The increasingly frequent use of radiation therapy (RT) and systemic chemotherapy (CT) in the treatment of breast carcinoma requires surgical pathologists and cytologists to be familiar with the variable histologic changes initiated by these agents. Both treatment modalities can cause severe epithelial abnormalities, which are difficult to distinguish from carcinoma. The progression or regression of these histologic abnormalities in nonneoplastic breast tissue have not been extensively evaluated. Our study used 120 post-RT biopsy or mastectomy specimens from 117 patients (3 had bilateral carcinoma treated with RT) yielding 120 specimens. The interval from post-RT to biopsy or mastectomy ranged from 1 to 229 months with 25 of the specimens obtained 1 to 12 months after cessation of RT and 95 more than a year post-RT. Twenty-seven specimens were from >6 years after RT. The histologic features of pretreatment and posttreatment specimens were graded (0–3) blindly for each histologic feature to include stromal vascular and fibroblastic changes and epithelial cell changes of the terminal duct lobular unit and extralobular ducts as well as terminal duct lobular unit fibrosis/atrophy. The changes between the pre- and post-RT grades were all statistically significant (P < 0.05) using multiple nonparametric statistical methods and the parametric Student t test. The specimens obtained within the first year post-RT were compared with those from >1 year post-RT, >3 years post-RT, and >6 years post-RT. None of the histologic features evaluated showed significant changes over the various time intervals regardless of the statistical method used. The absence of regression of the radiation-induced histologic changes over time mandates the surgical pathologist be alert to the possibility of RT or CT even without that therapeutic history.
An estimated 211,300 new cases of invasive female breast cancer are expected to occur in the United States during the year 2003 and 55,700 new cases of in situ breast carcinoma are projected. 7 Breast-conserving therapy consisting of tumor excision followed by radiation is used with increasing frequency and has been demonstrated to be equally effective as mastectomy. 1,12,14,23,30 However, recent studies 13 have questioned the long-term (25 year) survival advantage from radiation therapy (RT) among some groups of treated women.
This shifting of therapeutic modalities from mastectomy to include radiation and/or systemic chemotherapy (CT) mandates pathologists understand and be alert to the variable histologic changes increasingly introduced by these therapies. Many of these patients require posttherapeutic biopsy or mastectomy because of the development of new clinically or mammographically detectable lesions within the breasts. Of course, CT can produce changes that may mimic the cytologic and histologic changes resulting from RT. 2,6,15,19,24
Other authors 10 have emphasized that histologic studies of radiated normal breasts are scarce and urgently needed. Bentzen and Dische 3 have compared the clinical aspects of the late morbidity of RT to Damocles' sword. Many clinical studies of late morbidity of the radiated breast have been published, 4,5 but few if any studies have focused on the late changes of the histologic features of the normal breast tissue.
The published literature 27 suggests that there is absence of progression or regression of the radiation-induced histologic changes of the nonneoplastic breast tissue regardless of the interval from the termination of RT to the time the post-RT specimen was obtained. These studies also suggested that the use of adjuvant CT produced no differences. These variables were carefully evaluated in our larger series of cases to determine if repair in post-RT injury is operative in ameliorating or enhancing the RT-induced histologic changes over both short and long periods.
The atypical epithelial cells and stromal changes produced by RT injury can be challenging to differentiate from carcinoma. It is imperative that the surgical pathologist have a thorough understanding of the many facets of RT and CT injury of the breast. This discernment is complicated by the observations that characteristic pathognomonic individual features of radiation effect are absent in any mammalian tissue. However, when analyzed together, the histologic changes are usually strongly suggestive of radiation injury. 10
Radiation-induced histopathologic changes have been described in several scientific papers 16,27 and books. 10,11,26 Some of these sources also contain excellent photomicrographs. This current study of 120 cases concentrates on the effects of time.
From the Departments of *Pathology and †Radiation Oncology (J.E.S.), Penrose Hospital, Colorado Springs, CO, and the ‡Department of Education, Colorado State University, Fort Collins, CO.
Reprints: Gene H. Moore, MD, Department of Pathology, Penrose Hospital, Box 7021, Colorado Springs, CO 80933 (e-mail: genemoore@Centura.org).