INTRODUCTION
Lumbar interbody fusion is a well-established surgical technique used to restore spinal stability and decompress neural structures in patients with degenerative or isthmic spondylolisthesis [1]. With the growing emphasis on minimally invasive strategies, unilateral biportal endoscopic transforaminal lumbar interbody fusion (BE-TLIF) has gained increasing attention due to its ability to minimize muscle trauma, reduce intraoperative blood loss, and shorten postoperative recovery time compared to conventional open or tubular minimally invasive approaches [1-3]. This technique enables effective decompression and interbody fusion through 2 small portals under continuous endoscopic visualization [2,4].
In BE-TLIF, the choice and placement of the interbody cage are critical determinants of surgical success. Interbody cages serve to restore disc height (DH), maintain segmental lordosis, and promote fusion, all of which are influenced by cage morphology and positioning [5-7]. Among the available options, banana-shaped and straight cages are most commonly utilized in BE-TLIF [8,9]. While their respective biomechanical properties and clinical applications have been investigated in conventional and minimally invasive TLIF [8-10], their roles in BE-TLIF remain underexplored.
To date, no comparative studies have specifically addressed how these 2 cage types perform within the context of BE-TLIF. As such, the clinical and radiological implications of cage morphology in BE-TLIF are not yet well understood. Understanding these differences is essential to optimize surgical planning and outcomes, particularly in settings where cage selection may be constrained by availability or cost [11-14].
This study aims to compare the clinical and radiological outcomes of BE-TLIF using banana-shaped versus straight interbody cages in patients with single-level spondylolisthesis.
MATERIALS AND METHODS
1. Study Design and Ethics
This study was designed as a retrospective cohort analysis conducted at the University Medical Center, Ho Chi Minh City. The investigation included patients who underwent single-level unilateral BE-TLIF between January 2023 and May 2024. The study protocol was approved by the Institutional Review Board of the hospital (approval number: 70/GCN-HĐĐĐ), and all procedures were performed in accordance with the Declaration of Helsinki. Given the retrospective nature of the study, informed consent for participation was waived; however, written informed consent for surgical treatment and the use of anonymized clinical data was obtained from all patients prior to surgery.
2. Patient Selection
Patients were eligible for inclusion if they met the following criteria: (1) age ≥18 years; (2) diagnosis of single-level degenerative or isthmic spondylolisthesis of Meyerding grade I or II; (3) persistent low back pain and/or radiculopathy refractory to at least 6 weeks of conservative treatment; and (4) minimum clinical and radiological follow-up of 12 months.
Exclusion criteria included: (1) previous lumbar spine surgery; (2) presence of spinal infection, trauma, tumor, or congenital deformity; and (3) requirement for multilevel fusion.
A total of 34 patients meeting these criteria were identified and divided into 2 groups based on the type of interbody cage used during BE-TLIF: banana-shaped cage (n=17) or straight cage (n=17). The selection of cage type was determined by implant availability at the time of surgery. Specifically, due to procurement and inventory constraints at the study institution, only one type of cage—either banana-shaped or straight—was available during certain time periods. As a result, patients were not randomized and were allocated to each group based on the time window in which they underwent surgery.
This pragmatic, resource-driven allocation reflects real-world practice in settings with limited access to implant options and served as the rationale for conducting this study, which aimed to assess whether comparable clinical and radiological outcomes could be achieved irrespective of cage morphology under such constraints.
3. Surgical Technique
All procedures were performed under general anesthesia with the patient in the prone position. Two endoscopic portals were created under fluoroscopic guidance: a working portal positioned at the intersection of the target disc level and the lateral border of the pedicle, and a viewing portal placed 2–3 cm cranially or caudally, depending on the surgeon’s dominant hand. The working portal incision was approximately 10 mm, and the viewing portal 5 mm in length. The triangulation technique was applied to establish the endoscopic field, with saline irrigation facilitating visualization (Figure 1).
