Clinical Study of Unilateral Extraforaminal Lumbar Interbody Fusion (ELIF) Revision For Lumbar Disc Herniation After Primary Discectomy


 Background: Lumbar disc herniation (LDH) is a common spinal disorder. The discectomy with non-fusion operation is widely used. When the revision is needed, the options for revision way is still controversial. This study aims to introduce unilateral extraforaminal lumbar interbody fusion (ELIF) revision surgery, and to investigate the clinical efficacy and complication of Unilateral Extraforaminal Lumbar Interbody Fusion (ELIF) in revising primary discectomy for lumbar disc herniation. Methods:16 patients with incomplete removal and recurrence herniation of lumbar disc after minimally invasive treatment without fixation were treated by ELIF with unilateral pedicle screw(UPS) from April 2016 to October 2020. All those patietns including 11 male and 5 female aged 29-65 years were analyzed retrospectively. The clinical effects were evaluated by operation time, intraoperative blood loss, postoperative blood volume of drainage and complications. The Visual Analogue Scale(VAS), Oswestry Disability Index (ODI) were documented before, after surgery and at last follow--up. Results: The operation time was 95.73±10.5 min, the bleeding volume was 201.5±27.6 ml. Postoperative blood volume of drainage was 50.7±6.3 ml. 2 patients suffered dura tear. All patients were followed up for 12-26 months with 15.7 on average. VAS and ODI scores significantly improved at the preoperative, postoperative and the 12th month’s follow up (P < 0.05). Conclusion: The application of ELIF with unilateral fixation is a satisfied way to revise primary discectomy for lumbar disc herniation. However, the dura tear induced by scar tissue adhesion needs to be noticed.


Background
Lumbar disc herniation (LDH) is a common spinal disorder which mainly caused by degeneration and overloading pathological factors, and the incidence of LDH is increasing [1] . Herniated intervertebral disc tissue stimulates nerve roots or (and) dural sac resulting in lower lumbar and leg pain and neurological symptoms [2] . Surgical treatment will be recommended when the systematic conservative treatment failed.
The non-fusion operation for LDH is widely used during the treatment for its advantages of minimally invasive, satis ed e cacy and quick recovery. Traditional partial laminectomy with discectomy(TPLD), microscope discectomy (MD),microendoscopic discectomy(MED) and percutaneous endoscopic lumbar discectomy(PELD) were classical non-fusion methods for treating LDH in last several decades [3] .
However, there are still some issues after non-fusion surgery involving incomplete removing the nucleus pulposus, recurrence of intervertebral disc herniation or failed back surgery syndrome (FBSS), et al. The options for revision way is still controversial. [4,5] . This study aims to introduce unilateral extraforaminal lumbar interbody fusion (ELIF) revision surgery for lumbar disc herniation after discectomy without fusion at rst time. To our current knowledge, this kind of application has not been reported.

Inclusion and Exclusion Criteria
Inclusion criteria: Patients with lumbar disc herniation have underwent surgery of TPLD, MD, MED, PELD.
Pathological factor in single segment caused unilateral symptom after non-fusion surgery. Followed up after ELIF surgery more than one year.

Surgical Procedure
Patient was put in a prone position under general anesthesia. The upper and lower pedicle shadows of lumbar segment on pathological side were identi ed and marked on skin by AP view uoroscopy. A longitudinal incision between the upper and lower pedicle shadows was made after conventional surgery area skin disinfection. Super cial fascia and lumbar dorsal fascia were opened, and the Wiltse approach (the gap between the longissimus and multi dus muscle) was used in muscle layer to reach the outer edge of superior and inferior facet joints. Two self-design retractors opened soft tissues to identi ed pedicle screw entering point on the upper vertebral body. This one pedicle screw was rstly placed after correct preparation. There self-design retractors were properly placed under the assistance of already implanted screw to create a square surgical space. Proper resection of the ventral part of superior articular process of lower vertebral body to enlarged Kambin triangle space for intraspinal and intervertebral space management in the next step. Operation was performed to remove of residual or reherniated nucleus pulposus and hypertrophic scar tissue induced by original surgery. Subsequently, discectomy, endplate preparation, bone graft and cage implantation were carried out step by step. Lower pedicle screw was placed like the rst one, and one pre-bent titanium rod was put and locked on screws after longitudinal proper compression. Position of cage, internal xation and lumbar lordosis were evaluated by intraoperative uoroscopy. The drainage tube was placed, and wound was sutured layer by layer ( Figure 1. ELIF procedure) .

Rating Parameters
Duration of surgery, intraoperative blood loss, postoperative drainage were recorded. ODI and Vas were recorded one day before operation, 2 weeks and 12 months after operation. The complications both in perioperative and follow-up period were documented.

Statistical Analysis
The SPSS 21.0 software was used for statistical analysis. The measurement data were recorded as mean ± standard deviation(χ̅ ±s). The t test was used to compare the scores of ODI and VAS before and after ELIF surgery. P-value <0.05 was regarded as signi cant statistical difference.

