Combined Anterior-Posterior Decompression and Fusion for Cervical Spondylotic Myelopathy

Discussion

Cervical myelopathy is a common yet frequently underdiagnosed disease, owing to the fact that many patients remain asymptomatic even after experiencing degenerative changes in the spinal column.^14-16 The additive effects of spondylosis, osteophyte formation, ligamentous hypertrophy, and listhesis lead to progressive canal and intervertebral foraminal compromise, ultimately producing the clinical syndromes of myelopathy and radiculopathy.¹⁷ The characteristic symptoms of CSM are known to have an insidious onset. In the early stages, patients note a subtle gait disturbance and later experience manual dexterity reductions and upper extremity dysesthesias.¹⁸ As the condition progresses and conservative management fails, surgical intervention is sought.

Nevertheless, the pursuit of surgical treatment for CSM remains somewhat controversial. Some authors have found no statistically significant difference between conservative and surgical management of mild to moderate CSM,¹⁹ whereas others have found that surgically treated patients had much better outcomes than their medically treated counterparts.²⁰ In 2010, Scardino and colleagues²¹ reported that CSM patients who were bedridden and/or wheelchair-bound with seemingly irreversible myelopathy were capable of neurologic improvement after surgical intervention. At the very least, what remains clear is that untreated CSM is known to follow an unpredictable course, with the condition deteriorating faster for some patients than others.²²Traditional anterior or posterior approaches, which can be used in the majority of cases of cervical spondylosis and/or radiculopathy, have been compared extensively.^23,24 The inverse relationship concerning the integrity of an anterior construct and the number of levels fused is a well-established clinical finding.^3,4,8,25-28 Laminectomy with fusion is not without its disadvantages: Cervical instability secondary to mechanical loss of posterior cervical support, and subsequent post-laminectomy kyphosis, is a common complication.²³ In cases in which more stability is required, the combined anterior-posterior approach is more promising than either approach alone. This technique has its roots in the treatment of several thoracolumbar spine disorders, including infections, scoliosis, trauma, and tumors.^29-31 More recently, the technique has been applied to CS disorders.

In 2008, Gok and colleagues³² retrospectively compared the results of anterior-only fusion and CAPDF for CSM. Forty-six patients underwent anterior surgery only, and 21 underwent CAPDF. The groups’ complication rates were similar: 28.6% (anterior only) and 24% (CAPDF); the incidence of ASD was lower in the combined group. Song and colleagues³³ conducted a similar study in 2010. They compared anterior fusion alone and CAPDF in treating degenerative cervical kyphosis. Results were strongly in favor of the combined technique, as it led to “greater correction of sagittal alignment, a better maintenance of correction angle, a higher rate of fusion, a lower rate of subsidence and lower complications.” Both studies established that, in a select group of patients, the benefits of CAPDF outweighed the risks. These findings, combined with our study’s findings of no major complications and the transience of minor complications, suggest CAPDF should not be considered too invasive or risky.

The results of our study also mirror those of 3 other studies on the use of CAPDF for CS disorders. In 1995, McAfee and colleagues³⁴ reported on a group of 100 patients with follow-up of 2 years or more. In most cases, the surgical indication was trauma, but neoplasm, infection, rheumatoid arthritis, and CSM were found as well. Outcomes were very favorable: improvement in a previous neurologic deficit (57/75 patients), ability to walk again (21/35 patients), no new neurologic deficits, and no hardware failures. In 2000, Schultz and colleagues³⁵ retrospectively reviewed the cases of 72 patients who underwent CAPDF for a variety of complex CS disorders. Two of the 72 experienced transient neurologic deficits, and, though the immediate complication rate was relatively high (32%), the long-term complication rate was down to 5%. In 2009, Konya and colleagues³⁶ retrospectively reviewed the cases of 40 patients who underwent CAPDF, primarily for CSM. Within 1 week after surgery, neurologic deficits were reduced in 36 patients; by 1 year after surgery, neurologic deficits were reduced in all 40 patients, and fusion was achieved in 39. These 3 studies^34-36 helped establish CAPDF of the CS as a viable and effective procedure that can be performed within a single day.