Following soft tissue dissection, ipsilateral decompression was performed using a high-speed burr, Kerrison punches, and osteotomes. Contralateral decompression was achieved via an “over-the-top” approach. The inferior articular process and part of the superior articular process were removed to access the intervertebral disc space. The ligamentum flavum was excised to expose the dura and nerve roots. Disc material and cartilaginous endplates were removed under endoscopic visualization until bleeding subchondral bone was exposed, confirming sufficient preparation for fusion (Figure 2).
Autologous bone graft, obtained during decompression, was inserted into the disc space using a bone funnel. Interbody cage placement was guided by endoscopy and fluoroscopy. For banana-shaped cages (TSPACE PEEK, B. Brown, Germany), the cage was introduced laterally and rotated transversely to achieve anterior placement (Figure 3). For straight cages (PLIF PEEK, L&K Biomed, Korea), the device was inserted diagonally across the disc space toward the contralateral anterior annulus to optimize sagittal alignment and endplate coverage (Figure 4). Representative cage designs are illustrated in Figure 5.
Following cage placement, bilateral percutaneous pedicle screws and connecting rods were inserted under fluoroscopic guidance. Final inspection confirmed decompression and cage positioning, after which a drainage catheter was placed and incisions were closed in layers.
4. Clinical Evaluation
Clinical outcomes were assessed using the visual analogue scale (VAS) for back and leg pain and the Oswestry Disability Index (ODI). These assessments were performed preoperatively and at 1, 3, 6, and 12 months postoperatively. All evaluations were conducted by a single trained clinical research assistant who was blinded to the cage type used.
5. Radiological Evaluation
Radiographic outcomes were assessed using standard lateral standing radiographs of the lumbar spine obtained preoperatively and at postoperative intervals (2 days, 1 month, 3 months, 6 months, and 12 months). Measured parameters included:
• DH: the vertical distance between the midpoints of the superior and inferior endplates at the index level.
• Segmental lordosis angle (SLA): the Cobb angle between the superior endplate of the cranial vertebra and the inferior endplate of the caudal vertebra at the operative segment.
• Lumbar lordosis angle (LLA): the Cobb angle from the superior endplate of L1 to the superior endplate of S1.
Cage position was assessed on standard lateral standing radiographs and computed tomography (CT) scans using 2 methods:
• Sagittal plane analysis (central point ratio, CPR): CPR was defined as the ratio between the distance from the posterior edge of the superior endplate to the center of the cage, and the total length of the endplate in the sagittal plane (Figure 6). A CPR value greater than 0.5 indicated anterior cage placement.
• Axial plane analysis using a 3 × 3 grid system: On axial CT images, the vertebral body was divided into 9 equal zones to classify the cage position (Figure 7). This method allowed qualitative and semiquantitative analysis of cage placement relative to anatomical landmarks.
Cage subsidence was defined as vertical migration of more than 2 mm into adjacent vertebral endplates, assessed using serial radiographs or CT scans. Fusion status was evaluated on CT scans obtained 12 months postoperatively, and graded according to the modified Bridwell fusion criteria. Grades I and II were considered indicative of successful fusion.
All radiographic measurements were independently performed by a fellowship-trained orthopedic spine surgeon blinded to the cage type. Measurements were repeated twice using PACS (picture archiving and communication system)-integrated digital tools, and the average value was used for analysis.
6. Statistical Analysis
All statistical analyses were performed using IBM SPSS Statistics ver. 21.0 (IBM Co., USA). Continuous variables were tested for normality using the Shapiro-Wilk test. Depending on data distribution, comparisons between the 2 groups were made using the independent samples t-test or Mann-Whitney U-test for continuous variables, and the chi-square test or Fisher exact test for categorical variables.
Data were presented as mean±standard deviation for continuous variables, and as frequencies and percentages for categorical variables. A p-value <0.05 was considered statistically significant for all comparisons.
