Abstract
Background
Posterior hemivertebra resection, combined with anterior column reconstruction by short segmental fusion, is a widely accepted approach in the treatment of congenital scoliosis (CS) caused by hemivertebra. Both autogenous cancellous bone and titanium mesh cages (TMCs) are commonly used for anterior column reconstruction. However, TMCs have gained more attention due to their stability, biocompatibility, and competent mechanical properties. This study compared the clinical and biomechanical performance of autogenous bone and TMCs in CS patients.
Methods
We retrospectively reviewed 117 CS patients with hemivertebra between T10 and L2 who underwent posterior hemivertebra resection, short segmental internal fixation (2–4 segments) and anterior column reconstruction from 2008 to 2020. Based on the methods of anterior column reconstruction, patients were divided into the autogenous bone group and the TMC group, and further subgrouped by preoperative sacral tilt. Coronal and sagittal deformity parameters, intraoperative parameters and complications were analyzed. Additionally, we developed two finite element models of the T10-L2 thoracolumbar spine with a TMC and autogenous bone graft respectively, and evaluated the maximum von Mises stress and range of motion in both models.
Results
In patients without severe sacral tilt, the main curve showed a significantly greater reduction immediately after surgery in Group 2a (TMC group) compared to Group 1a (autogenous bone group). However, the curve correction was similar between the two groups at the lastest follow-up. In patients with severe sacral tilt, the ues of TMCs significantly prevented the extention of main curve length after surgery (p = 0.04). All 6 reported complications occurred in the autogenous bone group. Finite element analysis revealed that patients with autogenous bone experienced more stress on the screws and rods compared to those with TMC.
Conclusion
TMC is a suitable implant for anterior column reconstruction after thoracolumbar hemivertebra resection and provides comparable deformity correction as autogenous bone and a lower risk of complications.
Level of evidence
Level III.
Supplementary Information
The online version contains supplementary material available at 10.1186/s13018-025-06433-4.
Keywords: Hemivertebra resection, Thoracolumbar, Anterior column reconstruction, TMC, Retrospective study, Finite element analysis
Introduction
Hemivertebra (HV) is a common form of congenital spinal dysplasia [1]. Some types of HV progress quickly and may cause severe scoliosis [2]; lumbar HV, in particular, could cause compensatory sacral tilt [3]. The current standard treatment for HV includes HV resection with short segmental fusion [4]. As the anterior column of the spine bears greater load and has richer vascularity than the posterior structures [5], anterior column reconstruction is necessary following osteotomy.
Autogenous cancellous bone graft is traditionally used for anterior column reconstruction because of its low immunogenicity and strong osteogenic potential. However, this method has notable limitations, including complications at donor sites [6], insufficient mechanical strength, and limited graft-endplate contact surface [7]. Recently, titanium mesh cages (TMCs) filled with cancellous bone have been introduced to the treatment of hemivertebra to provide more rigid reconstruction and facilitate bone fusion reported in previous studies [7, 8]. Nonetheless, it remains controversial whether TMCs exhibit superior performance in anterior column reconstruction, especially due to postoperative subsidence of TMCs.
A previous study [9] verified the merits of TMCs for improving deformity correction and maintaining spinal alignment. Wang et al. [3] further demonstrated that TMCs help restore coronal balance and preserve segmental mobility in patients with severe sacral tilt (> 10°). However, larger cohort studies should further clarify whether TMCs outperform autogenous bone in anterior column reconstruction after HV resection.
Hemivertebra-induced deformities at the thoracolumbar junction (T10-L2) are typically characterized by pronounced kyphosis, necessitating the use of TMCs for anterior column reconstruction. In this study, we retrospectively compared the clinical outcomes of TMCs versus autogenous bone in anterior column reconstruction of thoracolumbar spine. In addition, a finite element study was conducted to evaluate the mechanical performance between two reconstruction methods.
