Ductal carcinoma in situ (DCIS) of the breast has rapidly grown in number of cases over the past several decades, due to increases in mammographic screening. It currently represents up to 20% of newly diagnosed breast cancer [1]. DCIS is a biologically heterogeneous group of noninvasive lesions in the breast, characterized by proliferative malignant ductal cells limited to existing ductal units, without invasion through the basement membrane. In the last few years, the increasing occurrence of very small, radiologically detected subclinical lesions has led some authors to consider most DCIS as a possibly indolent disease and consequently propose a less intensive treatment. However, DCIS has a potential for progression to invasive carcinoma, usually within the first 10 years after initial diagnosis, which characterizes DCIS as a preinvasive or precursor lesion and as a continuum along the mammary neoplastic transformation process [2].

Historically mastectomy was considered the therapy of choice, since having only 1–2% of local recurrence and associated with a survival rate of 98% [3, 4], but it is currently considered as overtreatment in many cases, especially in the presence of small, non-palpable lesions. Nowadays, breast conserving therapy, consisting of breast-conserving surgery with adjuvant radiation therapy, is considered the standard of care with regard to local management for eligible patients.

No randomized trials directly comparing mastectomy versus breast-conserving surgery with radiation therapy have been performed in patients with DCIS. Therefore, the efficacy of breast-conserving therapy is often extrapolated from randomized trials in patients with early-stage invasive cancers, which have confirmed that this treatment is not associated with inferior outcomes and provides equivalent survival to mastectomy. Four randomized trials investigated the role of the addition of postoperative radiation therapy after breast-conserving surgery and showed a significantly reduced risk for local recurrences when adjuvant radiotherapy was administered. These four randomized trials are the NSABP B-17 trial , the EORTC 10853 trial , the SweDCIS trial , and the UK/ANZ DCIS trial . The NSABP B-17 trial [5] enrolled 818 patients with localized DCIS treated by lumpectomy that achieved tumor-free margins between October 1985 and December 1990. These patients were randomly assigned to the lumpectomy-only group or to the lumpectomy followed by radiotherapy group. Endpoints of this study included invasive ipsilateral breast tumor recurrence, DCIS-breast tumor recurrence, contralateral breast cancers, overall survival, breast cancer-specific survival, and survival after invasive ipsilateral breast tumor recurrence. Median follow-up was 207 months. Radiation therapy reduced invasive ipsilateral breast tumor recurrence by 52% in the lumpectomy followed by radiotherapy group compared to the lumpectomy-only group, with an hazard ratio of risk = 0.48, 95% confidence interval = 0.33–0.69 with P < 0.001. The 15-year cumulative incidence of invasive ipsilateral breast tumor recurrence was 19.4% for the lumpectomy -only group and 8.9% for the lumpectomy followed by radiotherapy group. The 15-year cumulative incidence of all contralateral breast cancers was 10.3% for the lumpectomy-only group and 10.2% for the lumpectomy followed by radiotherapy group. The invasive ipsilateral breast tumor recurrence was associated with increased mortality risk (hazard ratio of death = 1.75, 95% confidence interval = 1.45–2.96 with P < 0.001), whereas recurrence of DCIS was not. Twenty-two of 39 deaths after invasive ipsilateral breast tumor recurrence were attributed to breast cancer. Among all patients, the 15-year cumulative incidence of breast cancer death was 3.1% for the lumpectomy -only group and 4.7% for the lumpectomy followed by radiotherapy group.

In the EORTC 10853 trial [6], 1010 patients with DCIS treated with breast-conserving surgery were randomly assigned to adjuvant radiotherapy (507 patients) or no further treatment (503 patients).

Twenty-six patients (5%) randomly assigned to the radiotherapy group received a boost. The median follow-up time was 15.8 years. Radiotherapy reduced the risk of any local recurrence by 48%, with a hazard ratio of 0.52 and 95% confidence interval from 0.40 to 0.68, with P = 0.001. The 15-year local recurrence-free rate was 69% in the group of patients who underwent a local excision only, compared to 82% in the radiotherapy group, while the 15-year invasive local recurrence-free rate was 84% in the local excision-only group and 90% in the patients assigned to adjuvant radiotherapy, with a hazard ratio of 0.61 and 95% confidence interval from 0.42 to 0.87. An overall salvage mastectomy rate after local recurrence was lower in the radiotherapy group (13%) rather than in the local excision-only group (19%).

