Keywords
cervical PEEK cage - cervical interbody cage - zero-profile anchored spacer - ROI-C - adjacent segment disease - self-locking stand-alone cage
Introduction
Anterior cervical discectomy and fusion (ACDF) in degenerative disc disease has widely been accepted nowadays after modifications of surgical techniques originated by Smith and Robinson[1] and by Cloward.[2] Autologous bone graft use provides high percentage of fusion rate over allografts.[3] However, autologous bone grafts obtained from the anterior iliac crest are associated with significant donor site morbidity and complications including severe acute pain, hematoma formation, infection, meralgia paresthetica, and chronic pain.[4]
Cervical interbody cages have been developed and applied in clinical practice to eliminate complications at the donor site. There are two basic types of cages: threaded hollow cylindrical cages (Cloward type procedure) and rectangular cages (Smith–Robinson type procedure). The threaded cages are introduced and screwed through the endplates of the vertebral bodies, whereas the rectangular cages mimic the intervertebral space dimensions and in accordance with the anatomy of the endplates. The cages are made of titanium, carbon fiber, or polyetheretherketone (PEEK).[5]
[6] Currently, PEEK cages have widely been used because the elastic modulus of the PEEK cage is similar to that of human cortical bone that may help to decrease stress shielding and increase bony fusion.[7]
Anterior plate system was added in ACDF using with autologous iliac crest graft or cages, especially in multilevel surgery, to increase fusion rates and reduce the problem of graft extrusion and collapse.[8] Nonetheless, some studies reported complications such as plate migration, screw loosening, or back-out, soft tissue injury and dysphagia.[9]
[10] Bioabsorbable anterior cervical plating then has been introduced to reduce some of the long-term complications and imaging artifacts associated with titanium instrumentation.[11] However, it is still not routinely used because of higher costs compared with the titanium plate.
Stand-alone cage concept, initiated by Bagby,[12] used in the human spine since 1988, was started with a stainless-steel basket implant for cervical spine surgery in horses, working with veterinarians. Stand-alone cages have been designed to avoid hardware-related complications as mentioned earlier. However, this concept was warned due to high incidence of subsidence, especially titanium cervical cage.[13] Stand-alone cervical PEEK cage for single- to two-level degenerative disc disease has been used with accepted incidence of subsidence.[14] Subsequently, zero-profile cage-plate device with locking screws has been developed and used clinically to reduce incidence of subsidence and dysphagia. However, there was some reports about device-related complications such as loosening of screws, and device malposition with a screw threatening the vertebral artery.[15]
[16]
The ROI-C implant system (Zimmer Biomet, Austin, Texas, United States) is a new type of stand-alone anchored spacer designed for implant as an intervertebral spacer through the anterior cervical approach. This system is composed of an anatomical PEEK cage and two VerteBRIDGE self-locking plates.[17]
[18]
The purpose of the present study was to assess clinical and radiological outcomes of the ROI-C device for cervical adjacent segment disease.
Materials and Methods
Subjects
The patients harboring cervical adjacent segment disease treated with zero-profile anchored spacer (ROI-C) were retrospectively reviewed. This study was approved by the ethical committee of our institute. All patients were operated on by a single neurosurgeon (P.I.). Indications for surgical treatment were radiculopathy, myelopathy, and radiculomyelopathy. Before surgery, all patients obtained magnetic resonance imaging and plain radiographs including anteroposterior (AP), lateral, lateral flexion, and extension views.
Surgical Technique
Surgical procedure was performed using standard Smith–Robinson approach. A transverse incision was performed onto right-or left-sided opposite to previous incision. Discectomy and decompression were performed under an operating microscope in all cases. The posterior longitudinal ligament was only explored in cases suspected free disc fragments according to preoperative imaging. Only the cartilaginous portion of the vertebral endplates was carefully removed with preservation of the bony layer of endplates. Under fluoroscopic guidance, the cage trial was used for sizing the cage. The chosen cage was filled with demineralized bone matrix (InterGro, Zimmer Biomet, Austin, Texas, United States) or Triosite bone graft (Zimmer Biomet, Austin, Texas, United States). Triosite bone graft substitute is a bioactive calcium phosphate ceramic composed of hydroxyapatite and tricalcium phosphate. After insertion of the cage, two VerteBRIDGE plates were inserted into cranial and caudal vertebral bodies along caudal and cranial slots through the implant holder, respectively. Following removal of the implant holder, fluoroscopy was used to confirm the appropriate position of the cage in AP and lateral views ([Fig. 1]). The proper position of cage was confirmed by tantalum alloy radiologic position marker. Operative time and intraoperative blood loss were recorded.
