INTRODUCTION
Degenerative lumbar diseases, including disc herniation and spinal stenosis, are common. When accompanied by moderate-to-severe foraminal stenosis or spondylolisthesis, simple decompression alone may carry a high risk of residual symptoms or recurrence, and lumbar interbody fusion (LIF) is often considered as a treatment option. Posterior LIF (PLIF) is a conventional surgical technique [3,22]. PLIF provides direct central and bilateral foraminal decompression and allows the correction of spinal instability through a familiar posterior approach. Despite variations among individual techniques, it has shown fusion rates exceeding 90%, and remains one of the most commonly performed fusion procedures [22,38]. However, the midline posterior approach may cause paraspinal muscle injury and is often associated with greater intraoperative blood loss. During cage insertion, retraction of the dural sac and exiting nerve root (ENR) is required by an assistant, and the limited working space increases the risk of neural injury and restricts cage height and angle selection [29]. With the growing interest in minimally invasive surgery among surgeons and patients, advances in endoscopic instruments and techniques have expanded the application of endoscopic LIF. Both biportal endoscopic transforaminal LIF (BE-TLIF) and full endoscopic TLIF (FE-TLIF) have been introduced previously [1,2,7-9,33]. Initially, FE-TLIF was performed in a single cage [17,30]. However, single-cage placement has been shown to provide less stability during lateral bending, flexion, and extension than double-cage placement [16]. To overcome this limitation, Lee et al. [18] reported the technical feasibility of double-cage FE-TLIF in 2022. However, no studies have evaluated the clinical outcomes, radiological results, or complications associated with double-cage FE-TLIF. Although several studies have compared the outcomes of single-cage FE-TLIF with those of conventional PLIF, TLIF, minimally invasive TLIF, and even BE-TLIF [5,6,21,31,38], to date, no study has analyzed the 1-year outcomes of double-cage FE-TLIF or directly compared them with those of PLIF. Therefore, this study aimed to investigate and analyze these aspects by comparing the clinical and radiological outcomes and complications of double-cage FE-TLIF with those of PLIF.
MATERIALS AND METHODS
Study design and subjects
This retrospective cohort study was approved by the Institutional Review Board (IRB) of Chosun University Hospital prior to its initiation (IRB No. 2025-06-012-001). The clinical and radiological data of patients who underwent double-cage FETLIF or PLIF for degenerative lumbar disease between March 2021 and February 2024 at our institution were retrospectively reviewed. All data were collected retrospectively from electronic medical records and radiological examinations.
Inclusion criteria were as follows : 1) single-level degenerative lumbar disease requiring fusion confirmed on imaging; 2) presence of Meyerding grade I spondylolisthesis; 3) low back pain and/or radicular pain with neuropathic symptoms; 4) pain refractory to at least 12 weeks of conservative treatment; 5) availability of bone mineral density (BMD) data obtained using dual-energy X-ray absorptiometry; and 6) a minimum follow-up duration of 12 months.
Exclusion criteria were as follows : 1) multilevel lumbar disease; 2) history of previous surgery at the same level; 3) concomitant spinal infection, tumor, fracture, or tuberculosis; 4) Meyerding grade ≥II spondylolisthesis; and 5) incomplete 12-month follow-up data.
A total of 115 patients who met the inclusion criteria were enrolled in this study. Among them, 58 patients underwent double-cage FE-TLIF and 57 underwent PLIF.
Surgical technique
PLIF
Under general anesthesia, the patient was placed in the prone position on a radiolucent table. A midline incision was made over the index level, and subperiosteal dissection was performed to expose the spinous processes, laminae, and bilateral facet joints. A self-retaining retractor was positioned to minimize the paraspinal muscle ischemia. Standard laminectomy and partial or complete facetectomy were performed to decompress the thecal sac and the exiting and traversing nerve roots. After removing the hypertrophied ligamentum flavum and the osteophytes, bilateral annulotomies were performed near the posterior longitudinal ligaments. Sequential discectomy and careful endplate preparation were performed while preserving the subchondral bone. The disc space was packed with a mixture of autologous bone, allografts, and demineralized bone matrix (DBM). Two interbody cages (polyetheretherketone [PEEK] or 3D-printed titanium), selected according to the trial size, were inserted bilaterally into the grafted disc space in the anteromedial paramedian direction under fluoroscopic guidance. After meticulous hemostasis, a drain was inserted, and the wound was closed in layers. Subsequently, percutaneous transpedicular screw fixation was performed under fluoroscopic guidance.
