INTRODUCTION
Spine surgery has transformed over the last decade due to advancements in approach to spine and technology, which gives patients confidence in treatment and outcome. Minimal invasive approach is often perceived as superior and easily marketable due to small incision, fast recovery, and less pain [1]. Posterior lumbar interbody fusion (PLIF) procedure, from open surgery to minimally invasive surgery, started in the early 19th and 20th centuries, which is performed through a 2-cm incision and microscope. This technique was later popularized by Williams in 1978 [2].
Minimally invasive transforaminal lumbar interbody fusion (MIS-TLIF) has shown equivocal results with the conventional open approach in terms of patient-reported clinical outcomes and surgical complications. In contrast, other studies also indicate MIS-TLIF to be superior for lesser back pain, societal cost, and early return to work at the cost of higher revision and readmission rates [3]. Some studies do not advocate MIS over open surgery for lumbar disc diseases, which makes this study relevant for further research [4]. Among the other advantages are faster recovery, less pain, and shorter hospital stay, which motivates patients for early return to a pain-free daily life.
In this study, we have evaluated the functional and radiological parameters of MIS-TLIF surgery operated for degenerative lumbar diseases with a long-term follow-up of minimum 5-year. Our secondary aim was to analyse the long-term complications in patients operated on with single or double-level MIS-TLIF surgeries.
MATERIALS AND METHODS
1. Study Design
This study was conducted following due approval from the Scientifc and Ethics committee. A detailed informed consent was taken from all the patients included in the study. This was a retrospective analytical study conducted in 180 patients who were operated on between March 2009 and February 2014 and were followed up until February 2024 with a minimum follow-up of 5 years. All patients were operated on for MIS-TLIF with percutaneous pedicle screw fixation at a single center by a single surgeon. Patients aged more than 18 years, spinal stenosis, spondylolisthesis, degenerative disc disease from L1 to S1 with single- or double-level involvement, and a minimum follow-up of 5 years were included in the study. Patients with infection, trauma, tumour, more than 2-level involvement, previously operated, and irregular follow-up were excluded from the study. Among 180 consecutive patients who underwent single- or double-level MIS-TLIF surgery, 111 patients were followed up for a minimum of 5 years or more, and 94 patients were followed up for a minimum of 10 years or more.
2. Surgical Procedure
Under general anaesthesia, patients were positioned prone on a radiolucent table. Real-time fluoroscopic guidance was used to identify anatomical landmarks such as pedicles and spinous processes. A 2- to 3-cm long incision was made approximately 4–5 cm lateral to the midline on the affected side. However, in some cases, bilaterally 2- to 3-cm long incisions were kept instead of 2 separate holes for the pedicle screws, especially at L5–S1 levels, where 2 pedicle screws were easier to insert from the same incision. If bilateral decompression was needed, it was done unilaterally with the top technique. A probe was inserted to locate the facet joints targeted for fusion under fluoroscopic guidance. Sequential dilators (METRx, Medtronic, USA) were used to create a minimally invasive path, and a tubular retractor was placed over the facet joint to access the decompression. Only the affected facet unilaterally was removed using a high-speed burr, Kerrison punch, and L-shaped osteotome. After removal of the thickened ligamentum flavum to decompress the dural sac and nerve root, discectomy and end plate preparation were done. Standard 22-mm cage (Solco-Sara) was used, with bone graft collected during surgery used for interbody fusion from the same incision. Contralateral nerve root decompression was achieved from the same incision by tilting the tubular retractor under fluoroscopy guidance by unilateral laminectomy and bilateral decompression method when required. After satisfactory decompression and fusion tubular retractor was removed.
Percutaneous pedicle screws were introduced under fluoroscopic guidance via guide wires inserted through Jamshidi needles. Ipsilateral side same skin incision was used for screw insertion on the same side; however, 2 separate 8-mm incisions were used on the contralateral side for screw and rod insertion. Later, wound and skin were closed using Vicryl subcuticular sutures (Figure 1).