Almost one in three nonirradiated women developed a local recurrence after local excision for DCIS, and radiotherapy reduced this risk by a factor of two.

The differences in local recurrence in both arms did not lead to differences in breast cancer-specific survival or overall survival. Patients with invasive local recurrence had a significantly worse breast cancer-specific survival and overall survival compared with those who did not experience recurrence, but the long-term prognosis was good and independent of the given treatment.

In the SweDCIS trial [7], 1067 women in Sweden from 1987 to 1999 were randomly assigned to postoperative radiotherapy or control. The main outcome was new ipsilateral breast cancer events, and distant metastasis-free survival analyzed according to intention to treat. In this study, 64 ipsilateral breast events occurred in the patients who underwent a radiotherapy treatment and 141 in the control group, corresponding to a risk reduction of 16.0 percentage points at 10 years and a relative risk of 0.40. In the radiotherapy group 59.4% and in the control group 45.4% of the ipsilateral events were invasive. The authors showed that radiotherapy reduced the risk of invasive and in situ events similarly. A total of 18 events of metastatic breast disease and breast cancer deaths occurred in the radiotherapy group and a total of 15 in the control group, but there was no statistically significant difference in distant metastasis-free survival. In this study, radiation therapy has proven more effective in women older than 60, compared to women younger than 50. The age effect was not confounded by focality, lesion size, completeness of excision, or detection mode. However the data regarding the effect of age come from a subgroup analysis and should be considered with caution.

In the UK/ANZ DCIS trial [8], 1701 patients, of which 1694 eligible for analysis, were randomized to receive radiotherapy, tamoxifen, or both: 912 patients chose to enter into two by two randomization, to radiotherapy and tamoxifen (242 patients) and to tamoxifen only (224 patients) or to radiotherapy only (220 patients) and to not treatment (226 patients); 782 chose to enter into randomization to one of the treatment; 664 chose radiotherapy and were only randomized to receive tamoxifen or not; and 118 chose tamoxifen and were only randomized to receive radiotherapy or not.

The radiotherapy dose was 50 Gy in 25 fractions over 5 weeks; the boost to the surgical bed was not administered; tamoxifen was prescribed at a dose of 20 mg daily for 5 years. In the patients randomly assigned to tamoxifen, the authors found an absolute 10-year reduction of 3.9% for all ipsilateral events and of 2.3% in all contralateral events, with an absolute 10-year reduction of 6.5% for all new breast events. Patients randomized to radiotherapy had fewer new breast events, with an absolute reduction of 12.6%; radiotherapy significantly reduced all ipsilateral events, whereas no effect was reported in relation to contralateral events. Furthermore there was no significant difference in new breast events between patients randomly assigned to radiotherapy and tamoxifen and those randomized to radiotherapy alone; instead the differences are significant among patients randomized to receive radiotherapy plus tamoxifen compared to those randomized to receive tamoxifen only.

Within the four prospective randomized studies, the greatest benefit was observed in patients with high-grade lesions, with positive margins, and in the elderly (age > 50 years); instead a statistically significant difference did not emerge in the incidence of distant metastases and overall survival. A significant increase in mortality from cardiovascular diseases, in the group of patients undergoing postoperative radiotherapy, was not found, except in the UK/ANZ trial, in which, however, the number of cases was very limited.

Also the results of three meta-analysis showed a greater local control when the conservative surgery was followed by radiotherapy. From the meta-analysis of Viani et al. [9], it emerged that the addition of radiation therapy to lumpectomy results in approximately 60% reduction in breast cancer recurrence, in the absence of benefit for survival or distant metastases compared to excision alone. Patients with high-grade DCIS lesions and positive margins benefited most from the addition of radiotherapy. In this meta-analysis it was reported higher rates of contralateral breast cancer in the group of patients undergoing a postoperative radiotherapy; in the subsequent two meta-analyses in which the four clinical trials were examined after a longer follow-up period, the differences were minimal and not statistically significant.

