Comparative Analysis Between Single and Double 3-Dimensional Printed Titanium Cages: 1-Year Outcomes After Unilateral Biportal Endoscopic Transforaminal Lumbar Interbody Fusion

Ji Yeon Kim; Hyun Jin Hong; Su Yong Choi; Dong Chan Lee; Hyeun Sung Kim; Dong Hwa Heo

doi:10.21182/jmisst.2024.01970

J Minim Invasive Spine Surg Tech > Volume 10(Suppl 2); 2025 > Article

Kim, Hong, Choi, Lee, Kim, and Heo: Comparative Analysis Between Single and Double 3-Dimensional Printed Titanium Cages: 1-Year Outcomes After Unilateral Biportal Endoscopic Transforaminal Lumbar Interbody Fusion

Original Article

Journal of Minimally Invasive Spine Surgery and Technique 2025; 10(Suppl 2): S225-S234.

Published online: July 31, 2025

DOI: https://doi.org/10.21182/jmisst.2024.01970

Comparative Analysis Between Single and Double 3-Dimensional Printed Titanium Cages: 1-Year Outcomes After Unilateral Biportal Endoscopic Transforaminal Lumbar Interbody Fusion

Ji Yeon Kim¹

, Hyun Jin Hong²

, Su Yong Choi¹

, Dong Chan Lee²

, Hyeun Sung Kim³

, Dong Hwa Heo³

¹Department of Neurosurgery, Spine Center, Seran General Hospital, Seoul, Korea

²Department of Neurosurgery, Spine Center, Wiltse Memorial Hospital, Anyang, Korea

³Department of Neurosurgery, Spine Center, Harrison Spinatus Hospital, Seoul, Korea

Corresponding Author: Su Yong Choi Department of Neurosurgery, Spine center, Seran General Hospital, 256, Tongil-ro, Jongno-gu, Seoul 03030, Korea Email: tigrissy@hanmail.net

Received December 8, 2024 Revised January 31, 2025 Accepted February 14, 2025

This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (https://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Objective

This study aimed to illustrate the techniques of unilateral biportal endoscopic transforaminal lumbar interbody fusion (UBE-TLIF) using double cages and compare surgical outcomes with those using a single cage.

Methods

We retrospectively analyzed 62 patients who underwent single-level UBE-TLIF using 3-dimensional (3D)-printed titanium cages (29 with a single cage and 33 with double cages). Radiological parameters, including Bridwell fusion and subsidence grading, were assessed via x-ray and computed tomography at 6 months and 1 year. Clinical outcomes were measured using the visual analogue scale (VAS) for lower back and leg pain, as well as the Oswestry Disability Index (ODI).

Results

At the 6-month follow-up, the overall fusion rate (grades I + II) was significantly higher in the double-cage group (86.2% vs. 100%, p=0.04); however, no significant difference was noted at the 1-year follow-up (89.6% vs. 100%). The double-cage group showed lower cage subsidence rates at both follow-ups (3% vs. 31%, p=0.01; 12.1% vs. 38%, p=0.03). The double-cage group exhibited significantly greater improvements in leg pain (VAS: 7.6 to 1.7 vs. 7.2 to 1.4, p=0.03) and ODI (28.6 to 10.8 vs. 28.3 to 9.4, p=0.01) at 1 year.

Conclusion

UBE-TLIF with 3D-printed titanium cages achieved a favorable fusion rate at the 1-year follow-up, regardless of whether single or double cages were used. However, double cages accelerated fusion at the 6-month follow-up and significantly reduced cage subsidence at both the 6-month and 1-year follow-ups. These benefits contributed to improved VAS scores for leg pain and ODI at the final follow-up.

Key Words: Spinal fusion, Lumar vertebrae, Endoscopy, Titanium

INTRODUCTION

Spinal fusion is a common procedure used to alleviate symptoms of lumbar degenerative diseases, such as spondylolisthesis and foraminal stenosis, through various surgical techniques. Minimally invasive transforaminal lumbar interbody fusion (MIS-TLIF) has been developed to reduce muscle damage and improve patient outcomes, resulting in shorter hospital stays and faster recovery [1,2]. Recently, unilateral biportal endoscopic transforaminal lumbar interbody fusion (UBE-TLIF) was introduced as a minimally invasive technique utilizing spinal endoscopy to further minimize tissue trauma [3-6]. While UBE-TLIF with a single cage has shown favorable clinical and radiological outcomes [7-11], the narrow contact area between the cage and the bony endplate is a significant drawback, increasing the risk of cage subsidence [12]. This limitation can lead to complications such as severe subsidence and pseudarthrosis [8].

