Comparison of clinical outcomes achieved with oblique lateral interbody fusion with or without supplementary instrumentation in patients with single-level lumbar degenerative disease

Background Oblique lateral interbody fusion (OLIF) is applied often to treat degenerative disc disease in the lumbar spine. Stand-alone OLIF prevents morbidities associated with supplemental xation and is less expensive. However, it remains controversial whether stand-alone OLIF is sucient to avoid subsidence for single-level diseases. Additionally, bilateral pedicle screw (BPS) and bilateral transfacet screw (BTS) xation are well-established posterior xation methods that can offer improved biomechanical stability. But the comparison of clinical outcomes of OLIF with and without supplementary instrumentation is lack. Methods We retrospectively examined 20 patients who underwent single-level stand-alone OLIF for symptomatic lumbar degenerative disease at L1–L5 (SA group). Groups of patients treated with OLIF plus BPS (n = 20, BPS group) or BTS (n = 20, BTS group) were matched for age, sex, diagnosis, operative level, body mass index, and bone mineral density. The disk height index (DHI), segmental lordotic (SL) angle, and lumbar lordotic (LL) angle were measured preoperatively and at 3 days and 6 months postoperatively. Clinical outcomes were evaluated.


Introduction
Oblique lateral interbody fusion (OLIF) is a minimally invasive retroperitoneal surgical procedure commonly used for treating degenerative disease in the lumbar spine [1,2]. The OLIF procedure was rst described in 1997 [3,4] and evolved from techniques such as direct (extreme) lateral transpsoas interbody fusion (DLIF or XLIF) and anterior lumbar interbody fusion (ALIF). During OLIF, the disc space is accessed from the space between the psoas and major vessels of the abdomen without splitting the psoas muscle.
The OLIF technique aims to avoid the major vascular/visceral injuries that occur during ALIF; the paraspinal/soft tissue trauma that occurs during transforaminal lumbar interbody fusion, posterior lumbar interbody fusion, and posterolateral fusion; and the lumbar plexus nerve injuries that occur during DLIF and XLIF. The reported bene ts of OLIF include decreased blood loss, improved postoperative pain due to less muscle retraction and smaller incisions, a reduced infection rate, a shorter length of hospital stay, an increased disc space height, a higher fusion rate, and avoidance of the need for further surgery [5,6].
The e ciency of OLIF is partly due to the extension-distraction moment arm applied to the anterior and middle columns of the spine. The resultant ligamentotaxis and restoration of disc height can indirectly decompress the spinal canal and neuroforamina and alter the spinal alignment to relieve claudication, radicular symptoms, and back pain [7]. Graft subsidence into one or both of the adjacent vertebral bodies may compromise this indirect decompression and correction of the lumbar lordotic curvature, causing the return of symptoms, new-onset back pain, neurological de cits, and, in more severe cases, fracture of the vertebral body itself [8][9][10]. In the multicenter survey involving 155 patients, Abe et al. observed a subsidence rate of 18.7% that resulted in the need for reoperation in 1.9% of patients [11].
Stand-alone OLIF prevents morbidities associated with supplemental xation and is less expensive. However, without posterior column support, the rates of interbody cage subsidence and restenosis may be increased [12]. Liu et al. suggested that OLIF with a cage alone may be insu cient to provide adequate stability for 3 or more levels of fusion, which could contribute to subsidence [13]. To the best of our knowledge, it remains controversial whether stand-alone OLIF is su cient for the single-level condition [14,15]. A variety of lateral and posterior xation methods are available to supplement OLIF with varying capabilities to restrict range of motion (ROM) [13,16]. Bilateral pedicle screw (BPS) xation is a common technique that can offer improved biomechanical stability in all directions [17]. BPS is associated with a slight, but signi cant increase in complications due to the risk of screw malpositioning and wound healing problems and is relatively more expensive [18]. Bilateral transfacet screw (BTS) xation is a common and well-established posterior xation method [19]. Biomechanical studies have demonstrated that the equivalent stiffness of the BTS technique is comparable to that of the BPS technique [20,21]. BTS also provides reliable and safe results, including symptom relief and a low complication rate [22]. Thus, the research supports that percutaneous BTS is an attractive surgical alternative for treating single-level spinal fusions. In the present study, we used these two techniques for supplemental xation in patients undergoing OLIF.
The present study aimed to determine stand-alone OLIF could provide satisfactory clinical outcomes for patients with lumbar degenerative diseases treated using single-level fusion and to compare the radiographic and clinical outcomes after stand-alone OLIF with those achieved using OLIF with BPS or BTS as supplementary instrumentation.

