Anterior Controllable Antedisplacement and Fusion (ACAF) Technique for the Treatment of Multilevel Cervical Spondylotic Myelopathy With Spinal Stenosis (MCSMSS)

Study Design: Retrospective study. Objective: To investigate the clinical effect of anterior controllable antedisplacement and fusion (ACAF) technique for the treatment of multilevel cervical spondylotic myelopathy with spinal stenosis (MCSMSS), compared with hybrid decompression fixation (HDF). Summary of Background Data: A retrospective analysis of 94 cases with MCSMSS was carried out. Fifty-four patients were treated with ACAF, whereas 40 patients were treated with HDF. Methods: The operation time, intraoperative blood loss, postoperative complications, Japanese Orthopedic Association score, Neck Disability Index score, parameters at axial computed tomography, cervical curvature and the Kang grade were compared between 2 groups. Results: The patients were followed up for 12–17 (15.6±1.6) months. Compared with HDF, ACAF group achieved better decompression according to computed tomography measurement and Kang grade (P<0.05), and recovered to a greater cervical Cobb angle (P<0.05). However, Japanese Orthopedic Association score and Neck Disability Index showed no significant difference 1 year after surgery (P>0.05). In addition, ACAF presented longer operation time (P<0.05) and similar intraoperative blood loss (P>0.05), compared with HDF. In terms of complications, ACAF produced less incidences of cerebrospinal fluid leakage, implant complication, epidural hematoma, and C5 palsy compared with HDF. Conclusions: ACAF is an effective method for the treatment of MCSMSS. In comparison to HDF, ACAF has the advantages of more sufficient decompression, more satisfactory cervical curvature, and lower incidence rates of complications.


BACKGROUND
Multilevel cervical spondylotic myelopathy with spinal stenosis (MCSMSS) has been a complex cervical disorder, characterized by gradual spinal cord compression involving 3 or more cervical segments. The typical feature of MCSMSS is the compression at the levels of both intervertebral disk and vertebra. 1 Surgery has been actively recommended for patients with intolerable symptoms and suspected neurological damage. 2 However, considering the multilevel compression, the optimal choice of surgery remains controversial. 3 For single-level compression, anterior cervical discectomy and fusion (ACDF) has been a "gold standard." 4,5 However, it is difficult for ACDF to achieve sufficient decompression when faced with the compression behind vertebra. As to the anterior cervical corpectomy decompression and fusion (ACCF), although its decompressive effect is obvious, the high occurrence rate of surgery-related complications significantly compromise its clinical benefits. 6 Anterior controllable antedisplacement and fusion (ACAF) has been initially proposed to treat patients with cervical ossification of the posterior longitudinal ligament. [7][8][9][10] In this technique, the affected vertebrae and compressive components are not resected but isolated and moved anteriorly to reconstruct the anterior column of the cervical spine. As the isolated vertebrae can be used as autografts with intervertebral cages, cervical spine can acquire satisfactory stability under the premise of sufficient decompression after surgery. Given the satisfactory outcomes of decompression of the cord and the enlargement of the spinal canal by ACAF, we introduced this technique to solve the clinical problems of MCSMSS. However, no study has been published to investigate the surgical outcomes of ACAF on patients with MCSMSS. Therefore, this study aimed to investigate the clinical effect of ACAF technique for the treatment of MCSMSS, and compare its outcomes with hybrid decompression fixation (HDF, the combination of ACDF and ACCF). 11

Patients Population
This study was approved by the institution review board of our hospital, and all patients signed the informed consent.
From January 2017 to April 2019, 99 patients diagnosed with MCSMSS and admitted to the Second Department of Spine Surgery at the Changzheng Hospital in Shanghai were included. X-ray, computed tomography (CT), and magnetic resonance imaging (MRI) of cervical spine were taken before operation. Inclusion criteria: (1) Diagnosed as cervical spinal stenosis (the sagittal diameter of the spinal canal is <12 mm) by radiologic imaging, and the conservative treatment is ineffective for more than 6 months; (2) cervical spondylotic myelopathy involving segments ≥ 3. Exclusive criteria: (1) deformity, ankylosing spondylitis, rheumatoid arthritis, and other diseases involving the cervical spine; (2) cervical spine trauma or surgical history; (3) severe osteoporosis.
Two patients were excluded for history of cervical spine trauma, 2 were excluded due to previous cervical spine surgery, and 1 was excluded owing to severe osteoporosis. Ultimately, totally 94 patients were included in this study with 54 cases by ACAF and 40 cases by HDF.