Materials and methods
Patient characteristics
We retrospectively reviewed the clinical data of 117 patients with a single HV located at T10-L2 who underwent posterior HV resection at our center between 2008 and 2020. Patients were divided into 2 groups according to the method of anterior column reconstruction: Group 1 (TMC group): 67 patients who underwent anterior column reconstruction using TMCs filled with cancellous bone; Group 2 (autogenous bone group): 50 patients who underwent anterior column fusion using a mixture of autogenous cancellous bone harvested from the resected HV and NovaBone@ (NovaBone product, LLC, US). Based on preoperative sacral tilt, we subdivided the patients without severe sacral tilt (defined as a sacral tilt angle larger than 10°3) into Group 1a (35 patients with TMCs) and Group 2a (34 patients with autogenous bone). Patients with severe sacral tilt were classified into Group 1b (28 patients using TMCs) and Group 2b (9 patients using autogenous bone). Demographic information and intraoperative parameters were collected and compared among groups.
Indications
The decision to use TMCs and autogenous bone graft for hemivertebra short segment fusion depends on the severity of the deformity, the size of the osteotomy gap, and the need for postoperative stability. Indications for TMCs [9]: (1) large hemivertebrae that leave a gap ≥ 0.5 cm after compression correction following resection; (2) posterolateral hemivertebrae that may lead to significant kyphoscoliosis, where anterior structural reconstruction can serve as a fulcrum to enhance kyphosis correction; (3) hemivertebrae between L5 and S1, even if semi-segmented, that cause significant trunk shift, requiring early surgical intervention to prevent severe coronal imbalance and avoid long fusion to the lumbar spine. Indications for autogenous bone graft: (1) small osteotomy gaps or cases requiring rapid fusion, as autogenous bone promotes fusion with its osteogenic activity; (2) small hemivertebrae and minimal kyphoscoliosis.
Surgical procedure [9]
After placing the patients in a prone position, a midline incision was made and subperiosteal dissection was performed to expose the HV and adjacent structures. Once the location of the HV was confirmed by fluoroscopy, pedicle screws were placed at the adjacent levels. The posterior elements of the HV, consisting of the lamina, lower and upper facets and the transverse processes, were then removed to expose the nearby nerve roots and pedicles. In cases where the HV was in the thoracic segments, the rib head and proximal part of the excess rib on the convex side should be removed to expose the lateral cortex of the HV. With temporary rods stabilizing the spine, the vertebral body of the HV was removed along with the cartilage endplate, the cartilage tissue at the contralateral side, and the upper and lower intervertebral discs. Then, the temporary rods were replaced by the pre-contoured rods, and either TMCs or autogenous cancellous bone graft were placed in the gap through a posterolateral approach. Stepwise compression was then appliedto correct the deformity, with confirmation via fluoroscopy. Throughout the whole procedure, sensory evoked potentials (SEPs) and motor evoked potentials (MEPs) were continuously monitored. Two patients who underwent surgery using either autogenous bone or a TMC were shown in Fig. 1.
Fig. 1.
Two cases of titanium mesh cage and autogenous bone to treat hemivertebra. A female patient was diagnosed with congenital scoliosis caused by hemivertebra between L2 and L3 when she was 2 years old (A, B). She underwent posterior hemivertebra resection, short segmental fusion from L2 to L3, and anterior column reconstruction with autogenous bone (C, D). A male patient was diagnosed with congenital scoliosis caused by hemivertebra between T12 and L1 when he was 3 years old (E, F). We performed posterior hemivertebra resection, short segmental fusion from T11 to L2, and anterior column reconstruction with a titanium mesh cage (G, H)
Measurements of patients without severe sacral tilt
Group 1a and Group 2a were included to assess the correction of deformity. The following variables were evaluated: (1) Cobb angle of the main curve; (2) standardized sagittal HV Cobb: the Cobb angle between the lower endplate of the upper vertebra and upper endplate of the lower vertebra of the HV on lateral (L) radiograghs minus standard segmental Cobb according to the previous reference [10]; (3) coronal alignment (CAlign): measured as the horizontal distance between the midpoint of C7 and S1; (4) sagittal vertical axis (SVA); (5) Hadjacent: the height between the upper endplate of the upper vertebra and the lower endplate of the lower vertebra of the HV on L radiograghs.
Measurements of patients with severe sacral tilt
Groups 1b and 2b were included for comparison of postoperative progression in patients with severe sacral tilt. The main curve lengths, defined as the numbers of segments from the upper end vertebra (UEV) to the lower end vertebra (LEV), were recorded at different time points during follow-up.
