Long-term Outcomes of Multilevel Anterior Cervical Osteotomy and Posterior Instrumentation for OPLL-Induced Myelopathy With Cervical Kyphosis
Article information
Abstract
Graphical Abstract
Abstract
Objective
To analyze long-term clinical and radiological outcomes after multilevel anterior osteotomy with posterior instrumentation in patients with ossification of posterior longitudinal ligament (OPLL)-induced myelopathy and cervical kyphosis.
Methods
Patients who underwent multilevel anterior osteotomy with posterior instrumentation for OPLL-induced myelopathy and cervical kyphosis and had a minimum of 5-year follow-up were included. Clinical outcomes (Japanese Orthopaedic Association score system for cervical myelopathy [C-JOA], 12-item Short Form health survey [SF-12], Neck Disability Index [NDI]) and radiological parameters (C2–7 lordosis, center of gravity of the head [CGH]-C7 sagittal vertical axis [SVA], T1 slope) were analyzed at the preoperative, immediate postoperative, and latest follow-up timepoints.
Results
Twenty-eight patients were included. The average follow-up period was 66.4 months. All clinical outcome parameters showed significant improvement. C-JOA, SF-12, and NDI showed significant improvement at latest follow-up (p<0.001). C2–7 lordosis increased significantly immediately postoperatively (-6.0°±10.4°) compared to preoperatively (+9.2°±9.6°), and was largely maintained at latest follow-up (-5.7°±9.4°). T1 slope significantly increased between the immediate postoperative timepoint (21.9°±7.7°) and latest follow-up (24.2°±9.5°) (p=0.046). CGH-C7 SVA significantly increased between the immediate postoperative timepoint (22.7±14.8 mm) and latest follow-up (32.2±22.6 mm) (p=0.046).
Conclusion
Multilevel anterior osteotomy with posterior instrumentation is a safe and effective surgical option for OPLL-induced myelopathy with kyphotic cervical alignment. Future studies are required to investigate the forward tilting of cervical spine over time after surgery.
INTRODUCTION
Ossification of the posterior longitudinal ligament (OPLL)-induced myelopathy can cause irreversible neurological damage and is an indication for surgery [1]. Surgical strategies for multilevel cervical OPLL include posterior laminoplasty/laminectomy and fusion for indirect decompression and multilevel anterior osteotomy for direct decompression [2]. Posterior approach is usually preferred in patients with >3 affected spinal levels and high age. However, as the posterior approach relies on posterior drift of the spinal cord, it is ineffective in cases with kyphotic cervical alignment and OPLL with canal occupancy ratio of >60% [3]. Anterior surgery allows for direct removal of OPLL and cervical realignment and hence is more suitable for these cases. The K-line index is also useful in deciding between anterior and posterior approaches. Anterior approach is preferred in K-line (-) cases as the posterior approach can further aggravate preexiting kyphosis [4].
Although a combined approach with anterior osteotomy and posterior stabilization has been reported to be the most appropriate for multilevel OPLL with kyphosis [5], no previous study has analyzed long-term outcomes after this surgery. As such, there is a lack of evidence regarding maintenance of clinical improvement and cervical realignment over long term after surgery. The purpose of the study was, therefore, to fill this gap in the literature by assessing long-term clinical and radiological outcomes of multilevel anterior osteotomy and posterior stabilization for multilevel OPLL with kyphotic cervical spine.
MATERIALS AND METHODS
1. Study Design and Population
This was a retrospective review of prospectively collected data approved by the Institutional Review Board (IRB) of Wooridul Spine Hospital (IRB No. 2012-100). Consecutive patients who underwent multilevel anterior osteotomy plus posterior stabilization for multilevel OPLL-induced myelopathy with kyphotic cervical spine alignment at a single spine hospital between January 2012 and January 2016 were included. All patients had a minimum of 5-year follow-up. Patients with lordotic sagittal alignment (C2–7 lordosis<0°), K-line (+), osteoporosis (bone mineral density (T-score<-2.5), and history of cervical operation were excluded. Osteoporotic patients (bone mineral density T-score<-2.5) were excluded due to the increased risk of implant-related complications and correction loss, even with posterior fixation. Although posterior fixation enhances stability, significant osteoporosis still poses high risk for subsidence and implant loosening, potentially confounding the results. Thus, we excluded these patients to maintain homogeneity and ensure validity of outcomes. Patients with cervical kyphotic deformity, myelopathy symptoms, and K-line negative status uniformly underwent combined anterior multilevel osteotomy and posterior fixation. The decision was consistent across the included cohort, following standardized surgical protocols followed at our institution. Patient selection flowchart is shown in Fig. 1.