Surgical Procedure
For each patient, surgical details, expected benefits, potential risks, and complications of ACAF and HDF were clearly informed before operation. The surgical procedure was determined based on experience of surgeons, intraoperative condition, and wishes of patients. To avoid fusion-related complications caused by implantation of long-segment titanium mesh, patients expected to resect 3 vertebrae by ACCF were treated with ACAF; to reduce the risk of cerebrospinal fluid (CSF) leakage caused by the direct resection of vertebrae, patients were treated with ACAF when the posterior longitudinal ligament (PLL) and the dura mater were tightly adhered and difficult to separate by nerve hook during intraoperative exploration.
ACAF group: (1) After general anesthesia, supine position was taken, cervical oblique incision was made at the anterior right side to expose deep structures. (2) The responsible intervertebral disk, anterior and posterior edge osteophyte were removed until the PLL was exposed, and the superior and inferior endplates were scraped to make the surfaces smooth. Then, at the cephalic and caudal ends of the planned hoisting segments, the PLL was bitten in intervertebral space, whereas the rest of PLL was not deal with. (3) According to the requirement for decompression, the proper thickness of anterior part of the vertebra was removed, and the space was reserved for the responsible segment to be hoisted forward. (4) According to the decompression width needed, the grooving position was selected on the anterior surface of vertebra, which was usually the inner edge of Luschka joint. Grinding drill was used to dig deep along the grooving lines on both sides, and the bone was chisel away to reach the cortex of the posterior wall of the vertebra. The posterior wall was bitten on one side. The other side was temporarily retained to maintain the stability of the vertebra. As to bleeding of the exposed cancellous bone of anterior surface of vertebrae and grooves, the use of piezosurgery or high-speed grinding drill can significantly reduce the bleeding. Meanwhile, the application of bone wax can also obtain satisfying hemostasis effect. (5) Intervertebral fusion cages, precurved titanium plate, and screws fixed on caudal and cranial vertebrae were installed to fix the plate, and then the posterior wall of the vertebra on the other side was resected to make the vertebra free. (6) The screws fixed on free vertebrae were tightened to pull them forward. At this time, compression mater would move forward together until the vertebrae and the titanium plate were closely joined. Finally, the incision was flushed, hemostasis, and drainage were performed, and suture was finished layer by layer. The diagram of the procedure of ACAF is presented in Figure 1.
HDF group: ACCF was performed at the segments with severe compression, and ACDF was performed at other segments. The specific steps are as follows: (1) ACCF: Exposure process was the same as ACAF group.
Degenerative disks and osteophytes in the intervertebral space was removed, until reaching hook joints on both sides. Subtotal resection of adjacent vertebrae was performed, and osteophytes were removed. Titanium cages filled with broken bones were placed after decompression. Finally, titanium plate and screw were installed. (2) ACDF: Other processes were the same as ACCF. After discectomy, the osteophytes at posterior margin of the vertebra were removed, and the intervertebral space was expanded by distractor to normal height. Intervertebral fusion cage was selected and inserted into the intervertebral space. Titanium plate and screw fixation was finally performed.
Postoperative management was as ACAF's. All operations were performed by 3 surgeons from the same spine surgery team. Negative pressure drainage tubes were placed and pulled out 24-48 hours after operation. All patients were fixed with external cervical bracket for 3 months.

Functional Evaluation
Before and 1 year after operation, neurological function was evaluated by Japanese Orthopedic Association (JOA) score and Neck Disability Index (NDI) score.

Axial CT Measurement
By utilizing the measuring tools embedded in software of picture archiving system, 3 parameters were measured on axial CT images before and after operation to evaluate the surgical effects. The definitions of the parameters were as follows: In patients accepted ACAF, the data were measured at the median level of each vertebra being hoisted, whereas the corresponding data were measured at the vertebrae being resected in patients accepted HDF. In addition, the segments not hoisted or not resected were not involved in the calculation.
The measuring diagram is depicted in Figure 2.
The Cobb angle method was used to evaluate cervical curvature on the lateral images of x-ray, and the angle is formed by vertical lines of the upper edge of C 2 and the lower edge of C 7 .

