecisely to the neurologic level, the proportion of spinal cord affected by edema was directly related to the severity of initial neurologic injury. Schaefer et al41 correlated the MRI appearance of the spinal cord on admission to the change in total motor index score (MIS) in 57 patients. Patients with hemorrhagic spinal cord lesions showed no statistical improvement in MIS at follow-up. The group of patients with small areas of edema (less than 1 vertebral segment in length) demonstrated the largest improvement in MIS (72% recovery), whereas larger areas of edema showed intermediate recovery of MIS (42%).41 Flanders et al37 assessed the prognostic capabilities of MRI in forecasting motor recovery in 104 cervical SCI patients. Individual manual muscle test scores were compiled for the upper and lower extremities both at the time of admission and 12 months after injury. A motor recovery rate for the upper and lower extremities was also determined. The injured spinal cord segment on MRI was measured using a unique method that quantiﬁed spinal cord hemorrhage and edema by length and location relative to known anatomic landmarks. Lesion
Imaging of spinal cord injury length was directly proportional to neurologic impairment at the time of injury (P o 0.001). In addition, spinal cord hemorrhage was associated with the most severe injuries (P o 0.001). Although improvement in motor function after 1 year was observed in all patients, subjects with spinal cord hemorrhage on MRI had lower initial motor scores and had less improvement than those without hemorrhage. Nonhemorrhagic MRI lesions were associated with signiﬁcantly higher motor recovery rates in the lower and upper extremities and had a higher proportion of useful muscle function. Multiple regression analysis was used to determine the contribution of MRI in predicting the outcomes parameters of motor function independent of the initial clinical evaluation. Initial motor scores, the presence of hemorrhage and the length of edema were independent predictors of ﬁnal motor score and the proportion of muscles with useful function at one year. The addition of the MRI parameters to the initial clinical information improved the statistical power of the SCI model by 16% for the upper extremities and 34% for the lower extremities.37 Flanders et al44 also compared the MRI parameters of edema and hemorrhage to a standardized measurement of disability (FIM). A total of 4 distinct motor scales from the FIM assessment were determined at the time of admission to rehabilitation and subsequently at discharge from rehabilitation. The individual motor scales included tasks related to self-care, sphincter control, mobility, and locomotion. Patients without spinal cord hemorrhage on MRI had signiﬁcant improvement in self-care and mobility scores compared to patients with hemorrhage. The upper limit of the lesion (edema) correlated with admission and discharge self-care, admission mobility, and locomotion scores. Edema length correlated negatively with all FIM scores at admission and discharge. Moreover, at the time of admission to rehabilitation, all patients were completely dependent on equipment or caregivers to perform the FIM tasks. At the time of discharge, only patients with nonhemorrhagic MRI lesions improved to a modiﬁed dependence category.44 Boldin et al45 did a prospective analysis of 29 SCI patients by comparing an absolute measurement of the size of the injured segment on a postoperative MRI to the initial clinical examination and changes in long-term neurologic status. The authors also found that the presence of intramedullary hemorrhage had a higher association with a complete neurologic deﬁcit and patients with hemorrhages that measured greater than 4 mm in cranial-caudal length showed no clinical improvement at follow-up. Both the length of edema and hemorrhage were shown to be predictive variables for complete injuries. Patients with hemorrhages measuring o4 mm had incomplete injuries upon admission and showed clinical improvement at follow-up. Although their patient cohort was small and the authors were unable to control for time to clinical follow-up or time to imaging, their data suggest that there may be an absolute threshold for lesion size that predicts neurologic recovery.45 Boghosian et al46 correlated the NLI with the anatomic location of the spinal cord lesions on the MRIs of 109 cervical spinal cord injured patients. The authors found a statistically
437 signiﬁcant correlation between the location of the upper margin of spinal cord edema and hemorrhage as well as the lesion epicenter. The upper boundary of hemorrhage showed a stronger correlation than either edema or lesion epicenter. The lesion length showed no statistical signiﬁcance with NLI. Lesion epicenter and edema length were the best predictors of NLI. The implication of this work is that some MRI measures may be used as an objective measure of the NLI when determination by clinical examination is either inaccurate or unavailable.46
DTI in SCI Although MRI is the best imaging modality for the evaluation of spinal cord parenchyma, conventional MR techniques do not appear to differentiate edema from axonal injury. They are therefore limited to providing anatomic information about the spinal cord parenchyma. The water content or hemorrhagic content does not necessarily reﬂect the status of the white matter tracts and consequently the functional status of the spinal cord is not well assessed. As with white matter tracts in the brain, anisotropy in the spinal cord is probably a result of diffusion barriers encountered as water moves in the direction perpendicular to the ﬁbers. Figure 5 shows DTI in a normal individual with axial fractional anisotropy (FA) maps and DTI tractography. These barriers are believed to be cellular membranes and myelin sheaths, which result in a low transverse apparent diffusion coefﬁcient (tADC). As water diffuses longitudinally in the spinal cord, these diffusion barriers are not encountered, and the longitudinal ADC is therefore large in comparison to tADC. Using either diffusion-weighted imaging or DTI techniques, the preferred direction of anisotropic water diffusion in spinal cord white matter tracts has been shown in numerous ex vivo47-53 and in vivo54-61 experimental studies, as well as in vivo human studies,62-70 to be parallel, or longitudinal, to the long axis of the axons. A critical goal of spinal cord imaging research is a noninvasive quantiﬁable predictor of axon loss. Schwartz et al71 have shown that the natural variation of differing axon morphometric parameters (including axon density, axon spacing, and axon diameter) between normal spinal cord tracts signiﬁcantly correlates with different directional water diffusion values. Ford et al72 showed that alterations in ADC values were more sensitive than conventional MR techniques in detecting experimental SCI. Following injury, tADC values increased and longitudinal ADC values decreased in both normal and abnormal appearing white matter. These changes resulted in decreased anisotropy. These results imply that there are consequences of SCI that dramatically alter axon structure without changing water content or T2, and therefore would not be detected by conventional MR imaging. Nevo et al51 have shown that measurement of ADC values and anisotropy can be used to quantify SCI and neuroprotection. Changes in ADCs in spinal cord white matter have been correlated with behavioral recovery following cervical lateral funiculus lesion and transplantation of ﬁbroblasts genetically modiﬁed to express brainderived neurotrophic factor.48
K. Talekar et al.
Figure 5 Spinal cord DTI. Normal study. (A) Normal color coded axial FA maps. (B) DTI tractography. (Color version of ﬁgure is available online.)
