Posterior Surgery for Cervical Myelopathy: Indications, Techniques, and Outcomes




This article details the controversies associated with the different treatment strategies in patients with cervical spondylotic myelopathy. The natural history, incidence, pathophysiology, physical examination, and imaging findings are discussed followed by the indications, techniques, and outcomes of patients treated with posterior cervical decompression via decompressive laminectomy, laminectomy and instrumented fusion, and laminoplasty.


Cervical spondylosis is the most common cause of acquired disability in patients over the age of 50 and cervical spondylotic myelopathy (CSM) is the most common progressive spinal cord disorder in patients over the age of 55. Despite this, controversy remains in the literature regarding the optimal treatment regimen and optimal timing of treatment when caring for patients with CSM, particularly when the disease process is in its early stages. This is likely due to the paucity of well-designed prospective studies with validated outcome measures that clearly delineate patient populations and identify the best course of treatment.


Surgical treatments to decompress the neural elements in patients with CSM that is moderate to severe (modified Japanese Orthopaedic Association [mJOA] score <12) or progressive in nature have been shown to significantly improve functional status and pain when compared with nonoperatively treated patients. Nonoperative modalities, however, have shown maintenance of or even improvement in functional status in patients under 75 years old with mild CSM (mJOA score >12). Nonoperative treatment modalities should be entertained before embarking on surgical decompression in patients with mild CSM. Even though the traditional natural history of myelopathic patients is a slow, stepwise decline, there is a subset of patients who maintain a steady state without further neurologic decline with nonoperative treatment. Conversely, there is a subset of patients who have shown benefit from surgical decompression even in the setting of mild myelopathy, and factors associated with stabilization/improvement of functional outcomes include advanced age, long duration of symptoms, and/or rapid progression of symptoms.


Additionally, controversy exists surrounding the optimal surgical approach for patients with CSM: anterior versus posterior decompression. Both have been shown beneficial. Typically, the anterior approach is used when shorter decompressions are required and the posterior approach is used when longer (3+ segment) decompressions are needed. Regardless of the approach, the goal of surgery in myelopathic patients is adequate decompression of the neural elements to halt progression of symptoms and improve patient symptomology. This article discusses in detail the techniques, outcomes, and associated complications for each of 3 specific techniques used to treat cervical myelopathy from a posterior approach.


Incidence and etiology


The true incidence of CSM at present is unknown yet it remains the most common cause of spinal cord dysfunction in patients over the age of 55. This is likely due to the subtle findings in patients living with mild myelopathy or patients not seeking care. What is known is that as the population ages and life spans increase the incidence of patients presenting with CSM will likely continue to rise.


CSM is a clinical diagnosis made by identifying long-tract signs in the upper and lower extremities and is largely a result of direct mechanical compression on the spinal cord. Compressive effects on the spinal cord have a multitude of causes, grouped as static or dynamic factors. Static factors represent narrowing of the spinal canal secondary to acquired conditions (eg, spondylosis) and/or developmental conditions (eg, congenitally narrow canal). Dynamic factors, such as segmental vertebral instability, also play a role in spinal cord compression and may aggravate the underlying static factors, resulting in repetitive cord compression and neurologic injury. Additionally, Breig and colleagues have shown that morphologic changes occur within the spinal cord with flexion and extension that may also play a role in the development of myelopathy. As the cervical spine flexes the spinal cord stretches and it may be prone to pathologic compression if it abuts spondylotic changes ventrally. As the cervical spine extends the spinal cord thickens and it may be prone to pathologic compression as the ligamentum flavum infolds or it may abut the lamina in congenitally stenotic patients.


The static and dynamic factors that cause spinal cord compression may have additive affects that cause reversible (mild–moderate myelopathy) and irreversible (severe myelopthy) damage to the neurons, axons, and glial cells of the spinal cord. Histologic studies on autopsy specimens have shown loss of neurons and vacuolar degeneration in the gray matter along with demyelination, myelin fragmentation, and swelling of the axons in white matter.