Fig. 1 Surgical technique of the ROI-C cervical cage with VerteBRIDGE plating technology. (A) Cervical cage filled with demineralized bone matrix (DBM). (B, C) VerteBRIDGE anchor plate. During insertion of (D) cage trial, (E) cervical cage (F) upper VerteBRIDGE, (G) lower VerteBRIDGE, and (H) removal of the implant holder under fluoroscopic guidance. (I) Intraoperative photograph shows complete insertion of two VerteBRIDGE plates into both cranial and caudal slots of cage.
Postoperatively, dexamethasone was administered as an intravenous preparation at a dose of 4 mg. every 6 hours for 1 day. Soft cervical collar was used in each patient for 1 to 2 weeks. Plain radiographs including AP, lateral, lateral flexion, and extension views were performed before discharge, at 6, and 12 months.
Radiological Measurements
The cervical curvature was assessed by modified method from the study of Profeta et al.[19] A straight line (Line A) was drawn from the inner border of the odontoid tip to the inferoposterior border of C7. Another line (Line B) was drawn from the inner border of the odontoid tip to the inner border of middle part of the C4 body. The angle was measured in degree between the Line A and Line B on the pre- and postoperative images ([Fig. 2A]). The intervertebral height (IH) was measured by drawing the line from the middle portion of inferior border of the upper level of correspondent cervical spine to the middle portion of the superior border of the lower level (Line C). The length of the inferior border of the C2 was used as the reference line (Line D). We calculated the proportion of the Line C/Line D (IH ratio). If the calculated proportion at the postoperative period was higher than the preoperative, then the calculated value inferred the increase of the segmental height at the postoperative period ([Fig. 2B]). Fusion was defined according to the following criteria: (1) the absence of motion on flexion–extension radiographs and the absence of any dark halo around a cage on both AP and lateral radiograph; (2) presence of bridging trabecular bone anterior or posterior to the cage. Cage subsidence was defined as migration of cage into the superior and/or inferior vertebral body of more than 2 mm. All images were evaluated independently by the experienced neuroradiologist (S.H) using Picture Archiving and Communication System (FUJIFILM, Stamford, Connecticut, United States).
Fig. 2 (A) Scheme for the evaluation of the lordosis. A straight line is drawn from the posterior border of the C2 to the posterior–inferior border of C7 (Line A) and another line from the posterior border of C2 to the posterior border of C4. The angle between these two lines was measured. (B) Scheme for the evaluation of the intervertebral height. Line C is the line that drawn from the middle portion of inferior border of the upper level to the superior border of the lower level. Line D is the length
Clinical Evaluation
Clinical outcome was assessed before and after surgery using a 10-point visual analog scale (VAS) with endpoint anchors of “no pain” and “severe pain,” Modified Japanese Orthopedic Association (JOA) scoring system for cervical myelopathy,[20] and the Neck Disability Index (NDI).[21] Independent observer reviewed hospital charts of all patients. Clinical scores were collected preoperatively, on the day before discharge, at 1, 3, 6, and 12 months. All probable complications including dysphagia, postoperative infection, hematoma, and other device-related complications were also reviewed.
Statistical Analysis
Statistical analysis was performed using Statistical Package for the Social Sciences (SPSS) for Windows, version 16.0 (SPSS Inc., Chicago, Illinois, United States). Age of patients, operative time, estimated blood loss, IH ratio, degree of cervical curvature, VAS, modified JOA, and NDI score were presented as median with interquartile range (p25–p75). Other variables were presented as frequencies and percentages. Comparison between preoperative and postoperative clinical and radiographic data were analyzed using Wilcoxon signed-ranks test. A probability value (p-value) less than 0.05 was considered statistically significant.
Results
Between January 2017 and June 2019, 15 patients, including 3 men (20%) and 12 women (80%) with median age 57 years, range 51 to 67 years, were diagnosed with cervical adjacent segment disease and treated by ACDF with the ROI-C systems. Patient demographics, symptoms, and operative characteristics are shown in [Table 1]. Seven patients presented with radiculopathy, two with myelopathy, and six with radiculomyelopathy. Types of previous operation were Smith–Robinson procedure with iliac bone graft in one patient ([Fig. 3]), Cloward procedure in five ([Fig. 4]), cage-plate construct in eight ([Fig. 5]), and stand-alone PEEK cage in one patient. Single-level ACDF was performed in 11 patients, two-level (skip level) in 4 patients ([Fig. 6]), resulting in the treatment of 19 levels. The C3 to 4 segment was performed in four (21.1%) patients, C4 to 5 in six (31.5%), C5 to 6 in one (5.3%), and C6 to 7 in eight (42.1%) patients. All patients were followed up for at least 12 months after surgery. ACDF of supra-adjacent segment was performed in 10 levels, and infra-adjacent segment in 9 levels. Median operative time was 107 minutes (interquartile range [p25–p75]: 89–105 minutes) and median estimated blood loss was 10 (5–20) mL.
