Total en bloc spondylectomy for primary tumors of the lumbar spine.

This was a retrospective clinical study.This study aimed to evaluate our institution's experience with total en bloc spondylectomy (TES) in patients treated for primary lumbar spine tumors and investigate postoperative clinical outcomes.TES is a widely accepted by spinal and musculoskeletal surgical oncologists and results in favorable health-related quality of life outcomes. However, this procedure still imposes major risks and complications.The cases of TES performed for primary lumbar spine tumors between 1993 and 2015 were retrospectively analyzed. Primary outcome measures were the rates of perioperative complications and reoperation for instrumentation failure.We enrolled 30 patients (13 men and 17 women; median age and follow-up, 38 years and 87 months, respectively). Three, 7, and 5 cases involved previous radiotherapy, intralesional resection, and chemotherapy, respectively. The most common tumor was giant cell tumor (14 cases) followed by osteosarcoma (4 cases) and plasmacytoma (3 cases). The median estimated blood loss was 1450 mL, and the median operative time was 11 hours. At least 1 perioperative complication occurred in 26 patients (86.7%), with the most common being postoperative muscle weakness (24 patients, 80.0%) followed by surgical site infection and postoperative cerebrospinal fluid leakage (7 patients, respectively; 23.3% each). Revision surgery for instrumentation failure was required in 6 patients (20.0%) at a median of 33 months after the index TES. Four patients experienced local tumor recurrence (13.3%), and their 10-year disease-free rate was 75.0%.TES is a feasible and effective procedure for primary lumbar spine tumors, but the risks of perioperative complications and late instrumentation failure should be acknowledged. Surgical oncologic outcomes were good, especially in patients who underwent TES as their first surgical treatment. Therefore, being familiar with the indications for TES and the surgical technique is important.

Introduction

Total en bloc spondylectomy (TES) is a relatively aggressive surgical treatment for spinal neoplasms [ and involves en bloc removal of an entire vertebral body and posterior elements to achieve tumor resection with negative margins. This procedure results in low recurrence rates and favorable health-related quality of life outcomes [ and is now widely accepted by spinal and musculoskeletal surgical oncologists. With advances in surgical techniques,[ TES indications have been expanded to include patients with extra-compartmental or consecutive multilevel spinal tumors.[ However, the procedure still imposes major risks including spinal cord injury, pleural effusion, and postoperative cerebrospinal fluid leakage.[ Additionally, late instrumentation failure reportedly occurs in approximately 40% of patients undergoing TES.[ Due to the unique anatomy of the lumbar region, lumbar spine TES remains a challenge, and the close relationship between the vertebrae and abdominal structures results in the risk of a major vessel, lumbar plexus, or bowel injury. Unlike thoracic spine TES performed with transection of nerve roots, lumbar spine TES usually necessitates extensive nerve root dissection with frequent retraction to preserve lower extremity motor function. Thus, TES in the lumbar spine typically requires an anterior-posterior combined procedure, while thoracic spine TES can often be performed using a solely posterior approach. For these reasons, the majority of published accounts describing lumbar spine TES performed for primary tumors are case reports or small case series.[ Therefore, this study aimed to assess our institution's experience with TES performed in cases of primary aggressive benign and malignant lumbar spine tumors and investigate the rates of postoperative complications and revision surgery for instrumentation failure and oncological outcomes.

Materials and methods

Study design and inclusion criteria

We performed a retrospective chart review of all cases of surgically treated primary malignant or locally aggressive benign spinal tumor at our institution between 1993 and 2015. The patients who underwent TES for treatment of a tumor primarily located in the lumbar spine (L1–L5) were included. In cases of solitary plasmacytoma for which the primary treatment is local radiotherapy,[ TES was performed where severe spinal instability or neurologic impairment was identified. The patients who underwent anterior en bloc corpectomy alone or en bloc resection of only the posterior elements were excluded. Our institution's ethics committee approved the study, and informed consent was obtained from all patients.

