Surgical Management of Cervical Ossification of the Posterior Longitudinal Ligament: Natural History and the Role of Surgical Decompression and Stabilization

Patrick A. Sugrue, M.D.; Jamal McClendon Jr., M.D.; Ryan J. Halpin, M.D.; John C. Liu, M.D.; Tyler R. Koski, M.D.; Aruna Ganju, M.D.

Disclosures

Neurosurg Focus. 2011;30(3):e3 

In This Article

Discussion

When evaluating patients with OPLL, preoperative planning is essential. Careful assessment of the imaging studies includes determining the extent of ossification and the direction of the surgical approach (anterior vs posterior). Diagnostic imaging in OPLL is usually multimodality. Typically, when patients present for neurosurgical evaluation, an MR imaging study has already been performed for evaluation of neck pain or arm pain. The appearance of OPLL on MR imaging differs depending on the extent of progression of the ossification process. For example, early OPLL is often confined to the area behind the disc space with some extension to the adjacent vertebral bodies or end plates; in such a case, OPLL appears very similar to spondylosis or disc herniation. Hypertrophied and calcified ligament appears hypointense on unenhanced MR images, but following the administration of gadolinium, ossified PLL enhances. Disc herniations do not enhance. In fact, particularly advanced cases of OPLL will display evidence of fat deposition and signs of bone marrow production. Advanced continuous OPLL may demonstrate the development of a Haversian canal system within the calcified ligament.[5] Wang et al.[34] studied T2 signal intensity ratios between intramedullary areas posterior to compressive ossified PLL compared with the intramedullary signal behind the C7–T1 disc space where there was no compression in patients with myelopathy prior to laminoplasty. They concluded that patients with low signal-intensity ratios had better surgical outcomes than those with high signal-intensity ratios and evidence of pyramidal signs.[34] Given the likely progression of untreated OPLL we advocate for early surgical intervention.

Computed tomography can help to confirm OPLL first suspected on MR imaging; CT can often help to better define the extent of disease seen on MR imaging and identify any foraminal component or the degree of stenosis. Likewise, OPLL in its early stages may be missed on CT sagittal reconstructions and these images should be correlated with axial imaging. Ossification of the ligamentum flavum can also be associated with OPLL and can be seen best on CT.[5] The longitudinal extent and circumferential location of both anterior and posterior ossification and subsequent canal or foraminal stenosis aid in formulating the appropriate surgical plan. CT myelography often helps to define the extent of spinal cord compression and the urgency of surgical intervention. Dynamic flexion-extension imaging can also play a role when the surgeon is considering stabilization. It is important to point out the importance of utilizing CT reconstructions in assessing OPLL to best understand the 3D anatomy of the disease. Chang et al.[2] have demonstrated the increased intra- and interobserver reliability of CT reconstructions compared with plain radiographs or even axial CT alone.

Patient age and medical comorbidities also influence the decision to use an anterior or posterior approach and whether to perform stabilization. Epstein[5] reports that in comparison to stand-alone posterior procedures, anterior procedures produce better short- and long-term improvement in neurological outcomes. The work of Kawano et al.[18] provides support for this notion, demonstrating improved outcomes in the short-term and long-term for patients undergoing anterior corpectomy compared with those undergoing posterior procedures. Anterior decompression of the spine seems to be the definitive surgical approach for a compressive lesion ventral to the spinal cord. However, OPLL can often present as segmental or continuous type of lesion spanning multiple levels. Good results have been reported for 1- and 2-level cervical corpectomies; however, corpectomies of 3 levels or more are fraught with complications including graft fracture, graft pistoning, graft dislodgement, instrumentation failure, and pseudarthrosis.[4] Dalbayrak et al.[4] have introduced the idea of the "skip" corpectomy, whereby the C-5 vertebral body is left intact and C-4 and C-6 corpectomies are performed for cervical spondylitic myelopathy and OPLL. They conclude that the C-5 body adds additional stability to the anterior cervical construct and the procedure still provides adequate decompression. The biomechanical complications are compounded by high rates of postoperative dysphagia in patients who have undergone multilevel cervical corpectomy.