The relationship between the size of TMCs and HV
To assess the relationship between the titanium cage size and preoperative measurements, the following measurements were measured: (1) AP/L-HV height: the vertical maximum height of HV on anterior–posterior (AP) and L radiographs; (2) AP/L-HV width: the horizontal width and anteroposterior diameter of HV on AP and L radiographs; (3) cage height/diameter. Stepwise regression was applied for the analysis.
Statistical analysis
Continuous variables are presented as the median (lower quartile, upper quartile). For continuous variables, the Mann‒Whitney U test and Friedman test were used to test the significance of independent and dependent samples respectively. The chi-square test was used to test the significance of categorical variables. Significant differences were revealed by p < 0.05.
Development of the 3D finite element model
A finite element model of the T10-L2 thoracolumbar spine segments was developed based on the CT reconstruction data from a 2-year-old boy (weight: 14 kg, height: 85 cm) with an HV between T12 and L1. The original CT DICOM file was imported into Mimics 21.0 (Materialise Inc., Belgium) for geometric contour extraction of the T9-L4 vertebral body, using image threshold segmentation, filling and erasing. Afterward, the models were further modified by surface construction, patching, griding, fitting in Geomagic 2017 (Geomagic Inc., USA). The spine models were then imported into SolidWorks, where two distinct models were created: one involving HV resection, short segmental fixation from T11 to L1, and anterior reconstruction with a TMC filled with cancellous bone (Model 1, Fig. 2A–D), and the other with cancellous bone alone (Model 2, Fig. 2E–H).
Fig. 2.
The three-dimensional finite element models established in this study. We simulate posterior hemivertebra resection, short segmental fixation from T12 to L1, and anterior column reconstruction with a titanium mesh cage (A–D, Model 1) or autogenous bone (E–H, Model 2)
The models of endplates, intervertebral discs, cartilage and growth plates were developed using extrusion and split functions in SolidWorks 2021 (Dassault Systems Inc., France). The models of screws, rods, TMC and cancellous bone (represented by a cylinder with identical size as the cage) were also developed in SolidWorks 2021 according to the actual surgical instruments. Ligament reconstruction was simulated with spring unit. Since the cancellous bone filled in the TMC serve primarily as an inducer for bone fusion without providing mechanical support, we only simulate the TMC in Model 1. The detailed modeling procedure is listed in Supplementary 1. The material properties and parameters for the ligaments are shown in Supplementary Tables 1 and 2. Because of the similar mechanical properties of cancellous bone and NovaBone@, we assumed a cylinder with material properties matching cancellous bone to simulate in the fillings in the procedure. Previous mechanical tests have comfirmed that the cancellous bone filling in the TMC serves only as an fusion inducer for without mechanical support (Supplementary Fig. 1).
Finite element analysis (FEA)
In the analytical procedure, Ansys 17.0 was used to mesh the network and assign material properties to different parts of the structure. The element type was a first-order element. All mesh divisions used tetrahedral mesh, and all boundary conditions were set as binding. After fixing the bottom of the models, a 70-N vertical axial preload and a 4-N·m moment were applied to simulate body weight and thoracolumbar spinal motions, respectively. Due to the absence of existing references describing the model with an HV between T12 and L1, we validated the rationality of the model by comparing the range of motion (ROM) of T9-T12 and L1-L4 segments from our study to the reference data [11]. For the FEA, the nodes on the inferior endplate of L2 in both Model 1 and Model 2 were fixed in all directions as boundary conditions. Our study analyzed the vertical axial load and motions of flexion/extension, lateral bending, and axial rotation. We also evaluated the ROM and the maximum von Mises stress on the implants and the bone surfaces near the implants.
Results
Demographic information
Table 1 presents the demographic information of the patients included in this study. A total cohort of 117 patients was included in this study. The follow-up time was significantly longer in Group 2, and there was more fully-segmented HV in Group 1. However, age, sex, HV location and other demographic characteristics were comparable between two groups.
Table 1.