To address these concerns, the use of larger or multiple cages has been explored to enhance stability and facilitate solid interbody fusion [13], as demonstrated in novel UBE-TLIF techniques by Heo et al. [7] and Eum et al. [14,15]. The double-cage approach may mitigate the disadvantages of single-cage implants by increasing the contact area, thereby reducing the risk of subsidence and enhancing overall support. Pao [16] reported that stable interbody implants with larger footprints can expedite the fusion process.

However, comparative studies between single and double cages are necessary to validate the advantages of double-cage implants in UBE-TLIF. Thus, we conducted this comparative study to analyze the outcomes of single versus double-cage insertion in patients undergoing UBE-TLIF, focusing on radiological outcomes related to fusion status and cage subsidence. Additionally, we provide a detailed outline of the surgical procedure for UBE-TLIF utilizing double cages to enhance feasibility for readers.

MATERIALS AND METHODS

1. Study Patients

This study received approval from the ethics committee of Wiltse Memorial Hospital (NR-IRB 2023-W05). From January 2021 to December 2021, we retrospectively analyzed 62 patients who underwent single-level UBE-TLIF using 3D-printed titanium cages (single cage, n=29; double cages, n=33). Two experienced UBE spine surgeons performed the surgeries at a single center, with one surgeon using a single cage and another using double cages according to consecutive patients. We utilized straight bullet-type 3D-printed titanium cages (GS Medical, Korea), fabricated through selective laser melting 3D printing (DMP 350, 3D Systems, USA), featuring pore sizes of 500–1100 µm and a porosity of 70%–80%. The single-cage group employed long TLIF (28-mm length, 11-mm width) cages, while the double-cage group used cases for posterior lumbar interbody fusion (24-mm length, 11-mm width) (Figure 1A and B). We utilize the autologous bone grafts harvested during facetectomy and laminectomy. The cage hole is filled with autologous bone graft and demineralized bone matrix (DBM).

Inclusion criteria were: (1) persistent low back and radiating leg pain despite 3 months of conservative treatment, (2) single-level endoscopic TLIF performed for spinal stenosis, spondylolisthesis, segmental instability, or recurrent disc herniation, and (3) a minimum of 12 months of follow-up. We excluded cases with severe low bone mineral density (T score < -3.0), those requiring multiple level fusion, or patients who underwent concomitant decompression at adjacent levels.

2. Surgical Procedures

1) Unilateral biportal endoscopic facetectomy, discectomy and endplate preparation for TLIF

Biportal endoscopic systems were used, including a 4-mm diameter, 0° endoscope, a toolkit set, a customized scope retractor [17], and a working sheath [18] (MD&Company, Korea) (Figure 1C). Continuous irrigation with normal saline was used, and a radiofrequency wand (ArthroCare, USA) was employed to ablate soft tissue. Two ipsilateral skin incisions were made along the lateral border of the pedicles which also served as the insertion sites for pedicle screws (Figure 2A). After serial dilation. The endoscope and surgical instruments were advanced through the trocar and working sheath to reach the initial docking area (Figure 2A). A total facetectomy of the ipsilateral inferior and superior articular processes (SAPs), along with a laminotomy, was performed using endoscopic drills, punches, and osteotomes. The harvested bone was saved for use as interbody fusion material. For neural decompression, the ipsilateral ligamentum flavum and foraminal ligament were completely removed. Contralateral decompression was performed in cases involving bilateral symptomatic spinal or foraminal stenosis. Annulotomy was completed using an endoscopic knife, drill, and punches. A scope retractor, mounted on the trocar, was used for neural retraction during endplate preparation and cage insertion (Figure 1C). Curved dissectors or curettes were used to separate the cartilaginous endplate from the bony endplate, which was subsequently removed.

The surgical steps, from skin incision to endplate preparation, are the same for both groups; however, distinct procedures are required to accurately position the cage during single cage (Figure 2C) and double-cage (Figure 2D) insertions.

2) Unilateral biportal endoscopic single-cage insertion (Supplementary Video Clip 1)

The size of the interbody cage was determined using a trial cage. Harvested bone chips mixed with residual DBM were inserted into the intervertebral disc space using a customized funnel, followed by the oblique insertion of the cage under endoscopic guidance (Figure 2B). Endoscopic visualization allowed for immediate correction of improper cage positioning, and the cage was repositioned using cage impactors as necessary (Figure 2C).