Recorded data

Patients’ data including age, sex, history of intralesional resection, previous chemotherapy and radiotherapy, and tumor histology were obtained from clinical notes. Surgical data such as the approach, resected vertebrae, operative time, estimated blood loss (EBL), and occurrence of surgical complications were gathered from operative and follow-up clinical notes. Our primary outcome measures were the rates of perioperative complications and reoperation for instrumentation failure. The follow-up protocol included an examination with x-ray imaging, a computed tomography scan, and magnetic resonance imaging every 3 months for the first year, every 6 months for the second year, and yearly thereafter. Secondary outcome measures included local recurrence and disease-free survival.

Statistical analysis

General descriptive statistics were performed for the study population. Data are presented as proportions or median values with interquartile ranges (IQR). A simple logistic regression analysis was performed to investigate the relationship between specific patient and operative parameters and instrumentation failure occurrence. Results are presented as odds ratios (OR) with 95% confidence intervals (CI). All statistical analyses were performed using SPSS version 20 (IBM Corp., Armonk, NY). A probability value (P) of <.05 was considered statistically significant.

Results

Patient data

A total of 30 patients were included in this study, and Table 1 shows their characteristics. The median age at the time of surgery was 38 years (IQR: 32–48), and 13 patients were men (43.3%). Three (10.0%) and 5 (16.7%) patients had received previous radiotherapy or chemotherapy, respectively, and 7 (23.3%) had a history of intralesional resection. The most common tumor was giant cell tumor (14 cases, 46.7%) followed by osteosarcoma (4 cases, 13.3%); plasmacytoma (3 cases, 10.0%); aggressive hemangioma, chondrosarcoma, and chordoma (2 cases, respectively; 6.7% each); and hemangiopericytoma, synovial sarcoma, and osteoblastoma (1 case, respectively; 3.3% each). Twenty-four patients (80.0%) underwent single-level TES, 2 (6.7%) underwent 2-level TES, and 4 (13.3%) underwent 3-level TES. The median follow-up time in all patients was 87 months (IQR: 37–119).

Table 1

Clinical and surgical characteristics in 30 patients with primary lumbar spine tumors treated with total en bloc spondylectomy.

Surgical procedure

The lumbar spine tumor TES technique was previously described.[ Twenty-two cases (73.3%) required a combined anterior-posterior approach. Twelve and 10 cases were single-stage and 2-stage surgeries, respectively. The remaining 8 patients underwent TES with a single-stage solely posterior approach. Anterior retroperitoneal and anterior transperitoneal approaches were utilized in 16 (53.3%) and 6 patients (20.0%), respectively, with the aid of a general surgeon. Anterior reconstruction was performed using a titanium mesh cage enclosing a locally obtained autograft or autologous iliac bone in all patients. Pediculotomies were accomplished using a T-saw. Posterior reconstruction was performed using a combination of pedicle screws, sacral screws, iliac screws, rods, and transverse connectors (Fig. 1). The median EBL was 1450 mL (IQR: 670–3100), and the median operative time was 11.0 hours (IQR: 9.9–19.6).

Figure 1

A giant cell tumor of the lumbar spine in a 38-year-old female. T1-gadolinium-enhanced magnetic resonance imaging shows the L4 vertebral tumor extending to neighboring vertebrae (A) and largely expanding outside the vertebral body. B, Postoperative imaging after total en bloc spondylectomy with a posterior-anterior combined approach shows the paravertebral tumor and the vertebral bodies of L3 and L4, and half of L5 were removed en bloc with a marginal margin. C, Images show the spinal reconstruction with a titanium mesh cage enclosing an autologous bone graft after posterior instrumentation (D and E).

