Central Cord Syndrome

Abstract Douglas D. Nowak, MD Central cord syndrome is the most common type of incomplete Joseph K. Lee, MD spinal cord injury. This syndrome most often occurs in older persons with underlying cervical spondylosis caused by a Daniel E. Gelb, MD hyperextension mechanism. It also occurs in younger persons who Kornelis A. Poelstra, MD, PhD sustain trauma to the cervical spine and, less commonly, as a Steven C. Ludwig, MD result of nontraumatic causes. The upper extremities are more affected than the lower extremities, with motor function more severely impaired than sensory function. Central cord syndrome presents a spectrum, from weakness limited to the hands and forearms with sensory preservation, to compete quadriparesis with sacral sparing as the only evidence of incomplete spinal cord injury. Historically, treatment has been nonsurgical, but recovery is often incomplete. Early surgical treatment of central cord syndrome remains controversial. However, recent studies have shown beneﬁts, particularly of early surgery to decompress the spinal cord in patients with pathologic conditions revealed by radiography or MRI.

Dr. Nowak is Orthopaedic Surgery Resident, New York-Presbyterian Hospital, Columbia University pinal cord injuries (SCIs) are clas- Knowledge of the organization of Medical Center, New York, NY. Ssified as complete or incomplete. the spinal cord and the pathophysiol- Dr. Lee is Orthopaedic Surgery ogy of CCS is essential in selecting Resident, New York-Presbyterian The American Spinal Injury Associa- Hospital, Columbia University tion (ASIA) defines complete injury the optimal treatment method. Non- Medical Center. Dr. Gelb is as the absence of sensory and motor surgical management is sufficient in Associate Professor of function below the level of injury.1 some instances, but recent research Orthopaedics, Department of has shown potential benefit from Orthopaedics, University of Conversely, with incomplete injury, Maryland Medical Center, Baltimore, some neurologic function remains surgical management. MD. Dr. Poelstra is Assistant below the level of injury. Incomplete Professor of Orthopaedics, injuries include central cord syn- Department of Orthopaedics, Anatomy University of Maryland Medical drome (CCS), anterior cord syn- Center. Dr. Ludwig is Associate drome, posterior cord syndrome, Professor and Chief of Spine and Brown-Séquard syndrome. Knowledge of spinal cord anatomy Surgery, Department of is critical in understanding CCS. CCS is the most common type of in- Orthopaedics, University of The spinal cord fills approximately complete SCI, comprising 15% to 25% Maryland Medical Center. 50% of the canal in the cervical and of all cases.2-4 This syndrome was first Reprint requests: Dr. Ludwig, thoracolumbar spine. Cerebrospinal described by Schneider et al5 in 1954. Department of Orthopaedics, fluid, epidural fat, and dura sur- University of Maryland Medical Classic CCS presents as underlying cer- round the cord and fill the remainder Center, Suite S11B, 22 South vical spondylosis in the older patient Greene Street, Baltimore, MD of the canal space. A myelomere is (aged >60 years) who sustains a hyper- 21201. the segment of the cord from which extension injury without any evidence a nerve root arises; each is located J Am Acad Orthop Surg 2009;17: of damage to the bony spine. CCS also 756-765 one level above the same-numbered occurs in younger persons who sustain vertebral body in the cervical and Copyright 2009 by the American higher-energy trauma resulting in spi- upper thoracic regions (eg, the C5 Academy of Orthopaedic Surgeons. nal fractures or instability. nerve root myelomere is at the level

756 Journal of the American Academy of Orthopaedic Surgeons Douglas D. Nowak, MD, et al of the C4 vertebral body). Figure 1 The neural elements within the spinal cord are arranged geographically. The long tracts extending from the brain are arranged peripherally and are composed primarily of white matter. The peripheral white matter is abundant in the cervical spine because that region includes the long tracts to the cervical, thoracic, lumbar, and sacral levels. The more central gray matter contains the lower motor neurons. The main descending motor pathway is the lateral corticospinal tract. The upper motor neuron originates in the contralateral cerebral cortex, decussates in the midbrain, and de- scends on the ipsilateral lateral pe- riphery of the spinal cord. The upper motor neuron then synapses with its Illustration of the cervical spinal cord (axial cut). Note the orientation of the corresponding lower motor neurons lateral corticospinal tract and dorsal column tracts (fasciculus gracilis, in the anterior horn of the gray mat- fasciculus cuneatus), with the sacral structures being more peripherally ter. The lateral corticospinal tract has located and the cervical structures more centrally located. This view illustrates why central cord syndrome preferentially affects the upper traditionally been thought to be ar- extremities. C = cervical, L = lumbar, S = sacral, T = thoracic. (Adapted with ranged with the cervical structures permission from Gupta MC, Benson DR, Keenan TL: Initial evaluation and more centrally located and the sacral emergency treatment of the spine-injured patient, in Browner BD, Jupiter JB, structures more peripherally located Levine AM, Trafton PG, eds: Skeletal Trauma: Basic Science, Management, and Reconstruction, ed 3. Philadelphia, PA, Saunders, 2003, pp 685-707.) (Figure 1). Whether this lamination exists is controversial.6 Hand and forearm musculature are primarily The major ascending sensory path- cord and ascend in the contralateral lat- supplied by the large motor axons of ways include the posterior column eral spinothalamic tract. Conversely, the lateral corticospinal tract.6,7 The tracts (fasciculus gracilis, fasciculus cu- proprioception and vibratory sensation ventral corticospinal tract is a minor neatus) and the smaller lateral spinotha- ascend ipsilaterally in the posterior descending motor pathway. The mo- lamic tracts (Figure 1). Sensory neuron column of the spinal cord and cross tor fibers in the ventral corticospinal cell bodies are located in the dorsal root only after reaching the brain stem. tract do not decussate in the mid- ganglion, and sensory input enters the Similar to the lateral corticospinal tract, brain, and these fibers descend on posterior horn of the gray matter. Pain the dorsal columns are arranged the contralateral side of the spinal and temperature input immediately such that the sacral structures are more cord. cross to the opposite side of the spinal peripherally located and the cervical

