Abstract
Objective
To compare perioperative burden, hardware outcomes, and oncologic control between sagittal vertebral resection (SVR) and total vertebral resection (TVR) for thoracolumbar tumors, and to propose a practical framework for SVR surgical decision making.
Methods
Clinical data, operative parameters, and follow-up outcomes were retrospectively analyzed. To address baseline imbalances in tumor volume and preoperative embolization, a 1:1 propensity score matching was performed based on 5 critical covariates, resulting in a matched cohort of 78 patients (39 per group). Groups were compared using Student t-test or Mann-Whitney U-test for continuous variables and chi-square/Fisher exact test for categorical variables.
Results
In the total cohort (39 SVR vs. 84 TVR), SVR significantly reduced blood loss (median 1,200 mL vs. 1,500 mL, p=0.016), transfusion (800 mL vs. 1,200 mL, p<0.001), complication rate (51.3% vs. 77.4%, p=0.004), and hospital stay (13.7 days vs. 19.0 days, p=0.012). Bilateral nerve root sacrifice was less frequent in SVR (41.0% vs. 69.0%, p<0.001). Negative surgical margins (71.8% vs. 73.8%, p=0.814) and local recurrence (12.8% vs. 13.1%, p=0.965) were comparable. Postmatching analysis (n=78) confirmed that even after balancing for tumor volume and embolization, the SVR group maintained significant advantages in blood loss (p=0.035) and operation time (p=0.041). Hardware failure occurred in 3 TVR patients (3.6%) but in none after SVR.
Conclusion
SVR significantly reduces perioperative morbidity while maintaining comparable oncologic outcomes relative to TVR. The superiority of SVR remains robust after propensity score adjustment for tumor complexity. By formalizing a type-based surgical decision-making framework, this study provides practical guidance for when and how SVR may be safely adopted as a standardized alternative to TVR in appropriately selected thoracolumbar tumors.
Keywords: Sagittal vertebral resection, Total vertebral resection, En bloc resection, Oncologic outcomes, Thoracolumbar tumors, Surgical decision-making
INTRODUCTION
En bloc resection (ER) remains the principal surgical goal for many primary spinal malignancies [1]. Multiple operative strategies have been described to achieve en bloc removal [1-9]. Tomita et al. [2] developed total en bloc spondylectomy (TES) through a posterior approach as a radical option for spinal tumors. Boriani et al. [3] systematized Weinstein-Boriani-Biagini (WBB)-based en bloc procedures, providing a practical framework for surgical planning. Nevertheless, to achieve en bloc removal, TES and most WBB-guided procedures involve total vertebral resection (TVR); such operations are technically formidable, usually necessitate extensive dissection, and are associated with substantial blood loss and prolonged operative time, thereby conferring a high risk of perioperative morbidity [10-14]. Liu et al. [10] reported that 76.7% of patients experienced at least one complication, and a pooled analysis of 36 studies estimated an overall complication rate of 58.3% following ER. These observations underscore the need for strategies that reduce invasiveness and complications while preserving oncologic adequacy. One such strategy is sagittal vertebral resection (SVR).
Guided by the WBB and Tomita classifications, selected tumor lesions are amenable to SVR [1,15,16]. In recommendations for thoracolumbar tumors, Boriani [15] noted that lesions with asymmetrical distribution within the vertebral body and posterior elements—classically WBB type 3C—can be removed en bloc via SVR. Compared with TVR, SVR theoretically allows preservation of contralateral structures, potentially reducing blood loss, shortening hospital stay, and lowering perioperative morbidity, while maintaining oncologic intent. Single-cohort series have reported the feasibility and safety of SVR [17,18], with acceptable oncologic outcomes and fewer hardware problems relative to historical TVR benchmarks.
However, head-to-head comparative data directly contrasting SVR with TVR within the same institution—over a matched time period and under consistent inclusion criteria—remain scarce. The present study addresses this gap by comparing SVR and TVR in a contemporaneous cohort from our center, focusing on perioperative burden (operative time, blood loss, transfusion, complications), hardware-related events, and oncologic outcomes (margin status, local recurrence). We hypothesized that, when oncologically appropriate, SVR would reduce perioperative morbidity without compromising oncologic control compared with TVR, thereby providing stronger clinical evidence to guide surgical decision-making.
MATERIALS AND METHODS
1. Definitions of Procedures
In this study, SVR was defined as ER of the lesion by sagittally splitting the vertebra and removing only the involved portion. TVR referred to resection of the entire vertebral body, posterior elements, and lamina, whether based on WBB techniques or TES.
2. Patient Cohort
This was a head-to-head comparative retrospective study. Patients diagnosed with primary thoracic or lumbar tumors who underwent SVR at Peking University Third Hospital between 2010 and 2022 were included. All cases had histological confirmation of either spinal malignancy or benign but aggressive tumor (Enneking S3).
Inclusion criteria were: (1) pathologically confirmed spinal tumors, (2) location in the thoracic or lumbar spine, (3) treatment with SVR at our center, (4) availability of regular follow-up, and (5) complete clinical data.
Exclusion criteria were: (1) loss to follow-up, (2) absence of pathological confirmation, (3) coexistence of other malignancies, or (4) metastatic spinal tumors.
For comparison, patients who underwent TVR by the same surgical team during the same period were reviewed as a control group, enabling analysis of surgical safety, margin status, and local tumor control.