Physical examination and imaging


Patients presenting with myelopathy often complain of issues with balance, gait disturbances, and loss of fine motor skill and hand dexterity. These symptoms are often first noted by family members who notice a wide-based gait and a diminishing ability to use buttons or zippers. Patients may also complain of urinary urgency, frequency, and/or hesitation but rarely note incontinence. Axially based neck pain and radicular symptoms are also common concomitant complaints.


Signs of myelopathy on physical examination include gait instability, positive Romberg sign, hyperreflexia, inverted radial reflex, Hoffmann signs, pathologic clonus, and upgoing toes on plantar stimulation. Ono and colleagues described a characteristic dysfunction of the hand that was observed in patients with cervical spinal cord compression. They noted loss of power of adduction, extension of the ulnar 2 or 3 fingers, and an inability to grip and release rapidly with these fingers. Ono and colleagues termed these changes, myelopathy hand , which seemed due to pyramidal tract involvement. Changes in pain, temperature, and proprioception may also be noted. If spinal cord compression is present above the C3 level, a scapulohumeral reflex may be noted.


Depending on the level of suspicion and magnitude of presenting signs/symptoms, advanced imaging may be necessary. In the absence of any red flags (tumor, trauma, infection, and/or neurologic injury), the acquisition of upright plain radiographs, including dynamic flexion/extension views, is recommended initially. If red flags are present, advanced imaging (CT and/or MRI) should proceed without delay.


Plain radiography may portray evidence of spondylotic changes, including disk space collapse, uncovertebral joint hypertrophy, facet arthropathy, and vertebral endplate sclerosis but the commonality of these findings should be kept in mind. The presence of an ossified posterior longitudinal ligament (OPLL), congenital stenosis, or dynamic instability may be more diagnostic in relation to myelopathy.


MRI remains the gold standard in evaluating the soft tissues of the cervical spine, including the disks, ligaments, and neural elements. Because of the sensitivity of MRI in visualizing abnormalities of the soft tissues, it is important to correlate MRI findings with patients’ signs/symptoms. Intramedullary spinal cord changes on MRI have been correlated with histopathological findings by Ohshio and colleagues. They described abnormally high T2-weighted image signal intensities that appeared nonspecifically in mildly altered lesions or areas with edema. These lesions may resolve with time. A more ominous finding is when a low T1-weighted image in addition to a high T2-weighted image signal intensity appears in the gray matter. This represents severely altered lesions with necrosis, myelomalacia, or spongiform change. Additionally, abnormally high T1-weighted image intensities in the white matter also appear in severely altered lesions.


Currently, the findings of cord signal changes on MRI cannot be correlated with either the prognosis of patients with CSM or their outcome if operative decompression is elected. There are ongoing studies looking at diffusion tensor MRI sequences in hopes of not only providing prognostic and outcome data but also delineating the timing of intervention controversy that exists today.


When evaluating patients with CSM, CT scanning with myelography plays a vital role, especially if contraindications to performing an MRI scan exist (pacemaker, foreign bodies, and so forth). Even though CT myelograms are an invasive procedure, they delineate osseous sources of compression from soft tissue sources that may not be well visualized on MRI scanning, as seen in the setting of OPLL. Shafaie and colleagues compared the inter-rater reliability between CT myelography and MRI scanning and found it only moderately good. The investigators tended to more severely grade the degree of central and neuroforaminal stenosis with CT myelogram, which more clearly delineated the osseus pathology. They concluded that CT myelogram and MRI are complementary studies not exclusive studies.




Physical examination and imaging


Patients presenting with myelopathy often complain of issues with balance, gait disturbances, and loss of fine motor skill and hand dexterity. These symptoms are often first noted by family members who notice a wide-based gait and a diminishing ability to use buttons or zippers. Patients may also complain of urinary urgency, frequency, and/or hesitation but rarely note incontinence. Axially based neck pain and radicular symptoms are also common concomitant complaints.


