Comparison of En Bloc Resection and Intralesional Excision for Re‐resection of Giant Cell Tumors of the Spine

Abstract

Objective

Re‐resection of spinal giant cell tumors is an exceedingly difficult procedure. Moreover, the prognosis of patients with en bloc resection or intralesional excision for re‐resection has rarely been reported. This study aimed to compare the prognostic value of en bloc resection with that of intralesional excision in patients undergoing re‐resection for giant cell tumors of the spine.

Methods

This retrospective analysis evaluated patients who underwent revision surgeries for relapse of giant cell tumors of the spine at our center between January 2005 and January 2021. Local progression‐free survival represents the duration between en bloc resection or intralesional excision and tumor recurrence. Neurological recovery, survival rates, local control, and complications were evaluated. The Kaplan–Meier estimator was used for survival analysis.

Results

A total of 22 patients (nine men and 13 women) with a mean age of 34.1 (range 19–63) years were included. Significant statistical differences were found in the local tumor recurrence rate between patients treated with en bloc resection and those treated with intralesional excision (p < 0.05). The 5‐ and 10‐year local progression‐free survival rates were both 90% in the en bloc resection group, while in the intralesional excision group, the 5‐year local progression‐free survival rate was 80% with a 10‐year rate of 45.7%. The en bloc resection group had a lower local tumor recurrence rate than that of the intralesional excision group (p < 0.05), but the former had a higher rate of complications (p = 0.015).

Conclusions

This study revealed a low local recurrence rate in patients who underwent en bloc resection for giant cell tumors, while the perioperative complication rate was high.

Our findings highlight the trade‐off between lower recurrence rates with en bloc resection and the higher risk of perioperative complications, helping clinicians make informed decisions when choosing the most appropriate surgical technique for the patients undergoing re‐resection for giant cell tumors of the spine.

Introduction

Giant cell tumors (GCTs) of the spine have a high postoperative recurrence rate. ¹ The clinical outcomes following surgery depend on various factors, including the preoperative diagnostic method, tumor localization, extent of disease, and the choice of surgical technique for tumor resection. ² Due to the limited value of clinical and radiological features, the diagnosis of GCTs is mainly based on the histological examination of computed tomography (CT)‐guided core needle biopsy samples. ³ However, some local hospitals still perform open or resection biopsies, which may result in incomplete tumor resection. Patients with incompletely resected spinal GCTs require special attention due to the possibility of recurrences. ⁴ Achieving complete resection of spinal tumors remains a challenge for many surgeons, and in some cases, curettage is selected as the primary treatment despite its association with extremely high recurrence rates. ⁵ The unique anatomical structure of the spine, as well as severe scar hyperplasia, make revision surgery particularly difficult. ⁶ Therefore, complete resection of spinal GCTs can be challenging, and reports on this subject are rare.

In 2009, the Spine Oncology Study Group conducted a systematic literature review and strongly recommended en bloc resection for the treatment of spinal GCTs whenever technically feasible. Although en bloc resection for spinal GCTs demonstrates a low local tumor recurrence rate, ⁷ Boriani et al. ⁸ did mention that the choice of en bloc resection must be balanced with the inherent risks of the procedure. Jia et al. ⁹ reported that total spondylectomy, whether performed using the total piecemeal method or total en bloc spondylectomy (TES), could prolong disease‐free survival compared to subtotal spondylectomy. However, the outcomes of TES have not yet been compared to those of non‐en bloc gross total resection.

Although en bloc resection is considered an appropriate treatment for GCTs, intralesional excision is performed in cases where TES is anatomically unfeasible or when the patient has undergone previous surgeries. There are few published treatment options for incomplete resection of spinal GCTs, likely due to a lack of cases and long‐term follow‐up data. Hence, it is important to investigate the neurological recovery, survival rates, local control, and complications associated with en bloc resection and intralesional excision.

The purpose of this study was to (i) compare the clinical outcomes of en bloc resection and intralesional excision and (ii) assess whether the prognosis of en bloc resection is better than that of intralesional excision for re‐resection of GCTs of the spine.