Results
The ELIF operation was successfully performed for all patients. The operation time was 95.73±10.5 min, the bleeding volume was 201.5±27.6 ml. Postoperative blood volume of drainage was 50.7±6.3 ml. Two patients suffered dura tear and no cauda equina nerve or nerve root injury has happened. The incision suture was removed 12-14 days after operation, and one dura tear patient was removed suture at17 th day after operation because of cerebrospinal uid leakage. Patients were followed up for 12-26 months with 15.7 months on average. The scores of ODI and VAS at one day before operation, 14 days ( average discharge days) and 12 months after operation were signi cantly improved (Table 1). No fusion or xation failure and cage subsidence was observed during whole follow-up period. Although it was not essential after intervertebral fusion, two young patients have required to remove their internal xations at 1.5 and 2 years after operation according to their own wishes. (Figure 2. one classic case) 6±4.9*** Note * is for P 0.05,** is for P 0.01,*** is for P 0.001.

Discussion
Lumbar disc herniation is a disorder of high incidence at present. The protrusion or prolapse of nucleus pulposus compress nerve root or cauda equina leading to low back and leg pain, spinal non-structural scoliosis and even cauda equina syndrome, et al [6] . Surgical treatment always be recommended when the systematic conservative treatment does not work. Discectomies like TPLD, MD, MED and PELD are classic and universal applications [7] [8] . These methods have obvious advantages of minor tissues injury, less blood loss, fewer scars and faster recovery due to early postoperative pain relief. However, incomplete removing the nucleus pulposus, LDH recurrence or FBSS about these methods are inevitable [9] [10] . The LDH recurrence rate about discectomy is about 15%, and this rate increases with the observation time going longer [11] . OD/MD, MED, PELD have overall complication rates of 16.8% / 16.1%, 21.2% and 5.8%, respectively [12] . The middle and long-term complications after discectomy involving the height of intervertebral space lost, hypertrophy facet joints and instability of spine also can cause the re-surgical symptoms. Therefore, revision surgery for lumbar disc herniation after primary discectomy deserves further concern.
The formation of scar tissue is a common condition after the primary surgery [13] . To deal with these postsurgical scar tissues is a crucial step during the revision surgery. It is summarized that these scar tissues are classi ed as extraspinal and intraspinal canal types. Posterior approach is adopted for TPLD, MD and MED to treat LDH. Once the revision surgery is need for patients who have underwent these primary surgeries, PELD is an option according to the previous report [14] . PELD from lateral posterior approach could totally avoid extraspinal canal scar tissues and partially avoid extraspinal canal scar tissues. However, this PELD revision may meet failure sometimes. In this study, a 37 years old male presented with low back and leg pain after endoscopic MED for L5/S1 disc herniation in July 2015. There was a relapse of nucleus pulposus caused serious symptom in August 2017. PELD was performed for revision in October. It was found that the herniated nucleus pulposus incomplete removal caused by the intraoperative extradural scar tissue adhesion Figure 3 . Additionally, multiple non-fusion surgeries for spinal column also has potential instability risk.
Lumbar interbody fusion is still the gold standard to relieve neurovascular compression and reconstruct lumbar sequence [15] . At present, many fusion technologies are used clinically including posterior lumbar interbody fusion (PLIF), transforaminal lumbar interbody fusion (TLIF), anterior lumbar interbody fusion (ALIF), oblique lateral interbody fusion (OLIF) and so on, which are effective methods for the treatment of lumbar [16] . In 1992, Kabins [17] rst proposed unilateral TLIF pedicle screw for intervertebral fusion. This unilateral xation can not only reduce the damage of spinal structure, but also save the medical cost. In 2002, Phillips [18] rst introduced the extraforaminal lumbar interbody fusion (ELIF). This surgical method revealed satisfactory curative effect, immediate postoperative stability and high fusion rate in the treatment of lumbar degenerative diseases. In view of its obvious advantages, scholars have achieved satisfactory results in exploring the mechanical research and clinical reports of unilateral ELIF [19] .
It is found that , as compared with PLIF,TLIF, ALIF, OLIF, the ELIF is an appropriate revision method for the recurrence or residual of lumbar disc herniation after initial non-fusion surgery. Wiltse approach enters the gap between the longissimus and multi dus muscle to reach the outer edge of superior and inferior facet joints, which causes minor injury and avoids scar tissue like PELD. Scholars has reported that Wiltse approach can signi cantly reduce postoperative pain and reduce the incidence of postoperative paraspinal muscle atrophy and fat liquefaction [20] . In this study, VAS score of preoperative low back pain and ODI score signi cantly decreased from (6.3±2.9 and 70.9±15.3) to (3.4±0.9 and 40.4±9.3) respectively at the rst assessment after operation. In addition to relieving the nerve compression in spinal canal, the protection of soft tissue during operation provided a certain effect. With preserving the facet joints through extraforaminal approach, only partial resection the ventral bone of superior articular process to enlarged Kambin triangle space for the next step's intraspinal and intervertebral management. Meanwhile, pedicle screws and cage under ELIF surgical led can be inserted with a larger abduction angle. Additionally, Retaining the mechanical support including the posterior ligament complex and the middle column can improve the immediate stability and signi cantly reduce the in uence for adjacent segments. Therefore, the unilateral xation of ELIF can achieve satis ed e cacy in lumbar degenerative diseases including LDH revision surgery , and ELIF obtains signi cantly higher stability than TLIF in unilateral xation [21] . The well initial stability was not only an important factor for interbody fusion, but also a guarantee that the patient's symptoms can be continuously improved in one year VAS 2.6±0.8, ODI 14.6±4.9 . Apart from the symptom improvement, the amount of bleeding, during the ELIF revision, was 201.5±27.6 ml and the drainage volume was 50.7±6.3 ml. Therefore, unilateral xation of ELIF is an effective and feasible revision way.
