Clinical Evaluation

As for any patient, a complete clinical evaluation including a directed history and physical examination is the essential first step in the work-up of a patient with degenerative cervical disease. The history and physical examination should be directed toward establishing or confirming the diagnosis and understanding the impact of the disorder on the patient’s quality of life. Signs and symptoms of cervical radiculopathy or myelopathy or frequently a combination of the two may be present.

As the diagnosis is confirmed and the course of the disease established, the presence or absence of factors that may affect the surgical strategy is determined. The patient is queried regarding the presence, nature, location, quality, and intensity of the neck pain. Exacerbating and ameliorating factors are identified. A history of previous neck or cervical spine surgery is elicited. The patient’s occupation and any serious avocations are determined; of particular interest are activities that involve speaking or singing. The patient is specifically interrogated regarding any swallowing dysfunction. Any of these factors may favor the use of a posterior approach.

Imaging

Adequate imaging evaluation usually includes anteroposterior, lateral, and lateral flexion-extension cervical radiographs (including the occiput through T1) and either magnetic resonance imaging (MRI) or computed tomography (CT), the latter with or without preceding myelography. In some cases both MRI and CT will be desirable. Oblique radiographs, flexion-extension MRI, or other studies may be obtained in specific cases.

The three most important factors in determining whether a posterior (dorsal) approach to cervical degenerative disease is appropriate or feasible are the location of the disease (dorsal or ventral), the extent of disease (number of spinal segments involved), and the alignment of the cervical spine (lordotic, straight, or kyphotic). Compression that is purely dorsal or dorsolateral is generally best relieved with surgery from a posterior approach. A common scenario involves a patient in the seventh or eighth decade with a thick, buckling ligamentum flavum causing multisegmental spinal cord compression. Occasionally, such a patient is initially seen after a fall with an acute neurological deficit (Fig. 279-1).

FIGURE 279-1 Multilevel cervical spondylosis with an acute neurological injury after minor trauma.

Pathology that extends over many cervical segments, particularly if it crosses into the thoracic spine, is often treated preferably via a posterior approach. Degenerative spinal disease superimposed on a congenitally narrow canal (<12 mm, identified on a lateral radiograph by near-superimposition of the posterior cortex of the lateral masses and the spinolaminar line) is one such situation. Another condition that often involves multiple segments at initial evaluation is ossification of the posterior longitudinal ligament (OPLL) (Fig. 279-2). The typically extensive, confluent disease seen with OPLL ossification is often optimally approached posteriorly, particularly because the ventral dura is frequently annealed to the ossified ligament and a spinal fluid leak is unavoidable if an anterior approach is used.

FIGURE 279-2 Midline sagittal magnetic resonance image (A) and computed tomography scan (B) of a patient with symptomatic multilevel, confluent ossification of the posterior longitudinal ligament. Such patients are often best treated with a posterior approach.

The alignment of the cervical spine can be an important factor in determining the surgical approach. In general, a posterior approach is best suited for treating patients with straight or lordotic alignment. With all other factors constant, dorsal decompression (laminectomy) is less effective in relieving compression in a kyphotic spine than in a lordotic spine. Another consideration is the need to correct the patient’s alignment in the sagittal plane. A posterior-only approach may not be the optimal method for correcting alignment of the cervical spine. The methods of posterior fixation offer limited resistance to pullout and few salvage options. Anterior interbody grafting may provide a superior option for the restoration of segmental lordosis.

Indications for Surgical Treatment

Surgical treatment of cervical degenerative disorders, such as CSM or radiculopathy, may be indicated in patients with progressive signs of myelopathy, progressive or severe weakness in a cervical myotome, or intractable radicular pain and correlative imaging. The natural history of CSM is variable and unpredictable for any particular patient.³ Recommendations may be tailored to the patient based on age, symptoms, and findings on electromyography (EMG) or radiography, among other factors.⁴ Once progression has been demonstrated, however, it is unlikely that the process will halt completely or reverse spontaneously.⁵ In these circumstances, a frank, clear discussion should be held with the patient and family regarding the potential risks associated with surgical intervention versus the risks associated with continued observation.

Anesthesia and Positioning

General anesthesia is used for all posterior cervical surgery at our institution. If two 20-gauge or larger peripheral intravenous lines are placed, central venous access is usually unnecessary. An arterial line is established to allow close monitoring of arterial blood pressure. A Foley catheter is inserted, and sequential compression devices are placed on the legs.

If Lhermitte’s sign or other symptoms develop with cervical extension, awake fiberoptic intubation is performed. Conversely, if the patient is able to achieve adequate extension without symptoms, intubation with direct larygoscopy is preferred. Meticulous attention is paid to systolic blood pressure during induction and until the spinal cord is decompressed. In the absence of contraindications, the anesthesiologist is instructed to maintain systolic blood pressure at greater than 120 mm Hg to optimize spinal cord perfusion.

We use the prone position exclusively for posterior cervical surgery. Based on surgeon preference, either an electric operating table with padded bolsters and a rigid Mayfield head holder or a Jackson spine table with Gardner-Wells traction is used. With the Jackson table, a bivector traction setup with two ropes is used to permit the patient to be positioned initially in relative cervical flexion to facilitate the decompressive portion of the procedure and placement of the fixation points. The weight is moved to the upper rope to enhance cervical lordosis before placement of the rods and performing the arthrodesis (Fig. 279-3). A horseshoe head holder is placed 2 to 3 cm away from the patient’s face to serve as a backup should a problem occur with the traction.

FIGURE 279-3 Bivector traction is used to facilitate the decompressive portion of the procedure and allow the establishment of normal alignment for instrumentation. A, During the exposure and decompression, the weight (15 lb) is on the lower rope. B, The weight is moved to the upper rope prior to final placement of instrumentation to restore cervical lordosis.

