Introduction

It is now well recognized that term “ductal carcinoma in situ” (DCIS) encompasses a heterogeneous group of lesions clinically, radiographically, histologically, immunophenotypically, and at the molecular level [1, 2]. If left untreated, some cases of DCIS will progress to invasive breast cancer whereas some, perhaps even the majority among certain types, will not [3]. For many years, a major goal of clinical research has been to distinguish patients with DCIS who are more likely to progress to invasive cancer and who, in turn, require more treatment, from those in whom the lesion is unlikely to progress and who require less treatment, or perhaps even no treatment beyond the diagnostic biopsy.

More recently, the management of patients with DCIS has come under even greater scrutiny amidst concerns about overdiagnosis and overtreatment [4,5,6,7,8], particularly at a time when de-escalation of local and systemic therapy for patients with invasive breast cancer is becoming increasingly common [9,10,11,12,13]. The purpose of this article, therefore, is to review the issues related to de-escalation of therapy for patients with DCIS.

Epidemiology and natural history

Most cases of DCIS in current clinical practice are asymptomatic and present as microcalcifications detected on screening mammograms. The impact of mammographic screening on the number of DCIS cases detected over the last several decades has been dramatic. For example, there was an almost ninefold increase in the age-adjusted incidence of DCIS in 2007 relative to 1997 [14]. DCIS currently accounts for ~20–25% of newly diagnosed breast “cancers” with over 60,000 new cases each year in the U.S. [6, 8]. One major question of clinical concern is: how much of this represents overdiagnosis (i.e., lesions that are identified but are biologically innocuous and will never harm the patient)?

DCIS is found at autopsy in up to 14.7% of women dying from other causes (the median prevalence in these studies is 8.9%) [3, 15] and in 0.1–1.1% of women undergoing reduction mammoplasty [16]. These autopsy and reduction mammoplasty studies undoubtedly underestimate the prevalence of DCIS in the population because of the limited sampling of breast tissue in these studies. Therefore, the true prevalence of DCIS in the general population is unknown.

The natural history of DCIS is poorly defined. A few, very small, retrospective reviews of surgical breast biopsies that were originally diagnosed as benign but in which DCIS was found on subsequent histologic review have been used to gain insight into the natural history of this lesion [3, 17,18,19,20] (Table 1). Because these biopsies were initially interpreted as benign, the patients received no treatment beyond the diagnostic surgical excision. However, it is important to note that in these cases, neither the extent of the lesion nor the adequacy of excision were known. In these studies, the reported frequency of subsequent invasive breast cancer ranged from 20 to 53%. These data indicate that if left untreated, some but not all DCIS lesions will progress to invasive breast cancer and that DCIS is, therefore, best regarded as a non-obligate precursor to invasive breast cancer. However, the proportion of DCIS cases that will progress to invasive cancer among the small, mammographically-detected lesions often encountered in current clinical practice is unknown.

Table 1 Retrospective studies evaluating the natural history of ductal carcinoma in situ.
Full size table

Current treatment

Since DCIS poses no threat to life, the major purpose of treatment is to prevent the development of an invasive breast cancer (i.e., prophylactic treatment). In current clinical practice there are three main treatment options for DCIS: mastectomy, breast conserving surgery followed by radiation therapy, and breast conserving surgery alone. Any of these may be accompanied by adjuvant endocrine therapy [21]. In recent years, the combination of breast conserving surgery followed by radiation therapy has become the most frequently used treatment approach to women with DCIS. It is important to note that none of these treatment options offers a survival advantage over the others.

Until the early 1990s, mastectomy was the standard treatment for women with DCIS. However, as increasing numbers of patients with invasive breast cancer were being managed with breast conserving surgery and radiation therapy rather than mastectomy in the late 1980s and early 1990s, it seemed paradoxical to treat patients with DCIS with mastectomy while offering patients with invasive breast cancer the option of breast conserving therapy. No randomized clinical trial has ever compared mastectomy to breast conserving surgery and radiation therapy in women with DCIS. However, four randomized clinical trials have compared breast conserving surgery with radiation therapy to breast conserving surgery alone. A meta-analysis of these four trials from the Early Breast Cancer Trialists Collaborative Group demonstrated that the addition of radiation therapy to breast conserving surgery reduced the risk of local recurrence by ~50%, and further, that there was no subgroup of patients with DCIS that did not benefit from the addition of radiation therapy [22]. A subsequent randomized clinical trial from the NSABP (NSABP B24) demonstrated that the addition of tamoxifen to the combination of breast conserving surgery and radiation therapy reduced the risk of local recurrence by ~30% [23], and that the benefit of tamoxifen was limited to women whose DCIS was estrogen receptor (ER) positive [24]. Other studies have shown that the aromatase inhibitor anastrozole similarly reduced the risk of local recurrence in women with DCIS treated with breast conserving surgery and radiation therapy (IBIS II) or was superior to tamoxifen in women <60 years of age (NSABP B35) [25, 26].