In the earliest case report which used diffusion-weighted imaging in acute human SCI it was noted that diffusion values decreased acutely at the site of injury, potentially due to cellular and axonal swelling.73 In another case report of a patient with syringomyelia, DTI was able to identify spared white matter around the periphery of the syrinx, underscoring the potential for visualizing spared white matter following trauma.74 In patients with spondylosis and spinal cord compression, it has been seen that diffusion MRI improves sensitivity to cervical myelopathy, however, there have been conﬂicting reports of both increased and decreased ADC values, and it may be that the age and clinical severity of a lesion may be important in relating the imaging ﬁnding to pathophysiology.75,76 DTI may also be able to determine the degree and directionality of glial scarring in the gray following injury, which may go undetected with conventional MRI.77,78 As some current therapies are focused on decreasing the degree of glial scarring following injury, DTI may provide an important noninvasive outcome measure. Although the majority of published works on the application of DTI in SCI use animal models, there are limited published series that illustrate the utility of DTI in human SCI.78-87 Ellingson et al83,84 reported signiﬁcant decreases in FA and MD in a group of chronic spinal cord injured patients compared to normal controls. MD was measurably lower throughout the spinal cord in the injured group and FA reduction was indirectly related to clinical severity. In a small clinical series, Shanmuganathan et al85 demonstrated the feasibility of clinical DTI in acute SCI by reporting a consistent change of DTI parameters in 20 SCI patients compared to normal controls using a standard clinical MRI unit. Whole cord ADC values were signiﬁcantly lower in patients and both the ADC values and FA values were decreased at the site of injury compared with controls. Interestingly, the authors reported a decrease in regional ADC values remote from the site of injury suggesting that the DTI parameters can vary in normal appearing spinal cord on conventional MRI. This supports
the concept that DTI may have greater value in mapping the full extent of injury in conjunction with features from conventional MRI. In a subsequent study, Cheran et al86 found statistically signiﬁcant differences in mean diffusivity (MD), FA (FA), radial (RD), and longitudinal diffusivity (AD) for hemorrhagic and nonhemorrhagic SCI patients compared to controls. For nonhemorrhagic SCI, the investigators found strong correlations between admission motor scores (total MIS) and average MD, FA, RD, and AD at the injury site. This same relationship did not hold for hemorrhagic SCI. In a small cohort of human cervical SCI patients, Chang found that DTI indices correlated better than conventional anatomic MRI.87 While MRI and DTI will likely never provide the same high level of granularity as that of a good quality neurologic examination, it does offer three distinct advantages over the INSCSCI assessment: (1) speed, (2) objectivity and (3) direct visualization of the end organ (the spinal cord). Pharmacologic companies that are engaged in development of SCI therapeutics not only need a reliable measure of neurologic function (ie, the clinical examination) but they also rely upon MR imaging to demonstrate the intrinsic changes to the spinal cord immediately after injury and temporally after direct administration of a therapy. As such, it may be more suitable to select a patient for a clinical trial primarily on the basis of the MRI and DTI ﬁndings and secondarily on the initial neurologic assessment.
Chronic Cervical Spondylotic Myelopathy Cervical spondylosis is a common disorder affecting predominantly the elderly. The term spondylosis encompasses the range of degenerative changes that occur in the structures forming the spine including disc degeneration, disc herniation, degenerative spondylolysthesis, uncovertebral and facet arthropathy and ligamentum ﬂavum infolding. These
Imaging of spinal cord injury degenerative changes have a potential to narrow the spinal canal and thereby impinge or compress upon the neural elements—the spinal cord and the nerve roots—leading to cervical spondylotic myelopathy (CSM) and radiculopathy. Cervical spondylosis is an extremely common disorder with radiographic evidence of degenerative changes in the cervical spine seen in as many as 80% of those over the age of 75 years. The incidence and prevalance of cervical spondylotic myelopathy has been estimated to be 41 and 605 per million in North America, respectively and it is the commonest nontraumatic cause of spinal cord disorder in adults.88 The onset of spondylosis and CSM is insidious with a slow progression leading to difﬁculty in its clinical detection. This is further compounded by the difﬁculty in predicting the progress of myelopathy and its irregular correlation with degree of severity of spondylosis. Imaging plays an important role in grading the severity of spondylosis and detecting cord abnormalities suggesting myelopathy.
Pathophysiology of CSM Although the exact pathophysiology for CSM remains uncertain, the central cause for its development is compressive injury to the cord. This results from static factors leading to spinal canal stenosis, dynamic factors which involve structures producing strain on the cord with cervical motion leading to repetitive direct mechanical injury to the cervical cord and possibly a component of spinal cord ischemia. The static factors consist of canal stenosis produced by the various structures that form the spinal canal. Canal compromise is caused by various factors: (1) Anteriorly—disc herniation or bulges, osteophytes. (2) Anterolaterally—uncovertebral joint hypertrophy or osteophytes. (3) Posteriorly—facet hypertrophy, ligamentum ﬂavum infolding (Fig. 6). These factors decrease the CSF reserve around the spinal cord and contribute to mechanical injury. The canal can be developmentally small in caliber, which is an independent risk factor for development of CSM. Edwards
439 and LaRocca89 predicted a canal size o10 mm as a high risk factor for development of myelopathy, those with a canal size of 13-17 mm were less prone to myelopathy but were more prone to symptomatic cervical spondylosis, and those with a canal size 417 mm were asymptomatic. Morishita et al90 proposed on MRI that a congenital sagittal diameter o13 mm was a signiﬁcant risk factor for development of symptomatic disease. Cervical motion can aggravate spinal cord damage precipitated by static compression. The canal dimension changes with motion and gets narrower in ﬂexion and extension. In ﬂexion, the spinal cord lengthens and occupies the anterior portion of the spinal canal.91 The cord can then get stretched over the disc and osteophytes.92 In extension, buckling of ligamentum ﬂavum causes reduction in subarachnoid space posteriorly and overall greater decrease in the spinal canal dimension compared to ﬂexion.93 Also, this is compounded by the fact that in extension, the spinal cord shortens with a slight increase in the spinal cord thickness further increasing risk of mechanical injury to the cord.94 Apart from extrinsic compression during motion, abnormal motion can also produce shear and strain forces within the cord contributing to cord injury.94 On a cellular level, these factors result in direct injury to neurons and glia and lead to a cascade of events including ischemia, excitotoxicity, and apoptosis.95 This pathophysiology is not unlike what is postulated for traumatic SCI though the time course in CSM is over months to years. Grossly, the central cord, in particular the gray matter of the spinal cord and the anterior horn cells are most severely affected in CSM, especially with severe compression of the cord. Among the white matter tracts, the lateral corticospinal tracts are most prone to injury even with minor compression. The posterior column and more so the anterior column white matter are more resistant to compressive injury.96 In comparison, demyelinating disease preferentialy involve the peripheral white matter rather than the central grey.
