Background: Prospective application of the French Sarcoma Group (FSG) method of surgery reporting in soft tissue sarcoma (STS) in a single centre.
Methods: Patients with primary STS of the extremities or trunk wall consecutively operated at the same institution from January 1996 to December 2002 were evaluated for local recurrence (LR). There were 205 patients, with AJCC/UICC stages III and IV in 51% of cases. Resection types according to FSG were R0 in 147, R1 in 53 and R2 in five cases. Radiotherapy was delivered in 163 patients and chemotherapy in 103. Multivariate analysis was performed. Overall five-year survival was 75%. Median follow-up for surviving patients was 53 months.
Results: Actuarial five-year LR incidence was 13% in 200 patients with gross resection (R0+R1), it was 7% in R0 and 30% in R1 patients (p<0.0001). At univariate analysis, significant prognosticators for LR were age, histotype, tumour invasion, grade and resection type R. At multivariate analysis, resection R1 (relative risk (RR) 4.3, p=0.001) and grade 3 (RR 3.9, p=0.013) independently predicted LR. Combining these two variables produced three prognostic groups for LR: group 0 (no factor, n=70), group 1 (one factor, n=94) and group 2 (two factors, n=36) with five-year LR of 4%, 12% and 39%, respectively (p=6.4x10(-7)).
Conclusion: This first prospective evaluation of surgery reporting in STS evidences a fourfold, highly discriminating difference in LR between resections R0 and R1.