Methods

Study Design and Eligibility Criteria

We retrospectively reviewed the clinical data of all patients with incompletely resected GCTs who underwent total gross tumor resection of the entire osseous compartment at our center between January 2005 and January 2021. The inclusion criteria were: (i) the GCT primary site was in the thoracic and/or lumbar spine, (ii) initial surgery was an incomplete resection, and (iii) the follow‐up period was at least 18 months. The exclusion criteria were: (i) the primary tumor was located at the cervical spine or sacrum and (ii) an inability to undergo total gross tumor resection.

Twenty‐two patients were included in the study. Ethical approval was granted by the ethics committee of our institution (IRB00006761‐M2017374), and written informed consent was obtained from all patients. The extracted data included patient age, sex, neurological symptoms, tumor location, number of affected segments, adjuvant radiotherapy, and follow‐up information including local tumor recurrence, distant metastases, and survival at 5 and 10 years (%).

En Bloc Resection or Intralesional Excision for Reresection

All primary surgical procedures were performed at a local hospital (Figure 1), and each patient's pathology was verified using a postoperative specimen. All cases were classified as Enneking stage III GCTs. Our center's surgical approach for revising GCTs was determined by considering the tumor's size, location, and impact on surrounding neurovascular structures, as determined through radiological data for each patient. Spinal structures with tumor involvement were partially removed during the primary surgical procedure. Extracapsular tumor separation, including the entire osseous compartment, was the first step in total gross tumor resection. Tumor‐free exposed margins were achieved by keeping the tumor capsule intact prior to resection. Scarring at the surgical site was observed as a result of previous surgery. Following pedicle osteotomy, total tumor excision was performed using TES if it was feasible to separate the scar between the lamina, articular process, and dural sac. When the residual laminae and articular processes were grossly invaded by the tumor, an intralesional excision technique was employed to excise the posterior spinal structures. In addition, if the tumor invaded the canal spines at these locations, the adhesion between the tumor, surgical scar, and dura within the spinal canal was notably tight. Prior to removing the vertebral body, it was imperative to extract the tumor in large fragments and separate the scar to ensure complete liberation of the dural sac. In such cases, gross total excision was performed using the intralesional excision technique. All patients underwent selective arterial embolization preoperatively, followed by en bloc resection (Figure 2) or intralesional excision as the procedure of choice to achieve total gross tumor resection of the entire osseous compartment. All patients were treated according to the same follow‐up protocol. Clinical follow‐up was conducted at 3, 6, 12, 18, and 24 months postoperatively, and annually thereafter.

FIGURE 1.

L1–L3 giant cell tumor in a 26‐year‐old man. A resection biopsy performed 6 months prior at a local hospital resulted in incomplete resection. Anteroposterior and lateral lumbar spine radiographs (A, B) performed at a local hospital after the resection biopsy are shown. Coronal CT (C) and transverse CT (D) show the bone destruction of the tumor. Sagittal MRI (E) and transverse MRI (F) show paravertebral soft tissue tumors. CT, computed tomography; MRI, magnetic resonance imaging.

FIGURE 2.

Revision surgery of a representative patient at our center. The patient was transferred to our hospital and underwent en bloc resection to achieve total gross tumor resection. CT imaging (A, B) of the surgically resected specimen shows en bloc resection of the tumor. The patient was followed up for 2 years after the procedure at our center. The anteroposterior and lateral radiographs (C, D) show no instrumentation failure. Coronal CT (E) and sagittal CT (F) show no 3D‐printed prosthesis subsidence. MRI (G) shows no tumor recurrence, and most cerebrospinal fluid leakage has been absorbed. CT, computed tomography; MRI, magnetic resonance imaging.

Statistical Analysis

Quantitative data are expressed as the median (range), and qualitative data as counts and percentages. Univariate analysis was performed using the Mann–Whitney U test for continuous variables and Pearson's chi‐square or Fisher's exact tests for categorical variables. The clinicopathological characteristics were evaluated based on their potential prognostic value for local progression‐free survival (LPFS), which was defined as the time interval from total gross tumor resection to local tumor relapse. In cases of death or loss to follow‐up, the date of the last available follow‐up was censored. Kaplan–Meier survival analyses were performed to evaluate LPFS. The significance level was set at p < 0.05. Statistical calculations were performed using SPSS software (version 22.0; IBM Corp., Armonk, NY, USA).

