Results of Cervical Laminoplasties

Because the majority of laminoplasty operations are performed to treat cervical myelopathy, the preoperative and postoperative conditions are frequently scored using the JOA scoring system for cervical myelopathy (Table 157-2). Several studies comparing cervical corpectomy and several different laminoplasty techniques determined that both techniques are effective in treating cervical myelopathy. The recovery rates for myelopathy were no different statistically between either procedure.^15–17

Table 157-2 Japanese Orthopaedic Association Score

Comparison of the single-door versus double-door laminoplasty techniques shows no significant difference in the outcome between these two groups.¹⁸

Comparison of laminectomy, anterior decompression and fusion, and laminoplasty showed no significant difference in outcome as measured by change in JOA scores.^4,19–23 A comparison of subtotal corpectomy and laminoplasty showed identical results from the two procedures at 1- and 5-year follow-up evaluations. There may be a significant difference in outcome in the special case of massive ossification of the posterior longitudinal ligament, where over 50% of the canal is occupied by ossified posterior longitudinal ligament (OPLL). In this condition, the outcome was much better for the anterior approach patients because none of them worsened neurologically, as compared with a 33% neurologic complication rate for the laminoplasty patients.²⁴

Laminoplasties can provide rapid improvement of myelopathic symptoms within the first postoperative year as measured via JOA myelopathy scores (Tables 157-2 and 157-3, and Figs. 157-1 and 157-2). Factors predicting recovery have been noted. Urinary bladder function recovers less than other symptoms. Patients older than 60 years improve less than younger patients. Preoperative JOA score, Pavlov ratio, and compression ratio also affect outcomes. Lower JOA scores, lower Pavlov ratios, and lower compression ratios all correlate with less improvement after open-door laminoplasties.^3,25–28 Abnormal spinal cord signal on preoperative magnetic resonance imaging (MRI) scanning has been associated with poor postoperative prognosis.²⁹ Local kyphosis exceeding 13 degrees is a crucial risk factor for poor recovery after laminoplasty. If kyphosis exceeding 13 degrees is detected, either anterior or posterior fusion should be considered³⁰ (Fig. 157-3). Transverse area of the spinal cord at the level of maximum compression is also related to outcome, since areas of 42.6 mm² correlated with excellent outcomes and areas of 31 mm² correlated only with good outcomes in postoperative JOA scores.³¹ It appears that the ideal enlargement of the canal is 4 mm along an anterior posterior line, and that a 3-mm posterior cord shift is required for clinical improvement.⁶ Reoccurrence of myelopathy postoperatively due to posterior shifting of the cord into the area of split lamina has been reported. This rare event usually occurs in the first few days after laminoplasty and responds to laminectomy of the affected segments.³²

Table 157-3 Recovery Rate

FIGURE 157-1 Compression ratio is the sagittal cord diameter divided by the transverse cord diameter times 100.

FIGURE 157-2 The Pavlov ratio, also known as the Torg-Pavlov ratio. The ratio is the canal size divided by the vertebral body size. This is determined by using the lateral plain radiograph, and measuring the distance between the posterior vertebral body line and the spinolaminar line, divided by the distance between the anterior vertebral body line and the posterior vertebral body line. The normal value is 1.0 or greater, and stenosis is denoted by a ratio of 0.82 or less.

FIGURE 157-3 Method of measuring lordosis: the angle formed by a line parallel to the posterior body of C2 and a line parallel to the posterior body of C7. A positive value indicates lordosis, a negative-value kyphosis.

Risk factors for failure of adequate decompression with laminoplasty have been determined in cases of OPLL (Fig. 157-4). Maximum thickness of the ossification of more than 7 mm, lordosis of less than 10 degrees, or any kyphosis are significantly correlated with continued cord contact in the postlaminoplasty state. If these risk factors are present, then anterior decompression should be considered.^26,33,34 It is recommended that patients be treated surgically as soon as myelopathy is detected. Patients with OPLL usually have improved JOA scores in the postlaminoplasty state, and those last for at least 5 years before myelopathy progresses. Ossification of the posterior longitudinal ligament progresses in 73% of patients after laminoplasty, and younger patients with the mixed or continuous type of OPLL progress most often.³³ Progression of OPLL cannot be predicted after laminoplasty, but if symptomatic progression occurs, an anterior approach may be considered. Late neurologic deterioration may be seen in 18% to 60% of cases after cervical laminoplasty, and may be due to instability or development of kyphosis.³⁵

FIGURE 157-4 Ossified posterior longitudinal ligament types: continuous, segmental, mixed, and localized.

