Safety and Efficacy of Surgery for Primary Tumors of the Spine

Daryl R. Fourney, Charles G. Fisher, and Stefano Boriani

Introduction

The rarity of primary spine tumors and the slow growth rate of many types (e.g., chordoma, low-grade chondrosarcoma, giant cell tumor) make them difficult to study, especially with regard to the results of surgery and other treatments. The overwhelming majority of studies are retrospective, and the few prospective studies are limited by a short follow-up time.¹ It is difficult to compare older studies because of loose interpretation of the terminology and a lack of standardization in staging and management. Even in individual case series, patient management often varies over time due to advances in diagnosis, surgical technique, and adjuvant therapies. All of these factors have combined to yield a heterogeneous group of published reports,²^–⁹ but has not obscured the principal fact that an oncologically valid approach to these diseases leads to improved disease control and survival.¹⁰

The work of Weinstein, Boriani, and Biagini²^,³^,¹¹^,¹² and others⁵^,⁶ has shown that the principles of musculoskeletal oncology originally developed by Enneking¹³^,¹⁴ can be applied to the spine. The principles of primary tumor management are outlined in Table 2.1, and these principles are revisited throughout this book, with variations in how they are addressed based on the pathology, the location and extent of disease, and patient factors (e.g., medical fitness, exposure to previous therapies, and treatment preference). It is important to emphasize the extensive nature of en-bloc resection surgery, its inherent morbidity, and its cost. In fact, the high rate of intraoperative and early postoperative complications are often weighed against the probability of survival and local control, begging the question of whether en-bloc surgery is justified in the spine.¹^,⁵^,⁷

Efficacy is a broad concept that evaluates the overall value or usefulness of an intervention. It includes an analysis of clinical outcomes and resource utilization. For primary spine tumors, the main clinical outcome of interest is disease control and, ideally, cure. This chapter assesses the efficacy of en-bloc or oncologically appropriate surgery for primary spine tumors to achieve disease control, and discusses the potential for morbidity and strategies to promote safety.

Long-Term Results of Surgery

Recently, the first population-based study on malignant primary osseous tumors of the spine that had sufficient power to evaluate the role of surgical resection on survival was reported by Mukherjee et al.¹⁵ The Surveillance, Epidemiology, and End Results (SEER) registry was analyzed for cases of histologically confirmed primary (nonmetastatic) tumor of the mobile spine or pelvis over a 30-year period (1973–2003). A major limitation of this study was that the surgical margins and the functional outcome of surgery were not reported. The authors performed a multivariate analysis, and after adjusting for age, radiation therapy, and extent of local tumor invasion, surgical resection was independently associated with significantly improved survival for chordoma (hazard ratio [HR], 0.617; 95% confidence interval [CI], 0.25–0.98), chondrosarcoma (HR, 0.153; 95% CI, 0.07–0.36], osteosarcoma (HR, 0.382; 95% CI, 0.21–0.69), and Ewing’s sarcoma (HR, 0.494; 95% CI, 0.26–0.96). Radiation therapy was associated with prolonged survival for patients with osteosarcoma and chordomas, but not for patients with Ewing’s sarcoma and chondrosarcoma.

Table 2.1 Principles and Pitfalls that Affect the Efficacy of Surgery for Primary Tumors

Principle	Pitfall
High index of suspicion	• Diagnosis after intralesional resection with contaminated tumor margins
Biopsy before treatment (percutaneous CT-guided core biopsy)	• Biopsy tract not incorporated in later excision
	• Fine-needle aspiration limits ability to assess tissue architecture
Consistent terminology	• Lack of precision describing surgical approach or margins
Staging with Enneking and WBB? classification	• Enneking inappropriate surgery associated with higher rates of local recurrence and mortality
Multidisciplinary management/experienced centers	• All aspects of care not fully appreciated (e.g., pathology, local and systemic staging, surgical feasibility, technical points to reduce surgical morbidity, need for adjuvant and neoadjuvant therapy, soft tissue coverage, patient preferences, etc.)
Long term follow-up	• Delayed diagnosis of recurrence many years after excision of slow-growing tumor types (e.g., chordoma, low-grade chondrosarcoma)

Abbreviations: CT, computed tomography; WBB, Weinstein–Boriani–Biagini.

