Preoperative vs. postoperative radiotherapy in the treatment of soft tissue sarcomas: a matter of presentation

Purpose: Radiotherapy for soft tissue sarcoma is typically preoperative or postoperative, with advocates of each. In this study, the relationship of the sequencing of radiotherapy and surgery to local control was examined. Methods and Materials: The cohort consisted of 453 patients with Grade 2–3 malignant fibrous histiocytoma, synovial sarcoma, or liposarcoma treated from 1965–1992. Retroperitoneal sarcomas were excluded. Median follow-up was 97 months. There were 3 groups of patients that were classified by the treatment administered at our institution: preoperative radiotherapy to a median dose of 50 Gy given before excision at MDACC (Preop; n = 128); postoperative radiotherapy to a median dose of 64 Gy given after excision at MDACC (Postop; n = 165); and radiotherapy to a median dose of 65 Gy without excision at MDACC (RT Alone; n = 160). Those in the RT Alone Group had gross total excision at an outside center prior to referral. Results: Histological classification, whether locally recurrent at referral, and final MDACC margins were independent determinants of local control in Cox proportional hazards multivariate analysis using the entire cohort. The type of treatment was not significant; however, tumor status at presentation (gross disease vs. excised) affected these findings greatly. Gross disease treated with Preop was controlled locally in 88% at 10 years, as compared to 67% with Postop (p = 0.01). This association was independently significant for patients treated primarily (not for recurrence). In contrast, for those presenting after excision elsewhere, 10-year local control was better with Postop (88% vs. 73%, p = 0.07), particularly for patients treated primarily (91% vs. 72%, p = 0.02 in univariate analysis; p = 0.06 in multivariate analysis). Re-excision at MDACC (Postop) resulted in enhanced 10-year local control over that with RT Alone (88% vs. 75%, p = 0.06), and was confirmed to be an independent predictor in multivariate analysis (p = 0.02). Conclusion: Local control was highest with Preop in patients presenting primarily with gross disease, and with Postop in patients presenting primarily following gross total excision. The data suggest that 50 Gy is inadequate after gross total excision, possibly due to hypoxia in the surgical bed.