Further genetic analysis of 91 sibling pairs of patients with OPLL suggests that the genetic locus for OPLL may be situated close to the HLA region on chromosome 6p.^6,⁷ Patients with DISH, who commonly have OPLL, also test positive for the same HLA antigen. The COL11A2 gene, which is located within the class II major histocompatibility complex region of chromosome 6, is believed to be the gene responsible for OPLL. Patients with OPLL have a significantly higher incidence of genetic abnormalities in the COL11A2 region.⁸ COL11A2 encodes the α₂ chain of type XI collagen, which is a minor collagen in the cartilage matrix and may be involved in controlling the growth of type II collagen. Overproduction of type XI collagen has been seen in histologic studies of some surgical specimens of OPLL.⁹

Other genetic loci associated with OPLL include the collagen 6A1 gene (COLA1). COLA1 is located on chromosome 21q, and genome-wide linkage and linkage disequilibrium analysis of 142 affected sibling pairs has identified this gene locus as the candidate gene for the pathogenesis of OPLL.¹⁰ COLA1 encodes the α₁ chain of type VI collagen.¹¹ Overproduction of type VI collagen may serve as a scaffold for chondrocyte infiltration and endochondral bone growth.

The animal model of OPLL, the tiptoe walking (ttw) mouse, demonstrates progressive calcification of spinal ligaments. A single-base mutation in the nucleotide pyrophosphate gene (NPPS) is thought to be responsible for the development of ligament ossification in this model. NPPS encodes for a protein that inhibits the mineralization and calcification of cartilaginous tissue. Loss of the NPPS protein may cause unregulated ossification of the spinal ligaments. In humans, NPPS has three isoforms, with NPP1 being located on chromosome 6. A single-base polymorphism in NPP1 is more common in patients with cervicothoracic OPLL than in those with only cervical OPLL, perhaps indicative of more severe and progressive disease.¹²

Growth Factors

The posterior longitudinal ligament in patients with OPLL demonstrates an enhanced potential for osteogenesis, thus suggesting that intrinsic growth factors may be integral in the pathogenesis of OPLL. Bone morphogenetic proteins (BMPs) are a class of growth factors involved in the induction of new bone formation. The nonossified ligaments of OPLL patients and their relatives are found to have elevated concentrations of BMP.¹³ Histologic evaluation of OPLL surgical specimens demonstrates the presence of BMP-2 in the ossified matrix, chondrocytes, and fibroblasts in cartilaginous areas adjacent to the OPLL.^14,¹⁵ Furthermore, BMP-2 mRNA was isolated from the spinal ligaments of patients with OPLL and not from control subjects.¹⁵ Zinc finger protein 145, a regulator of BMP expression and an early initiator of osteoblastic differentiation, is also upregulated in OPLL patients. Comparatively, the MSX2 gene, which expresses a protein that inhibits osteoblastic differentiation and mineralization of fibroblasts, is downregulated in OPLL patients.

Other growth factors associated with OPLL include transforming growth factor-β, which is present in the ossified matrix and chondrocytes in cartilaginous tissue adjacent to OPLL. Osteocalcin synthesis, which reflects the osteoblastic phenotype of cells, is observed in the supernatant of the posterior longitudinal ligament of patients with OPLL and not from patients with non-OPLL cervical spondylosis.¹⁶ Insulin-like growth factor, connective tissue growth factor, parathyroid hormone, platelet-derived growth factor, retinoid, estrogens, and interleukin-1 are also associated with the pathogenesis of OPLL.

Metabolic Factors

OPLL is associated with a high incidence of adult-onset obesity and impaired glucose tolerance.^17,¹⁸ Among 535 patients with OPLL, 28% were diabetic and 18% qualified as borderline diabetic.¹⁹ In patients with known OPLL, the extent of ossification is significantly correlated with the fasting serum insulin level.¹⁷ In a study from the Japan Collaborative Epidemiological Study Group for Evaluation of Ossification of the Posterior Longitudinal Ligament of the Spine Risk,²⁰ adult-onset obesity and non–insulin-dependent diabetes were identified as independent risk factors for OPLL. When 69 eligible patients with OPLL were compared with control patients without spinal disorders matched for age and gender, the OPLL group had a significantly higher proportion of patients with diabetes and a higher body mass index.

From the same study, additional dietary habits associated with increased risk for the development of OPLL included frequent consumption of pickled foods. Alternatively, diets rich in chicken and soy products were correlated with a decreased risk for OPLL. No association between cigarette or alcohol consumption and OPLL was observed. Other reports have suggested that hypoparathyroidism and hypophosphatemic rickets are related to a higher incidence of OPLL.¹⁹

Other Enthesopathies

The enthesopathies are characterized by progressive inflammation of tendons and ligaments and their association with bone. OPLL is of primary concern to neurosurgeons and orthopedic spine surgeons because of the location of the posterior longitudinal ligament within the spinal canal and the risk for cord compression with hypertrophy and calcification. Other major spinal enthesopathies include DISH, OALL, and OLF. DISH is an ossifying condition that affects the axial and appendicular skeleton and is commonly also known as ankylosing hyperostosis. With spinal involvement, the anterolateral vertebral column is primarily affected. DISH most commonly occurs in adults, and it affects 3% of those older than 40 years and 15% of those older than 65 years.^21,²² Because ossification occurs outside the spinal canal, patients are generally asymptomatic; however, an extensive calcified cervical mass may cause dysphagia or thoracic outlet syndrome and require surgical resection.²³ DISH and OPLL may be genetically linked, with the same HLA antigen that tests positive in patients with OPLL occurring in those with DISH as well. Fifty percent of patients with DISH concurrently have OPLL, whereas 24% of those with OPLL also demonstrate DISH.¹⁹

AS is an immune-mediated inflammatory condition that results in progressive erosion of the intervertebral disks and apophyseal joints. Joint destruction is eventually replaced by ankylosis and subsequent autofusion of the spine. Patients with AS also suffer from diffuse osteoporosis, which leads to an increased susceptibility to fractures after minor trauma and a high risk for spinal deformity. AS is genetically linked to the HLA-B27 antigen. OPLL reportedly occurs in 3.5% to 29% of patients with AS.^24,²⁵ AS, an important entity for those concerned with spinal disease, is discussed separately in Chapter 284.

