Independent predictors for local recurrence following surgery for spinal metastasis

Background: Local recurrence of spinal metastasis after surgical resection is relatively common. We sought to determine risk factors and independent predictors for local recurrence after primary surgical resection of spinal metastasis. Methods: Demographic and clinical variables were collected for patients who underwent surgery for spinal metastasis June 2005 to June 2011. Primary outcome of interest was local recurrence. Significant associations between covariates of interest and recurrence were identified using the chi-square test. Multivariable logistic regression models for recurrence risk were fit and adjusted for potential confounders. Results: A total of 99 patients were analyzed. Mean time to metastatic recurrence was 9.8 months. Thirty-two patients (32.3 %) had local recurrence of metastatic disease following initial surgery. Patients who underwent radiotherapy had significantly higher recurrence rates than patients who did not (39.2 % vs. 12.0 %, respectively; P = 0.012). Patients with metastatic disease affecting more levels had significantly lower recurrence rates. On multivariate analysis, older age was an independent predictor of decreased likelihood of local recurrence. Melanoma was the only cancer type independently associated with higher risk for recurrence. Patients with recurrence had significantly higher 1- and 2-year survival rates than patients without recurrence. Median length of survival was longer in the recurrent group as well. Conclusions: Other than melanoma, covariates significantly associated with recurrence were factors likely associated with increased survival, including less-extensive spinal disease and radiotherapy. Thus, longer survival time following surgery likely results in a greater chance for local recurrence. As advancements in treatment provide prolonged survival, local recurrence rates will likely increase.