Fig. 3 Adjacent segment disease at C6 to 7 following Smith–Robinson procedure C5 to 6. (A) Sagittal T2-weighted magnetic resonance imaging of the cervical spine shows C6 to 7 disc herniation with compressing of the spinal cord. Sequential lateral radiographs of the cervical spine at (B) before surgery, (C) 2 days, (D) 6 months, and (E) 1 year after surgery reveal improvement in intervertebral disc height. The bony trabecular pattern of C6 to 7 level is noted.
Fig. 4 Adjacent segment disease at C3 to 4 following Cloward's operation C4 to 6. (A) Sagittal T2-weighted magnetic resonance imaging of the cervical spine shows kyphotic change and C3 to 4 disc herniation compressing the spinal cord. Sequential lateral radiographs of the cervical spine at (B) before surgery, (C) 2 days, (D) 6 months, and (E) 1 year after surgery reveal improvement in lordotic curve. The bony trabecular pattern of C3 to 4 level is noted.
Fig. 5 Adjacent segment disease at C3 to 4 following anterior plate construction C5 to 7. Sequential lateral radiographs of the cervical spine at (A) before surgery, (B) 2 days, (C) 6 months, and (D) 1 year after surgery show improvement in intervertebral height. The bony trabecular pattern at C3 to 4 level is noted at 1 year after surgery. The screw dislodgement (arrowhead) at C7 vertebral body was noticed and was removed during the operation.
Fig. 6 Adjacent segment disease at C4 to 5 and C6 to 7 following anterior cage-plate construction C5 to 6. Sequential lateral radiographs of the cervical spine at (A) before surgery, (B) 2 days, (C) 6 months, and (D) 1 year after surgery show improvement of intervertebral heights and good lordosis. The bony trabecular pattern of both levels is noted. There is subsidence at C6 to 7 level.
Table 1
Summary of preoperative and operative data
Preoperative characteristics
|
Patients, n
|
15
|
Age (y), median (IQR)
|
57 (53–70)
|
Gender, n (%)
|
Male
|
3 (20%)
|
Female
|
12 (80%)
|
Symptoms, n (%)
|
Radiculopathy
|
7 (46.7%)
|
Myelopathy
|
2 (13.3%)
|
Radiculopathy and myelopathy
|
6 (40.0%)
|
Operative characteristics
|
Type of previous surgery, n (%)
|
Smith–Robinson procedure (iliac bone)
|
1 (6.7%)
|
Cloward procedure
|
5 (33.3%)
|
Cage-plate construct
|
8 (53.3%)
|
Stand-alone PEEK cage
|
1 (6.7%)
|
Number of operative levels
|
One level, n (%)
|
11 (73.3%)
|
Two levels, n (%)
|
4 (26.7%)
|
Operated level
|
C3–4, n (%)
|
4 (21.1%)
|
C4–5, n (%)
|
6 (31.5%)
|
C5–6, n (%)
|
1 (5.3%)
|
C6–7, n (%)
|
8 (42.1%)
|
Adjacent segment
|
Supra-adjacent segment, n (%)
|
10 (52.6%)
|
Infra-adjacent segment, n (%)
|
9 (47.4%)
|
Operative time(minutes), median (IQR)
|
107 (89–105)
|
Estimated blood loss (mL), median (IQR)
|
10 (5–20)
|
Abbreviations: IQR, interquartile range; PEEK, polyetheretherketone.
Comparison between preoperative and postoperative clinal and radiological outcomes was shown in [Table 2]. The median pre- and postoperative VAS pain scores were 8.00 and 1.00, respectively. There was a significant relief of cervical pain after surgery (p < 0.05). The median preoperative modified JOA score was 12.00 and the mean postoperative scores increased to 16.50. The differences between the pre- and the postoperative JOA scores were statistically significant (p < 0.05). The median pre- and postoperative NDI scores were 17.00 and 3.00, respectively, which represented a significant improvement of the postoperative quality of life (p < 0.05).