Clinical Findings

OPLL most commonly occurs in the cervical spine of middle-aged or elderly adults and is uncommon in children and adolescents. The frequency of OPLL begins to increase after the age of 40, and most patients are initially seen in the sixth decade. It predominantly affects the cervical spine, with 70% of cases occurring in this region, and generally affects two to four levels. The remaining 30% of cases are generally divided between the upper thoracic and upper lumbar regions. Cervical OPLL occurs in twice as many males as females; however, thoracic OPLL is more frequent in women.

Clinically, OPLL may be asymptomatic, but when symptoms do develop, patients have signs and complaints consistent with myelopathy or radiculopathy. Patients may have neck pain, paresthesias, radicular pain, motor weakness, and hand dyscoordination. With severe disease, long-tract signs such as spasticity, gait disturbance, and bowel and bladder dysfunction arise. Neurological deficits occur secondary to direct mechanical compression of the spinal cord as a result of the OPLL lesion occupying space within the spinal canal. Compression also causes perturbed vascular flow with subsequent spinal cord ischemia. Ventral stenosis may lead to greater loss in the corticospinal tracts with progressive motor dysfunction. Chronic injury may result in gliosis and myelomalacia of the spinal cord. Generally, symptoms follow a slow, insidious course; however, in the setting of even minor trauma with hyperextension of the neck, patients may have acute spinal cord injury.

The cervical myelopathy attributable to OPLL is classified according to either the Nurick grade or the Japanese Orthopedic Association (JOA) score. The Nurick classification is composed of six grades (0 to 5) based on the severity of radiculopathy and myelopathy. A patient who is neurologically intact or has only mild radiculopathy is grade 0; grade 6 correlates with severe myelopathy or quadriplegia. The JOA classification is a 17-point composite score in which points are assigned based on hand and leg motor function, upper and lower extremity sensory function, trunk sensory function, and bladder continence.²⁶ A modification of the JOA score incorporates a manual muscle test of shoulder and elbow strength, and a western modification has also been developed because the scoring of upper extremity function is based in part on the ability to use chopsticks.

Diagnosis

OPLL is diagnosed by the presence of an ossified ligament on radiographic imaging. It is often identified on lateral plain radiographs as a longitudinal hyperdense area of calcification along the dorsal margin of the vertebral body. Based on lateral radiographs, OPLL is classified into one of four subtypes (Fig. 282-1A to D). The continuous type is characterized by a single contiguous ossified ligament that spans two or more adjacent vertebrae. The segmental type is the most common form and has fragmented lesions located behind the vertebral bodies that do not cross the disk space. The mixed type is a combination of both continuous and segmental lesions. Rarely, OPLL is a localized variant that is confined to only the area posterior to the disk space.

FIGURE 282-1 Illustrations demonstrating subtypes of ossification of the posterior longitudinal ligament: A, continuous; B, segmental; C, mixed; D, localized.

Lateral plain films are used to determine the occupying ratio and the space available for the spinal cord (SAC). The occupying ratio is the maximum anteroposterior thickness of the ossified ligament divided by the diameter of the spinal canal at the corresponding level. An occupying ratio of greater than 40% is associated with a high risk for myelopathy. In a clinical series of patients with OPLL, Matsunaga and colleagues found that all patients with greater than 60% stenosis suffered myelopathy.²⁷ Alternatively, in patients with less than 60% stenosis, there was no correlation between the occupying ratio and Nurick grade. The SAC is determined by measuring the anteroposterior distance between the dorsal aspect of the ossified lesion and the posterior margin of the lamina at a constant tube-to-film distance of 150 cm. In one clinical series, all patients with a SAC of less than 6 mm had myelopathy, whereas no patients with a SAC of 14 mm or greater demonstrated myelopathy.²⁸ For patients with a SAC of between 6 and 14 mm, there was no correlation between SAC and the degree of neurological compromise.

Additional imaging modalities such as CT and MRI may facilitate the diagnosis and management of OPLL. CT is particularly helpful for better characterizing the morphology of the ossified lesion in cross section, identifying potential dural involvement, and improving visualization of the cervicothoracic junction. CT axial and reconstructed images and CT myelography may facilitate the detection of thoracic OPLL or OLF. MRI is useful for imaging the spinal cord and nerve roots; however, differentiation of OPLL from conventional cervical spondylosis may not be possible. Although early OPLL with bone marrow and fat within the ligament may appear as hyperintense foci, mature OPLL lesions generally appear hypointense on MRI. MRI, however, may demonstrate evidence of changes in spinal cord signal indicative of edema, gliosis, or myelomalacia. The presence of these imaging findings serves as an indirect measure of neurological status and may facilitate treatment planning and assessment of patient prognosis.

