Anterior cervical discectomy and fusion (ACDF) is a common and accepted treatment for the patient with cervical disc herniation with radiculopathy. Successful results have been reported since the original descriptions of the operative techniques of Smith and Robinson¹ and Bailey and Badgley² and the refined techniques and instruments developed by Cloward.^3,4 Although single-level ACDF has a generally high fusion rate by several techniques, argument continues as to the added benefit of plating for treatment of single-level disease. Proponents of plating cite reduction in interbody graft migration, diminished rates of pseudarthrosis, improved immediate postoperative segmental rigidity, reduction in kyphotic angulation, and the lack of need for cervical orthosis.

Alternatives to Ventral Cervical Plating

Anchored Interbody Spacers

Several recent devices have come to market that share the properties of being a low-profile interbody spacer that allows for fixation to the adjacent vertebral bodies without a formal anterior cervical plate (Fig. 232-1). The polyetheretherketone (PEEK) Prevail cervical interbody device (Medtronic Sofamor Danek, Memphis, TN) is an interbody device with rostral and caudal midline flanges that allow for a screw interface with the ventral corner of the adjacent vertebral body. The device utilizes nitinol wires as a locking mechanism to prevent screw backout.

FIGURE 232-1 Low-profile cervical interbody devices. The polyetheretherketone Prevail device (left) and Zero-P (right) both utilize angled screws to anchor the interbody spacer to the adjacent vertebral bodies without the addition of a traditional ventral cervical plate.

The Zero-P device (Synthes, West Chester, PA) has similar properties but does not require preparation of the anterior surface of the vertebral body. This “anchored spacer” consists of a polyetheretherketone interbody cage with a ventral titanium housing that allows for anchoring screws to be placed diagonally into the adjacent vertebral bodies. Scholz et al.¹⁶ recently examined the biomechanical properties of this stand-alone cervical interbody device. This device demonstrated decreased range of motion compared to interbody cage alone (no fixation) and no significant difference when compared to the addition of a locking or dynamic anterior cervical plate.

Cervical Total Disc Arthroplasty

In the absence of significant coexisting cervical disc disease at other levels, total disc arthroplasty is another option for treatment of cervical disc herniation with radiculopathy. Mummaneni et al.¹⁷ recently presented the 3- to 5-year results of a prospective randomized trial of ACDFP versus arthroplasty with the Prestige device (Medtronic Sofamor Danek). Of 541 patients included in the study, 276 underwent discectomy and disc arthroplasty. Neck disability index and neck pain scores were significantly better in the arthroplasty group at 3 years but were similar to the ACDFP group at 5 years. The arthroplasty devices maintained a mean of 7.1 degrees of flexion-extension motion, and the arthroplasty group had fewer secondary surgeries than did the ACDFP group.

Anterior cervical discectomy with interbody fusion (ACDF) is a very common surgical procedure that can be used to treat multiple cervical pathologies. ACDF surgery is commonly used to treat radiculopathy and myelopathy due to degenerative cervical disease. ACDF was initially done without instrumentation and demonstrated very good results in treating radiculopathy and myelopathy.¹ However, pseudarthrosis and graft extrusion are known complications of ACDF.² To address these complications, plating was added to ACDF, as first described by Bohler and Gaudermak in 1964.³

Description of Procedure

The following is a description of our technique for ACDF.

The patient is placed supine on the operating room table. Once the anesthesiologist’s endotracheal tube and lines and are appropriately secured, the patient’s head should be placed in a neutral position or in slight extension. This can be done with a small shoulder roll or by dropping the head of the table. The head can be placed on a donut to secure it in position or in traction if desired. The shoulders should be pulled down toward the feet either with tape or with wrist straps if necessary to visualize the lower cervical levels. To avoid injury to the brachial plexus, care should be taken not to pull too tightly. Electrophysiologic monitoring can be used to minimize the risk for iatrogenic injury due to positioning, decompression, and instrumenting as desired by the primary surgeon.⁴

External landmarks can be used to estimate cervical spine levels, which should then be confirmed with fluoroscopy. The hyoid bone is close to the C2-3 interspace. The top of the thyroid cartilage is close to C3-4, while the inferior border can be estimated as C4-5. The cricoid ring is roughly at the C5-6 level. One finger breadth above the clavicle is estimated as C7-T1.⁵

