Abstract
Introduction
Spinal hemangiomas are the most common tumor of the spine, being identified in 2 to 27% of patients at autopsy and on radiographic and magnetic resonance imaging studies. The majority of spinal hemangiomas are identified incidentally with only 1% presenting with symptoms, including pain, pathologic fracture, or neurologic compromise. Because of the rarity of symptomatic presentation, optimal surgical treatment of, and risk factors for local recurrence and mortality associated with, symptomatic spinal hemangiomas remain unclear. This multicenter cohort study aims to quantify rates of local recurrence and mortality following surgical treatment of symptomatic spinal hemangiomas and to identify prognostic variables for local disease control and death.
Patient and Methods: AOSpine knowledge forum tumor investigators created an ambispective database of surgically treated patients with symptomatic spinal hemangiomas. Patient data pertaining to demographics, clinical presentation, diagnosis, treatment, cross-sectional survival, local recurrence, and perioperative morbidity were collected. Tumors were classified according to Enneking and Weinstein–Boriani–Biagini. Descriptive statistics were summarized and Kaplan–Meier curves illustrated mortality and recurrence. Fisher exact and log-rank tests identified prognostic variables for recurrence and mortality.
Results
Between 1996 and 2012, 68 patients (mean age = 51 years, SD = 16), 43 females and 25 males, underwent surgical treatment for a spinal hemangioma with 23% (n = 13) classified as benign aggressive. Epidural tumor extension was present in 55% of patients (n = 33). Pain was a presenting symptom in 82% (n = 54), 31% (n = 20) had a pathologic fracture, and 37% (n = 24) were neurologically compromised. Overall, 23 patients (35%) underwent preoperative embolization. Enneking appropriate surgery was performed in 79% of patients (n = 44). Adjuvant radiotherapy was employed in 10% of patients (n = 7). The local recurrence rate was 3% (n = 2) with no recurrences in patients undergoing marginal or wide resection. Mortality because of the spinal hemangioma was not observed during the study period (mean follow-up = 3.9 years, SD = 3.8). Prior subtotal intralesional resection was performed in one of the two patients who had a local recurrence.
Conclusion
This is the largest multicenter surgical cohort of spinal hemangiomas. Although symptomatic spinal hemangiomas frequently present with epidural disease and neurologic compromise, they remain a benign tumor. Thus, formal en bloc resection, with an associated risk and increased morbidity, is not required, and excellent rates of local control and long-term survival can be obtained when strict oncologic treatment principles are followed during the index surgery.