3) Unilateral biportal endoscopic double-cage insertion (Supplementary Video Clip 2)

The SAP in the foraminal space should be completely removed until the pedicle is exposed, in order to provide sufficient space for double-cage insertion. Bone chips were placed in the contralateral disc space (Figure 3A), and the first cage was inserted obliquely across the midline (Figure 3B). The first cage could be replaced more contralaterally to secure the enough space for bone graft packing and second cage insertion. When the first cage reposition is insufficient in the severely narrowed disc space, additional trial cage insertion helps to moving the first cage contralaterally (Figure 3C). After additional bone chips were placed in the ipsilateral and central disc space (Figure 3D), the second cage was inserted into the ipsilateral disc space (Figure 3E). We recommend inserting the second cage more vertically along the apophyseal ring of the intervertebral disc space to enhance cage stability.

In cases with narrow disc space, bone graft packing in the contralateral disc space may obstruct the insertion of the first cage. Therefore, preemptive bone grafting before the insertion of the first cage is not recommended. Additional bone grafting can be performed between the 2 cages if sufficient residual space is available after cage insertion

3. Percutaneous Pedicle Screw Fixation

In addition to the 2 initial incisions used for the biportal technique, 2 additional skin incisions were made for percutaneous pedicle screw fixation.

4. Postoperative Management

All patients in this study were advised to use a rigid lumbar orthosis for 3 months following surgery. Patients with osteoporosis (T score < -2.5) were managed with oral or intravenous bisphosphonates along with vitamin D; however, anabolic agents for osteoporosis, such as teriparatide and abaloparatide, were not utilized.

5. Clinical Data Collection

Clinical information was collected preoperatively, postoperatively, and at 6 months and final follow-ups, in both the inpatient ward and outpatient clinic. Patient characteristics, including sex, age, follow-up duration, type of surgery, and any complications, were documented. Outcomes were evaluated using the visual analogue scale (VAS) for back and leg pain, as well as the Oswestry Disability Index (ODI) to assess disability and pain response.

6. Radiological Data Collection

Dynamic radiography and computed tomography (CT) scans were performed at 3 months, 6 months, and 1 year to assess fusion and subsidence. However, the 3-month CT images were not utilized for fusion evaluation, as trabecular bone formation or remodeling processes were often not clearly defined in most cases. Radiography is considered unreliable for assessing the fusion state in patients with titanium cages. Therefore, interbody fusion was evaluated based on CT images and classified according to the Bridwell fusion grading system. Imaging features such as bony bridges, trabecular bone formation, lucency, and graft absorption are more clearly observable on CT scans [10,19]. The Bridwell fusion grades are classified as follows [20]: grade I, fused with trabecular bone formation; grade II, graft intact, not fully remodeled and incorporated, with no lucency present; grade III, graft intact, with potential lucency at the top and bottom of the graft; and grade IV, fusion absent, with collapse or resorption of the graft (Figure 4). Fusion rates were calculated by combining grade I and II cases.

Subsidence was measured on standing neutral lateral radiographs, assessing parallel endplates at the index level, and categorized as follows [10,21]: grade 0, 0%–24% loss of postoperative disc height; grade I, 25%–49% collapse; grade II, 50%–74% collapse; and grade III, 75%–100% collapse. The preoperative disc height at the index level was calculated by averaging the anterior, middle, and posterior intervertebral disc heights. The difference between the preoperative disc height and the height of the inserted cage was analyzed to assess the influence of taller cages on cage subsidence. Bone mineral density was evaluated prior to surgery using dual x-ray absorptiometry to assess the impact of bone quality on cage subsidence and interbody fusion.

7. Statical Analysis

Statistical analysis was performed using the Wilcoxon rank-sum test and Fisher exact test, with SAS ver. 9.4 (SAS Institute Inc., USA). A p-value of <0.05 was considered statistically significant.

RESULTS

A total of 62 patients were included in the study, with 29 patients in the single-cage group and 33 in the double-cage group, all undergoing UBE-TLIF. The mean age was 63.5±8.1 years in the single-cage group and 65.4±10.5 years in the double-cage group (Table 1). The mean follow-up duration was comparable between the groups, with 14.2±1.2 months for the single-cage group and 14.0±1.0 months for the double-cage group. The average hospital stay was 12.1±2.7 days in the single-cage group and 11.5±2.8 days in the double-cage group. The mean operative time was not significantly longer in the double-cage group (195.8±59.2 minutes) compared to the single-cage group (207.6±47.9 minutes, p=0.29) (Table 1). The most commonly treated level in both groups was L4–5, with spondylolisthesis as the predominant diagnosis (Table 1).