RESULTS
1. Patient Demographics and Perioperative Characteristics
A total of 34 patients were included, with 17 patients in each group (banana-shaped cage vs. straight cage). Baseline characteristics, including age, sex, body mass index, and bone mineral density, were comparable between groups (p>0.05). The majority of procedures were performed at the L4–5 level in both groups (Table 1).
There were no statistically significant differences in operative time, estimated blood loss, or hospital stay between the 2 groups (Table 2). The average cage height used was also similar.
2. Radiographic Outcomes
Both groups demonstrated an increase in DH and SLA following surgery. Although the banana-shaped cage group showed slightly greater improvements in DH and SLA at all time points, these differences were not statistically significant (p>0.05) (Table 3).
Changes in LLA were variable, with no consistent pattern of superiority between groups. At 12 months, the mean DH gain was 2.18±1.85 mm in the banana-shaped group and 1.71±1.93 mm in the straight cage group.
3. Subsidence
Cage subsidence was observed in 4 patients (23.5%) in each group. The mean subsidence depth was slightly lower in the banana-shaped cage group (4.50±1.00 mm vs. 5.50±1.73 mm), though this difference was not statistically significant (p=0.317). All cases of subsidence occurred within the first 2 months following surgery (Table 4).
4. Fusion Status
At the 12-month follow-up, solid fusion was achieved in 100% of patients in the banana-shaped cage group and in 94.1% of patients in the straight cage group. One patient in the straight cage group was graded as Bridwell grade II, and one as grade III. No grade IV fusion failures were observed. The difference in fusion grading between groups was not statistically significant (p=0.485) (Table 5).
5. Cage Position
Analysis of cage positioning revealed significant differences in anterior and posterior placement between groups. Straight cages were more frequently positioned in the anterior third of the vertebral body (76.5% vs. 41.2%, p=0.037), while posterior positioning occurred only in the straight cage group (41.2% vs. 0%, p=0.007). CPR was not significantly different between groups (0.52±0.09 vs. 0.55±0.11, p=0.480) (Table 6, Figure 6).
Axial positioning assessed by the 3 × 3 grid method demonstrated broader anterior coverage in the straight cage group and predominant central–lateral placement in the banana-shaped cage group (Figure 7).
6. Clinical Outcomes
Both groups showed significant improvement in back and leg pain as well as functional scores over time. At 12 months, the mean VAS score for back pain was 0.71±0.59 in the banana-shaped cage group and 0.88±0.86 in the straight cage group (p>0.05). Leg pain VAS scores were similarly improved with no between-group differences (Table 7).
ODI scores decreased significantly in both groups from baseline to final follow-up. The mean ODI at 12 months was 17.76%±5.75% in the banana-shaped group and 15.59%±9.75% in the straight cage group (p=0.297) (Table 8).
DISCUSSION
The BE-TLIF technique has emerged as a promising minimally invasive surgical option for managing lumbar spondylolisthesis. Compared with conventional open or tubular minimally invasive TLIF, BE-TLIF offers several theoretical advantages, including reduced muscle disruption, enhanced visualization, and faster postoperative recovery [1,2,15]. As this technique gains popularity, implant selection—particularly the morphology and positioning of the interbody cage—plays an increasingly critical role in optimizing surgical outcomes.
Our study evaluated and compared clinical and radiographic outcomes between banana-shaped and straight cages in single-level BE-TLIF (Table 8). Both cage types resulted in significant improvements in back and leg pain, as well as functional status over 12 months, with no statistically significant differences observed between groups in VAS or ODI scores. These findings are consistent with previous literature on minimally invasive and open TLIF procedures, suggesting that clinical success may be more dependent on surgical technique than cage design [3,5].
Radiographically, banana-shaped cages demonstrated a slight advantage in restoring DH and SLA, likely due to their anterior placement tendency and inherent curvature. However, these differences were not statistically significant. The technical challenge of rotating banana-shaped cages into an ideal anterior position within the confined BE-TLIF corridor may contribute to this result. In contrast, straight cages were more consistently positioned anteriorly, as reflected by higher CPR values. This supports the theoretical advantage of straight cages in endoscopic procedures, where controlled trajectory and ease of manipulation are essential [3,16].