Natural History

The natural history of OPLL is characterized by slowly progressive and variable growth of the ossified lesion. In a 2-year radiographic follow-up study of 13 patients with OPLL managed conservatively, the growth rate of the ossified lesion varied highly from patient to patient; however, mean growth was 4.07 mm longitudinally and 0.67 mm in thickness per year.²⁹ The natural history clinically appears to be dependent on the initial neurological status at diagnosis. Matsunaga and coworkers studied 167 patients in whom OPLL was diagnosed and treated conservatively for an average of 11 years.²⁸ Of the 140 patients who were initially asymptomatic, in only 18% did myelopathy subsequently develop during the study period. Alternatively, of the 27 patients who were initially myelopathic and refused to undergo surgery, 44% experienced worsening of their neurological status during the study period. Of all 167 patients managed conservatively, myelopathy either developed or worsened in only 22% despite the fact that 42% demonstrated radiographic growth of their ossified lesion during the study period.

In a subsequent study with a minimum of 10 years’ follow-up, Matsunaga and associates evaluated 304 patients with OPLL who were managed conservatively for 10 to 30 years.²⁷ Of the patients who were initially asymptomatic, only 17% became myelopathic during the study period. The Kaplan-Meier estimate of the myelopathy-free rate in patients without myelopathy at initial evaluation was 71% at 30 years. Of patients who were myelopathic initially, however, 64% deteriorated neurologically during the follow-up period. Radiographic evaluation revealed that patients with minimal stenosis on initial assessment rarely progressed to greater than 60% stenosis during follow-up examination. The results of these studies suggest that in only a small percentage of patients in whom OPLL is diagnosed and is initially asymptomatic will myelopathy eventually develop during their lifetime.

Management

Conservative management of patients with OPLL is reserved for those who are asymptomatic and who radiographically demonstrate a small degree of stenosis. Additionally, nonoperative management may be appropriate for patients with advanced age or significant medical comorbidities, those who are poor surgical candidates, or those who have severe chronic neurological deficits that are presumably irreversible. MRI evidence of spinal cord myelomalacia may facilitate assessment of the prognosis for neurological recovery with surgical intervention. Conservative treatment consists of oral nonsteroidal agents, epidural steroid injections, and potentially a cervical orthosis for patients with neck pain.

Surgical decompression is indicated for patients with myelopathic or radicular findings such as paresthesias, radicular pain, motor weakness, hand dyscoordination, gait disturbance, spasticity, and bowel or bladder dysfunction. Early treatment of mild to moderate myelopathy, especially in younger patients, may prevent the development of severe irreversible neurological deficits. Patients with mild symptoms but who have a SAC of less than 6 mm, an occupying ratio of greater than 40%, or MRI evidence of changes in spinal cord signal may also be considered candidates for surgical decompression.

Severely compromised patients, however, may not benefit from surgical treatment. In a study of patients with a minimum of 10 years’ follow-up, surgery was no more effective than conservative measures in patients with a poor Nurick grade 5 (wheelchair bound or bedridden).²⁷ Surgical decompression proved better than conservative management only in patients with a moderate Nurick grade 3 or 4 (able to ambulate with assistance but unable to work). However, this lack of improvement may be due to the spinal cord already being permanently damaged from traumatic compression or ischemia. Additionally, posterior surgery alone instituted in severely compromised states may not provide adequate neural decompression.

Some contend that prophylactic surgery may be indicated in younger patients with mild symptoms but severe stenosis radiographically. With a longer life expectancy, younger patients have a greater risk of suffering even minor cervical trauma that may result in permanent neurological injury. Patients with a narrowed SAC of less than 10 mm have a higher risk for the development of myelopathy after trauma, with up to 23% of patients with OPLL becoming acutely or increasingly symptomatic after a minor event.³⁰ In a study of patients with OPLL manifested as acute cervical cord injury, the level of the injury was frequently at the caudal edge of the ossified lesion, probably secondary to hyperextension of the neck with minor trauma.³¹

Once the decision to proceed with surgical intervention is made, an appropriate surgical approach must be selected. A variety of anterior and posterior procedures have been described for direct or indirect decompression of the spinal cord in the setting of OPLL (Fig. 282-2). Anterior cervical approaches consist of direct decompression by either surgical resection or “floating” of the ossified lesion, along with reconstruction of the anterior spinal column and fusion. Posterior cervical approaches achieve indirect decompression by increasing SAC through either laminectomy or laminoplasty without directly addressing the ossified lesion. Even though several studies have failed to demonstrate any significant difference in outcome between anterior and posterior surgery,³²^–³⁵ anterior approaches are associated with longer operative time, increased blood loss, and more frequent complications.³⁵

FIGURE 282-2 A 63-year-old man with myelopathy from ossification of the posterior longitudinal ligament was treated by multilevel laminectomy and posterior instrumented fusion. The osteophytes within the spinal canal can be seen on the radiograph, as well as the obvious ossification of the anterior longitudinal ligament consistent with diffuse idiopathic skeletal hyperostosis.

Anterior Surgery

Anterior procedures for cervical OPLL allow definitive resection or mobilization of the ossified lesion and direct decompression of the spinal cord (Fig. 282-3A to I); however, they are frequently complicated by significant epidural bleeding, leakage of cerebrospinal fluid from ossified dural involvement, and potential injury to the spinal cord or nerve roots. Alternatively, an anterior floating method has been described in which a generous corpectomy is performed laterally such that the transverse dimensions extend at least 20 to 25 mm.^34,³⁶ Particular attention is directed toward identifying the uncinate processes to demarcate the extent of lateral resection. Once the ossified lesion is released at its lateral, cranial, and caudal margins with the use of a high-speed drill and a diamond bur, the calcified plaque rises ventrally to decompress the spinal cord. With the anterior floating method, the ossified lesion is not resected but instead is allowed to migrate anteriorly into the corpectomy defect. After decompression, either by resection of the OPLL or with the anterior floating method, anterior strut grafting with or without instrumentation is performed. Other complications described with anterior decompression include graft dislodgement, pseudarthrosis, and C5 nerve root palsy.

