Staged surgical strategy of symptomatic lumbar adjacent-segment degeneration: full-endoscopic decompression versus extended posterior interbody fusion

Background Symptomatic adjacent-segment degeneration (ASD) is a common complication after lumbar fusion surgery. We want to evaluate the clinical and imaging outcome of full-endoscopic lumbar decompression (FELD) in comparison with extended posterior lumbar interbody fusion (PLIF) for ASD after lumbar fusion surgeries and develop a staged revision strategy. Methods This retrospective study enrolled 65 patients with ASD who treated with FELD (n=31) or extended PLIF (n=34) between January 2014 and January 2018. Clinical outcome evaluations were performed preoperatively, at 3, 12, 24 months postoperatively, including Oswestry Disability Index (ODI) score and Japanese Orthopaedic Association (JOA) scores for function assessment, visual analog scale (VAS) scores for low-back pain and leg pain. Imaging outcome evaluations were performed preoperatively, immediately after surgery and at the last follow up, including disc height (DH) and lumbar lordosis (LL). Results The ratio of patients with adjacent segmental instability in the PLIF group was signi�cantly higher than that in the FELD group (p<0.05). There was no signi�cant difference in other baseline data for FELD and PLIF groups (p>0.05). Mean operative time, blood loss and length of hospital stay were signi�cantly decreased for the FELD group (p<0.001). For patients with stable adjacent segments, there was no signi�cant difference in preoperative and postoperative low-back pain, leg pain and function between 2 groups (p>0.05). The patients in PLIF group had restored DH after surgery, which was signi�cantly higher than FELD group (p<0.05). For patients with unstable adjacent segments, the low-back and leg pain of PLIF group were signi�cantly relieved than that in FELD group within 24 and 3 months after surgery (p<0.05). The function of PLIF group was signi�cantly improved within 12 months after surgery (p<0.05). The DH was only signi�cantly restored within 3 months after surgery (p<0.05). Recurrence was found in 2 (6.5%, p>0.05) patients in FELD group, while no patient in PLIF group had recurrence. Conclusions FELD could achieve satisfactory safety and e�cacy for the treatment of stable ASD, which was not worse than PLIF. With less trauma and faster recovery, FELD may be an alternative surgical treatment for stable ASD.


Introduction
Symptomatic adjacent-segment degeneration (ASD) is a common complication after lumbar fusion surgery, with a high rate ranging from 5.2% to 18.5% [1].It is de ned as the recurrence of low-back pain and leg pain associated with the degeneration at the adjacent segments after a symptom-free period.Spondylolisthesis, instability, herniated nucleus pulposus, osteophytic spurs, and spinal stenosis (SS) have been considered to be the main pathological changes of ASD [2,3], and lumbar disc herniation (LDH) was the most common pathological change among them.Revision surgery has become a regular choice after a failed conservative treatment.
The extended posterior lumbar interbody fusion (PLIF) was performed as a general salvage surgery in lumbar ASD after fusion surgery.However, tissue scars resulting from previous surgery and ASD caused by the damage of vertebral motion segments was considered to be associated with further complications [4].Percutaneous endoscopic lumbar discectomy (PELD) was a minimally invasive surgery used for the treatment of LDH [5,6].With further development of PELD, Kim et al. applied full-endoscopic lumbar decompression (FELD) for patients with SS and achieved satisfactory clinical outcomes [7,8].It directly enlarged the spinal canal, decompressed the spinal cord under local anesthesia and avoided repeated damage to the posterior and paraspinal structures, which made FELD a potential alternative to open lumbar surgery.Minimally invasive technique is effective for preserving facet joint, however, indications must be carefully considered when scoliosis or instability are present [9].
The surgical management of symptomatic lumbar ASD remains controversial, and researches comparing outcomes of revision surgery by means of FELD with extended PLIF are rare.Our study has performed FELD as revision surgery for symptomatic ASD, and the results are satisfactory.Therefore, we attempted to evaluate the clinical and radiologic outcomes of FELD versus extended PLIF for symptomatic ASD resulting from posterior fusion surgery in this study.