Segmental motor paralysis occurs in 7% to 20% of patients after laminoplasty. Radiculopathy most often appears within 2 weeks of surgery, and is evenly split between the hinge and open-door side of the laminoplasty when using the open-door technique. The C5 nerve root appears to be particularly predisposed to postoperative radiculopathy. Performing preoperative electromyogram (EMG) may prevent postoperative C5 palsy. If radiculopathy is found on EMG, selective microforaminotomies should be performed during the laminoplasty procedure.^36,37 There is evidence to suggest that segmental motor paralysis may be due to changes in the central gray matter of the spinal cord, rather than in the nerve root. In those cases, there is a definite delay from the end of surgery to the onset of deficit, and MRI scan may show a high-intensity zone in the spinal cord on T2 images.³⁸ The mechanism is unknown, but may be attributable to surgical trauma or possibly due to rapid reperfusion into the ischemic central gray of the cord.³⁹ Other causes of severe postoperative neck pain or motor paralysis may be due to malposition of osteotomized lamina, surgical trauma, or reclosure of opened lamina.² Some patients may develop kyphosis after laminoplasty, and those who develop kyphosis tended to have a poor recovery. Additionally, patients who had preexisting kyphosis had less reversal of myelopathy.^40,41 Some degree of lordosis is lost in 87% of patients after laminoplasty, and this appears to be related to detachment of the semispinalis cervicis muscle on C2, and the nuchal ligament at C6–C7. For this reason, it is recommended that the semispinalis cervicis be left attached whenever possible.^42,43 Laminoplasty techniques have been developed to spare the muscular and ligamentous attachments at the spinous processes. Despite this modification, 26.9% of patients continued to complain of axial neck pain postoperatively, and no good data yet exist to indicate that the extra time spent with this technique is beneficial to patient outcome.⁴¹ One retrospective study of double-door laminoplasty with reconstruction of the extensor musculature found that there was no influence in either the development or resolution of axial pain symptoms. This suggests that muscle reconstruction techniques may have merit in decreasing axial pain postoperatively after laminoplasty. It appears that preventing denervation of the paraspinal muscles by limiting dissection to the lateral border of the facet complex, reconstructing the extensor musculature, avoiding damage to the facet joint when making the lateral gutter hinges, and providing muscular exercise in the postoperative period all contribute to less axial neck pain.⁴⁴

Fusion spontaneously occurs in patients after laminoplasty and may involve multiple levels. Intervertebral body fusion has been reported in 18% of patients who have undergone laminoplasty, and interlaminar fusion has been seen in 80% of patients who have undergone laminoplasty.⁴⁵

Neck pain exists in up to 80% of patients after cervical laminoplasty, and seems to be greater in those who have more operated levels, and in those who have less range of motion postoperatively.⁴⁶ This may be related to damage to the facet capsules, or the paravertebral muscles. The nuchal muscles may be reduced to 80% of their preoperative size in those patients who have undergone laminoplasty.⁴⁷ This is likely due to denervation of the musculature from dissecting lateral to the facet joints, and may play a role in persistent postoperative neck pain. Neck pain appears to correlate with an imbalance in the ratio of strength between the flexor and extensor muscles in the postlaminoplasty state.⁴⁸ Comparing laminoplasty to anterior fusion for the incidence of postoperative chronic neck and shoulder pain reveals the incidence is 60% in the laminoplasty group as compared with 19% in anterior cervical fusion.¹¹

Occupational recovery after cervical laminoplasty for OPLL depends on the severity of preoperative myelopathy, and the degree of postoperative recovery. Return to work rates vary depending on job demands, with 95% of light sedentary laborers, 50% of standing laborers, and 35% of heavy laborers returning.⁴⁹

Long-term follow-up of open-door laminoplasty indicates that the canal size increases 48% initially, and decreases slightly to 40% after 5 years. The average loss of motion on flexion is 35%, and the average loss of motion on extension is 57%.⁵⁰ Approximately 8% of patients have preserved neck motion after laminoplasty. Ten years after laminoplasty by the double-door technique, 78% of patients maintain their initial improvement in degree of myelopathy.⁵¹ Causes of late deterioration in patients with OPLL who have undergone laminoplasty include thoracic myelopathy from ossification of the yellow ligament, minor trauma, and rarely progression of OPLL at the operative levels.^2,45 Adjacent segment degeneration can occur after laminoplasty, but compared with anterior cervical fusion, laminoplasty may present fewer predispositions for development of adjacent segment degeneration.⁵²

Types of Cervical Laminoplasty

Five basic types of laminoplasty exist, but only three have gained in popularity. The earliest form of laminoplasty was called the Z-plasty type (Fig. 157-5). In this type of laminoplasty the spinous processes are removed, and the lamina are thinned to the inner cortical region. An opening is made along the long axis of the lamina, and two lateral troughs are made. The canal is augmented by then shifting the lamina and securing them to each other to maintain the new alignment. This procedure has been largely abandoned as technically demanding without superior results, as compared with the other techniques. The open-door or Hirabayashi-type laminoplasty utilizes a thinned hinge on one side of the lamina, and a complete cut through the lamina on the opposite side (Fig. 157-6). The double-door or Kurokawa-type laminoplasty uses hinges fashioned bilaterally just medial to the junction of the lamina and facet, and a midline bicortical cut (Fig. 157-7