Postoperative complications were minimal in both groups. Magnetic resonance imaging detected surgical site hematomas in 5 patients from the single-cage group and 2 from the double-cage group; all cases were managed successfully with conservative treatment. Transient dysesthesia due to irritation of the exiting nerve root was observed in the double-cage group and resolved with conservative management. Additionally, a dural tear occurred in one patient during the insertion of the first cage in the double-cage group, which was successfully repaired endoscopically (Table 1).

Osteopenia was present in both groups, with no significant difference in T scores identified between them (Table 1).

Fusion rates, based on the Bridwell fusion grading system, were higher in the double-cage group at the 6-month follow-up, with 86.2% of patients in the single-cage group and 100% of patients in the double-cage group achieving successful fusion (grades I + II) (Table 2). Although no significant difference was found in grade I fusion rates between the 2 groups (p=0.110), the single-cage group had 4 cases of grade III fusion, while the double-cage group had no cases of grade III or IV (Table 2).

By the 1-year follow-up, fusion rates had increased to 89.6% in the single-cage group while remaining at 100% in the double-cage group, resulting in an insignificant difference between the 2 groups (p=0.100). Three cases of grade III fusion were still present in the single-cage group, but notably, no grade IV fusion cases were observed in either group.

Cage subsidence was significantly lower in the double-cage group compared to the single-cage group throughout the follow-up period. At the 6-month follow-up, total cage subsidence (grades I, II, and III) was observed in 31% of the single-cage cases, in contrast to only 3% in the double-cage group (p<0.001). Severe cage subsidence (grade II or III) occurred in 3 patients in the single-cage group, compared to one patient in the double-cage group at 6 months. By the 1-year follow-up, total cage subsidence had increased from the 6-month assessment and remained significantly higher in the single-cage group (38%) compared to the double-cage group (12.1%, p=0.03).

The preoperative disc height did not significantly differ between the 2 groups. However, the double-cage group utilized significantly taller cages than the single-cage group (p=0.01) (Table 2). While the increase in disc height was greater in the double-cage group, this difference was not statistically significant (p=0.30) (Table 2).

Both groups demonstrated significant improvements in VAS scores for back pain by the final follow-up, with no significant differences observed between the single-cage and double-cage groups (Table 3). However, the double-cage group exhibited significantly better outcomes in terms of VAS scores for leg pain and the ODI at the final follow-up (p=0.03, p=0.01, respectively) (Table 3).

DISCUSSION

Endoscopic TLIF, whether performed via uniportal or biportal endoscopy, is designed to minimize tissue damage while achieving direct decompression of the bilateral spinal canal [22]. However, several critical issues associated with endoscopic TLIF remain, including technical challenges and inadequate interbody fusion. Water-based endoscopic TLIF may compromise interbody fusion by flushing out bone graft material from the intervertebral disc space. Despite these concerns, recent systematic reviews have shown that interbody fusion outcomes with endoscopic techniques can be favorable and comparable to those of open lumbar fusion [8]. Meticulous preparation of the endplates under endoscopic visualization has been identified as a critical factor in achieving adequate interbody fusion, even in studies that utilized single, narrow-footprint cages.

To enhance interbody fusion, research efforts have focused on optimizing cage materials, dimensions, and grafting substances. Solid fusion and reduced cage subsidence are directly influenced by the increased contact area between the cage and the vertebral endplate, as well as the volume of bone grafting. To increase the cage-bone contact area, 2 strategies have been employed: the use of a large footprint cage or multiple small footprint cages. Heo et al. [23] reported UBE-TLIF utilizing a large cage designed for oblique lumbar interbody fusion. A large cage placed across the bilateral apophyseal ring significantly reduces cage subsidence, and the increased cage-bone contact area enhances the fusion rate. However, this technique presents challenges in safely inserting the large cage through a narrow working space. Preemptive bone grafting within the disc space may obstruct the repositioning of the cage, thereby limiting the sufficient bone grafting.

