As discussed in other chapters in this book, radiation therapy in the adjuvant setting plays an important role in the management of patients with ductal carcinoma in situ (DCIS). Multiple previous studies have established the role of postoperative radiotherapy at reducing the risk of local recurrence after breast-conserving surgery [1–3]. There exist, however, limited data for the role of postmastectomy radiation in women with DCIS, and much of the rationale for postmastectomy radiation in the few DCIS patients who receive this treatment is extrapolated from experiences in women with invasive disease treated surgically with mastectomy.

Even within the context of invasive disease, however, the indications for postmastectomy radiation remain somewhat controversial. The rationale for radiation treatment in this setting, as in the adjuvant treatment of invasive breast cancer after lumpectomy, is to not only mitigate the potential of disease recurrence from disease reservoirs within the chest wall and regional lymphatics but also to eliminate any microscopic disease that might serve as the nidus for distant metastasis. For patients with DCIS, in whom adjuvant treatment after lumpectomy is primarily pursued for local control, the rationale for postmastectomy radiotherapy is similarly more restricted with local control as the primary aim. In both settings, to ensure net benefit from treatment, one must begin by identifying which patients have significant risk of harboring residual microscopic disease that might be eradicated by radiotherapy. The adequate and appropriate identification of patients who are at significant risk of harboring such microscopic disease, however, remains an area of controversy.

In this chapter, we begin by reviewing the role of postmastectomy radiation for invasive breast cancer, including the randomized trials that have established its role in the management of node-positive patients and the retrospective studies that have sought to identify subgroups of node-negative patients who also might benefit. The chapter describes the studies that identified surgical margin status as a risk factor for locoregional recurrence in node-negative patients with invasive cancer. It then proceeds to discuss in detail the few retrospective studies that have explored risk factors for local recurrence after mastectomy for DCIS. Because many of these studies have considered margin status as a risk factor, it reflects on the challenges of margin assessment before reflecting briefly on the limited data regarding outcomes in patients with DCIS treated with postmastectomy radiotherapy. Finally, treatment techniques and expected toxicities in this setting, again extrapolating from the much larger experience in the invasive cancer setting, are summarized. Ultimately, the chapter concludes that the decision regarding postmastectomy radiotherapy in patients with DCIS is challenging, given the limited data available, but that through extrapolation from existing studies, physicians may guide patients to make appropriate choices that reflect their personal values and preferences in this setting.

Before discussing the more limited role of postmastectomy radiation therapy for DCIS, it is important to examine the rationale and experience in the more common setting of invasive disease . Initial randomized trials evaluating the efficacy of postmastectomy radiation in women with invasive breast cancer were limited by increased toxicity and late complications associated with more primitive radiation techniques. Those early studies clearly showed an improvement in locoregional control with the administration of postoperative radiotherapy, but this did not translate into an overall survival benefit given the higher rates of noncancer-related mortality likely associated with the side effects of radiation treatment techniques that, at the time, did not spare the dose to surrounding critical structures including the heart and lungs [4–6]. This, coupled with the lack of effective systemic treatments at the time, meant that effective locoregional control of disease was not translated into an overall survival benefit [1, 7–12]. Despite the limitations to the early postmastectomy radiation trials in women with invasive breast cancer, subsequent trials in a more modern era that incorporated increasingly sophisticated radiation treatment techniques and effective systemic therapies have shown both a locoregional disease control benefit as well as a survival benefit for appropriately selected patients [13, 14]. These trials, which included primarily lymph-node-positive patients as well as locally advanced, lymph-node-negative patients, provide the evidence upon which current recommendations and guidelines are based [1, 2, 15] .

One such study, from investigators in Denmark, evaluated the role of postmastectomy radiation in premenopausal women with invasive breast cancer and clearly demonstrated a substantial reduction on locoregional recurrences (from 32 to 9 % at 10 years) and an overall survival benefit at 10 years of 10 % (from 45 % with no radiation to 54 % with radiation, P-value < 0.001) [16]. Perhaps just as importantly, analysis of the patients and tumor characteristics in this study identified risk factors associated with local recurrence and overall survival that remain foundational to the current indications for postmastectomy radiation in the invasive disease settings. In this study, multivariate analysis identified tumor size, number of involved lymph nodes, grade, age, and the use of radiation therapy as all being significantly associated with, and independent predictors of, outcome in these patients [16]. Subsequent analysis of the data failed to show a difference in survival in patients with left-sided versus right-sided disease and there was no excess risk of ischemic heart disease or death in irradiated versus nonirradiated patients [17]. Thus, in an era of more sophisticated radiation treatment techniques, the abrogation of a survival benefit seen in previous trials was lost once radiation treatment techniques were adapted to limit heart and lung toxicity .