Dr. Gelb or a member of his immediate family serves as a board member, owner, officer, or committee member of AO Spine North American Education committee; has received royalties from Globus Medical; is a member of a speakers’ bureau or has made paid presentations on behalf of Synthes; serves as a paid consultant to or is an employee of Alphatec Spine and Synthes; has received research or institutional support from Synthes; and has stock or stock options held in Alphatec Spine. Dr. Poelstra or a member of his immediate family is a member of a speakers’ bureau or has made paid presentations on behalf of and serves as a paid consultant to or is an employee of DePuy. Dr. Ludwig or a member of his immediate family has received royalties from DePuy and Globus Medical; is a member of speakers’ bureau or has made paid presentation on behalf of AO, DePuy, Globus Medical, Stryker, and Synthes; serves as a paid consultant to or is an employee of AO, DePuy, Globus Medical, Stryker, and Synthes; has received research or institutional support from Synthes; has stock or stock options held in Globus Medical; and has received nonincome support (such as equipment or services), commercially derived honoraria, or other non–research-related funding (such as paid travel) from AO, DePuy, Globus Medical, Stryker, and Synthes. Neither of the following authors nor a member of their immediate families has received anything of value from or owns stock in a commercial company or institution related directly or indirectly to the subject of this article: Dr. Nowak and Dr. Lee.

December 2009, Vol 17, No 12 757 Central Cord Syndrome

5,12,13 Figure 2 matic before injury. Vertebral fractures and dislocations typically are absent in persons in this age group with CCS. The cervical cord can be injured by direct compression from buckling of the ligamenta flava into an already narrowed spinal canal. Younger persons with congenital cervical stenosis also are at increased risk of sustaining CCS as a result of hyperextension injury. CCS and its variants, including transient quadriplegia, cervical cord neurapraxia, and burning hand syndrome, have been reported in several football players with congenital stenosis.14-16 Football players are at increased risk because of the de- Illustration of central cord syndrome (CCS) in the cervical spinal cord (axial cut). The colored area is affected in cases of CCS. Note that the sacral mands of the game, particularly structures are more peripheral in the dorsal columns and the lateral tackling. The terms “transient quad- corticospinal tract; thus, those structures are preferentially spared in persons riplegia” and “cervical cord neura- with CCS. C = cervical, L = lumbar, S = sacral, T = thoracic. (Adapted with permission from Gupta MC, Benson DR, Keenan TL: Initial evaluation and praxia” are used interchangeably in emergency treatment of the spine-injured patient, in Browner BD, Jupiter JB, the literature to describe significant, Levine AM, Trafton PG, eds: Skeletal Trauma: Basic Science, Management, sometimes complete, upper and and Reconstruction, ed 3. Philadelphia, PA, Saunders, 2003, pp 685-707.) lower extremity weakness and sensory disturbances that typically last 10 or 15 minutes and then resolve on structures are more centrally located and diving injuries. CCS can also re- their own. (Figure 1). sult from nontraumatic causes, such In younger persons without pre- as spinal epidural abscess.11 existing stenosis or spondylosis, a The classic presentation of CCS in- Mechanism and higher-energy traumatic mechanism volves an older patient with underlying Pathophysiology is required to cause CCS. Often, the cervical spondylosis who sustains an in- mechanism consists of a severe spinal SCI can be divided into two patho- jury as the result of hyperextension of column injury with an associated physiologic phases: primary and sec- the head and neck relative to the tor- fracture-dislocation, resulting in an 5,8,9,12 ondary. Primary injury occurs at the so. The hyperextension mechan- unstable spine.8,9,17 A third subset of time of the inciting trauma. Such in- ism often seems to be very mild but patients includes younger persons jury can be caused either directly by in the setting of cervical spondylosis with traumatic disk herniation that excessive flexion, extension, and/or can result in marked neurologic in- results in CCS in the absence of spi- rotation of the spinal cord or indi- jury. Persons with CCS have a nal fracture or instability.17 rectly by displaced bone or disk ma- smaller sagittal diameter of the cervi- CCS originally was theorized to con- terial having an impact on the spinal cal spinal canal compared with the sist of injury to the central gray matter cord. Secondary injury occurs after average, and >90% of patients with and the central portion of the long the inciting traumatic event. Such in- CCS aged >40 years have been tracts, with preservation of the periph- jury is caused by an incompletely un- shown to have underlying cervical eral structures (Figure 2). Injury to the derstood complex reaction involving spine conditions, such as spondylosis central gray matter and cord hemor- a combination of an inflammatory with osteophyte formation, canal rhage were thought to be the main response and neuronal cell apoptosis stenosis, and ossification of the causes of CCS.5 However, recent (ie, programmed cell death). posterior longitudinal ligament.8 Al- studies have shown that the lateral Trauma is the most common cause though some patients experience cortical spinal tract in the mid to up- of CCS.2,4,5,8-10 CCS occurs most often minor symptoms, the underlying cer- per cervical spine contains the main after motor vehicle accidents, falls, vical spondylosis often is asympto- area of pathologic abnormality.6,18,19