The neck is then prepped and draped in the usual sterile fashion. A transverse incision over the level of interest should be adequate for most surgical exposures, though if a particularly long ventral construct is being performed, then a longitudinal incision along the medial border of the sternocleidomastoid muscle provides greater exposure.⁶

The transverse incision should be just off midline and extend laterally to the sternocleidomastoid. The incision is normally performed on the left side to reduce the risk of injury to the recurrent laryngeal nerve, though it is unclear whether this actually reduces the risk.⁷ The skin is opened with a knife, exposing the platysma. The platysma is then divided and undermined. After the sternocleidomastoid (SCM) has been identified and the carotid has been palpated, a plane medial to SCM and carotid sheath and lateral to the trachea and esophagus should be identified. With the use of blunt dissection, this avascular plane is followed down to prevertebral fascia. The omohyoid muscle may obstruct this approach at the C5 level and can be divided without significant consequences (Fig. 232-2).

FIGURE 232-2 The corridor for anterior cervical discectomy and fusion approaches.

(Used with permission from Barrow Neurological Institute.)

Once the prevertebral fascia has been reached, a Kittner sponge is used to bluntly dissect the fascia free. At this point, the cervical vertebral bodies, disc spaces, and longus colli muscles can be visualized. A spinal needle is inserted into the disc space, and a fluororadiograph is obtained to confirm the appropriate level. After the appropriate level has been confirmed, the disc space is marked with coagulation. While the longus colli muscles are still intact, midline is marked with electrocautery on the bone.

By using cautery, the longus colli muscles are then detached medially to laterally, and the retractor system can be placed. Placement of the retraction blades on the longus colli muscles is important to reduce risk of injury to the underlying carotid and esophagus. One theory about injury to the recurrent laryngeal nerve is that it is a compressive injury of the endolaryngeal segment of the nerve between the retractor and the endotracheal tube rather than a direct dissection injury. Apfelbaum et al. reported that if the endotracheal tube cuff was deflated after the retractor blades were placed and was then reinflated, their rate of injury went from 6.4% to 1.7%.⁸

Once the retractor system is in place, distraction pins, if being used, can be placed into the vertebral body. Distraction pins should not be used on extremely osteoporotic bone because of the risk for fracturing when the distracting forces are applied. If distraction pins are not being used, gentle cervical traction can be applied. The cervical discectomy can then be started. This can be done either by using an operative microscope or with surgical loops, depending upon the surgeon’s preference. The annulus of the disc is entered sharply using a no. 11 blade and a small rectangle is cut out. By using a combination of instruments including curettes, pituitary, and Kerrison rongeurs, the disc material is removed down to the posterior longitudinal ligament. To assist with visualization, the ventral caudal lip of the rostral vertebral end plate can be removed with a Kerrison rongeur or use of the distraction pins. In performing the discectomy, care should be taken not to go too far laterally, as this puts the vertebral artery at risk.

We favor opening the posterior longitudinal ligament to provide a more complete decompression. This can be done with an up-angled curette. Then by using a 1- or 2-mm Kerrison rongeur, the rest of the posterior longitudinal ligament can be removed, as can any dorsal osteophyte complexes, to adequately decompress the cord. The neural foramen should also be decompressed as needed. Care should be taken in particularly stenotic patients, as the act of inserting a Kerrison rongeur can injure the cord. In severe stenosis, a high-speed drill can be used to drill down the dorsal osteophyte complex until it is eggshell thin. Then a curette can be used to fracture the remaining lip into the emptied disc space.

Once the decompression is complete, the vertebral end plates are prepared. All the soft tissue is removed, and the end plate is decorticated to help maximize the probability for fusion. The decortication can be done with a combination of high-speed drill and curette. Care is taken to avoid violating the end plate, as this increases the risk of the graft pistoning into the vertebral body, which decreases the strength of the construct.