FE-TLIF
The patient was placed in the prone position on a radiolucent table. Under C-arm fluoroscopic guidance, a single 10-12-mm skin incision was made at the appropriate entry point to perform a posterolateral approach (facet resection FE-TLIF) [4]. Following serial dilation, a full endoscopic system (iLESSYS® Delta HD Laminoscope®; joimax GmbH, Karlsruhe, Germany; working length, 125 mm; outer diameter, 10.0 mm; working channel diameter, 6.0 mm; optic angle, 15°) with a compatible working sheath was docked at the area around the lateral facet and the superior articular process (SAP). Bone work was performed around the inferior articular process and isthmus of the lamina using an endoscopic burr, followed by the removal of these structures. Unilateral laminotomy for bilateral decompression was then performed to decompress the thecal sac and bilateral traversing nerve roots. After partial resection of the SAP to expose the disc space, an annulotomy was performed. Sequential discectomy and meticulous cartilaginous endplate preparation were performed using various endoscopic instruments. Following the replacement of the working sheath with a glider, sequential disc distraction was performed to restore the disc height, and the optimal cage height and angle were determined using trial spacers. After filling the interbody space with an autograft and allograft, an interbody cage (PEEK or 3D-printed titanium) was packed with autologous bone and DBM and inserted across the midline to the contralateral side. Subsequently, a second cage was inserted ipsilaterally. In this study, double-cage FE-TLIF was selectively applied only at L3-4 and lower lumbar levels. After meticulous hemostasis, a drain was placed, and percutaneous transpedicular screw fixation was performed under C-arm fluoroscopic guidance, as in the PLIF procedure (Figs. 1 and 2).
Outcome measurement and evaluation
To evaluate the perioperative outcomes, the following parameters were assessed and compared between the two groups : operative time, intraoperative estimated blood loss, cage type and size, and intraoperative cage malposition caused by surgical errors. In the FE-TLIF group, intraoperative blood loss was calculated by subtracting the amount of irrigation fluid used from the total volume of the suction bottle, whereas in the PLIF group, blood loss was determined directly from the total suction volume.
For clinical outcomes, the length of hospital stay, Macnab’s criteria, Visual analog scale (VAS) scores for back and leg pain, and the Oswestry disability index (ODI) were evaluated at baseline and 1 week, 1 month, 6 months, and 12 months postoperatively and compared between the two groups.
Radiological assessments were performed using whole-spine standing radiographs obtained preoperatively and at 12 months postoperatively. Segmental lumbar lordosis angle (SLA), total lumbar lordosis angle (TLA), and mean disc height (DH) were measured and analyzed from these images. The TLA was defined as the angle between the upper endplate of L1 and the upper endplate of S1, whereas the SLA was measured as the angle between the upper endplate of the upper vertebra and the lower endplate of the lower vertebra at the operated segment. The mean DH was calculated as the average anterior DH and posterior DH. In addition, postoperative 1-year computed tomography (CT) scans were used to assess fusion status and cage subsidence. Fusion status was classified according to the Bridwell grading system, in which grades I and II were considered fusion, and grades III and IV were defined as non-fusion, and the results were compared between the groups. Cage subsidence was measured on sagittal CT images as the degree of cage sinking into the adjacent vertebral endplate in the operated segment [37]. Subsidence was categorized as no subsidence, ≤2 mm, and clinically significant subsidence (>2 mm) was compared between the groups.
Postoperative complications, including epidural hematoma, infection, dural tear, and revision surgery were also investigated. Epidural hematomas and infections were recorded only when revision surgery was required. Dural tears were classified as those that healed with conservative treatment or those that required reoperation.
Statistical analysis
Data were analyzed using R statistical software (version 4.2.2; R Foundation for Statistical Computing, Vienna, Austria). Continuous variables are expressed as mean±standard deviation and categorical variables as numbers (%). Between-group comparisons of continuous variables were performed using Welch’s t-tests, and within-group changes were analyzed using paired t-tests. Categorical variables were compared using chi-squared or Fisher’s exact tests, as appropriate. To minimize baseline imbalances in clinical scores, improvement from baseline (ΔVAS and ΔODI) was calculated and compared between groups. Fusion (grades I-II vs. III-IV) and clinically relevant subsidence (>2 mm) were further analyzed using multivariable logistic regression adjusted for age, surgical level (L3-4, L4-5, L5-S1), accompanying spondylolisthesis, and BMD (T-score). Statistical significance was defined as p<0.05.