Results

Demographic Data

The mean follow‐up time was 93.4 (median: 80.5; range: 21–192) months. The study group included nine men and 13 women with a mean age of 34.1 (range 19–63) years. During the initial examination at our center, two or more contiguous vertebral levels were affected in 15 patients. Thirteen patients received denosumab for at least 4 weeks before surgery. None of the patients underwent chemotherapy.

Outcomes of En Bloc Resection and Intralesional Excision

In 12 patients, the tumors were removed by en bloc resection, with nine further requiring intralesional pediculectomy for complete resection; intralesional excision was used in the other 10 patients (Table 1). A titanium mesh cage filled with allograft was used for anterior reconstruction in 11 patients. A 3D‐printed artificial vertebral body was used for reconstruction in the remaining 11 patients treated after 2017.

TABLE 1.

Summary of patient clinical characteristics and details of GCTs treatments

N/Sex /Age (years)	Site	SCI (pre/post)	Denosumab	Approach	Surgery	Prosthesis	Radiation	Follow‐up (months)	Recurrence (months)	Status	Complications
1/F/20	T5–6	E/E	Yes	P + AL	TES	3D	Yes	94	No	NED	Pleural tear, CSF
2/M/28	T8–10	B/E	Yes	P	TES	3D	Yes	75	No	NED	Pleural tear
3/F/25	C7–T3	B/D	Yes	P + A	TES	3D	Yes	77	No	NED	CSF, C7 nerve root palsy
4/F/22	T6–8	C/E	Yes	AL + P	TES	3D	Yes	58	No	NED	Pleural tear
5/F/31	T12	E/E	Yes	P	TES	3D	Yes	81	24	AWD	CSF
6/F/34	T10	E/E	Yes	P	TES	3D	No	23	No	NED	Pleural tear
7/M/26	L1–3	E/E	Yes	AL + P	TES	3D	No	29	No	NED	CSF
8/M/45	T9–12	C/E	Yes	P	TES	3D	Yes	21	No	NED	Pleural tear, CSF
9/M/19	T2–4	E/E	Yes	P	TES	Cage	Yes	64	No	NED	Pleural tear
10/F/19	L4	‐	Yes	P + AL	TES	3D	Yes	34	No	NED	None
11/F/44	L4	‐	Yes	P + A	TES	3D	Yes	24	No	NED	CSF
12/M/25	T11–L1	D/E	Yes	AL + P	TES	3D	Yes	114	No	NED	Wound infection
13/F/63	L2	‐	No	P	EIE	Cage	Yes	80	No	NED	Wound infection
14/M/45	T9	D/E	No	P	EIE	Cage	Yes	155	72	AWD	Pleural tear, thoracic duct
15/M/43	T10–12	C/E	Yes	P + AL	EIE	Cage	Yes	79	No	NED	CSF, deep vein thrombosis
16/F/48	L4	‐	No	P + A	EIE	Cage	Yes	192	No	NED	Common iliac vein and peritoneum
17/M/36	T3–5	E/E	No	AL + P	EIE	Cage	Yes	180	54	NED	Pleural tear, CSF, chylous duct
18/F/36	T6–8	C/E	No	P	EIE	Cage	Yes	144	78	AWD	None
19/M/21	T5–6	E/E	No	P	EIE	Cage	Yes	120	96	TRD	None
20/F/34	T10–11	C/D	No	P	EIE	Cage	Yes	99	No	LFU	CSF, pleural tear
21/F/49	T10–11	C/E	No	P	EIE	Cage	Yes	180	No	NED	None
22/F/37	T1–2	E/E	No	P + A	EIE	Cage	Yes	132	18	A	None

Abbreviations: A, anterior; AL, anterolateral; AWD, alive with disease; Cage, titanium mesh cage; CSF, cerebrospinal fluid leakage; EIE, extracapsular intralesional excision; F, female; GCT, giant cell tumor; LFU, lost to follow‐up; M, male; mos, months; NED, no evidence of disease; P, posterior; SCI, spinal cord injury; TES, total en bloc spondylectomy; TRD, tumor‐related death; 3D, three‐dimensionally printed artificial vertebral body.