Figure 6 Schematic diagram at the level of a cervical intervertebral disc. Note: the narrowing of the central canal by the disc-osteophyte complex and the ligamentum ﬂavum infolding. The neural foramen on the left is narrowed by osteophytic spurring from the uncovertebral and facet joints.
Cervical spondylosis has an insidious onset and slow progression. It commonly starts of as neck pain, which in itself is a nonspeciﬁc symptom and may even be caused by various structures outside the cervical spine. The source for neck pain related to cervical spondylosis can be from the nerves innervating paraspinal muscles, intervertebral discs, and facet joints.97 Narrowing of the spinal canal and neural foramina can lead to impingement and compression on the nerve roots, which can lead to radiculopathy wherein the symptoms would occur in a speciﬁc dermatomal distribution in the upper extremity. The symptoms may be sensory, most frequently pain or motor loss corresponding to the involved nerve root. Cervical myelopathy can lead to varied symptoms and signs depending predominantly on the area of affection of the cord. Crandall and Batzdorf described 5 broad categories of cervical spondylotic myelopathy based on the location of insult in the cord: (1) Transverse lesion syndrome, in which all the motor
K. Talekar et al.
“Torg-Pavlov ratio,” or the “canal-to-body ratio.” Measuring canal diameter using Torg-Pavlov ratio may have value in predicting myelopathy. Yue et al101 showed a signiﬁcant difference in T-P ratio between patients with and without CSM (0.72 vs 0.95 with P o 0.001). However, the ratio has a low positive predictive value due to variability in vertebral size in the population.102 A narrower vertebral size will underestimate severity of canal narrowing by this ratio whereas a wider vertebral body will overestimate narrowing. The utility of this ratio has been disputed by many and a study performed on CT by Blackley et al103 to detect the correlation of this ratio with true canal diameter found this ratio to be a poor predictor of canal stenosis. Although radiographs and particularly CT give a fair idea of the osseous anatomy and severity of canal narrowing, they fail to provide the detailed visualization of the narrowing produced by the soft tissue components of disc and ligaments and are not sensitive in providing accurate information of canal dimensions. MRI with its exquisite soft tissue resolution is the imaging modality of choice for depiction of cervical spondylosis as well as changes in the cervical spinal cord (Fig. 7). A routine MRI protocol for evaluation of cervical degenerative disease would usually consist of imaging in 2 planes— sagittal and axial. The protocol used in our institution consists of sagittal T1, T2, and STIR sequences. Axial T2 sequence at the level of the intervertebral disc allows the ability to further localize the site and cause of canal and foraminal narrowing. Axial T2 is obtained with both spin echo and gradient sequences. Spin echo sequence is better for evaluation of cord
and sensory tracts were involved. (2) Motor system syndrome, in which corticospinal tracts and anterior horn cells were involved leading to spasticity. (3) Central cord syndrome, in which motor and sensory deﬁcits affected the upper extremities more severely than the lower extremities. (4) BrownSéquard syndrome, which consisted of ipsilateral motor deﬁcits with contralateral sensory deﬁcits. (5) Brachialgia and cord syndrome, which consists of radicular pain in the upper extremity along with motor or sensory long-tract signs.98 Though this compartmentalization may not hold always hold true and a mixed pattern of deﬁcit may be produced.
Imaging of Cervical Spondylosis and CSM The primary role of imaging is determination of the static factors leading to spinal canal narrowing. The bony component of the narrowing can be detected by radiographs or more accurately by computed tomography (CT). Radiographs and CT can depict vertebral subluxation, disc space narrowing, which indicates underlying disc degeneration, osteophytes, uncovertebral, and facet arthropathic changes leading to central canal and foraminal narrowing. The sagittal canal dimension measured on a lateral cervical radiograph is not reliable indicator of bony canal dimension and is prone to measurement error due to magniﬁcation and positioning issues. To partially offset these issues, Torg et al99 and Pavlov et al100 proposed measurement of the ratio of the sagittal diameter of the cervical canal divided by the corresponding diameter of the vertebral body, also referred to as the
Figure 7 Cervical spinal stenosis. (A) Plain lateral radiograph of the cervical spine demonstrating osseous changes of cervical spondylosis—multilevel disc space narrowing and osteophytes. (B) Corresponding sagittal T2W MRI of the same patient demonstrating the associated soft tissue components—disc (anterior arrows) and ligaments (posterior arrows) that lead to multilevel spinal canal stenosis.
