INTRODUCTION
Minimally invasive spinal surgeries, including endoscopic spinal surgeries, are becoming increasingly popular. In particular, the use of biportal endoscopic spine surgery (BESS) is increasing [1,2]. This technique merges the benefits of traditional posterior approaches with those of minimally invasive surgery. It involves accessing the surgical site through 2 skin incisions, each spaced 1 cm apart, facilitating the utilization of both hands via 2 distinct portals [3,4]. The application of biportal endoscopy to lumbar interbody fusion surgery provides a safer approach for interbody cage insertion [5-9], minimizing damage to the paraspinal muscles, improving early postoperative clinical outcomes, enabling faster rehabilitation, and reducing complications related to blood transfusion [5,10]. Despite these advantages, there are concerns about the fusion rate and cage subsidence when using biportal endoscopy. The disadvantages of BESS include a smaller amount of available autogenous bone and the potential loss of fusion material due to continuous irrigation. Hence, the selection of suitable cage material, which impacts both fusion rate and cage subsidence, is crucial.
Polyetheretherketone (PEEK) and titanium are the most commonly used cage materials with distinct advantages and disadvantages. PEEK cages have the advantages of being radiolucent and biocompatible, with an elastic modulus similar to that of bone [11]. Titanium cages have good biocompatibility and better fusion rates [12,13], although they have a heightened risk of subsidence owing to their high elastic modulus. Recently, the incorporation of 3-dimensional (3D)-printing technology has resulted in a reduction in the elastic modulus [14]. Additionally, the creation of a porous structure has improved osteoinduction. Moreover, the surface area of the intervertebral cage affects the fusion rate and potential subsidence. Some lumbar interbody fusion studies have reported lower rates of cage subsidence associated with larger cages [15-17].
Therefore, to ensure appropriate interbody cage selection and to improve surgical outcomes, this study aimed to compare fusion rate, incidence and severity of cage subsidence, and other radiographic parameters according to cage material and size in biportal endoscopic transforaminal lumbar interbody fusion (BE-TLIF).
MATERIALS AND METHODS
1. Study Design and Patient Characteristics
In this retrospective cohort study, we reviewed the medical charts and radiographic image analyses of 121 patients who underwent BE-TLIF between April 2019 and February 2023. This study was conducted in accordance with the guidelines outlined in the Declaration of Helsinki. The study was approved by the Institutional Review Board of Hallym University Kangnam Sacred Heart Hospital, Hallym University College of Medicine (approval number: 2024-01-002). The Institutional Review Board of Hallym University Kangnam Sacred Heart Hospital granted an exemption from obtaining informed consent. The inclusion criteria were as follows: (1) age 50–80 years; (2) refractory pain that was not controlled by conservative treatment for >3 months; and (3) single-level lumbar degenerative disease requiring fusion surgery. The exclusion criteria were as follows: (1) multilevel disease; (2) a history of prior surgery at the same level; (3) other pathologic conditions, such as fracture, tumor, infection, and inflammatory disease; (4) osteoporosis with a T score of -2.5 or lower; (5) endplate injury that occurred intraoperatively; or (6) lack of 1-year follow-up data.
Eventually, 86 patients were analyzed after excluding 35 patients: 19 lost to follow-up, 13 with intraoperative endplate injury, 1 with dural tear, and 2 with revision surgery due to suspected nonunion and incomplete decompression. Of these, 55 were in the 3D-printed cage group, and 31 were in the PEEK cage group. Patients with 3D-printed cages were divided into 2 groups: those with regular-sized cages (<400 mm2 area, n=31) and those with large cages (>400 mm2 area, n=24) (Fig. 1). The patients were categorized as follows: group A, regular 3D-printed titanium cage; group B, regular PEEK cage; and group C, large 3D-printed titanium cage.