Demographic information of the patients
| Subject | Group 1 | Group 2 | P value |
|---|---|---|---|
| Number | 67 | 50 | |
| Age (years) | 3.9 (2.9, 5.7) | 3.0 (2.0, 6.3) | 0.33 |
| Gender | 27F/40 M | 23F/27 M | 0.54 |
| Follow-up (months) | 12.0 (5.0, 24.0) | 31.0 (3.0, 60.0) | 0.01 |
| HV location | |||
| Thoracic segments | 17 | 15 | 0.58 |
| Lumbar segments | 50 | 35 | |
| HV type | |||
| Fully-segmented | 49 | 22 | 0.00 |
| Semi-segmented | 18 | 28 | |
| Total fixed segments | |||
| 2 | 36 | 27 | 0.20 |
| 3 | 17 | 7 | |
| 4 | 14 | 16 | |
| HV height | |||
| Pre- AP | 14.2 (12.7, 16.1) | 13.5 (11.3, 15.3) | 0.24 |
| Pre- L | 14.9 (11.7, 17.7) | 13.0 (10.3, 16.9) | 0.25 |
| HV width | |||
| Pre- AP | 13.6 (12.0, 17.2) | 12.6 (11.1, 16.0) | 0.15 |
| Pre- L | 16.9 (14.2, 20.1) | 14.7 (12.3, 20.7) | 0.26 |
| Surgery duration (min) | 225.0(205.0, 240.0) | 207.5(181.3, 230.0) | 0.02 |
| Amount of bleeding (mL) | 150.0(50.0, 200.0) | 100.0(0.0, 250.0) | 0.15 |
Pre-, preoperative parameters; HV, hemivertebra; AP, on anteroposterior radiographs; L, on lateral radiographs;
Comparison of deformity correction
For patients without severe sacral tilt, the comparison of deformity correction is shown in Table 2. Immediately after surgery, the main curve correction in Group 2a was significantly better than that in Group 1a, but showed comparable results at the last follow-up. Additionally, the coronal and sagittal balance of the two groups was comparable at any time point. The height reconstructed by either TMCs or autogenous bone showed no significant deviation. For patients with severe sacral tilt, as shown in Table 3, the extension of main curve length showed a significant deviation in Group 2b.
Table 2.
Deformity correction of the patients without severe sacral tilt
| Subject | Group 1a | Group 2a | P value |
|---|---|---|---|
| Main curve (°) | |||
| Pre- | 36.7 (30.0, 42.9) | 34.9 (30.9, 40.4) | 0.40 |
| Post- | 10.6 (3.9, 14.2) | 6.9 (2.8, 9.4) | 0.04 |
| Δ1 | 26.3 (23.3, 32.6) | 26.7 (23.2, 32.1) | 0.77 |
| 6 months post- | 14.1 (9.0, 19.5) | 11.4 (6.8, 21.8) | 0.44 |
| Δ2 | 23.6 (17.5, 28.8) | 23.2 (15.7, 28.0) | 0.56 |
| FU- | 16.1 (7.9, 20.9) | 9.1 (3.7, 18.7) | 0.09 |
| Δ3 | 23.6 (16.1, 29.1) | 22.7 (15.7, 29.5) | 0.95 |
| Standardized sagittal HV Cobb (°) | |||
| Pre- | 18.9 (8.9, 33.9) | 18.5 (4.0, 30.0) | 0.64 |
| Post- | 10.1 (4.4, 15.3) | 8.3 (3.4, 15.6) | 0.72 |
| Δ1 | 7.9 (− 2.7, 23.7) | 6.2 (− 4.5, 20.1) | 0.57 |
| 6 months post- | 10.2 (4.1, 18.7) | 5.6 (0.4, 15.7) | 0.24 |
| Δ2 | 10.9 (− 2.5, 23.8) | 14.4 (− 0.6, 20.9) | 0.97 |
| FU- | 9.6 (3.7, 18.0) | 6.9 (3.2, 17.0) | 0.40 |
| Δ3 | 9.3 (− 3.6, 22.9) | 8.0 (− 6.5, 21.0) | 0.75 |
| CAlign (mm) | |||
| Pre- | 10.4 (6.6, 19.8) | 10.7 (5.0, 17.8) | 0.61 |
| Post- | 12.0 (4.9, 25.9) | 14.5 (5.1, 23.1) | 0.93 |
| Δ1 | − 2.6 (− 7.2, 5.9) | − 3.4 (− 9.5, 7.4) | 0.93 |
| 6 months post- | 7.0 (3.2, 15.9) | 9.5 (3.5, 12.7) | 0.96 |
| Δ2 | 5.6 (− 2.5, 8.6) | 5.6 (− 2.4, 13.6) | 0.40 |
| FU- | 11.8 (4.5, 20.3) | 7.2 (2.6, 15.8) | 0.27 |