Imaging of spinal cord injury signal and gradient sequence is better for evaluation of osteophytes. On the anterior aspect of the cervical spinal canal, disc disease and osteophytes are the primary causes for canal narrowing. The normal disc demonstrates central hyperintense signal on T2 because of presence of hydrophilic glycosoaminoglycans. Disc degeneration begins with desiccation of the disc. The disc loses its normal hyperintensity on T2. There can be accompanying degenerative endplate changes in the vertebral bodies. Disc height narrowing and bulging of the annulus occur with more advanced degeneration. A rent in the disc referred to as an annular ﬁssure is often a precursor to disc herniations. Contrary to disc bulge, herniations involve less than one-fourth (901) of the disc circumference. Disc herniations with a base wider than dome are referred to as “protrusions” whereas those with a base narrower than the dome are referred to as “extrusions.” Less commonly, the herniated disc material may lose contact with the parent disc and is referred to “sequestered” disc. In the cervical spine, disc material is often accompanied by marginal osteophytes from the vertebral body and these are referred to as “disc-osteophyte” complexes (Fig. 8). The location of the bulge or herniation and its severity may correlate with neural compression. The centrally located disc herniations can impinge on the ventral cord or the ventral nerve roots. The more peripherally
441 located disc herniations in the foramina or extraforaminal locations can impinge on the exiting nerve roots. Foraminal stenosis, apart from disc herniations is often contributed by osseous narrowing due to osteophytes from the uncovertebral and facet joints. On the posterolateral aspect, the ligamentum ﬂava which run along the lamina of the cervical vertebra can show infolding decreasing the CSF space on the posterolateral aspect of the cord and with greater severity may impinge on the cord. Another component that can produce narrowing from the anterior aspect of the spinal canal is ossiﬁcation (and thickening) of the posterior longitudinal ligament. Though pathophysiologically a distinct entity, OPLL also produces symptoms due to chronic compression of the cervical spinal cord. Apart from evaluation of the anteroposterior spinal canal diameter on the sagittal sequences, various additional quantitative measurements like the transverse area of the spinal cord, compress ratio, maximal canal compromise, and maximal spinal cord compression can be calculated on MRI.104 Numerous grading systems have been proposed to grade the severity of the canal compromise. Muhle et al105 proposed a numeric grading system consisting of grade 0, normal; grade 1, partial obliteration of the anterior or posterior subarachnoid space; grade 2, complete obliteration of the anterior or posterior subarachnoid space; and grade 3, cervical cord
Figure 8 MR axial imaging of disc disease. Axial GRE images demonstrates diffuse disc bulge (A) and a focal disc protrusions (C and D) (arrows). Additional presence of uncovertebral hypertrophy (B) resulting in foraminal stenosis (arrow heads), commonly referred to “disc-osteophyte” complex in the cervical spine. GRE, gradient echo.
K. Talekar et al.
442 compression or displacement (Fig. 9). Kang et al106 suggested a modiﬁcation of this grading system replacing absence or presence of cord deformation as criteria for grades 1 and 2 instead of partial or complete obliteration of the CSF spaces which they considered less reliable. Recently, combined task forces of the North American Spine Society, the American Society of Spine Radiology and the American Society of Neuroradiology have suggested a simple technique for grading spinal stenosis that can be graded as mild, moderate, or severe, if the canal is narrowed by less than a third, one-third to twothirds, or greater than two-thirds of the original diameter respectively; this can also be used to grade spinal stenosis.107 Evaluation of the dynamic component of canal narrowing can be performed with imaging in ﬂexion and extension sequences.108 This would better deﬁne the dynamic component of compression but has not gained widespread acceptance, partly due to the time consumed for performing additional sequences and a lower yield in the general patient population. However, certain cases with suspected higher grade stenosis should beneﬁt from additional imaging in ﬂexion and extension. Kinematic cine MRI sequences depicting real time motion changes in the spinal canal using fast EPI sequences is also a promising tool in better understanding the dynamic component of the disease. Flexion MRI may also aid diagnosis of Hirayama disease, which classically manifests in adolescent males with progressive weakness that stabilizes after
several years. It is characterized by asymmetric muscle weakness and atrophy in the C8-T1 distribution. MR imaging ﬁndings reported include loss of attachment of the dura to the lamina, asymmetric lower cervical spinal cord atrophy, spinal cord T2 hyperintensity, loss of cervical lordosis in the neutral position, and forward displacement of the dura with ﬂexion MR imaging.109,110 MRI evaluates the cord by identifying the degree of spinal cord deformation and compression and detecting spinal cord signal abnormality. The normal cord has a rounded shape on axial imaging. Loss of CSF reserve around the cord with deformation of the cord and ﬂattening of its surface may indicate early cord compression. With chronicity and severity of cord compression, signal abnormalities develop in the spinal cord. T2 hyperintensity in the cord in early stages may represent reversible edema (Fig. 10). More advanced stages tend to have associated T1 hypointensity. This in combination with cord ﬂattening and volume loss is consistent with myelomalacia, which indicates irreversible injury111 (Fig. 11). CT accurately depicts osseous narrowing of the spinal canal compared to radiographs. It is more sensitive in depiction of ossiﬁcation of posterior longitudinal ligament compared to MRI (Fig. 12). It can have a complementary role to MRI in certain cases, especially in preoperative planning. CT myelography also plays a role in imaging patients, especially when the patient has a contraindication for getting an MRI study
Figure 9 Grading of severity of central canal stenosis. (A) Sagittal T2-WI and (B) sagittal STIR images in the same patient. Note: varying levels of central canal stenosis: Grade II at C6-C7 (effacement of CSF without cord deformity—small arrow on A), Grade III at C4-C5 (effaced CSF with cord compression—large arrow on A—and increased cord signal—arrow on B ) and Grade I at other levels. CSF, cerebrospinal ﬂuid.
Imaging of spinal cord injury (Fig. 13). However, intradural injection of contrast agent is an invasive procedure with associated risks and thus should not be considered a ﬁrst-line test in patients who are able to undergo MRI.
443 signiﬁcantly lower and ADC signiﬁcantly higher in cases with cord compression.113
Idiopathic Spinal Cord Herniation DTI in CSM Diffusion properties can be evaluated using quantitative indices such as the apparent diffusion coefﬁcient (ADC), mean diffusivity (MD) and fractional anisotropy (FA). The white matter tracts in the spinal cord have low ADC and high FA values because of tightly packed directional orientation of the ﬁber bundles. Wen et al112 detected that FA values were signiﬁcantly lower in the cervical spinal cord of patients with CSM compared to controls. Compared to T2 signal abnormality, FA values were better able to prognosticate the severity of the myelopathy. Kerkovsky et al also detected that FA values were signiﬁcantly lower and ADC values were signiﬁcantly higher at site of maximal compression in patients with CSM compared to healthy controls. In addition, among patients with CSM, symptomatic patients demonstrated lower FA values and higher ADC values empowering DTI to prognosticate the severity of CSM. Studies have shown FA values to be
Idiopathic spinal cord herniation is deﬁned as herniation of the spinal cord through a dural tear which occured without trauma or surgery. It may be asymptomatic, however, prolonged and severe cases can present with symptoms of myelopathy. It is more common in the mid-thoracic spine with the dural defect often noted in the ventral aspect of the spinal canal. The typical imaging feature is of a focally ventrally displaced deformed cord with prominent dorsal CSF space (Fig. 14). This makes a dorsal arachnoid cyst a close imaging differential and a CT myelogram study may be needed in difﬁcult case to differentiate these two entities.114
Summary Mechanical damage to the cord can occur in the setting of trauma, chronic compression or in the rare cases of cord
Figure 10 Cervical spondylotic myelopathy with cord edema. Sagittal T2 (A) and axial T2 (B) show a large disc protrusion at C5-C6 causing severe spinal canal narrowing, cord compression and cord edema. Note: the edema predominantly affects the central gray (double white arrows). In contrast, in a case of multiple sclerosis, the areas of T2 signal abnormality are almost exclusively seen involving the white matter ﬁber tracts—as seen in (C) where the black arrows point to foci of demyelination in the posterior and left lateral white matter columns.