Signs of myelopathy on physical examination include gait instability, positive Romberg sign, hyperreflexia, inverted radial reflex, Hoffmann signs, pathologic clonus, and upgoing toes on plantar stimulation. Ono and colleagues described a characteristic dysfunction of the hand that was observed in patients with cervical spinal cord compression. They noted loss of power of adduction, extension of the ulnar 2 or 3 fingers, and an inability to grip and release rapidly with these fingers. Ono and colleagues termed these changes, myelopathy hand , which seemed due to pyramidal tract involvement. Changes in pain, temperature, and proprioception may also be noted. If spinal cord compression is present above the C3 level, a scapulohumeral reflex may be noted.


Depending on the level of suspicion and magnitude of presenting signs/symptoms, advanced imaging may be necessary. In the absence of any red flags (tumor, trauma, infection, and/or neurologic injury), the acquisition of upright plain radiographs, including dynamic flexion/extension views, is recommended initially. If red flags are present, advanced imaging (CT and/or MRI) should proceed without delay.


Plain radiography may portray evidence of spondylotic changes, including disk space collapse, uncovertebral joint hypertrophy, facet arthropathy, and vertebral endplate sclerosis but the commonality of these findings should be kept in mind. The presence of an ossified posterior longitudinal ligament (OPLL), congenital stenosis, or dynamic instability may be more diagnostic in relation to myelopathy.


MRI remains the gold standard in evaluating the soft tissues of the cervical spine, including the disks, ligaments, and neural elements. Because of the sensitivity of MRI in visualizing abnormalities of the soft tissues, it is important to correlate MRI findings with patients’ signs/symptoms. Intramedullary spinal cord changes on MRI have been correlated with histopathological findings by Ohshio and colleagues. They described abnormally high T2-weighted image signal intensities that appeared nonspecifically in mildly altered lesions or areas with edema. These lesions may resolve with time. A more ominous finding is when a low T1-weighted image in addition to a high T2-weighted image signal intensity appears in the gray matter. This represents severely altered lesions with necrosis, myelomalacia, or spongiform change. Additionally, abnormally high T1-weighted image intensities in the white matter also appear in severely altered lesions.


Currently, the findings of cord signal changes on MRI cannot be correlated with either the prognosis of patients with CSM or their outcome if operative decompression is elected. There are ongoing studies looking at diffusion tensor MRI sequences in hopes of not only providing prognostic and outcome data but also delineating the timing of intervention controversy that exists today.


When evaluating patients with CSM, CT scanning with myelography plays a vital role, especially if contraindications to performing an MRI scan exist (pacemaker, foreign bodies, and so forth). Even though CT myelograms are an invasive procedure, they delineate osseous sources of compression from soft tissue sources that may not be well visualized on MRI scanning, as seen in the setting of OPLL. Shafaie and colleagues compared the inter-rater reliability between CT myelography and MRI scanning and found it only moderately good. The investigators tended to more severely grade the degree of central and neuroforaminal stenosis with CT myelogram, which more clearly delineated the osseus pathology. They concluded that CT myelogram and MRI are complementary studies not exclusive studies.




Rationale and options for posterior decompression


When considering surgical options in the treatment of patients with CSM, the principle goal is adequate decompression of the neural elements. The key factors to consider when deciding which approach is best suited for a patient are the number of levels requiring decompression, the cervical sagittal alignment, whether or not the anterior cervical spine is ankylosed (ankylosing spondylitis or diffuse idiopathic skeletal hyperostosis), the presence of OPLL, the presence of dynamic instability, and the anatomic location of the compressive structures.


The anterior approach is most commonly used when there are 3 or fewer levels involved, cervical lordosis has been lost, the spine is not ankylosed, or if dynamic instability is present. In patients with OPLL and severe compression (>60% canal compromise), it has been shown that anterior decompression via corpectomy has improved neurologic recovery but has increased complications associated with the anterior approach and may have increased risk of dural injury. The posterior approach should be considered.