In summary, while standard therapy for DCIS has largely been de-escalated from mastectomy to breast conserving surgery with radiation (with or without endocrine therapy), the addition of radiation therapy and even endocrine therapy to breast conserving surgery is likely overtreatment for some women with DCIS.

De-escalation of treatment

Attempting to further de-escalate treatment for women with DCIS may be accomplished by (1) the identification of women with DCIS at low risk for local recurrence; (2) evaluating the potential role of active surveillance for the management of patients with low risk DCIS; and (3) “de-escalation” of the diagnosis of DCIS by pathologists.

Identification of women with DCIS at low risk for local recurrence

Identification of risk factors for local recurrence of DCIS after breast conserving therapy provides insights into which patients may have “low risk” DCIS and may, therefore, be candidates for further de-escalation of therapy. Many studies have demonstrated that a variety of clinical factors, tumor factors, and treatment factors influence the risk of local recurrence of DCIS and/or the progression to invasive breast cancer (Table 2). Based on the results of these primarily retrospective studies, the combination of features that appears to most consistently characterize DCIS with a lower risk of recurrence and/or progression to invasive breast cancer includes older age, mammographic detection, small size, non-high grade histology, and negative margins of excision.

Table 2 Factors reported to influence risk of local recurrence in the treated breast after breast conserving therapy for ductal carcinoma in situ.
Full size table

There are three prospective studies that have attempted to determine if low risk DCIS lesions can be safely treated with breast conserving surgery without radiation therapy [27,28,29] (Table 3). Unfortunately, these prospective studies have been unable to identify a subset of patients with low risk DCIS treated with breast conserving surgery alone who have local recurrence rates of <10% after long-term follow-up based on conventional clinical and pathologic features. However, the views of what constitutes an acceptably low rate of local recurrence vary. For example, in the RTOG 9084 trial comparing breast conserving surgery and radiation therapy to breast conserving surgery alone for women with low risk DCIS [29], local recurrence rates in patients treated with breast conserving surgery alone were similar to the risk of subsequent breast cancer in women with lobular carcinoma in situ (~1%/year), a lesion that is managed conservatively (i.e., without attempt at surgical excision to negative margins and without radiation therapy).

Table 3 Local recurrence rates in prospective studies of low risk ductal carcinoma in situ treated with breast conserving surgery alone.
Full size table

More recent attempts to refine risk stratification among patients with DCIS have focused on the evaluation of gene expression profiling and biomarker analysis. The Oncotype DCIS score is a 12 gene RT-PCR-based assay that stratifies women with DCIS into low, intermediate, and high risk groups based on the gene expression signature. However, in two cohorts of patients (one derived from the ECOG E5194 prospective study of women with DCIS treated with breast conserving surgery alone and the other derived from the Ontario DCIS cohort) the 10-year local recurrence rates in the low risk groups were 12–13%, local recurrence rates that are too high for many clinicians to avoid recommending radiation therapy [30, 31]. A more recent meta-analysis combined data from these two cohorts and attempted to further refine the risk estimates provided by the Oncotype DCIS score using clinico-pathologic features (i.e., patient age and tumor size). In this analysis, a low risk group with a 10-year local recurrence rate of 7.2% was identified. This group was composed of women ≥50 years of age with DCIS lesions ≤1 cm who had a low risk DCIS score [32]. More recently, a signature that combines several biomarkers detected by immunohistochemistry with clinico-pathologic factors was reported to be predictive of recurrence risk and to predict benefit from radiation therapy [33]. Among patients treated with breast conserving surgery alone this assay (DCISionRT) identified a group of women with DCIS with an 8% risk of an ipsilateral breast event at 10 years in the initial population analyzed [33] and a 10% 10-year ipsilateral breast event risk in a subsequent validation study [34]. Taken together the results of these studies suggest that the combination of clinico-pathologic factors and biological features is better at identifying women with low risk DCIS than either alone.

Clinical trials of active surveillance for DCIS

In an effort to identify whether there is a subset of women with DCIS who require no treatment beyond the initial diagnostic core needle biopsy, several groups initiated randomized clinical trials to compare active surveillance to standard therapy for women with low risk DCIS. Until recently, three such trials were accruing patients: The LORIS trial in the United Kingdom, the LORD trial in continental Europe, and the Comparison of Operative vs Medical Endocrine Therapy (COMET) trial in the U.S. [6, 8, 35] (Table 4). In addition, the LORETTA trial in Japan is a prospective, single-arm trial of active surveillance and endocrine therapy for low risk DCIS [8, 36]. However, both the LORIS and LORD trials have recently been converted to registry trials due to lower than expected accrual. Therefore, the COMET trial is now the only prospective, randomized clinical trial that will be able to provide results comparing active surveillance to standard therapy for women with low risk DCIS.