Kang Grade
Kang MRI grading system was used to assess the degree of cervical spinal cord compression, 12,13 and the specific criteria is: grade-0, no spinal canal stenosis; grade-1, subarachnoid compression exceeded 50%; grade-2, spinal cord compression deformed; grade-3, spinal cord T2-weighted signal changes. All the parameters were measured by 2 senior spine surgeons independently similar to the method of axial CT measurement, and the average value of each parameter at each level was taken for independent calculation.
Statistical Methods SPSS 22.0 was used for statistical analysis. Measurement data were expressed as mean ± SD x s (¯± ). Paired t test was used for intragroup comparison. Two independent samples t test was for intergroup comparison, rank-sum test was for comparison of ranked data, and χ 2 test was for the categorical data comparison. Test level was α = 0.05.

Demographics
All the enrolled 94 cases were treated successfully, with significant relief of clinical symptoms and improvement of spinal cord function. No incision infection and nerve injury occurred perioperatively. Ninety-four cases were followed up for 12-17 (15.6 ± 1.6) months.
Among 94 cases, there were 45 cases in the ACAF group, including 35 males and 19 females, aged 36-72 (53.9 ± 9.1) years; 40 cases in the HDF group, including 22 males and 18 females, aged 38-75 (56.1 ± 9.6) years. There was no significant difference regarding sex, age, and operative levels between the 2 groups. The demographic data were shown in Table 1.

Axial CT Parameters
No statistical differences was observed in terms of the preoperative diameter of spinal canal, decompression width, and axial area of spinal canal in 2 groups. However, after surgery, compared with HDF group, patients in ACAF group acquired better decompression at C3, C4, C5, and C6 levels from the aspect of diameter of spinal canal, decompression width, and spinal canal area. The data were shown in Table 2.

Clinical Outcomes
There was no significant difference in preoperative JOA score (P = 0.105), NDI (P = 0.833), Cobb angle (P = 0.578), and Kang grade (P = 0.261) between the 2 groups. One year after operation, both groups acquired obvious improvement in JOA score (both P < 0.001) and NDI (both P < 0.001), whereas no significant difference of these 2 scores existed between 2 groups (P = 0.120 for JOA; P = 0.150 for NDI). As to morphologic parameters, ACAF achieved better recovery of Cobb angle (P < 0.001) and Kang grade (P = 0.015) compared with HDF. Although, patients in ACAF group experienced longer operation time (P < 0.001), there was no significant  difference in intraoperative blood loss between the 2 groups (P = 0.214) with regard to complications, 5 patients (12.5%) presented with CSF leakage, 2 (5.0%) with epidural hematoma, 2 (5.0%) with C5 palsy, and 3 (7.5%) with dysphagia in HDF group. In addition, at the final follow-up, 4 patients (10.0%) presented with implantrelated complications in HDF group, including 2 with mesh subsidence and 2 with delayed union. In contrast, ACAF group demonstrated only 1 case (1.9%) of CSF leakage and 4 cases (7.4%) of dysphagia, without other complications. The data are shown in Tables 1 and 3.

Representative Cases Case 1
Complained of numbness of both arm with gait disturbance for 3 years. The numbness had aggravated 2 months before admission. ACAF was performed at C3/4, C4/5, C5/6, and C6/7. After the operation, the patient showed dramatical recovery of neurological function. Sufficient decompression of the spinal cord was achieved according to imaging. The JOA score increased from 7 to 11. There were no complications during follow-up. The imaging examination is shown in Figure 3.

Case 2
Presented with spastic weakness of both hands and gait instability for 2 years. The symptoms progressively have worsened in the past 3 months before presentation. ACDF was performed at C4/5 level, and ACCF at C5-C7 level. After the operation, the patient's myelopathy showed marked recovery. Preoperative and postoperative JOA score was 8 and 12, respectively. No complications were observed during follow-up. The imaging examination is shown in Figure 4.