2. Radiographic Parameters
Radiographic measurements were acquired using preoperative, immediate postoperative, and 1-year postoperative wholespine standing radiographs. The immediate postoperative radiograph was obtained 1- or 2-day postsurgery. Radiographs were analyzed for changes in lumbar lordosis, anterior and posterior disc heights, segmental lordosis, intervertebral disc angle, and intervertebral foraminal height in the operated segment. Radiographic parameters were measured using commonly used methods. Cage subsidence has been measured using various methods in previous studies. We evaluated the 1-year postoperative sagittal computed tomography (CT) images (2-mm slices) to assess cage subsidence. Cage subsidence was defined as sinking >2 mm in the cephalad and/or caudal endplates of the operated segment. This method was used to evaluate the incidence rate of cage subsidence. We also evaluated the severity of cage subsidence. The severity of the level was considered to be the maximum subsidence value measured by the above method. To evaluate cage position, the central point ratio (CPR) [18], we measured the ratio of the distance between the cage midpoint and the posterior superior corner of the lower vertebra, divided by the length of the superior endplate of the lower vertebra (Fig. 2). Anterior positioning was indicated when CPR was >0.5. Fusion status was also analyzed using 1-year postoperative CT. Fusion was classified according to the Bridwell classification [19], with classifications I and II defined as fusion and III and IV as nonfusion. The radiographic measurements were performed by 2 blinded independent orthopedic surgeons who did not participate in this study. For fusion grading, if the evaluators’ opinions differed, a final consensus was obtained between the authors.
3. Surgical Procedure
All surgeries were performed by a highly experienced singlespine surgeon. Under general endotracheal anesthesia, the patients were placed in a knee-flexed prone position on a radiolucent spinal table. The procedure was carried out using the multiportal approach technique [8]. Initially, central portals were created for central canal decompression and contralateral facet release. A single portal (M-portal) was created immediately superior to the midline of the intervertebral disc space. Another portal (M´ portal) was made approximately 1.5- to 2-cm caudal to perform central canal decompression through unilateral laminectomy and bilateral decompression. Decompression proceeded until reaching the contralateral facet, where contralateral facetectomy was performed to enhance segmental mobility and prevent subsidence during cage insertion. Subsequently, 2 additional portals (P+2 [L] and [R]) for the far-lateral approach were created approximately 2 cm laterally from the lateral margin of the pedicle. Ipsilateral facetectomy and foraminotomy were performed using these portals. This approach enables bilateral facetectomy, facilitating the collection of a relatively substantial amount of autobone, typically averaging 6–9 mL. Subsequently, the disc space was prepared. At this stage, using real-time imaging with biportal endoscopy allows for disc space preparation on the contralateral side, which is challenging to visualize in minimally invasive TLIF (MIS-TLIF). Cage insertion was then performed using all 4 portals. Before cage insertion, bone grafting was performed by mixing 3 mL of demineralized bone matrix (DBM) with autobone obtained from laminectomy and facetectomy. Initially, the inside of the cage was filled with fusion material, and any remaining material was packed on the anterior side of the disc space before cage insertion. The cage was inserted as parallel as possible to the coronal plane. Percutaneous pedicle screw fixation was uniformly carried out in all patients, with intraoperative C-arm imaging employed for confirmation (Figs. 3–5).
Both the 3D-printed titanium (regular size: EIT cellular titanium cage, DePuy Synthes, West Chester, PA, USA; large size: 3D XTLIF CAGE, Endovision Co., Ltd., Daegu, Korea) and PEEK cages (Lospa Is TLIF cage system, Corentec Co., Ltd., Seoul, Korea) were used for interbody fusion. Cage size was classified based on the surface area using a 400-mm² standard. The surface area was calculated by multiplying the length and width of each cage. The cage height was tailored to match the disc space of the patient. A percutaneous pedicle screw system (Lospa Is MIS Spinal System; Corentec Co., Ltd.) was used for fixation. For achieving solid fusion, a consistent amount of 3 mL of DBM (OSG DBM Syringe, OSSGEN Co., Ltd., Daegu, Korea) was uniformly used in all patients.
4. Statistical Analyses
For continuous variables, each group was first tested for normality using the Kolmogorov-Smirnov test to determine the normality of data distribution. For normally distributed data, 1-way analysis of variance (ANOVA) was used to compare continuous variables, with post hoc tests when significant differences were found. One-way ANOVA is a hypothesis test employed to ascertain whether there exists a notable difference among the means of 3 or more groups. Chi-square tests were used to compare categorical variables such as fusion and subsidence rates between the 2 groups. Student t-test was used to assess the differences in the severity of cage subsidence. Statistical significance was set at p<0.05. All statistical analyses were performed using IBM SPSS Statistics ver. 26.0 (IBM Co., Armonk, NY, USA).