Patients and indications
This retrospective study examined the patients who underwent single-level OLIF at L1-L5 in Beijing Ji Shui Tan Hospital between January 2018 and April 2020. The inclusion criteria were as follows: (1) symptomatic lumbar degenerative disease refractory to conservative treatments, (2) degenerative changes mainly concentrated in one segment of the lumbar spine with no or mild degeneration on the adjacent level, and (3) lack of facet fusion, severe lumbar stenosis, developmental lumbar stenosis and free disc fragment/facet cyst in imaging, and an improvement in low back and/or leg pain of greater than 50% during rest. The exclusion criteria were as follows: (1)  QCT was performed to assess patients' BMD preoperatively. Postoperative data were collected on day 3 and at 6 months after surgery.

Surgical techniques
All procedures were performed by three senior spine surgeons who had experience in the OLIF procedure. The choice of surgical techniques for a certain patient was based on the experience and preference of the surgeons, because there is no de nitive consensus regarding the optimal supplementary instrumentation.

OLIF procedure
After general anesthesia intubation, the patient was placed in the lateral decubitus position, and uoroscopy was performed to identify the target intervertebral disc. A 6-cm skin incision was made in the lateral abdominal region parallel to the iliac crest. The abdominal muscles were dissected sequentially with a muscle-splitting approach. The retroperitoneal space was then accessed by blunt dissection, and the peritoneal content was mobilized anteriorly. The space between the psoas major and the aorta was bluntly dissected using ngers. After exposure of the target intervertebral disc and the lateral side of the adjacent vertebral body, uoroscopy was performed to con rm the proper level before proceeding with interbody fusion, and the tubular retractor system was docked. The intervertebral disc was removed, and both the endplates were prepared for cage insertion. Subsequently, an OLIF polyetheretherketone cage of appropriate size (Clydesdale Spinal System, Medtronic, 18-mm width, or Oracle system, Synthes, 22-mm width) lled with allograft synthetic bone materials containing bone morphogenetic protein (BMP) was inserted, and the optimal position was con rmed using uoroscopy. Finally, the abdominal muscle planes were closed sequentially without placement of a drainage catheter.

Robot-assisted percutaneous BPS/BTS procedures
The patient was placed in the prone position after the OLIF procedure. Robot-assisted procedures were performed using the TiRobot system (TINAVI Medical Technologies, Beijing, China). The patient tracker was percutaneously anchored at the spinal process. An alternative method was used for the BTS procedure in some patients. Preparation and draping were done at the beginning, and the patient was kept in the lateral decubitus position after OLIF cage insertion. The patient tracker was percutaneously anchored at the iliac crest.
Fluoroscopic images produced by the C-arm (ARCADIS Orbic 3D C-arm, Siemens Medical Solutions, Erlangen, Germany) were transferred to the robotic workstation. After the preoperative planning of pedicle screw instrumentation using the TiRobot workstation, the robotic arm was instructed to move to the chosen trajectory. When the accuracy of guidance was less than 0.5 mm and became stable, a tiny incision over the trajectory was made. The guiding cannula was inserted into the incision at the distal end of the robotic arm until the bony surface was touched. Then, the guiding pin held by a drill bit was placed along the guiding cannula to the optimal depth. A uoroscopic re-scan by the C-arm was performed, and the position of the guiding pin was evaluated. Once optimal positioning of the guiding pin was con rmed, the pilot hole was tapped, followed by insertion of cannulated pedicle screws or cannulated Herbert screws. Afterward, the positions of screws were evaluated and con rmed using uoroscopy. The incisions were closed without drainage.
On the rst day after the operation, patients were asked to perform out-of-bed activities with the help of a lumbar brace. Also, all patients were required to wear the lumbar brace for 3 months postoperatively.

Radiologic analysis
Radiologic outcomes were assessed by collecting anteroposterior and lateral radiographs at 3 days and 6 months postoperatively. The disc height was expressed according to the disc height index (DHI), based on the method of Inoue from standing lateral radiographs [23]. The DHI was calculated as the mean of the anterior, middle, and posterior disc heights, divided by the sagittal diameter of the overlying vertebral body at the midvertebral level. The segmental lordotic (SL) angle was de ned as the angle between the superior endplate of the upper vertebra and the inferior endplate of the lower vertebra at the corresponding level. The global lumbar lordotic (LL) angle was de ned as the angle between the upper endplate of the L1 vertebra and the upper endplate of the S1 vertebra ( Fig. 1).

