A Long-Term Control Study of the Treatment of Cervical Radiculopathy by Percutaneous Endoscopic Posterior Cervical Discectomy and Anterior Cervical Decompression, Bone Graft Fusion and Internal Fixation


 Background: Cervical spondylotic radiculopathy (CSR) is very common all over the world. However, there are only a few reports about the efficacy of percutaneous endoscopic posterior cervical discectomy (PEPCD) in the treatment of CSR. Anterior cervical decompression and fusion (ACDF) and PEPCD which is a better way to treat CSR need further study.Methods：From January 2015 to December 2016. A retrospective study of 70 patients undergoing surgery for CSR (33 using PEPCD and 37 using ACDF). The intra-operative parameters, neck disability index (NDI), neck and arm visual analog scale (VAS) score were used to assessed clinical outcome. Radiological outcomes were assessed by measuring cervical 2-7 (C2-7) lordosis, disc height index (DHI), and degree of degenerative changes at the corresponding level.Results: The mean follow-up period was 48.5 month (36 –66 months). Two groups can significantly improve the clinical symptoms. There was no significant difference between the two groups in clinical results (VAS, NDI). Compared with preoperative, the lordosis of C2-7 increased significantly at the last follow-up. At the last follow-up, C2-7 lordosis in the ACDF group was significantly higher than that in the PEPCD group. In PEPCD, compared with preoperative, there was no significant difference in the DHI, but the degree of disc degeneration was significantly increased at the last follow-up.Conclusion: For patients with CSR, PEPCD had similar clinical early and intermediate outcomes when compared with ACDF, with the advantages of minimally invasive. PEPCD is a sufficient and safe supplement and alternative to conventional surgery.


Introduction
Cervical spondylotic radiculopathy (CSR) refers to a series of secondary pathological changes that occur on the basis of degenerative changes in the intervertebral disc. These pathologic factors interact with intervertebral disc degeneration, oppressing the corresponding nerve roots, thereby eliciting clinical manifestations of pain and dysfunction [1,2]. The disease is slow onset, and more common in the population aged 40-60, with long-term unhealthy posture. The natural history of most patients is selflimited [3]. The symptoms were relieved after conservative treatment in most of the patients. When conservative treatment does not work, surgery is required [4].
Traditional surgical treatment of CSR includes ACDF [5][6][7]and posterior cervical intervertebral incision, the "keyhole" surgery [8,9]. ACDF was initiated by Robinson et al. in the 1950s [5], which is widely used in the treatment of cervical spondylosis, and gradually become a standard for treatment of cervical disc disease [6,7,10]. While ACDF showed a certain degree of prosthesis sedimentation, dislocation and false joint formation and other complications [11][12][13]. Compared with ACDF, posterior surgery has the following advantages: First, because the back of the cervical spine without signi cant vascular and nerve structure, so the possibility of damage to vascular nerves is low [14]; Second, posterior surgery can be effective decompression on the proliferation of articular processes and prominent intervertebral disc induced neural root canal stenosis , to avoid the ACDF fusion failure, prosthesis sedimentation and pseudarthrosis and other risks [15]. However, the open posterior cervical spine surgery has many disadvantages, such as large trauma, bleeding, wide range of paraspinal muscles need to be stripped, neck pain and dysfunction [16]. In addition, there is a loss of the lordosis of the cervical spine and the potential risk of sagittal deformity after long segment decompression [17].
With the development of minimally invasive spine surgery, the development of endoscopy has made remarkable achievements [18][19][20][21]. Micro Endoscopic Discectomy (MED) [19,22] and percutaneous endoscopic assisted intervertebral foramen expansion and nucleus removal [21,23] are widely used at present. MED can use the microscope to amplify the eld, so that the operation is more accurate.
Compared with the traditional open surgery, the muscle stripping range is greatly reduced, short hospital stay, less intraoperative bleeding and less postoperative complications [24]. Due to surgery need to use the channel for a long time to distract the muscles, postoperative muscle pain and muscle spasms may occur, and the risk of dural rupture is relatively high [25]. At present, almost all of the lumbar disc herniation can be done through the endoscope [21]. PEPCD is an extension and progression of posterior cervical MED technique. Using transcutaneous endoscopic technique, the eld is directly projected through the electronic imaging system to the display screen, making the surgical trauma smaller, more accurate, less bleeding, shorter hospital stay, and fewer complications [21,23,26,27]. Limited paper reported effect of PEPCD on the treatment CSR. According to intra-operative parameters, clinical outcomes, complication and recurrence, ACDF and PEPCD which is a better way to treat CSR need further study. Moreover, there are few reports about the di culties of localization and intraoperative bleeding during PEPCD operation.
The purpose of this retrospective study was to (1) provide comparative data regarding clinical outcomes using well-established patient-reported outcome measures in patients undergoing surgery for CSR using either PEPCD or ACDF. (2) discusses the problems of rapid location and intraoperative bleeding during PEPCD operation.