3. Radiological Outcome
All the patients were evaluated on the basis of x-rays taken preoperative, postoperatively, and subsequent follow-up. Lumbar lordosis (LL) and pelvic incidence (PI), disc height, and foraminal height were measured in the lumbar lateral radiograph. Intervertebral fusion was measured using a sequential lateral radiograph [5]. The Bridwell interbody fusion grading system was used on plain radiographs. Grade 1: fused with bony bridging, presence of trabecular remodelling; grade 2: not fully bony bridged, remodelling but without radiolucency above or below the cage; grade 3: radiolucency present at the top or bottom of the cage and screw; grade 4: fusion absent, with false motion. In this series, Bridwell grades 1 and 2 were defined as fused, and Bridwell grades 3 and 4 as nonfused.
4. Clinical Outcome
In this study, demographic information, age, sex, past operative history, follow-up period, and level of surgery were recorded. Operation time, estimated blood loss, and hospital stay parameters were observed. Different complications and revisions required were noted. Clinical outcome was measured using the visual analogue score (VAS), Oswestry Disability Index (ODI) score was recorded.
Clinical and radiological parameters were assessed preoperatively, postoperatively at stitch removal, and 5-year and 10-year follow-ups (Figure 2). All measurements were performed using Picture archiving and communication system (PACS) software on dedicated PACS workstations. Disc height was evaluated on plain lateral radiographs, along with PI at the operated segment and overall LL.
Radiological adjacent segment degeneration (ASD) on follow-up was analysed using plain x-rays. It was considered positive if postoperative comparisons with preoperative radiographs revealed a reduction in disc height of ≥3 mm on a neutral lateral x-ray, vertebral slip increase of ≥3 mm, or a posterior opening angle increase of ≥ 5° on a flexion lateral x-ray.
Data analysis was conducted using SPSS ver. 12.0 (SPSS Inc., USA). A p-value of 0.05 was considered the threshold for statistical significance. Interobserver agreement of all analyses was done by 2 independent spine surgeons.
RESULTS
A total of 180 patients underwent transforaminal lumbar interbody fusion (TLIF), including 94 single-level and 86 double-level procedures. The mean age of the cohort was 53.3±10 years, with no significant difference between single-level (52.1±10.0 years) and double-level (53.6±11.0 years) groups (p=0.35). Gender distribution was comparable (single‐level: 60 male/34 female; double‐level: 58 male/28 female).
Mean follow-up durations did not differ significantly between groups at 5 years (104.6±34.0 months vs. 105.9±35.0 months, p=0.80), or 10 years (140±30 months vs. 142±30 months, p=0.66).
In the single-level group, L4–5 (n=41) and L5–S1 (n=53) level were fused, whereas the double-level group underwent either L3–4 & L4–5 (n=41) or L4–5 & L5–S1 (n=45) level fusion.
Operative time was significantly shorter for single‐level TLIF (126.0±10.6 minutes) compared to double‐level (156.0±24.0 minutes, p=0.01). Estimated blood loss was lower in single‐level cases (111.0±19.5 mL vs. 180.0±220 mL, p=0.05), and hospital stay was reduced (3.6±0.5 days vs. 4.8±0.7 days, p=0.04).
Overall postoperative complication rates were low and similar between groups (single: 5 [2.8%] vs. double: 4 [2.2%], p=0.99). Specific complications included cage back-out (n=3), hematoma (n=2), infection (n=1), root sleeve injury (n=1), screw breakage (n=1), and screw loosening (n=1). Revision surgery was required in 5 patients (2.8%). Adjacent-level disease occurred in 16 patients (8.8%), with no significant difference between single-level (5.0%) and double-level (3.9%) groups (p=0.80) (Table 1).