Some surgeons prefer a sitting position for posterior cervical foraminotomies, particularly when the procedure is performed with a minimally invasive technique. Advantages of this position include improved fluoroscopic visualization of the cervicothoracic junction and a nearly bloodless field. Disadvantages include the risk for air embolism and, possibly, a somewhat longer setup time. If a sitting position is used, a precordial Doppler probe and an end-tidal CO₂ monitor are recommended to facilitate early detection of an air embolus.

General anesthesia is maintained during the operation with a total intravenous anesthesia technique that allows optimal monitoring of somatosensory evoked potentials (SSEPs) and motor evoked potentials (MEPs). Most often this means that the use of volatile inhaled anesthetics is avoided or minimized and general anesthesia is maintained with propofol (Diprivan), fentanyl, or sufentanil and nitrous oxide.

Spinal Cord Monitoring

SSEPs and MEPs are monitored routinely in patients with myelopathy undergoing a posterior procedure. Patients with purely radicular symptoms may be monitored depending on the specific pathology and neurological status of the patient, among other factors. In patients with severe compression, radiographic instability, or a rapidly progressive clinical course, we generally record baseline SSEPs after intubation before turning the patient to the prone position. Prepositioning baseline MEPs are not obtained because of the residual effect of the muscle relaxant given to facilitate intubation.

Although we routinely monitor SSEPs and MEPs in these patients, the utility of this strategy and the appropriate response to changes in signal quality, amplitude, or latency in different clinical scenarios are sometimes unclear. We have, on occasion, identified problems with positioning of the arms and shoulders based on monitoring changes. During decompression of a stenotic foramen it is not unusual to see transient nerve root irritation on free-running EMG that has no apparent clinical postoperative correlation. An evidence-based review found conflicting data regarding the use of intraoperative improvements in electrophysiologic data for clinical prognostication.⁶

The surgeon must, for each individual situation, determine the response to the decrement or loss of SSEP or MEP signals during a posterior cervical procedure. The anesthesiologist and neurophysiologist are queried regarding any recent changes in anesthetic technique and the nature of the identified changes. Initial steps may include increasing blood pressure, inspecting the surgical field for any compression of the spinal cord, and reversing any changes in spinal alignment. These and any further steps are taken at the surgeon’s discretion.

Exposure

The surgeon should maintain meticulous hemostasis during surgical exposure of the posterior cervical spine. This is facilitated by careful surgical technique. After the skin incision is made and the subcutaneous tissue is divided to the cervicodorsal fascia, spending a few moments to identify the avascular plane between the paracervical muscles will pay dividends in achieving a quicker, drier approach. One technique to facilitate identification of this plane is to use the cutting rather than the cautery setting of the monopolar cautery; it will be readily apparent if the dissection proceeds out of plane and into the adjacent muscle. This avascular plane extends ventrally to the spinous processes.

The posterior bony elements are exposed subperiosteally with monopolar cautery. The exposure extends to the junction of the lamina and the lateral masses. The lateral masses and facet joints are not exposed only if a decompressive procedure is to be performed. If motion-preserving decompression with intrasegmental implants is planned, such as a laminaplasty with plate reconstruction, the lateral masses are exposed while taking care to preserve the facet capsules. When a laminectomy with arthrodesis is planned, the lateral masses and facet joints are exposed and the articular surfaces removed at all levels of the planned fusion. Regardless of the procedure, it is essential to try to preserve the dorsal midline soft tissue structures connecting the vertebrae at the ends of the operative segments to the adjacent nonoperative segments, particularly the interspinous and supraspinous ligaments. This posterior tension band is important for maintenance of alignment of the transitional segments from the operated to unoperated regions.

Minimally invasive techniques involving sequential muscle dilation and tubular retractors may be used for unilateral one- and two-level foraminotomies. A longitudinal paramedian incision is planned under fluoroscopic guidance. The incision is made and continued ventrally through the fascia. Under fluoroscopic control, the localizing needle and sequentially larger muscle dilators are introduced. The ideal exposure is centered over the intervertebral space with the medial half of the lateral mass in the lateral half of the operative field. A microscope or endoscope is used for illumination and magnification.

Foraminotomy/Diskectomy

The posterior keyhole foraminotomy as first described by Scoville has a role in the treatment of isolated foraminal stenosis or lateral disk herniation.⁷ Patients with symptoms referable to a single cervical nerve root are the best candidates for this procedure. Bone removal is tailored to the nature of the pathology. In patients with spondylosis and osteophytic foraminal stenosis, a keyhole-shaped decompression extending from the lateral thecal sac to the lateral border of the pedicle is performed to achieve complete decompression of the exiting nerve root. For the removal of lateral soft herniated disks, a smaller amount of bone can be removed to preserve the majority of the facet joint.

A herniated cervical disk impinges on the exiting nerve root that corresponds to the pedicle caudal to the disk space (e.g., a C5-6 herniated disk causes C6 impingement). The foraminotomy is performed at the appropriate interspace just proximal to the pedicle of the level caudal to the disk herniation (Fig. 279-4). The first part of the foraminotomy procedure consists of a small laminotomy and flavectomy to expose the lateral dura and the origin of the exiting nerve root. The foraminotomy is extended laterally by using a high-speed drill to thin the ventromedial superior facet. Bone removal is completed with small straight and up-going curets. A minimal amount of the inferior facet is removed. After the axilla of the exiting nerve root is identified and a portion of the root approximately 2 to 3 mm in length is exposed, a micro–nerve hook is used to dissect parallel with the nerve root along its caudal edge. With removal of less than half the facet joint, up to 5 mm of the nerve root may be exposed.⁸