High Fusion Rates with Structured Titanium TLIF Cages: A Retrospective 1-Year Study with and Without Adjacent Level Dynamic Stabilization

Abstract

Background: Structured titanium (ST) cages are designed to enhance osseointegration and fusion in lumbar interbody procedures. However, clinical and radiological outcomes following TLIF using ST cages—particularly with or without adjacent-level dynamic stabilization (DSS)—have not been widely reported. Objective: To evaluate 12-month fusion outcomes and patient-reported outcomes (PROMs) after TLIF with structured titanium cages, comparing cases with and without adjacent-level DSS. Methods: In this retrospective cohort study, 82 patients undergoing TLIF with ST cages were analyzed—41 with hybrid instrumentation (TLIF + DSS) and 41 with TLIF alone. PROMs (ODI, VAS for back and leg pain, EQ-5D-5L) were assessed preoperatively and at 12 months. Fusion was assessed via CT scans at 12 months. Results: PROMs significantly improved over time in both groups (p < 0.001 for ODI, VAS back, VAS leg), but there were no significant differences between the hybrid and non-hybrid groups. Overall, the interbody fusion rate was 84%. Complete fusion was observed in 84% of the hybrid group and 80% of the TLIF-only group (p = 0.716), with very low rates of non-union. Conclusions: Structured titanium cages demonstrated excellent 1-year fusion rates and supported significant clinical improvement after TLIF. The addition of dynamic stabilization had no measurable effect on patient-reported or radiological outcomes at 12 months. Long-term studies are needed to assess any potential effect of DSS on adjacent segment disease.

1. Introduction

Spinal fusion surgery for the treatment of degenerative spinal diseases has advanced significantly over the past several decades [1]. Initially, posterior spinal fusion was performed using non-instrumented posterolateral techniques [2]. Today, interbody fusion combined with pedicle screw and rod fixation represents the standard of care. Harms and Jeszensky introduced the transforaminal lumbar interbody fusion (TLIF) technique in the early 1990s as an evolution of the classic posterior lumbar interbody fusion (PLIF) [3]. TLIF enables interbody fusion via a posterior route without the need for anterior access and facilitates nerve root decompression by partially resecting the articular processes and widening the intervertebral foramen [3]. However, recent systematic reviews have reported fusion rates ranging from 73% to 94% [4,5].

Modern interbody implants consist of a highly biocompatible titanium alloy (TiAl6V4) and are often filled with autologous or allogenic bone graft material. Structured titanium (ST) cages characterized by high porosity (up to 70%) and textured surfaces have demonstrated improved osseointegration and fusion potential in preclinical and early clinical studies [6,7,8]. Several clinical studies suggest that porous titanium implants may reduce subsidence and enhance early fusion compared to PEEK cages, particularly in posterior lumbar fusion techniques [9,10]. Recently, topping-off techniques—defined as combining rigid fusion with adjacent-level hybrid or dynamic stabilization—have gained attention as a potential strategy to prevent adjacent segment disease (ASD), a common long-term complication of lumbar fusion [11]. However, the application of ST cages within hybrid constructs that incorporate adjacent-level dynamic stabilization systems (DSS) in TLIF procedures has not been thoroughly investigated.

To the best of our knowledge, this is the first study to compare clinical and radiographic outcomes of a hybrid TLIF technique employing high-porosity structured titanium cages with adjacent-level dynamic stabilization to those of conventional TLIF. This study aims to evaluate whether the addition of dynamic stabilization confers any short-term clinical or radiological advantages at 12 months postoperatively.

2. Materials and Methods

We conducted a retrospective analysis of demographic, perioperative, and patient-reported outcomes in individuals aged 18 years or older who underwent single- or multi-level TLIF at a single center between 2019 and 2021.

2.1. Surgical Technique

All patients underwent minimally invasive TLIF in the prone position using a Wilson frame. Percutaneous pedicle screws were inserted under intraoperative computed tomography (CT)-guided navigation (AIRO, Brainlab AG, Feldkirchen, Germany). The fixation was either non-hybrid—with polyaxial pedicle screws (Everest Spinal System, K2M, Inc. or DIPLOMAT^®, Signus, Medizintechnik GmbH, Alzenau, Germany) and a titanium alloy (Ti6Al4V) cage (MOBIS^® II ST Spinal Implant, SIGNUS Medizintechnik GmbH, Alzenau, Germany; Figure 1A) filled with synthetic bone graft substitute—or hybrid—combining standard instrumentation with DSS. In the hybrid group, polyaxial pedicle screws were used alongside a topping-off system (Hybrid Performance System HPS™ 2.0, RTI Surgical, Alachua, FL, USA) at the adjacent segment. Care was taken to place the pedicle screws in parallel to ensure correct alignment of the dynamic stabilizer. The surgical procedure involved osteotomy of the facet joint and removal of the superior articular process, followed by discectomy and endplate preparation. The titanium cage was then packed with a synthetic bone graft substitute and inserted into the intervertebral space.