Among 16 patients (23 adjacent-level segments observed), of the 180 patients initially enrolled, 111 (61.7%) completed 5-year radiographic follow-up, and 94 (52.2%) completed 10-year follow-up. At 5 years, ASD was identified in 6 patients (2 single-level [S] and 4 double-level [D] fusions), rising to 10 patients (7 S, 3 D) at 10 years. Radiographically, disc height reduction was seen in 2 proximal and 2 distal segments at 5 years versus 1 proximal and 4 distal segments at 10 years (p=0.98). Posterior angle increases >5° occurred in 1 proximal and 1 distal segment at 5 years compared with 2 proximal and 1 distal segment at 10 years (p=0.55); and vertebral slip >3 mm was noted in 2 proximal and 2 distal segments at 5 years and in 2 proximal and 3 distal segments at 10 years (p=1.00). Overall, radiographic ASD affected 10 segments at 5 years (5 proximal, 5 distal) and 13 segments at 10 years (5 proximal, 8 distal) - corresponding to event rates of 5.4% and 14.5%, respectively (p=0.09). Segment reoperations for ASD were performed in 4 segments (3 proximal, 1 distal) by 5 years and in 7 segments (4 proximal, 3 distal) by 10 years (p=0.08) (Table 2).
PI remained essentially unchanged over time, averaging 51.85°±4.43° preoperatively, 51°±5.8° immediately postoperative, 51.23°±6.35° at 5 years (p=0.371) and 51.04°±6.32° at 10 years (p=0.401), with no significant difference between the 5- and 10-year values (p=0.98). LL likewise showed minimal variation, from 51.12°±9.53° before surgery to 50.0°±6.7° after surgery and then measuring 50.92±8.08° at 5 years (p=0.490) and 51.27°±8.29° at 10 years (p=0.434), again with no difference between follow-up intervals (p=0.20). Patients experienced dramatic pain relief, with mean VAS scores falling from 7.79±0.78 preoperatively to 3.5±0.79 postoperatively, then to 1.40±0.51 at 5 years and 1.67±0.87 at 10 years (both p<0.05), and a small but not significant change between 5 and 10 years (p=0.14). Functional disability (ODI) improved from 48.52%±5.86% before surgery to 23.0%±5.6% immediately after, then to 20.80%±2.97% at 5 years and 20.63%±3.09% at 10 years (both p<0.001), with no significant change between the 2 follow-ups (p=0.53). Disc height increased from 8.4±1.3 mm preoperation to 9.1±1.2 mm postoperation, and then measured 9.2±1.2 mm at 5 years and 9.6±1.2 mm at 10 years (p=0.1), without a significant difference between 5- and 10-year measurements (p=0.15) (Table 3).
At the 5-year follow-up (n=111; complications=8, noncomplications=103) and the 10-year follow-up (n=94; complications=9, noncomplications=85), we compared spinopelvic alignment (PI–LL), pain reduction (ΔVAS), and disability (ODI) between patients with and without complications using Welch t-tests. The heatmap of p-values (Figure 3) shows no significant differences in PI–LL or lordosis at either time point (all p>0.1). Pain reduction remained directionally worse in the complication group—ΔVAS was higher (indicating less improvement) at both 5 years (p=0.282 ↑) and 10 years (p=0.328 ↑)—though neither reached statistical significance. Disability scores were also numerically greater in the complication cohort at 5 years (p=0.403 ↑) and 10 years (p=0.170 ↑) without significant differences. These results suggest that early postoperative complications exert a durable negative effect on long-term pain trajectories, despite minimal impact on structural alignment or functional disability.
Radiological fusion was assessed using the Bridwell grading system at 5-year and 10-year follow-up intervals. At the 5-year follow-up (n=111), Grade 1 fusion was observed in 86.5% of patients, while grade 2 fusion was seen in 9.9%, resulting in an overall fusion rate (grade 1 + 2) of 96.4%. Only a small proportion of patients demonstrated grade 3 (2.7%) or grade 4 (0.9%) changes. By the 10-year follow-up (n=69), the proportion of patients with grade 1 fusion increased to 94.2%, with grade 2 observed in 4.35%, yielding an overall fusion rate of 98.6% (Figure 4). Grades 3 and 4 remained infrequent at 1.45% each. These findings indicate a progressive improvement in solid bony fusion over time, with sustained long-term outcomes and a very low incidence of nonunion. Statistical analysis using the chi-square test confirmed a significant shift in fusion grades over time (chi-square=50.98, p<0.0001). Pairwise post hoc tests revealed significant differences between each time point (p<0.01), supporting the conclusion that fusion quality improves with longer follow-up. These findings demonstrate the long-term efficacy of the surgical procedure in achieving solid bony fusion (Table 4). In our study, at 10-year follow-up cage subsidence was found in 9 patients (5%) with no significant complaint.