758 Journal of the American Academy of Orthopaedic Surgeons Douglas D. Nowak, MD, et al

MRI studies have failed to show any Figure 3 evidence of cord hemorrhage.8,19,20 Two autopsy studies found no evidence of lower motor neuron injury or cord parenchyma hemorrhage; however, diffuse injury to the large- diameter motor axons in the lateral corticospinal tract was noted.18,19 Patients with CCS exhibit wallerian degeneration of the axonal tracts distal to the zone of injury in the lateral corticospinal tracts.18 In a feline model of SCI, large-diameter axons were shown to be more susceptible to injury than smaller axons.21 Trom- bly and Guest11 presented the only report of neuromonitoring of a case of CCS. The motor-evoked potentials were more severely affected than Axial (A) and sagittal (B) T2-weighted magnetic resonance images of a 72- year-old man with underlying cervical spondylosis who developed central the somatosensory-evoked poten- cord syndrome after a hyperextension injury. tials, and the hand musculature, particularly the abductor pollicis brevis, was most severely affected. Thus, in- Short tau inversion recovery jury to the large myelinated axons of Diagnosis and Evaluation (STIR) sequences can complement the lateral corticospinal tract seems to be the main cause of the deficits Rapid and accurate diagnosis is essen- the sagittal sections. Hyperintense associated with CCS. This explains tial in the patient with suspected SCI. signal within the parenchyma of the the primary involvement of the fine A full radiographic evaluation should cervical spinal cord is typically dem- motor movements of the distal upper be done, typically consisting of cross- onstrated on T2-weighted MRI and extremity.6,7 table lateral, AP, and open-mouth STIR sequences in patients with CCS presents on a spectrum, from odontoid views. Coronal and sagittal CCS. This finding is consistent with weakness limited solely to the hands reconstruction CT scans can be ob- edema without any evidence of pa- and forearms with sensory preserva- tained to gain a better understanding of renchymal hemorrhage.19,20 T2- tion, to complete quadriparesis with bony injury and to detect injuries that weighted imaging and/or STIR se- sacral sparing as the only evidence of are not obvious on plain radiographs. quences are critical in evaluating incomplete SCI. The upper extremi- CT is also helpful in assessing possible injury to the anterior and posterior ties are more severely affected than injuries to the occipitocervical and cer- soft tissues, such as the intervertebral the lower extremities. In particular, vicothoracic junctions. In trauma cen- disk and posterior ligamentous ten- the hands and forearms are most af- ters, CT frequently is performed either sion band. T2-weighted MRI may fected. Bladder dysfunction, typically in addition to or instead of traditional also reveal prevertebral hyperinten- urinary retention and bowel and sex- radiography. Because of the approxi- sity, which has been shown to be a ual dysfunction, may be present in mately 10% to 20% incidence of non- predictor of spinal instability.20 Un- more severe cases. Motor function contiguous spine fractures, when a cer- derlying cervical spondylosis and return, if any occurs, proceeds in a vical spine fracture is identified, the stenosis can be assessed with plain caudad to cephalad manner. Toe entire spine often is imaged with either and advanced imaging studies. flexors are the first to return, fol- plain radiography or CT.22,23 MRI In evaluating a patient with SCI, it lowed by the toe extensors, and then can be useful in further assessing the is essential to determine the extent of the structures innervated by the lum- presence of soft-tissue injury or cord neurologic injury (ie, incomplete ver- bar cord (eg, those that enable ankle compression. MRI evaluation typi- sus complete). Incomplete injuries dorsiflexion). Recovery is usually cally includes axial, coronal, and have a greater chance for neurologic less complete in the upper extremi- sagittal sections of T1- and T2- recovery, whereas motor recovery is ties than in the lower extremities. weighted images (Figure 3). achieved in only 3% of patients with