RESULTS
Baseline characteristics
A total of 115 patients were included in this study. The FE-TLIF group consisted of 58 patients (25 men and 33 women) with a mean age of 72.4±7.3 years (range, 54-91). The PLIF group consisted of 57 patients (23 men and 34 women) with a mean age of 69.1±10.7 years (range, 49-85). No significant differences were observed between the two groups in terms of age, sex, body mass index, smoking status, surgical level, accompanying spondylolisthesis, BMD, or follow-up duration (Table 1). All patients underwent single-level surgery.
Perioperative outcomes
The operative time and estimated blood loss were significantly lower in the FE-TLIF group than in the PLIF group. The distribution of the cage types (PEEK or 3D-printed titanium) showed no significant differences between the groups. Although the cage length was 26 mm in both groups, the cage height and angle were slightly but significantly greater in the FE-TLIF group (height, 11-15 mm; angle, 4°-8°) than in the PLIF group (height, 7-13 mm; angle, 0°-8°). Intraoperative cage malposition requiring readjustment occurred in two patients in the FE-TLIF group, while none were observed in the PLIF group; however, this difference was not statistically significant (Table 2).
Clinical outcomes
At baseline, no significant differences were observed between the two groups in the VAS scores for lower back and leg pain or ODI scores. At 1 week and 1 month postoperatively, the VAS scores for lower back pain were significantly lower in the FE-TLIF group than in the PLIF group. At 1 week, the VAS scores for leg pain and ODI scores tended to be lower in the FE-TLIF group; however, the differences were not statistically significant. At 1 month, the VAS scores for leg pain were comparable between the groups, and the ODI score showed a slightly greater improvement in the FE-TLIF group; however, the difference was not statistically significant. At 6 and 12 months postoperatively, no significant differences were observed between the groups in the VAS scores for lower back or leg pain, ODI scores, or MacNab’s criteria. Both groups demonstrated clinically significant improvements in all outcome measures (Table 3).
The degree of improvement from baseline in VAS scores for lower back and leg pain, as well as ODI, was significantly greater in the FE-TLIF group at 1 week postoperatively. At 1 month, the FE-TLIF group still showed a tendency toward greater improvement in these parameters, although the differences were not statistically significant. The remaining results showed trends similar to the absolute score comparisons, and the minimal clinically important difference values were comparable between the two groups (Table 4).
Radiological outcomes and complications
At the 12-month follow-up, fusion rates were comparable between the two groups. After adjusting for age, surgical level, the presence of low-grade spondylolisthesis, and BMD, the fusion statuses remained statistically similar. Similarly, the incidence of cage subsidence >2 mm showed no significant difference between the groups, even after multivariate adjustment. Both procedures achieved significant improvements in the SLA, TLA, and DH at 12 months. The improvement in DH was slightly greater after FE-TLIF, suggesting a modest but statistically significant advantage for DH restoration and maintenance.
Postoperative complications were rare and comparable between the groups. Epidural hematoma and infection requiring revision occurred only in the PLIF group; no such cases were observed after FE-TLIF. Dural tears were managed conservatively in all FE-TLIF cases and required reoperation in one case. The overall revision rate tended to be higher in the PLIF group, although the difference was not significant (Table 5).
Illustrative case 1 for double-cage FE-TLIF
A 79-year-old male patient presented with a bilateral leg tingling sensation and motor weakness. Magnetic resonance imaging (MRI) revealed severe stenosis and degenerative changes at the L4-5 level. The patient underwent double-cage FE-TLIF at L4-5 using a 3D-printed titanium cage. Postoperative MRI confirmed adequate central decompression, and follow-up radiographs at 12 months demonstrated improvements in TLA, SLA, and mean DH compared with the preoperative values. The patient’s primary symptoms improved significantly, and no major complications were observed (Fig. 3).