Materials And Methods
A retrospective analysis was performed on 70 patients diagnosed with CSR and underwent cervical spine surgery. From January 2015 to December 2016, 412 cases of cervical spine surgery performed by an experienced doctor (Xiaojian Ye) were collected and stored in the Changzheng Hospital Orthopedic Medical Database. According to the following inclusion and exclusion criteria, 70 patients diagnosed with single-segment CSR were treated with PEPCD or ACDF. medical history and physical examination, NDI and VAS (neck and upper limb) score, cervical X-ray, MRI and CT were used to determine the clinical outcome. All patients were scheduled for follow-up at 3, 6, and 12 month and after each year postoperatively through an outpatient or telephone consultation.

Inclusion/exclusion Criteria
The inclusion criteria for this study were: (1) older than 20 years old. (2) a clear diagnosis of singlesegment CSR, and at least 8 weeks of strict conservative treatment is invalid. (3) The surgical segment is a single segment in C2-T1 and is operated by ACDF or PEPCD.
The exclusion criteria for this study: (1) Existence of cervical surgery contraindications. Under general anesthesia, the patient is supine in the operating bed, and the neck is in the extended position to increase the intervertebral height of the target segment. Use C arm machine for positioning. In the right side of the neck skin for a length of about 3 cm transverse incision, and layer by layer exposed to the target intervertebral space. Removal of the intervertebral discs and the compression of the spinal cord and nerve roots, including prominent nucleus pulposus, osteophyte and posterior longitudinal ligament.
Scrape the cartilage lamina, but to protect the bone plate to prevent the sedimentation of the implant. Preoperative evaluation and intraoperative measurement to select the appropriate size of the Zero-P (Johnsonn). Zero-P was lled with autogenous bone or arti cial bone, and then use Zero-P's own screws to x. Rinse the surgical area, stop bleeding, rinse again, and then sutured the incision layer by layer. Complete surgery. After the operation, the patients were treated with dehydrating agents and neurotrophic drugs. The second day after surgery under the protection of the neck support to get out of bed activities, and neck support needs to wear 2-4 w.