The graft is then inserted into the disc space. Autograft or allograft can be used. Autograft has a slightly higher fusion rate, quoted as being between 83% and 97%, but there is the risk for donor site complications, and the allograft fusion rate of 82% to 94% is nearly as good.⁹ The graft is slightly oversized and the vertebral bodies are slightly distracted so that the graft is under compression, which helps with fusion. Distracting the vertebral bodies can be done with distractor pins, gentle cervical traction, or disc spacers. Care is taken not to oversize the graft so that the vertebral bodies are not overdistracted. Once the graft is in place, the distraction force is removed, and a radiograph is taken to evaluate alignment and graft position.

When the graft is in place, an appropriately sized plate is chosen. The shortest plate that spans the graft is used because a longer plate increases the risk for adjacent-segment disease, owing to violation of the vertebral disc interspace at the adjacent levels. The cervical plate should be contoured to fit flush against the vertebral body. Any prominent osteophytes that prevent the plate from lying flush should be removed. Care should be taken to avoid violating the cortical bone in removing the osteophyte complexes because that will reduce the strength of the instrumentation.

The plate should be placed in the midline—straight up and down. Unicortical screws are often used. The screws should be toed in, or directed medially in the horizontal direction, and toed out, or directed away from the graft, in the vertical direction to help reduce pull-out. The screws should not violate the end plate, as this will predispose the patient to adjacent-segment disease (Fig. 232-3).

FIGURE 232-3 Anteroposterior (A) and lateral (B) views of instrumented anterior cervical discectomy and fusion.

(From Rhee JM, Riew KD: Dynamic anterior cervical plates. J Am Acad Orthop Surg 15:640–646, 2007.)

Once the plate has been secured, a final fluoroscopy image can be obtained to confirm placement. The wound is then copiously irrigated. After hemostasis, if there is any concern regarding bleeding, a small drain can be left behind. The platysma and dermis should then be closed. Dermabond or an absorbable suture is used on the skin. A hard Miami J collar may be used, but often this is not required.

Dynamic versus Fixed

During the last 20 years, there have been many developments in the design of the plating systems. The different plating systems have different retractors and distractors but, most important, different screw plate interfaces. The major choice in selection of the plate system is whether a dynamic or fixed system is chosen. In a fixed system, the screw-plate interface is constrained. This provides for a more stable construct, but it also stress-shields the graft and diminishes load sharing between plate and bone graft. Grafts are best incorporated into constructs when they are subjected to compressive forces. This can be achieved by oversizing the graft and distracting the vertebral bodies, but a fixed anterior cervical plate will still shield the graft from the normal daily compressive forces that help bones to remodel (Fig. 232-4).

FIGURE 232-4 Biomechanical differences of force vectors in the uninstrumented (A) and instrumented (B).

(From Rhee JM, Riew KD: Dynamic anterior cervical plates. J Am Acad Orthop Surg 15:640–646, 2007.)

Dynamic plating systems allow movement at the screw-plate interface. This may decrease the rigidity of the construct, but it allows compressive forces to run through the graft. In theory, this should help to promote fusion. However, several studies comparing the fusion rate between dynamic and fixed plate systems show similar fusion rates. One study demonstrated slightly better outcomes from a pain perspective with the dynamic plates,¹³ but in general, the few studies that have been completed showed minimal difference between fixed and dynamic plates.¹⁴

Over the past few decades, ventral cervical spine decompression and fusion surgery using the Smith-Robinson technique¹ has become commonplace. Anterior cervical discectomy allows direct decompression of the spinal canal and neural foramen in the treatment of myelopathy and cervical radiculopathy.^1,2 The question arises, “What should be placed in the disc space after total discectomy?”

Currently, most surgeons prefer to perform anterior cervical discectomy with fusion (ACDF) using an interbody graft supplemented with an anterior ventral plate to restore disc height, provide immediate structural stability, and maintain cervical lordosis. ACDF produces highly predictable, favorable, and reproducible results.^3–5 However, limitation of the cervical range of motion and the potential of accelerating adjacent-segment disease (ASD) are drawbacks with ACDF. Biomechanical⁶ and clinical^7–10 data both demonstrate limited range of motion and probable progression of ASD in the cervical spine after arthrodesis. Hilibrand et al.⁸ demonstrated an ASD rate of 2.9%/year following ACDF. Other studies by Katsuura et al.⁹ and Goffin et al.⁷ reported cumulative ASD rates as high as 30% to 50% over several years.