Technical Features of HDF
The purpose of HDF is to use ACDF in slightly affected segments and ACCF in the severe ones, 14 so as to exploit superiorities of these 2 techniques, including preserving vertebrae and maintaining the stability of cervical spine much better after decompression. Compared with multilevel ACCF, HDF can implant shorter titanium mesh to decrease the risks of mesh subsidence and fusion failure to some extent. 4,15 The satisfactory outcomes of decompression and fixation has also been confirmed. 14,16 However, HDF still adopts the strategy of corpectomy in response to the compressive lesion behind the vertebrae bodies. During the procedure, the operator has to repeatedly probe into the vertebrae-dura interface with surgical instruments to bite off osteophytes, which increases the risk of causing disturbance to the dura and spinal cord, 17 especially in cases with multilevel compression. 18

Technical Features of ACAF
In ACAF, vertebrae of affected segments with posterior osteophytes are considered as a whole, namely vertebrae-compression complex (VCC), which should be disassociated from the other part of vertebrae through grooving at both sides, and it can help to reconstruct the anterior wall of the spinal canal. By moving VCC forward, the expansion of spinal canal and retention of bone of vertebrae can be achieved simultaneously. During the operation, the forward movement of VCC can be monitored by intraoperative x-ray, which could ensure that the whole process of hoisting and spinal cord antedisplacement can be controlled, and the "anatomical reduction" of spinal cord can be achieved. The spinal cord is decompressed by ACAF without any anterior or posterior shift, which we called "in situ decompression of the spinal  cord." In addition, the whole procedure of ACAF is completed outside the spinal canal, without any instruments into the canal, which could significantly decrease the risk of neural element injury.
The ideal hoisting by ACAF is to restore the normal sagittal diameter of spinal canal and relieve the spinal stenosis, whereas incomplete hoisting would lead to incomplete decompression 19 ; while excessive hoisting (excessive removal of bone of the anterior part of vertebra) can also increase the risk of fusion-related complications and may result in dural tear and CSF leakage when the dura adheres to the posterior edge of vertebra. According to the degree of spinal stenosis, the thickness of the removed vertebra is generally 3-4 mm. To achieve full fusion of bone graft, ACAF should be performed to ensure that the sagittal diameter of the remaining vertebra after resection exceeds 12 mm. This is partly because 12 mm is the length of anterior-posterior diameter of the fusion cage commonly used in cervical surgery, and by doing so, the vertebrae with sagittal diameter exceeding 12 mm can achieve fully contact with fusion cages to reach the maximum fusion area. In addition, 10-12 mm is the shortest size of the general mass-produced vertebral screws, and the adequate sagittal diameter of vertebra can provide the reliable holding force for screw placement. Therefore, considering that the average sagittal diameter of human cervical vertebra is about 18 mm, the thickness of osteotomy should not exceed 6 mm generally. 7 However, the actual expansion of spinal canal will be affected by the curvature of the precurved titanium plate, which indicates that the greater the curvature is, the more the vertebrae move forward when hoisting. 19 Therefore, the increase of sagittal diameter of spinal canal will be slightly larger than the resected thickness of vertebra. In conclusion, the thickness of vertebra needed to be removed should be determined by taking into account the degree of spinal stenosis and the curvature of titanium plate. If necessary, the intraoperative O-arm x-ray machine can also be used to confirm the hoisting condition.
In contrast, according to clinical experience, the grooving position is roughly at the inner edge of Luschka joint, which not only ensures the enough decompression width (the effective range almost reaches the medial wall of the pedicle), but also avoids the injury of vertebral artery. Compared with the HDF operated in the anterior median region of cervical spine, ACAF operation mainly focuses on the region around Luschka joint, which is far away from spinal cord and its artery traveling area, constructing a "safe space" for spinal cord and reduces the risk of spinal cord injury. 20 ACAF retains and hoists vertebrae to complete direct decompression and to maintain the mechanical stability of cervical spine, which makes it unnecessary to perform large scale osteotomy when dealing with multilevel spinal stenosis. 21 However, diseases with destruction of vertebrae may be surgical contraindications of ACAF. What is more, the vertebrae with damaged structure cannot provide enough holding force for screws, which may lead to failure of hoisting; on the other hand, damaged bone may increase the risk of fusion-related complications. Therefore, patients with cervical infection, cervical cancer, cervical fracture, and severe osteoporosis are not suitable for ACAF.