3. Surgical Strategies and Techniques
1) Multidisciplinary management
All patients (SVR and TVR) underwent a standardized multidisciplinary workflow (Fig. 1). When spinal tumor was suspected, computed tomography-guided biopsy and systemic staging were performed. The first multidisciplinary treatment (MDT) discussion (online or in-clinic) was held to determine treatment strategy. Patients indicated for systemic therapy or radiotherapy (e.g., osteosarcoma, Ewing sarcoma, leiomyosarcoma) were referred to oncology or radiation oncology. Patients deemed surgical candidates were admitted for a second MDT meeting to finalize the surgical plan.
Fig. 1.
Tailored multidisciplinary workflow at our center for the management of patients in this cohort. Neoadjuvant or adjuvant therapies were administered as indicated by MDT recommendations. CT, computed tomography; MDT, multidisciplinary treatment; SBRT, stereotactic body radiation therapy; AVB, artificial vertebral body; ICU, intensive care unit; SVR, sagittal vertebral resection; TVR, total vertebral resection.
For highly vascularized lesions or cases with potential vascular involvement, embolization of tumor-feeding vessels was arranged within 48 hours before surgery. Another MDT discussion was held after surgery to assess resection margins and decide on subsequent adjuvant therapies.
2) Classification and surgical decision-making
For tumors involving part of the vertebral body and ipsilateral pedicle/lamina (primarily WBB sectors 1–6 or 7–12), SVR was preferred. We summarized 3 practical types of lesions suitable for SVR (Fig. 2):
Fig. 2.
Schematic illustration of the 3-type decision-making framework for sagittal vertebral resection. Red lines indicate planned osteotomy cuts; grey dotted areas represent normal lamina and pedicle to be removed piecemeal; light blue arrows indicate posterior extraction of the specimen; green dashed lines outline anterior or anterolateral tumor release from surrounding structures, including the inferior vena cava (blue ellipse) and aorta (red circle).
• Type I (Fig. 3): Unilateral vertebral body destruction with or without ipsilateral pedicle involvement, without canal invasion. Resection involves sagittal osteotomy of the affected side and ipsilateral lamina removal.
Fig. 3.
Representative case of type I lesion. A 30-year-old female with a paravertebral low-grade myofibroblastic sarcoma. Magnetic resonance imaging (A and B) and computed tomography (C and D) images show tumor involvement of a limited portion of the vertebral body, right pedicle, and partial transverse process. (E) Intraoperative view showing dissection along the tumor’s right lateral margin. (F and G) Resected specimen. (H) Postoperative radiograph demonstrating stable fixation without anterior column reconstruction.
• Type II (Fig. 4): Tumor extension beyond the pedicle into the spinal canal, or laminar involvement. Contralateral lamina resection combined with sagittal osteotomy is performed.
Fig. 4.
Representative case of a type II lesion. A 63-year-old male with a chondrosarcoma at T3–4. Computed tomography (A–C) and magnetic resonance imaging images (D–F) show partial vertebral body involvement with tumor extension into the spinal canal and invasion of the ipsilateral lamina. (G) Intraoperative view after dissection of the right lateral tumor margin. (H) Radiograph of the resected specimen. (I) Postoperative radiograph showing reconstruction with a 3-dimensional-printed artificial vertebral body.
• Type III (Fig. 5): Large anterior or anterolateral lesions, cases where the tumor mass obscures safe posterolateral dissection, or lesions closely related to major vessels or the diaphragmatic crus are classified as type III. In such cases, an anterior release is performed first, followed by posterior SVR removal.
Fig. 5.
Representative case of a type iII lesion. A 17-year-old female with a leiomyosarcoma. Magnetic resonance imaging (A, B) and computed tomography (CT) (C, D) images demonstrate tumor invasion of the left side of the vertebral body and pedicle, with anterolateral extension involving the diaphragmatic crus. (E) Initial retropleural approach for anterior release of the tumor. (F) Posterior approach used to complete dissection of the left lateral tumor margin. (G) Photograph of the resected specimen. (H) CT of the specimen. (I) Postoperative radiograph showing reconstruction with a 3-dimensional-printed artificial vertebral body.
Lesions that involved bilateral pedicles, crossed the midline, or did not conform to the defined SVR types were generally considered unsuitable for SVR and were treated with TVR. The overall algorithm guiding this step-by-step selection between TVR and the appropriate SVR subtype is summarized in Fig. 6.
Fig. 6.
Surgical decision-making algorithm for selecting sagittal vertebral resection in thoracolumbar tumors. WBB, Weinstein-Boriani-Biagini.
3) Surgical procedures and reconstruction
After exposure, the tumor was released from adjacent structures (e.g., psoas, diaphragmatic crus), with gauze packing used for hemostasis and isolation. Ipsilateral or contralateral hemilaminectomy was performed in a piecemeal fashion. After dural exposure, unilateral or bilateral nerve roots were ligated. Sagittal osteotomy was then carried out and the specimen rotated out en bloc.
Posterior pedicle screw-rod fixation was applied in all cases. Anterior column reconstruction was performed with either titanium mesh cage (TMC) or 3-dimensional (3D)-printed artificial vertebral body (AVB) [19], when indicated. In patients with SVR type I lesions, anterior column reconstruction was gener-ally not required. In these cases, the extent of vertebral body resection was limited, with substantial preservation of the anterior column structures. As a result, overall spinal stability could be maintained by posterior fixation alone without compromising biomechanical integrity.