FIGURE 282-3 A 37-year-old Vietnamese dentist had severe acute myelopathic findings after a surfing accident. Sagittal (A) and axial (B) magnetic resonance images demonstrated severe but focal spinal cord compression, which on computed tomography (CT) (C-E), was definitively demonstrated to be ossification of the posterior longitudinal ligament (OPLL). He underwent anterior cervical corpectomy because of (1) his young age (and high risk for recurrence of symptoms with posterior decompression), (2) the focal compression, and (3) his desire to avoid postoperative neck pain. F and G, Postoperative lateral and anteroposterior radiographs demonstrating anterior decompression and reconstruction with a fibula graft and anterior plate. H and I, CT reconstruction demonstrates resection of the OPLL lesion.

Posterior Surgery

Unlike anterior surgery, posterior procedures achieve indirect decompression by removing or elevating the dorsal elements to increase the space in the spinal canal. The increased SAC allows the cord to migrate dorsally from the ventral compression. Posterior decompression is primarily effective for patients with preserved cervical lordosis. Cervical kyphosis tethers the spinal cord over the vertebral bodies and OPLL and thus does not allow the cord to move away from the ventral pathology.

Cervical laminectomy without fusion may be performed in patients with a stable spinal column; however, particular attention to maintaining greater than 50% of the facet joint is necessary to preserve stability. Laminectomy with posterior fusion may be indicated for patients who require generous facetectomies or foraminotomies or for those who have questionable spinal stability (Fig. 282-4A to C). Complications associated with cervical laminectomy include postlaminectomy membrane and scar formation, delayed kyphosis, and C5 nerve root palsy.

FIGURE 282-4 A 55-year-old man complained of severe myelopathy and was found on magnetic resonance imaging (A) to have multiple levels of spinal cord compression because of ossification of the posterior longitudinal ligament. B and C, Because of the multiple levels of compression, he underwent posterior decompression with laminectomy and instrumented fusion as a result of the nonlordotic stature of his cervical spine. The laminectomy was performed in the same fashion as laminoplasty, with en bloc removal of the posterior elements in a “lobster tail” fashion (D) to avoid the need for placement of Kerrison rongeurs into the spinal canal, which could cause neural injury.

Because patients with OPLL being treated surgically often have a significant degree of compression, en bloc laminectomy is recommended. With en bloc laminectomy or “lobster tail” removal of the lamina, bilateral gutters are drilled along the lamina-facet junction across the segments that require decompression. The ligamentum flavum is then resected along these lateral margins with a 2-mm Kerrison rongeur. Finally, disconnection of the interspinous ligaments and the ligamentum flavum at the cranial and caudal extents of the decompression allows en bloc removal of the intervening lamina and ligamentum (Fig. 282-4D). In so doing, one avoids repeatedly inserting the Kerrison rongeur into the central portion of the spinal canal, where the cord is the most compromised.

In 1971, Oyama and Hattori introduced an alternative posterior approach—the Z-plasty or laminoplasty.³⁷ Laminoplasty achieves decompression by elevating without removing the posterior elements from the spinal cord, thereby allowing the cord to migrate dorsally. Shortly thereafter, Hirabayashi and coauthors reported a revised laminoplasty procedure, the expansile open-door or “single-door” laminoplasty.^38,³⁹ Several different techniques and subsequent modifications of cervical laminoplasty have been developed (Fig. 282-5A to F). The expansile open-door laminoplasty technique has since been augmented with graft spacers or the addition of hardware (or both) for maintaining the “open door.” The “French door” laminoplasty creates an enlargement of the median canal by splitting the spinous process, elevating the lamina bilaterally, and interposing a graft spacer to maintain the “open double door.”

FIGURE 282-5 Illustrations demonstrating various posterior decompression and laminoplasty techniques. A, En bloc laminectomy. B, Hattori’s Z-plasty. C, Hirabayashi’s expansile open-door laminoplasty. Suture crosses the spinous process and the hinge-side facet to maintain the open door. D, Instrumented ungrafted open-door laminoplasty. A small screw-plate construct is affixed to the lamina and facet to maintain the open door. E, Instrumented grafted open-door laminoplasty. Bone graft is lodged between the open door and the facet and instrumented with a screw-plate construct. F, Kurokawa’s French-door laminoplasty. Bone graft is interposed between the split spinous process to maintain the open double door.

The benefit of laminoplasty over laminectomy is maintenance of the dorsal tension band and preservation of the posterior elements as a barrier against postlaminectomy membrane formation. For adequate decompression, 4 to 5 mm of canal expansion in the anteroposterior diameter or a 50% increase in overall canal area is necessary. This corresponds to greater than 3 mm of dorsal cord migration, which correlates with optimal clinical outcomes.⁴⁰ Despite these theoretical advantages of laminoplasty, studies have failed to demonstrate any conclusive benefit of any laminoplasty technique over cervical laminectomy.⁴¹