Table 2
Summary of clinical and radiological data
|
Pre-op
|
Post-op
|
p-Value[a]
|
VAS score, median (IQR)
|
8.00 (7.00–9.00)
|
1.00 (1.00–2.00)
|
0.001
|
Modified JOA score, median (IQR)
|
12.00 (10.75–13.00)
|
16.50 (16.00–17.00)
|
0.027
|
NDI score, median (IQR)
|
17.00 (16.00–23.00)
|
3.00 (2.00–5.00)
|
0.001
|
IH ratio, median (IQR)
|
0.34 (0.29–0.39)
|
0.41 (0.36–0.46)
|
< 0.001
|
Curvature (degree), median (IQR)
|
4.00 (2.00–7.00)
|
5.00 (2.00–8.00)
|
0.026
|
Abbreviations: IH, intervertebral height; IQR, interquartile range; JOA, Japanese Orthopedic Association; NDI, Neck Disability Index; PEEK, polyetheretherketone; VAS, visual analog scale.
a Wilcoxon signed ranks test.
Of the 15 patients, the median pre- and postoperative cervical lordosis was 4.00 and 5.00 degree, respectively. The difference was statistically significant (p < 0.05). In 19 operated levels, the median pre- and postoperative IH was 0.34 and 0.41, respectively. The difference of pre- and postoperative IH was statistically significant (p < 0.05).
Successful fusion was achieved in 15 patients (100%). Of 19 operated levels, subsidence occurred in 1 level (5.3%) and breakage of VerteBRIDGE plate occurred in 2 (5.3%) from 38 plates but produced no symptom ([Figs. 6D], [7]). There was only one case of mild dysphagia, which resolved in less than 2 weeks. One patient had dysphagia from the loosening of screw from previous surgery and resolved from symptoms after removal of this screw ([Fig. 5A]). No esophageal injury occurred in our study.
Fig. 7 Adjacent segment disease at C4 to 5 following anterior cage-plate construction C5 to 7. Sequential lateral radiographs of the cervical spine (A) before surgery, (B) at 2 days, (C) 6 months, and (D) 1 year after surgery. The breakage of the lower VerteBRIDGE (arrowheads) is noted.
Discussion
Self-locking stand-alone cages have been increasingly used in ACDF to avoid complications associated with anterior cervical plates.[22] The ROI-C cage is a cervical interbody cage composed of radiolucent PEEK optima using the VerteBRIDGE double anchoring system for initial and long-term stability. It was designed for having zero profile without hardware protruding outside the vertebral bodies.[23]
Adjacent segment degeneration is defined as radiological changes at levels adjacent to a previous cervical fusion. However, adjacent segment disease refers to development of new symptoms correlating with adjacent segment degeneration.[24] Following the anterior cervical fusion, the predicted prevalence of symptomatic adjacent changes is 13.6% at 5 years and 25.6% at 10 years of follow-up.[25]
The exact pathophysiologic mechanism of adjacent segment degeneration remains unclear. The use of plate and screws is likely to accelerate degenerative changes in adjacent segments.[26] Based on kinematics stimulation study in comparison between ACDF with zero-profile spacer and cage-plate construct by Li et al,[27] they found that cage-plate construct may have an impact on the biomechanics of the adjacent level after a single-level ACDF, probably explaining the decreasing incidence of adjacent segment degeneration in patients using zero-profile spacer. Zhou et al[28] compared a self-locking stand-alone cage (ROI-C system) and anterior plate for ACDF in long-term follow-up and found that adjacent segment degeneration in cage group was significantly less than those in the plate group.
In case of previous ACDF with plate and screws and adjacent segment disease, the insertion of the ROI-C system is no need to remove or expose an existing adjacent plate.[17] In the present study, two VerteBRIDGE plates could be inserted into the vertebral bodies with no need to remove an existing plate and screws in all patients underwent previous ACDF with cage-plate construct.
Several studies[7]
[17]
[18]
[23]
[28]
[29]
[30] including our study demonstrated that the ROI-C interbody cage with VerteBRIDGE anchoring plates achieved a high rate of fusion. The top and bottom of the implant have teeth that fit into the bone, making the cage more stable.[28] The self-locking VerteBRIDGE plates ensure excellent primary stability of implant and promote early fusion.[7] Using the ROI-C implant, solid fusion was achieved at mean time of 4.5 to 6.9 months.[28]
[30] In the current study, all patients achieved complete interbody fusion within 1 year follow-up.