Complications

Twenty-six patients (86.7%) developed at least 1 perioperative complication (Table 2), with the most common being postoperative lower extremity muscle weakness (24 patients, 80.0%) followed by surgical site infection (SSI) and postoperative cerebrospinal fluid (CSF) leakage (7 patients, respectively; 23.3% each). Postoperative muscle weakness occurred in all cases of TES at L3 or below and in half of those at L1 or L2. Of the 24 patients with postoperative muscle weakness, 19 completely recovered within 6 months postoperatively, and all patients could walk without any support at the last follow-up appointment. All 7 patients with an SSI required reoperation for debridement, and 1 required an exchange of spinal instrumentation. All SSIs had resolved by the final follow-up examination. No patient experienced a fatal pulmonary embolism or major vessel injury.

Table 2

Perioperative complications in 30 patients after total en bloc spondylectomy for a primary lumbar spine tumor.

Instrumentation failure and oncological outcomes

Revision surgery for instrumentation failure was required in 6 patients (20.0%) at a median time of 33 months (IQR: 29–39) after the index TES. Logistic regression analysis showed no significant factors associated with instrumentation failure (Table 3). During the follow-up period, local tumor recurrence occurred in 4 patients (13.3%), and all had undergone previous intralesional resection. Three (1 and 2 cases of chondrosarcoma and osteosarcoma, respectively) of the 4 patients with a local recurrence died from the disease, while 1 locally recurrent giant cell tumor was controlled with denosumab. Four patients (13.3%) died from their disease, and the 10-year disease-free rate was 75.0% (Fig. 2).

Table 3

Factors associated with instrumentation failure after primary lumbar spine tumor total en bloc spondylectomy.

Figure 2

Kaplan–Meier survival curves show disease-free survival rates in all patients.

Discussion

We investigated our institution's 22-year experience with TES in patients treated for primary tumors located in the lumbar spine and assessed postoperative clinical outcomes to augment the currently limited number of published accounts describing this challenging procedure.

Sciubba et al[ reported outcomes in 23 patients who underwent TES for an aggressive or malignant primary tumor of the lumbar spine. All but 1 underwent a 2-stage posterior-anterior approach; the median total EBL was 3200 mL, and the median total operative time was 18.5 hours. Fifteen patients developed at least 1 perioperative complication (65.2%). There were 6 cases of wound infection and ileus (26.1% each), 4 of deep vein thrombosis with pulmonary embolism (17.4%), and 3 of CSF leakage (13.0%). There were 9 cases of instrumentation failure requiring revision surgery (39.1%) at a median time of 23 months after the index spondylectomy.[ Compared to the complication rate reported by Sciubba et al,[ the postoperative complication rate in our series was considerably higher at 86.7% versus 65.2%. The most common complication was postoperative muscle weakness (24 patients, 80.0%). Further, postoperative muscle weakness occurred in all cases of TES at L3 or below. We hypothesize that the incidence of muscle weakness after TES at L1 and L2 was lower than after TES at L3 or below because no muscle is dominated by the L1 or L2 nerve root alone and compensation by the other nerve roots masked the weakness.[ In previous reports, postoperative muscle weakness after lumbar TES was insufficiently investigated or not mentioned. The cause of the weakness must be extensive nerve root dissection with frequent retraction and detachment of the iliopsoas muscles. As discussed above, lumbar spine TES is performed through a surgical window formed by widely dissected and retracted nerve roots, the development of which strains the nerve roots. Although this is an inevitable feature of lumbar spine TES, the majority of patients who developed lower extremity weakness recovered within 6 months postoperatively, and all patients were able to walk at their final follow-up appointment. However, postoperative muscle weakness is an important problem, and future improvement is necessary.