Applying Enneking’s Principles to the Spine

A predominantly prospective cohort study by Fisher et al ⁶ of patients with primary spine tumors from four spine centers in Canada treated between 1982 and 2008 was analyzed to determine whether applying Enneking’s principles to the surgical management of primary bone tumors of the spine significantly decreases local recurrence or mortality. Cases were divided into two groups based on surgical margins: Enneking appropriate (EA) and Enneking inappropriate (EI). There were 147 patients with a median follow-up of 4 years (range, 2–7 years). There was a significantly higher risk of local recurrence after EI surgery, and there was a strong correlation between the first local recurrence and mortality with an odds ratio of 4.69 (p < 0.0001). EI surgery was associated with a higher risk of mortality (HR, 3.10; p = 0.0485) compared with an EA approach. The most significant limitation of this study was that the EI cohort had more high-grade tumors; however, the EI group also had a significantly higher rate of adjuvant therapy, which should bias toward the null hypothesis, thus strengthening the conclusions.

Feasibility of Obtaining Acceptable Margins

In a systematic review published in 2009, Yamazaki et al¹ assessed the early efficacy of en-bloc resection techniques by determining the number of wide/marginal resections versus the number of intralesional resections. The results varied significantly among studies. All of the studies included in the systematic review reported the postoperative margins, but not the expected margins based on the preoperative surgical staging. The one exception was a prospective study by Fisher et al.⁵ Among the 26 patients with primary spine tumors reported by Fisher et al., post operative margins were wide in 15, marginal in four (all planned wide), and intralesional in seven (two planned wide, one planned marginal). Thus, Weinstein–Boriani–Biagini (WBB) staging accurately predicted the margins in 19 of 26 cases (73%). However, if the general goal of surgery was attainment of a wide or marginal en-bloc resection (i.e., excluding the four cases that were expected to be intralesional), the WBB staging accurately predicted the attainment of wide or marginal en-bloc resection in 23 of 26 cases (88%). In experienced hands, en-bloc resection with acceptable margins is achievable if WBB staging determines that it is feasible.

Long-Term Results of En-Bloc Resection for Chordoma and Chondrosarcoma

Boriani et al² reviewed 52 consecutive chordomas of the mobile spine over a 50-year period. The series included a retrospective review of 15 cases treated prior to 1991 and a prospective group of 37 cases treated from 1991 to 2002. All patients were staged and subjected to one of five protocols: (1) radiation therapy with or without palliative surgery (n = 10); (2) intralesional extracapsular excision (n = 8); (3) intralesional excision plus radiation therapy (n = 16); (4) attempted en-bloc resection, intralesional or contaminated, combined with radiation therapy (n = 8); and (5) en-bloc resection (n = 10). Every patient treated with radiation therapy alone or intralesional excision had recurrence within 2 years of surgery. Intralesional extracapsular excision plus radiation therapy also had a very high recurrence rate (12 of 16, mean 30 months) but three patients were continuously disease-free (mean, 52 months) and five patients were alive with stable disease (mean, 69 months) at last follow-up. Twelve of 18 patients who received en-bloc resection were continuously disease-free for an average of 8 years (range, 48–155 months). For the six patients who recurred, it is notable that all of them had previously undergone surgery or had documented contaminated tumor margins. The only treatment protocol that resulted in greater than 5 years of continuously disease-free survival was en-bloc resection with marginal or wide margins.

Boriani et al³ also retrospectively reviewed 32 cases of chondrosarcoma within the mobile spine. The average follow-up period was 81 months (range, 2–236 months). Recurrences occurred in 3 of 14 patients treated with en-bloc resection, compared with 100% of 18 patients treated with intralesional curettage.

A systematic review published in 2009 by Boriani et al⁴ compared the effects of wide/marginal (en-bloc) resection with the effects of intralesional resection on local recurrence and survival for chordomas and chondrosarcomas of the spine. They also evaluated the effects of radiation therapy. The authors reported significantly decreased local recurrence and death for both tumor types after en-bloc resection, as compared with intralesional resection (p < 0.0001). The odds ratio for local recurrence with intralesional resection was 10.25. For chordoma, the odds ratio for death if local recurrence occurred was 15.03.

Radiation therapy is generally recommended in the event of intralesional or incomplete resection of chordoma or chondrosarcoma, but fairly high doses (60 to 65 Gy equivalents) are required.⁴ Superior dose distribution may be achieved with proton beam therapy or stereotactic body radiosurgery (SBRT).¹⁶^–¹⁸ The published series using these treatments are mainly limited to skull base and cervical spine lesions, but we expect more studies utilizing SBRT for mobile spine chordomas and chondrosarcomas in the coming years.

