Anterior Column Reconstruction of the Thoracolumbar Spine with a Modular Carbon-PEEK Vertebral Body Replacement Device: Single-Center Retrospective Case Series of 28 Patients

Abstract

Background: Carbon-fiber-reinforced polyetheretherketone (CFR-PEEK) vertebral-body replacements (VBRs) aim to mitigate subsidence, minimize imaging artifacts, and facilitate radiation planning while preserving fusion potential. We assessed the safety and efficacy of a novel modular, titanium-coated CFR-PEEK VBR (Kong^®) for anterior column reconstruction (ACR) in the thoracolumbar spine. Primary question: Does the implant safely and effectively achieve and maintain kyphosis correction after ACR for trauma and neoplasms? Methods: A single-center retrospective case series was performed on 28 patients who underwent thoracolumbar ACR with the Kong^® VBR for fractures or tumors (2020–2021). The primary outcome was the bi-segmental kyphotic angle (BKA). Secondary outcomes were screw loosening, cage height loss, fusion rate, subsidence, and tilting. Clinical status was recorded with Odom criteria, Karnofsky Performance Status (KPS), and AOSpine PROST. Results: Twenty-eight patients (mean age, 61 yr; 33% female; mean follow-up, 17.7 mts) were studied. Mean postoperative BKA correction was 16.5° (p = 0.006) and remained 14.5° at final follow-up (p = 0.008); loss of correction was 2.0° (p = 0.568). Subsidence, cage height, and sagittal tilt were unchanged. Fusion (Bridwell grade I/II) was observed in 95% on CT. One deep surgical-site infection occurred. At final follow-up, 91% of patients were graded “excellent” or “good” by Odom. KPS improved by 20 points (p = 0.031), and mean AOSpine PROST was 56.9. Conclusions: Single-center early results indicate that the modular titanium-coated CFR-PEEK VBR is a safe, effective adjunct for thoracolumbar ACR in trauma and neoplasm, providing durable kyphosis correction, mechanical stability and high fusion rates and grants for improved follow-up imaging quality.

1. Introduction

The reconstruction of the anterior and middle spinal columns following corpectomy is essential for restoring spinal stability; however, stabilizing the anterior column remains challenging due to complications such as construct failures, including cage dislocation, non-fusion, segment degeneration, and iatrogenic deformity [1,2]. Early attempts at anterior column reconstruction (ACR) employed bone autografts and femoral allograft struts and rings, but these methods often failed to maintain long-term construct stability [3,4]. The advent of titanium implants, initially as mesh cages [5] and later as expandable devices, led to significant improvements in stabilization, fusion rates, and kyphosis correction. Modular implants have further enhanced applicability and adaptability [2].

Polyetheretherketone (PEEK) implants have gained popularity in interbody fusion and vertebral body replacement (VBR) due to their elastic modulus comparable to cancellous bone, potentially reducing subsidence. Moreover, their radiolucency allows for enhanced follow-up imaging and facilitates postoperative radiation therapy [6,7]. Conversely, PEEK lacks direct osseointegration potential due to its lower bioactivity with bone, leading to proposals for adding bioactive coatings [6]. Titanium-coated carbon fiber-reinforced PEEK (CFR-PEEK) implants theoretically provide improved osseointegration capabilities while retaining the advantages of a reduced elastic modulus and radiolucency. In clinical practice, pedicle screws made from CFR-PEEK have shown a risk profile and rates of implant failure similar to those of traditional metallic implants, with the added benefits of enhanced postoperative imaging quality and greater accuracy in calculating radiotherapy doses compared to titanium counterparts [8,9,10]. Although vacuum plasma-sprayed titanium-coated CFR-PEEK cages have proven effective in lumbar interbody fusion [11], only one study with limited follow-up has evaluated an expandable titanium-coated CFR-PEEK VBR device for ACR to date, underscoring the need for further assessment in different clinical settings [12].

The primary objective of this study was to assess the safety and efficacy of a new modular, expandable, titanium-coated CFR-PEEK VBR (Kong^®) in preserving local kyphosis correction in the thoracolumbar spine for spinal defects due to trauma and spinal neoplasms. Secondary objectives included radiographic evaluation of cage and construct stability, fusion rate, complications, artifacts in post-op CT imaging and clinical outcomes.