Illustrative case 2 for double-cage FE-TLIF
A 60-year-old female patient presented with weakness in left dorsiflexion and radiating bilateral leg pain. MRI revealed foraminal stenosis at the L5-S1 level. The patient underwent double-cage FE-TLIF at L5-S1 using PEEK cages. Postoperative MRI confirmed adequate neural decompression at L5-S1. Follow-up radiographs at 12 months demonstrated improvements in TLA, SLA, and mean DH compared to the preoperative measurements. The patient showed marked improvement in left dorsiflexion and bilateral leg pain, and no major complications occurred (Fig. 4).
DISCUSSION
PLIF has long been established as a reliable surgical technique with numerous studies reporting significant postoperative improvements in clinical outcomes [14,28]. Radiological outcomes have also been reported to be favorable, with consistently high fusion rates and acceptable levels of cage subsidence [6,20,26]. Yang et al. [36] compared the clinical and radiological outcomes of PEEK and 3D-printed titanium cages in PLIF and demonstrated excellent results in both groups. These findings collectively support the idea that PLIF provides stable and reliable fusion, largely independent of the cage material type. With the increasing preference for minimally invasive spinal surgery among patients and surgeons, endoscopy-assisted TLIF techniques have continued to evolve and gain clinical relevance. Early endoscopic lumbar fusion techniques introduced the trans-Kambin approach, which preserves the facet joint during fusion. This technique, known as trans-Kambin endoscopic fusion, was later renamed full-endoscopic facet-sparing TLIF (FE fs-TLIF) [4,11,24]. This approach allows direct decompression of the ipsilateral ENR, but provides only indirect decompression of the central traversing nerve root. With the development of larger and shorter endoscopic systems, a new posterolateral approach called FE facet-sacrificing TLIF was introduced [17]. This approach, also known as FE facet-resecting TLIF (FE fr-TLIF), enables direct decompression of the ipsilateral ENR and the central traversing nerve root [4]. Unlike FE fs-TLIF, which accommodates only a long single cage, FE fr-TLIF permits the insertion of two cages, similar to those used in PLIF or minimally invasive TLIF procedures [17,30,32]. Although both techniques yield excellent clinical and radiological outcomes [13,27,33-35], FE fr-TLIF is considered a more advanced technique because it provides a wider surgical field, allows the insertion of double cages for improved interbody support, offers greater biomechanical stability, and enables direct central canal decompression [12,17,19,23,33]. In addition, according to previous studies, the double-cage configuration provides a broader endplate contact surface and improved load distribution, which may offer potential radiological stability advantages over a single-cage design [16,18,25]. For these reasons, all patients in the FE-TLIF group in our study underwent the double-cage FE fr-TLIF technique, which allows direct decompression of both the central traversing nerve root and ipsilateral ENR as well as the insertion of two interbody cages. However, double-cage insertion in a single direction may be technically demanding and potentially associated with increased risk at high lumbar levels, such as L1-2 or L2-3. Accordingly, in our institution, double-cage FE-TLIF was selectively applied to mid- to lower-lumbar segments (L3-S1), while single-cage techniques were preferred for upper lumbar levels.
In our study, no significant differences were observed in baseline characteristics between the two groups. Regarding perioperative outcomes, the operation time and estimated blood loss were significantly lower in the FE-TLIF group than in the PLIF group. This difference is likely attributable to the smaller surgical corridor used in FE-TLIF, which results in less soft tissue injury and shorter duration of hemostasis and wound closure. In all FE-TLIF cases, the fr-TLIF technique was applied, and the same 26-mm cage length was used as in the PLIF group; therefore, no comparison based on cage length was made. However, the cage height and angle were slightly greater in the FE-TLIF group than in the PLIF group. In PLIF, annulotomy and discectomy are performed adjacent to the dural sac, and cage insertion requires simultaneous retraction of both the dural sac and ENR by an assistant. This setup introduces variability depending on the assistant’s technique and may limit the use of larger cages. Excessive retraction may also increase the risk of nerve injury and postoperative dysesthesia. In contrast, during FE-TLIF, the surgeon inserts a two-tip glider in a cranial-caudal orientation and then rotates it by 90° in the medial-lateral direction, allowing safe docking into the interbody disc space while protecting both the traversing nerve root and ENR. The first cage is advanced through the glider into the disc space, during which the glider expands slightly, just enough to provide minimal controlled retraction of the neural elements. As a result, cage insertion can be performed safely and efficiently without significant assistance, permitting the use of cages with greater heights and lordotic angles than in PLIF. This safe insertion maneuver was recently described as the GUARD (Glider Used As a Rotary Device) technique [4,10,18]. Intraoperative cage malposition occurred in two patients in the FE-TLIF group and in none of the patients in the PLIF group. Both incidents involved lateral insertion of the second cage. One case was corrected endoscopically, whereas the other required conversion to an open procedure for removal and reinsertion. Although the difference between the groups was not statistically significant due to the small number of cases, such malpositions can likely be prevented through careful surgical manipulation.