The advantage of disc arthroplasty is the ability to preserve motion while maintaining stability.¹¹ Theoretically, the maintenance of motion after discectomy results in lower stresses on the adjacent disc space. Two-year data available from Food and Drug Administration (FDA) trials shows that artificial disc spacers maintain normal segmental motion at the index surgical level and have a lower reoperation rate than ACDF.^12–14 Further studies are pending regarding the long-term benefits of disc arthroplasty as it relates to ASD.

Preoperative Planning

Before considering arthroplasty, the surgeon should review all appropriate imaging including anteroposterior (AP), lateral, flexion-extension radiographs, MRI, and/or CT of the cervical spine (Figs. 232-5A and B). There are several arthroplasty devices to choose from in the United States (three are currently FDA approved for single-level placement).^12–14 The Prestige ST (Medtronic, Memphis, TN) incorporates a stainless steel ball-and-trough design. Vertebral body screws are used to fix the device’s position in the interspace. The device is considered semiconstrained with regard to motion (Fig. 232-6). The ProDisc-C (Synthes Spine, Paoli, PA) device has two cobalt chrome alloy end plates and an ultra-high-molecular-weight polyethylene inlay in a ball-and-socket configuration. Instead of vertebral screws, it uses a central keel for primary fixation. This device is also semiconstrained (Fig. 232-7). The third device approved by the FDA is the Bryan Cervical Disc (Medtronic). This artificial disc spacer consists of two titanium alloy shells articulating with a polyurethane core. It is held in the disc space with only a tight fit. Bone ingrowth eventually bonds the metallic device faces to the vertebrae. The device is considered unconstrained¹⁶ (Fig. 232-8).

FIGURE 232-5 Sagittal (A) and axial (B) T2-weighted MRI views in a 40-year-old woman with neck pain and left arm radicular pain. These scans show a central disc herniation with left foraminal stenosis at C5-6. The facet joints appear to be well preserved, and there was no listhesis on dynamic radiographs (not shown). Postoperative flexion (C) and extension (D) radiographs following C5-6 discectomy and placement of the Prestige ST artificial disc spacer. There is preservation of normal motion. Notice that the device spans almost the entire interbody space. The patient reported resolution of her preoperative neck and arm pain.

FIGURE 232-6 The Prestige ST artificial disc spacer.

(Courtesy of Medtronic Sofamor Danek, Memphis, TN)

FIGURE 232-7 The ProDisc-C artificial disc spacer.

(Courtesy of Synthes Spine, Paoli, PA).

FIGURE 232-8 The Bryan Cervical Disc artificial disc spacer.

(Courtesy of Medtronic, Memphis, TN; the Bryan Cervical Disc System incorporates technology developed by Gary K. Michelson, MD.)

Patient Counseling

Artificial disc spacers are relatively recent additions to the spine surgeon’s armamentarium. Patients should be counseled to understand that these are new devices with limited data regarding long-term (decades) use in the United States. We typically discuss alternative procedures such as ACDF and posterior foraminotomy with the patient.

The risks of implant failure, new or residual radiculopathy, migration, subsidence, and reoperation should also be discussed. Many devices contain stainless steel, which can cause significant artifact on MRI scans. Patients should be aware of the possible need to undergo a more invasive CT myelogram instead of MRI for workup of new neck problems. Second-generation artificial discs that contain components that are MRI-compatible are currently undergoing FDA IDE clinical trials

The treatment of cervical disc herniation with resultant radiculopathy has evolved over the past half century. Once managed through a primarily dorsal approach, anterior cervical discectomy (ACD) has become the favored approach for the treatment of this disease process and is associated with excellent symptom resolution and a relatively low rate of complications. In fact, ACD for nerve root decompression is one of the most common procedures performed by neurosurgeons today.

Since its introduction, ACD has been described in conjunction with placement of an interbody graft for fusion.^1–3 Originally, this graft was harvested from the patient’s iliac crest; hence, this additional surgical step contributed its own set of risks and complications. Years later, surgeons began advocating discectomy without fusion for the treatment of this disease process. The results were fairly good, and complications associated with graft harvest and placement were avoided.^4,5 With the introduction of various interbody graft materials and the improvement in the rate of fusion associated with these materials, many neurosurgeons now routinely perform fusion procedures. However, conclusive data are lacking to support the superiority of interbody fusion compared to a simple discectomy for the treatment of radiculopathy related to a cervical disc herniation. We therefore review the existing data in the hope of clarifying this controversial issue.