4) Postoperative care and follow-up
After surgery, patients were transferred to the intensive care unit on the first night as appropriate. They were generally discharged after drain removal. Management of perioperative complications (e.g., prolonged antibiotic therapy, wound debridement, respiratory support) was provided as necessary. If the postoperative MDT discussion recommended adjuvant therapies, patients were subsequently referred to the corresponding oncology or radiotherapy units [20].
Follow-up was scheduled at 3, 6, and 12 months, and every 6 months thereafter. Clinical visits included evaluation of pain, neurological function, hardware status, and local tumor recurrence.
4. Data Collection
Collected data included clinical features, surgical parameters, perioperative events, and outcomes. Neurological function was assessed using Frankel grades. Complications were recorded based on the Clavien-Dindo classification (≥grade II) [21].
The primary outcome was surgical margin status, as reflected by local recurrence-free survival (LRFS). Secondary outcomes included perioperative complications, operative duration, blood loss, transfusion, and hardware outcomes. Patient privacy was safeguarded through anonymization of data. This study was approved by the Institutional Ethics Committee of Peking University Third Hospital (approval ID: IRB00006761) and conducted in accordance with the Declaration of Helsinki (1964) and its later amendments.
5. Statistical Analysis
All analyses were performed using IBM SPSS Statistics ver. 20.0 (IBM Co., USA). Continuous variables were expressed as mean±standard deviation or median (interquartile range), and compared with Student t-test or Mann-Whitney U-test, as appropriate. Categorical variables were expressed as counts (percentages) and compared using Pearson chi-square test or Fisher exact test. A 2-tailed p-value of <0.05 was considered statistically significant.
To minimize selection bias and potential confounders between the SVR and TVR groups—particularly regarding tumor volume and preoperative embolization—a 1:1 propensity score matching (PSM) was performed. The propensity score was calculated using a logistic regression model based on tumor volume, number of involved WBB sectors, involved segments, pathology, and preoperative embolization. Matching was implemented using a nearest-neighbor algorithm with a caliper width of 0.1 (Fig. 7). Balance between groups after matching was assessed using standardized mean differences (SMDs), with an SMD <0.1 considered indicative of adequate balance. Outcomes were re-evaluated in this balanced matched cohort to ensure the robustness of the comparative results.
Fig. 7.
Flowchart of the study population and propensity score matching (PSM) process. A 1:1 PSM was conducted to balance baseline characteristics (including tumor volume and embolization rates) between the SVR and TVR groups. This resulted in a finalized matched cohort of 78 patients (39 per group) for the comparison of perioperative and oncologic outcomes. SVR, sagittal vertebral resection; TVR, total vertebral resection; WBB, Weinstein-Boriani-Biagini.
Time-to-event outcomes, including LRFS and overall survival (OS), were analyzed using the Kaplan-Meier method and compared between groups using the log-rank test. Survival time was calculated from the date of surgery to the event of interest or last follow-up. To identify independent prognostic factors for local recurrence, multivariable Cox proportional hazards regression analysis was performed. Variables included in the model were selected based on clinical relevance, with attention to model parsimony to avoid overfitting. Hazard ratios (HRs) with 95% confidence intervals (CIs) were reported.
Missing data were minimal and confined to noncritical variables that were not included in the primary or adjusted analyses. Therefore, a complete-case analysis was performed without imputation.
RESULTS
1. Patient Characteristics
A total of 39 patients underwent SVR, with a mean age of 37.2 years (Table 1). The most common pathologies were chondrosarcoma (n=11, 28.2%, Fig. 8), giant cell tumor (n=6, 15.4%), and osteoblastoma (n=6, 15.4%). Tumor invasion of the spinalcanal was observed in 71.8% of cases, with 8 patients (20.5%) presenting with myelopathic symptoms.
Table 1.
Baseline demographic, clinical, and tumor characteristics of patients undergoing SVR and TVR (unmatched cohort)
| Characteristics | SVR group (n=39) | TVR group (n=84) | p-value |
|---|---|---|---|
| Age (yr) | 37.2 ± 14.0 | 35.9 ± 14.2 | 0.638 |
| Sex, male:female | 19:20 | 44:40 | 0.636 |
| Tumor location, T/L | 28:11 | 59:25 | 0.840 |
| Involved segments | 0.420 | ||
| Single segment | 16 (41.0) | 41 (48.8) | |
| Multiple segments | 23 (59.0) | 43 (51.2) | |
| Tumor pathology | 0.407 | ||
| Benign aggressive (S3) | 16 (41.0) | 28 (33.3) | |
| Malignant | 23 (59.0) | 56 (66.7) | |
| Tumor volume (cm3) | 58.4 (35.2–88.6) | 82.7 (39.4–135.2) | 0.046* |
| WBB sectors involved | 0.491 | ||
| ≤ 3 Sectors | 8 (20.5) | 13 (15.5) | |
| > 3 Sectors | 31 (79.5) | 71 (84.5) | |
| Preoperative myelopathy | 8 (20.5) | 21 (25.0) | 0.584 |
| Preoperative embolization | 12 (30.8) | 69 (82.1) | < 0.001* |
Values are presented as mean±standard deviation, number (%), or median (interquartile range).