RESULTS
1. Demographic Data
We analyzed data from 86 patients. Among them, 31 received regular-sized 3D-printed titanium cages (group A), 31 received PEEK cages (group B), and 24 received large 3D-printed titanium cages (group C). No significant differences were noted in age, body mass index (BMI), bone mineral density, operation level, cage height, or CPR among the 3 groups (Table 1).
2. Radiographic Parameters
All radiographic parameters were measured using simple radiographs preoperatively, immediately postoperatively, and at 1 year postoperatively. We compared changes from preoperative to immediate postoperative measurements and from immediate postoperative to 1-year postoperative measurements for each variable but found no significant differences among the 3 groups (Table 2).
3. Fusion Rate and Cage Subsidence
The fusion rate was highest in group C, followed by group A and then group B, although the differences were not statistically significant. Fusion grade was also not significantly different among the 3 groups. The incidence of cage subsidence showed no significant difference between groups A and B, as well as between groups B and C. However, it was notably lower in group C than in group A despite using the same cage material. The cage subsidence severity was lowest in group C followed by groups B and A. group C had significantly lower cage subsidence severity than groups A and B (Table 3).
DISCUSSION
BE-TLIF is a viable option with good surgical outcomes [5-8], although the limited availability of autogenous bone and potential loss of fusion material due to continuous irrigation raises concerns about fusion, and cage size could potentially influence subsidence. In this study, we compared the fusion rate, subsidence, and other radiographic parameters based on the material and size of the cage used in BE-TLIF. We found that cage materials did not affect the 1-year postoperative outcomes, but cage subsidence was markedly reduced with larger cages in BE-TLIF.
The surgeon sequentially selected cages during the surgery, progressing from group A to C. Intraoperative endplate injury was observed in 13 cases: 6 in group A, 4 in group B, and 3 in group C, notably occurring more frequently during the early learning curve period (Table 4). Performing BE-TLIF in a limited endoscopic surgical field poses challenges, especially before mastering the learning curve. It is speculated that increased bony endplate injury during cage insertion, may contribute to cage subsidence. This phenomenon is not unique to BE-TLIF but can also occur in existing MIS-TLIF procedures [20]. Cases of intraoperative endplate injury that occurred during the learning curve were excluded from the study as they would have biased the study analyzing the effect of cage size and material on subsidence.
Cage subsidence following lumbar interbody fusion surgery has significant implications for postoperative decompression, particularly foraminal restenosis, and can influence fusion rates, as shown in some studies [21,22]. Restoration of disc height via cage height is critical in achieving indirect decompression of foraminal stenosis. However, if the cage subsides, reducing the disc height, restenosis may occur, leading to radiculopathy recurrence. Typical risk factors for cage subsidence include conditions such as osteoporosis, higher BMI, multifidus muscle atrophy, poor endplate bone quality, posterior cage positioning, and use of an excessively high cage [23,24]. Additionally, the use of a smaller cage surface area may increase cage subsidence risk, whereas the use of a larger surface area may decrease this risk [25,26].
Cage material type affects cage subsidence significantly. PEEK cages have the advantage of being radiolucent, allowing for easy postoperative assessment of fusion status, and given their high biocompatibility, they have few adverse effects [11,27]. Additionally, the elastic modulus of PEEK is similar to that of bone, reducing the likelihood of subsidence. In contrast, titanium cages exhibit strong corrosion resistance and stiffness, thereby providing excellent immediate postoperative stability. They are also osteoinductive and have favorable fusion rates [12,13,28]. However, with their high elastic moduli, the cage subsidence risk increases. Some previous meta-analyses comparing cage materials in lumbar interbody fusion reported that PEEK and titanium cages showed similar fusion rates but higher subsidence rates with titanium cages [29,30]. However, recent advances in 3D-printing technology have allowed the development of porous materials, and in vitro experiments have shown that a specific pore size is associated with more osteoblasts [31]. Application of 3D-printing technology allows implementation of this osteoblast-friendly pore size. Thus, the elastic modulus was reduced, and osteoinduction was further enhanced. Therefore, 3D-printed titanium cages may increase the fusion rate and reduce cage subsidence. A recent study comparing PEEK and 3D-printed titanium cages reported that 3D-printed titanium cages yielded a better fusion rate [32-35]. Similarly, in our study, the fusion rate in the 3D-printed titanium cage group was higher than PEEK, but not significantly.