DISCUSSION
Minimally invasive spine surgery was introduced in the late 20th century using retractors, microscopes, and image intensifiers [6]. Since then, the technique has evolved significantly through the development of various approaches. In 2002, Foley and Lefkowitz [7] detailed their adaptation of the TLIF technique, employing tissue-sparing mechanisms. Their method incorporated 1-inch paramedian incisions for tubular retractor insertion and percutaneous pedicle screw fixation.
As global life expectancy rises, the prevalence of degenerative lumbar spine conditions similarly increases, driving greater demand for effective and durable fusion techniques [8,9]. Minimally invasive transforaminal lumbar interbody fusion (MIS-TLIF) was developed to minimize soft-tissue disruption, reduce perioperative morbidity, and accelerate recovery compared to open procedures [10-12]. Although numerous studies have documented favorable short- and mid-term outcomes for MIS-TLIF, data on its long-term efficacy—particularly over a decade—remain scarce [13,14]. Moreover, concerns persist regarding adjacent segment disease, cage subsidence, and the need for revision surgeries, all of which may impact sustained clinical benefit [15-17]. Our study provides a comprehensive 10-year analysis of 180 patients undergoing single- or double-level MIS TLIF, focusing on fusion rates, clinical outcomes, radiographic changes, cage subsidence, ASD incidence, and revision procedures.
Sustained reductions in pain and disability are paramount for evaluating fusion success. Our cohort demonstrated marked improvement in VAS back pain (7.8±0.8 to 1.7±0.9) and ODI (48.5±5.9 to 20.6±3.1) over 10 years (p<0.001). These results parallel long-term studies: Park et al. [18] reported stable VAS/ODI improvements at 5 years, and Rouben et al. [19] at 49 months, and extend findings by confirming maintenance of clinical benefit at a decade.
In our cohort, the 5-year fusion rate was 96.4%, reaching 98.6% at 10 years. These results underscore the reliability of MIS-TLIF in achieving solid arthrodesis, relying on local autograft from the same surgery. Our 10-year fusion rates align with or exceed those in Park et al. [18] (97.7% at 5 years) and Hong et al. [20] (85.5% at 5 years in MIS TLIF group).
ASD remains a key concern after fusion surgery. In our 10-year follow-up, radiographic ASD (R-ASDeg) appeared in 8.6% (23 segments) of segments, while symptomatic ASD requiring intervention (O-ASD) occurred in 4.1% (11 segments). Comparatively, open PLIF studies report higher R-ASDeg rates of 68%–75% and O-ASD rates of 9.9%–15% at similar follow-up intervals [21-23]. The lower O-ASD rate in our MIS-TLIF cohort may reflect preservation of paraspinal musculature and ligamentous integrity, which reduces hypermobility at adjacent levels [24]. Biomechanical studies demonstrate that tubular retractor-based approaches preserve facet and soft-tissue attachments, minimizing stress transfer to neighboring segments [25].
Cage subsidence—a potential cause of loss of disc height and neural foraminal narrowing—occurred in 5 percent of patients by 10 years. These subsidence events were mild (<2-mm endplate breach) and did not significantly worsen VAS or ODI scores, suggesting that limited subsidence may be clinically tolerable if initial placement achieves adequate lordosis and endplate contact [26,27]. Our findings concur with Oh et al. [28] (subsidence > 3 mm without clinical sequelae), emphasizing meticulous endplate preparation and use of appropriately sized cages to mitigate subsidence risk.