December 2009, Vol 17, No 12 759 Central Cord Syndrome

Figure 4 ing consists of perianal sensation, rec- tal tone, and activity of the great toe flexor. Spinal shock can occur after severe SCI. This is defined as a state of complete areflexia and usually re- solves within 24 hours of the time of injury.10 The completeness of the neurologic injury cannot be deter- mined until the spinal shock has resolved. The return of the bulbocaver- nosus reflex heralds the end of spinal shock. The clinical test assesses the integrity of the intact S3-S4 arc and is performed by squeezing the glans penis, placing pressure on the clito- ris, or tugging on a Foley catheter. An intact reflex will result in contraction of the anal sphincter. An accurate system of clinical neurologic assessment and recording in- American Spinal Injury Association Standard Neurologic Classiﬁcation of fluences treatment decisions, allows Spinal Cord Injury Worksheet.1 (Reproduced with permission from the Ameri- for reliable serial monitoring, and can Spinal Injury Association.) provides prognostic information. The ASIA worksheet details the Table 1 physical examination (Figure 4). Motor Grading Scale for Suspected Spinal Cord Injury1 Each of the 28 dermatomes can be assessed bilaterally for sensory func- Grade Description tion to pinprick and light touch. 0 Total paralysis Normal sensation is grade 2, altered 1 Palpable or visual contraction is grade 1, and absent is 0. Ten key 2 Active movement, full range of motion, gravity eliminated muscle groups, five in the upper ex- 3 Active movement, full range of motion, against gravity tremities and five in the lower ex- 4 Active movement, full range of motion, against some resistance tremities, are tested bilaterally and 5 Active movement, full range of motion, provides normal resistance graded on a standard scale of 0 to 5 points (Table 1). The motor scores are summed to obtain the ASIA mo- complete injury during the first 24 must test touch and pinprick sensa- tor score (maximum, 100). The ASIA hours and never after 24 to 48 tion in the perianal area as well as has classified the level of impairment hours.10,24 Complete injuries origi- voluntary contracture of the external from complete SCI to varying levels nally were defined as the absence of anal sphincter. of incomplete SCI (Table 2). The motor and sensory function more Sacral sparing is an important indi- ASIA defines the level of injury as than three levels below the zone of cator of incomplete SCI because it sig- the most caudal level that has intact injury. Incomplete injuries, in contra- nifies at least partial continuity of the motor and sensory function on both distinction, involve some preserva- long white-matter tracts (ie, corticospi- sides of the body. tion of motor or sensory function be- nal and spinothalamic) from the conus low the level of injury. The ASIA has medullaris to the cerebral cortex. At the more recently redefined complete time of initial evaluation of a patient Management SCI as the absence of sensory and who has sustained SCI, sacral sparing motor functions in the lowest sacral may be the only neurologic function Nonsurgical segments (ie, S4-S5).1 Thus, to diag- present to differentiate incomplete from In most cases of CCS, the patient ex- nosis SCI as complete, the physician complete SCI. Evaluation of sacral spar- periences considerable neurologic

760 Journal of the American Academy of Orthopaedic Surgeons Douglas D. Nowak, MD, et al and motor recovery without surgical Table 2 intervention.5,8,12,25-28 The work con- 5 American Spinal Injury Association Impairment Scale for Spinal Cord ducted by Schneider et al in 1954 Injury1 suggested that the natural history of Grade Description CCS led to spontaneous recovery and that surgical management was A Complete: no motor or sensory function is preserved in the sacral unnecessary and possibly harmful. segments (S4-S5) Optimal medical management, early B Incomplete: sensory function is preserved below the level of injury, but immobilization, and possibly intra- no motor function is preserved venous steroids have improved the C Incomplete: motor function is preserved below the level of injury, but more than half the affected muscles have a grade ≤3 overall prognosis of CCS.5,25-28 D Incomplete: motor function is preserved below the level of injury and at Proper medical management of the least half the affected muscles have a grade ≥3 person with CCS requires that the E Normal: motor and sensory function are normal patient be placed in intensive care during the initial period after injury. Central venous and indwelling arterial catheters can be used for close of any neuroprotective agent, includ- milestones are met, from outpatient monitoring of hemodynamics and re- ing steroids.29 Administration of ste- therapy. sponses to therapy. A Swan-Ganz roids may adversely affect patient catheter may be necessary. Maintain- outcome, especially in patients with Surgical ing adequate blood pressure (mean penetrating injuries. Surgery is an adjunct to medical arterial pressure >85 mm Hg) by vol- Any patient with suspected CCS management in the patient with ume resuscitation supplemented by should be promptly immobilized CCS; all of the aforementioned prin- vasopressors, if needed, has been with a hard cervical orthosis (eg, ciples of medical management should shown to improve neurologic out- Philadelphia collar) to prevent fur- be included in the treatment regimen 25 come. Blood pressure augmentation ther motion injury. The hard collar is regardless whether surgical interven- presumably maximizes spinal cord used for at least 6 weeks or until tion is indicated. Older studies sug- perfusion and limits secondary in- neck pain has resolved and asso- gested that surgical management of jury. ciated neurologic improvement is CCS might be detrimental and inef- Although controversial, intrave- noted. Patients with no evidence of fective. Schneider et al5 and Morgan nously administered methylpredniso- axial skeletal instability after com- et al32 showed that decompression lone is the most commonly used plete radiographic evaluation are laminectomy did not improve patient pharmacologic treatment for com- kept in the hard cervical orthosis and neurologic status. One patient woke plete and incomplete SCI. The Na- are mobilized early after medical sta- up quadriplegic after surgical de- tional Acute Spinal Cord Injury bilization. Patients with unstable compression.5 However, nonsurgical Study II and III trials established the fractures or dislocations can be management of CCS has not been standard dosing of a 30 mg/kg bolus treated surgically or with application successful in certain patients, and 26,27 followed by 5.4 mg/kg/hr. The in- of cervical tongs and/or halo rings in more recent research has suggested fusion is continued for 24 hours the emergency setting to facilitate that surgical intervention may be when started within 3 hours of the skeletal traction and early closed re- beneficial in certain subsets of per- time of injury or for 48 hours when duction.12,30 Traction provides the sons with CCS.13,17,20,30,32-35 started between 3 and 8 hours from most urgent form of spinal cord de- the time of injury. No proven benefit compression; early closed reduction Indications of steroid administration has been is associated with neurologic im- Spinal instability is the only absolute in- shown when the patient presents >8 provement.30,31 dication for surgical intervention. Spi- hours from the time of injury.26,27 Early mobilization and rehabilitation nal instability has been defined as an- However, a recent and extensive lit- with physical therapy and occupational gular displacement >11° compared with erature review, which was formu- therapy are essential once the patient is an adjacent vertebra or vertebral body lated into clinical practice guidelines medically stable. Retraining hand func- translation >3.5 mm.36 Instability on the early acute management of tion and gait are the main goals. Many can worsen secondary injury to the adults with SCI, showed that there patients benefit from initial intensive in- spinal cord by acting as a dynamic currently is no evidence for the use patient rehabilitation and, after certain factor, further damaging the cord.