| Δ3 | 1.6 (− 4.7, 8.2) | 2.0 (− 5.6, 9.9) | 0.76 |
| SVA (mm) | |||
| Pre- | 18.6 (8.4, 35.3) | 24.0 (10.1, 43.2) | 0.30 |
| Post- | 32.4 (12.1, 51.8) | 37.3 (17.7, 55.7) | 0.44 |
| Δ1 | − 8.7 (− 30.1, 16.4) | − 13.1 (− 24.9, 4.3) | 0.70 |
| 6 months post- | 16.7 (9.5, 35.7) | 29.5 (11.3, 44.7) | 0.13 |
| Δ2 | − 1.5 (− 18.7, 9.3) | 4.8 (− 17.5, 23.2) | 0.48 |
| FU- | 15.8 (7.1, 33.1) | 23.9 (9.6, 40.8) | 0.33 |
| Δ3 | 9.3 (− 3.6, 22.9) | 8.0 (− 6.5, 21.0) | 0.75 |
| Hadjacent (mm) | |||
| Pre- | 35.6 (30.6, 41.0) | 36.1 (29.3, 45.8) | 1.00 |
| Post- | 33.3 (30.1, 36.6) | 29.8 (27.3, 37.3) | 0.10 |
| Δ1 | 2.6 (− 0.4, 6.4) | 2.6 (− 1.7, 7.6) | 0.74 |
| 6 months post- | 31.3 (28.5, 34.1) | 29.1 (25.5, 36.0) | 0.57 |
| Δ2 | 5.4 (1.1, 9.1) | 2.9 (0.2, 10.3) | 0.82 |
| FU- | 31.1 (29.5, 35.5) | 33.3 (27.5, 38.2) | 0.65 |
| Δ3 | 3.2 (-1.4, 9.0) | 3.1 (− 5.5, 9.5) | 0.51 |
Pre-, preoperative parameters; post-, postoperative parameters; Δ1, defined as the main curve Cobb angle before surgery minus that immediately after surgery; 6 months post-, parameters 6 months after surgery; Δ2, defined as the main curve Cobb angle before surgery minus that at 6 months after surgery; FU-, the latest follow-up parameters; Δ3, defined as the main curve Cobb angle before surgery minus that at the latest follow-up
Table 3.
Main curve length in patients with severe sacral tilt
| Subject | Post- | 6 months post- | FU- | P value* |
|---|---|---|---|---|
| Group 1b | 4.0(3.0, 4.0) | 3.5(3.0, 4.0) | 3.5(3.0, 4.0) | 0.16 |
| Group 2b | 2.0(2.0, 2.0) | 4.0(3.0, 4.0) | 3.5(3.0, 4.0) | 0.04 |
Post-, postoperative parameters; 6 months post-, parameters 6 months after surgery; FU-, the latest follow-up parameters
*p refers to the comparison between FU- and post- value
Intraoperative comparison and complications
The operation time in Group 1 was significantly longer than that in Group 2 (p = 0.02). Nonetheless, no significant differences was found in the amount of bleeding between the two groups (p = 0.15). All 6 reported complications occurred in the autogenous bone group, including 4 cases related to internal fixation failure and 2 patients with poor incision healing.
The parameters of TMCs
No distortion of TMCs was observed at 6 months after surgery (p = 0.23) or at the last follow-up (p = 0.22) compared to the parameters immediately after surgery. Subsidence of TMCs occurred in some patients, but none experienced significant correction loss or implants failures. Regression analysis showed that the postoperative height of the TMC was positively correlated with the height of the HV in AP radiography, and the postoperative diameter of the TMC was positively correlated with the height of the HV in L radiography.
Model validation
The ROMs of the T9-T12 model and L1-L4 model in all directions were comparable to the data reported by Busscher [11], as shown in Fig. 3. Model 1 consisted of 420,489 nodes and 226,254 cells, while Model 2 consisted of 421,296 nodes and 225,011 cells.
Fig. 3.