At 10 years postoperatively, patients who experienced early complications continue to demonstrate significantly less pain relief (ΔVAS) compared to those without complications, whereas spinopelvic alignment (PI–LL, lordosis) and disability (ODI) measures remain statistically equivalent between groups. This sustained disparity in pain recovery underscores the durable impact of perioperative insults—such as infection, hematoma, or hardware irritation—on the central sensitization and neuroinflammatory processes that contribute to chronic back pain decades after surgery [29]. Notably, the absence of structural divergence at 10 years suggests that routine radiographic parameters may fail to capture the full spectrum of complication-related morbidity, reinforcing pain-focused assessments as critical for long-term monitoring [30]. These results align with long-term cohort studies showing that initial complication status is a stronger predictor of pain trajectories than baseline deformity or postoperative alignment [31]. Accordingly, optimizing perioperative complication avoidance and implementing early, targeted pain-modulating interventions should remain priorities to improve not only short-term recovery but also the enduring quality of life for spinal fusion patients.
Our 10-year data validate MIS-TLIF as a durable fusion technique with high arthrodesis rates, sustained pain relief, and low revision requirements. Strategies to minimize ASD, such as preserving posterior tension bands and optimizing sagittal alignment-should be integrated into patient selection and surgical planning. Emerging technologies, including robotic assistance, augmented reality-guided cage placement, and expandable interbody devices, warrant investigation for further reducing complications and improving precision.
Recent publications continue to expand on these findings. A 2024 randomized trial comparing MIS-TLIF and modified open TLIF showed superior preservation of paraspinal muscle integrity and reduced blood loss in the MIS group, with no compromise in fusion rates [32]. Another 2024 meta-analysis found that while endoscopic TLIF may reduce blood loss further, MIS-TLIF had a lower complication rate and similar functional outcomes [33]. Additionally, in cases of isthmic spondylolisthesis, MIS-TLIF resulted in faster early recovery and higher patient satisfaction compared to open TLIF [34]. These cumulative findings further validate our results and affirm that MIS-TLIF remains a front-line surgical option in lumbar degenerative disease.
Despite its strengths, our study has limitations. Its retrospective design and single-center setting may introduce selection bias and limit external validity. Surgical techniques and instrumentation have evolved over the past 10 years, potentially affecting outcomes. Future prospective, multicenter randomized trials are needed to corroborate these findings and explore cost-effectiveness and patient-reported metrics in greater depth.
One of the shortcomings of this study was that we treated those cases with MIS-TLIF that had more symptoms on one side than the other and 1- or 2-level involvement. On the other hand, our open TLIF cases had more than 2 levels of involvement, and symptoms were bilaterally equal in both lower extremities, which required bilateral facetectomies for the decompression. Additionally, open TLIF cases had more severe canal narrowing or near-zero spinal canal diameter at the affected levels. Therefore, we believe the comparison between open and MIS-TLIF cases in our series would not be fair and give a correct picture of the surgical outcome. Instead, we hope that our case series would add value to the long-term results of MIS-TLIF, with a minimum follow-up of 5 years, and would further add value for future studies.
CONCLUSION
Our single- and 2-level MIS-TLIF cohort of 180 patients shows that minimally invasive techniques can achieve and maintain high fusion rates (98.6% at 10 years) while providing lasting pain relief (mean VAS reduction from 7.8 to 1.7) and improved function (ODI from 48.5 to 20.6). Rates of radiographic ASD (8.6%), symptomatic ASD (4.1%), and cage subsidence (5%) remained low and mostly without clinical impact, with minimal overall revision needs. These results highlight the benefit of muscle-sparing, tubular retractor-based approaches in preserving paraspinal health and reducing stress on neighboring segments. Despite the limitations of a retrospective, single-center study, our data support MIS-TLIF as a durable, low-morbidity fusion method for lumbar degenerative disease. Moving forward, prospective multicenter trials and new technologies—such as robotic guidance and expandable interbody devices—should be pursued to improve outcomes and enhance patient selection.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