Although many physicians have achieved favorable results with single-day surgery, the decision to operate in a sequential or staged manner remains controversial. Some anecdotally claim CAPDF poses a greater operative risk to the patient. In 1991, the continuous procedure was found to involve less blood loss and shorter LOS while providing for better correction of severe spinal deformity in patients with scoliosis and rigid kyphosis.³⁷ Three more recent comparative studies examining the same issue in the treatment of CS diseases found staging did not reduce the complication rate and may in fact have been associated with higher complication rates, more blood loss, and longer total operative time and LOS.^10,11,38 Our study’s lower blood loss, shorter LOS, and lower major complication rate relative to the combined groups in all 3 of those studies are most likely attributable to our operating on a lower mean number of spinal levels and our restricting the surgical indication to CSM. The positive short-term outcomes and low rate of long-term complications in our study, combined with the data from these 3 comparative studies, suggest that same-day surgery is superior to staged surgery. A staged operation should be considered only if the patient cannot tolerate long periods under general anesthesia.

Many have advocated extending fusion down to T1 to prevent ASD at the C7–T1 disk space.^35,39,40 We decided against this approach for 2 reasons. First, at C7, lateral mass screws were always chosen over pedicle screws. When possible, shorter lateral mass screws were used at this level, making C7 much less rigid. Second, the C7–T1 facet capsule was maintained to preserve joint integrity. We suggest extending fusion down to T1 only if there is prior evidence of spinal disease and/or listhesis at C7–T1. Although long-term (many-year) follow-up is often desired, we specifically assessed short-term (3-month) outcomes. We have anecdotally found that degree of improvement often follows a predictable course after 3-month follow-up. If myelopathy resolves even to a small extent during the first 3 postoperative months, later improvement will likely follow an upward course. Conversely, if myelopathy does not improve during the first 3 months, further improvement is much less likely.

This trend in neurologic improvement likely is directly related to degree of myelopathy before surgery. Patients with CSM generally experience symptoms over an extended period and try conservative management before any surgical consultation. Although spinal ischemia is often resolved by decompression, permanent ischemic damage to the cord is not uncommon. In this setting, postoperative neurologic improvement is minimal or even nonexistent, and decompression is preventive rather than curative. In our study, 1 patient had no subjective improvement after surgery. At 3-month follow-up, magnetic resonance imaging showed notable myelomalacia without residual spinal cord compression. We attribute the failure of the ischemic changes to resolve to long-standing preoperative damage to the cord. Nevertheless, surgery stabilized the myelopathy and prevented further ischemic damage and clinical deterioration.

As is the case with any operation, patients must be carefully selected for CAPDF. Indications for CAPDF, as described by Kim and Alexander,⁷ include acute spinal trauma, post-laminectomy kyphosis, kyphotic deformity with intact posterior tension band, multilevel spondylosis and OPLL, and preexisting risk factors for pseudarthrosis. Clearly, the severity of each varies, and the pathologies are not mutually exclusive. We emphasize that these indications provide only a guideline for performing CAPDF, and patients must be selected on a case-by-case basis. All the patients in our study were symptomatic and exhibited significant compression of the spinal cord anteriorly and posteriorly at multiple levels. Several presented with concomitant pathologies, such as cervical kyphotic deformity, congenital spinal stenosis, and OPLL. In each case, the indication for surgical intervention was undoubted. We sought both to improve the patient’s baseline symptoms and to prevent further damage to the spinal cord.

This study had its limitations. First, its retrospective design predisposed it to a higher degree of bias. Second, because CAPDF is not commonly performed, the sample size was relatively small. Third, although it provided a descriptive analysis of CAPDF for CSM, the study did not use a direct comparison group to establish whether treatment within a single day or staged treatment was more beneficial for our cohort in particular. On the basis of prior experience and observation, we think performing the operation within a single day is much more beneficial for the patient. Our discussion of studies that have compared same-day and staged surgery supports this observation. Therefore, staged treatment was not recommended to our patients.