2.2. Data Collection

Data were collected through the NeuroSpine Institute SPINE Registry, a prospective observational cohort registry that collects demographic, perioperative, and patient-reported outcome measures (PROMs) from participants who were >18 years of age and undergoing surgical treatment for spine-related diseases from four surgeons (G.C., P.T., M.K., and A.M.) across three sites in Perth, Western Australia (St John of God [SJOG] Hospital Murdoch, SJOG Subiaco, and Mount Hospital). The registry was approved by the SJOG Murdoch Human Research Ethics Committee (HREC 1776) and validated by the respective ethics boards at SJOG Subiaco and Mount Hospital. All patients provided written or electronic informed consent. Data included demographic characteristics (age, sex, body mass index (BMI), and smoking status) and patient-reported outcome measures (PROMs). Perioperative data included the surgical indication, neurological deficit, the number of operated levels, surgery duration, estimated blood loss, intra-operative complications, and postoperative surgical complications. PROMs included the Oswestry Disability Index (ODI), the Visual Analogue Scale (VAS) for back and leg pain, and the EQ-5D-5L for health-related quality of life (HRQoL). Assessments were performed preoperatively and at 26 and 52 weeks postoperatively.

A total of 82 participants were included in the study. The mean age was 62.1 years and 57% were female. Of these, 41 patients underwent TLIF alone and 41 received hybrid instrumentation with TLIF and adjacent-level dynamic stabilization (Figure 2).

2.3. Radiographic Outcomes:

All patients underwent routine CT scans 12 months postoperatively. Fusion rates were assessed using thin-slice (≤1 mm) sagittal and coronal reconstructions from 12-month postoperative CT scans. Two independent reviewers (a board-certified radiologist and a fellowship-trained spine surgeon) evaluated the scans for continuous trabecular bone bridging across the intervertebral space. Complete fusion was defined as solid bony bridging with no lucency at the implant interface; partial fusion as incomplete bridging with visible bone formation; and non-union as the absence of trabecular continuity [12]. Discrepancies were resolved by consensus.

2.4. Statistical Analyses

Statistical analyses were performed using the stats package in the R environment using the “Cherry Blossom” Release (3c53477a, 9 March 2023) for Windows. Unless otherwise indicated, descriptive statistics are reported as mean (standard deviation, SD). Group comparisons for demographic and perioperative variables were conducted using Fisher′s Exact, chi-squared, t-tests, and Wilcoxon rank-sum tests as appropriate. Patient-reported outcomes were analyzed using mixed two-way ANOVA to assess the interaction between follow-up timepoints and surgical technique (TLIF with or without DSS). Assumptions of data normality, homogeneity of variances, homogeneity of covariances, and sphericity were confirmed using Shapiro−Wilk, Levene’s, Box’s M, and Mauchly’s tests, respectively. All outliers were retained in the analysis. Holm’s correction was applied to pairwise comparisons of significant main effects. Fusion rates were compared using Fisher’s exact test.

3. Results

3.1. Surgical Characteristics and Complications in Non-Hybrid Versus Hybrid Stabilization

A statistically significant difference in surgical indications was observed between the non-hybrid and hybrid groups (p = 0.04). Radiculopathy was more prevalent in the non-hybrid group (73%) compared to the hybrid group (46%). Conversely, claudication was more frequent in the hybrid group (37%) than in the non-hybrid group (15%). Low back pain was also more commonly reported in the hybrid group (15%) compared to the non-hybrid group (7%). Primary surgery was more common in the hybrid group (90%) than in the non-hybrid group (71%). Revision procedures at the same level (following prior decompression or nucleotomy) were performed in 27% of non-hybrid cases versus 7% in the hybrid group. The median operative time was significantly longer in the hybrid group, by approximately 1 h, compared to the non-hybrid group (p < 0.001). Estimated blood loss did not differ significantly between groups (p = 0.08). The overall perioperative complication rate was comparable between the two groups (

Table 1. Demographic, clinical features, and surgical data of the study participants.