To address the limitations associated with large cage insertion, we developed a technique for inserting double cages through a unilateral approach. This method ensures a wider cage-bone contact area and allows for adequate packing of bone grafts. The posterior lumbar interbody fusion cage has a narrow width (11 mm), allowing for the placement of taller cages within the constrained disc space while reducing the risk of excessive neural retraction [7,16]. Once the first cage is inserted, the disc space is expanded, facilitating the smoother insertion of the second cage. In this study, the double-cage group utilized significantly taller cages than the single-cage group (11.5±1.5 mm vs. 10.6±1.3 mm, p=0.01) (Table 2).

A key factor in enhancing fusion is the volume of bone graft placed within the intervertebral disc space. While grafting within the cage is essential, external bone grafting plays an even more critical role. To maximize bone graft volume, we employed a technique where grafts were initially placed in the contralateral disc space, and the first cage was subsequently inserted to retain the graft material. This initial cage placement also expands the disc space, facilitating the insertion of additional bone grafts before placing the second cage [16]. Consequently, this approach ensures effective packing of bone graft material, even in significantly collapsed disc spaces. However, during single TLIF cage insertion, the volume of bone graft utilized is constrained by the patient's preoperative disc height. Excessive graft material placed prior to cage insertion can hinder optimal cage positioning. Furthermore, additional bone grafting following cage placement may lead to graft migration into the neuroforaminal area, potentially resulting in neural compression.

Cage subsidence could be influenced by other factors, including bone marrow density and cage height [24]. Excessive expansion of disc height by using taller cages can increase the risk of cage subsidence. In this study, despite the greater cage height and increased disc height in the double-cage group, cage subsidence was significantly lower in this group. This suggests that a wider cage-bone contact area provides stable support for the bony endplate, effectively preventing subsidence, even with the use of taller cages. Recent studies have highlighted the advantages of using 3D-printed titanium cages due to their superior fusion outcomes, which are associated with improved bone-cage interface contact compared to traditional polyetheretherketone (PEEK) cages. Kim et al. [19] reported 1-year outcomes comparing 3D-printed titanium cages with PEEK cages, showing similar overall fusion rates (grades I + II) for both types (95.0% vs. 93.0%). However, a significantly higher rate of grade I fusion was observed in the 3D-printed titanium group (77.5% vs. 51.2%, p=0.013).

In our study, the 1-year fusion rate for the single-cage group was 89.6%, with a grade I fusion rate of 65.5%. These results are comparable, though slightly lower, than those reported by Kim's study on 3D-printed titanium cages, which demonstrated an overall fusion rate of 95% and a grade I fusion rate of 77.5%. In contrast, the double-cage technique achieved substantial improvements, attaining a 100% overall fusion rate and an 84.8% grade I fusion rate. These impressive outcomes may be attributed to the characteristics of the 3D-printed titanium cage, coupled with the benefits provided by the double-cage approach.

The improved radiological outcomes associated with double-cage insertion may lead to better clinical results, particularly concerning leg pain, as assessed by the VAS, and overall disability, as measured by the ODI, at the final follow-up (Table 3). At this follow-up, the single-cage group experienced significant cage subsidence, with grade II subsidence observed in 13.8% of cases, as well as poor fusion grades, with grade III fusion seen in 10.3%. These issues may lead to recurrent radiculopathy due to restenosis and increased motion within the fusion segment. Such recurrent symptoms could negatively impact clinical outcomes, particularly in relation to leg pain.

The UBE-TLIF using a double-cage insertion method was first described by Heo et al. [7], who positioned the first cage contralaterally and the second cage ipsilaterally, employing PEEK cages. Following this, Pao [16] introduced an alternative technique that involved inserting the first cage ipsilaterally and the second cage contralaterally, utilizing both PEEK and titanium cages. Each technique presents distinct advantages and disadvantages, which may vary based on the surgical conditions of the disc space and transforaminal space. The volume of 3D-printed titanium cages is greater than that of PEEK cages of the same size due to their micro-spike design [24]. Consequently, we recommend inserting the first cage contralaterally to ensure adequate space for the second cage when utilizing 3D-printed titanium cages. If the first cage is inserted ipsilaterally and positioned more medially than intended, the space for the second cage insertion—between the dural sac and the cage—can become very narrow. This situation increases the risk of significant neural retraction injury during the insertion of the second cage. Thus, we advise starting with the contralateral cage insertion, followed by the placement of the second cage in the ipsilateral disc space, particularly in cases involving narrow disc heights and transforaminal spaces. Furthermore, this approach is especially beneficial at levels with a larger facet width, such as L4–5 and L5–S1 levels [13]. However, it may increase the complexity of the procedure and may not be feasible at upper lumbar levels.