Complications associated with laminoplasty include a 5% to 10% risk for postoperative C5 nerve root palsy.^33,^42,⁴³ The development of C5 nerve root palsy is probably due to dorsal migration of the spinal cord and subsequent traction on the exiting nerve roots. With the C5 nerve root being at the apex of the lordosis, it is susceptible to greater tethering and tension. A variable rate of axial neck and shoulder pain has been described after laminoplasty. This pain is frequently reported as affecting the ipsilateral shoulder of the hinge side of a single-door laminoplasty, which may be due to subsequent intersegmental fusion. Cervical laminoplasty is associated with a significant decrease in range of neck motion postoperatively in most series.⁴¹ Although delayed kyphosis develops less frequently after laminoplasty than after laminectomy, a significant percentage of patients have a relative loss of lordosis postoperatively.⁴¹

Surgical Decision Making: Anterior versus Posterior Approach

Ultimately, the surgical treatment of symptomatic patients with OPLL requires selecting the appropriate technique, whether it be an anterior or posterior approach. Both methods have been demonstrated in the literature to be effective in achieving surgical decompression with a high rate of neurological improvement.^33,^34,⁴²^–⁴⁷ However, the paucity of the literature directly comparing anterior and posterior decompression poses an oftentimes difficult decision in selecting the appropriate procedure for a given patient. General guidelines that factor in specific patient characteristics can be applied on a case-by-case basis to facilitate providing patients with a reasonable surgical option that optimizes clinical results while minimizing surgical morbidity (Table 282-1).

TABLE 282-1 Factors Involved in Selecting Anterior versus Posterior Decompression for Treatment of Ossification of the Posterior Longitudinal Ligament

	ANTERIOR APPROACH	POSTERIOR APPROACH
Canal compromise	>60%	<60%
Age	Younger Higher lifetime risk for progression of disease	Older Lower lifetime risk for progression of disease
Medical comorbidities	Prefer patients with low surgical risk	More amenable for patients with higher surgical risk
Number of levels involved	Focal or short-segment disease	Multilevel disease
Cervical alignment	Kyphotic, neutral, or lordotic	Lordotic
Previous surgery	Difficult if previous anterior surgery	Amenable to patients with previous anterior surgery
Positioning	Amenable to patients who are unable to be positioned prone	Only acceptable for patients who can tolerate prone positioning
Axial neck pain	Amenable to patients with existing neck pain	Prefer patients without preexisting neck pain
Osteoporosis	Prefer patients with adequate bone density	Amenable to patients with osteoporosis
Cosmesis	May result in a visible scar	Generally a well-concealed scar
Voice	Relatively contraindicated in singers, speakers	Amenable to singers, speakers

Anterior procedures are primarily of benefit by providing the ability to perform direct surgical resection of the OPLL. Therefore, anterior surgery is recommended for patients who will gain from total removal of the offending lesion, including in particular those with greater than 60% canal compromise (Fig. 282-6A to E),⁴⁸ as well as younger patients, who may enjoy a longer lifetime benefit with a reduced risk for progression of OPLL at the operated levels. Anterior surgery is likewise recommended for patients with preexisting cervical kyphosis because these patients fare poorly with posterior indirect decompression. Anterior decompression with reconstruction and fusion may also help restore cervical lordosis and spinal alignment.

FIGURE 282-6 A-C, A 61-year-old Hispanic man with severe ossification of the posterior longitudinal ligament and greater than 60% compromise of the canal. D, Because of the severity of compression, the patient underwent multilevel anterior corpectomy, which was necessarily followed by posterior supplemental fixation (E).

Anterior surgery is best suited for short-segment or focal compression; however, multilevel anterior corpectomies have been performed with good clinical outcomes. An anterior approach may also be performed in patients with concomitant symptomatic DISH to allow simultaneous treatment of both pathologies. Patients who cannot tolerate prone positioning are also recommended to undergo anterior surgery. A relative indication for anterior surgery occurs in patients with preexisting axial neck pain because these patients will probably have worsening chronic neck pain after a posterior approach and may in fact benefit from anterior stabilization and fusion.

Posterior approaches are recommended for patients with significant multilevel involvement (three or more vertebral levels) or those who have high surgical risk and may not tolerate an anterior procedure. When compared with anterior decompression, posterior surgery is generally associated with shorter operative times and less blood loss. Posterior surgery is also ideal for patients who have previously undergone an anterior approach, thereby avoiding operating through scar and adhesions. Posterior decompression is generally recommended for patients with preserved cervical lordosis, although some contend that patients with neutral alignment may still benefit from laminoplasty.

Posterior decompression should be performed in patients with osteoporosis or low bone density because such patients are at high risk for subsidence of the graft or implant failure with an anterior procedure. A relative indication for posterior surgery occurs in patients who are professional singers or speakers; such patients may not accept even a small risk for recurrent laryngeal nerve palsy. Finally, posterior surgery generally results in a well-concealed incision, whereas an anterior neck scar may present cosmetic issues for some patients.

Surgical Results

Anterior Surgery

Anterior cervical decompression plus fusion in patients with OPLL provides excellent outcomes in both short- and long-term evaluation. In a series of 107 patients with at least 2 years of follow-up, 89% of those with myelopathy exhibited significant postoperative improvement in neurological status.⁴⁵ The fusion rate was 97% despite often performing multilevel procedures. The rate of cerebrospinal fluid leakage was high (20%), and it was the primary complication. Two patients required reoperation for growth of OPLL at a nonoperated site. In only 1 patient did postoperative C5 nerve root palsy develop. In a long-term study of 30 patients monitored for a minimum of 10 years after anterior cervical decompression, 53% demonstrated excellent improvement in ambulatory function (based on Okamoto’s classification of walking disability).⁴⁶ Twenty-seven percent had good recovery, 17% were unchanged, and only 3% had a poor outcome. Postoperative neurological deterioration occurred in 13%; however, no cases were due to continued growth of OPLL at the operated site.