ACD with or without fusion employs the approach described by Cloward, Robinson et al., and Smith and Robinson.^1–3 The discectomy is performed under magnification, with resection of the posterior longitudinal ligament and extruded disc fragments. Foraminotomies are performed to fully decompress the nerve roots. Finally, in the fusion procedure, an interbody graft is inserted, and its position is confirmed under fluoroscopy. A cervical plate spanning the interspace may then be placed.

In 1983, Rosenørn et al. published a prospective randomized study evaluating ACD with and without fusion.⁶ Sixty-three patients with symptomatic cervical disc herniations were randomized either to discectomy alone (DE, 32 patients) or to discectomy and fusion using freeze-dried bone grafts (DEF, 31 patients). Clinical outcome was graded as “excellent” if patients returned to work without symptoms, “good” if they returned to work with minor symptoms, “fair” if they returned to work with the same symptoms or had to change occupations, and “poor” if they did not return to work and their status was unchanged or worse. At both 3-month and 12-month follow-up examinations, the clinical outcomes for DE were significantly better than those associated with DEF. However, if the “excellent” and “good” results were grouped together, the difference in outcomes between DE and DEF was insignificant at 12 months. There were no significant differences in length of hospital stay between the two groups, and there was no difference in the complication rate.

Another prospective randomized study presented in 1998 by Savolainen et al. compared three different surgical techniques for single-level cervical root compression: (1) discectomy alone, (2) discectomy and fusion with autologous bone graft, and (3) discectomy and fusion with bone graft plus plating.⁷ After 4 years of follow-up, there were no significant differences between the three groups in the achievement of good clinical outcomes (i.e., no symptoms): 76% for discectomy, 82% for discectomy + graft, and 73% for discectomy + graft/plate. There were no poor outcomes in the discectomy group compared to 4% poor outcomes in both of the other groups (results not significant). All patients who underwent grafting, plating, or both grafting and plating achieved a solid fusion, with 90% of all simple discectomy patients achieving solid fusion as well. Finally, there were no significant differences in the development of kyphosis among the three groups (63% for discectomy, 41% discectomy + graft, 44% discectomy + graft/plate). The authors therefore argued that interbody fusion seemed unnecessary for single-level cervical nerve root compression.

In 1999, Dowd and Wirth reported their experience after enrolling 84 patients in a prospective randomized trial comparing ACD with ACD and fusion (ACDF).⁸ Patients who underwent placement of iliac crest bone graft were most likely to develop fusion (97% vs. 70%). At final follow-up (average 4.5 years), however, there were no differences between the two groups in terms of patient satisfaction or return to preoperative activity.

Proponents of fusion propose that placement of an interbody graft after discectomy prevents collapse of the disc space, segmental kyphosis, and the resultant neck pain and avoids the consequent narrowing of the neural foramina and nerve root impingement. To some degree, Murphy et al. supported this proposal with radiologic studies.⁹ They evaluated the foraminal area before and after ACD alone or after discectomy with interbody graft placement and fusion. As expected, the foraminal area decreased significantly after discectomy alone (P < .001) and increased after placement of an interbody graft. However, when the authors compared absolute changes in disc height between the two groups, the difference was not significant. Most important, they found no significant correlation between foraminal area and clinical outcome.

Oktenoglu et al. compared disc height and foraminal area at index and adjacent levels and evaluated symptom relief of patients undergoing ACD or ACDF with plating.¹⁰ Patients were examined preoperatively and in the immediate and 1-year postoperative periods. There were no significant differences between the two groups in terms of disc height and foraminal areas at adjacent levels. At 1 year, both groups showed decreases in disc height at the index level compared to preoperative values, but the ACDF with plating group maintained greater disc height (P = .007). Both groups showed a decrease in foraminal area, with no significant difference between ACD and ACDF with plating (P = .603). Importantly, pain scores for arm pain decreased similarly and significantly for all patients after surgery, regardless of type of surgery.