This study has several limitations. Surgeon-related factors may impact radiological outcomes, including the methods of endplate preparation, the amount of bone graft packing, and the choice of cage type. Although we excluded cases with poor bone mineral density (T score < -3.0), multiple level fusions, and concomitant decompression surgeries at adjacent levels, other confounding variables, such as body weight and daily activities, may still influence the surgical results. Furthermore, long-term follow-up studies are necessary to evaluate the clinical implications, including adjacent segment disease.

CONCLUSION

Supplementary Material

Supplementary Video Clips 1-2 are available at https://doi.org/10.21182/jmisst.2024.01970.

Supplementary video clip 1.

Unilateral biportal endoscopic transforaminal interbody fusion using a single cage at the L3–4 left.

Supplementary video clip 2.

Unilateral biportal endoscopic transforaminal interbody fusion using a double cage at the L4–5 right.

NOTES

Conflicts of interest

JYK, HSK, and DHH are member of the Editorial Board of Journal of Minimally Invasive Spine Surgery & Technique, are the author of this article. However, they played no role whatsoever in the editorial evaluation of this article or the decision to publish it. All Authors have no conflict of interest to declare.

Funding/Support

This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Figure 1.

The 3-dimensional (3D)-printed titanium cages and specialized instruments. (A) A 3D-printed titanium cage with a length of 28 mm was used for the single-cage group. (B) A 24-mm cage was utilized for the double-cage group. (C) A scope retractor, mounted on the endoscopic trocar (asterisk), is employed for neural protection and soft tissue retraction.

Figure 2.

Intraoperative x-rays and photographs. (A) The endoscope and instruments are positioned at the initial docking points. Two skin incisions are made at the lateral borders of the pedicles (red ovoid indicates the endoscopic portal; yellow ovoid indicates the working portal; dotted circle denotes the pedicle). (B) The lateral x-ray image depicts the cage insertion step. The scope retractor (asterisk) ensures protection of the neural structures during the procedure. (C) Intraoperative photographs following the insertion of the single cage. (D) Intraoperative images after the insertion of the double cage.

Figure 3.

Illustrations of double-cage insertion. (A) Bone graft is placed in the contralateral disc space using a customized funnel. (B) The first cage is inserted, crossing the midline to encompass the bone graft in the contralateral disc space. (C) A trial cage is then inserted to secure the space for the second cage and to accommodate additional bone grafting. (D) Additional bone graft is subsequently packed. (E) The second cage is positioned in the ipsilateral disc space.

Figure 4.

Computed tomography (CT) images illustrating the fusion status according to the Bridwell Fusion Grading System. (A) Grade I fusion at 6 months in the single-cage group, demonstrating a clearly visible bony bridge (red asterisk). (B) Grade I fusion at 1 year in the single-cage group. (C, D) Grade II fusion at 1 year in the single-cage group. (E, F) Grade III fusion at 1 year in the single-cage group, with severe cage subsidence (red asterisk) and endplate lucency (red arrowhead) observed. A periscrew halo was also noted (red arrow). (G) Grade I fusion at 6 months in the double-cage group.

Table 1.

Patient information

Characteristic	Single cage (n=29)	Double cages (n=33)	p-value
Sex, male: female	12:17	20:13	-
Age (yr)	63.5±8.1 (47–78)	65.4±10.5 (38–88)	0.45
Follow-up period (mo)	14.2±1.2 (13–17)	14.0±0.0 (12–17)	0.46
Hospital stays (day)	12.1±2.7 (8–22)	11.5±2.8 (7–18)	0.46
Operation time (min)	207.6±47.9 (155–220)	195.8±59.2 (120–240)	0.29
Canal decompression
Unilateral:bilateral	18:11	15:18
Operated level
L3–4	4	2	-
L4–5	16	23	-
L5–S1	9	8	-
Diagnosis
Recurrent disc herniation	4	5
Spinal stenosis	11	8
Spondylolisthesis	14	20
BMD (T score)	-1.0±1.4 (-2.8 to 0.4)	-1.0±1.3 (-2.7 to 3.2)	0.89
Complications
Hematoma	5	2
Screw malposition	1	-
Postoperative dysesthesia of exiting nerve root	-	2
Dural tear	-	1

Values are presented as number or mean±standard deviation (range).

BMD, bone marrow density.