Clinical assessment
The operating time, estimated blood loss, and height, length, and width of the OLIF cage were recorded.
Before surgery and at the 6-month follow-up, self-assessment of disability and pain was provided by patients using the Oswestry Disability Index (ODI), Japanese Orthopaedic Association (JOA) scoring system, and Visual Analog Scale (VAS) for back/leg symptoms. Any intraoperative and postoperative complications were also recorded.

Statistical analysis
The χ 2 -test and Fisher exact test were used to evaluate between-group differences in gender, diagnosis, operative level, complication, and the height, length, and width of the OLIF cage. One-way analysis of variance (ANOVA) was used to identify differences among the groups for age, BMI, BMD, DHI, SL, LL, operative time, estimated blood loss, VAS score, JOA score, and ODI score. Differences in variables across the three time points of the study were identi ed by paired t-tests for DHI, SL, LL and the VAS, JOA, and ODI scores. The independent-samples t-test was used to evaluate the differences in operative time and estimated blood loss in subgroups of the BTS group. Statistical analysis was performed using SPSS (version 21.0; SPSS, Chicago, IL, USA). The level of statistical signi cance was set as P < 0.05.

Radiographic outcomes
In all three groups, the DHI was improved immediately after surgery and then deteriorated signi cantly by the 6-month follow-up (Table 2). Other than a signi cant decrease in the SL angle from 3 days to 6 months postoperatively in the SA group (P = 0.047), no changes in the SL and LL angles were observed in any group over the course of the study (Table 2). Values are mean ± standard deviation (SD).
The changes in DHI, SL angle, and LL angle (ΔDHI, ΔSL, and ΔLL) were compared among the three groups. No signi cant differences in the ΔSL and ΔLL values were found, and no signi cant difference in the ΔDHI value from before surgery to 3 days after surgery was observed. However, from 3 days to 6 months after surgery, the values for the degree of graft subsidence (ΔDHI) in the SA and BTS groups were signi cantly greater than that in the BPS group (Table 2). Notably, the ΔDHI values during this time for the SA and BTS groups did not differ signi cantly.

Clinical outcomes
The operative time was signi cantly less in the SA group than in the other two groups (P = 0.000), and the estimated blood loss volume was signi cantly higher in the BPS group than in the SA and BTS groups (P = 0.004; Table 3). No signi cant differences in the height, length, and width of the OLIF cage were observed among the three groups. Values are mean ± standard deviation (SD) or as otherwise indicated.
In the BTS group, surgery was performed in the prone position for 9 patients and in the lateral decubitus position for 11 patients. The operative time did not differ between these subgroups, nor did the estimated blood loss (Table 4). Values are mean ± standard deviation (SD).
All patients experienced signi cant improvements in the VAS score for back pain, the VAS score for leg pain, the JOA score, and the ODI score by 6 months after surgery compared with those before surgery.
The changes in these scores (ΔVAS back pain, ΔVAS leg pain, ΔJOA and ΔODI) from before surgery to 6 months after surgery were compared among the three groups. The improvement in the VAS score for back pain was signi cantly less in the SA group than in BPS and BTS groups, as were the JOA score and ODI score. The improvements in the VAS score for back pain, JOA score, and ODI score were similar in the BPS and BTS groups. The improvement in the VAS score for leg pain did not differ among the three groups (Table 5). Complications included psoas weakness, left thigh numbness, paralytic ileus, sympathetic dysfunction, and super cial wound infection (Table 6). No patients required reoperation during the 6-month follow-up.
The incidence of complications did not differ signi cantly among the three groups.