Flowchart explaining patient selection. CSM, cervical spondylotic myelopathy; OPLL, ossification of posterior longitudinal ligament; CL, C2–7 lordosis; CGH, center of gravity of the head; SVA, sagittal vertical axis; C-JOA, Japanese Orthopaedic Association score system for cervical myelopathy; SF-12, 12-item Short Form health survey; NDI, Neck Disability Index; ASD, adjacent segment degeneration.
2. Surgical Technique
A standard transverse incision along the anterior neck crease was typically made. For extensive 4-level surgeries, the incision length was extended vertically along the medial border of the sternocleidomastoid muscle to adequately expose the surgical field. Grades 3 and 4 anterior osteotomies were performed according to the Ames cervical osteotomy classification [6]. Grade 3 is a partial or complete corpectomy and grade 4 is a complete uncovertebral joint resection to the transverse foramen. Grade 4 osteotomy was performed to ensure sufficient release of bilateral foramen, and grade 3 osteotomy was performed in continuous type OPLL-induced myelopathy patients. Posterior instrumentation was done in form of lateral mass or pedicle screws and rods. We utilized lordotic cages made from polyetheretherketone to achieve and maintain the lordotic curvature. These cages were typically available in angles ranging from 6° to 8° and filled with autologous bone graft harvested during osteotomies. The anterior approach (osteotomies, decompression, and cage insertion) was performed first. Following repositioning, posterior fixation was performed for added stability. The posterior procedure was strictly limited to instrumentation without additional osteotomies. All surgeries were completed in a single session with intraoperative repositioning between the anterior and posterior procedures.
3. Data Collection
Following data were collected: (1) Demographic variables: age, sex, follow-up duration. (2) Operative variables: osteotomy grade and levels, posterior stabilization levels, lower instrumented vertebra (LIV). (3) Clinical outcomes: patient reported outcome measures (PROMs) included Japanese Orthopaedic Association score for cervical myelopathy (C-JOA), 12-item Short Form health survey (SF-12) score, and Neck Disability Index (NDI). C-JOA was used to evaluate the clinical recovery in myelopathy, SF-12 was used to evaluate the patient’s quality of life, and NDI was used to evaluate pain and disability of the neck. These PROMs were recorded preoperatively and at latest followup. (4) Radiological outcomes: (i) Cervical alignment parameters: C2-7 lordosis, center of gravity of the head (CGH)-C7 sagittal vertical axis (SVA) [7], T1 slope (Fig. 2). These were measures on lateral radiographs at the preoperative, immediate postoperative, and latest follow-up timepoints. For C2-7 lordosis, (+) denoted kyphotic angle and (-) denoted lordotic angle. (ii) Subsidence and implant failure: these were assessed on radiographs and computed tomography (CT) at the latest follow-up. Subsidence was defined as >3 mm loss in anterior or posterior interbody space height when comparing immediate postoperative and latest follow-up radiographs. (iii) Adjacent segment degeneration (ASD): presence of ASD was evaluated on magnetic resonance imaging by assessing the disc height, disc signal intensity, disc protrusion, central and foraminal stenosis, and narrowing of disc space.
Radiologic parameters. ① C2–7 lordosis: angle between the lower endplate of the C2 vertebra and lower endplate of the C7 vertebra. ② Center of gravity of the head-C7 sagittal vertical axis: interval between the plumb line of the anterior margin of the external auditory canal and the posterior-cranial corner of the C7 vertebral body. ③ T1 slope: angle formed by drawing a line along the superior endplate of T1 and horizontal reference line.
4. Statistical Analysis
Statistical analyses were performed using IBM SPSS Statistics ver. 19.0 (IBM Co., USA). Changes in the postoperative radiologic parameters and clinical outcome parameters were compared using paired t-tests. A p-value of <0.05 was deemed as statistically significant.
