Mammary angiosarcoma. The prognostic significance of tumor differentiation.

Mammary angiosarcoma appears to be a heterogeneous group of neoplasms morphologically, in which growth pattern or grade is an important prognostic factor. In this report, 63 patients whose slides were seen in consultation and/or who were treated at Memorial Sloan-Kettering Cancer Center (MSKCC) are described. Included are 32 MSKCC patients from an earlier study (with additional follow-up for 15 who are alive) and 31 new cases. Twenty-five (40%) patients had low-grade (Type I) lesions (including the only man in the series), 12 (19%) patients had intermediate-grade (Type II) tumors, and 26 (41%) patients had high-grade (Type III) angiosarcomas. Because prognosis is significantly related to the type of angiosarcoma, the lesion should be thoroughly studied microscopically to establish tumor type. Estimated probabilities of disease-free survival 5 years after initial treatment were as follows: Type I: 76%; Type II: 70%; and Type III: 15%. The median length of disease-free survival also was related to tumor type (Type I: greater than 15 years; Type II: greater than 12 years; and Type III: 15 months). Pretreatment duration of the lesion and primary tumor size were not significantly related to the risk of recurrence or to survival. Total mastectomy is recommended for most mammary angiosarcomas. Adjuvant chemotherapy was used in 31 of 63 (49%) patients and, in this group, recurrences developed in 14 (45%) patients. Among 32 (51%) patients not treated with adjuvant chemotherapy, 18 (56%) had recurrences. Whereas the few long-term survivors with Type III tumors received adjuvant chemotherapy, recurrences developed in most patients with comparable lesions who were given this treatment. Because Type III angiosarcoma has a poor prognosis, such treatment may be appropriate for these patients. We were unable to demonstrate that adjuvant chemotherapy significantly improved prognosis for patients with Type I and Type II tumors, although recurrences tended to be less frequent among treated patients.