A Novel System for the Surgical Staging of Primary High-grade Osteosarcoma: The Birmingham Classification

Background Chemotherapy response and surgical margins have been shown to be associated with the risk of local recurrence in patients with osteosarcoma. However, existing surgical staging systems fail to reflect the response to chemotherapy or define an appropriate safe metric distance from the tumor that will allow complete excision and closely predict the chance of disease recurrence. We therefore sought to review a group of patients with primary high-grade osteosarcoma treated with neoadjuvant chemotherapy and surgical resection and analyzed margins and chemotherapy response in terms of local recurrence. Questions/purposes (1) What predictor or combination of predictors available to the clinician can be assessed that more reliably predict the likelihood of local recurrence? (2) Can we determine a better predictor of local recurrence-free survival than the currently applied system of surgical margins? (3) Can we determine a better predictor of overall survival than the currently applied system of surgical margins? Methods This retrospective study included all patients with high-grade conventional osteosarcomas without metastasis at diagnosis treated at one center between 1997 and 2012 with preoperative chemotherapy followed by resection or amputation of the primary tumor who were younger than age 50 years with minimum 24-month followup for those still alive. A total of 389 participants matched the inclusion criteria. Univariate log-rank test and multivariate Cox analyses were undertaken to identify predictors of local recurrence-free survival (LRFS). The Birmingham classification was devised on the basis of two stems: the response to chemotherapy (good response = ≥ 90% necrosis; poor response = < 90% necrosis) and margins (< 2 mm or ≥ 2 mm). The 5-year overall survival rate was 67% (95% confidence interval [CI], 61%–71%) and 47 patients developed local recurrence (12%). Results Intralesional margins (hazard ratio [HR], 9.9; 95% CI, 1.2–82; p = 0.03 versus radical margin HR, 1) and a poor response to neoadjuvant chemotherapy (HR, 3.8; 95% CI, 1.7–8.4; p = 0.001 versus good response HR, 1) were independent risk factors for local recurrence (LR). The best predictor of LR, however, was a combination of margins ≤ 2 mm and a less than 90% necrosis response to chemotherapy (Birmingham 2b HR, 19.6; 95% CI, 2.6-144; p = 0.003 versus Birmingham 1a; margin >2 mm and more than 90% necrosis HR, 1). Two-stage Cox regression model and higher Harrell’s C statistic demons