RESULTS
Twenty-eight patients were included. The mean age was 59.9±10.2 years. Twenty-four patients (85.7%) were male and 4 (14.3%) were female. The mean follow-up period was 66.4±8.6 months. Twenty-one patients (75%) underwent grade 3 osteotomy and 7 patients (25%) underwent grade 4 osteotomy. Posterior instrumentation ranged from 3 to 5 spinal segments. The LIV was C7 in 24 patients (85.7%) and C6 in 4 patients (14.3%). The average total operative time was 210±60.2 minutes, with a mean estimated blood loss of 73.2±55.6 mL (Table 1).
Demographic data (n=28)
C2–7 lordosis significantly increased at the immediate postoperative timepoint (-6.0°±10.4°) compared to preoperative (-9.2°±9.6°) (p=0.025). This improvement in C2–7 lordosis was maintained at the latest follow-up (-5.7°±9.4°) and no significant difference was seen in C2–7 lordosis between the immediate postoperative and latest follow-up timepoints (p=0.842). T1 slope did not increase significantly at the immediate postoperative timepoint (21.9°±7.7°) compared to preoperative (19.5°±8.8°) (p=0.103). However, statistically significant increase was seen in T1 slope at the latest follow-up (24.2°±9.5) compared to immediate postoperative (p=0.046). CGH-C7 SVA did not increase significantly at the immediate postoperative timepoint (22.7±14.8 mm) compared to preoperative (19.1±21.3 mm) (p=0.326). However, statistically significant increase was seen in CGH-C7 SVA at the latest follow-up (32.2±22.6 mm) compared to immediate postoperative (p=0.046) (Table 2). None of the patients showed ASD or subsidence until the latest follow-up.
Changes in radiologic parameters
C-JOA score significantly improved from 10.7±0.4 preoperatively to 15.8±0.4 at latest follow-up (p<0.001). SF-12 score significantly improved from 34.1±3.6 preoperatively to 44±4.1 at latest follow-up (p<0.001). NDI score significantly improved from 38.6±6.6 preoperatively to 15.9±4.5 at latest follow-up (p<0.001) (Table 3). There were no preoperative or postoperative complications, such as hematoma, screw malposition, esophageal injury, thoracic duct injury, vessel injury, and recurrent laryngeal nerve injury.
Changes in clinical outcome parameters
1. Illustrative Cases
1) Case 1 (Ames cervical osteotomy classification grade 3 - partial or complete corpectomy)
A 48-year-old gentleman was diagnosed with OPLL at C4–6 causing myelopathy (Fig. 3). He underwent Ames anterior cervical osteotomy grade 3 at C4–6 levels with posterior instrumentation at C3–7. C2–7 lordosis increased to -1.1° at the immediate postoperative timepoint compared to +5.2° preoperatively (-1.1°) and remained almost unchanged (-0.9°) at latest followup (Fig. 4A–C). T1 slope increased from 20.7° preoperatively to 26.3° at the immediate postoperative timepoint and 29.4° at latest follow-up (Fig. 4A–C). CGH-C7 SVA increased from 44.3 mm at the immediate postoperative timepoint to 47.1 mm at latest follow-up (Fig. 4B, C). There was no subsidence or implant failure until latest follow-up (Figs. 3F and 4C). C-JOA, SF-12, and NDI scores improved postoperatively.
Computed tomography (CT), magnetic resonance imaging (MRI), and myelogram of case 1. Sagittal images of preoperative CT (A), MRI (B), and coronal image of MRI myelography (C) showed ossification of posterior longitudinal ligament at C4–6 level and kyphotic cervical alignment. Postoperative sagittal images of CT (D) and MRI (E) showed adequately decompressed with correction of kyphosis.
Radiographs of case 1. Lateral (A–C) and anteroposterior (D–F) views. Ames anterior cervical osteotomy grade 3 at C4–6 levels with posterior instrumentation at C3–7. (A–C) C2–7 lordosis increased to -1.1° at the immediate postoperative timepoint compared to +5.2° preoperatively (-1.1°) and remained almost unchanged (-0.9°) at latest follow-up. (A–C) T1 slope increased from 20.7° preoperatively to 26.3° at the immediate postoperative timepoint and 29.4° at latest follow-up. (B, C) CGH-C7 SVA increased from 44.3 mm at the immediate postoperative timepoint to 47.1 mm at latest follow-up. Disc height increased on the immediate postoperative anteroposterior (AP) image (E) compared with the preoperative AP image (D), and no evidence of subsidence or implant failure was observed until the latest follow-up (C, F). CGH, center of gravity of the head; SVA, sagittal vertical axis; preop, preoperative; postop, postoperative; AP, anteroposterior.