Validation of the model. We calculated the range of motion of the T9-T12 Model (A) and L1-L4 Model (B) during flexion/extension, lateral bending and axial rotation. The black bars indicate the standard deviation of each result. Furthermore, we compared our results with previous studies
Comparisons of stress and ROM
Overall, the maximum von Mises stress on the TMCs was apparently larger than that on the autogenous bone, as shown in Fig. 4 and Table 4. Consequently, the maximum von Mises stress on the upper and lower adjacent vertebrae was significantly greater in Model 1. In contrast, the maximum von Mises stress on the posterior screws and rods was lower in Model 1 (1788.7 MPa max) than in Model 2 (2131.30 MPa max), as shown in Fig. 5 and Table 4. In addition, as shown in Fig. 6, the ROM was larger in Model 2, with the maximum ROM observed in lateral bending movement.
Fig. 4.
The FEA results of von Mises stress (MPa) on the implants of anterior column reconstruction. We calculated the maximum von Mises stress on the titanium mesh cages (A–F) in Model 1 and that on the autocancellous bone (G–L) in Model 2
Table 4.
The von Mises stress (MPa) in model 1 and model 2
| Flexion | Extension | Left lateral bending | Right lateral bending | Left axial rotation | Right axial rotation | ||
|---|---|---|---|---|---|---|---|
| The von Mises stress (MPa) on the implants | |||||||
| Titanium cage | 260.27 | 120.15 | 464.32 | 439.60 | 368.61 | 372.80 | |
| Autograft | 5.17 | 2.09 | 3.35 | 3.95 | 5.67 | 4.20 | |
| The von Mises stress (MPa) on the screws and rods | |||||||
| Model 1 | 381.51 | 426.90 | 1788.70 | 1780.50 | 468.93 | 446.42 | |
| Model 2 | 500.97 | 555.26 | 2131.30 | 1823.00 | 442.78 | 431.85 | |
| The von Mises stress (MPa) on the upper and lower adjacent bone surface | |||||||
| Model 1 | Upper adjacent bone | 92.04 | 39.69 | 53.63 | 47.54 | 72.36 | 90.87 |
| Lower adjacent bone | 142.83 | 62.88 | 243.36 | 177.27 | 146.07 | 166.45 | |
| Model 2 | Upper adjacent bone | 5.55 | 3.20 | 32.59 | 25.74 | 12.80 | 14.56 |
| Lower adjacent bone | 5.44 | 5.69 | 16.25 | 16.01 | 5.70 | 6.05 | |
Fig. 5.
The FEA results of von Mises stress (MPa) on the posterior screws and rods. We calculated the maximum von Mises stress on the posterior screws and rods in Model 1 (A–F) and Model 2 (G–L)
Fig. 6.
The FEA results of ROM. We calculate the range of motion in Model 1 and Model 2 during flexion/extension, lateral bending and axial rotation under a 70-N vertical axial preload and a 4-N·m moment
Discussion
This study found that TMCs are an effective and safe option for anterior column reconstruction after hemivertebra resection and short segmental fusion in a cohort of 117 patients. Only in recent years has TMC implantation following osteotomy been gradually accepted and gained popularity. A previous report [9] highlighted the importance of TMCs for adequate correction in the coronal and sagittal planes in patients who underwent posterior HV resection with anterior column reconstruction and short segmental fusion. Zhang et al. [12] also comfirmed that the use of TMCs would not increase the rate of mechanical complications, such as loss of correction and internal fixation failure. Our results further demonstrated that TMCs with autogenous bone had a similar effect on deformity correction in patients without severe sacral tilt. In addition, the risk of mechanical complications was lower with TMCs than compared to autogenous bone alone.
The FEA results also validated this conclusion. With the TMC bearing more load, the screws and rods experienced less loading, thereby reducing the chance of rod breakage and screw loosening, consistance with findings in the previous study [13]. ROM is a key indicator of spinal stability, with higher ROM reflecting lower stability after interbody fusion [14], which is recognized as a risk factor for screw loosening or pulling out [15]. As shown in Fig. 6, the ROM of Model 2 is larger than that of Model 1, indicating that the TMC provided higher stabilitythan autogenous bone alone and may reduce the risk of implant failure.