Matsuoka and colleagues studied 63 patients who underwent the anterior floating method with a minimum of 10 years’ follow-up.³⁴ Overall, the average JOA score improved from 8.3 preoperatively to a peak of 14.2 at 5 years’ follow-up. The average postoperative JOA score was maintained for up to 10 years; however, a slight decline to 13.5 occurred at final follow-up. Accordingly, the recovery rate, which is calculated as

peaked at 68.5% at 5 years and was stable for up to 10 years. At last evaluation, the recovery rate had decreased slightly to 59.3%. Radiographic studies have revealed that between 4 and 8 weeks postoperatively, the ossified lesion had successfully migrated anteriorly away from the spinal cord.

There were no neurological complications as a result of surgery, and cerebrospinal fluid leaks occurred in only 5.1% of patients. Eight percent required subsequent posterior decompression an average of 8 years after the initial anterior surgery. The authors found that factors correlated with good or complete recovery after surgery included younger age, shorter duration of myelopathy, and larger preoperative cross-sectional area of the spinal cord. The degree of spinal canal stenosis and the thickness of the ossified lesion were not related to outcome.

Posterior Surgery

Laminectomy

Kato and colleagues studied 44 patients for an average of 14 years after laminectomy.⁴⁹ Overall, the average JOA score improved from 7.6 preoperatively to 11.5 at 1 year postoperatively. JOA scores remained stable for up to 5 years, but they declined to 10.3 at last examination. Accordingly, the recovery rate improved to 40% at 1 and 5 years postoperatively but decreased to 30% at final follow-up. Late neurological deterioration occurred in 19% of patients and was predominantly due to minor trauma or the development of thoracic OPLL or OLF. Only 1 patient (2%) had a decrease in neurological function because of growth of OPLL at the operated levels.

Laminoplasty

Several studies have evaluated the long-term clinical results (5 to 14 years postoperatively) in patients who underwent laminoplasty for OPLL.^33,⁴²^–⁴⁴ ^,47 Average JOA scores improved from 8.3 to 9.2 preoperatively to 13.5 to 14.2 at 3 to 10 years postoperatively. Studies that evaluated patients beyond 10 years observed a slight decline in JOA score to 12.3 to 13.1 at last examination. Accordingly, recovery rates peaked at 59% to 63.1% at 3 to 10 years postoperatively and subsequently declined to 47.9% to 54% at last examination when monitored for more than 10 years.

In the study by Chiba and associates, late neurological deterioration, defined as greater than a 1-point decrease in JOA score, was observed in 30% of patients. Patients who met the criteria for late neurological deterioration had an average age of 77 years.⁴² Iwasaki and colleagues³³ and Ogawa and coauthors⁴³ reported that 15% to 16% demonstrated a late decrease in JOA score of greater than 2 points, with only 4% of patients in both series deteriorating because of progression of OPLL at the operated level. Primarily, the decreased neurological function was due to the de novo development of thoracic OPLL or OLF. Concordant with these findings, Chiba and coauthors reported that patients maintained a stable degree of canal expansion from 1 year postoperatively until final examination. The authors observed that the canal increased from 13.5 mm before surgery to 18.3 mm at 1 year and was sustained at 17.9 mm at last radiographic evaluation, thus suggesting that late deterioration was not due to progression of disease at the operated segments.⁴²

Surgical complications included a 5% to 10% incidence of C5 or C6 nerve root palsy.^33,^42,⁴³ Axial neck pain occurred in 25% to 31% of patients, with 41% to 53% complaining of shoulder pain. Other complications encountered included cerebrospinal fluid leaks, postoperative hematoma, and infection. Several studies observed that patients exhibited a significant loss of range of motion after laminoplasty.⁴²^–⁴⁴ Decreased cervical lordosis was also seen after laminoplasty⁴²; however, it remains unclear whether the change in cervical alignment correlated with diminished recovery of neurological function.^43,⁴⁴ The development of kyphosis, though, does appear to be related to a higher incidence of postoperative shoulder pain.⁴³

Factors correlating with better clinical outcome included younger age at the time of surgery and higher preoperative JOA score, particularly for patients with severe myelopathy at initial evaluation.^33,⁴³ Patients with moderate preoperative myelopathy fared better with a shorter duration between the onset of symptoms and initiation of surgery.⁴³ In 2007, Iwasaki and colleagues reanalyzed their case series and found that the morphology of the ossified lesion, particularly a hill-shaped ossification, correlated with worse clinical outcomes.⁵⁰ Despite comparable preoperative JOA scores, patients with a preoperative occupying ratio of less than 60% had higher postoperative JOA scores than did those with an occupying ratio of greater than 60%. Segmental versus nonsegmental OPLL was not found to affect postoperative JOA scores or the degree of neurological recovery.⁵¹

Anterior versus Posterior Surgery

Few studies have directly compared anterior and posterior surgery for the treatment of cervical OPLL. Unfortunately, these studies are generally retrospective in nature and reflect a change in institutional preference rather than a randomized comparison between two techniques. Wada and coworkers performed a long-term study of patients treated with anterior subtotal corpectomy versus laminoplasty for cervical spondylotic myelopathy.⁵² They found no significant difference in neurological recovery at 1 and 5 years postoperatively and at last examination. The laminoplasty group had a higher incidence of axial neck pain and greater loss of range of motion. The corpectomy group suffered from a longer operative time, increased blood loss, and pseudarthrosis (26%). The anterior cervical fusion group also demonstrated a 54% rate of radiographically evident adjacent segment degeneration; however, in only 4% did clinical symptoms develop as a result.