The Cochrane Database of Systematic Reviews [10] showed a 50% reduction in local recurrence in DCIS patients with postoperative radiotherapy with similar rates of reduction noted for invasive and noninvasive recurrences.

The Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) meta-analysis confirmed the reduction of the rate of ipsilateral breast tumor recurrence compared to surgical resection only (12.9% versus 28.1%), with an absolute reduction in local recurrence at 10 years of 15.2% with the addition of radiotherapy. The impact of postoperative radiation therapy compared to surgery alone was similar in terms of both invasive and in situ local recurrence: 6.9% versus 15.4% and 6.5% versus 14.9%, respectively. The benefit of the radiation therapy was independent of diagnostic modality (clinical or radiological), age, type of surgery ( lumpectomy or quadrantectomy), nuclear grade, the presence of comedo necrosis, architectural subtype, tumor size, margin status (free, close or unknown), and tamoxifen administration [11].

Risk factors for local recurrences can be stratified into three groups: clinical factors, histopathological factors, and treatment-related factors.

The clinical risk factors are represented by the clinical presentation and patient age. Diagnostic modalities and mammographic characteristics are very important because symptomatic DCIS patients, with, for example, skin retraction and sero-hemorrhagic nipple discharge, have a higher local recurrence risk than those for whom the disease was radiologically diagnosed.

In the EORTC 10853 trial [6], the local recurrence relative risk was 1.48 for DCIS detected by clinical examination compared to mammography-detected lesions, after both local excision (27% versus 16%) and local excision followed by postoperative radiation therapy (17% versus 11%). Finally, an accurate analysis of microcalcification subtype in the SweDCIS trial showed that “crushed stone and casting type ” were associated with a higher histopathological grade and more extensive disease, and the relative risk of local recurrence for the casting type was 2.1 [7].

The young age, generally considered under 40 or 50 years old, is one of the most important parameters related to the incidence of local recurrence, both in the clinically palpable DCIS cases and in the hidden forms, diagnosed with mammography. This leads some centers to have a more aggressive approach in young women, including a higher mastectomy rate and a more frequent use of a boost after whole-breast irradiation in order to obtain a better local control.

In the SweDCIS trial , the authors found an interaction by age and effect of radiotherapy that indicated a lower effect of radiotherapy in the young patients. The cumulative incidence in the radiotherapy arm was 20% in the youngest age group, falling to 8% among those age 65 and older. There was thus a modest absolute risk reduction in younger women (6%) and a substantial reduction (18%) in older women [7].

In the NSABP B-17 trial , women younger than 45 years showed a 2.1-fold increased risk of invasive ipsilateral breast tumor recurrence compared with women aged 65 years and older at diagnosis; women aged 45–64 years also showed an increased risk of invasive ipsilateral breast tumor recurrence relative to women aged 65 years and older [5].

The histopathological factors are tumor size, nuclear grade, the state, and the magnitude of the margins.

In the literature, the DCIS size is characterized by a wide variation in assessing, recording, and reporting. In the EORTC 10853 trial [6], only 25% of the lesions were measurable with precision, with clear dimensions expressed in mm. In the survey of Cutuli et al., indeed, the DCIS size was identified in 97% of the patients: lesions <10 mm, 10–20 mm, and >20 mm were found in 41%, 27%, and 32% of the cases, respectively [12].

In general the correlation between DCIS size and local recurrence still remains poorly documented. However, a recent report from Alvarado et al. [13] showed a significant influence of tumor size on local relapse, with 5-year recurrence rates of 5.6% versus 2.2% for lesions over and under 15 mm, respectively.

In a Chinese observational study of tumor subtype, treatment and outcome of breast carcinoma in situ, the authors showed a decreased overall survival only in patients with a tumor size >50 mm [14].

Nuclear grade was analyzed in many cohort of patients, and it is well known that about one third of DCIS appears with complex histologic patterns, including the presence of varying nuclear grades within the same lesion. High nuclear grade has always been correlated with an increased rate of local recurrence.