The posterior approach is best suited when 3 or more levels are involved, when cervical lordosis is preserved, when the spine is ankylosed, and possibly in the setting of OPLL (discussed previously). Cervical sagittal alignment is crucial to note because a posterior decompression in a kyphotic cervical spine does not allow the spinal cord to migrate posteriorly and, if the kyphosis progresses, further compression on the cord may ensue causing worsening neurologic decline. Within this subset, there are 3 possible techniques discussed in detail: (1) decompressive laminectomy alone, (2) laminectomy and fusion, and (3) laminoplasty.




Posterior techniques for cervical myelopathy


Decompressive Laminectomy


Historically, laminectomy has been successfully used to treat CSM, with many case series showing successful outcomes. Posterior decompression was and is largely reserved for multilevel stenosis, with one- and two-level disease treated anteriorly. The standard laminectomy length is 3 to 5 levels, because this allows room for the spinal cord to migrate away from anterior structures in addition to removing dorsal compression. The concern with removal of this number of laminae and spinous processes without concomitant fusion is the loss of stability, allowing late kyphotic collapse.


The literature on laminectomy for CSM is replete with level-3 evidence and little evidence above this level. There has emerged a clear expectation for modest improvement in some patients treated in this fashion with loss of that improvement in a subset of patients. Arnold and colleagues found an initial 77% improvement in neurologic status, although at late (8-year) follow-up, that was maintained in only 52% of patients. Likewise, Bishara reported improvement in 56% of the 59 patients in their report at 5 years, which dropped to 51% improvement at the 10-year follow-up, with no reported instability in any of the patients. Carol and Ducker reported somewhat better long-term improvement rates. They evaluated 125 patients who underwent laminectomy compared with another 81 patients who underwent anterior decompression and fusion and 10 patients with combined approaches. The laminectomy group compared favorably with 68% improvement rate at a mean 10-year follow-up. Kato and colleagues, like the other investigators, reported an early recovery rate better than long-term improvement. They found a 44% recovery rate 1 year after laminectomy, although by 10 years, this had been maintained in only 32.8% of the 44 patients followed.


Over the past several decades, this technique has fallen out of favor, largely because of the increased rate of complications in comparison with other surgical techniques. Postlaminectomy kyphosis is the most concerning complication and seems to occur in 10% to 45% of patients. Matsunaga and colleagues compared 37 patients treated by laminectomy alone with 64 patients who underwent laminoplasty, with a mean follow-up of greater than 5 years. They found postoperative kyphosis rates of 35% in the laminectomy group and only 7% in the laminoplasty group. Likewise, Kato and colleagues found 47% of patients developed postoperative kyphosis, although this did not correlate with neurologic deterioration. Perez-Lopez and colleagues found a 24% rate of postlaminectomy kyphosis in the 19 patients who underwent laminectomy compared with only 7% in the 17 who underwent laminectomy and fusion, whereas Hamanishi and Tanaka found a postoperative kyphosis rate of only 17% of the 35 patients who underwent laminectomy alone and no difference from the fusion group. They did select those deemed to have preoperative instability to fusion, so the 2 groups were not comparable.


The rate of postlaminectomy kyphosis seems to increase in patients with straightened or kyphotic spines preoperatively. The risk may be lower in patients with advanced degenerative changes, fully collapsed disks, and decreased motion in the cervical spine. Postlaminectomy kyphosis may lead to further neurologic deterioration.


Patients with multiroot radiculopathy tend to do less well with laminectomy alone. This may in part be because of the hesitation to remove too much facet if a fusion is not planned, for fear of creating instability. A more significant facetectomy is often required to adequately decompress the roots, so decompressive laminectomy alone is not the correct procedure when significant foraminal decompression is needed.


