Unilateral Tubular Approach for Bilateral Laminotomy vs Transforaminal Lumbar Interbody Fusion for Lumbar Spinal Stenosis

Background: Endoscopic techniques is considered to be the standard tissue sparing approach for surgical advantages, rapid rehabilitation, and instability prevention. PEID techniques were introduced to treat lumbar spinal stenosis. However, its effectiveness and safety in the treatment of moderate and severe lumbar stenosis is uncertain. Methods: Between June 2014 and June 2016, patients suffering from lumbar spinal stenosis underwent a bilateral laminotomy through a unilateral tubular approach. Demographics, intraoperative data, and patient-reported outcomes were ascertained Results:There were no signicant differences in patient demographics between the two groups. The statistically signicant improvements in leg and back VAS scores were reported on the ipsilateral and the contralateral side (p<0.05, respectively). Analysis of VAS and ODI scores demonstrated no signicant differences in improvement on both groups. There were no difference in radiographical and intraoperative outcomes during the follow-up period. Conclusions: Regarding the intraoperative data, and patient-reported outcomes, the PEID is a safety, and ecacy minimally invasive technique. This observation suggests that the PEID techniques were excellent or showed good results.


Introduction
Lumbar spinal stenosis (LSS) is caused by discal, bony, ligamentary, or capsular structures. LSS is the most common disease accompanied by nerve compression in the intervertebral foramen, lateral recess, or central vertebral canal, resulting in low back pain, leg pain, or neurogenic claudication 1 . The association between the extent of LSS in imageology examination and the clinical symptoms is unequivocal 2 . Fusion is routinely adopted to limit postoperative instability in patients with LSS 3 . Adjacent level degeneration caused by instability may result in deterioration of symptoms 4,5 . In therapeutic terms, decompression with instrumented fusion is su cient in elderly patients meet the indication for lumbar spine surgery 6 . The traditional approach involves bilateral paraspinal muscle dissection, extensive removal of the spinous processes, supraspinous, and interspinous ligaments, wide laminectomy, increasing the incidence of complications such as infection, neurologic de cits, and adjacent segment degeneration 7 . These perioperative complications have been demonstrated to be associated with various problems, such as a prolonged decubation, increased incidence of spinal instability, and increased reoperation rate 8 . Further, failures of operationattribute to paraspinal muscle atrophyand persistent back pain, caused by excessive intraoperative dissection of the paraspinal muscule anddenervation injury 9 .
To overcome these problems following the conventional approach, we adopt the " BUDA with PELD" technique which describes bilateral decompression via a ipsilateral approach to treat LSS 10 .We present here a case series of patients who underwent " BUDA with PELD" procedures for the treatment of severe LSS and assess the clinical outcomes of this procedure.

Study Design
Electronic medical record of 105 patients who underwent a BUDA with PELD and TLIF procedure from June 2014 to June 2017were identi ed.The basic demographic and clinical data were as follows: age, gender, body mass index (BMI), smoking status, Charlson Comorbidity Index (CCI), chief complaint, surgical levels, comorbidity. This retrospective case study was approved by the ethicalreview committee of our institution (Grant No.K-20190036-W). All patients signed a consent preoperatively for participation in this clinical studyinvolving a BUDA with PELD and TLIF procedure for LSS.All the patients suffered from low back pain were diagnosed with moderate or severe LSS via computed tomography (CT) scans and magnetic resonance imaging (MRI). The main cause of back pain was identi ed by a CT scan and whether the back pain was caused by symptomatic LSS was determined by an MRI examination. The operations were performed by two surgeons with abundant experience in lumbar surgery.