For patients with severe sacral tilt, the utilization of TMCs could prevent the progression of scoliosis after surgery. Wang et al. [3] first suggested that for patients with severe sacral tilt, it was preferred to undercorrect scoliosis with TMCs to compensate for the residual lumbosacral curve. If the scoliosis was thoroughly corrected, the upper instrumented vertebra (UIV) tilt and deviation would be increased, leading to deterioration of deformity and extension into the proximal mobile spine. Our results also supported this hypothesis: at the last follow-up, patients with autogenous bone showed an extension of the main curve length (p = 0.04), whereas the deformity segments remained stable in patients with TMCs.
A prolonged operative time due to the additional shaping of TMCs may be a potential disadvantage. Our results also showed that Group 1 had a significantly longer operative time than Group 2. Therefore, we explored the relationship between the size of TMCs and the preoperative HV size by regression analysis. Estimating the proper dimensions of the TMCs before surgery may reduce the operation time.
Nonetheless, it is worth noting that the implantation of TMCs carries the risk of cage subsidence after surgery. Previous study has reported the risk factors for the subsidence of TMCs in lateral lumbar interbody fusion [16]: including low bone density, age 65 and female. In our study, cage subsidence was observed in some patients, but none experienced significant correction loss and implant failure. This can be attributed to the fact that all our patients were young children without osteoporosis and have good bone metabolism for timely fusion. Additionally, the preoperative deformity was not severe, and the postoperative stability was relative good.
There are still some limitations in this study: (1) As a retrospective cohort study, some confounding factors would have affected the statistical analysis. Therefore, further randomized controlled trials or case-matched studies are needed to better address this topic; (2) Since TMCs have only recently been applied in anterior column reconstruction after hemivertebra resection, the overall follow-up time for patients with TMCs is relatively short. A longer follow-up time is required to further validate the effectiveness and safety of TMCs; (3) As a retrospective study, it was challenging to adhere strictly to the indications for TMCs. Therefore, a prospective study with well-defined indications for TMCs should be conducted in the future.
Conclusion
Both TMCs and autogenous bone are effective for anterior column reconstruction after hemivertebra resection and short segmental fusion. The comparison of clinical outcomes and the FEA confirmed that the TMCs reduce the stress on screws and rods, leading to a lower mechanical complication rate. In addition, for patients with severe sacral tilt, TMCs could serve as a fulcrum for undercorrection and prevent postoperative progression of deformity.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
We would like to thank every author for their great efforts to this research.
Abbreviations
- TMC
Titanium mesh cage
- HV
Hemivertebra
- SEP
Sensory evoke potential
- MEP
Motor evoked potential
- SVA
Sagittal vertical axis
- UEV
Upper end vertebra
- LEV
Lower end vertebra
- UIV
Upper instrumented vertebra
- ROM
Range of motion
Author contributions
ZL and BH contributed equally to this work and share first authorship. SW and JZ contributed equally to this work and share corresponding authorship. ZL and YD coordinated and supervised data collection, conducted the data analysis.ZL conducted the data analysis and drafted the manuscript. BH provided imaging data and advice in 3D model development. SW and JZ conceived and obtained funding of this study, provided critical comments and revised the manuscript. All authors have read and agree to the published version of the manuscript.
Funding
This work was supported by the National High Level Hospital Clinical Research Funding (2022-PUMCH-B-001) and National Key Research and Development Program of China (2023YFC2507700). Funding source did not play a role in the investigation.
Data availability
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
Data for this study were collected following institutional review board approval (I-25PJ1254) and had consents to each participant.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Zhiyi Li and Bingtai Han have contributed equally to this work.
Contributor Information
Shengru Wang, Email: wangshengru@foxmail.com.
Jianguo Zhang, Email: jgzhang_pumch@yahoo.com.