Cage positioning has been shown to influence biomechanical load distribution, subsidence, and fusion success [4,12,17,18]. In this study, axial CT analysis revealed differing placement patterns: banana-shaped cages were more commonly located in the middle and lateral regions, while straight cages exhibited broader anterior coverage. Despite these positional variations, subsidence rates and fusion outcomes were comparable between groups, with all subsidence cases occurring within the first 2 postoperative months. This may reflect the importance of endplate preparation and load-bearing placement near the apophyseal ring rather than cage shape alone [11,19,20].
Our findings are further supported by previous comparative studies in TLIF. Kim et al. [3] demonstrated that curvilinear cages improved segmental lordosis, but emphasized that surgical technique and cage positioning were more critical than cage shape. Choi et al. [5] also reported similar clinical outcomes between cage types in MIS-TLIF, with increased subsidence noted in banana-shaped cages—potentially due to central placement. While these results originated from conventional or MIS-TLIF, their relevance extends to BE-TLIF when considering cage trajectory and load-bearing characteristics.
Importantly, BE-TLIF introduces distinct technical considerations compared to traditional TLIF, including a bimanual triangulation setup, constrained working corridors, and reliance on continuous endoscopic and fluoroscopic guidance. These nuances may affect cage maneuverability and final positioning. In this context, the observed equivalence in clinical outcomes between banana-shaped and straight cages highlights the adaptability of BE-TLIF and reinforces that precise placement and proper endplate preparation are likely more influential than implant morphology.
Fusion outcomes were excellent in both groups, with 100% fusion in the banana-shaped cage group and 94.1% in the straight cage group. The use of standardized surgical techniques and autologous bone grafting likely contributed to these favorable results. Additionally, the preservation of subchondral bone under endoscopic visualization may have minimized the risk of cage subsidence and facilitated successful arthrodesis.
Taken together, our findings support the notion that both banana-shaped and straight cages are effective options in BE-TLIF when inserted with appropriate technique. The choice of cage can therefore be guided by anatomical considerations, surgeon preference, and implant availability without compromising short-term outcomes. Nonetheless, future prospective, multicenter studies with longer follow-up and stratification by spinal level and bone quality are warranted to validate these results and explore their long-term implications.
This study has several limitations. First, the retrospective design inherently introduces the risk of selection and information bias, despite strict inclusion criteria and blinded outcome assessment. Second, the allocation of patients to cage groups was not randomized but based on implant availability during specific time periods. This pragmatic approach reflects real-world practice in resource-limited settings; however, it introduces a potential selection bias, as unmeasured confounding variables may have influenced surgical timing and outcomes.
Third, the sample size was relatively small (n=17 per group), which limits the statistical power to detect subtle differences and may affect the generalizability of the findings. Fourth, all surgeries were performed by a single experienced surgeon at a single center, which may introduce operator-dependent bias and limit external validity. Fifth, we did not stratify outcomes by spinal level (e.g., L3–4, L4–5, L5–S1), which may influence cage performance due to anatomical variations.
Finally, the 12-month follow-up period may not fully capture long-term outcomes such as adjacent segment degeneration, implant longevity, or delayed complications. To address this, extended follow-up of the current patient cohort is ongoing, and a prospective study with larger sample size, stratified analysis, and standardized cage allocation is currently in development.
CONCLUSION
Our findings suggest that both banana-shaped and straight interbody cages are viable options in BE-TLIF for single-level lumbar spondylolisthesis, with comparable short-term clinical and radiological outcomes. Cage morphology did not appear to influence fusion success or subsidence, underscoring the importance of precise cage placement and meticulous endplate preparation. These results offer practical guidance for surgical planning, especially in resource-limited settings. To validate and expand upon these findings, future prospective multicenter studies with larger cohorts, stratification by spinal level and bone quality, and long-term follow-up are warranted.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