Discussion
Subsidence is a common complication following various lumbar interbody fusion surgeries [24] and a matter of great concern for OLIF in particular, as the surgery relies primarily on indirect decompression achieved by restoration of disc height and sagittal alignment [5]. The reported incidence of subsidence with OLIF has ranged from 4.4-21.6% [25]. The development of subsidence is considered to be a multifactorial process, and studies have investigated several potential risk factors, including low BMD, disc space overdistraction, insu cient cage width, construct length, endplate violation, use of osteobiologics, and supplemental xation [9,[26][27][28][29][30][31][32]. Inconsistent results likely resulting, at least in part, from a lack of a uniformity in the methodology used to measure and report subsidence have made it di cult to understand this radiological phenomenon [7,10,33,34]. In the present study, subsidence was de ned as any compromise of either endplate due to the cage and was recorded based on speci c numeric measurements of cage settling, which can easily be discerned and measured radiographically. Based on the conclusion of previous studies that subsidence is an early postoperative event that does not progress signi cantly beyond 3 months postoperatively [10,35], we used postoperative 6 months as the last follow-up date. Over-distraction is an important factor for subsidence; however, to the best of our knowledge, there has been no standard for this issue until now. In the present study, we compared the changes in disc height before and immediately after the operation, which showed that there was no statistical difference, so as to make cage subsidence comparable between the three groups.
Among the patients in the present study, the OLIF procedure was an effective method for restoring disc height, regardless of the xation pattern used, which was consistent with the results of previous studies[36-38]. In all three groups, loss of disc height had occurred by 6 months after surgery. Notably, in one previous study, subsidence was reported as an expected occurrence rather than a complication [35].
However, the subsidence in the BPS group was signi cantly less than that in either of the other two groups, indicating that supplementary xation and a speci c pattern can help to avoid a loss in disc height. In recent decades, the use of supplemental xation in OLIF has been the subject of ongoing debate. Previous studies showed that a su ciently distracted intervertebral space following discectomy can be stabilized for multidirectional movement by tension forces of the residual annulus and ligaments [14,39,40]. Also, the authors supporting stand-alone OLIF have stated that it avoids the morbidity associated with supplemental screw xation and is less expensive. Conversely, Fogel et al.
reported that a stand-alone cage decreased the ROM by only about 23% compared with the normal spine and signi cantly increased the anterior-posterior (interbody) displacement in a L4-L5 spondylolisthesis cadaver model [41]. Liu et al. used a nite element model to evaluate the biomechanics of three-level lateral interbody fusion with and without supplementary instrumentation and found that stand-alone lateral interbody fusion could not provide adequate ROM restriction, whereas lateral cages with bilateral pedicle screw and rod xation provided favorable biomechanical stability [13]. Additionally, stand-alone OLIF generated signi cantly higher endplate stress than did OLIF with supplemental instrumentation, which may increase the risk of cage subsidence. Nevertheless, their results can only be applied to multilevel fusion. Whether supplemental instrumentation should be applied to prevent subsidence in single-level OLIF remains controversial. Malham et al. proposed that patients with one-or two-level disease, normal bone density, and no obvious instability were candidates for stand-alone OLIF [7]. In contrast, in the present study, the subsidence in the SA group was signi cantly greater than that in the BPS group at the 6-month follow-up, indicating that stand-alone OLIF may not be su cient for maintaining disc height even in single-level disease, which is consistent with the result of Choi and Sung [35].
However, studies have reported greater blood loss and longer operative time for OLIF with supplementary BPS compared with stand-alone OLIF due to the requirement of additional posterior surgery [39,43]. These results were consistent with our ndings. As a traditional method for posterior xation, BTS has several advantages over BPS, such as fewer incisions, being minimally invasive, having less in uence on the adjacent segments, and less cost, as only two screws are used at each spinal segment [44]. In the present study, the operative time tended to be shorter when the operation was performed with patients in the lateral decubitus position versus the prone position, but the difference was not signi cant, probably due to the small sample sizes of the subgroups. Previous studies have demonstrated that BTS provides equivalent stiffness against segmental movements compared with BPS [20,44]. Chin et al. even showed that BTS offers greater stability than pedicle screws in the motion of exion [21]. However, in the current study, the subsidence rates of the BTS and SA groups were similar and signi cantly higher than that of the BPS group, indicating that although additional BTS can make the overall construct stiffer, this construct is unlikely to achieve more resistance to axial compression than stand-alone OLIF. The screw placement was done under the guidance of the TiRobot system. Previous studies at our center have demonstrated a high accuracy of BPS and BTS with assistance by this robotic system [45,46], and thus, accuracy was not a concern in this study.