2. Materials and Methods

2.1. Study Design and Patient Selection

With approval from the Cantonal Ethics Committee (Kantonale Ethikkommission Bern [KEK]), we conducted a retrospective case series analysis of demographic, perioperative, and radiographic data on patients aged 18 years and older who underwent spinal surgery, using the Kong^® VBR (icotec ag, Altstätten, Switzerland; Figure 1) at our institution. This analysis included our first 28 consecutive cases treated for trauma or tumors in the T9 to L4 region between January 2020 and December 2021. We excluded patients with infectious conditions, those without general consent, those with incomplete medical records, and those with follow-up periods shorter than six months.

2.2. Radiographic Evaluation

Standing plain radiographs were obtained preoperatively, immediately postoperatively, at 6 weeks, 3 months, 1 year, and 2 years, with annual follow-up thereafter if indicated. After 12 months, most patients underwent computed tomography (CT). Kyphosis correction was evaluated by measuring the bi-segmental kyphotic angle (BKA) with the Cobb method; the angle was defined between the upper endplate of the cranial intact vertebra and the lower endplate of the caudal intact vertebra [13]. Measurements were recorded preoperatively, postoperatively, and at the last available follow-up (≥6 months). For single-segment stabilization (hemicorpectomy), only the affected segment was evaluated, whereas in 3-segment stabilization all treated segments were assessed. Subsidence (decrease in construct height) and sagittal cage angulation were determined with the method of Schnake et al. [2]. Screw loosening was evaluated with the Epsilon-angle method of Aghayev et al. to minimize measurement bias [14] (Figure 2). Cage-expansion integrity was assessed with the previously described cage-height coefficient [15]. Anterior bony fusion was graded on available CT scans and plain radiographs according to the Bridwell I–IV scale [16,17].

To evaluate metallic artifacts from carbon cages versus titanium screws, we performed artifact quantification on all available CT scans (Siemens Biograph Vision Quadra PET/CT with iMAR; Siemens Healthineers, Forchheim, Germany) using axial bone-window reconstructions (1 mm slice thickness). We adapted methods from Guggenberger et al. [18] and Bamberg et al. [19]. For titanium screws, we selected the slice with the greatest hypodense streak and placed a circular ROI in the artifact to measure mean Hounsfield units (HU), standard deviation (SD), and diameter (mm). We obtained reference HU/SD from an unaffected slice in identical tissue (e.g., muscle/vertebral body). For carbon cages, we obtained measurements at the cage center in the axial mid-slice, with reference values from adjacent tissue. The rationale is that visually, due to the central locking mechanism, this was the only location where artifacts were uniquely produced by the cage, since the screws are usually above and below its center. While the tantalum spikes at the endplates also cause artifacts, in their corresponding slices, the chief artifacts are produced by screws. We quantified noise as SD in background air. Qualitative evaluation included artifact extent (0 = none to 4 = massive) and diagnostic value (0 = fully diagnostic to 4 = non-diagnostic) scores on a 0–4 Likert scale [18,19].

2.3. Clinical Outcomes

Intraoperative complications were extracted from operative reports to evaluate implantation safety. Perioperative and postoperative complications were recorded from follow-up notes. Clinical outcome was assessed postoperatively with the Odom criteria, a 4-point scale (excellent, good, fair, poor) widely used after spinal surgery [20,21]. At the final follow-up, trauma patients completed the AOSpine Patient-Reported Outcome Spine Trauma (PROST) questionnaire [22]. In tumor patients, the Karnofsky Performance Status (KPS) was documented preoperatively and at the final follow-up [23].

