En bloc resections for primary spinal tumors in 20 years of experience: effectiveness and safety.

BACKGROUND CONTEXT: Many studies have demonstrated that en bloc surgical resection of primary spinal tumors with adequate margins results in improved local disease control and survival compared with intralesional excision. Nevertheless, the use of this procedure is under debate because most of the current evidence is provided by small and heterogeneous series of cases.
PURPOSE: To validate the application of en bloc resection for the treatment of aggressive benign and primary malignant spinal tumors. STUDY
DESIGN: This is a prospective cohort study. PATIENT SAMPLE: From August 1990 to March 2010, 103 consecutive patients affected by primary spinal tumors were enrolled in the study. All patients were submitted to the same clinical and imaging workup. OUTCOME MEASURES: Analysis of local recurrence (LR) and tumor-related mortality, reliability of preoperative surgical planning, and morbidity and mortality. In addition, the effects of possible predictors of these events were studied.
METHODS: The parameters for the effectiveness and safety of en bloc resections performed on primary spinal tumors were considered as primary end points of this study, and two research questions were formulated. The analysis of the procedure effectiveness considered the identification of possible predictors of LR and tumor-related mortality. Information about safety is collected so as to clarify the possibility to respect the preoperative planning and to identify possible predictors of morbidity and mortality. Data from clinical and imaging examination were collected in a database and were used to answer the proposed research questions.
RESULTS: All patients were followed for a minimum of 24 months or until death. At the final assessment, 69 patients resulted with no evidence of disease with a mean follow-up of 100 months. Among the 103 patients, tumor recurred in 22 cases with a mean follow-up period of 39 months after surgery. A Cox regression multivariate analysis shows that marginal and intralesional resections are independent predictors of LR (hazard ratio [HR] 9.45, 95% confidence interval [CI] 1.06-84.47 and HR 38.62, 95% CI 4.67-319.21, respectively, compared with wide resection) and tumor-related mortality (in particular, HR 17.10, 95% CI 3.80-77.04 for intralesional resection compared with the wide one). The same analysis demonstrates that en bloc resection performed in recurrent cases or patients previously submitted to open biopsy (nonintact cases) have a LR risk higher than intact cases (HR 3.45, 95% CI 1.38-8.63). The success rate of en bloc resections in achieving adequate margins is 82.4%, and Weinstein-Boriani-Biagini surgical staging can also predict the margins in a high percentage of cases (75.7%). Complications occurred in 41.7% of patients with a higher rate observed in the nonintact group and for surgery with a double-approach or multisegmental resections. The mortality rate related to surgery complications was 1.9%, whereas tumor-related mortality was 15.5%.
CONCLUSIONS: Statistical analysis of the long-term results referred to 103 patients affected by aggressive benign and malignant primary spine tumors indicates that an en bloc resection is associated with a high rate of complications. Nevertheless, it decreases the risk of LR and tumor-related mortality. En bloc resection is a highly demanding procedure but can be performed to an acceptable degree of safety.