Advantages of ACAF in the Treatment of MCSMSS
(1) Favorable recovery of CSF space and spinal cord morphology CSF is the direct environment for the survival of spinal cord. Normal flow of CSF plays an important role in buffering concussion and maintaining normal metabolism of spinal cord, and disturbance of the pulsatile CSF flow has demonstrated a high correlation with the severity of myelopathy. 22 Once compression occurs, CSF band would be affected before spinal cord. Chang and Shibuya thought that in CSM patients, CSF flow disorder can not only reflect the degree of dural sac compression, but also change the fluid pressure on the surface of the spinal cord, causing damage to the spinal cord. 23,24 Yun found that in Kang grade 0 and grade 1 patients, the flow of CSF was free and smooth. In group of grade 2, the flow of CSF in the dorsal side of spinal cord tends to disappear. As for grade 3, CSF flow motion is absent in both dorsal and ventral sides. 22 Therefore, the reconstruction of CSF band should be a vital part of surgery to acquire a better prognosis. There was 1 case with CSF leakage in the ACAF group, which may be attributed to the retrovertebral adhesion and excessive hoisting. To further avoid dural tear, the hoisting distance should be reasonably planned preoperatively according to the degree of spinal stenosis to avoid excessive hoisting, so as to make the movement of dura mater as smooth and soft as possible. In this present study, it is observed that the Kang grade of ACAF group decreased significantly, which means the recovery of CSF band is better. The flow of CSF is cyclical and continuous. So, if 1 segment is compressed, the circulation of CSF will be interrupted. In ACAF, by pulling the VCC, 3-4 vertebrae can move forward as a whole, and the compression at disk and vertebral levels would be relieved simultaneously. Furthermore, the vertebrae after antedisplacement can be involved in the reconstruction of anterior column. So, the vertebrae can provide protection as anterior wall of spinal canal to avoid excessive antedisplacement of the dura sac. 25,26 (2) Fewer complications (A) C5 palsy Owing to the limited visual field of ACCF during operating, the true decompression width of spinal canal is smaller than the slotting width of the anterior vertebral surface, resulting in incomplete decompression at the outlet of the nerve root and C5 palsy. 27 However, extremely wide anterior decompression may still contribute to the occurrence of C5 palsy. 28 The mechanisms are considered to be excessive dural expansion and anterior shift of spinal cord. 29 In ACAF, the width of decompression is fully guaranteed by vertically grooving by grinding drill. As a result, when VCC moves forward, the medial part of the outlet of nerve root will be greatly expanded to achieve significant decompression. Meanwhile, the over anterior shift of spinal cord is prevented by VCC as well. 30 On the basis of the above reasons, ACAF hold an edge of lower incidence of C5 palsy. (B) Postoperative hematoma and CSF leakage When venous plexus and dura mater adhere to surrounding ossified mater, rash operation may lead to hematoma and CSF leakage. 31 In ACAF, the operation does not directly involve the dura mater and intraspinal venous plexus. When there exists adhesion between dura and ossified mater, after VCCs moving forward, these structures still maintain the previous connecting relationship, which can reduce the risk of dura tear and hematoma. 8 Furthermore, the antedisplacement of VCC will pull the dura forward, keep its tension to a certain extent, which is just like a tent, thus blocking the compression of spinal cord by the potential hematoma.

Limitation
Owing to its retrospective nature and short-time follow-up, the results of this study might be limited. A more convincing clinical conclusion would be drawn if a larger asymptomatic cohort is included.

CONCLUSIONS
The results of this study showed that ACAF technique can expand spinal canal volume by anterior displacement of the compressive substance in the treatment of MCSMSS. It can safely and effectively relieve the spinal cord compression and spinal canal stenosis, and significantly improve cervical curvature with favorable recovery of CSF band. In addition, ACAF can reduce the rates of complications such as CSF leakage and epidural hematoma that commonly occur in traditional anterior operation. In conclusion, ACAF is a satisfactory optional surgical method for the treatment of MCSMSS.