Another trial from the Danish group, this time in postmenopausal women, also not only demonstrated a locoregional disease control benefit, with local recurrences at 10 years reduced from 35 % without radiation to 8 % with radiation, but also confirmed an absolute overall survival benefit at 10 years of 9 % (from 36 to 45 %, P-value 0.03) [18]. A similarly designed Canadian study in postmenopausal women also showed a 20-year survival advantage with the administration of postmastectomy radiotherapy [19]. Perhaps most compellingly, the Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) published meta-analyses that demonstrated the benefit of local control with adjuvant radiation therapy after mastectomy [3]. In this meta-analysis in more than 8000 women who underwent mastectomy and axillary clearance for lymph-node-positive disease, the 5-year local recurrence risk was reduced by the addition of adjuvant radiation from 22.8 to 5.8 % with an associated reduction in 15-year breast cancer mortality risk of 5.4 % (from 60.1 to 54.7 %). Additionally, there was a 4.4 % absolute overall survival benefit at 15 years (from 59.8 to 64.2 %) which indicates that despite the technical limitations of the radiation treatment techniques at the time (and the associated cardiac and lung toxicity), postmastectomy radiotherapy for patients with lymph-node-positive disease provided a significant survival benefit. While the 2005 publication of the EBCTCG meta-analysis suggested a local control benefit for women with node-negative disease, subsequent updated analyses have not demonstrated a clear improvement of any end point, including local control, for women with node-negative breast cancer treated with mastectomy on these trials if restricted to the subgroup of patients who had complete axillary dissection [3, 13] .

These studies have established postmastectomy radiation as the standard of care for patients with invasive breast cancer and at least four positive lymph nodes. Given advances in systemic therapy and low rates of locoregional recurrence even without radiotherapy in selected patients in retrospective modern series, the role of postmastectomy radiation for patients with 1–3 positive lymph nodes is more controversial [20–22]. These studies also suggested that with more modern radiation techniques, postmastectomy radiotherapy can improve locoregional control without any adverse impact on survival. These findings led to increased interest in identifying additional patients, including those with node-negative invasive disease or DCIS, who may also benefit from postmastectomy radiation therapy under certain circumstances .

In addition to the postmastectomy patients with lymph-node-positive disease who clearly benefit from radiation in the adjuvant setting, there remain additional groups of women with invasive disease in whom postmastectomy radiation may be considered. The strongest evidence supporting the use of postmastectomy radiotherapy in node-negative invasive cancer patients has come from the Danish trials, which included patients with T3 and T4 node-negative disease. In those patients, locoregional recurrence was reduced from 17 to 3 % in premenopausal women and from 23 to 6 % in postmenopausal women [22]. Additionally, there was a survival advantage (from 70 to 82 %) in the premenopausal women [22] with the addition of postmastectomy radiation in node-negative invasive breast cancer with large primary tumors. However, this impact was not observed in the Oxford EBCTCG meta-analysis of the impact of postmastectomy radiotherapy in node-negative patients who received mastectomy and axillary dissection [13]. Moreover, retrospective analyses of patients with T3N0 disease have suggested that the risk of locoregional recurrence may be relatively low, and additional selection criteria may be necessary in order to identify a population of node-negative patients who really do have sufficient risk of locoregional recurrence to merit postmastectomy radiotherapy [23, 24].

Several retrospective studies have sought to further explore locoregional recurrence risks in patients with node-negative disease who undergo mastectomy. Such studies have identified possible additional risk factors for locoregional recurrence in lymph-node-negative patients who underwent mastectomy that may have implications in the setting of DCIS. These include young patient age, large tumor size, high nuclear grade, triple negative subtype, and close or positive surgical margins [25–28]. Of greatest relevance when considering patients with DCIS may be those studies which identified close or positive surgical margins as a possible indication for postmastectomy radiotherapy in node-negative invasive cancer [23].

Surgical margins in particular have received much recent attention. Initial indications regarding the effect of margin status on postmastectomy recurrence in women with invasive breast cancer comes from data from British Columbia. In this study, there was a higher likelihood of recurrence in early-stage patients treated with mastectomy with positive surgical margins that was age dependent (all recurrences were in women less than 50 years old) [25]. Additional data from investigators at Harvard suggested that close margins, T2 or larger tumors, premenopausal status, and lymphovascular invasion (LVI) put patients at higher risk for locoregional recurrences, and these patients may benefit from postmastectomy radiation [28]. These older studies, however, included patients treated before the advent of more sophisticated approaches to systemic therapy that may themselves reduce locoregional recurrence risk. In a more recent series from Boston, risks in invasive cancer patients with positive margins were higher than in those with close or negative margins, but considerably lower than in older studies; locoregional recurrence rates at 5 years increased from 1.9 % for women with negative margins to 6 % in women with margin-positive disease [29]. Women with close margins (< 2 mm) had a 1.5 % risk of locoregional recurrence at 5 years [29].