Variable	Total n = 82	Non-Hybrid (TLIF Only) n = 41	Hybrid (TLIF + DSS) n = 41	p (Fisher’s Exact)
Age (years), median (IQR)	62.1 (17.6)	58.2 (18.9)	65.2 (10.5)	0.009 a
Body mass index a, median (IQR)	29.4 (7.5)	28.7 (9.2)	30.4 (6.2)	0.14 a
Female, n (%)	47 (57)	21 (51)	26 (63)	0.37
Current smoker, n (%) b	12 (15)	7 (17)	5 (12)	0.55
Radicular deficit, n (%)	28 (34)	17 (41)	11 (27)	0.24
Surgical Indications, n (%)				0.04
Radiculopathy	49 (60)	30 (73)	19 (46)
Claudication	21 (26)	6 (15)	15 (37)
Low Back pain	9 (11)	3 (7)	6 (15)
Spondylolisthesis	3 (4)	2 (5)	1 (2)
Surgery type, n (%)				0.04
Primary	67 (82)	30 (71)	37 (90)
Re-operation (same level, decompression or nucleotomy)	15 (17)	11 (27)	3 (7)
Surgery duration (min), median (IQR)	180 (90)	120 (60)	180 (60)	<0.0001 a
Estimated blood loss (mL), median (IQR)	500 (300)	400 (200)	500 (400)	0.08 a
Post-operative complications, n (%)	30 (37)	13 (32)	17 (41)	0.49
Superficial infection	5 (6)	2 (5)	3 (7)
Complicated superficial infection	10 (12)	3 (7)	7 (17)
Infection requiring surgical debridement	6 (7)	5 (12)	1 (2)
Extruded graft requiring revision	2 (2)	0	2 (5)
Repair Cerebrospinal Fluid leak	1 (1)	1 (2)	0
Screw loosening	1 (1)	0	1 (2)
Screw malpositioning requiring revision	1 (1)	0	1 (2)
Adjacent level surgery	1 (1)	1 (2)	0

Data are presented as mean (SD) unless otherwise indicated; IQR, interquartile range; DSS, Dynamic stabilization technique; p-values calculated by Fisher’s Exact. ^a. Wilcoxon rank-sum test with sample estimates (95% confidence interval) are presented for continuous variables. ^b smoking data unavailable for 5 participants; 4 non-hybrid participants and 1 hybrid participant.

).

3.2. Significant Clinical Improvements over Time, Independent of Stabilization Technique

There were no statistically significant differences in PROMs between the non-hybrid and hybrid groups at any follow-up timepoint (p > 0.05; Supplementary Table S1). Across the entire cohort, significant improvements were observed in health-related quality of life, as measured by the Oswestry Disability Index (ODI) and EQ-5D-5L index scores, as well as in back and leg pain severity, as measured by Visual Analogue Scale (VAS) scores, from baseline to 12-month follow-up (p < 0.05; Figure 3).

3.3. No Differences in Fusion Rates Between Stabilization Techniques

Fusion outcomes were assessed using postoperative CT imaging obtained at one year following surgery. In the hybrid group, CT scans were available for 37 participants (90%), with complete interbody fusion observed in 31 cases (84%), partial fusion in 5 cases (14%), and non-union in 1 case (3%). In the non-hybrid group, 35 participants (85%) underwent CT assessment. Complete fusion was evident in 28 cases (80%), partial fusion in 3 cases (9%), and non-union in 4 cases (11%). No statistically significant difference in fusion outcomes was found between the two groups (p = 0.716) (Figure 4).

4. Discussion

This study evaluated short-term outcomes following TLIF using ST cages, with or without the addition of adjacent-level DSS. The findings demonstrated favorable clinical and radiological outcomes at 12-month follow-up. The overall interbody fusion rate was 84%, with comparable outcomes between the TLIF-only and hybrid groups. These results are consistent with or slightly exceed the fusion rates reported in recent systematic reviews, which range from 73% to 94% [4,5]. Wang et al. reported high fusion rates (≈89%) and favorable postoperative alignment outcomes using 3D printed Ti6Al4V cages in posterior lumbar interbody fusion, with clinical improvement comparable to PEEK. Similarly, Lee et al. summarized that 3D printing has enabled customizable cage geometry and porosity, improving mechanical compatibility and osteointegration. These features are increasingly seen as essential for high-quality spinal implants [8,9]. No statistically significant differences in fusion rates or PROMs were observed between the TLIF-only and hybrid TLIF + DSS groups.

ST cages are designed to enhance osseointegration through a combination of high porosity, roughened surfaces, and mechanical properties that approximate cancellous bone. These characteristics are intended to facilitate early and robust bone ingrowth, improve mechanical interlock at the bone–implant interface, and mitigate stress shielding. Their porous architecture—often exceeding 60–70%—facilitates vascularization and supports the migration of osteogenic cells into the cage, mimicking the structure of trabecular bone [6]. Moreover, ST cages typically have a lower elastic modulus (~2–5 GPa) than solid titanium (~110 GPa), approximating the elastic modulus of cancellous bone (0.1–2 GPa). This compatibility may reduce stress shielding and improve load transfer to adjacent vertebral endplates, further stimulating bone remodeling and fusion [13]. Roughened or laser-structured surfaces increase surface energy and mechanical interdigitation, thereby enhancing osteoconductivity and early stabilization [14].