Age, follow-up period, hospital stays, operation time: Wilcoxon rank-sum test.

Table 2.

Radiological outcomes of interbody fusion

Fusion grading scale	Single cage (n=29)	Double cage (n=33)	p-value
Bridwell fusion grading system
6 Months			0.11
Grade I	9 (31)	13 (39.4)
Grade II	16 (55.2)	20 (60.6)
Grade III	4 (14.8))	0 (0)
Grade IV	0 (0)	0 (0)
Fusion rate (grades I + II)	25 (86.2)	33 (100)	0.04*
12 Months			0.09
Grade I	19 (65.5)	28 (84.8)
Grade II	7 (24.2)	5 (15.2)
Grade III	3 (10.3)	0 (0)
Grade IV	0 (0)	0 (0)
Fusion rate (grades I + II)	26 (89.6)	33 (100)	0.10
Subsidence grading	Grade 0/grade I/grade II/grade III
6 Months			-
Grade 0	20	32
Grade I	6	1
Grade II	3	0
Grade III	0	0
Subsidence^†	9 (31)	1 (3)	0.01^*
12 Months			-
Grade 0	18	29
Grade I	7	4
Grade II	4	0
Grade III	0	0
Subsidence^†	11 (38)	4 (12.1)	0.03^*
Height of disc and cage (mm)
Cage height	10.6±1.3 (8–12)	11.5±1.5 (8–13)	0.01^*
Preoperative disc height	6.5±2.9 (3.5–11.9)	6.9±2.2 (4.1–12.2)	0.46
Disc height increment (cage height-disc height)	4.1±2.6	4.6±1.8	0.30

Values are presented as number (%) or mean±standard deviation (range).

Fusion and subsidence grades: Fisher exact test due to small sample sizes, Height of disc and cage: Wilcoxon rank-sum test.

^*p<0.05, statistically significance differences.

^†Subsidence includes grades I, II, and III.

Table 3.

Clinical outcomes

Variable	Single cage (n=29)	Double cage (n=33)	p-value
VAS_back pain
Preoperative	7.8±0.8	7.6±1.0	0.45
Final follow-up	2.2±0.5	2.0±0.6	0.33
VAS_leg pain
Preoperative	7.6±1.2	7.2±1.9	0.83
Final follow-up	1.7±0.6	1.4±0.5	0.03^*
ODI
Preoperative	28.6±4.2	28.3±2.8	0.71
Final follow-up	10.8±2.0	9.4±1.9	0.01^*

Values are presented as mean±standard deviation.

VAS, visual analogue scale; ODI, Oswestry Disability Index.

VAS and ODI: Wilcoxon rank-sum test.

^*p<0.05, statistically significance differences.

REFERENCES

1. Bin Abd Razak HR, Dhoke P, Tay KS, Yeo W, Yue WM. Single-level minimally invasive transforaminal lumbar interbody fusion provides sustained improvements in clinical and radiological outcomes up to 5 years postoperatively in patients with neurogenic symptoms secondary to spondylolisthesis. Asian Spine J 2017;11:204–12.

2. Yang Y, Liu ZY, Zhang LM, Pang M, Chhantyal K, Wu WB, et al. Microendoscopy-assisted minimally invasive versus open transforaminal lumbar interbody fusion for lumbar degenerative diseases: 5-year outcomes. World Neurosurg 2018;116:e602–10.

3. Heo DH, Park CK. Clinical results of percutaneous biportal endoscopic lumbar interbody fusion with application of enhanced recovery after surgery. Neurosurg Focus 2019;46:E18.

4. Heo DH, Son SK, Eum JH, Park CK. Fully endoscopic lumbar interbody fusion using a percutaneous unilateral biportal endoscopic technique: technical note and preliminary clinical results. Neurosurg Focus 2017;43:E8.

5. Kang MS, You KH, Choi JY, Heo DH, Chung HJ, Park HJ. Minimally invasive transforaminal lumbar interbody fusion using the biportal endoscopic techniques versus microscopic tubular technique. Spine J 2021;21:2066–77.

6. Kim HS, Wu PH, Lee YJ, Kim DH, Jang IT. Technical considerations of uniportal endoscopic posterolateral lumbar interbody fusion: a review of its early clinical results in application in adult degenerative scoliosis. World Neurosurg 2021;145:682–92.

7. Heo DH, Hong YH, Lee DC, Chung HJ, Park CK. Technique of biportal endoscopic transforaminal lumbar interbody fusion. Neurospine 2020;17(Suppl 1):S129–37.