Table 4 Randomized clinical trials comparing active surveillance to standard treatment for low risk ductal carcinoma in situ.
Full size table

The COMET trial is a prospective, randomized non-inferiority trial comparing guideline concordant care with active surveillance for women with low risk DCIS diagnosed on a breast core needle biopsy. This trial is open to women ≥40 years of age with low or intermediate grade DCIS that is ER positive and/or progesterone receptor positive. Two pathologists must agree on the diagnosis. The presence of comedo necrosis in the DCIS was initially an exclusion criterion, but was subsequently permitted when it became clear that there was poor interobserver agreement among pathologists on the definition of comedo necrosis [37]. Enrollment in the COMET trial is also open to women who have had a surgical excision for DCIS with positive margins but in whom the DCIS otherwise meets enrollment criteria, as well as to women with atypical ductal hyperplasia bordering on DCIS on a core needle biopsy. The use of endocrine therapy is optional. The accrual goal for this trial is 1200 patients; as of June 23, 2020, 537 patients have been enrolled. The primary clinical endpoint of the trial is the development of an ipsilateral invasive breast cancer within 2 years; rates of ipsilateral breast cancer at 5, 7, and 10 years will also be evaluated.

Clinical trials of active surveillance raise the question of how often patients with a core needle biopsy showing DCIS have concurrent invasive carcinoma in the breast. A meta-analysis of 52 studies containing 7350 cases of DCIS diagnosed on core needle biopsy in which a subsequent surgical excision was performed reported a pooled upgrade rate to invasive cancer of 25.9%. However, the more clinically germane question is: what is the upgrade rate for patients who would have been eligible for the clinical trials of active surveillance? A number of studies have attempted to address this question by identifying patients with DCIS who underwent surgical excision but who would have been eligible for one or more of the active surveillance trials based on trial entry criteria and then determining the frequency of invasive cancer in the surgical specimens (Table 5). While some of these studies are limited by small patient numbers, upgrade rates to invasive breast cancer range from 6 to 24% [38,39,40,41]. The clinical significance of these findings remains to be determined, but these studies demonstrate that a proportion of patients enrolled in the active surveillance trials of DCIS will have undiagnosed invasive carcinoma.

Table 5 Upgrade to invasive cancer among patients with ductal carcinoma in situ on core needle biopsy who underwent surgical excision but who would have been eligible for active surveillance trials.
Full size table

De-escalation of the diagnosis of DCIS by pathologists

Given the trend toward de-escalation of treatment, it seems prudent for pathologists to assist in this by attempting to “de-escalate” the diagnosis of DCIS in borderline cases. While the most recent World Health Organization Classification of Breast Tumours reiterates the traditional quantitative criteria of Page and of Tavassoli and Norris for the distinction between atypical ductal hyperplasia and low grade DCIS (i.e., 2 spaces and 2 mm, respectively), it further indicates that “these thresholds are arbitrary and should be used as general guidelines. Because both of these systems were developed on the basis of findings in excisional biopsies, the criteria should be applied with caution, and the WHO Classification of Tumours Editorial Board recommends a conservative approach when lesions of limited extent are identified, particularly in core needle biopsies, in which the entire lesion may not be visualized” [42]. For example, Fig. 1 illustrates core needle biopsies from four different patients containing atypical, low grade, ductal pattern proliferations that range in extent from 2 spaces (Fig. 1a) to 2.5 mm (Fig. 1d). Each of these lesions had been diagnosed as DCIS, all four patients underwent mastectomy, and none had DCIS in the mastectomy specimen. If these cases had been given diagnoses that fell short of DCIS (e.g., severely atypical ductal hyperplasia, severely atypical intraductal proliferation bordering on DCIS, etc.), it is likely that none of these patients would have undergone mastectomy which, in retrospect, was overtreatment for all of these patients. Thus, cases of this type illustrate the value of using the conservative approach to atypical ductal lesions of limited extent as currently recommended by WHO [42]. They further emphasize that, given the concerns about the overdiagnosis and overtreatment of DCIS, pathologists can help mitigate the problem by taking a conservative approach for borderline lesions, especially in core needle biopsies.

Fig. 1: Atypical intraductal proliferative lesions encountered in breast core needle biopsies in which the differential diagnosis was atypical ductal hyperplasia vs low nuclear grade ductal carcinoma in situ.
figure 1

The lesion extents are a 2 spaces; b 1 mm; c 2 mm; d 2.5 mm. All of these lesions were diagnosed as ductal carcinoma in situ on the core needle biopsy specimen and all patients underwent mastectomy. There was no residual lesion identified in the mastectomy specimen of any of these patients.

Full size image

Conclusions

Concerns about overdiagnosis and overtreatment of DCIS have led to efforts to de-escalate therapy. The combination of breast conserving surgery and radiation therapy has replaced mastectomy as the most common form of treatment for DCIS, but this still likely represents overtreatment for many patients. While the reproducible identification of low risk DCIS has remained elusive, ongoing clinical trials of active surveillance of patients with presumed low risk DCIS and their correlative science components will provide important information on the feasibility of further de-escalating therapy for DCIS, provide new insights into factors associated with the progression of DCIS to invasive breast cancer, and perhaps identify new targets for prevention.