Outcomes of Perioperative Complication

Seventeen patients (77.3%) developed at least one perioperative complication, and the complication rate was higher for en bloc resection than for intralesional excision (p = 0.015). The most common complication was pleural tear, which occurred in nine patients (45%), leading to pleural effusion and atelectasis in seven of those patients. The second common complication was cerebrospinal fluid leakage caused by dural tears, which occurred in eight patients (40%). The third common complication was wound infection, which occurred in two patients (10%), requiring debridement and antibiotic treatment. One case each of nerve root palsy, deep vein thrombosis, ligation of the chylous duct after injury, and ligation of the thoracic duct after laceration was observed. One patient sustained tear in the common iliac vein and peritoneum during anterior osteotomy, and the laceration was repaired after tumor resection (Table 1).

Outcomes of Adjuvant Radiation

Twenty (90.9%) patients received adjuvant radiation therapy after the revision procedure at our center. Twelve patients (54.5%) presented neurological symptoms before the revision surgery, which improved 6–12 months after the procedure.

Outcomes of Tumor Recurrence

One patient was lost to follow‐up at 99 months, and one died of tumor recurrence with multiple osseous metastases 120 months after the procedure. None of the patients developed lung metastases during follow‐up. The 5‐ and 10‐year LPFS rates were both 90.0% in patients who underwent en bloc resection. In comparison, the 5‐year LPFS rate was 80.0% in patients treated with intralesional excision, and the 10‐year LPFS rate was 45.7% (Figure 3). There was a statistically significant difference in the local recurrence rate between patients who underwent en bloc resection and those who underwent intralesional excision (p = 0.043).

FIGURE 3.

Kaplan–Meier analysis of cumulative local progression‐free survival of patients. The local progression‐free survival of patients undergoing total en bloc spondylectomy with blue line is higher than that of patients undergoing extracapsular intralesional excision with green line.

Discussion

In this series of 22 surgical cases, the complication rate with en bloc resection was higher than that with intralesional excision. However, the local recurrence rate was significantly lower in patients who underwent en bloc resection than in those who underwent intralesional excision.

Clinical Outcomes of Reresection

Revision surgery for re‐resection of spinal GCTs, particularly following multiple recurrences, poses significant challenges due to the altered local anatomy of the spine and neurological deficits that create considerable surgical obstacles. Moreover, preexisting spinal cord damage is common prior to revision surgery, and tumor invasion into adjacent spinal segments further complicates the procedure. In our study sample, 65.2% of patients presented with severe neurological deficits and involvement of two or more adjacent levels. Therefore, as previously described, the potential risks and benefits must be carefully evaluated during surgical planning. ¹⁰

The risks associated with TES are considerably high and include spinal cord injury, pleural effusion, and postoperative cerebrospinal fluid leakage. Sciubba et al. ¹¹ reported a perioperative complication rate of 65.2% following TES. In a study by Ming et al., ¹² 28.6% of patients who underwent intralesional excision experienced perioperative complications. Yokogawa et al. ¹³ reported a postoperative complication rate of 60% among 25 patients with Enneking stage III spinal GCTs, including 13 who underwent TES and 12 who underwent total piecemeal spondylectomy. However, these studies did not statistically compare the complication rate between TES and total piecemeal spondylectomy.

Whether TES or intralesional resection is used for the treatment of spinal GCT relapse, the procedure is technically demanding and, therefore, the complication rates may be high for both procedures. In our study sample, 12 patients underwent en bloc resection, with intralesional excision of the entire osseous compartment performed in the other 10 patients. Common complications included pleural tears, cerebrospinal fluid leakage, and wound infections. Patients who underwent en bloc resection exhibited higher complication rates than those who underwent intralesional excision.