The cage surface area is another important factor affecting the fusion rate and cage subsidence. Regular and large cages were used in this study. Larger 3D titanium cages resulted in a better fusion rate than regular cages of the same material, but this was not statistically significant. Both the cage subsidence rate and severity of subsidence of the large cage were significantly lower than those of the regular cage. A previous biomechanical study of cage type in lateral lumbar interbody fusion showed that wider cages with larger surface areas had greater biomechanical stability [36,37]. With a wider surface contact area, the pressure per unit area will likely decrease, potentially reducing the force between the cage and endplate. Consequently, the cage subsidence is expected to decrease. Additionally, in a study comparing regular-sized and large cages, less cage subsidence was observed in the large-cage group, similar to our study [15-17].
The anatomical structure of the vertebral body also influences cage subsidence. Because the endplate is more rigid in the peripheral region, called the epiphyseal ring, than in the central region of the vertebral body, a cage with a large footprint should be used to ensure proper positioning of the cage on both epiphyseal rings [38]. With a large cage, pedicle screw interference can occasionally stop cage subsidence, suggesting that a cage that extends to the screw position may increase the bone density during subsidence, potentially serving as a subsidence-preventive measure. Therefore, both the incidence and severity of cage subsidence are likely reduced in large cages.
In most previous studies comparing BE-TLIF and other minimally invasive TLIF procedures, no significant differences were found in surgical outcomes [5,39,40]. Despite concerns about the washout of the fusion material, BE-TLIF may yield similar results because of the detailed endplate preparation and real-time biportal endoscopic view. However, Park et al. [41] reported that BE-TLIF yielded less favorable results in terms of definite fusion than conventional posterior lumbar interbody fusion, possibly due to inadequate autogenous bone harvesting, bone material washout, and the likelihood of complications such as cage subsidence or retropulsion. In a meta-analysis comparing minimally invasive TLIF (MIS-TLIF) and open TLIF, fusion rates of 80.5% and 91.1% were reported for MIS-TLIF and open TLIF, respectively [42]. In MIS-TLIF, similar to that in BE-TLIF, the amount of autogenous bone harvested during surgery was lower than that harvested during conventional open surgery. Concerns have been raised that the fusion rate could be lowered because of a limitation during endplate preparation, given the narrow working space using a tubular retractor. In comparison, BE-TLIF, which is similar to open lumbar surgery, is performed in a wide and free working space because of the advantages of biportal endoscopy. Additionally, because endplate preparation can be performed more carefully under direct endoscopic vision, a higher fusion rate is expected. Specifically, employing the far multiportal approach of biportal endoscopy allows for a relatively safer insertion of a large footprint cage through a more lateral corridor. Therefore, we consider BE-TLIF to be a viable option for proficient surgeons. Moreover, using a 3D-printed large titanium cage, as in this study, reduces concerns regarding conventional MIS-TLIF and can improve radiographic outcomes.
This study had some limitations. First, it was a retrospective study. Second, because this was a single-institution study involving a single surgeon, the sample size was small. Due to the small sample sizes in all 3 groups, there is a potential for selection bias stemming from an imbalance among them. BE-TLIF is a relatively new surgical technique; consequently, there has not been sufficient time to accumulate a substantial number of cases. Third, a clinical evaluation was not performed. Specifically, because fusion rate and cage subsidence were confirmed only radiographically, it was difficult to ascertain whether it contributed to symptom manifestation. However, we aimed to confirm the occurrence of radiographic cage subsidence and bony fusion. Therefore, whether a higher subsidence incidence yields worse clinical outcomes remains unclear. Compared with previous studies that evaluated cage subsidence through simple radiography, the cage subsidence rate in this study was slightly higher, possibly because the evaluation involved CT [21,22]. Additionally, cage artifacts appear to influence the measurement of cage subsidence using CT. Future studies incorporating long-term follow-up and randomized controlled trials with larger sample sizes are required. Despite these limitations, this study represents an investigation of cage material and size in terms of fusion rate, subsidence, and other radiographic parameters in BE-TLIF, which have not been reported concurrently to date.
CONCLUSION
The study findings revealed no significant difference in fusion rate according to cage material and size in BE-TLIF; however, cage subsidence was reduced with larger cage use. Once the insertion of a cage with a large footprint is familiarized with biportal endoscopic assistance, a 3D-printed large titanium cage could be a feasible option to reduce cage subsidence and increase the stability of the surgical segment during the follow-up period.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