It is well established that improper sagittal alignment of the spine results in ine cient energy use and maximizes muscle tension due to exhaustive bracing and spine instability, contributing to adjacent-level disease [47,48]. In a review of 12 retrospective and 2 prospective studies including 1266 levels in 476 patients, Costanzo et al. showed that lateral interbody fusion is effective when the lumbar lordosis and sagittal balance correction goals were less than 10° and 5 cm, respectively, but the results for sagittal balance restoration using the technique were inconsistent [49]. Chen et al. compared the alteration and maintenance of SL and LL in lateral interbody fusion with and without supplementary xation [39]. They found that lateral interbody fusion alone and with supplementary xation could increase the postoperative SL and LL angles, but BPS could maintain the LL angle but not the SL angle for 2 years after surgery. In the present study, OLIF with or without supplemental xation improved the SL angle to 0.3°~2.3° on average, with no signi cant difference among the groups. At the 6-month follow-up though, the reduction in the SL angle was signi cant only in the SA group. Overall, research to date has produced inconsistent ndings regarding local and global sagittal balance restoration after OLIF, and further studies are needed to address this issue.
The questionnaire-based clinical outcomes in our cohort showed signi cant improvements in symptoms and function at 6 months after surgery compared with preoperative values, regardless of the xation pattern, which was consistent with previous studies[6, 50,51]. The change in the VAS score for leg pain did not differ among the three groups, indicating that indirect decompression can be achieved and maintained by the stand-alone technique. However, the improvements in the VAS score for back pain, the JOA score, and the ODI score were worse in the SA group compared with the other two groups. Back pain and related disability may have resulted due to the instability caused by the lack of posterior column support. In a nite element analysis, stand-alone OLIF generates signi cantly higher endplate stress compared with that of supplemental instrumentation, which results in an increase in the risk of cage migration and development of clinical symptoms [13]. Our results con rm this limitation of the SA technique. Interestingly, although BTS cannot offer more resistance to subsidence compared with standalone OLIF, the symptomatic and functional improvement with BTS were better. These ndings were consistent with other studies that investigated the clinical outcomes with BTS as supplementary xation for lumbar interbody fusion and observed promising results [22,52]. This can be explained by the additional stability provided by the transfacet screws. In a study of the effect of supplemental BTS on the stability of stand-alone ALIF under physiologic compressive preloads, Phillips et al. found that the standalone cage is likely to be less stable and supplemental BTS can enhance the stability of the motion segment, particularly during conditions of low compressive preloads [53]. Moreover, the results of our study also indicated no obvious correlation between subsidence and clinical outcomes. This is consistent with the results of previous studies comparing subsidence to nal clinical outcomes in which no clear relationship was observed [12,54,55]. Marchi et al. graded subsidence and proposed that high-grade subsidence (> 50% into the vertebral endplates) could lead to persistent back pain or radiculopathy and a need for revision surgery [10]. Similar results were reported by Tempel et al., who reported a strong correlation between subsidence grade and the risk of revision surgery [55]. Because the sample size of the current study was small, we did not grade the subsidence. A further prospective, randomized, and controlled study is needed to clarify the correlation between subsidence grade and clinical outcomes.
Because of the novelty of OLIF, there is a paucity of data regarding its complications. In a multicenter study including 155 patients, Abe et al. reported a complication incidence of 29.5% [11], and a systematic review reported 1.5% intraoperative and 9.9% postoperative complications among 1453 patients treated with OLIF [5]. A recent analysis of OLIF-related complications identi ed only 20 research studies of su cient quality and relevance, and reported overall combined rates of 5.7% for postoperative psoas weakness, 8.7% for thigh numbness, and 3.3% for paralytic ileus [25]. These rates were similar to those in the present study (psoas weakness 3.3%, thigh numbness 13.3%, and paralytic ileus 1.7%), in which the overall complication rate was 23.3%. These complication rates did not differ signi cantly among the groups, but these ndings should be interpreted with caution due to the relatively small sample sizes.
A major limitation of this study was its retrospective design. Additionally, the study was a single-center study with a small sample size. Although we matched the patients according to main parameters, complex baseline differences may have been confounding factors for clinical and radiographic outcomes among the groups. Differences between surgeons could be a potential confounding factor in this study, because the surgeries were performed by three senior spine surgeons. Although all three surgeons had received standardized training for the OLIF procedure and implemented standardized surgical maneuvers, and there was no statistical difference between the numbers of cases in the different treatment groups for the three surgeons, this confounding factor cannot be completely ruled out. Another limitation was the lack of dynamic radiographic evaluation at the 6-month follow-up. Thus, no conclusions regarding instability or union can be made. Ideally, fusion status should be assessed by CT scan at one year after surgery. Future studies should focus on the stability status after OLIF rather than subsidence only.

Conclusions
The stand-alone OLIF procedure was associated with greater subsidence and poorer clinical outcomes compared with OLIF plus supplementary BPS. The addition of BTS to OLIF could not decrease the degree of subsidence, but provided comparable clinical outcomes to OLIF with BPS. Therefore, BTS could be considered as a substitute for BPS as supplemental xation with OLIF.

Consent for publication
Informed consent was obtained from all individual participants involved in the study.

Availability of data and materials
The datasets generated and/or analyzed during the current study are not publicly available due to the data is con dential patients' data but are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.  C. The LL angle was de ned as the Cobb angle formed between the upper end plate of L1 vertebra and the upper end plate of S1 vertebra.