2.4. Surgical Technique

All patients first underwent posterior stabilization with a pedicle screw–rod system, with or without decompression, depending on the underlying pathology and clinical presentation. Corpectomy was performed via mini-open thoracotomy for thoracic pathologies and via lateral mini-open lumbotomy for lumbar pathologies. The extent of corpectomy was tailored to the pathology: hemicorpectomy was performed for mono-segmental cases (e.g., A3-type fractures according to the AO Spine Trauma Classification with severe upper endplate destruction and kyphosis > 30°), whereas complete corpectomy was necessary for bi- or tri-segmental pathologies. The cage size and endplate configuration were intraoperatively customized to the individual defect (see Figure 1). In trauma cases, after corpectomy and adjacent endplate preparation, the cage bed was filled with local cancellous autograft; in tumor cases, allograft bone chips were used. Correct cage positioning was confirmed intraoperatively with fluoroscopy.

2.5. Statistical Analyses

All analyses were carried out in R (v 4.3.1; R Foundation for Statistical Computing) within RStudio (2023.03.0 + 386, Boston, MA, USA). Continuous variables were examined for normality with the Shapiro–Wilk test. Normally distributed data (bi-segmental kyphotic angle, sagittal tilt, Epsilon angle, construct height) are expressed as mean ± SD and were compared between paired time points with two-tailed paired Student t-tests. Non-normal continuous data (cage-height coefficient, Karnofsky Performance Status, AOSpine PROST) are reported as median (IQR) and were analyzed with two-tailed Wilcoxon signed-rank tests. Ordinal and categorical variables (Odom grades, Bridwell fusion grades, complications) are presented as counts and percentages without inferential statistics. Statistical significance was defined as p < 0.05. All R code was independently audited by a biostatistician to confirm accuracy and reproducibility.

3. Results

3.1. Cohort Characteristics

Thirty-three consecutive cases were screened; five were excluded—three for inadequate follow-up and two for infection—yielding a case series of twenty-eight patients: fifteen with high-energy fractures, seven with osteoporotic fractures, and six with tumors (

Table 1. Patient demographics and clinical characteristics, including fracture classifications, neurological status, and surgical details, stratified by pathology (trauma, osteoporotic fractures, tumors) and overall cohort.

Variable	Trauma (n = 15)	Osteoporotic (n = 7)	Tumors (n = 6)	Total (n = 28)
Age, mean ± SD a (years)	54.5 ± 18.4	77.7 ± 9.8	62.5 ± 14.0	60.5 ± 18.6
Female sex, n (%)	5 (33.3%)	3 (42.9%)	2 (33.3%)	10 (35.7%)
Fracture Classifications
AO Spine Classification
Type A	9	—	—	9
Type B	4	—	—	4
Type C	2	—	—	2
AO Spine OF b Classification
OF 3	—	1	—	1
OF 4	—	6	—	6
SINS c Classification
Score 0–6	—	—	1	1
Score 7–12	—	—	3	3
Score 13–18	—	—	2	2
Initial Neurological Impairment (ASIA d Grade)
ASIA A	1	0	0	1
ASIA B	0	0	0	0
ASIA C	3	0	2	5
ASIA D	0	0	2	2
ASIA E	11	7	2	20
Number of Spinal Levels Instrumented
Less than 3 levels	0	1	0	1
3 levels	8	3	2	13
More than 3 levels	7	3	4	14
Staging of Surgeries
1-stage surgery	2	5	4	11
2-stage surgery	13	2	2	17

^a SD, standard deviation; ^b AO, Arbeitsgemeinschaft für Osteosynthesefragen (Association for the Study of Internal Fixation); OF, Osteoporotic Fracture; ^c SINS, Spinal Instability Neoplastic Score; ^d ASIA, American Spinal Injury Association; n, number of patients.

). The cohort included ten women (36%); the mean age was 60.5 ± 18.6 years (range, 19 to 89 years), and the mean radiographic follow-up was 17.7 ± 7.7 months (range, 9 to 36 months).

Eleven reconstructions (39%) were completed in a single stage and seventeen (61%) in two stages. All trauma patients received titanium pedicle screws configured as two percutaneous monoaxial, four percutaneous polyaxial, six open monoaxial, and three open polyaxial constructs. Every osteoporotic case was treated with titanium polyaxial screws—three percutaneously and four via an open approach—with cement augmentation in all seven patients. All tumor cases were instrumented through an open approach; one patient received titanium monoaxial screws and five received carbon-fiber-reinforced PEEK polyaxial screws. Titanium rods were used in every case except three tumor patients. Screw cement augmentation was performed in eight patients (seven osteoporotic and one tumor). In the trauma cohort, posterior hardware was electively removed in nine patients (32%) after at least six months, once fusion had been confirmed on CT.