Patient population
This is a retrospective study performed after achieving the approval of the Ethic Committee of The Peking university Third Hospital, and patient outcomes were collected independently from informed consent.Case assignment was from the result of the whole section discussion before surgery and eventually determined by experienced surgeons.
From January 2014 to December 2017, 65 consecutive patients underwent FELD (n=31) and extended PLIF (n=35) for symptomatic ASD were enrolled.Radiographic examination, computed tomography, and magnetic resonance imaging were applied for preoperative diagnosis and postoperative evaluation.

Inclusion and exclusion criteria
The eligibility criteria were 1) previous posterior fusion surgery for lumbar degenerative disease; 2) suffered from low-back pain or radicular pain in the lower extremities; 3) failed conservative treatment for at least 3 months; 4) ASD including disc herniation and stenosis con rmed by radiologic images; 5) underwent FELD or extended PLIF revision surgery; 6) with at least 24 months of follow-up.
The exclusion criteria were: revision for infection, lumbar fracture, breakage of instrumentation, pseudarthrosis and other causes.

FELD
There were 2 approaches for endoscopic revision surgeries of lumbar ASD.Transforaminal (TF) approach was applied in patients with ASD at L4-L5 or upper level and interlaminar (IL) approach was applied in L5-S1 level.Special instruments were designed and constructed for foraminotomy and laminectomy, including outer working cannula, trepan, and inner working cannula (Endosurgi, China) (Fig 1).All procedures were carried out under local anesthesia with mixed anesthetic (2% lidocaine 15ml, 1% ropivacaine 10ml and saline 20ml).
The key points of IL approach included: 1) su cient decompression of bony structure to expose the cranial and caudal insertion points of interlaminar ligamentum avum; 2) en-bloc resection of the unilateral hyperplastic ligamentum avum to expose the axillary and under-shoulder space of S1 nerve (Fig 3).

Extended PLIF
Extended PLIF was performed with the patient in the prone position under general anesthesia.The previous fused level and adjacent degenerated level were exposed from the standard middle incision.A standard decompression was carried out in the adjacent segments after laminectomy.The extension of xation and interbody fusion were performed then.

Outcome measurements
The general characteristics of the patients were recorded, including operative time, blood loss, hospital stay.The clinical outcome measures were evaluated preoperatively, 3, 12, and 24 months postoperatively, including visual analog scale (VAS) for low-back and leg pain, Oswestry Disability Index (ODI) and Japanese Orthopaedic Association (JOA) scores for functional disability.The radiographic outcome was measured preoperatively, immediately after surgery, and at the last follow-up, including the disc height (DH), lumbar lordosis (LL), and adjacent-segmental stability.DH was average of anterior and posterior heights of the disc and measured on the lateral radiographs.LL was measured between the upper endplate of the L1 and S1 vertebra.Radiographically instability was diagnosed, if spondylolisthesis >4mm (anterior or posterior translation) was presented at functional imaging, segmental kyphosis >10°, hypermobility in rotation >15°, complete herniated disc, degradation in Weiner Classi cation of two or more grades, complete collapse of intervertebral space, lateral translation >3mm and disc wedging >5° occurred at follow-up imaging at the adjacent level above or below the instrumentation [11].Measurements were performed by a single observer in Picture Archiving and Communication System (PACS) (GE, USA).

Statistical analysis
The data was analyzed using SPSS (version 26.0, Chicago, Illinois, USA).Statistical analysis was performed using the t test for continuous variables or the chi-square test for dichotomous variables.These 2 tests were used to compare the demographic parameters and perioperative and postoperative parameters between 2 groups.Statistical signi cance was de ned as p<0.05.

Baseline characteristics
Table 1 demonstrates the baseline characteristics between the FELD group and PLIF group.There were no signi cant differences in terms of follow-up period, age, duration of symptoms, interval time, sex, BMI, location of ASD, level of ASD, and previous fusion levels (p>0.05).Signi cant difference was found in adjacent-segmental stability (20/11, stable/unstable, in FELD group versus 12/22, stable/unstable, in PLIF group) between the 2 groups (p=0.019).