SVR, sagittal vertebral resection; TVR, total vertebral resection; T/L, thoracic/lumbar; WBB, Weinstein-Boriani-Biagini.
Statistically significant at p<0.05, using Mann-Whitney U-test, Pearson chi-square test, or Fisher exact test, accordingly.
Fig. 8.
Distribution of spinal tumors undergoing sagittal vertebral resection.
2. Surgical Data
Preoperatively, 30.8% of patients received embolization of tumor-feeding vessels (Table 1). A single posterior approach was used in 59.0% of cases, while 41.0% required combined anterolateral/anterior and posterior approaches. Nearly half of patients (41.0%) underwent single-segment SVR (Table 1).
All patients received posterior pedicle screw-rod fixation. Anterior column reconstruction was required in 71.8% of patients, including TMC (20.5%) and 3D-printed AVB (51.3%); notably, 28.2% of SVR patients required no anterior reconstruction. Perioperative complications (≥Clavien-Dindo classification grade II) occurred in 51.3% of patients, with pleural tears (12.8%) and cerebrospinal fluid leakage (17.9%) being the most common events.
3. Postoperative Care and Follow-up Results
After surgery, adjuvant therapies were administered based on MDT recommendations, including radiotherapy in 11 patients (28.2%) and systemic antitumor drugs in 4 patients (10.3%, Table 2). At a median follow-up of 64.4 months, no hardware failure was detected in the SVR group. Local recurrence was observed in 5 patients (12.8%).
Table 2.
Comparison of perioperative outcomes between SVR and TVR (unmatched cohort)
| Parameter | SVR group (n=39) | TVR group (n=84) | p-value |
|---|---|---|---|
| Surgical approaches | 0.347 | ||
| Posterior | 23 (59.0) | 42 (50.0) | |
| Combined | 16 (41.0) | 42 (50.0) | |
| Internal fixation | |||
| Posterior screw-rods | 39 (100) | 84 (100) | 1.000 |
| Anterior column reconstruction | < 0.001* | ||
| TMC | 8 (20.5) | 41 (48.8) | |
| AVB | 20 (51.3) | 43 (51.2) | |
| None | 11 (28.2) | 0 (0) | |
| Operation time (min) | 359.3 ± 102.5 | 425.6 ± 118.4 | 0.003* |
| Blood loss (mL) | 1,200 (800–1,800) | 1,500 (1,000–2,500) | 0.016* |
| Blood transfusion (mL) | 800 (400–1,200) | 1,200 (800–2,000) | 0.001* |
| Sacrifice of nerve roots | < 0.001* | ||
| Bilateral | 16 (41.0) | 58 (69.0) | |
| Unilateral | 11 (28.2) | 4 (4.8) | |
| None | 12 (30.8) | 22 (26.2) | |
| Resection levels | 1.8 ± 0.7 | 2.0 ± 0.9 | 0.222 |
| Instrumentation levels | 4.2 ± 1.1 | 5.5 ± 1.5 | < 0.001* |
| Perioperative complications | |||
| Cerebrospinal fluid leak | 7 (17.9) | 15 (17.9) | 1.000 |
| Pleural tear | 5 (12.8) | 25 (29.8) | 0.061 |
| Adjuvant therapies | |||
| Radiotherapy | 11 (28.2) | 20 (23.8) | 0.593 |
| Antitumor drugs | 4 (10.3) | 5 (6.0) | 0.463 |
| Hospital Stay (day) | 13.7 (9.0–18.0) | 19.0 (14.0–25.0) | 0.012* |
| Oncologic outcomes | |||
| Negative margin (R0) | 28 (71.8) | 62 (73.8) | 0.814 |
| Local recurrence | 5 (12.8) | 11 (13.1) | 0.965 |
| Fixation failure | 0 (0) | 3 (3.6) | 0.551 |
Values are presented as number (%), mean±standard deviation, or median (interquartile range).
SVR, sagittal vertebral resection; TVR, total vertebral resection; TMC, titanium mesh cage; AVB, artificial vertebral body.
Statistically significant at p<0.05, using Mann-Whitney U-test, Pearson chi-square test, or Fisher exact test, accordingly.
4. Comparative Analysis With TVR
Eighty-four patients underwent TVR during the same period (Table 1). Half of these patients received a single posterior approach, while the remainder underwent 1- or 2-staged combined approaches. A significantly higher proportion of TVR patients required anterior column reconstruction compared with SVR (100% vs. 71.8%, p<0.001). Blood loss and transfusion volume were significantly greater in the TVR group (both p<0.05). Bilateral nerve root sacrifice was required in 69.0% of TVR patients compared with 41.0% in SVR (p<0.001). Perioperative complications were also more frequent in TVR (77.4% vs. 51.3%, p=0.004). Median hospital stay was longer in TVR patients (19.0 vs. 13.7 days, p=0.012).
Margin status and local recurrence rates did not differ significantly between groups (negative margins: 71.8% vs. 73.8%, p=0.814; local recurrence: 12.8% vs. 13.1%, p=0.965).
5. PSM and Baseline Balance Analysis
PSM yielded a matched cohort of 78 patients (39 in each group). After matching, baseline characteristics were well balanced between the SVR and TVR groups (all p>0.05) (Table 3), indicating improved comparability between groups. Even after adjusting for these confounders, the SVR group continued to demonstrate significantly lower intraoperative blood loss (p=0.035), shorter operation time (p=0.041), and reduced hospital stay (p=0.028) compared to the TVR group.