K. Talekar et al.
Figure 11 Cervical spondylotic myelopathy with myelomalacia. Sagittal T2 (A) and sagittal T1 (B) MRI of the cervical spine demonstrate hyperintense T2 and slight hypointense T1 signal in the spinal cord at C3-4 disc level (white arrow). Axial T2 GRE (C) and axial T2 spin echo (D) demonstrate a central disc protrusion and the cord signal abnormality (thick double white arrows on C and D, respectively) predominantly involving the central gray matter of the cord.
Figure 12 Ossiﬁcation of posterior longitudinal ligament (OPLL). Sagittal and axial CT (A and B) and MRI (C and D) depicting OPLL as a cause of canal narrowing (arrows).
Imaging of spinal cord injury
Figure 13 CT cervical myelogram. Multilevel central canal narrowing, most severe at C6-7 (arrow) where the CSF reserve is lost both anteriorly and posteriorly. CSF, cerebrospinal ﬂuid.
herniation. MRI is the modality of choice for evaluation of cord injury. In the acute traumatic setting, areas of hyperintensity on T2-WI probably reﬂecting edema and particularly hypointensity on GRE images reﬂecting hemorrhagic changes correlate with clinical ﬁndings and worse prognosis. Similarly, abnormal signal in the cord with CSM—usually prominent in the central gray matter—heralds important cord damage.
Figure 14 Idiopathic spinal cord herniation. Focal ventrally displaced cord with prominence of the dorsal subarachoid space. This appearance is typical for cord herniation through a dural defect. When the cord deformity is less marked than this a dorsal arachnoid cyst can be considered as a differential.
1. National spinal cord injury statistical center: Spinal Cord Injury: Facts and Figures at a Glance 2008. Birmingham: University of Alabama, 2008. www.spinalcord.uab.edu 2. Berkowitz M, O'Leary P, Kruse D, et al: Spinal Cord Injury: An Analysis of Medical and Social Costs. New York: Demos Medical Publishing Inc., 1998 3. Bracken MB, Shepard MJ, Collins WF, et al: A randomized, controlled trial of methylprednisolone or naloxone in the treatment of acute spinalcord injury. Results of the second national acute spinal cord injury study. N Engl J Med 322(20):1405-1411, 1990 4. Nesathurai S: Steroids and spinal cord injury: Revisiting the NASCIS 2 and NASCIS 3 trials. J Trauma 45(6):1088-1093, 1998 5. Hurlbert RJ, Moulton R: Why do you prescribe methylprednisolone for acute spinal cord injury? A canadian perspective and a position statement. Can J Neurol Sci 29(3):236-239, 2002 6. Kwon B, Oxland T, Tetzlaff W: Animal models used in spinal cord regeneration research. Spine 27:1504-1510, 2002 7. Kwon B, Tetzlaff W, Grauer J, et al: Pathophysiology and pharmacologic treatment of acute spinal cord injury. Spine J 4:451-464, 2004 8. The National Spinal Cord Injury Statistical Center. Facts and Figures at a Glance. Birmingham: University of Alabama, 1999. 9. Blight A, DeCrescito V: Morphometric analysis of experimental spinal cord injury in the cat: The relationship of injury intensity to survival of myelinated axons. Neuroscience 19:321-341, 1986 10. McDonald J, Sadowsky C: Spinal-cord injury. Lancet 359:417-425, 2002 11. McDonald J, Althomsons S, Hyrc K, et al: Oligodendrocytes are highly vulnerable to AMPA/kainate receptor-mediated excitotoxicity. Nat Med 4: 291-297, 1998 12. Choi D: Exocitotoxic cell death. J Neurobiol 23:1261-1276, 1992 13. Emery E, Aldana P, Bunge M, et al: Apoptosis after traumatic human spinal cord injury. J Neurosurg 89:911-920, 1998 14. Sakamoto A, Ohnishi S, Ohnishi T, et al: Relationship between free radical production and lipid peroxidation during ischemia-reperfusion injury in the rat brain. Brain Res 53(554):186-192, 2001 15. Kirshblum SC, Burns SP, Biering-Sorensen F, et al: International standards for neurological classiﬁcation of spinal cord injury (revised 2011). J Spinal Cord Med 34(6):535-546, 2011 16. adley MN, Walters BC, Aarabi B, et al: Clinical assessment following acute cervical spinal cord injury. Neurosurgery 72:40-53, 2013 17. Burns AS, Lee BS, Ditunno Jr JF, et al: Patient selection for clinical trials: The reliability of the early spinal cord injury examination. J Neurotrauma 20(5):477-482, 2003 18. Mulcahey MJ, Gaughan J, Betz RR, et al: The International Standards for Neurological Classiﬁcation of Spinal Cord Injury: Reliability of data when applied to children and youths. Spinal Cord 45(6):452-459, 2007 19. Anderson K, Aito S, Atkins M, et al: Functional recovery measures for spinal cord injury: an evidence-based review for clinical practice and research. J Spinal Cord Med 31(2):133-144, 2008 20. Itzkovich M, Gelernter I, Biering-Sorensen F, et al: The Spinal Cord Independence Measure (SCIM) version III: Reliability and validity in a multi-center international study. Disabil Rehabil 29(24):1926-1933, 2007 21. Waters RL, Yakura JS, Adkins RH, et al: Recovery following complete paraplegia. Arch Phys Med Rehabil 73(9):784-789, 1992 22. Waters RL, Adkins RH, Yakura JS, et al: Motor and sensory recovery following complete tetraplegia. Arch Phys Med Rehabil 74(3):242-247, 1993 23. Ditunno Jr JF, Stover SL, Freed MM, et al: Motor recovery of the upper extremities in traumatic quadriplegia: A multicenter study. Arch Phys Med Rehabil 73(5):431-436, 1992 24. Ishida Y, Tominaga T: Predictors of neurologic recovery in acute central cervical cord injury with only upper extremity impairment. Spine (Phila Pa 1976) 27(15):1652-1658, 2002. [discussion 1658] 25. Marino RJ, Ditunno Jr JF, Donovan WH, et al: Neurologic recovery after traumatic spinal cord injury: Data from the model spinal cord injury systems. Arch Phys Med Rehabil 80(11):1391-1396, 1999