The second most common complications were SSI and CSF leakage, each seen in 23.3% of patients, rates similar to those of previous reports.[ Patients often undergo chemo- or radiotherapy before TES, and both have been shown to impair wound healing.[ Additionally, a large skin incision, extensive damage to soft tissues, and a sizable dead space inhibit wound healing.[ Hayashi et al[ found that a combined approach to TES was an independent risk factor for SSI, and long operative time has also been associated with SSI after TES. Therefore, lumbar spine TES is thought to impose a high risk of SSI because it is a lengthy operation and often requires an anterior/posterior combined approach. However, the median total operative time in our study was approximately 11 hours, shorter than that of previous reports, but we did not observe a corresponding reduction in infection rate. Although some have suggested a posterior-only approach as a means to reduce operative time,[ a combined approach is often necessary, especially at L2 or below, to ensure surgical safety. Although no major vessel, nerve root, or bowel injuries occurred in our series, there have been reports of major vessel injury during lumbar spine TES.[ The high risk of major vessel injury in reoperation and post-irradiation cases is considered to be a result of tissue adhesion and scarring. Particularly in these cases, adding an anterior approach is important to avoid injury to major vessels.[

The rate of instrumentation failure requiring revision surgery was 20.0% in this study. Although logistic regression analysis showed no significant factors associated with instrumentation failure, the failure rate after multilevel resection was much higher than single vertebra resection (50.0% vs 12.5%), a result similar to that of previous reports.[ The lack of statistical significance was probably due to our small sample size. Robust spinal reconstruction, for example, using 3 or more rods or extending the instrumented vertebra level, should be considered to achieve bony fusion in patients who require multilevel resections.[ Although radiation therapy was not identified as a risk factor for instrumentation failure in our study, 2 previous studies reported irradiation as a significant risk factor.[ Despite radiotherapy being associated with instrumentation failure, the use of this treatment modality cannot always be avoided as it serves as adjuvant therapy for tumor treatment. However, radiation therapy should only be used after careful treatment planning, and indiscriminate irradiation must be avoided.

Sciubba et al[ reported only one local recurrence in their study of 23 cases (4.3%) of lumbar TES for primary tumors. In our study, 4 local recurrences occurred (13.3%), but all occurred after an intralesional surgery where it was difficult to achieve tumor-free surgical margins. In contrast, no local recurrences occurred in patients who underwent TES as the first treatment, so local control of primary lumbar tumors can be said to be excellent after TES.

This study has some limitations. First, this is a single arm study in a single institute, and the relatively small sample size could have caused it to be underpowered and unable to detect statistical significance in some analyses. Larger studies with comparative groups are needed to further validate and generalize our findings. Second, because of the retrospective design of this study, some complications might have been overlooked due to the collected data obtained from electronic medical records were not designed to address this study. Third, changes in the oncological and surgical treatment of several tumors during this study, such as the use of denosumab for giant cell tumors,[ may have affected the results. Despite the recognized limitations, to our knowledge, this study included the largest number of patients after lumbar TES to date. We found that TES in the lumbar spine could result in favorable oncological outcomes while maintaining lower extremity function, albeit with a high risk for complications and instrumentation failure.

In conclusion, TES is a feasible and effective procedure for the treatment of primary tumors of the lumbar spine, but the risks of perioperative complications and late instrumentation failure should be acknowledged. Although postoperative transient lower extremity muscle weakness is an almost inevitable feature of lumbar spine TES, oncologic outcomes were good, especially in patients who underwent TES as their first surgical treatment. Therefore, being familiar with the indications for TES and the surgical technique is important.

Acknowledgments

We would like to thank Editage (www.editage.jp) for English language editing.

Author contributions

Conceptualization: Takaki Shimizu, Norio Kawahara.

Data curation: Takaki Shimizu, Satoru Demura, Satoshi Kato, Katsuhito Yoshioka, Noriaki Yokogawa, Katsuro Tomita.

Formal analysis: Takaki Shimizu.

Investigation: Takaki Shimizu.

Methodology: Takaki Shimizu.

Supervision: Hideki Murakami, Hiroyuki Tsuchiya.

Writing – original draft: Takaki Shimizu.

Writing – review & editing: Hideki Murakami.