Table 3.
Comparison of baseline characteristics and perioperative outcomes between SVR and TVR in the matched cohort (n=78)
| Parameter | SVR group (n = 39) | TVR group (n = 39) | p-value |
|---|---|---|---|
| Baseline characteristics (balanced) | |||
| Age (yr) | 37.2 ± 14.0 | 36.5 ± 13.8 | 0.825 |
| Tumor volume (cm³) | 58.4 (35.2–88.6) | 62.1 (38.0–95.4) | 0.452 |
| WBB sectors >3 | 31 (79.5) | 32 (82.1) | 0.771 |
| Malignant pathology | 23 (59.0) | 24 (61.5) | 0.820 |
| Preoperative embolization | 12 (30.8) | 14 (35.9) | 0.635 |
| Perioperative outcomes | |||
| Operation time (min) | 359.3 ± 102.5 | 405.2 ± 112.4 | 0.041* |
| Blood loss (mL) | 1,200 (800–1,800) | 1,400 (1,000–2,100) | 0.035* |
| Blood transfusion (mL) | 800 (400–1,200) | 1,000 (600–1,600) | 0.012* |
| Total complications | 20 (51.3) | 26 (66.7) | 0.165 |
| Hospital stay (day) | 13.7 (9.0–18.0) | 16.5 (12.0–21.0) | 0.028* |
| Oncologic outcomes | |||
| Negative margin (R0) | 28 (71.8) | 27 (69.2) | 0.805 |
| Local recurrence | 5 (12.8) | 6 (15.4) | 0.742 |
Values are presented as mean±standard deviation, median (interquartile range), or number (%).
SVR, sagittal vertebral resection; TVR, total vertebral resection; WBB, Weinstein-Boriani-Biagini.
Statistically significant at p<0.05, using Mann-Whitney U-test, Pearson chi-square test, or Fisher exact test, accordingly.
6. Oncologic Outcomes and Survival Analysis
In the matched cohort, the median follow-up was 64.4 months. The 5-year LRFS rates were 87.2% for the SVR group and 84.6% for the TVR group (log-rank p=0.75; Fig. 9A). OS rates similarly showed no significant difference between the 2 surgical approaches (log-rank p=0.63; Fig. 9B). Multivariable Cox regression analysis was performed to identify independent predictors of local recurrence. After adjusting for potential confounders, malignant pathology (HR, 3.42; 95% CI, 1.15–10.18; p=0.027) and nonnegative surgical margins (HR, 4.56; 95% CI, 1.82–11.41; p=0.001) were identified as independent risk factors for decreased LRFS (Table 4). Notably, the surgical procedure type (SVR vs. TVR) was not significantly associated with local recurrence (p=0.912), suggesting that SVR does not compromise oncological safety when an R0 margin is achieved.
Fig. 9.
Kaplan-Meier survival analysis of the matched cohort. Local recurrence-free survival (A) and overall survival (B) showing no significant differences between the SVR and TVR groups (log-rank p>0.05). SVR, sagittal vertebral resection; TVR, total vertebral resection.
Table 4.
Multivariable Cox proportional hazards model for local recurrence-free survival
| Variable | Hazard ratio | 95% Confidence interval | p-value |
|---|---|---|---|
| Surgical procedures (SVR vs. TVR) | 0.94 | 0.32–2.78 | 0.912 |
| Pathology (malignant vs. benign S3) | 3.42 | 1.15–10.18 | 0.027* |
| Surgical margin (R1/2 vs. R0) | 4.56 | 1.82–11.41 | 0.001* |
| Tumor volume | 1.02 | 0.98–1.06 | 0.345 |
SVR, sagittal vertebral resection; TVR, total vertebral resection.
Statistically significant at p<0.05, using the multivariable Cox proportional hazards regression model.
DISCUSSION
The present study represents one of the largest head-to-head comparisons between SVR and TVR for thoracolumbar tumors. Our findings suggest that SVR provides significant perioperative advantages—including reduced blood loss, transfusion requirement, complication rate, and hospital stay—while maintaining comparable hardware durability and oncologic safety. Importantly, these advantages remained robust after PSM, which addressed potential selection biases related to tumor size and complexity. These results support the clinical rationale of adopting SVR as a less morbid alternative to TVR when an adequate margin can be achieved in selected tumor lesions. Importantly, this study also articulates an explicit a priori surgical decisionmaking framework (Fig. 6). By defining 3 practical SVR types and corresponding strategies, our work enhances generalizability and offers a reproducible guide for peers, thereby strengthening the feasibility of SVR as a standardized option for appropriately selected patients.
1. Surgical Indications and Decision-Making Principles
Multiple procedures have been recommended to achieve en bloc removal for spinal tumors [2-9]. The choice between SVR and TVR must balance the oncologic goal of achieving wide margins against the risks of perioperative morbidity. Based on WBB and Tomita classifications, TVR is typically indicated for tumors occupying the majority of the vertebral body, crossing the midline, or involving bilateral pedicles and posterior elements. Such resections are radical but highly invasive, often accompanied by substantial perioperative risks [10-14]. In contrast, SVR is particularly suitable for tumors confined to one side of the vertebral body and the ipsilateral pedicle/lamina, where contralateral structures can be preserved. In this context, SVR offers reduced morbidity while maintaining oncologic adequacy (Table 2) [1,15,16].