2) Case 2 (Ames cervical osteotomy classification grade 4 - complete uncovertebral joint resection to the transverse foramen)
A 62-year-old lady progressive paraparesis for 2 years. Imaging showed C3–6 OPLL. Patient underwent grade 4 osteotomy at C3–6 with posterior instrumentation. C2–7 lordosis increased to -3.9° at the immediate postoperative timepoint compared to +10.8° preoperatively (-1.1°) and remained almost unchanged (-3.8°) at latest follow-up (Fig. 4A–C). T1 slope increased from 22.8° preoperatively to 26.2° at the immediate postoperative timepoint and 31.3° at latest follow-up (Fig. 5A–C). CGH-C7 SVA increased from 3 mm at the immediate postoperative timepoint to 10.5 mm at latest follow-up (Fig. 5B, C). There was no subsidence or implant failure until latest follow-up (Figs. 2F and 5C). C-JOA, SF-12, and NDI scores improved postoperatively.
Radiographs of case 2. Grade 4 osteotomy at C3–6 with posterior instrumentation. (A–C) C2–7 lordosis increased to -3.9° at the immediate postoperative timepoint compared to +10.8° preoperatively (-1.1°) and remained almost unchanged (-3.8°) at latest follow-up. (A–C) T1 slope increased from 22.8° preoperatively to 26.2° at the immediate postoperative timepoint and 31.3° at latest follow-up. (B, C) CGH-C7 SVA increased from 3 mm at the immediate postoperative timepoint to 10.5 mm at latest follow-up. Disc height increased on the immediate postoperative anteroposterior (AP) image (E) compared with the preoperative AP image (D), and no evidence of subsidence or implant failure was observed until the latest follow-up (C, F). CGH, center of gravity of the head; SVA, sagittal vertical axis; preop, preoperative; postop, postoperative; AP, anteroposterior.
2. Best Correction Case
Cervical lordosis correction from +15.0° (preoperative) to -12.0° (postoperative) with favorable T1 slope (minimal increase from 19° to 21°). This optimal result was likely due to the patient’s adequate preoperative cervical flexibility and moderate kyphosis severity.
3. Worst Correction Case
Minimal cervical lordosis correction from +3.0° (preoperative) to +1.0° (postoperative) with considerable increase in SVA and T1 slope. This was likely attributable to severe preoperative stiffness, extensive degenerative changes, and poor bone quality that limited the achievable correction.
DISCUSSION
OPLL needs surgical intervention when presenting with cervical myelopathy. Previous studies have shown that laminoplasty is sufficiently effective in patients with multilevel cervical myelopathy [7-9]. However, the effectiveness of laminoplasty is dependent on posterior drift of the cord and therefore it is not an appropriate option for patients with kyphotic cervical alignment, especially >11°. Laminoplasty has also been shown to worsen malalignment after surgery [10-12]. In this study, we performed multilevel anterior osteotomy combined with posterior instrumentation for myelopathy caused by multilevel OPLL in presence of kyphotic alignment. The findings suggested that the surgery resulted in favorable clinical and radiological outcomes.
The study demonstrated a significant improvement in C2–7 lordosis at the immediate postoperative timepoint that was maintained until the latest follow-up. However, it is important to note that the average correction in the lordotic angle with anterior osteotomy of 3 levels or more was only about 15°. This insufficiently corrected angle is most likely due to failure of setting target values for various indicators such as horizontal gaze, chinbrow angle, and CGH-C7. Surgeries were performed with the aim of correcting the kyphotic alignment to a lordotic alignment. Although insufficient angle correction was obtained, the corrected angle was found to be well maintained at long-term follow-up. This maintenance of alignment correction was likely due to the stability provided by posterior instrumentation. Additionally, none of our patients had implant failure or subsidence over long term. Previous studies reported high rates of subsidence after multilevel anterior osteotomy [13,14]. We believe that this contrast in findings is due to supplementation with posterior stabilization in the current study.