Anterior corpectomy has been shown to have significant benefits for these patients, but many surgeons are concerned about the risk of durotomy and postoperative CSF leak due to the adherent ossified PLL and the potential for continued neurological deterioration due to insufficient decompression.[3,6,23,28,35] In the series reported by Chen et al.,[3] only 4 patients suffered a CSF leak, and there was no evidence of neurological decline in patients who underwent anterior corpectomy. The authors conclude that these outcomes are the result of a lack of traumatic manipulation of the spinal cord and protection of the epidural vascular plexus. They used specially designed microdissectors to separate the ossified PLL from the dura and they also used the popular "floating" technique when there was evidence of dural ossification. Hida et al.[9] studied the CT characteristics of patients with OPLL and demonstrated some key findings that help determine the amount of dural ossification and thus the potential risk of durotomy and CSF leak at the time of surgery. They studied 21 patients with radiographic evidence of OPLL and correlated their findings with intraoperative evidence of dural defect. They describe double- and single-layer OPLL. Twelve patients were found to have a double layer of ossification on CT at the level of the thickest ossified PLL and 10 patients were found to have a single calcified layer. Of the 12 patients with double-layer OPLL, 10 of them were found to have dural defects, whereas 9 of the 10 patients with single-layer OPLL had intact dura.[9] Utilizing these findings on preoperative CT can help determine which areas of the ossified PLL should be left in place at surgery. In our experience, calcified ligament can be removed safely. When removing the calcified ligament, one must be careful to dissect in the appropriate plane. That plane can often be found by sliding a blunt nerve hook under the lateral aspect of the ligament. Utilizing the appropriate plane minimizes the risk of neurological injury, CSF leak, and injury to the engorged epidural venous system. In the event that the ligament is adherent to the dura, certain "bone islands" can be left in place. If the area of adherent calcified ligament is totally disconnected circumferentially, then it is unlikely to impose any compression on the spinal cord. Repair of an anterior durotomy can be challenging, particularly in the setting of calcified dura. If a CSF leak occurs it should be repaired immediately. If CSF drains from the thecal sac, the engorged epidural venous plexus will likely bleed significantly, making the procedure even more technically challenging.

Although in our experience anterior decompression and reconstruction can be done effectively and safely for 2-level disease, 3-level corpectomies are often not well tolerated due to postoperative dysphagia as well as the biomechanical challenges of such a large construct. With that in mind, we find that the most important factor in choosing an anterior or posterior approach is cervical lordosis. In particular, patients who lack cervical lordosis may benefit from an anterior approach when possible. In cases in which posterior decompression is required, however, we typically perform a stabilization procedure in our institution due to concern about progression of OPLL as a result of dynamic instability or progression of sagittal deformity. Posterior decompression alone has been shown to accelerate progression of OPLL ventrally. Takatsu et al.[32] demonstrated an increase in the rate of deterioration of patients who underwent posterior decompression alone—laminectomy or laminoplasty—compared with those who did not undergo any surgical intervention for OPLL. Likewise, when the decompression extends to the cervical-thoracic junction, stabilization should extend across the cervical-thoracic junction to T-2 to minimize the risk of a junctional kyphosis.

In the US, most cases of OPLL are treated using a posterior approach, as it can be a less technically than an anterior corpectomy. However, the need for posterior stabilization is controversial. Laminoplasty has been used to achieve posterior decompression while preserving motion, but patients with a straight or kyphotic cervical spine are at risk for progression of cervical kyphosis, and laminoplasty is contraindicated in this population. Patients with reversal of cervical lordosis who undergo laminoplasty have been shown to experience progression of kyphotic deformity as well as progression of ligamentous ossification postoperatively.[3,13,17,22]

Laminectomy without stabilization or laminoplasty can be an effective surgical option in specific circumstances. Iwasaki et al.[14,15] have studied long-term outcomes in patients who underwent cervical laminoplasty for OPLL and concluded that the most significant predictors for poor neurological outcome following laminoplasty are "hill-shaped" ossification, lower preoperative JOA score, postoperative change in cervical alignment, and older age at time or surgery. Likewise, they report worse neurological outcomes for patients with an occupying ratio of 60% or greater.