December 2009, Vol 17, No 12 761 Central Cord Syndrome

Figure 5 Although no guidelines currently are available for the timing of decompression in cases of acute CCS, studies suggest that early decompression is feasible and may result in improved outcomes, especially in the patient with progressive neurologic deterioration.30,33-35 However, there is no evidence to support when, or even if, surgery should be performed on a patient who shows neurologic improvement.

Procedures Surgical options are dictated by the pathologic abnormality. Imaging techniques such as CT and MRI al- low the surgeon to identify the site of compression. The extent of surgery is individualized for each patient, and Preoperative (A) and postoperative (B) lateral plain radiographs of the same multilevel stenosis is a typical presen- patient as in Figure 3. The patient underwent C3-C6 anterior decompression tation (Figure 5). An anterior, a pos- and fusion with instrumentation because of early neurologic deterioration. terior, or a combined approach is used to relieve pressure on the cord.

Assessment of stability can be diffi- Timing Clinical Outcomes cult, and integrity of the discoliga- The timing of surgical intervention in mentous complex is the key factor in persons with CCS is controversial, with Nonsurgical determining the stability of the spinal two exceptions: (1) overt spinal insta- Currently, no prospective, randomized 37 motion segment. Surgical interven- bility with acute dislocation in which studies have been published comparing tion for CCS without spinal early reduction and fixation should be nonsurgical with surgical treatment of instability is controversial. No evi- performed and (2) in rare cases of CCS. However, several studies have re- dence currently supports decompres- progressive neurologic deficit. Recent ported good results with nonsurgical sion in a patient who demonstrates studies have attempted to elucidate the management.4,5,9,10,28 Recovery is neurologic improvement. Patients optimal surgical protocol for CCS. gradual and often incomplete and is with persistent cord compression, With traumatic CCS, early surgical in- related to the severity of the injury. failure of motor recovery, or pro- tervention (≤24 hours of injury) per- Pain typically is not a major sequela longed neurologic plateau or dete- formed on patients with pathologic ab- of CCS. The progression of neuro- rioration, however, may benefit normality (ie, fracture-dislocation, logic and motor recovery usually be- from surgical intervention.17,19,22,32,34,35 acute herniated disk) confirmed with gins in the lower extremities, contin- Cord compression can occur as a re- radiography or MRI was associated ues with improved bladder and sult of a herniated disk, an epidural with greater motor and neurologic re- bowel control, and ends with upper hematoma, or bone fragments invad- covery than was late surgical interven- extremity control. Restoration of ing the canal. Early removal of the tion (>24 hours after injury).30,34 La hand function is variable; in some offending spinal cord compression Rosa et al35 conducted a systematic patients, hand function does not re- may prevent the progression of analysis of the literature and con- cover. Hand function impairment is chronic myelopathic changes and cluded that early surgical interven- the most common long-term disabil- may lead to improved recovery and tion (≤24 hours of injury) was as- ity associated with CCS.4,9 Although overall function.17,19,32 Underlying sociated with better neurologic most patients achieve major neuro- spinal stenosis and spondylosis are outcomes than either delayed surgi- logic improvements, some face con- further relative indications for surgical intervention or nonsurgical man- siderable long-term functional im- cal management. agement. pairment.