Perioperative characteristics
The perioperative characteristics of patients are summarized in Table 2. Signi cant differences were found in operative time (63.4±38.9min in FELD group versus 142.4±39.6 min in PLIF group) between 2 groups (p<0.001).Intraoperative blood loss was signi cantly less for FELD group (9.0±8.9 ml) than in the PLIF group (241.5±100.4ml, p<0.001).Patients in the FELD group had shorter time to stay at hospital (4.6±2.4 days in FELD versus 8.9±4.1 days in PLIF group, p<0.001).

Clinical and radiologic outcomes
For patients with both stable and unstable adjacent segments, no signi cant difference was found in terms of preoperative low-back and leg pain VAS, ODI score, and JOA scores between 2 groups (p>0.05).For patients with stable adjacent segments, no signi cant difference was found in terms of postoperative low-back and leg pain VAS, ODI score, and JOA scores between 2 groups (p>0.05,Table 3).However, for patients with unstable segments, the lowback and leg pain VAS were signi cantly higher for FELD group than in the PLIF group within 24 months and 3 months (p<0.05,Table 4).Meanwhile, the ODI score was signi cantly less and the JOA scores was signi cantly higher for FELD group than in the PLIF group within 12 months (p<0.05,Table 4).
For patients with both stable and unstable adjacent segments, no signi cant difference was found in terms of preoperative DH and LL between 2 groups (p>0.05).For patients with stable adjacent segments, the DH was signi cantly restored for PLIF group than in the FELD group immediately after surgery and at the last follow-up (p<0.05,Table 5).However, for patients with unstable adjacent segments, the DH was only signi cantly restored for PLIF group than in the FELD group immediately after surgery (p<0.05,Table 6).For patients with both stable and unstable adjacent segments, no signi cant difference was found in postoperative LL between 2 groups (p>0.05).

Complication and recurrence
No complication was occurred in 2 groups until the last follow-up.Recurrence was found in 2 (6.5%, p>0.05) patients in FELD group and further treatment was required, while no patient in PLIF group had recurrence.

Discussion
Typically, revision for ASD after lumbar fusion was performed via a posterior approach with or without extension of the instrumentation and fusion [12].The extended PLIF can achieve a satisfactory decompression of the spinal cord that results from ASD.However, traditional open spine surgery may lead to laminectomies, muscle damage, ligamentum avum excision and nerve retraction, which are the main causes of spinal instability and epidural space scarring [13].
FELD has signi cant advantages over conventional open lumbar surgery in the above aspects, including performing the surgery under local anesthesia, preserving normal posterior and paraspinal structures, less intraoperative blood loss, less postoperative pain, and short hospital stay.In this study, FELD also showed a shorter mean operative time, less intraoperative blood loss, and less hospital stay than extended PLIF.Furthermore, endoscopic surgery doesn't go through the posterior scar tissue, which may avoid the access-related complications for the revision surgery [14].
In previous study, the use of PELD for patients with recurrence of disc herniation had been discussed [15], while the use of PELD for lumbar ASD was rarely mentioned.A prospective cohort study performed by Ba Z et al. compared percutaneous endoscopic transforaminal procedure (PE-TF) and PLIF in treating disc herniation at adjacent segments after lumbar fusion surgeries, the clinical outcomes of 2 approaches were similar, but PE-TF showed advantage in traumatization, cosmetology, hospitalization time, and blood loss [16].In tradition, disc protrusions without response to conservative treatment is identi ed as the limited indication for PELD.Major neurological de cit, segmental instability, spondylolisthesis, extruded disc prolapse, narrow spinal canal, and lateral recess are all considered as relative contraindications [17].It was reported foraminoplastic superior vertebral notch approach with reamers in PELD may offer safe and e cacious results in the relatively limited indications[18], including high grade migration, high canal compromise, and combined foraminal stenosis, which gave a technical support for the treatment of ASD after lumbar fusion.Furthermore, Xiong et al. showed a safe and minimally invasive way to treat lumbar stenosis combined with disc herniation by using TESSYS-ISEEs technique [10].The TESSYS-ISEEs technique, used in TF approach of FELD, enlarged the narrow intervertebral foramen and spinal canal, which achieved the decompression of the nerve roots and spinal cord for patients with symptomatic lumbar ASD.Meanwhile, the IL approach showed a satisfactory outcome as TF approach in decompression of lumbar lateral recess stenosis [19].
Spinal instability is a frequently cited justi cation for fusion surgery, the true clinical relevance of spinal instability is poorly understood [20].Treatment with minimally invasive technique in patients with segmental instability still remains controversial [9].When we choose the surgical method, the clinical manifestations and radiographic outcomes of patients need to be considered simultaneously.If the patient's main symptoms are radiating pain in the lower extremities due to nerve compression or intermittent claudication, the main purpose of surgery is decompression.If the patient's primary symptom is mechanical low-back pain due to the lumbar instability, the primary purpose of surgery is fusion.Therefore, in this study, we comprehensively analyzed the clinical manifestations and radiographic outcomes to determine whether the patients performed FELD to achieve decompression or extended PLIF to achieve fusion.The patients were divided into 2 groups based on the radiographic stability of responsible segment next to the fusion segments.For the patients with stable ASD, FELD yielded clinical outcomes not worse than extended PLIF even if the extended PLIF could achieve a more effective restoration in DH.For the patients with unstable ASD, extended PLIF could get better results in relieving low-back pain and leg pain, and improving the function of ASD.Meanwhile, extended PLIF could only achieve a more effective restoration of DH within 3 months.Therefore, we believe that the e cacy of FELD in treating stable lumbar ASD is reliable and a better restoration in DH may be found in patients with stable adjacent segments under extended PLIF.
Furthermore, the recurrence after surgery should also be noted.In previous studies, the recurrence rate after FELD was reported to be 0%-6.9%[21][22][23].In contrast, recurrence rate after extended PLIF was relatively low [16,24].In this study, 2(6.5%) patients underwent FELD had recurrence, while there was no patient underwent extended PLIF had recurrence.Although recurrence rate may be higher in FELD group, no signi cant difference was found.Meanwhile, no complication was occurred in 2 groups until the last follow-up.Therefore, although the recurrence rate is relatively higher in patients with treatment of FELD, the safety of FELD in treating stable lumbar ASD is still reliable.
Although our study showed some reliable results, there were a few limitations.First, since it was a single-centered retrospective study and our sample size was relatively small, the generalizability of our ndings could be limited.A randomized control trail may be more convincing.In addition, the follow-up period is relatively short and a long-term follow-up of a larger sample will be required in the future.

