Desmoid Tumors: A 20-Year radiotherapy experience

From 1964 through 1984, 45 patients were referred for radiation therapy for desmoid tumor. Fourteen patients had inoperable lesions, or gross residual disease after incomplete resection. Thirty-one patients received postoperative XRT for positive margins or concern about the adequacy of the margin. The minimum follow-up was 2 years, maximum 22 years, median 7.6 years. No patient was lost to follow-up. The primary site was head and neck in 5, upper extremity in 10, chest wall and back in 8, abdomen 2, pelvis 4, and lower extremity 16. All patients were treated with megavoltage radiation therapy using shrinking field techniques. Large fields received a median dose of 50 Gy in 25 fractions. Boost fields were used in the majority of patients to deliver an additional dose of 7 to 27 Gy. The range of total doses was 50 to 76.2 Gy. Three patients received a boost with neutrons. Analysis of patients with inoperable or gross residual showed tumor control in 10 of 14 with a median follow-up of 9.4 years. Resolution of gross disease occurred at a range of i to 64.3 months with a median of 9 months, There was no evidence of a higher probability of ultimate control at higher doses. Tumor control was equal for men and women. The ten patients with local control had doses from 50 to 76.2 Gy whereas the four patients with in field failures had tumor doses of 57 to 66.4 Gy. There was no difference in median dose for patients with local control (60.3 Gy) versus those with tumor recurrence (60 Gy). For subclinical disease, 31 patients receiving postoperative or preoperative XRT had a 77 percent probability of local control in spite of the history of multiple tumor recurrences; local control was achieved in 8 of 9 with negative or uncertain margins and 16 of 22 with positive margins. An analysis of local control as a function of the number of operations revealed that patients referred for adjuvant radiotherapy with no more than two operative procedures had an 88 percent probability of local control, versus 66 percent for more than two operative procedures. All grade 3 complications (defined as requiring surgical intervention or prolonged hospitalization) occurred with doses above 60 Gy. Management of recurrences was successful in 8 of the 11 patients and no patient has died of tumor. Since we find no evidence of a dose-control relationship, but a clear dosecomplication relationship, we currently recommend 50–55 Gy for both gross as well as subclinical disease after surgical ex