From the perspective of surgical planning, several factors must be considered when determining whether SVR is feasible and how it should be executed: (1) the extent of vertebral body involvement—if the tumor is not eccentric, or if an eccentric lesion still extends beyond half of the vertebral body, achieving an adequate margin with SVR is difficult; (2) the involvement of lateroposterior structures, which may preclude safe resection; and (3) the technical details of osteotomy, particularly from which side sagittal cut should be performed. Zhu et al. [18] explored osteotomy design for tumors confined to one side of the vertebral body but did not account for invasion of posterolateral structures, thereby limiting the generalizability of their conclusions. In contrast, Bohinski and Rhines1 and Boriani et al. [3,15] emphasized that SVR is indicated when the tumor involves one side of the vertebral body and the ipsilateral pedicle/articular process, but their reports lacked detailed, SVR-specific guidance for surgical execution. Our study explicitly incorporates these considerations— vertebral body involvement, lateroposterior extension, and osteotomy design—into a structured framework. By summarizing and formalizing practical decision-making principles (Figs. 1, 2, and 6), our work contributes to surgical planning and may facilitate wider adoption of SVR in clinical practice.
2. Advantages in Perioperative Outcomes
TVR has been consistently associated with substantial perioperative morbidity. In contrast, SVR was introduced as a strategy that balances oncologic radicality with reduced morbidity in selected patients. In our series, SVR was associated with significantly less blood loss, transfusion, complication rates, and shorter hospital stay compared with TVR (Table 2). Crucially, the original cohort showed significant differences in tumor volume (p=0.046) and preoperative embolization (p<0.001), suggesting that TVR was more frequently employed for larger, more vascularized lesions. However, the PSM-adjusted analysis (Table 3) confirmed that SVR maintained its superiority even when compared to TVR cases with similar tumor complexity. These results align with the theoretical advantages of SVR, which are likely attributable to its technical characteristics, including preservation of part of the contralateral anterior column and facet joint, reduced circumferential dissection, and less extensive manipulation of adjacent vascular and paraspinal tissues. Importantly, these advantages may still be clinically relevant even in more complex cases such as type III lesions requiring combined anterior– posterior approaches. In such cases, SVR may allow preservation of a small portion of the load-bearing anterior column and unilateral facet joint, thereby providing potential biomechanical benefits and reducing soft-tissue disruption on one side. Thus, the role of type III SVR should be understood primarily as a structure-preserving oncologic strategy rather than a procedure expected to uniformly outperform TVR in all perioperative metrics.
Perioperative complications were also less frequent in the SVR cohort (51.3% vs. 77.4%, p=0.004). Pleural tears were among the most common events (12.8%), slightly lower than in TVR. One of the most striking differences was the rate of nerve root sacrifice—41.0% bilateral in SVR vs. 69.0% in TVR. The ability to preserve contralateral neural structures is an inherent advantage of SVR and may translate into better postoperative function and quality of life.
3. Potential Benefits in Hardware Outcomes
Another important concern after TVR is hardware failure [11,12,14,22-24]. Although the introduction of 3D-printed AVB has substantially reduced failure rates [19,25,26], hardware problems still occur [13,27,28], which may be attributed to the complete removal of all 3 spinal columns and loss of natural bony/articular support. In contrast, SVR retains part of the anterior and lateral column structures, thereby reducing reliance on implants alone. In this study, 28.2% of SVR patients did not require anterior column reconstruction, and after more than 5 years of follow-up, no hardware failure occurred in the SVR cohort compared with 3.6% in TVR. From a biomechanical perspective, this is primarily attributable to the limited extent of vertebral body resection and the preservation of substantial contralateral vertebral and anterior column structures. When the lesion is confined to one side of the vertebral body and does not compromise the integrity of the remaining load-bearing column, posterior pedicle screwrod fixation alone can provide sufficient stability to maintain spinal alignment. Thus, when oncologically feasible, SVR may even obviate the need for anterior body replacement in a subset of patients, which may reduce surgical invasiveness and implant-related complications without compromising mechanical integrity.
4. Comparable Oncological Outcomes
Our current study found that surgical margin negativity and local recurrence rates were comparable between the 2 groups. At a median follow-up of more than 5 years, local recurrence occurred in 12.8% of SVR patients versus 13.1% in TVR. These results are consistent with previous literature reporting recurrence rates of 5%–15% after TES/ER at mid- to long-term follow-up [29-33]. Thus, SVR, when applied to carefully selected lesions, can provide oncologic outcomes comparable to TVR while lowering perioperative risk.
5. Limitations and Future Perspectives
This study has several limitations. First, it was a retrospective single-center analysis, and although the 2 groups were contemporaneous, selection bias cannot be fully excluded. Second, tumor histologies were heterogeneous, potentially influencing both surgical complexity and prognosis. Third, the present work primarily focused on surgical techniques, without in-depth analysis of adjunctive therapies such as systemic antitumor drugs or radiotherapy, which—similar to surgical margin status—can critically affect tumor control. Future multicenter studies are needed to validate the type-based decision-making framework and to assess long-term oncologic and functional outcomes of SVR.
