Post-radiotherapy vascular proliferations in mammary skin: A clinicopathologic study of 11 cases

Abstract

Post-radiotherapy atypical vascular lesions (AVL) in mammary skin show significant clinical and histopathologic overlap with well-differentiated angiosarcoma (AS) and pose a considerable diagnostic and managerial challenge when encountered. We review Stanford's experience with diagnosing AVL and formulate a clinicopathologic approach to these lesions. We performed a clinicopathologic study on 11 cases that were initially diagnosed as AVL and examined whether there are specific clinical or histopathologic features that delineate AVLs from well-differentiated AS. Clinically, all patients were women with a mean age of 68.1 years, had a history of infiltrating breast carcinoma, and were treated by excision with postoperative radiation therapy. All lesions were located in mammary skin within the prior radiation field. The clinical presentation included erythema, telangiectasias, papules, plaques, and nodules. All patients were diagnosed with AVL on initial biopsy. Six patients showed no recurrence or progression of disease following incomplete excision with no further therapy (3/6) or re-excision with negative margins (3/6). The remaining 5 patients were shown to have AS in the re-excision specimen. The patients diagnosed with AS were older and had a shorter interval from radiation as compared to those who did not experience an adverse outcome. Histologically, all initial biopsy specimens were transected and were characterized by complex, anastomosing vascular proliferations with dilated spaces. Each case was morphologically evaluated according to the AVL criteria of Fineberg and Rosen. Three cases met all of the criteria for AVL, and these patients showed no progression of disease. The remaining cases met most but not all diagnostic criteria for AVL and showed some features of AS, but fell short of a definitive diagnosis of AS, including the 5 cases that were subsequently diagnosed as angiosarcoma. This retrospective study utilized a small number of cases from a single consultation service; therefore, some inherent selection bias may exist. We could not identify unequivocal clinical or histologic criteria that allows for a sharp separation between AVL and AS. Dermatologists and pathologists need to be aware of the overlap between AVL and well-differentiated AS and all patients who receive a diagnosis of AVL should undergo complete excision with close clinical follow-up and biopsy of any new lesions.