Given that much of the previously mentioned data consider the possible role of surgical margin status as a risk factor for local recurrence of DCIS after mastectomy, it is important to reflect on the additional challenges that arise with regard to the accuracy and reproducibility of margin assessment. Recent consensus guidelines published by the joint Society of Surgical Oncology-American Society for Radiation Oncology consensus panel have nicely summarized the challenges associated with accurately defining margin status [36]. While the consensus guidelines deal with appropriate margins governing the need for re-excision for invasive breast cancer, the challenges regarding margin assessment apply equally to patients with DCIS. One such challenge involves the processing of specimens after resection. Upon removal of the breast tissue, there is flattening of the specimen either from extrinsic compression of the sample or because of a lack of support from the surrounding tissue ex vivo. This leads to artificial narrowing of the margin and may falsely categorize as specimen having close or positive margins [37]. Further confounding of margin status occurs as superficially applied ink (either during surgery or during postsurgical sample pathologic processing) penetrates deeper into tissue, again artificially narrowing the margin. Methods of margin assessment may itself influence the rates of margin positivity, as data suggest that shaved margin assessment results in the categorization of many positive margins that would have been called negative by the inking method, thus leading to increased rates of re-excision and mastectomy [38]. Furthermore, margin status assessment is by necessity a highly selective process as true cell-by-cell assessment of the margin is impractical. Thus, random sampling of margin status is routinely performed but often only examines a tiny fraction of the total margin and thus may miss areas of close or positive margins [39]. Finally, margin status assessment is performed on fixed, inked, sectioned samples in two dimensions and clearly is unable to appreciate three-dimensional architecture. Thus, margin status on any given slide in two dimensions may not accurately reflect true margin status in three-dimensional space a few millimeters superficial or deep to the sectioned tissue. It is therefore important to consider these inherent limitations when examining the data regarding margin status postmastectomy for women with DCIS.

As mentioned previously, there are no prospective, randomized data assessing the efficacy of postmastectomy radiation in women with DCIS . The limited data that do exist suggest that postmastectomy radiation is effective in controlling local disease with exceptionally low rates of local recurrence after postmastectomy radiation, even in women with positive surgical margins. The largest published series comes from investigators at the University of Pennsylvania who tracked the outcomes of 287 women treated with postmastectomy radiation between 1978 and 1992 [40]. Of these 287 women, 1 % (three patients) had DCIS and underwent mastectomy because of diffuse microcalcifications on screening mammography. All three women subsequently received postmastectomy radiation because of positive surgical margins after mastectomy. All patients were clinically node negative, and one of the women did undergo axillary lymph node dissection with 0 out of 10 lymph nodes positive. All women received between 42.75 and 50 Gy in 1.8–2.25 Gy daily fractions delivered to the chest wall (one with 1-cm bolus applied every other day) utilizing tangential fields without regional nodal irradiation. None of the patients received adjuvant endocrine or systemic chemotherapy . With a median follow-up time of 7.4 years (7.1–19.4 years), all patients were alive and disease-free with no evidence of local or distant recurrence. There were no contralateral breast events and the authors reported no significant long-term side effects. However, given the small numbers of patients with DCIS included, it is difficult to draw conclusions from these findings .

Given the limited data regarding the role of postmastectomy radiation for DCIS, treatment guidelines have been developed based on institutional experience and have been borrowed from experience and principles gleaned from trials in the invasive cancer space. Indeed, data from multiple randomized trials involving thousands of patients are unequivocal in demonstrating that adjuvant radiation significantly decreases the rates of local recurrence after lumpectomy or mastectomy for invasive breast cancer and also significantly decreases local recurrence after lumpectomy for DCIS. More limited data also demonstrate excellent local control rates when adjuvant radiation is utilized in the postmastectomy setting, even for women with positive margins. The important unanswered question, however, is whether patients with close or positive surgical margins after mastectomy for DCIS are at sufficiently high risk to justify the use of the adjuvant postmastectomy radiation. As noted above, different retrospective series have generated widely varying estimates of risk in patients with DCIS who have close or positive margins after mastectomy.