Morbidity Rate of En-Bloc Resection

In a systematic evaluation of complications after surgery for primary spine tumors, Yamazaki et al ¹ found tremendous variability in the reported rate of complications. Most studies were retrospective series and they reported lower rates of complications compared to the one prospective study.⁵ However, the retrospective studies had greater long-term follow-up, and thus late complications, such as instrumentation failure and tumor recurrence, were more frequently described.²^,⁵^,⁷^,¹⁹ Only two studies used previously published criteria²⁰ to classify complications.⁷^,¹⁹

Boriani et al²¹ retrospectively assessed morbidity in 134 patients who underwent en-bloc resection for spine tumors between 1990 and 2007, including 90 with primary tumors and 44 with metastatic tumors. The rate of complications was 35.1%. The risk of major complications was more than twice as high in patients with contaminated tumor margins (odds ratio [OR], 2.52; 95% CI, 1.01–6.30; p = 0.048). A separately published analysis of the same data showed a higher rate of complications in patients referred from another center after an open biopsy or surgery who subsequently developed recurrence (72%) compared with previously untreated patients (20%).¹⁹ On multivariate analysis, the independent predictors of major complications were multisegmental tumor (OR, 1.95; 95% CI, 1.07–3.56; p = 0.03) and use of a simultaneous combined anterior-posterior approach (OR, 3.79; 95% CI, 1.09–13.17; p = 0.036).²¹

For sacral tumors, Fourney et al⁷ found a correlation between higher level of sacrectomy and nerve root sacrifice, with an increased complication rate and longer inpatient stay. The incidence of wound complications has been reduced by the use of specialized flap closure techniques, and this is perhaps best exemplified after sacrectomy, where soft tissue defects may be particularly large.

Mortality Rate of En-Bloc Resection

In the systematic review by Yamazaki et al,¹ mortality rates varied from 0 to 7.7%. The most common cause of death related to surgery was respiratory failure. The mortality rate after en-bloc resection in the large series reported by Boriani et al²¹ was 2.2%, occurring in three patients: one intraoperative death due to injury of the vena cava, one postoperative pulmonary embolism, and one late dissection of the wall of the aorta 8 months after surgery.

Safety Promotion Strategies

The morbidity of surgical resection for spine tumors derives from several factors, but there are many effective prevention strategies (Table 2.2).

Preoperative angiographic embolization is a worthwhile consideration, especially for highly vascularized lesions such as giant cell tumor. Sufficient vascular access, including the placement of large-bore intravenous catheters, is necessary in order to administer large volumes of fluids and blood. Central venous pressure monitoring is recommended for most cases of en-bloc resection. We have generally avoided use of the cell saver in primary tumors because of the potential for tumor dissemination. Tranexamic acid appears to be a safe, effective, low-cost method to reduce blood loss during and after surgery. After induction of anesthesia, the loading dose is 2 g in 100 mL for adults and 30 mg/kg for children administered over 20 minutes, followed immediately by a maintenance dose of 1 g in 100 mL infused at a rate of 100 mg/h for adults and 1 mg/kg/h for children.²²

The length of these procedures, the degree of tissue resection, and the postoperative period of convalescence required are all risk factors for deep venous thrombosis and pulmonary embolism. We routinely utilize sequential pneumocompression stockings in the perioperative period and prophylactic low molecular weight heparin postoperatively.

Intraoperative complications are related to the manipulation of vital structures. The risk of injury increases in the case of reoperation or previously irradiation tissues, due to fibrous scar and tissue fragility. If a dural tear is encountered, immediate suturing with a muscular coverage is recommended. If a watertight closure cannot be obtained, cerebrospinal fluid (CSF) drainage is an option but carries its own risks (e.g., overdrainage, pneumocephalus, meningitis).

Table 2.2 Factors that Affect Morbidity After En-Bloc Resection for Primary Spine Tumors

Factor	Safety Promotion Strategy
Complex anatomy due to large tumor growth/extension	• Familiarity with anatomy • Experience with en-bloc resection of the spine/paraspinal structures
Excessive hemorrhage from the tumor or epidural vein	• Preoperative embolization • Tranexamic acid • Careful surgical technique
Manipulation or sacrifice of vascular or nervous structures	• Careful tissue handling with microsurgical techniques • Multimodality intraoperative neurophysiological monitoring
Epidural fibrosis due to previous surgery	• Prevent recurrence through careful staging and planning of the first resection
Wound dehiscence or infection	• Plan appropriate soft tissue coverage(flap if necessary) • Avoid operating through previously irradiated fields
Long-term hardware failure	• Circumferential reconstruction to enable a stable fusion

Manipulation of the spinal cord, especially in the thoracic spine, should be avoided. Careful microsurgical technique is required when handling nervous tissue. We recommend the use of multimodality intraoperative neurophysiological monitoring (somatosensory evoked potentials, motor evoked potentials, electromyography).

Late instrumentation failure has been reported in about 7% of cases.¹⁹ En-bloc resection entails significant instability, and given the goal of long-term disease-free survival, circumferential reconstruction and long-segment stabilization is recommended. Most importantly, meticulous planning and thought must go into biological fixation. With long-term survival anticipated, fusion must be achieved, and this can be difficult in a biologically hostile environment. Vascularized strut grafts and augmented soft tissue coverage are often necessary in large resections.