K. Talekar et al.
446 26. Frankel HL, Hancock DO, Hyslop G, et al: The value of postural reduction in the initial management of closed injuries of the spine with paraplegia and tetraplegia I. Paraplegia 7(3):179-192, 1969 27. Maynard FM, Reynolds GG, Fountain S, et al: Neurological prognosis after traumatic quadriplegia. Three-year experience of California regional spinal cord injury care system. J Neurosurg 50(5):611-616, 1979 28. Waters RL, Adkins RH, Yakura JS: Deﬁnition of complete spinal cord injury. Paraplegia 29(9):573-581, 1991 29. Consortium for Spinal Cord Medicine: Outcomes following traumatic spinal cord injury: Clinical practice guidelines for health-care professionals. J Spinal Cord Med 23(4):289-316, 2000 30. Oleson CV, Burns AS, Ditunno JF, et al: Prognostic value of pinprick preservation in motor complete, sensory incomplete spinal cord injury. Arch Phys Med Rehabil 86(5):988-992, 2005 31. Waters RL, Adkins RH, Yakura JS, et al: Motor and sensory recovery following incomplete paraplegia. Arch Phys Med Rehabil 75(1):67-72, 1994 32. Foo D, Subrahmanyan TS, Rossier AB: Post-traumatic acute anterior spinal cord syndrome. Paraplegia 19(4):201-205, 1981 33. Crozier KS, Graziani V, Ditunno Jr JF, et al: Spinal cord injury: Prognosis for ambulation based on sensory examination in patients who are initially motor complete. Arch Phys Med Rehabil 72(2):119-121, 1991 34. Katoh S, el Masry WS: Motor recovery of patients presenting with motor paralysis and sensory sparing following cervical spinal cord injuries. Paraplegia 33(9):506-509, 1995 35. Burns SP, Golding DG, Rolle Jr WA, et al: Recovery of ambulation in motor-incomplete tetraplegia. Arch Phys Med Rehabil 78(11): 1169-1172, 1997 36. Fawcett JW, Curt A, Steeves JD, et al: Guidelines for the conduct of clinical trials for spinal cord injury as developed by the ICCP panel: Spontaneous recovery after spinal cord injury and statistical power needed for therapeutic clinical trials. Spinal Cord 45(3):190-205, 2007 37. Flanders AE, Spettell CM, Tartaglino LM, et al: Forecasting motor recovery after cervical spinal cord injury: Value of MR imaging. Radiology 201(3):649-655, 1996 38. Flanders AE, Schaefer DM, Doan HT, et al: Acute cervical spine trauma: Correlation of MR imaging ﬁndings with degree of neurologic deﬁcit. Radiology 177(1):25-33, 1990 39. Marciello MA, Flanders AE, Herbison GJ, et al: Magnetic resonance imaging related to neurologic outcome in cervical spinal cord injury. Arch Phys Med Rehabil 74(9):940-946, 1993 40. Schaefer DM, Flanders A, Northrup BE, et al: Magnetic resonance imaging of acute cervical spine trauma. Correlation with severity of neurologic injury. Spine 14(10):1090-1095, 1989 41. Schaefer DM, Flanders AE, Osterholm JL, et al: Prognostic signiﬁcance of magnetic resonance imaging in the acute phase of cervical spine injury. J Neurosurg 76(2):218-223, 1992 42. Kulkarni MV, Bondurant FJ, Rose SL, et al: 1.5 tesla magnetic resonance imaging of acute spinal trauma. Radiographics 8(6):1059-1082, 1988 43. Miyanji F, Furlan JC, Aarabi B, et al: Acute cervical traumatic spinal cord injury: MR imaging ﬁndings correlated with neurologic outcome— Prospective study with 100 consecutive patients. Radiology 243(3): 820-827, 2007 44. 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. Am J Neuroradiol 20(5): 926-934, 1999 45. Boldin C, Raith J, Fankhauser F, et al: Predicting neurologic recovery in cervical spinal cord injury with postoperative MR imaging. Spine (Phila Pa 1976) 31(5):554-559, 2006 46. Boghosian G, Leypold BG, Flanders AE, et al. Predicting the neurological level of injury with MRI following cervical spinal cord injury. RSNA, 2006. 47. Pattany PM, Puckett WR, Klose KJ, et al: High-resolution diffusionweighted MR of fresh and ﬁxed cat spinal cords: Evaluation of diffusion coefﬁcients and anisotropy. Am J Neuroradiol 18(6):1049-1056, 1997 48. Schwartz ED, Shumsky JS, Wehrli S, et al: Ex vivo MR determined apparent diffusion coefﬁcients correlate with motor recovery mediated
by intraspinal transplants of ﬁbroblasts genetically modiﬁed to express BDNF. Exp Neurol 182(1):49-63, 2003 Schwartz ED, Chin CL, Shumsky JS, et al: Apparent diffusion coefﬁcients in spinal cord transplants and surrounding white matter correlate with degree of axonal dieback after injury in rats. Am J Neuroradiol 26(1):7-18, 2005 Schwartz ED, Cooper ET, Chin CL, et al: Ex vivo evaluation of ADC values within spinal cord white matter tracts. Am J Neuroradiol 26(2):390-397, 2005 Nevo U, Hauben E, Yoles E, et al: Diffusion anisotropy MRI for quantitative assessment of recovery in injured rat spinal cord. Magn Reson Med 45(1):1-9, 2001 Gulani V, Webb AG, Duncan ID, et al: Apparent diffusion tensor measurements in myelin-deﬁcient rat spinal cords. Magn Reson Med 45 (2):191-195, 2001 Weglarz WP, Adamek D, Markiewicz J, et al: Analysis of the diffusion weighted MR microscopy data of excised spinal cord of a rat on the basis of the model of restricted diffusion. Solid State Nucl Magn Reson 25(1-3): 88-93, 2004 Fraidakis M, Klason T, Cheng H, et al: High-resolution MRI of intact and transected rat