3.2. Radiographic Outcomes

Postoperatively, the bi-segmental kyphotic angle (BKA) improved by 16.5° ± 23.6° (p = 0.006; preoperative value, −11.6° ± 20.4°) (Figure 3,

Table 2. Changes in radiographic measurements over time. Bisegmental kyphotic angle and other measurement outcomes at preoperative, post-operative, and last follow-up, including statistical comparisons between time points.

Bisegmental Kyphotic Angle, Mean ± SD, in Degrees
Group	T1 a		T2 b		T3 c		p-Values d
Whole	−11.6 ± 20.4		4.9 ± 20.5		2.9 ± 18.0		T1 vs. T2: 0.006 * T2 vs. T3: 0.568 T1 vs. T3: 0.008 *
Trauma	−10.8 ± 16.7		−2.2 ± 16.7		−1.1 ± 18.7		T1 vs. T2: 0.234 T2 vs. T3: 0.837 T1 vs. T3: 0.178
Osteoporotic	−16.3 ± 23.8		5.5 ± 13.2		2.8 ± 13.6		T1 vs. T2: 0.009 * T2 vs. T3: 0.072 T1 vs. T3: 0.021 *
Tumor	−2.5 ± 41.7		45.0 ± 14.4		27.0 ± 4.2		T1 vs. T2: 0.248 T2 vs. T3: 0.236 T1 vs. T3: 0.466
Sagittal tilt, mean ± sd, in degrees
Group		T1		T3		p-values d
Whole		88.9 ± 9.7		88.1 ± 10.4		T1 vs. T3: 0.39
Epsilon angle, mean ± sd, in degrees
Group		T1		T3		p-values d
Whole		4.8 ± 4.0		5.0 ± 3.5		T1 vs. T3: 0.80
Construct height, mean ± sd, in mm
Group		T1		T3		p-values d
Whole		106.2 ± 12.9		105.7 ± 13.2		T1 vs. T3: 0.052
Cage height coefficient, median (IQR)
Group		T1		T3		p-values e
Whole		1.7 (0.36)		1.6 (0.35)		T1 vs. T3: 0.140

^a T1: Preoperative; ^b T2: before postoperative discharge; ^c T3: last follow-up; ^d paired student’s t-test for normally distributed variables; ^e Wilcoxon signed-rank test for non-normally distributed values; * indicates significance.

). At the final review the loss of correction was 2.0° ± 15.0° (p = 0.568), leaving a net gain of 14.5° ± 22.0° over baseline (p = 0.008). Sagittal tilt remained unchanged (mean difference, 0.83 ± 4.61°; p = 0.385). The Epsilon angle showed no interval change (mean difference, −0.17° ± 1.47°; p = 0.793), and no screw lysis > 2 mm was detected on radiographs or CT.

Construct height was stable (mean change, −0.5 ± 0.71 mm; p = 0.052), and the cage-height coefficient did not shift (median change, 0.02; p = 0.140) (Table 2). No cage collapse, expansion-mechanism failure, or loosening at the endplate–core interface occurred.

Twelve-month CT was obtained in eighteen patients and demonstrated Bridwell grade I or II fusion in seventeen (94.4%); one osteoporotic patient was grade III. On the latest plain radiographs, 26 of 28 patients (92.9%) were grade I or II and 2 (7.1%) were grade III, findings that mirrored the CT assessments (

Table 3. Distribution of fusion grades among patients evaluated using Bridwell’s classification on computed tomography (n = 18) and plain radiographs (n = 28).