Wiltse approach under quadrant channel is a minimally invasive and classic application for lumbar interbody fusion, but quadrant surgical instruments is of high cost. In this study, three self-designed retractors ( Fig. 4 A B, one with arc upper tip marked as α and two with arc lower head marked as β and γ) constructed a channel along Wiltse approach for the ELIF operation. First, two retractors α and β were maintained soft tissue channel to place one upper vertebral pedicle screw. The arc upper tip retractor α was placed on the outer edge of screw insertion point. The pedicle screw tail was used as a fulcrum to place the retractor β for blocking the dorsal soft tissue of lamina. The retractor γ was placed in inner inferior margin of lamina, which was against retractor β. (Fig. 4 C) Then, a stable and wide surgical eld was presented. The tension of skin reduced by relaxing the retractor in time to avoid skin ischemia. The lower pedicle screw was implanted after cage implantation. The visual eld occlusion by screw tails would happen during operation if two pedicle screw tails were inserted at the beginning . Therefore, three self-designed retractors combined with optimized pedicle screw placement order have constructed an ELIF revision channel with low cost and effective way.
Some issues and risks in ELIF for revision of non-fusion surgery need to be noticed as follow. To deal with the extradural scar tissue induced by original surgery is a crucial step. In this study, the scar tissue adhered to the ventral and lateral dura mater or never root were precisely separated by micro scissors.
During this sharp dissection procedure, dura mater and never root needed to be carefully medially retracted, and this slight pulling force created a space which let micro scissors to touch annulus brosus layer as lower as possible for separation. The fragility extent of the dura mater is increased after rst surgery. The risk of dura injury during revision operation should not be ignored. In this study, one patient had a relapse after 2 years of MED treatment for disc herniation, and underwent failure of PELD revision again. The dura tear occurred during the ELIF revision. The arti cial dura mater was cut and attached to the tear part. Postoperatively, patient was advised to keep supine position on bed, and his bed was caudally raised with 20 cm at height, and preventive application of antibiotics which was able to pass through blood-brain barrier was carried out. The amount and color of drainage uid were observed in time and intravenous liquid supplement was performed. When the volume of drainage began to reduce, intermittent clamp and release drainage tube was applicated combined with the observation of wound's aseptic dressing. The drainage tube was removed once the skin dressing was on dry condition under persistent clamp condition of drainage tube (at the 4 th day after surgery). The incision suture removal time was delayed according to the skin healing condition (17 days after surgery). ELIF retained the bone of posterior column structure. Therefore, the intervertebral autogenous bone implantation shortage was inevitable. Allogeneic bone was used in this study. However, lumbar disc herniation occurs mostly at L4/5 and L5/S1 lever, proper caudal enlarge soft tissue can create a tunnel beneath skin. It will be a feasible operation to harvest posterior superior iliac spine autogenous bone under this tunnel. The risk of nerve outlet root injury needed to be cautioned during intervertebral fusion cage implantation. In the selection of cage, too wide head shape should be avoided. The nerve outlet root needed to be identi ed and protected during the bone and cage implanted through the outside of intervertebral foramen.

Conclusion
This retrospective study concludes that: for patients with recurrent or residual lumbar disc herniation after discectomy without fusion treatment, unilateral ELIF revision through Wiltse approach has theoretical feasibility and is of actual advantages of minimally invasive, controllable risk, satis ed e cacy and quick recovery, which is worthy of further large sample study. Patients with FBSS and lumbar segment instability after discectomy without fusion can also be included in further study too. However, it is particularly worth mentioning that risk of management for hypertrophic scar tissue induced by original surgery should not be ignored, and dura tear and cerebrospinal uid leakage should be under predictable and reasonable management.  Figure 1 Wiltse approach between the longissimus and multi dus muscle (green arrow) The ventral part resection of superior facet joint of lower vertebral body red part to enlarged Kambin triangle space for intervertebral management.

Figure 3
Green arrow in A showed the recurrent prolapse of nucleus pulposus. Red arrow in A showed the partial lamina resection by the previous MED operation. Yellow arrow in B showed residual nucleus pulposus after PELD revision. Figure 4