Data Collection
Questions were asked to patients or their relatives over the phone and spine surgeons from our institutions completed the outcome measurements for them. Primary outcome included the Visual Analogue Scale (VAS),Oswestry Disability Index (ODI) and the modi ed Macnab Criteria(Macnab) weremeasured to patients preoperatively and postoperatively during the follow-up period. In addition, the estimated blood loss, procedure duration,and perioperative adverse events were evaluated. Secondary outcome including the incision size, the time for ambulation, and the time for hospitalization.
Operative Procedure All patients were treated by the standard procedures. The procedures were performed successfully under the guidance of a C-arm angiographic unit in all patients. The patients were placed on the operating table in a beach-chair position under cervical plexus block anesthesia or general anesthesia.Patients were positioned prone on a Jackson table. Under general anesthesia, a 2-cm skin incision was made overlying the target level on the posterior approach in a midline of back. Unilateral tubular approach for bilateral laminotomy, as described by as Alimi et al. 11 , was performed. The inferior edge of the lamina and the inferior edge and base of the spinous process were exposed. Then Ausculap channel expander was inserted.Removal of the thickened lamina with a 5 mm diameter grinding brick under micromanipulation.
When the lamina became thinner, a 5 mm diameter diamond bricks was used to remove the residual bone structure. (Figure1) A step by step decompression was performed starting with an ipsilateral partial laminotomy, using a bayonetted 2 and 3 mm Kerrisson punches. The tubular retractor was angled mediallyand the microscope and operating table were tilted tilting to the affected side to get clearer visualization of contralateral sublaminar structures. Then, the basal part of spinous process was removed using the drill. Neural root dissector protects dura and contralateral nerve roots and contralateral lamina was removed by a Kerrisson rongeur.Meticulous and complete decompression of the contra lateral recess and removal of the yellow ligament was accomplished. The nal step was returning the table and retractor to the initial position and completion of the ipsilateral decompression.(Figure2) Statistics Summary statistics reported as mean±standard deviation (SD). The demographic and surgical variables between the two groups were compared by univariate analysis. In addition, the categorical variables were tested by Fisher's exact test and the continuous variables were tested by Mann-Whitney test.The statistical signi cance level was P<0.05.

Results
A total of 25 patients underwent BUDA with PELD and 53 patients underwent TLIF were retrospectively analysed. No signi cant differences was found between the two groups with respect to age, gender, BMI and steroid medication and prior smoking history(p>0.05, Table 1). The BUDA with PELD group had signi cantly lower procedure duration, estimated blood loss and length of hospitalization than TLIF group (p<0.05, Table 2). No signi cant differences were found in complications or reoperations at 1-year follow-up between both groups. No serious intraoperative complications, including cerebrospinal uid leakage, spinal cord injury, were observed. In addition, no signi cant difference was observed in terms of anteroposterior diameter of spinal canal, spinal canal area, and adjacent segment disease. CT examination showed that lumbar spinal canal had been expanded by the surgery procedure, and we observed no sinking lamina or broken shaft. An MRI examination revealed that high signal intensity of the spinal cord decreased and no further spinal cord compressions. The MRI of radiological examinations shows canal size of lumbar spine between preoperative to postoperative late following-up.All patients reported signi cant improvement in neurological de cits within 2 weeks postoperationly.
The average VAS score, ODI score, and the Macnab showed a signi cant improvement at different time point postsurgery, showing satisfactory outcomes. However, table 3 indicated that no signi cant difference in VAS score and ODI score was found between the two groups in different follow-up time period (P>0.05, Table 3).