Regarding clinical outcomes, the FE-TLIF group demonstrated significantly lower VAS scores for back pain at 1 week and 1 month postoperatively. The length of hospital stay was significantly shorter in the FE-TLIF group. Moreover, 1 week postoperatively, the improvement in the VAS (back and leg) and ODI scores from baseline was significantly greater in the FE-TLIF group. These findings suggest that double-cage FE-TLIF offers advantages in early postoperative recovery compared to PLIF. The smaller incision and reduced soft tissue and muscle trauma in FE-TLIF likely contributed to faster pain relief and earlier functional recovery. However, with longer follow-up periods (6-12 months), tissue healing progressed in both groups, and the differences in clinical outcomes gradually diminished over time.
Regarding radiological outcomes, both the fusion rate and the incidence of cage subsidence >2 mm were slightly higher in the PLIF group than in the FE-TLIF group; however, the differences were not statistically significant even after adjustment. This finding suggests that FE-TLIF achieves radiological stability comparable to that of PLIF in terms of fusion and significant subsidence. Other radiological parameters, including SLA, TLA, and mean DH showed significant postoperative improvement in both groups. Although improvements in SLA and TLA were not significantly different between the groups, the increase in mean DH was slightly but significantly greater in the FE-TLIF group. As discussed in the perioperative outcomes section, this may be attributed to the FE-TLIF technique in which the two-tip glider allows for safe and efficient cage insertion with adequate neural protection and minimal retraction, enabling the use of larger and more lordotic cages.
No significant difference was observed in complications between the two groups; however, more complications were observed in the PLIF group. In the PLIF group, a total of nine complications were recorded, including two cases each of epidural hematoma and infection that required revision surgery, and five dural tears, among which five cases required reoperation. In contrast, the FE-TLIF group had no cases of epidural hematoma or infection requiring revision, and three dural tears, all of which were successfully managed conservatively without reoperation. There was one case of simultaneous cage subsidence and retropulsion, and this was the only case that required reoperation. At revision, insufficient interbody discectomy and endplate preparation were identified; after thorough discectomy and meticulous endplate preparation, the revision was completed successfully. Although nerve root herniation following a dural tear during endoscopic surgery can potentially cause severe neurological deficits, previous reports have shown that precise sealing with TachoSil (a collagen fleece) can allow recovery without open revision [15]. Similarly, in the present study, all three cases of dural tears in the FE-TLIF group were successfully treated with this technique and 5-7 days of absolute bed rest without the need for additional surgery.
However, this study had several limitations. First, this was a retrospective study conducted at a single institution, which may have introduced selection bias and limited the generalizability of the results. Second, although the baseline characteristics between the groups were comparable, potential confounding factors inherent to retrospective data collection could not be completely controlled. Third, the follow-up period was limited to 12 months, providing only short-term outcomes; therefore, the findings should be interpreted as preliminary. Fourth, because double-cage FE-TLIF was applied only at L3-S1 levels, the results may not be generalizable to upper lumbar segments. Long-term prospective studies with larger sample sizes and multicenter participation are warranted to confirm the durability and clinical relevance of double-cage full endoscopic TLIF compared to PLIF.
CONCLUSION
In this retrospective cohort study, double-cage FE-TLIF showed a shorter operative time, less blood loss, and fusion and subsidence outcomes comparable to those of PLIF. Early postoperative back pain improved more rapidly after FE-TLIF, although the long-term clinical and radiological results were similar. These findings indicate that double-cage FE-TLIF is a minimally invasive and biomechanically stable alternative to conventional PLIF for treating single-level degenerative lumbar diseases.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