Iwasaki and associates studied anterior decompression and fusion versus laminoplasty for an average of 6 years in patients with OPLL.⁴⁸ They observed no significant difference in postoperative JOA score or recovery rate between anterior and posterior procedures in patients with an occupying ratio of less than 60%. However, patients with an occupying ratio of 60% or greater fared significantly better after anterior decompression than after laminoplasty. The anterior surgical group had a maximum recovery rate of 64%, which decreased slightly to 54% at final examination. Comparatively, the laminoplasty group had a maximum recovery rate of just 34%, which significantly deteriorated to 14% at last evaluation. Of note, 26% of patients in the anterior decompression group did require an additional operation because of either graft-related complications, need for posterior stabilization, or subsequent laminoplasty for late neurological deterioration. Only one patient (2%) in the laminoplasty group required a second surgery, which was performed because of postoperative epidural hematoma.

Progression of Disease

A major issue in clinical outcome after either conservative or surgical intervention is the long-term potential for progression of the ossified lesion. Matsunaga and coworkers studied 167 patients with OPLL who were treated conservatively with an average follow-up of 11.2 years.²⁸ Radiographic evidence of axial progression of OPLL occurred in 42% of patients, with longitudinal growth occurring in 86%. Of the 70 patients with OPLL growth, de novo myelopathy or clinical worsening of their initial neurological status developed in 53%.

The reported incidence of OPLL progression after anterior decompression is variable and ranges from 36% to 64%.^34,⁵³ In a study of patients monitored for more than 10 years after undergoing anterior interbody fusion or subtotal corpectomy for OPLL, 87% demonstrated longitudinal growth of the OPLL, and 50% had an increase in thickness of the OPLL.⁴⁶ In a long-term study of patients who underwent the anterior floating method, only 8% demonstrated postoperative progression of OPLL at the operated level.³⁴ The low rate of OPLL progression may be due to disruption of the vascular supply to the ossified mass or elimination of intersegmental motion after fusion. At the levels adjacent to the operated field, however, 36% showed progression of OPLL either longitudinally or in thickness. Thirty percent had longitudinal growth adjacent to the operated level, with 25% demonstrating increased thickness of the OPLL adjacent to the operated segment.

OPLL progression is of particular concern in patients undergoing posterior indirect decompression. Because the ossified lesion is not resected, further growth of OPLL may lead to late neurological deterioration. Some have questioned whether posterior decompression surgery actually increases the rate of OPLL growth. Takatsu and colleagues radiographically studied OPLL progression in patients undergoing laminectomy, laminoplasty, and conservative treatment.⁵⁴ They found no significant difference in OPLL growth in the laminectomy and laminoplasty groups. OPLL progression, however, was more rapid in both surgically treated groups than in patients treated conservatively. Postsurgical changes in the microcirculatory and biomechanical environment may cause stimulation of OPLL development. However, patients in this series who were surgically treated may also have had more aggressive disease than those treated conservatively. Another study performed by Matsunaga and associates found that the rate of OPLL progression after laminoplasty was in fact comparable to that in patients treated nonsurgically.²⁷

Several studies have evaluated the risk for progression of OPLL after long-term follow-up of patients who underwent laminectomy or laminoplasty.^33,^49,^51,⁵⁵^–⁵⁸ Radiographic evidence of OPLL progression occurred in 70% to 73% of patients who were evaluated for a minimum of 10 years after posterior decompression.^33,^49,⁵⁸ In a multicenter study using a computer-assisted method for measuring OPLL progression, the incidence of OPLL growth at 1 and 2 years was 38.9% and 56.5%, respectively.⁵⁵ OPLL progression occurred more commonly in younger patients, with the average age of patients with OPLL growth being 52 to 54 years⁵⁶^–⁵⁸ as compared with 60 years in those who did not demonstrate significant OPLL growth.⁵⁸ Several studies have observed that OPLL progression occurs more commonly in the continuous and mixed types than in segmental OPLL.^33,^51,⁵⁶^–⁵⁸ Gender and type of posterior decompression were not correlated with increased risk for OPLL progression.⁵⁵

With regard to longitudinal growth, 64% to 75% of patients demonstrated longitudinal progression of OPLL after laminoplasty over 5 to 13 years of follow-up, with the average total growth being 9 to 13 mm.^33,^56,⁵⁸ Using computer-assisted radiographic analysis, Chiba and colleagues found that OPLL growth had occurred 1.5 mm rostrally and 1.3 mm caudally 1 year postoperatively.⁵⁵ At 2 years after surgery, the OPLL had progressed by 2.4 mm rostrally and 2.4 mm caudally. The OPLL had increased in thickness in 22% to 53% of patients at 5 to 13 years postoperatively.^33,^57,⁵⁸ The average total increase in thickness in these patients ranged from 1.3 to 3.5 mm during the follow-up period. Computer-assisted radiographic analysis found that OPLL thickness had increased by 1.1 mm at 1 year postoperatively and by 1.4 mm at 2 years.⁵⁵

Despite a 70% or greater rate of OPLL progression after posterior decompression, less than 10% of patients demonstrated worsening myelopathy or late neurological deterioration after surgery because of OPLL growth.^43,⁴⁹ The delayed neurological deterioration occurred predominantly as a result of trauma or the development of OPLL or OLF at a site other than the originally operated levels. Consequently, studies have failed to demonstrate an association between OPLL growth and the postoperative score-based recovery rate.^57,⁵⁸ Kawaguchi and coworkers found that 73% of patients after en bloc cervical laminoplasty had radiographic evidence of OPLL progression; however, only 7% of patients worsened as a result of increased thickness of the OPLL.⁵⁸ Other studies have observed that OPLL growth causes late neurological deterioration in only 3% to 4% of patients after posterior decompression.^33,⁴³ These findings suggest that although OPLL may continue to grow after posterior decompression, neurological recovery does not appear to be clinically affected by the slow progression.