spinal cord. Exp Neurol 153(2):299-312, 1998 Inglis BA, Yang L, Wirth 3rd ED, et al: Diffusion anisotropy in excised normal rat spinal cord measured by NMR microscopy. Magn Reson Imaging 15(4):441-450, 1997 Silver X, Ni WX, Mercer EV, et al: In vivo 1H magnetic resonance imaging and spectroscopy of the rat spinal cord using an inductivelycoupled chronically implanted RF coil. Magn Reson Med 46(6): 1216-1222, 2001 Elshaﬁey I, Bilgen M, He R, et al: In vivo diffusion tensor imaging of rat spinal cord at 7 T. Magn Reson Imaging 20(3):243-247, 2002 Bilgen M, Elshaﬁey I, Narayana PA: In vivo magnetic resonance microscopy of rat spinal cord at 7 T using implantable RF coils. Magn Reson Med 46(6):1250-1253, 2001 Madi S, Hasan KM, Narayana PA: Diffusion tensor imaging of in vivo and excised rat spinal cord at 7 T with an icosahedral encoding scheme. Magn Reson Med 53(1):118-125, 2005 Mogatadakala KV, Narayana PA: In vivo diffusion tensor imaging of thoracic and cervical rat spinal cord at 7 T. Magn Reson Imaging 27 (9):1236-1241, 2009 Fenyes DA, Narayana PA: In vivo diffusion characteristics of rat spinal cord. Magn Reson Imaging 17(5):717-722, 1999 Wheeler-Kingshott CA, Hickman SJ, Parker GJ, et al: Investigating cervical spinal cord structure using axial diffusion tensor imaging. Neuroimage 16(1):93-102, 2002 Tsuchiya K, Fujikawa A, Suzuki Y: Diffusion tractography of the cervical spinal cord by using parallel imaging. Am J Neuroradiol 26(2):398-400, 2005 Tsuchiya K, Fujikawa A, Honya K, et al: Diffusion tensor tractography of the lower spinal cord. Neuroradiology 50(3):221-225, 2008 Clark CA, Werring DJ, Miller DH: Diffusion imaging of the spinal cord in vivo: Estimation of the principal diffusivities and application to multiple sclerosis. Magn Reson Med 43(1):133-138, 2000 Holder CA, Muthupillai R, Mukundan Jr S, et al: Diffusion-weighted MR imaging of the normal human spinal cord in vivo. Am J Neuroradiol 21(10):1799-1806, 2000 Nagayoshi K, Kimura S, Ochi M, et al: Diffusion-weighted echo planar imaging of the normal human cervical spinal cord. J Comput Assist Tomogr 24(3):482-485, 2000 Bammer R, Augustin M, Prokesch RW, et al: Diffusion-weighted imaging of the spinal cord: Interleaved echo-planar imaging is superior to fast spin-echo. J Magn Reson Imaging 15(4):364-373, 2002 Dietrich O, Herlihy A, Dannels WR, et al: Diffusion-weighted imaging of the spine using radial k-space trajectories. MAGMA 12(1):23-31, 2001 Murphy BP, Zientara GP, Huppi PS, et al: Line scan diffusion tensor MRI of the cervical spinal cord in preterm infants. J Magn Reson Imaging 13 (6):949-953, 2001 Schwartz ED. Experimental techniques of spinal imaging. Spinal Trauma: Imaging, Diagnosis, and Management. pp. 373-407, 2007.
Imaging of spinal cord injury 72. Ford JC, Hackney DB, Alsop DC, et al: MRI characterization of diffusion coefﬁcients in a rat spinal cord injury model. Magn Reson Med 31(5):488-494, 1994 73. Sagiuchi T, Tachibana S, Endo M, et al: Diffusion-weighted MRI of the cervical cord in acute spinal cord injury with type II odontoid fracture. J Comput Assist Tomogr 26(4):654-656, 2002 74. Schwartz ED, Falcone SF, Quencer RM, et al: Posttraumatic syringomyelia: Pathogenesis, imaging, and treatment. Am J Roentgenol 173(2):487-492, 1999 75. Facon D, Ozanne A, Fillard P, et al: MR diffusion tensor imaging and ﬁber tracking in spinal cord compression. Am J Neuroradiol 26(6): 1587-1594, 2005 76. Demir A, Ries M, Moonen CT, et al: Diffusion-weighted MR imaging with apparent diffusion coefﬁcient and apparent diffusion tensor maps in cervical spondylotic myelopathy. Radiology 229(1):37-43, 2003 77. Schwartz ED, Duda J, Shumsky JS, et al: Spinal cord diffusion tensor imaging and ﬁber tracking can identify white matter tract disruption and glial scar orientation following lateral funiculotomy. J Neurotrauma 22 (12):1388-1398, 2005 78. Kelley BJ, Harel NY, Kim C, et al: Diffusion tensor imaging as a predictor of locomotor function after experimental spinal cord injury and recovery. J Neurotrauma 31:1362-1373, 2014 79. Tsuchiya K, Fujikawa A, Honya K, et al: Value of diffusion-weighted MR imaging in acute cervical cord injury as a predictor of outcome. Neuroradiology 48(11):803-808, 2006 80. Tsuchiya K, Katase S, Fujikawa A, et al: Diffusion-weighted MRI of the cervical spinal cord using a single-shot fast spin-echo technique: Findings in normal subjects and in myelomalacia. Neuroradiology 45 (2):90-94, 2003 81. Shen H, Tang Y, Huang L, et al: Applications of diffusion-weighted MRI in thoracic spinal cord injury without radiographic abnormality. Int Orthop 31(3):375-383, 2007 82. Beslow LA, Ichord RN, Zimmerman RA, et al: Role of diffusion MRI in diagnosis of spinal cord infarction in children. Neuropediatrics 39 (3):188-191, 2008 83. Ellingson BM, Kurpad SN, Schmit BD: Functional correlates of diffusion tensor imaging in spinal cord injury. Biomed Sci Instrum 44:28-33, 2008 84. Ellingson BM, Ulmer JL, Schmit BD: A new technique for imaging the human spinal cord in vivo. Biomed Sci Instrum 42:255-260, 2006 85. Shanmuganathan K, Gullapalli RP, Zhuo J, et al: Diffusion tensor MR