White and Fitzgerald tested the hypothesis that the degree of disc space settling and kyphosis after discectomy directly correlates with preoperative disc height, greater degrees of angulation and kyphosis being noted after resection of normal discs and with only minimal changes after resection of collapsed, degenerated discs.¹¹ Furthermore, they projected that with greater degrees of settling, angulation, and foraminal closure, the clinical outcomes after discectomy alone in patients with normal disc height would be worse than outcomes in patients with restored disc height after interbody graft placement. The authors evaluated preoperative and postoperative kyphosis in both groups (discectomy alone and discectomy plus fusion). Discectomy alone resulted in a greater degree of kyphosis if the preoperative disc space was more than 4 mm in height and resulted in only minimal change in sagittal alignment if the disc space was narrower than 4 mm. As clinical correlates to these findings, they noted an increase in adverse outcomes when small disc spaces were grafted and when larger disc spaces were left ungrafted. Of their 15 complications, 12 (80%) fell into one of these two categories. Therefore, they concluded that the decision regarding whether to graft should be based on preoperative evaluation of the height of the disc space on plain lateral radiographs, and they recommended placing a graft when disc height is greater than or equal to 4 mm.

Xie and Hurlbert also evaluated segmental kyphosis after ACD compared with after ACDF or ACDF with plating.¹² In their prospective randomized study of 42 consecutive patients with cervical radiculopathy, segmental kyphosis was noted in 75% of ACD patients after surgery compared with 17% before surgery. In contrast, there was no change in sagittal balance in the ACDF or ACDF with plating groups. Nonetheless, after 2 years of follow-up, there were no differences with respect to arm pain, neck pain, or interscapular pain among the three groups of patients (P > .05).

With the advent of synthetic interbody graft materials, the obvious negative factors associated with graft harvest (e.g., pain, potential graft site complications) could be avoided. Therefore, despite the lack of class I evidence to confirm a clinical benefit, the technique of fusion after discectomy continued to gain widespread acceptance. In 2008, Hauerberg et al. published their results evaluating ACD versus fusion with a titanium cage.¹³ They randomized 86 patients with cervical radiculopathy to undergo discectomy alone (46 patients) or to undergo ACDF with a Ray titanium cage (40 patients). Neither group was placed in an external orthosis postoperatively. After 2 years of follow-up, 86% of the patients who underwent titanium cage fusion reported a “good” clinical outcome compared to 77% of the discectomy alone group. Consistent with previous studies, this difference was not statistically significant. The authors also found no differences in self-reported satisfaction, arm pain, or neck pain between the two groups 3 months or 24 months after surgery. The rate of fusion was similar between the two groups: 81% in the discectomy alone group versus 83% in the titanium cage group.

In 2005, White and Fitzgerald provided a different angle from which to view this controversy.¹¹ Perhaps the real challenge is to develop an algorithm that predicts who may or may not benefit from a fusion procedure. Several factors could be included in such an algorithm, such as preoperative sagittal alignment, disc space height, single-level or multilevel disease, bone condition, and presence of concomitant neck pain¹⁴ (Fig. 232-9).

FIGURE 232-9 Algorithm for the treatment of radiculopathy. ACD, anterior cervical discectomy; ACDF, anterior cervical discectomy with fusion (bone graft); ACDFP, anterior cervical discectomy with fusion and plating; PCD, posterior cervical decompression. ^*Choice of procedure depends on the surgeon’s and patient’s preference.

(Used with permission from Theodore N, Sonntag VKH: Decision making in degenerative cervical spine surgery. Clin Neurosurg 48:260–276, 2001.)

In Sonntag’s experience, patients with single-level or two-level disease, strictly radicular symptoms, and no neck pain have done well after discectomy alone¹⁵ (Figs. 232-10 to 232-12). However, in patients with multilevel disease, loss of cervical curvature, or any evidence of instability, Sonntag performs discectomy followed by fusion. The presence of axial neck pain is also a strong indication for fusion, which provides excellent relief of symptoms. This trend was noted by Oktenoglu et al., who found that neck pain scores improved significantly only in patients who underwent interbody fusion (P = .008).¹⁰

FIGURE 232-10 Axial T2-weighted MRI shows stenosis of the spinal canal and left foramina at C5-6. This patient presented with left C6 radiculopathy and underwent ACD at C5-6 without an interbody graft or plate.