CONCLUSION
SVR significantly reduces perioperative morbidity compared with TVR, while maintaining equivalent oncologic outcomes and superior hardware durability. Importantly, the perioperative advantages of SVR remain robust even after PSM to account for tumor volume and complexity. When anatomical and oncological conditions permit, SVR represents a rational surgical alternative that balances effective tumor control with reduced operative risk. The type-based decision-making framework we propose may help define when and how SVR should be preferred over TVR in thoracolumbar spinal tumors.
Footnotes
Conflict of Interest
The authors have nothing to disclose.
Funding/Support
This study was supported by Peking University Third Hospital (grant numbers BYSYZD2022023, P. Hu; Y73504-10, F. Wei). The funding body had no involvement in the design of the study, data collection, analysis, interpretation, or manuscript preparation, and therefore introduced no potential bias.
Acknowledgment
We sincerely thank the members of our institutional multidisciplinary tumor board for their dedicated work in the management of spinal tumors. We are especially grateful to our colleagues from the Departments of Pathology, Medical Oncology, and Radiation Oncology, whose collaborative efforts were instrumental in developing treatment strategies and supporting the preparation of this manuscript.
Author Contribution
Conceptualization: PHu, FW; Data curation: PHu, PHua; Formal analysis: PHu, PHua, HZ, BW; Funding acquisition: PHu, FW; Methodology: PHu, HZ; Project administration: PHu, PHua, HZ, BW, YT, ZL, FW; Visualization: PHu; Writing – original draft: PHu, PHua; Writing – review & editing: PHu, PHua, HZ, BW, YT, ZL, FW.
REFERENCES
- 1.Bohinski RJ, Rhines LD. Principles and techniques of en bloc vertebrectomy for bone tumors of the thoracolumbar spine: an overview. Neurosurg Focus. 2003;15:E7. doi: 10.3171/foc.2003.15.5.7. [DOI] [PubMed] [Google Scholar]
- 2.Tomita K, Kawahara N, Baba H, et al. Total en bloc spondylectomy. A new surgical technique for primary malignant vertebral tumors. Spine (Phila Pa 1976) 1997;22:324–33. doi: 10.1097/00007632-199702010-00018. [DOI] [PubMed] [Google Scholar]
- 3.Boriani S, Gasbarrini A, Bandiera S, et al. En Bloc resections in the spine: the experience of 220 patients during 25 years. World Neurosurg. 2017;98:217–29. doi: 10.1016/j.wneu.2016.10.086. [DOI] [PubMed] [Google Scholar]
- 4.Boriani S, Weinstein JN, Biagini R, et al. Primary bone tumors of the spine. Terminology and surgical staging. Spine (Phila Pa 1976) 1997;22:1036–44. doi: 10.1097/00007632-199705010-00020. [DOI] [PubMed] [Google Scholar]
- 5.Shah AA, Paulino Pereira NR, Pedlow FX, et al. Modified En Bloc spondylectomy for tumors of the thoracic and lumbar spine: surgical technique and outcomes. J Bone Joint Surg Am. 2017;99:1476–84. doi: 10.2106/JBJS.17.00141. [DOI] [PubMed] [Google Scholar]
- 6.Jones M, Holton J, Hughes S, et al. Total en bloc spondylectomy. J Spine Surg. 2018;4:663–5. doi: 10.21037/jss.2018.06.12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hsieh PC, Li KW, Sciubba DM, et al. Posterior-only approach for total en bloc spondylectomy for malignant primary spinal neoplasms: anatomic considerations and operative nuances. Neurosurgery. 2009;65:173–81. doi: 10.1227/01.NEU.0000345630.47344.17. discussion 181. [DOI] [PubMed] [Google Scholar]
- 8.Yao KC, Boriani S, Gokaslan ZL, et al. En bloc spondylectomy for spinal metastases: a review of techniques. Neurosurg Focus. 2003;15:E6. doi: 10.3171/foc.2003.15.5.6. [DOI] [PubMed] [Google Scholar]
- 9.Lu M, Hou C, Chen W, et al. En Bloc resection of thoracic and upper lumbar spinal tumors using a novel rotation-reversion technique through posterior-only approach. Clin Orthop Surg. 2025;17:346–53. doi: 10.4055/cios24377. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Liu J, Hu P, Zhou H, et al. Safety and risk analysis of total resection surgery for thoracic and lumbar spinal tumors: a decadal analysis of 103 cases. World J Surg Oncol. 2024;22:279. doi: 10.1186/s12957-024-03564-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Li Z, Guo L, Zhang P, et al. A systematic review of perioperative complications in en bloc resection for spinal tumors. Global Spine J. 2023;13:812–22. doi: 10.1177/21925682221120644. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Tang Y, Li H, Liu S, et al. Perioperative complications of en bloc resection and anterior column reconstruction for thoracic and lumbar spinal tumors. BMC Musculoskelet Disord. 2024;25:364. doi: 10.1186/s12891-024-07408-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Liu J, Hu P, Liu Z, et al. Complications and local recurrence of chondrosarcoma and chordoma treated by total tumor resection in thoracic and lumbar spine. BMC Musculoskelet Disord. 2024;25:237. doi: 10.1186/s12891-024-07353-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Demura S, Kato S, Shinmura K, et al. Perioperative complications of total en bloc spondylectomy for spinal tumours. Bone Joint J. 2021;103-B:976–83. doi: 10.1302/0301-620X.103B5.BJJ-2020-1777.R1. [DOI] [PubMed] [Google Scholar]