Prior preclinical and clinical studies have demonstrated superior osseointegration and fusion performance of ST cages compared to bioinert polyetheretherketone (PEEK) cages, which lack inherent osseointegrative properties [7,15]. Clinical comparisons have suggested lower subsidence rates and improved load sharing in porous titanium cages, especially in stand-alone or hybrid applications [10]. In our cohort, the 84% overall fusion rate included a slightly higher number of non-unions in the TLIF-only group (4 cases) versus the hybrid group (1 case), although the difference was not statistically significant (p = 0.716). Despite the mechanical and biological advantages, longer-term follow-up studies are still needed to determine whether these material characteristics translate into improved implant longevity and reduced adjacent segment disease [16]. The marginally higher fusion rate observed in the hybrid group may suggest improved biomechanical load sharing, though this interpretation is limited by the small sample size.

Posterior DSS encompass a variety of technologies, including hybrid pedicle screw systems, interspinous devices, and facet replacement constructs [17]. These systems aim to reduce motion at adjacent levels while preserving physiological flexibility—particularly in patients with symptomatic adjacent segments. Importantly, our study was not designed to assess the long-term effects of DSS, especially regarding ASD. The decision to use dynamic stabilization in select cases was based on clinical symptoms and radiographic evidence of early adjacent segment degeneration. DSS was considered in patients with facet arthropathy, mobility-related discomfort at the adjacent level, or a high risk of postoperative overload. While the clinical effectiveness of DSS remains debated, biomechanical studies suggest it may reduce stress concentrations at transitional zones. This study was not powered to assess subgroup effects, but future prospective studies may clarify whether certain populations—such as older adults or patients with multilevel disease—benefit more from hybrid constructs [18,19]. A 2022 network meta-analysis by Chiou et al. also demonstrated that topping-off with devices such as DIAM and Dynesys significantly reduced radiographic and clinical ASD incidence compared to spinal fusion alone, while maintaining equivalent rates of back pain relief and reoperation [11]. The 12-month follow-up duration is insufficient to determine whether DSS contributes to preventing adjacent segment degeneration. Consequently, the long-term protective effects of DSS against ASD remain to be determined.

From a clinical standpoint, both groups experienced marked and statistically significant improvements across all PROMs—including ODI, VAS for back and leg pain, and EQ-5D-5L index scores—over the 12-month period.

However, since no significant between-group differences were observed, the observed benefits likely stem from the fusion technique and implant characteristics rather than the addition of dynamic stabilization. While some meta-analyses suggest potential advantages of hybrid constructs for symptom-specific relief (e.g., reduced leg pain) [20,21] —our findings do not indicate any added clinical benefit from DSS in the short term. A recent retrospective study by Fuster et al. also failed to show a significant difference in clinical outcomes between TLIF with and without DSS [22].

This study has several important limitations. Its retrospective design may introduce selection bias, particularly with regard to surgical indications and device selection. First, baseline differences between groups—such as older age and greater prevalence of claudication in the hybrid group versus more radiculopathy and revision surgeries in the non-hybrid group—may represent confounding variables. Second, although the overall complication and reoperation rate was low (4.8%), most events were minor (e.g., superficial infections), and no significant safety differences were identified between groups. Lastly, the limited 12-month follow-up restricts the interpretation of long-term outcomes, particularly regarding the hypothesized protective role of DSS against ASD. Spinopelvic parameters and sagittal alignment were not assessed, despite their relevance in long-term spinal balance.

While short-term outcomes were similar across groups, long-term data will be essential to assess whether DSS impacts the development of ASD or alters long-term functional trajectories. Although some literature supports the potential for DSS to mitigate ASD risk [23,24], these hypotheses require validation in long-term prospective trials.

5. Conclusions

Our findings demonstrate that structured titanium cages used in TLIF procedures yield high 12-month fusion rates and meaningful improvements in pain and functional outcomes. These results were consistent between patients treated with or without adjacent-level dynamic stabilization. Although DSS may provide long-term benefit, no short-term advantage was observed at the one-year mark. ST cages appear to be a reliable and effective option for achieving short-term success in TLIF. Long-term follow-up studies are warranted to determine the role of dynamic stabilization in preserving adjacent segment health.