PCPED:
PCPED was performed using the TESSYS TM system (Joimax, Germany). The patients were placed in the prone position on a customized plaster bed, and the neck was slightly buckled to increase the intervertebral space size of the target segment and reduce the overlap of small joints under general anesthesia. Double upper limbs close to the trunk, ngers toward the tail, the patient's head and double upper lip with a xed with tape ( gure.1 A). Use C arm machine for auxiliary positioning. Two 20 G long positioning needles were punctured at about 1.5-2 cm on the outside of the spinous process of the target gap ( gure.1 B), and locate the needle at the K point by uoroscopy (K point: the focal point of the upper lamina, the lower lamina and the connection of the medial margin of the adjacent facet gure.1D). A longitudinal skin incision with a length of about 0.8mm is positioned at the positioning needle, and the guide needle is inserted into K point, and the expansion sleeve is gradually arranged until the work channel. When the channel is in a good position ( gure.1C), take out the guide wire and the expansion sleeve, and put into the intervertebral disc endoscope. 0.9% of the saline hanging in the top of the patient about 1.5 m, adjust the ow rate about 50-100 ml / min, connected to the endoscopic device for continuous lavage. Radiofrequency ablation and hemostasis under endoscope. The lower edge of the upper lamina and the upper edge were removed by grinding and laminectomy (3mm of the medial margin of the facet joint). The range of bite is within 1/3 of the articular process ( Figure.1D Figure.2 A-D). Using the blue tweezers to remove some of the ligamentum avum, the epidural venous plexus was coagulated by bipolar radiofrequency electrode to maintain a clear visual eld and accurately identify nerve roots. According to the imaging data of the patients, the decision was made to expand the intervertebral foramen or remove the nucleus pulposus. When the intervertebral foramen stenosis, you can use the bite forceps and power grinding (Stryker, USA) to expand the intervertebral foramen. Decompression was performed until proximal and distal pedicles were con rmed longitudinally, a probe was then easily inserted through the foramen to con rm adequate neural decompression. When the disc is prominent, using nerve root probe to explore the nerve root axillary and nerve root shoulders, to determine the location of the nucleus pulposus; use nucleus pulposus gradually remove the prominent nucleus until the nerve root completely released ( Figure.2 C-E). The ber ring and the soft tissue around it were ablated by radio frequency electrode. When the dura mater beat was good, the decompression was complete. Use radio frequency to stop bleeding, rinse the area, take out the intervertebral foramen mirror and channel, suture the incision and complete the operation ( Figure.2 F). Because of the small incision of the PEPCD, it is not necessary to have a drainage tube after operation, and the silica gel can be placed for drainage and remove within 24 hours after operation. After the operation, the patients were treated with dehydrating agents and neurotrophic drugs. The second day after surgery under the protection of the neck support to get out of bed activities, and neck support needs to wear 5-7 d.

Intraoperative outcome Measure
According to the patient data and operation records, the operative segment, operation time, amount of bleeding, complications, LOHS (length of hospital stay), and time to return to normal work of the two groups were recorded and compared.

Postoperative outcome evaluation
The NDI, neck and arm VAS score, X-ray, MRI and CT were evaluated on the day before surgery and 3 m, 6 m, 12 m, 24 m, 36 m, 48 m, 60 m and the last follow-up after surgery. The Cobb angle of cervical was measured by C2-7 lordosis ( Figure.4 B) [28], and DHI was measured by Emery [29]. According to the modi ed P rrmann classi cation, the degree of cervical disc degeneration was evaluated before and after operation [30]. According to the preoperative and the last follow-up of the cervical dynamic X-ray lm to determine whether the existence of cervical instability: intervertebral angle> 10°or displacement> 3 mm to determine intervertebral instability.

Statistical Analysis
Mean outcome scores for NDI, neck and arm VAS were compared for each follow-up period. Statistical analysis was conducted with the Student t-test, with P<0.05 considered signi cant.

Patient Characteristics
As showed in the  There was no serious complication in both groups. Severe complications, including cerebrospinal uid leakage, paralysis, hoarseness, postoperative bleeding compression, infection and so on. In the ACDF group, there were ve patients with temporary swallowing pain and foreign body sensation and the symptoms were relieved within 2 w. one patient due to scar body induced neck appearance damage. In the PEPCD group, one patient had excessive bleeding during surgery, resulting in blurred vision, surgical methods were changed to cervical posterior MED. Symptom of another patient was not improved on the second day after surgery, and ACDF was performed on the third day. ACDF surgery con rmed that the target segment protrusion of patients with ossi cation, and nerve root canal stenosis. The operative failure rate of PEPCD group was 6.06%, which was signi cantly higher than that of ACDF group. There were no further complications, such as infection, spondylodiscitis or thrombosis.