(Used with permission from Barrow Neurological Institute.)

FIGURE 232-11 Sagittal T2-weighted MRI of the cervical spine shows a new disc herniation at C3-4 several months after the patient had undergone ACD at C5-6. The C6 radiculopathy had resolved, but the patient was experiencing new myelopathic symptoms.

(Used with permission from Barrow Neurological Institute.)

FIGURE 232-12 Sagittal T2-weighted MRI (A) and lateral view of cervical spine radiograph (B) after cervical discectomies at C3-4 and C5-6, both without placement of an interbody graft or plate. Note the solid fusion at both levels as well as the absence of a significant kyphotic deformity.

(Used with permission from Barrow Neurological Institute.)

Other indications for fusion include acute disc herniation with concomitant cervical vertebral body fracture, extensive vertebral body resection (large dorsal osteophytes), previous laminectomy, or when future laminectomy is anticipated.^15,16 As has been noted in several studies, postoperative transient interscapular and axial neck pain after ACD are common; however, without treatment, both usually resolve within a few weeks of surgery.

Finally, the economic costs of adding interbody fusion and instrumentation must be considered. In the absence of clear indications for fusion, such as instability, neck pain, or existing kyphotic angulation, simple discectomy can help to avoid the extra costs associated with a fusion procedure. These costs include the interbody graft, possible instrumentation, extra operative time, the orthosis, and intraoperative and follow-up radiographs to evaluate the position of the graft and fusion. Nonetheless, ACD with interbody fusion and plating has become almost universally accepted as the standard treatment for radiculopathy related to cervical disc herniation. It may take many more years of patient follow-up and evaluation to determine whether the costs associated with placing a graft and possible instrumentation are justified in regard to long-term outcomes and patient satisfaction.

Cervical disc herniation and degenerative spur are common causes of neurologic disability and pain syndrome in the neck and upper extremities. While a lesser number of patients develop spinal cord compression and myelopathy due to midline disc extrusion or spur formation, a far larger proportion develop nerve root compression and radiculopathy. Fortunately, the posterior longitudinal ligament acts as a barrier to direct dorsal displacement of the disc with compression of the spinal cord, favoring a lateral herniation subjacent to the nerve root.

Many patients with symptoms of radiculopathy will improve over 6 to 8 weeks with conservative management and observation. For those with significant residual symptoms, the surgical options fall generally into the following three categories:

The motives for selection of a ventral approach are as follows:

A variant of the ventral option is the more recent cervical disc arthroplasty. While it offers ventral access to a fragment or spur, in theory, the preservation of motion at the involved segment may be beneficial.

Dorsal laminoforaminotomy offers a number of significant advantages in the treatment of cervical radiculopathy; these are discussed next.^1,2

Although the exposure is dorsal to the nerve root, the expanse of nerve root decompression that can be achieved is extensive (Fig. 232-13). Even resection of less than 30% to 50% of the medial facet provides wide decompression of the nerve.² While I generally clear the fragment from ventral to the nerve root, there are abundant data to confirm that nerve root decompression alone is sufficient to treat the problem.² This is important in that in patients with spurs or fixed fragments, it is not necessary to be unduly aggressive with exploration and dissection ventral to the nerve root.

FIGURE 232-13 The subtotal laminar resection and approximately the medial one third of the facet afford thorough exposure of the nerve. The nerve origin is within 5 to 6 mm of the facet articular groove in most cases. Once the compression of the nerve has been thoroughly relieved by laminoforaminotomy, inspection ventral to the nerve can be undertaken if a soft disc is suspected.