Discussion
Lumbar spinal stenosis (LSS) is one of the most frequent etiologies of chronic back pain, neurogenic claudication, and lumbar radiculopathy. Bilateral dissection and retraction of the paraspinal muscles, wide laminectomy, medial facetectomy and foraminotomy were traditional decompression strategy involved in treatment of LSS. Excessive disruption of supraspinous/interspinous ligament complex and posterior bony contribute to spinal instability, which eventually increase the rate of surgical failures and revision surgery 12,13 . Meanwhile, minimally invasive laminectomy approaches have been widely recognized to be associated with decreased approach related morbidity.
Eventhough PEID was associated with less blood loss, narcotic use, and satisfactory clinical outcomes in previous studies. However, there is a lack of clinical evidence to com rm its safety and e cacy comparing PEID to TLIF. The current study shows that ULBD technique is a viable option for patients with moderate and severe LSS 14 . Absence of muscle dissection with the endoscope and instruments over the lamina and preservation of the soft tissue structures, could alleviate the muscle atrophy and spinal instability 15 . Furthermore, another advantage of ULBD lies in its mitigation the risk of adjacent segment fusions postoperative, while adjacent segment disease following fusion involves fusing a second level in traditional revision surgery 16 .Also, the reduction in duration of hospitalization was observed.The complication rates in the literature for TLIF have an average of 36,7%, whereas minimally invasive laminectomy approaches have 28,4% 17 . The present study found that patients underwent PEID approach show good clinical outcomes and pain scores at 3 and 12 months after operation. Even though decompression with instrumented fusion was widely used in the past, clinical experience demonstrate that decompression alone can lead to satisfactory patient-reported outcomes 18 .Complex MIS surgeries, such as microdiscectomy, transforaminal lumbar interbody fusion, are built on the foundation of simple laminotomy. At present, there is no consensus providing a treatment algorithm for LSS due to sagittal imbalance and severity of stenosis. While there is sort of a big academic debate on surgical indications of fusion for LSS, TLIF was proven to be improve lumbar function and spinal stabilization 3,19 . TLIF allow weight-bearing through the anterior column and would prevent the stress concentration that may increase the instability of the adjacent vertebra 20,21 .
PEID was reported to treat several degenerative lumbar disease, such as degenerative scoliosis and spondylolisthesis, lumbar stenosis, and thoracic disc herniations 14,22,23 . Previous reports have already demonstrated that ULBD is a effective and safe approach to treat degenerative spondylolisthesis 10 .Several clinical study focused attentionon the clinical e cacy of ULBD for the treatment of severe lumbar stenosis. Komp et al 24 conducted a prospective, randomized, controlled research and demonstrated that full-endoscopic ULBD is an is safe and effective strategy for treating LSS. In this study, 71 patients demonstrated constantly and obviously improved back pain and leg pain and daily activities. even though slight deterioration was observed.No patient suffered worsening of back pain. Postoperative pain symptom and analgesics, such as ibuprofen and paracetamol, show signi cant improvement.Furthermore, the rate of complications related to the procedure was signi cantly reduced during the follow-up. The maximum duration of hospitalization was shortened as a result of using the full-endoscopic technique. Referring to complications, we found the incidence of dural tears increased slightly in the in the ULBD group. Mobbs et al. showed that incidental durotomy rates was 3.7% in ULBD group and open surgery group 25 ,whereas incidence of dural tears in classic laminectomy may range from 5% to 15% 26 . Only 1 incidental durotomy happened in our study and we found it challenging to identi cate and suture the dural tears through a smaller surgical channel. In addition, postoperative lumbar instability was not observed in the PEID group. This study con rm that length of hospitalization and estimated blood loss were less in our operation, which is consist with the previous literature.
When it comes to clinical outcomes, the success rate demonstrated no signi cant difference between the 2 groups, although a slightly better patient-reported outcome after PEID can be seen; in the long term, PEID showed equally effection in improving pain and function when compared to fusion. ULBD facilitates neural decompression while preserving stabilizing osseoligamentous structures and may be uniquely suited for the treatment of LSS with concurrent mild to moderate degenerative deformity. Thus, the main advantages of this less invasive technique lie in the reduction of postoperative instability and more suitable for patients with lumbar spinal stenosis and severe osteoporosis. This study found that PEID did not have any advantage in patient-reported outcomes compared with TLIF techniques after surgery.
Additionally, this surgical procedure can be an effective treatment of other degenerative lumbar diseases.Thoracic disc herniations were rst reported by Yüce et al 23 to be alleviatied by ULBD approach.
A previous study reported that patients with thoracic disc herniation should undergo fusion 27 . However, fusion surgery in icts traumatic injury than ULBD, leading to a longer hospitalization. Other methods, such as physical therapy, medication treatment, and nerve root block could ease the back pain caused by disc herniation 28,29 . We strongly recommended PEID to alleviate pain if the conservative treatment has no any effect on patient.
Some limitations should be addressed in current study. First, this retrospectively study has less creditability. Hence, future prospective studies should be performed to evaluate the effectiveness and safety of the PEID procedure. Second, only 10 patients, a small sample size, were enrolled in this study, which makes the results inaccurate. Third, patients who underwent PEID procedure were not compared with control procedure in this study.

Conclusions
Our study demonstrated that PEID approach is an effective procedure for treatment of lumbar stenosis (LSS). The results of the present study demonstrated that patients underwent PEID procedure showed effective pain relief and functional improvement throughout the medium-term follow-up. Besides, there were no complications and the improvement in patient pain continued throughout the long-term followup. Large-scale prospective studies are required to assess the effectiveness of PEID procedure in the treatment of elderly patients with moderate to severe LSS. Most surgeons familiar with the procedure agree that the PEID approach appears to afford a more complete decompression of the contralateral thecal sac and nerve root, as compared to the ipsilateral approach, reducing iatrogenic injury of bone and soft tissue. Further, open surgery is recommended as a secondary approach if PEID procedure has proven unsuccessful.  A typical80-year-old female patient who underwent L4-5 ULBD.