Matz et al.[24] performed a detailed analysis of the literature and concluded that there is Class III evidence to support the use of laminoplasty in cervical spondylitic myelopathy or OPLL. They report 55%–60% JOA score improvement compared with nonoperative therapy. Furthermore, while the most common complication of laminoplasty is the development of a C-5 nerve palsy due to cord shift and nerve root stretch, there is preserved range of motion and Class III evidence of equivalence in functional improvement between laminoplasty, anterior cervical decompression and fusion, and laminectomy with arthrodesis.[24] For our patient population, laminoplasty or laminectomy alone are rarely used due to the concern about progression of OPLL or development of cervical kyphosis. Many patients with OPLL also have DISH. The combination of OPLL and DISH may provide some anterior stability to the construct and thus reduce the risk of developing progressive kyphosis in the setting of laminoplasty or laminectomy. However, we have not had that experience at our institution, and thus all patients with multilevel cervical decompression in the setting of a straight or kyphotic cervical spine also undergo stabilization.

The technique we employ for posterior cervical decompression typically involves an en bloc laminectomy. Using a high-speed drill, bilateral troughs are created at the junction of the lamina and lateral mass or lamina-facet line, thus releasing the lamina from the lateral mass. Once the troughs are drilled, a nerve hook is used to define the epidural space. Disconnection of the osteoligamentous structures is performed by using Kerrison punches to systematically complete the disassociation of lamina from the facets. Similarly, disconnection of the osteoligamentous structures must also be performed at the cephalad and caudal ends to remove the lamina en bloc. This process of removing the ligamentum flavum must be done carefully to avoid any tearing of the dura. Both the ligamentum and the dura may be calcified or adherent to one another. The lamina can later be used as autograft for fusion. Throughout the procedure, the mean arterial blood pressure is maintained above 80 mm Hg to provide appropriate spinal cord perfusion. Instrumentation is typically placed prior to decompression using a modified Magerl technique[16] for placement of lateral mass fixation and the free-hand technique described by Lenke[19] is used for thoracic pedicle screw placement when necessary. Complete arthrodesis of the facet complex is essential to optimize fusion. The use of osteobiologic agents is a controversial topic and is left to the discretion of the surgeon. While osteobiologics can enhance fusion rates in the posterior cervical spine, their use in the anterior cervical spine is not recommended due to the intense inflammatory response generated and the concern for compromising the patient's airway. The fusion rate in our case series is high and likely on a par with other institutions' series in large part due to the inflammatory nature of OPLL and high propensity for calcification and fusion.

Overall, the decision to use an anterior or posterior approach is a complicated one involving many factors. In our series, 67% of the patients underwent posterior decompression and fusion. Posterior stabilization is often a less technically demanding procedure that can be done safely and effectively. Chen et al.[3] retrospectively studied radiographic and clinical outcomes in 75 patients with severe OPLL as defined by 3 or more levels of OPLL with at least a 40% compromise of the cervical canal. The patients underwent anterior corpectomy and reconstruction, laminectomy and posterior instrumented stabilization, or laminoplasty. In this study they demonstrated a few important points. First, patients who underwent anterior corpectomy or laminectomy and instrumentation maintained a significantly greater cervical lordosis compared with those who underwent laminoplasty. Also, Chen et al.[3] demonstrated that neurological improvement measured by the JOA scale was significantly greater in patients who underwent anterior corpectomy than in those who underwent laminoplasty. For example, the mean JOA improvement percentage (± SD) for those in the 3 groups was as follows: anterior corpectomy 63.2 ± 15.2, laminectomy and stabilization 43.5 ± 12.7, and laminoplasty 25.1 ± 8.5. None of the patients in the anterior corpectomy group or the laminectomy and fusion group experienced neurological deterioration postoperatively, but 4 of 25 patients who underwent laminoplasty suffered neurological deterioration and progressive kyphotic deformity. The fact that some of the patients in the laminoplasty group experienced worsening neurological deficits helps support the idea that there is indeed an impact from a repeated mechanical stress and potential instability that may contribute to development of OPLL. The patients in this study who underwent instrumented stabilization did not suffer the same neurological decline. Whether the deterioration was from progressive kyphosis or from OPLL progression is not completely clear, but may be due to both processes leading to continued or worsening stenosis/compression. Although the literature suggests improved outcomes for anterior decompression and reconstruction, both groups of patients in our series who underwent anterior or posterior procedures did well neurologically. The decision to use an anterior or posterior approach was based on the number of levels involved, cervical lordosis, medical comorbidities, and patient age.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....