Although the lordosis correction was maintained and the construct remained stable, the entire cervical spine above the cervicothoracic junction (CTJ) was seen to gradually tilt forward over long term (Fig. 6). This was seen likely due to insufficient correction of C2–7 lordosis and T1 slope-lordosis mismatch. Hyun et al. [15] demonstrated that neck disability worsens with progressive cervical sagittal malalignment following cervical reconstructive surgery. They reported that T1 slope-lordosis mismatch >26.1° may indicate cervical sagittal malalignment. The current study demonstrated that T1 slope increased gradually over long term. As such, the possibility of worsening of cervical alignment due to increase in T1 slope and therefore T1 slope-lordosis mismatch over time should be considered.
Long-term changes of radiologic parameters. C2–7 lordosis is increased at the immediate postoperative timepoint (B) compared to preoperative (A). (C) Even after long-term follow-up, its value remains almost unchanged. (B, C) T1 slope gradually increased over long term. The value of center of gravity of the head-C7 sagittal vertical axis showed a significant increase from the immediate postoperative period (B) to the long-term follow-up (C). Preop, preoperative; postop, postoperative.
Inoue et al. [16] conducted a cross-sectional study of 388 asymptomatic subjects to assess age-related changes in T1 slope and found that it significantly increased with age. In a study of 478 subjects conducted by Lamas et al. [17], T1 slope was found to be significantly higher in patients >60 years compared to 40–60 years. Although these studies show that there is a natural progression in T1 slope with age, it is difficult to deduce the rate of progression as there is no evidence regarding it. In our study, the T1 slope increased significantly in 5 years after surgery which we believe is a higher rate than the natural progression. Xie et al. [18] conducted a study of 127 patients who underwent single or multilevel anterior cervical discectomy and fusion (ACDF) and found that there was no significant difference in T1 slope at the immediate postoperative and latest follow-up timepoints. They also reported that T1 slope had a significant correlation with loss of C2–7 lordosis after ACDF. The mean follow-up in this study was 43.7 months. Das et al. [19] conducted a retrospective cohort study of 265 patients with adult cervical deformity and found that the undercorrected cohort had significantly higher chances of distal junctional kyphosis (DJK). Ani et al. [20] conducted a retrospective review of 69 cervical deformity patients and similarly concluded that severe DJK patients had insufficient correction.
Apart from insufficient lordosis correction, the forward tilting of the cervical spine over time can also be due to an overload on the CTJ to bear the weight of the head. Normally, these loads are divided and distributed amongst the cervical spine segments. However, in cases with C6 or C7 as the LIV, these loads are supported mainly by the CTJ due to a fixed cervical spine. Additionally, weakness of the cervical extensor muscles may also be associated with this phenomenon. Yoon et al. [21] demonstrated that the cervical lordotic angle was positively correlated with the ratio of flexor to extensor muscle cross-sectional areas. They concluded that there is a significant relationship between cervical muscle imbalance, including extensor muscle weakness, and loss of cervical lordosis. It is not clear from the current study whether the increase in T1 slope is due to stoppage of instrumentation at C6–7 or due to a natural phenomenon caused by the aging process during the follow-up period.
This study has several limitations. The retrospective design limits the level of evidence and single-center data reduces generalizability. The sample size was small (n=28) and treatment options were heterogeneous. A subgroup analysis of outcomes based on type of osteotomy would have been valuable. However, due to the low number of patients in each osteotomy subgroup, this comparative analysis was not feasible. Future studies with larger cohorts are required for such subgroup analyses that can provide more granular information. Additionally, assessment of bony fusion would have been valuable considering this was a fusion procedure. However, we lacked the data for fusion assessment. Despite the limitations, the current study shows the effective correction and long-term maintenance after multilevel anterior osteotomy and posterior instrumentation for OPLL with kyphotic alignment.
CONCLUSION
Multilevel anterior osteotomy with posterior instrumentation is a safe and effective surgical option for OPLL-induced myelopathy with kyphotic cervical alignment. Future studies are required to investigate the forward tilting of cervical spine over time after surgery.
Notes
Conflict of Interest
The authors have nothing to disclose.
Funding/Support
This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Author Contribution
Conceptualization: SJK, JB; Data curation: SJK; Formal analysis: SJK; Methodology: SJK, JB; Project administration: SHL, JB; Visualization: SHL; Writing – original draft: SJK, PS; Writing – review & editing: SJK, PS, SHL, JB.