References
- 1.Sylvestre PL, Villemure I, Aubin CE. Finite element modeling of the growth plate in a detailed spine model. Med Biol Eng Comput. 2007;45(10):977–88. [DOI] [PubMed] [Google Scholar]
- 2.Mohanty S, Kumar N. Patterns of presentation of congenital scoliosis. J Orthop Surg (Hong Kong). 2000;8(2):33–7. [DOI] [PubMed] [Google Scholar]
- 3.Wang S, Zhang J, Zhang Y, Du Y, Yang Y, Lin G. Outcomes of posterior lumbar hemivertebra resection and short fusion in patients with severe sacral tilt. Neurospine. 2021;18(3):562–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Wang S, Zhang J, Qiu G, Li S, Yu B, Weng X. Posterior hemivertebra resection with bisegmental fusion for congenital scoliosis: more than 3 year outcomes and analysis of unanticipated surgeries. Eur Spine J. 2013;22(2):387–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Mummaneni PV, Haid RW, Rodts GE. Lumbar interbody fusion: state-of-the-art technical advances. invited submission from the joint section meeting on disorders of the spine and peripheral nerves. J Neurosurg Spine. 2004;1(1):24–30. [DOI] [PubMed] [Google Scholar]
- 6.Srivastava SK, Marathe NA, Bhosale SK, Raj A, Purohit S, Shaikh A, et al. Outcome analysis of anterior reconstruction with rib grafts in tuberculosis of the thoracic spine. Asian J Neurosurg. 2020;15(3):648–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Zhang HQ, Li M, Wang YX, Tang MX, Guo CF, Liu SH, et al. Minimum 5-year follow-up outcomes for comparison between titanium mesh cage and allogeneic bone graft to reconstruct anterior column through posterior approach for the surgical treatment of thoracolumbar spinal tuberculosis with kyphosis. World Neurosurg. 2019;127:e407–15. [DOI] [PubMed] [Google Scholar]
- 8.Cha JR, Hwang IY, Kwon SH, Chung HY. Autogenous bone grafts versue metal cage with allogenic bone grafts for post-corpectomy anterior column reconstruction in patients with infectious spondylitis. J Korean Neurosurg Soc. 2020;63(2):218–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Wang S, Zhang J, Qiu G, Li S, Zhang Y, Yang Y, et al. Posterior-only hemivertebra resection with anterior structural reconstruction with titanium mesh cage and short segmental fusion for the treatment of congenital scoliokyphosis: the indications and preliminary results. Spine (Phila Pa 1976). 2017;42(22):1687–92. [DOI] [PubMed] [Google Scholar]
- 10.Bernhardt M, Bridwell KH. Segmental analysis of the sagittal plane alignment of the normal thoracic and lumbar spines and thoracolumbar junction. Spine (Phila Pa 1976). 1989;14(7):717–21. [DOI] [PubMed] [Google Scholar]
- 11.Busscher I, van Dieën JH, Kingma I, van der Veen AJ, Verkerke GJ, Veldhuizen AG. Biomechanical characteristics of different regions of the human spine: an in vitro study on multilevel spinal segments. Spine (Phila Pa 1976). 2009;34(26):2858–64. [DOI] [PubMed] [Google Scholar]
- 12.Zhang B-b, Zhang T, Tao H-r, Wu T-l, Duan C-g, Yang W-z, et al. Neurological complications of thoracic posterior vertebral column resection for severe congenital spinal deformities. Eur Spine J. 2017;26(7):1871–7. [DOI] [PubMed] [Google Scholar]
- 13.Cho W, Wang W, Bucklen B, Ramos RDlG, Yassari R. The role of biological fusion and anterior column support in a long lumbopelvic spinal fixation and its effect on the S1 screw—an in silico biomechanics analysis. Spine. 2021;46(4):E250–6. [DOI] [PubMed] [Google Scholar]
- 14.Lu X, Zou F, Lu F, Ma X, Xia X, Jiang J. How to reconstruct the lordosis of cervical spine in patients with Hirayama disease? A finite element analysis of biomechanical changes focusing on adjacent segments after anterior cervical discectomy and fusion. J Orthop Surg Res. 2022;17(1):101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Zou D, Sun Z, Zhou S, Zhong W, Li W. Hounsfield units value is a better predictor of pedicle screw loosening than the T-score of DXA in patients with lumbar degenerative diseases. Eur Spine J. 2020;29(5):1105–11. [DOI] [PubMed] [Google Scholar]
- 16.Wu H, Shan Z, Zhao F, Cheung JPY. Poor bone quality, multilevel surgery, and narrow and tall cages are associated with intraoperative endplate injuries and late-onset cage subsidence in lateral lumbar interbody fusion: a systematic review. Clin Orthop Relat Res. 2022;480(1):163–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.