Radiologic Findings
In this study, there were no cases of cervical vertebral instability, cervical disc herniation, kyphosis or lordosis after operation in both group. The results of cervical curvature showed that the preoperative C2-7 lordosis of the ACDF and PEPCD group was 9.99 ± 3.66 and 10.91 ± 5.68 respectively. At the end of the follow-up, the lateral radiographs of the cervical spine showed that the C2-7 lordosis of the two groups were signi cantly increased (ACDF 18.56 ± 5.28 PEPCD 13.89 ± 3.35), and the ACDF group was signi cantly higher than that of PEPCD (P<0.05). At the last follow-up, MRI found that there were no signi cant changes in the adjacent intervertebral discs in the two groups. MRI showed that the intervertebral disc degeneration grade of PEPCD group was signi cantly increased from 3.  Figure.3 The changes of C2-7 lordosis in patients with ACDF and PEPCD before and at the last follow-up. A, B: cervical lateral radiographs of PEPCD at the time of preoperative and nal follow-up, respectively. C, D: cervical lateral radiographs of patients with ACDF before and at the last follow-up, respectively. E: C2-7 lordosis data for two groups of patients before and at the last follow-up. Data was mean ± SD, *P<0.05. Figure.4 48 years old male patient presented with a three dimensional CT after surgery. A: the arrow refers to the range of three-dimensional CT decompression on the 1 day after surgery; B: the arrow refers to the range of three-dimensional CT decompression at the end of the 2 year after surgery.