762 Journal of the American Academy of Orthopaedic Surgeons Douglas D. Nowak, MD, et al

Good prognostic factors in cases of independently, and 69% had bladder ASIA motor scores within 72 hours CCS include young age, preinjury em- control. Of the three surviving pa- of injury and at follow-up visits. The ployment, level of education, absence tients aged >70 years, only one could authors found a marked increase in of spinal cord signal abnormality shown ambulate independently, and none ASIA motor scores from a mean by MRI, higher initial ASIA motor had bladder control. score of 58.7 at the time of injury to 8 score, absence of spasticity, early mo- Ishida and Tominaga reported the a mean score of 92.3 at last follow- tor recovery, and good hand func- first prospective study of patients up. The best predictor of final ASIA 2,4,5,8,9,12,17,28,38 tion. Spinal column in- with CCS in 2002. Their study was motor score was the initial score at stability, degree of canal stenosis, limited to patients with weakness in the time of injury. Level of formal persistent spasticity, and medical co- the upper extremities only. At 2-year education was another noteworthy morbidities all correlate with poorer follow-up of the 22 patients, all of prognostic factor, with patients who neurologic recovery.2,20,28,39 whom were treated nonsurgically, had higher levels of education In their original reports in 1954 the authors observed that none had achieving greater recovery. At last and 1958, Schneider and col- any fractures or dislocations, 77% follow-up, 81% of patients reported leagues5,12 noted that of 17 patients had achieved full motor recovery, bowel and bladder continence, and treated medically, 2 died without im- 23% had mild dysfunction or weak- 86% were capable of independent provement, 14 achieved neurologic ness of the hands, and none had se- improvement but still had residual vere dysfunction. Motor and sensory ambulation. However, 59% had a deficits, and 1 regained complete recovery occurred rapidly during the large degree of spasticity, and 34% function. In 1971, Bosch et al2 pre- initial 3 weeks and, in most patients, expressed dissatisfaction with their sented one of the first studies that in- reached a plateau within approxi- symptoms. cluded a long-term follow-up period mately 6 weeks. The absence of ab- (4 months to 26 years) of patients normal MRI signal intensity was the Surgical with CCS who were treated nonsur- best predictor of recovery at final Patients with persistent cord com- gically. They noted that at least some follow-up, suggesting less severe in- pression, failure of motor recovery, return of neurologic function oc- jury in cases with normal MRI find- or prolonged neurologic plateau or curred in 75% of the 42 patients. In- ings. Early neurologic improvement deterioration may benefit from surgi- dependent walking improved from and younger age were also found to cal intervention.17,19,22,32,34,35,38 How- 19% immediately after injury to be significant contributors to im- ever, few data directly compare sur- 57% after rehabilitation. Likewise, proved outcome. Of note, neurologic gical and nonsurgical management. bladder control improved from 17% improvement during the first 6 Bose et al40 retrospectively com- to 53%. Importantly, the authors weeks was a stronger predictor of pared motor function recovery in 28 noted that only 43% of patients re- final neurologic function than was patients with CCS who were treated gained functional use of their hands admitting neurologic status. These either surgically or nonsurgically. A and that 24% of patients who ini- results are promising and show ex- greater degree of recovery was noted tially experienced neurologic im- cellent prognosis for nonsurgical in the surgical group. Surgery was provement reported a plateau, fol- treatment of the less severe form of performed on patients with spinal in- lowed by deterioration in neurologic CCS and for persons with symptoms stability and on those who had failed and functional recovery. This clinical limited to the upper extremities. to improve progressively after an ini- course suggested a chronic form of However, caution is required when tial period of improvement and who CCS marked by spasticity and py- attempting to extrapolate these re- had evidence of persistent compres- ramidal tract involvement. sults to more severe forms of CCS. sion. In 2000, Newey et al38 presented In 2005, Dvorak et al28 reported a Chen et al17 conducted an initial long-term (mean, 8.6 years), retro- rigorous study of patients with CCS. retrospective study and found two spective outcomes of 32 patients In contrast to the study conducted by groups that benefited from surgical with CCS who were treated nonsur- Ishida and Tominaga,8 Dvorak et al28 intervention: younger patients re- gically and reported on the differ- did not exclude the more severe gardless of radiographic abnormali- ences in recovery related to patient forms of CCS. Their prospective ties and older patients with clinically age. All six patients aged <50 years analysis had a minimum 2-year correlated encroaching cord lesions. could walk independently and had follow-up (average, almost 6 years), Chen et al13 subsequently conducted a bladder continence. In patients aged during which all patients underwent prospective study of 37 patients with 50 to 70 years, 77% could ambulate formal evaluation and calculation of preexisting cervical spondylosis and in-