Finally, the benefit of experience cannot be understated. The surgical anatomy is often complex due to tumor growth adjacent to the spinal cord and critical paraspinal structures. The entire treatment, from biopsy to resection to long-term follow-up, should ideally be performed at the same center, under the guidance of a multidisciplinary team led by an experienced spine oncology surgeon.

Avoiding Problems

Open Biopsy

Poorly planned biopsies increase the risk of local recurrence by tumor dissemination along fascial planes and the biopsy tract.⁸ In a series of patients with chordoma of the mobile spine and pelvis, Bergh et al⁹ found that performance of an invasive diagnostic procedure outside the index center and inadequate surgical margins were associated with tumor recurrence and tumor-related death. In patients with primary tumors of the sacrum, Fourney et al⁷ noted a 78% continuously disease-free survival for patients undergoing both biopsy and surgery at the index institution, compared with 55% for those patients who underwent a prior procedure at another institution.

Incisional biopsy or intralesional resection significantly increases the risk of local recurrence; therefore, transcutaneous computed tomography–guided trocar biopsy is recommended.¹ When there is a suspicion of a primary tumor, the surgeon who performs the definitive surgery should ideally be the one to perform or direct the biopsy procedure. Permanent marking of the biopsy site is recommended so that it can be identified and included in the excision. Although fine-needle aspiration provides cytomorphological features that may yield a diagnosis, a trocar needle may improve accuracy by analysis of all histological features of the tissue.¹^,²³^,²⁴

Inconsistent Terminology and Staging

Historically, the assessment and management of primary spine tumors have followed general principles of neurosurgery and orthopedics; however, the reluctance of spine surgeons to adopt Enneking’s principles of musculoskeletal oncology has largely been overcome in recent years. Nevertheless, it is still common to receive consultations regarding contaminated margins after a poorly conceived excisional biopsy or intralesional resection. It is important to emphasize the correct use of terminology and staging methods inherent to the planning and execution of oncologically sound procedures.¹¹^,¹⁴

Underestimating the Technical Challenges and Functional Sacrifices of Surgery

The management of primary tumors of the spine is challenging. Many are slow growing and may be extensive by the time of presentation (e.g., chordoma, chondrosarcoma), involve complex anatomy, and only partially respond to adjuvant therapies.²^,³^,⁵^–⁸^,¹⁹^,²¹ Local recurrence of chordoma in particular is difficult to palliate as it often leads to multiple aggressive recurrences, debilitation, and eventual death.²^,⁸^,¹⁰^,¹⁹

When staging determines that en-bloc excision is the procedure of choice, surgical planning needs to take into account not only the functional sacrifices necessary to achieve the oncological goals, but also the inherent morbidity of surgery and eventual health-related quality of life (HRQOL). Only one study has addressed this issue and found that at a mean follow-up of 3½ years after en-bloc resection for primary bone tumors the Short Form (SF)-36 mental component score was 51 (essentially normal) and the physical component summary 38.⁵

Chapter Summary

The risk of complications and recurrence is highest after revision surgery. Therefore, the first treatment is a major determinant of outcome.²¹ When there is a suspicion of primary spine tumor, referral to an experienced center is recommended, because the surgeon who performs the definitive excision should ideally perform (or at least direct) the biopsy.¹ This ensures that unrelated functional tissue is not contaminated with tumor. Contamination would necessitate the tissue’s inclusion in subsequent en-bloc resection—something that would not occur with properly planned biopsy.

Local and systemic staging using Enneking’s approach identifies patients who may be cured, or at least have a better chance at long-term disease control and reduced mortality with en-bloc resection. Oncologically sound margins are likely to be achieved when WBB staging determines that it is feasible.¹ However, the adverse-event profile of these surgeries is very high, even at experienced centers. Therefore, we recommend that only experienced, multi-disciplinary teams perform them. The need for careful patient selection including detailed preoperative counseling cannot be overstated.

Pearl

Enneking’s principles of oncological staging are valid with respect to primary spine tumors.

Adequate margins are feasible depending on the surgical staging (WBB system).

En-bloc resection has a high rate of complications and therefore should only be performed at experienced centers.

Multidisciplinary tumor boards are necessary to plan and carry out appropriate treatment.

Biopsy is ideally planned or performed by the surgical oncologist who will ultimately perform the en-bloc resection so that the biopsy tract can be incorporated in the resected specimen.

Pitfalls

A high index of suspicion for primary tumors is necessary to prevent incisional biopsy or intralesional resection that could make oncologically sound margins impossible to achieve.