In the EORTC 10853 trial [6], the 10-year local recurrence rates were 18%, 34%, and 35% for grade low, intermediate, and high, respectively, in the surgery-alone group and 9%, 23%, and 19% in the surgery with the addition of radiation therapy group. Also in the UK/ANZ DCIS trial, the high grade, in addition to large size and young age, was significant predictor to a high recurrence rate [8].

The Van Nuys team reported an increase of a 12-year local recurrence rate correlated with the nuclear grade: 13%, 23%, and 45% for low, intermediate, and high grade DCIS, respectively [15].

Many studies have shown a lower incidence of local recurrence in the presence of histologically negative margins, both after radiation therapy, than in patients treated with conservative surgery alone; however, the optimal margin distance remains a topic of debate. The definition of a negative margin varies widely from one study to another (1, 2, 3, 5, or 10 mm or untouched ink); also many reports on margin status are retrospective and lack a standardized assessment of margins, in terms of orientating, inking, and specimen sectioning. Furthermore the rate of close or uncertain margins widely varies among the series [2].

In the Dunne et al.’s meta-analysis of 22 studies, both retrospective and prospective randomized, including only patients treated with conservative surgery and radiotherapy, it has been a statistically significant reduction in the risk of local recurrence in the presence of negative margins ≥2 mm [16].

In the Wang et al.’s meta-analysis of 21 studies, both retrospective and prospective randomized, including patients treated with only conservative surgery or conservative surgery plus adjuvant radiotherapy, best results were seen with free margins ≥10 mm [17].

However, the advantage achieved with margins greater than 2 mm appears to be less pronounced in the subgroup of patients undergoing radiotherapy after conservative surgery. Therefore currently there is a broad consensus that margins ≥2 mm are adequate, when conserving surgery followed by radiotherapy.

Many authors have tried to identify a subgroup of patients with a presumably low risk of recurrence for whom radiotherapy could safety be omitted.

In these last years, many retrospective and prospective studies have evaluated outcome of DCIS patients with breast conservative surgery alone, without postoperative radiotherapy.

In a pooled analysis of the French Regional Cancer Center , 705 patients with breast DCIS were treated with excision alone between 1985 and 1995. The ipsilateral breast tumor recurrence in these patients was 32.4% for conserving surgery alone compared to 12.6% with the addition of radiation therapy [18]. Also in a review of Rakovitch et al. emerged a higher rate of local recurrence without radiotherapy: 19% versus 13% in over 3500 patients evaluated [19].

The prospective phase II trial of the Dana-Farber Cancer Institute [20] enrolled 158 patients affected by low-/intermediate-grade DCIS with tumor size ≤2.5 cm and surgical margins >1 cm. The use of tamoxifen was not permitted. The study was closed early due to a higher rate of local recurrence, 12% ipsilateral tumor recurrence at 5 years, with a local recurrence rate of 2.4% per patient year. In a subsequent publication, with a further follow-up, the local recurrence rates were 13% and 15.6% after 8 and 10 years, respectively.

The Eastern Cooperative Oncology Group (ECOG) E-5194 trial enrolled 670 patients with low/intermediate DCIS measuring ≤2.5 cm or high-grade DCIS with tumor size ≤1 cm and ≥3 mm or wider surgical margins in all cases. Tamoxifen was administered only in 30% of cases. The results of this trial showed that the rate of ipsilateral breast recurrence was 6.1% for low-/intermediate-grade DCIS and 15.3% for high-grade DCIS at 5 years and 15.4% for low-/intermediate-grade DCIS and 15.1% for high-grade DCIS at 10 years. The risks of developing an ipsilateral breast event and an invasive ipsilateral breast event increased over time through 12 years of follow-up. The 12-year rates of developing an ipsilateral breast event were 14.4% for cohort 1 and 24.6% for cohort 2, and the 12-year rates of developing an invasive ipsilateral breast event were 7.5% and 13.4%, respectively. No clearly defined plateau was observed for either cohort of patients [21, 22].