Discussion
In this study, we retrospectively analyzed the clinical data and surgical results of ACDF and PEPCD in the treatment of single-segment CSR. Our study found that PEPCD treatment of single-segment CSR was similar to ACDF in terms of early and mid-term effects. Both of which can effectively remove nerve root compression to relieve symptoms, which is similar to previous ndings [21,23,26,27]. Compared with ACDF, PEPCD can signi cantly reduce the surgical trauma (reduced blood loss, shortened hospital stay), to promote postoperative rehabilitation (Back to work / Full activity), reduce economic and social burden (lower hospitalization costs) [20,31]. We found that the ACDF and PEPCD can signi cant improve C2-7 lordosis, and the C2-7 lordosis improved more in the ACDF group when compared with that of PEPCD. In the PEPCD group, 2 patients due to intraoperative or postoperative symptoms did not signi cantly alleviate, thus changing the surgical procedure. There were no cases of revision surgery due to recurrence of symptoms. The failure rate PEPCD were signi cantly higher than those of ACDF group.
With the changes in modern life and working style, the incidence of neck and shoulder pain and other diseases caused by degenerative intervertebral discs is increasing [32]. CSR is a common cause of neck pain and upper limb pain [33]. Through the control of neck activities, physical therapy, analgesic, dehydration, neurotrophic drug therapy, symptoms can be alleviated [3,34]. For long-term, severe CSR patients, surgical treatment is required. With the development of minimally invasive endoscopic surgery in recent years, more and more people have reported the application of PEPCD in the treatment of cervical spondylosis [21,26,27,35,36]. PEPCD surgery not only have advantages of cervical posterior keyhole surgery and MED surgery, but also has a smaller trauma, better visual eld, more accurate decompression positioning, surgical operation more accurate and so on. With the further optimization of endoscopic devices, surgical operation is more simple. The cervical lamina is relatively at and there is a natural lamina intervertebral space, which helps to locate and place the endoscopic channel during operation. It is very practical.
In this study, patients with PEPCD were treated with a customized plaster bed based on the size of the patient. On one hand, the plaster bed was convenient for intraoperative X-ray uoroscopy. On the other hand, the plaster bed, which was customized according to the patient's body, was convenient for regulating the position of the patient and preventing intraoperative pressure. In the course of surgery, the patient's position is the head side raised 20-30° prone position. The aim is to reduce the pressure in the spinal canal venous plexus, thereby reducing intraoperative bleeding [37]. The patient's cervical spine is in a slightly exed position, widening the gap between the vertebrae and enlarging the operating space. Disc herniation, especially the prominent below the nerve root, will increase the nerve root tension. When the patient in general anesthesia, the neck muscles will be in a relaxed state. If the cervical spine in the excessive exion, nerve root will be excessive traction, which may increase the rate of iatrogenic injury. with. In this study, there was one case of intraoperative vertebral hemorrhage, which seriously affected the surgical eld of vision, and we had to change the operation plan nally. There are some ways to decrease bleeding in the literature. Using controlled hypotension [38,39], local vasoconstrictor [40] and new hemostatic biological materials [41] can reduce bleeding. It had been reported that the use of hot saline irrigation can effectively reduce the bleeding of functional endoscopic sinus surgery (FESS) [42].
Whether thermal saline can reduce the bleeding in PEPCD surgery, and thermal saline effect on spinal cord need further study.
For the upper and lower lamina decompression range, we believed that to minimize the destruction of the articular process while ensuring adequate decompression of the nerve root. In open surgery, biomechanical studies had shown that removal of more than 50% of articular processes can cause cervical instability [43]. PEPCD surgery reduced the stripping of ligaments and muscles around the spine, so the probability of cervical instability was reduced. In this study, three-dimensional reconstruction of CT showed that with the prolongation of time, the area of the cervical vertebral plate was gradually decreased. The results showed that the bited lamina had the trend of fracture healing under the in uence of biomechanics and internal microenvironment. This healing can further enhance the stability of the vertebral body and prevent the occurrence of postoperative vertebral instability. Therefore, we believe that the rst element in the PEPCD procedure is su cient laminectomy to achieve adequate decompression. Facet joint resection should be controlled within 50%, in the course of surgery it may be appropriate to enlarge the standard according to decompression.
The essence of PEPCD operation is the further optimization and extension of keyhole and MED. The technique of endoscopic technique is used to remove the small joints and the prominent nucleus pulposus through a more minimally invasive and more accurate technique, so as to achieve the purpose of nerve root canal enlargement and nerve root decompression. In this study, we con rmed that the clinical effect of PEPCD in the treatment of CSR is equivalent to that of ACDF, and the surgical trauma is smaller and the postoperative recovery is faster. PEPCD surgery is limited by its own technical characteristics, adapted to cervical disc herniation and small joint hyperplasia induced intervertebral foramen stenosis; contraindications include central cervical disc herniation, cervical spondylotic myelopathy, posterior longitudinal ligament ossi cation or cervical unstable and so on [20,44]. However, with the development of endoscopic techniques, PEPCD has also been applied to multi-segment cervical spondylosis [27], cervical spondylotic myelopathy, and achieved good early postoperative effect. In this study, the number of samples and the follow-up time were limited, and no re-protrusion was observed.
However, there was reports that the re-protrusion of PEPCD was 3.4% for 2 years fellow-up [21]. Reherniation after minimally invasive surgery has been a problem for spine surgeons. As surgery only removed the prominent nucleus pulposus, intervertebral disc rupture still exists, so the residual nucleus pulposus can still re-herniation through intervertebral disc rupture. It had been reported that the use of polymethyl methacrylate (PMMA) closure in the surgery can signi cantly reduce the probability of postoperative re-herniation [45]. Due to the limitations of material technology and lack of long-term follow-up, clinical application is not extensive. Due to the steep learning curve of PEPCD surgery, and the requirement of the ability of understanding the three-dimensional anatomical structure during PEPCD, the development of the operation is limited.

Conclusion
For patients with CSR, PEPCD had similar clinical early and intermediate outcomes when compared with ACDF, with less trauma and faster recovery. PEPCD can be used as an effective treatment for CSR ladder therapy, which is between conservative treatment and interbody fusion. Due to the limited number of cases, short follow-up time, long term clinical e cacy and recurrence need further study.   The changes of C2-7 lordosis in patients with ACDF and PEPCD before and at the last follow-up. A, B: cervical lateral radiographs of PEPCD at the time of preoperative and nal follow-up, respectively. C, D: cervical lateral radiographs of patients with ACDF before and at the last follow-up, respectively. E: C2-7 lordosis data for two groups of patients before and at the last follow-up. Data was mean ± SD, *P<0.05.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.