Other common complications of decompressive laminectomy for CSM include immediate neurologic loss, progressive long-term neurologic deterioration, hematoma, and wound infection. In an extensive review of 310 patients having undergone laminectomy over a 6-year time period, Halvorsen and colleagues report that 11.6% of patients deteriorated in their neurologic function immediately after surgery, 41.2% improved in function, and 47.2% were unchanged in neurologic function. More than half of the patients who experienced neurologic loss postoperatively had permanent changes. At final follow-up, only half of the patients undergoing laminectomy for CSM had improvement in their neurologic status. This is similar to other studies (discussed previously), with better early results and progressive loss of that improvement at 5-year to 10-year follow-up. Rates of postoperative hematoma are approximately 1% and are more likely in those patients with prior surgery, and postoperative wound infections occur in approximately 1% to 4% of patients after laminectomy.


In response to this concern of postlaminectomy kyphosis and progressive long-term loss of neurologic improvement, Yukawa and colleagues reported on a skip laminectomy technique and prospectively randomized 41 CSM patients to this technique or laminoplasty. Although only short-term (1-year) results are reported to date, they show no differences between groups. This technique may work well if the spine is neutral to lordotic, most of the compression is dorsal, and some dorsal migration of the cord is not required.


Multilevel laminectomy has been a long-used technique for decompressing the spinal cord in patients with CSM. Modest success rates are similar to other techniques, although with a moderate rate of late loss of neurologic improvement and postlaminectomy kyphosis, this technique has become less popular than the others covered in this review.


Laminectomy with Fusion


Given the rate of late neurologic deterioration and the moderate rate of postlaminectomy kyphosis (discussed previously), the concept of performing a posterior cervical fusion at the time of decompression gained interest. Early techniques included onlay bone grafting after decompressive laminectomy. Because of the increased risks of pseudoarthrosis and early kyphotic collapse after onlay bone grafting techniques in the posterior cervical spine, most surgeons advocate for a posterior instrumented fusion because it provides secure stabilization, does not interfere with decompression, and permits early mobilization of the patient. There are several techniques used for instrumented posterior cervical fusion, including sublaminar/facet wiring techniques as well as pedicle/lateral mass screw-rod techniques, but the latter is yet to be approved by the Food and Drug Administration. Pedicle/lateral mass screw-rod placement, however, is favorable due to ease of placement as well as providing rigid segmental fixation in the setting of a wide posterior laminectomy/foraminotomy. This is now the most common technique used for posterior stabilization.


The operative technique for the placement of screws in the posterior cervical spine typically involves the placement of pedicle screws/pars screws at C2, lateral mass screws at C3 through C6, and pedicle screws at C7 and below. Given the complex and variable neurovascular anatomy of the cervical spine, it is recommended that all patients undergo imaging studies to assess the neurovascular anatomy before applying posterior cervical screws for fixation.


Placement of lateral mass screws was initially described by Roy-Camille and Saillant with a recommendation to place the screws in a direction perpendicular to the lamina starting at the midpoint of the lamina. This technique was later modified by Magerl in an effort to increase length of fixation and improve the safety and efficacy of screw placement to avoid the exiting nerve root (by aiming cephalad) and the vertebral artery (by aiming laterally). Their recommendations were to start slightly superiorly and medially to the midpoint of the lamina and aim 25° degrees and 45° in the cephalad direction. Several anatomic studies have been performed with reports of the Magerl technique having a higher incidence of potential nerve root injury and another study stating that the Roy-Camille technique is safer at C3 and C4 and the Magerl technique at C5 and C6 because of the varying angles of the facet joints. Both techniques were described with bicortical fixation and, with the use of locked polyaxial screws commonly used today, most surgeons place these unicortically to minimize the risk of nerve root injury while maintaining fixation strength ( Fig. 1 ).


Oct 6, 2017 | Posted by in ORTHOPEDIC | Comments Off on Posterior Surgery for Cervical Myelopathy: Indications, Techniques, and Outcomes

Full access? Get Clinical Tree

Get Clinical Tree app for offline access