8. Heo DH, Lee DC, Kim HS, Park CK, Chung H. Clinical results and complications of endoscopic lumbar interbody fusion for lumbar degenerative disease: a meta-analysis. World Neurosurg 2021;145:396–404.

9. Kim JE, Choi DJ. Biportal endoscopic transforaminal lumbar interbody fusion with arthroscopy. Clin Orthop Surg 2018;10:248–52.

10. Kim JY, Hong HJ, Kim HS, Heo DH, Choi SY, Kim KM, et al. Comparative analysis of uniportal and biportal endoscopic transforaminal lumbar interbody fusion in early learning stage: technical considerations and radiological outcomes. J Minim Invasive Spine Surg Tech 2024;9(Suppl 1):S14–23.

11. Wu PH, Kim HS, Lee YJ, Kim DH, Lee JH, Jeon JB, et al. Uniportal full endoscopic posterolateral transforaminal lumbar interbody fusion with endoscopic disc drilling preparation technique for symptomatic foraminal stenosis secondary to severe collapsed disc space: a clinical and computer tomographic study with technical note. Brain Sci 2020;10:373.

12. Kettler A, Schmoelz W, Kast E, Gottwald M, Claes L, Wilke HJ. In vitro stabilizing effect of a transforaminal compared with two posterior lumbar interbody fusion cages. Spine (Phila Pa 1976) 2005;30:E665–70.

13. Lee DC, Kim JY, Kim TH, Park CK. Unilateral approach with two-cage insertion for full endoscopic transforaminal lumbar interbody fusion: technical report. Acta Neurochir (Wien) 2022;164:1521–7.

14. Eum JH, Park JH, Song KS, Lee SM, Suh DW, Jo DJ. Endoscopic extreme transforaminal lumbar interbody fusion with large spacers: a technical note and preliminary report. Orthopedics 2022;45:163–8.

15. Heo DH, Eum JH, Jo JY, Chung H. Modified far lateral endoscopic transforaminal lumbar interbody fusion using a biportal endoscopic approach: technical report and preliminary results. Acta Neurochir (Wien) 2021;163:1205–9.

16. Pao JL. Biportal endoscopic transforaminal lumbar interbody fusion using double cages: surgical techniques and treatment outcomes. Neurospine 2023;20:80–91.

17. Kim JY, Heo DH. Biportal endoscopic cervical open-door laminoplasty to treat cervical spondylotic myelopathy. Acta Neurochir (Wien) 2024;166:182.

18. Kim JY, Kim KM, Choi SY, Heo DH, Kim HS, Hong HJ, et al. Advanced technique of 360° decompression for thoracic spondylotic myelopathy using the biportal endoscopic posterior approach. J Minim Invasive Spine Surg Tech 2024;9:102–9.

19. Kim DY, Kwon OH, Park JY. Comparison between 3-dimensional-printed titanium and polyetheretherketone cages: 1-year outcome after minimally invasive transforaminal interbody fusion. Neurospine 2022;19:524–32.

20. Bridwell KH, Lenke LG, McEnery KW, Baldus C, Blanke K. Anterior fresh frozen structural allografts in the thoracic and lumbar spine. Do they work if combined with posterior fusion and instrumentation in adult patients with kyphosis or anterior column defects? Spine (Phila Pa 1976) 1995;20:1410–8.

21. Marchi L, Abdala N, Oliveira L, Amaral R, Coutinho E, Pimenta L. Radiographic and clinical evaluation of cage subsidence after stand-alone lateral interbody fusion. J Neurosurg Spine 2013;19:110–8.

22. Wu PH, Kim HS, Jang IT. A narrative review of development of full-endoscopic lumbar spine surgery. Neurospine 2020;17(Suppl 1):S20–33.

23. Heo DH, Jang JW, Park CK. Enhanced recovery after surgery pathway with modified biportal endoscopic transforaminal lumbar interbody fusion using a large cage. Comparative study with minimally invasive microscopic transforaminal lumbar interbody fusion. Eur Spine J 2023;32:2853–62.

24. Patel NA, O’Bryant S, Rogers CD, Boyett CK, Chakravarti S, Gendreau J, et al. Three-dimensional-printed titanium versus polyetheretherketone cages for lumbar interbody fusion: a systematic review of comparative in vitro, animal, and human studies. Neurospine 2023;20:451–63.