Grade	Description of Fusion	n (%)
Bridwell classification (computed tomography)		Total n = 18
I	Definite (fused with remodeling and trabeculae)	14 (78%)
II	Probable (graft intact, not fully remodeled and incorporated through; no lucencies)	3 (17%)
III	Probably no (graft intact, but a definite lucency at the top or bottom of the graft)	1 (5%)
IV	No (definitely not fused with resorption of bone graft and with collapse)	0
V	Could not be assessed	0
Bridwell Classification (plain radiograph)		Total n = 28
I	Definite (fused with remodeling and trabeculae)	18 (64%)
II	Probable (graft intact, not fully remodeled and incorporated through; no lucencies)	8 (29%)
III	Probably no (graft intact, but a definite lucency at the top or bottom of the graft)	2 (7%)
IV	No (definitely not fused with resorption of bone graft and with collapse)	0
V	Could not be assessed	0

).

Artifact analysis was conducted on CT scans from all patients with available imaging. Quantitative metrics for carbon-fiber-reinforced PEEK cages and titanium screws are summarized in

Table 4. Summary of CT Artifact Metrics for Carbon Cages and Titanium Screws.

Metric	Carbon Cage (Mean ± SD a)	Titanium Screw (Mean ± SD)	p-Value
Artifact Width (mm)	1.4 ± 0.5	2.6 ± 0.8	0.0004
Artifact HU b	−46.8 ± 39.7	−360.0 ± 104.1	<0.0001
Artifact SD	57.2 ± 49.8	127.7 ± 47.7	0.0010
Absolute Difference (HU)	109.4 ± 56.6	409.5 ± 107.0	<0.0001
Relative Density (%)	−87.6 ± 95.1	−946.2 ± 581.3	0.0001
Implant HU	352.7 ± 23.4	3068.6 ± 3.9	<0.0001

^a SD: standard deviation; ^b HU: Hounsfield unit.

. Artifact width (ROI diameter) was 1.4 ± 0.5 mm for cages and 2.6 ± 0.8 mm for screws (p = 0.0004). Mean artifact HU was −46.8 ± 39.7 HU for cages and −360.0 ± 104.1 HU for screws (p < 0.0001). Artifact SD was 57.2 ± 49.8 for cages and 127.7 ± 47.7 for screws (p = 0.0010). Absolute HU difference from reference tissue was 109.4 ± 56.6 HU for cages and 409.5 ± 107.0 HU for screws (p < 0.0001). Relative density was −87.6 ± 95.1% for cages and −946.2 ± 581.3% for screws (p = 0.0001). Implant HU was 352.7 ± 23.4 HU for cages and 3068.6 ± 3.9 HU for screws (p < 0.0001). Reference HU was 62.6 ± 33.7 for cages and 49.5 ± 22.7 for screws (p = 0.2638). Reference SD was 48.2 ± 32.8 for cages and 45.7 ± 28.7 for screws (p = 0.6734). Noise SD was 50.6 ± 42.3 HU for both (p = NaN).

Qualitative assessment was performed for cages only. Artifact extent score was 0.2 ± 0.4 on a 0–4 scale. Diagnostic value score was 0.0 ± 0.0 on a 0–4 scale. Figure 4 and Figure 5 illustrate the excellent CT and MRI image quality afforded by the low-artifact carbon-fiber PEEK implants.

3.3. Clinical Outcomes

Median Karnofsky Performance Status in the six oncologic cases improved by 20 points from baseline to final review (p = 0.031; Table 4). Among the 22 non-tumor patients, Odom grades improved serially, and 20 (91%) rated the result “excellent” or “good” at last follow-up (Table 4). The 15 high-energy–trauma patients recorded a mean AOSpine PROST of 56.9 ± 21.3, indicating moderate postoperative function (Table 4). Neurologic status improved in seven of the eight patients who presented with a deficit (four traumatic, four neoplastic;

Table 5. Serial outcome measures recorded preoperatively and at 2-, 6-, and final-month follow-ups: Odom grades (non-tumor, n = 22), AOSpine PROST (trauma, n = 15), Karnofsky Performance Status (tumor, n = 6), and ASIA grade evolution (neurologic deficit, n = 8).