imaging in cervical spine trauma. Am J Neuroradiol 29(4):655-659, 2008 86. Cheran S, Shanmuganathan K, Zhuo J, et al: Correlation of MR diffusion tensor imaging parameters with ASIA motor scores in hemorrhagic and nonhemorrhagic acute spinal cord injury. J Neurotrauma 28: 1881-1892, 2011 87. Chang Y, Jung T, Yoo DS, et al: Diffusion tensor imaging and ﬁber tractography of patients with cervical spinal cord injury. J Neurotrauma 27:2033-2040, 2010 88. Nouri A, Tetreault L, Singh A, et al: Degenerative cervical myelopathy: Epidemiology, genetics, and pathogenesis. Spine 40(12):E675-E693, 2015 89. Edwards WC, LaRocca H: The developmental segmental sagittal diameter of the cervical spinal canal in patients with cervical spondylosis. Spine 8(1):20-27, 1983 90. Morishita Y, Naito M, Hymanson H, et al: The relationship between the cervical spinal canal diameter and the pathological changes in the cervical spine. Eur Spine J 18(6):877-883, 2009 91. Breig A, Turnbull I, Hassler O: Effects of mechanical stresses on the spinal cord in cervical spondylosis. A study on fresh cadaver material. J Neurosurg 25(1):45-56, 1966 92. Panjabi M, White 3rd A: Biomechanics of nonacute cervical spinal cord trauma. Spine 13(7):838-842, 1988 93. Parke WW: Correlative anatomy of cervical spondylotic myelopathy. Spine 13(7):831-837, 1988
447 94. Henderson FC, Geddes JF, Vaccaro AR, et al: Stretch-associated injury in cervical spondylotic myelopathy: New concept and review. Neurosurgery 56(5):1101-1113, 2005. [discussion 1101-13] 95. Baptiste DC, Fehlings MG: Pathophysiology of cervical myelopathy. Spine J 6:190S-197S, 2006 96. Fehlings MG, Skaf G: A review of the pathophysiology of cervical spondylotic myelopathy with insights for potential novel mechanisms drawn from traumatic spinal cord injury. Spine 23(24):2730-2737, 1998 97. Rao R: Neck pain, cervical radiculopathy, and cervical myelopathy: pathophysiology, natural history, and clinical evaluation. J Bone Joint Surg Am Vol 84-A(10):1872-1881, 2002 98. Crandall PH, Batzdorf U: Cervical spondylotic myelopathy. J Neurosurg 25(1):57-66, 1966 99. Torg JS, Pavlov H, Genuario SE, et al: Neurapraxia of the cervical spinal cord with transient quadriplegia. J Bone Joint Surg Am Vol 68 (9):1354-1370, 1986 100. Pavlov H, Torg JS, Robie B, et al: Cervical spinal stenosis: Determination with vertebral body ratio method. Radiology 164(3):771-775, 1987 101. Yue WM, Tan SB, Tan MH, et al: The Torg–Pavlov ratio in cervical spondylotic myelopathy: a comparative study between patients with cervical spondylotic myelopathy and a nonspondylotic, nonmyelopathic population. Spine 26(16):1760-1764, 2001 102. Herzog RJ, Wiens JJ, Dillingham MF, et al: Normal cervical spine morphometry and cervical spinal stenosis in asymptomatic professional football players. Plain ﬁlm radiography, multiplanar computed tomography, and magnetic resonance imaging. Spine 16:S178-S186, 1991 103. Blackley HR, Plank LD, Robertson PA: Determining the sagittal dimensions of the canal of the cervical spine. The reliability of ratios of anatomical measurements. J Bone Joint Surg Br Vol 81(1):110-112, 1999 104. Karpova A, Craciunas S, Chua SY, et al: Accuracy and reliability of MRI quantitative measurements to assess spinal cord compression in cervical spondylotic myelopathy: A prospective study. Evid Based Spine Care J 1 (2):56-57, 2010 105. Muhle C, Metzner J, Weinert D, et al: Classiﬁcation system based on kinematic MR imaging in cervical spondylitic myelopathy. Am J Neuroradiol 19(9):1763-1771, 1998 106. Kang Y, Lee JW, Koh YH, et al: New MRI grading system for the cervical canal stenosis. Am J Roentgenol 197(1):W134-W140, 2011 107. Fardon DF, Williams AL, Dohring EJ, et al: Lumbar disc nomenclature: Version 2.0. Spine J 14(11):2525-2545, 2014 108. Zhang L, Zeitoun D, Rangel A, et al: Preoperative evaluation of the cervical spondylotic myelopathy with ﬂexion-extension magnetic resonance imaging: About a prospective study of ﬁfty patients. Spine 36(17): E1134-E1139, 2011 109. Hirayama K, Tokumaru Y: Cervical dural sac and spinal cord in juvenile muscular atrophy of distal upper extremity. Neurology 54(10): 1922-1926, 2000 110. Lehman VT, Luetmer PH, Sorenson EJ, et al: Cervical spine MR imaging ﬁndings of patients with Hirayama disease in North America: A multisite study. Am J Neuroradiol 34(2):451-456, 2013 111. Fernandez de Rota JJ, Meschian S, Fernandez de Rota A, et al: Cervical spondylotic myelopathy due to chronic compression: The role of signal intensity changes in magnetic resonance images. J Neurosurg Spine 6 (1):17-22, 2007 112. Wen CY, Cui JL, Liu HS, et al: Is diffusion anisotropy a biomarker for disease severity and surgical prognosis of cervical spondylotic myelopathy? Radiology 270(1):197-204, 2014 113. Kerkovsky M, Bednarik J, Dusek L, et al: Magnetic resonance diffusion tensor imaging in patients with cervical spondylotic spinal cord compression: Correlations between clinical and electrophysiological ﬁndings. Spine 37(1):48-56, 2012 114. Parmar H, Park P, Brahma B, et al: Imaging of idiopathic spinal cord herniation. Radiographics 28(2):511-518, 2008