Dorsal laminoforaminotomy is the “original” motion preservation technique. Although there have long been concerns that dorsal procedures, regardless of technique, impart some degree of spinal instability, there is substantial evidence to refute this claim. In vitro studies suggest that resection of less than 50% of the facet does not lead to clinically significant instability or progression of kyphosis.³ Significantly, a recent series of 162 cases treated with laminoforaminotomy failed to show a significant increase in kyphosis, even when 50% or more of a unilateral facet was resected.¹ Clarke et al. reviewed 303 patients with single-level dorsal foraminotomy, noting same-segment disease in only 3.2% of patients at 5 years and in 5% at 10 years.⁴ Although the ventral surgery proponents often portray the disc fusion option as less intrusive on spinal dynamics, in fact, the hypermotility and adjacent-segment disease after ventral cervical disc and fusion may reach a rate of 2.9% per year, with recurrent surgical procedures conducted in 7.0% of patients within 10 years of ventral cervical disc fusion procedures.⁵ Jagannathan et al. and Henderson et al. both had recurrent surgical procedure requirements of 5.0% or less at 10 years using a standard dorsal laminoforaminotomy approach.^1,2

The inherent risk ratio in dorsal laminoforaminotomy is competitive with all the ventral approaches. Since no fusion or arthrodesis is required, no risk of pseudarthrosis exists. Since no fusion is included in the procedure, no hypermotility alteration at adjacent segments occurs. The anatomic pathway to the target nerve does not involve dissection of any critical structures, in contrast to the small but real potential for injury to the recurrent laryngeal nerve, esophagus, or carotid and vertebral arteries that can occur with ventral exposures.

In light of current medical trends, cost analysis is a factor. No instrumentation or hardware is inserted with the dorsal approaches. Accordingly, the cost of implantable devices is zero. In contrast to disc arthroplasty, concerns about long-term device integrity are not an issue.

Technique

The procedure can be safely accomplished in either the prone or sitting position.^1,2 While more extensive monitoring is required in a sitting craniotomy procedure, the data suggest that in reasonably healthy patients, no Foley catheter, arterial line, or central venous catheter is needed in the sitting cervical laminoforaminotomy procedure.² I do not employ neurologic monitoring in cases of cervical radiculopathy.

In the initial exposure, a unilateral subperiosteal dissection is achieved through a moderate midline incision (Fig. 232-14). Minimal access techniques can be used to achieve similar exposure.⁶ It is important to limit the dissection to just lateral to the laminar facet groove. This prevents unnecessary disruption of the facet capsule.

FIGURE 232-14 The soft tissue of the facet capsule is elevated only at the medial 30% to 50% of the facet width. This will protect the integrity of the facet structure. Begin resection of the inferior laminar lip with a 2-mm cervical thin foot plate Kerrison rongeur or a drill. Generous resection of the laminar lip as well as the superior rim of the inferior lamina (area in green circle) allows efferent exposure of the dura and the proximal nerve root at its origin from the thecal sac.

The nerve will exit within 5 to 7 mm of the articular groove of the facet; accordingly, the target zone for the laminoforaminotomy will be the laminar facet groove (for medial-lateral planning) and the facet articular groove (for superior-inferior planning). It is preferable to begin the laminotomy somewhat medial to the laminar facet groove, as this will allow exposure of the lateral thecal sac (see Fig. 232-13). Once the thecal sac has been identified with a removal of the medial facet (usually less than 30% of the facet), the origin of the nerve root will become evident. (see Fig. 232-13). It is now necessary only to extend the foraminotomy directly dorsal to the nerve root to achieve full and effective decompression. Beginning the dissection medially allows a carefully targeted and minimized facet resection by removing only the portion of facet that is directly dorsal to the nerve.

If a soft disc fragment is lodged ventral to the nerve, the nerve can be mobilized in a superior direction under magnified vision. Occasionally, the fragment will be around the shoulder of the nerve root. A shallow incision in the fragment in the direction of the nerve can be made to release the fragment. A small but important 4% to 6% subgroup may have a conjoint nerve root origin, with the motor root in the ventral-inferior position. The conjoint root, when stretched over a disc fragment, can be subtle to visualize. Being certain of the full movement of the edge of the nerve root dura, magnification, and attention to this anatomic variant will reduce the risk of nerve injury. (Fig. 232-15).

FIGURE 232-15 In most cases, the soft fragments are located in the axilla of the nerve. Be certain that the caudal edge of the dura has been identified, and gently elevate the nerve in a superior direction with a microinstrument. Incise the fragment surface in the direction of the nerve (blue line). Only the fragment need be removed. I perform only the foraminotomy for hard spurs (see text).