Outcome Measure	Preoperative	2 Months	6 Months	Last
A. Odom’s Criteria (Non-Tumor Patients, n = 22)
Excellent, n (%)	—	4 (18.2%)	10 (45.5%)	13 (59.1%)
Good, n (%)	—	11 (50.0%)	10 (45.5%)	7 (31.8%)
Fair, n (%)	—	6 (27.3%)	2 (9.1%)	1 (4.5%)
Poor, n (%)	—	1 (4.5%)	0 (0.0%)	1 (4.5%)
B. AOSpine PROST a Score (Trauma Patients, n = 15)
Mean ± SD	—	—	—	56.9 ± 21.3
C. Karnofsky Performance Status (Tumor Patients, n = 6)
Median (IQR)	60.0 (15.0)	—	—	80.0 (15.0)
D. ASIA b Impairment Scale (Patients with Initial Deficits, n = 8)
Patient 1 (Trauma)	A	A	A	A
Patient 2 (Trauma)	C	D	D	D
Patient 3 (Trauma)	C	D	D	D
Patient 4 (Trauma)	C	D	D	E
Patient 5 (Tumor)	C	D	D	D
Patient 6 (Tumor)	C	E	E	E
Patient 7 (Tumor)	D	E	E	E
Patient 8 (Tumor)	D	E	E	E

^a AOSpine Patient Reported Outcome Spine Trauma; ^b American Spinal Injury Association impairment scale.

).

3.4. Complications

No intraoperative nerve or major vascular injuries occurred, and no secondary cage migration or construct failure was observed. Construct-adjacent vertebral fractures developed in 3 patients (10.7%), all >70 years. One tumor patient (3.6%) experienced a deep surgical-site infection 2 weeks postoperatively; the infection resolved after a single debridement without hardware exchange.

4. Discussion

This retrospective pilot case series evaluated a modular, expandable, titanium-coated CFR-PEEK vertebral-body replacement (Kong^®) for anterior column reconstruction after thoracolumbar trauma or tumor. The bi-segmental kyphotic angle improved by a mean 14.5° ± 22.0° from baseline (p = 0.008) and was preserved at final review. Subsidence was negligible, sagittal tilt unchanged, and no screw loosening or implant failure occurred. The device therefore provided durable kyphosis correction, high fusion rates, and a low complication burden, supporting its use as an alternative to conventional titanium cages.

4.1. Comparison with Existing Literature

Our radiographic results mirror those reported for expandable titanium cages. Schnake et al. noted a 2.2° loss of bi-segmental kyphotic angle (BKA) at 12 months and 6.5° at 60 months, with 2.2 mm of subsidence, after thoracolumbar fracture reconstruction with titanium cages [2]. Lange et al. recorded a 2.3° BKA loss and 2.9 mm subsidence at 12 months in a mixed trauma-tumor cohort treated with expandable titanium vertebral-body replacements [24]. The present series showed comparable maintenance of alignment and similarly minimal subsidence.

The lower elastic modulus of PEEK may further mitigate subsidence. In a randomized trial of cervical spondylotic myelopathy, Chen et al. reported subsidence >3 mm in 34.5% of titanium versus 5.4% of PEEK cage [25]. Schwendner et al. evaluated the same titanium-coated CFR-PEEK (Kong^®) cage in 17 thoracolumbar tumor cases and found a 3.3° BKA loss and 3.8% subsidence at 0.8-year mean follow-up [12], corroborating our findings.

4.2. Technical Aspects

Due to the better load distribution, subsidence is least when the cage spans the apophyseal ring and maximizes end-plate contact, as previously demonstrated in biomechanical experiments [26]. The Kong^® system supplies interchangeable endplates with varied footprints for intraoperative, anatomy-matched sizing. Its carbon-fiber PEEK core approximates the elastic modulus of cancellous bone, a particular advantage in osteoporotic bone. Reflecting these features, mean subsidence in our series was only −0.5 ± 0.7 mm (p = 0.052).