December 2009, Vol 17, No 12 763 Central Cord Syndrome complete cord injury who were oper- ment algorithm and the timing of sur- neural substrate for manual dexterity. ated on 2 to 14 days after injury. The gical intervention, if chosen, for any J Neurosci 1993;13:5105-5118. surgical group experienced more rapid nonprogressive neurologic deficit that 8. Ishida Y, Tominaga T: Predictors of neurologic recovery in acute central neurologic recovery than did the non- remains static. cervical cord injury with only upper surgical group. Recovery was signifi- extremity impairment. Spine (Phila Pa cantly slower for the nonsurgical group 1976) 2002;27:1652-1658. (P = 0.005), and the hospital stay was References 9. Penrod LE, Hegde SK, Ditunno JF Jr: Age effect on prognosis for functional longer. However, by 2 years, no signif- Evidence-based Medicine: Levels of recovery in acute, traumatic central cord icant difference was shown between the syndrome. Arch Phys Med Rehabil 1990; evidence are described in the table of 71:963-968. two cohorts (P = 0.06). Patients with contents. In this article, references 26 cervical stenosis affecting more than 10. Stauffer ES: Neurologic recovery and 27 are level I studies. References following injuries to the cervical spinal three vertebral levels experienced cord and nerve roots. Spine (Phila Pa 8, 13, and 33 are level II studies. poorer outcomes whether treated sur- 1976) 1984;9:532-534. Level III studies include references 9, gically or nonsurgically. 11. Trombly R, Guest JD: Acute central cord 10, 17, 25, 28, 30, 34, 35, and 40. Most studies that describe surgical syndrome arising from a cervical Level IV studies include references 2, epidural abscess: Case report. treatment are retrospective analyses Neurosurgery 2007;61:E424-E425. 4, 5, 11, 12, 14-16, 22-24, 31, 32, that are subject to selection biases and 37-39. Reference 29 is a level V 12. Schneider RC, Thompson JM, Bebin J: and confounding variables. No pro- The syndrome of acute central cervical study. spective randomized studies have spinal cord injury. J Neurol Neurosurg Psychiatry 1958;21:216-227. been performed comparing surgical Citation numbers printed in bold type indicate references published 13. Chen TY, Dickman CA, Eleraky M, with nonsurgical management. Thus, Sonntag VK: The role of decompression the true benefit of surgical treatment within the past 5 years. for acute incomplete cervical spinal cord is unclear. Well-designed prospective injury in cervical spondylosis. Spine 1. American Spinal Injury Association: (Phila Pa 1976) 1998;23:2398-2403. studies to examine the value and tim- Standard Neurological Classification of Spinal Cord Injury Worksheet. Atlanta, 14. Brigham CD, Adamson TE: Permanent ing of surgical intervention are GA, American Spinal Injury Association, partial cervical spinal cord injury in a needed. 2006. professional football player who had only congenital stenosis: A case report. 2. Bosch A, Stauffer ES, Nickel VL: J Bone Joint Surg Am 2003;85:1553- Incomplete traumatic quadriplegia: A 1556. Summary ten-year review. JAMA 1971;216:473- 478. 15. Finnoff JT, Mildenberger D, Cassidy CD: Central cord syndrome in a football CCS is the most common incomplete 3. Gupta MC, Benson DR, Keenan TL: player with congenital spinal stenosis: A SCI. With the increasing age of the pop- Initial evaluation and emergency case report. Am J Sports Med 2004;32: treatment of the spine-injured patient, in 516-521. ulation, physicians will be encounter- Browner BD, Jupiter JB, Levine AM, ing more patients with CCS. Thus, it is Trafton PG, eds: Skeletal Trauma: Basic 16. Ladd AL, Scranton PE: Congenital Science, Management, and cervical stenosis presenting as transient essential to understand the anatomy, Reconstruction, ed 3. Philadelphia, PA, quadriplegia in athletes: Report of two pathophysiology, and treatment of Saunders, 2003, pp 685-707. cases. J Bone Joint Surg Am 1986;68: 1371-1374. CCS. Although most patients with CCS 4. Merriam WF, Taylor TK, Ruff SJ, can expect gradual, albeit often incom- McPhail MJ: A reappraisal of acute 17. Chen TY, Lee ST, Lui TN, et al: Efficacy traumatic central cord syndrome. J Bone of surgical treatment in traumatic central plete, recovery with nonsurgical treat- Joint Surg Br 1986;68:708-713. cord syndrome. Surg Neurol 1997;48: ment, recent studies have shown poten- 435-440. 5. Schneider RC, Cherry G, Pantek H: The tial benefits of early surgery. All CCS syndrome of acute central cervical spinal 18. Jimenez O, Marcillo A, Levi AD: A patients should receive optimal medi- cord injury: With special reference to the histopathological analysis of the human mechanisms involved in hyperextension cervical spinal cord in patients with cal management and early immobiliza- injuries of cervical spine. J Neurosurg acute traumatic central cord syndrome. tion. In our practice, the primary sur- 1954;11:546-577. Spinal Cord 2000;38:532-537. gical indication is fracture with 6. Levi AD, Tator CH, Bunge RP: Clinical 19. Quencer RM, Bunge RP, Egnor M, et al: associated central cord injury. The frac- syndromes associated with Acute traumatic central cord syndrome: disproportionate weakness of the upper MRI-pathological correlations. tures typically present as extension- versus the lower extremities after cervical Neuroradiology 1992;34:85-94. distraction injury. Early surgical treat- spinal cord injury. Neurosurgery 1996; 38:179-183. 20. Song J, Mizuno J, Inoue T, Nakagawa ment is recommended in the absence of H: Clinical evaluation of traumatic spinal instability in the patient with neu- 7. Bortoff GA, Strick PL: Corticospinal central cord syndrome: Emphasis on terminations in two new-world primates: clinical significance of prevertebral rologic deterioration. Further research Further evidence that corticomoto- hyperintensity, cord compression, and is needed to determine the best treat- neuronal connections provide part of the intramedullary high-signal intensity on