Thoracic Disease

Though less common than cervical OPLL, thoracic OPLL presents a higher risk for severe myelopathy and potentially permanent neurological dysfunction. Patients with thoracic myelopathy secondary to OPLL may have worse neurological compromise than patients with cervical disease because of the smaller thoracic spinal canal and the natural thoracic kyphosis that drapes the spinal cord over the ventral pathology. Managing patients with thoracic OPLL and myelopathy is complicated by the absence of effective conservative measures and the generally rapid progression of neurological deterioration once symptoms arise. Surgical treatment of thoracic OPLL is technically challenging because of difficulty obtaining access to the ventral thoracic spinal column, inability to achieve indirect posterior decompression, a high risk for postoperative neurological deterioration, and the often concurrent existence of thoracic OLF.

Laminectomy for indirect posterior decompression of thoracic OPLL was studied by Tokuhashi and colleagues.⁵⁹ They performed intraoperative ultrasonography after laminectomy to assess for the presence of cerebrospinal fluid ventral to the spinal cord as an indication that the spinal cord had migrated posteriorly and was decompressed. In patients with positive ultrasound evidence of effective posterior decompression, the JOA score improved from 4.3 preoperatively to 8.6 at final follow-up, with a recovery rate of 62.7%. Alternatively, 67% of patients with ultrasonography that showed ineffective decompression demonstrated significant postoperative neurological deterioration.

To identify preoperative factors that predict optimal clinical outcomes, the authors compared the angle of thoracic kyphosis in patients with and without ultrasound evidence of decompression. They measured the angle from the superior posterior margin at the rostral level of the decompression and from the lower posterior margin at the caudal level of the decompression to the prominence of the maximum OPLL. They termed this the ossification-kyphosis angle and found that patients with effective decompression had an angle of less than 23 degrees. All patients with ineffective decompression had an ossification-kyphosis angle of greater than 23 degrees. The authors concluded that patients with an ossification-kyphosis angle of less than 23 degrees can be effectively treated with laminectomy alone; however, patients with an angle of greater than 23 degrees require anterior decompression.

Alternatively, Kawahara and colleagues described a method for circumspinal decompression of thoracic OPLL.⁶⁰ They suggested that anterior decompression is the most effective measure for relieving ventral compression of the thoracic cord; however, anterior decompression is technically challenging and is not sufficient in patients with concomitant OLF. Kawahara and coworkers described an initial posterior decompression in which laminectomy is performed at least one level above and below the length of the thoracic OPLL. Parallel gutters are then drilled on either side of the thecal sac ventrally into the vertebral bodies approximately 1 cm deep. Posterior segmental instrumentation is then placed and connected to rods that are underbent approximately 15 to 20 degrees less than the patient’s natural kyphosis to reduce the kyphosis angle and facilitate dorsal migration of the spinal cord.

After posterior decompression, imaging is performed at 3 weeks to assess for adequate decompression. If the thoracic spinal cord remains tethered over the ventral OPLL, thoracotomy is performed and the posterior third of the vertebral bodies at the affected levels are drilled to connect the two parallel gutters created in the initial posterior procedure. The OPLL lesion is then resected from the thecal sac or allowed to float anteriorly into the corpectomy defect. Bone graft is then placed in the defect for anterior arthrodesis. In 11 patients who underwent circumspinal decompression and dekyphosis stabilization, the JOA score improved from 4.0 preoperatively to 9.1 at last follow-up. Only 1 patient demonstrated postoperative neurological deterioration.

Comparing various approaches for the treatment of thoracic OPLL, Yamazaki and coauthors evaluated patients who underwent laminectomy alone, laminectomy with posterior instrumented fusion, and OPLL extirpation via either thoracotomy or posterior resection of OPLL.⁶¹ The three groups had comparable mean preoperative JOA scores (3.4 to 3.6) and postoperative JOA scores at 1 year (7.7 to 7.9). However, a significant percentage of patients in the group that underwent laminectomy alone (39%) demonstrated late neurological deterioration, with a mean JOA score at final follow-up of 6.5. The groups that underwent OPLL extirpation and laminectomy with posterior fusion remained stable at final follow-up with a JOA score of 8 and no evidence of late deterioration. Both the laminectomy-alone and OPLL extirpation groups had patients in whom paralysis developed postoperatively. In the laminectomy-alone group, two of the three patients recovered with a secondary posterior instrumentation procedure to stabilize the spine. In the OPLL extirpation group, the three patients with postoperative paralysis had already been severely weak and wheelchair bound preoperatively. In no patients in the group that underwent laminectomy with posterior fusion did postoperative paralysis develop.

In analysis of their results, Yamazaki and colleagues recommended that laminectomy alone should not be performed for thoracic OPLL given the considerable risk for postoperative paralysis and late neurological deterioration.⁶¹ They suggest that patients who are amenable to posterior decompression should also undergo supplemental stabilization with a posterior instrumented fusion. Patients treated by OPLL extirpation demonstrated a high degree of neurological recovery that remained stable during follow-up, and this should be the primary surgical modality for patients who are myelopathic. In severely myelopathic patients who are nonambulatory, the authors recommend posterior decompression with instrumented fusion because they found that these patients are at high risk for postoperative paralysis from OPLL extirpation. Laminectomy with posterior fusion may also be considered for patients who are at high surgical risk when undergoing an anterior procedure.