4.3. Osseointegration and Fusion Rates

PEEK’s bio-inert nature has historically raised concerns about its osseointegration potential [27]. Prior studies evaluating uncoated PEEK VBRs for thoracolumbar reconstruction have reported similarly high success rates, such as the 96.3% fusion rate observed by Brandão et al. [28]. The current series observed a Bridwell grade I/II fusion rate of 95% on CT imaging with the titanium-coated CFR-PEEK cage. This result is favorable and is in the same ballpark as reported fusion rates reported for all-titanium constructs [29]. However, this study was not designed for direct comparison, and given the absence of a control group with titanium implants, conclusions on osseointegration performance relative to all-titanium devices cannot be drawn. In theory, the titanium layer supplies an osteoconductive surface. Conversely, preclinical studies have described gradual coating loss [30,31], and a meta-analysis reported no fusion advantage for coated versus uncoated PEEK implants [6]. In summary, the high rate of fusion observed here is a promising finding, but it requires validation through direct comparative studies.

4.4. Rationale for Use and Economic Considerations

The selection of a vertebral body replacement is guided by its clinical performance and economic impact. While the radiolucency of CFR-PEEK offers a clear advantage in spinal oncology and infection-related indications, facilitating imaging and radiation planning, its application in trauma and osteoporotic fractures in this cohort was primarily driven by mechanical properties, including footprint modularity for anatomical contact and a lower elastic modulus compared to titanium, which theoretically reduces subsidence risk in compromised bone. Outside oncologic or infection-related cases, where radiolucency is the primary benefit, the higher device cost necessitates selective use. Although a formal cost–utility analysis is beyond this study’s scope, the convergence of mechanical, oncologic, and institutional fiscal advantages justified adopting the CFR-PEEK device across our heterogeneous pathologies. This low-artifact profile may enable more accurate detection of complications such as subsidence or loosening. Formal cost–utility data are lacking and should be addressed prospectively.

4.5. Implantation Safety and Complications

The Kong^® cage showed an excellent safety profile: no intraoperative implant-related events and no hardware failures during follow-up. One deep surgical-site infection required debridement. The cage-height coefficient remained stable, confirming construct integrity [15]. From our early experiences, insertion of the cage in narrow anatomic situation (e.g., hemicorpectomy) can be challenging because of the pronounced anchoring spikes; precise sizing and gentle impaction are essential to avoid neural or vascular injury, especially for first-time users.

4.6. Limitations

The conclusions of this study must be interpreted in the context of several limitations. First, as a retrospective, single-center case series, the study is subject to selection bias and lacks a concurrent control group. The absence of a comparative arm—such as patients treated with all-titanium VBRs—precludes any definitive statements regarding the superiority of this CFR-PEEK device.

Second, the cohort was small and heterogeneous, comprising patients with distinct pathologies (trauma, osteoporosis, tumors). This heterogeneity may introduce confounding factors affecting outcomes, such as varying bone quality or healing potential across groups. Although subgroup data are presented, the small sample size limits statistical power for subgroup analyses. We combined tumor and osteoporotic fracture patients for certain outcome measures based on the shared clinical characteristic of compromised, osteopenic bone; however, this pragmatic grouping does not overcome the fundamental limitation of low statistical power.

Third, the mean radiographic follow-up of 17.7 months is adequate for assessing early outcomes such as implant stability and fusion initiation, but it is insufficient to evaluate long-term phenomena like delayed implant subsidence, adjacent segment disease, or hardware fatigue. Consequently, conclusions on long-term durability are preliminary and should be interpreted cautiously. Nevertheless, the goal of the investigated procedure is a fusion of minimum two vertebra, which was achieved in 95% of the cases. A late subsidence or failure of the VBR after solid fusion is very unlikely.

Finally, our commentary on the implant’s favorable imaging characteristics is based on qualitative radiological assessment. A quantitative analysis of imaging artifact was not performed but would be a valuable addition in future prospective studies. Ultimately, the promising results of this series require validation through larger, prospective, randomized controlled trials.

5. Conclusions

This study suggests that the modular, expandable, titanium-coated CFR-PEEK Kong^® VBR system is a safe and effective option for ACR in the thoracolumbar spine due to trauma or neoplasia. It achieves significant kyphosis correction, maintains sagittal alignment with minimal subsidence, and demonstrates high fusion rates. The system’s adaptability and favorable imaging characteristics due to low artifact formation offer additional advantages. While these single-arm data compare favorably with historical titanium-cage series, controlled trials are required to verify any superiority to titanium implants.