764 Journal of the American Academy of Orthopaedic Surgeons Douglas D. Nowak, MD, et al

magnetic resonance imaging. Surg mesylate for 48 hours in the treatment of 34. Guest J, Eleraky MA, Apostolides PJ, Neurol 2006;65:117-123. acute spinal cord injury: Results of the Dickman CA, Sonntag VK: Traumatic Third National Acute Spinal Cord Injury central cord syndrome: Results of 21. Blight AR, Decrescito V: Morphometric Randomized Controlled Trial. National surgical management. J Neurosurg 2002; analysis of experimental spinal cord Acute Spinal Cord Injury Study. 97(1 suppl):25-32. injury in the cat: The relation of injury JAMA 1997;277:1597-1604. intensity to survival of myelinated axons. 35. La Rosa G, Conti A, Cardali S, Cacciola Neuroscience 1986;19:321-341. 28. Dvorak MF, Fisher CG, Hoekema J, F, Tomasello F: Does early decompres- et al: Factors predicting motor recovery sion improve neurological outcome of 22. Henderson RL, Reid DC, Saboe LA: and functional outcome after traumatic spinal cord injured patients? Appraisal of Multiple noncontiguous spine fractures. central cord syndrome: A long-term the literature using a meta-analytical Spine (Phila Pa 1976) 1991;16:128-131. follow-up. Spine (Phila Pa 1976) 2005; approach. Spinal Cord 2004;42:503- 30:2303-2311. 512. 23. Vaccaro AR, An HS, Lin S, Sun S, Balderston RA, Cotler JM: 29. Consortium for Spinal Cord Medicine: 36. White AA, Punjabi MM: The problem of Noncontiguous injuries of the spine. Early Acute Management in Adults With clinical instability in the human spine: A J Spinal Disord 1992;5:320-329. Spinal Cord Injury: A Clinical Practice systematic approach, in White AA, Guideline for Health-Care Providers. Punjabi MM, eds: Clinical Biomechanics 24. Bohlman HH: Acute fractures and Washington, DC, Paralyzed Veterans of of the Spine, ed 2. Philadelphia, PA, dislocations of the cervical spine: An America, 2008, pp 21-23. Lippincott, 1990, pp 277-378. analysis of three hundred hospitalized patients and review of the literature. 30. Papadopoulos SM, Selden NR, Quint 37. Vaccaro AR, Hulbert RJ, Patel AA, et al: J Bone Joint Surg Am 1979;61:1119- DJ, Patel N, Gillespie B, Grube S: The subaxial cervical spine injury 1142. Immediate spinal cord decompression for classification system: A novel approach cervical spinal cord injury: Feasibility to recognize the importance of 25. Vale FL, Burns J, Jackson AB, Hadley and outcome. J Trauma 2002;52:323- morphology, neurology, and integrity of MN: Combined medical and surgical 332. the disco-ligamentous complex. Spine treatment after acute spinal cord injury: (Phila Pa 1976) 2007;32:2365-2374. Results of a prospective pilot study to 31. Katoh S, el Masry WS, Jaffray D, et al: assess the merits of aggressive medical Neurologic outcome in conservatively 38. Newey ML, Sen PK, Fraser RD: The resuscitation and blood pressure treated patients with incomplete closed long-term outcome after central cord management. J Neurosurg 1997;87:239- traumatic cervical spinal cord injuries. syndrome: A study of the natural history. 246. Spine (Phila Pa 1976) 1996;21:2345- J Bone Joint Surg Br 2000;82:851-855. 2351. 26. Bracken MB, Shepard MJ, Collins WF Jr, 39. Yamazaki T, Yanaka K, Fujita K, et al: Methylprednisolone or naloxone 32. Morgan TH, Wharton GW, Austin GN: Kamezaki T, Uemura K, Nose T: treatment after acute spinal cord injury: The results of laminectomy in patients Traumatic central cord syndrome: 1-year follow-up data. Results of the with incomplete spinal cord injuries. Analysis of factors affecting the second National Acute Spinal Cord Paraplegia 1971;9:14-23. outcome. Surg Neurol 2005;63:95-99. Injury Study. J Neurosurg 1992;76:23- 31. 33. Fehlings MG, Perrin RG: The timing of 40. Bose B, Northrup BE, Osterholm JL, surgical intervention in the treatment of Cotler JM, DiTunno JF: Reanalysis of 27. Bracken MB, Shepard MJ, Holford TR, spinal cord injury: A systematic review central cervical cord injury management. et al: Administration of methylpred- of recent clinical evidence. Spine (Phila Neurosurgery 1984;15:367-372. nisolone for 24 or 48 hours or tirilazad Pa 1976) 2006;31(11 suppl):S28-S35.

December 2009, Vol 17, No 12 765