Prognosis of Re‐Resection

Re‐resection of spinal GCTs carries a high risk of tumor recurrence owing to various factors associated with revision surgery, such as a history of prior treatment or an invasive diagnostic procedure performed at another hospital. These factors can contribute to poorer outcomes. Management of recurrent GCTs is difficult because of the scar formation and tumor range, as well as the fact that several of these patients cannot undergo total resection due to the vicinity of the tumor to vital anatomical structures. ¹⁴ Patients who experience a recurrence of spinal GCTs are at risk of two or more subsequent recurrences. ¹⁵ Treatment history is an independent prognostic factor for disease‐free survival in patients with spinal GCTs. ¹⁶ Moreover, it is impossible to apply TES to all patients requiring re‐resection of spinal GCTs. TES with intralesional pediculectomy has a good surgical outcome, even in patients with Enneking stage III spinal GCTs. ¹³ Removal of the entire osseous compartment, using either en bloc or piecemeal methods, can significantly reduce the recurrence rate of spinal GCTs. ¹⁷ Ma et al. ⁴ reported that patients with recurrent spinal GCTs treated using intralesional total spondylectomy could achieve a good prognosis. In our study, the 5‐ and 10‐year LPFS rates were significantly higher in patients treated with en bloc resection than in those treated with intralesional excision, indicating that a good prognosis can be achieved with en bloc resection. In addition, the local recurrence rate was lower for en bloc resection than for intralesional excision, further revealing that en bloc resection can attain better local tumor control than intralesional excision.

Intralesional resection for patients undergoing re‐resection of GCTs can be coupled with adjuvant therapies to reduce recurrence rates. Adjuvant therapies for GCTs of the bone include radiation, selective arterial embolization, and administration of denosumab, all of which have a potentially significant impact on the treatment of GCTs. ¹⁸ Denosumab only eliminates giant cells, cannot destroy tumor stromal cells, and cannot replace surgery and other treatments. Denosumab has been used as a preoperative treatment at our center since 2014. For patients requiring surgery, denosumab can control GCTs and may “harden the edge,” which is conducive to successful tumor resection. ¹⁹ Surgical treatment of GCTs frequently results in substantial intraoperative blood loss. Preoperative selective arterial embolization can reduce intraoperative blood loss, clear the surgical field, and facilitate maximal tumor removal. ²⁰ Patients with spinal GCTs should be treated with radiotherapy, especially in cases of revision surgery with intralesional resection. ¹⁴ In our study, all patients who underwent tumor intralesional resection received postoperative adjuvant radiotherapy. Zhou et al. ²¹ reported that the absence of adjuvant radiotherapy was a negative prognostic factor for local recurrence‐free survival in patients with Enneking stage III spinal GCTs.

Limitation and Strengths

The rarity of TES for the treatment of relapsed spinal GCTs means that this case series makes a significant contribution to the available evidence. However, the limitations of our study need to be acknowledged. First, the study sample size was small, limiting the power of the statistical analysis. Second, the study is a retrospective single‐center design; therefore, our findings will need confirmation from multiple centers in future studies.

En bloc resection is a feasible procedure in certain patients; however, the potential risks and benefits have to be carefully evaluated.

Conclusion

The local recurrence rate in patients who underwent en bloc resection for the re‐resection of GCTs was low. However, it is important to fully consider the high perioperative complication rate associated with this procedure.

Conflict of Interest Statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Author Contributions

All authors had full access to the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Conceptualization: H.Z., Z.L., and F.W. Methodology: H.Z., Y.T., Z.L., and F.W. Investigation: H.Z., P.H., Z.L., and F.W. Formal Analysis: H.Z., Y.T., S.Z., Z.L., and F.W. Resources: H.Z., X.L., Z.L., and F.W. Writing–Original Draft: H.Z., Z.L., and F.W. Writing–Review & Editing: H.Z., Z.L., and F.W. Visualization: H.Z., Z.L., and F.W. Supervision: H.Z. and F.W. Funding Acquisition: F.W.

Funding Information

This work was supported by grants from Key Clinical Projects of Peking University Third Hospital (BYSYZHKC2020114).

Ethics Statement

Ethical approval was granted by the Peking University Third Hospital Medical Science Research Ethics Committee (IRB00006761‐M2017374).

Data Availability Statement

The original contributions presented in the study are included in the article. Further inquiries can be directed to the corresponding author.