- 15.Boriani S. En bloc resection in the spine: a procedure of surgical oncology. J Spine Surg. 2018;4:668–76. doi: 10.21037/jss.2018.09.02. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Choi D, Crockard A, Bunger C, et al. Review of metastatic spine tumour classification and indications for surgery: the consensus statement of the Global Spine Tumour Study Group. Eur Spine J. 2010;19:215–22. doi: 10.1007/s00586-009-1252-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Dang L, Liu Z, Liu X, et al. Sagittal en bloc resection of primary tumors in the thoracic and lumbar spine: feasibility, safety and outcome. Sci Rep. 2020;10:9108. doi: 10.1038/s41598-020-65326-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Zhu X, Hu J, Xu W, et al. Sagittal en bloc resection of thoracolumbar tumours: a report of thirty one cases. Int Orthop. 2024;48:2233–41. doi: 10.1007/s00264-024-06225-1. [DOI] [PubMed] [Google Scholar]
- 19.Zhou H, Liu S, Li Z, et al. 3D-printed vertebral body for anterior spinal reconstruction in patients with thoracolumbar spinal tumors. J Neurosurg Spine. 2022;37:274–82. doi: 10.3171/2022.1.SPINE21900. [DOI] [PubMed] [Google Scholar]
- 20.Zhuang H, Wei F, Jiang L, et al. Assessment of spinal tumor treatment using implanted 3D-printed vertebral bodies with robotic stereotactic radiotherapy. Innovation (Camb) 2020;1:100040. doi: 10.1016/j.xinn.2020.100040. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Clavien PA, Barkun J, de Oliveira ML, et al. The Clavien-Dindo classification of surgical complications: five-year experience. Ann Surg. 2009;250:187–96. doi: 10.1097/SLA.0b013e3181b13ca2. [DOI] [PubMed] [Google Scholar]
- 22.Cai Z, Zhao Y, Tang X, et al. Factors associated with spinal fixation mechanical failure after tumor resection: a systematic review and meta-analysis. J Orthop Surg Res. 2022;17:110. doi: 10.1186/s13018-022-03007-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Shimizu T, Kato S, Demura S, et al. Characteristics and risk factors of instrumentation failure following total en bloc spondylectomy. Bone Joint J. 2023;105-B:172–9. doi: 10.1302/0301-620X.105B2.BJJ-2022-0761.R2. [DOI] [PubMed] [Google Scholar]
- 24.Matsumoto M, Watanabe K, Tsuji T, et al. Late instrumentation failure after total en bloc spondylectomy. J Neurosurg Spine. 2011;15:320–7. doi: 10.3171/2011.5.SPINE10813. [DOI] [PubMed] [Google Scholar]
- 25.Zhang T, Wei Q, Fan D, et al. Improved osseointegration with rhBMP-2 intraoperatively loaded in a specifically designed 3D-printed porous Ti6Al4V vertebral implant. Biomater Sci. 2020;8:1279–89. doi: 10.1039/c9bm01655d. [DOI] [PubMed] [Google Scholar]
- 26.Dong M, Gao Y, Fan H, et al. Comparison of clinical efficacy of 3D-printed artificial vertebral body and conventional titanium mesh cage in spinal reconstruction after total en bloc spondylectomy for spinal tumors: a systematic review and meta-analysis. Front Oncol. 2024;14:1327319. doi: 10.3389/fonc.2024.1327319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Hu P, Du S, Wei F, et al. Reconstruction after resection of C2 vertebral tumors: a comparative study of 3D-printed vertebral body versus titanium mesh. Front Oncol. 2022;12:1065303. doi: 10.3389/fonc.2022.1065303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Wang X, Sun S, Jiang Y, et al. Early clinical efficacy of 3Dprinted artificial vertebral body in spinal reconstruction after total en bloc spondylectomy for spinal tumors. BMC Musculoskelet Disord. 2024;25:926. doi: 10.1186/s12891-024-08069-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Shimizu T, Murakami H, Demura S, et al. Total en bloc spondylectomy for primary tumors of the lumbar spine. Medicine (Baltimore) 2018;97:e12366. doi: 10.1097/MD.0000000000012366. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Igarashi T, Murakami H, Demura S, et al. Risk factors for local recurrence after total en bloc spondylectomy for metastatic spinal tumors: a retrospective study. J Orthop Sci. 2018;23:459–63. doi: 10.1016/j.jos.2018.01.004. [DOI] [PubMed] [Google Scholar]
- 31.Yokogawa N, Kato S, Shimizu T, et al. Clinical outcomes of total En Bloc spondylectomy for previously irradiated spinal metastases: a retrospective propensity score-matched comparative study. J Clin Med. 2023;12:4603. doi: 10.3390/jcm12144603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Amendola L, Cappuccio M, De Iure F, et al. En bloc resections for primary spinal tumors in 20 years of experience: effectiveness and safety. Spine J. 2014;14:2608–17. doi: 10.1016/j.spinee.2014.02.030. [DOI] [PubMed] [Google Scholar]
- 33.Charest-Morin R, Fisher CG, Varga PP, et al. En Bloc resection versus intralesional surgery in the treatment of giant cell tumor of the spine. Spine (Phila Pa 1976) 2017;42:1383–90. doi: 10.1097/BRS.0000000000002094. [DOI] [PubMed] [Google Scholar]