Dysphagia is more common in patients treated with ACDF due to an interface with esophagus.[14] From the comparison study between the ROI-C group and the anterior plate group in treatment one- or two-level cervical spondylosis at 3 months postoperatively by Wang et al,[7] they revealed a significantly lower risk of dysphagia in the ROI-C group than that in the anterior plate group (0 vs. 27.3%). Similarly, Liu et al[31] compared the ROI-C group and anterior plate group for multilevel cervical spondylotic myelopathy and showed a higher incidence of dysphagia of 21.9% in contrast with that of 3.6% with ROI-C device at final follow-up. Dysphagia after ACDF with cage-plate construct may occur due to longer static retraction time, excessive dissection of soft tissue, installation of prevertebral titanium plate, and inevitably stimulating the anterior structure after surgery.[18] The ROI-C system can be totally implanted into the intervertebral space so that it avoids implant contact with the soft tissue in front of the cervical spine and any mechanical irritation of the esophagus, probably explaining the low incidence of dysphagia.[23]
[31] Compared with ACDF using a plate construction, the implant of ROI-C requires less dissection and less trauma to the surrounding soft tissue. The use of the curved anchoring plates instead of screws allows for smaller exposure.[17] Therefore, the ROI-C device was found to be superior to the anterior plate group in terms of operative time and blood loss.[7]
[28]
[31] Hofstetter et al[29] revealed significantly more swelling of the prevertebral space after implantation of an anterior locking plate compared with the ROI-C. Furthermore, the patients with postoperative dysphagia in the ROI-C group recovered more quickly than those in the plate group.[28] In our study, we avoided adhesion or scar from previous surgery by approach through the opposite site to previous incision. Fortunately, there was no esophageal complications in our series.
Subsidence seems to be a problem of concern with the use of stand-alone assisted interbody fusion cage.[14] The use of the stand-alone titanium cage had a high rate of subsidence.[13] The PEEK cage is more elastic than titanium, reducing the possibility of subsidence of the cage into the vertebral body.[8] Fujibayashi et al[32] speculated that significant subsidence occurred only with the wedge-type cage in comparison with anatomical-type cage. It is probably important for the cage to have a contact surface that approaches the anatomical curvature of the involved endplates as much as possible. In addition, they found that an over-sized cage had a high rate of subsidence. The ROI-C device system, an anatomical PEEK cage and two VerteBRIDGE self-locking plates, is designed for the elimination of the basic disadvantage of stand-alone cage.[17] Multiple studies,[17]
[18]
[22]
[23]
[28]
[30]
[33] including our study, showed low subsidence rate. Most patients with cage subsidence did not experience any symptoms and they achieved fusion in the final follow-up. Importantly, preservation of bony endplates and avoiding over-sized cage are imperative factors to avoid subsidence in our surgical technique. However, postoperative cage subsidence should be aware in patients with multilevel ACDF using self-locking stand-alone cage.[34]
[35]
According to the study of 90 patients who underwent ACDF with the ROI-C device by Lonjon et al,[22] they reported anchoring plate fracture in 5 (2%) from 249 plates with no clinical consequence. The mechanism of anchoring plate fracture remains unknown. In our study, there was the breakage of anchoring plate in 5.3%. Similarly, this device-related complication produced no symptoms and did not require surgery.
The ROI-C implant was associated with a significant benefit in treating noncontinuous level of cervical degenerative disc disease.[31] Lu et al[18] compared the clinical and radiological results of the ROI-C device and cage-plate construct for the treatment of noncontiguous bilevel or skip-level of cervical degenerative disc disease and found that the ROI-C group had a shorter operative time and significant lower incidence of dysphagia. Similar to skip-level cervical degenerative disc disease, supra- and infra-adjacent segments in the same patient were treated with ACDF using the ROI-C implants in four patients in the present study.
According to a meta-analysis study of comparison between zero-profile devices and cage-plate construct in ACDF with a minimum 2 years of follow-up by Sun et al,[36] they found that the zero-profile implant and cage-plate construct achieved comparable mid-term and long-term clinical and radiological outcomes, including the fusion rate, in ACDF. Additionally, the zero-profile group has obvious advantages in reducing intraoperative blood loss, improving the cervical lordosis angle, and reducing the incidence of postoperative dysphagia and adjacent segment degeneration.
Limitations
Our study has some limitations including the observational study without control group, small sample size, and relatively short follow-up. More patients and longer follow-up are acquired to further investigate the efficacy of the ROI-C system for the treatment of cervical adjacent disease. Due to concerning about radiation exposure, reliable computed tomography scan was not performed for the assessment of solid fusion in our study. However, the PEEK cage is radiolucent and easy to observe trabecular bone formation and evaluation of fusion status on radiographs.
Conclusion
The ROI-C implant system, a zero-profile interbody fusion cage with self-locking designed for stand-alone fusion without external plates and screws, is safe and effective for the treatment of cervical adjacent disease in patients with cervical radiculopathy and/or myelopathy. Significant improvement in clinical scores and low incidences of postoperative dysphagia and subsidence were obtained by this prosthesis. In addition, the ROI-C cage provides high rate of fusion, and restoration of the cervical lordosis and IH.