Conclusions
FELD could achieve satisfactory safety and e cacy for the treatment of stable lumbar symptomatic ASD, while extended PLIF had an advantage over FELD for patients with unstable lumbar symptomatic ASD.For patients with stable ASD, low-back pain and leg pain, lumbar function are improved without a signi cant difference between FELD and PLIF.Meanwhile, extended PLIF could achieve a more effective restoration of DH during the mid-term follow-up.For patients with unstable ASD, extended PLIF has better effect on relieving low-back pain and leg pain during the midterm follow-up.Extended PLIF could also achieve better effect on improving the function of ASD in the rst year after surgery.Meanwhile, extended PLIF could only achieve a more effective restoration of DH within 3 months.FELD may serve as an alternative surgical treatment for stable symptomatic lumbar ASD with less trauma and faster recovery, while extended PLIF still dominates the unstable symptomatic lumbar ASD.

Posterior lumbar interbody fusion (PLIF);
Oswestry Disability Index (ODI); Japanese Orthopaedic Association (JOA);     facet joint.e.The inner margin of L5 inferior articular process and S1 superior articular process was removed to expose the outer margin of ligamentum avum.f.The trepan was pointed to the superior margin of the S1 lamina.g.The superior margin of the lamina and articular process was removed to expose the caudal insertion point of 9 22.6±13.00.591 23.3±12.826.5±11.90.489 22.1±12.027.2±10.50.225 FELD, full-endoscopic lumbar decompression PLIF, posterior lumbar interbody fusion

Figures Figure 1
Figures

Table 1 .
Baseline data of patients in two groups

Table 3 .
Clinical data of patients with stable adjacent segments in two groups FELD, full-endoscopic lumbar decompression PLIF, posterior lumbar interbody fusion ODI, Oswestry Disability Index JOA, Japanese Orthopaedic Association VAS, visual analog scale

Table 4 .
Clinical data of patients with unstable adjacent segments in two groups

Table 6 .
Imaging parameters of patients with unstable adjacent segments in two groups