The introduction of national mammographic screening programmes and the increasing use of digital mammography and MRI have dramatically changed the clinical presentation of DCIS. Prior to this, DCIS made up a small proportion of all breast malignancy and was only diagnosed in patients presenting with a palpable mass, pathological nipple discharge, or occasionally found as an incidental biopsy finding [
Although the rates of all breast malignancy have increased with time, between 1980 and 1995, Western countries have experienced a four-fold “increase” in the incidence of DCIS specifically, particularly in women of screening age [
Data from a systematic review of 374 studies reported the pooled incidence of DCIS in the early 1970s as 5.8/100000 and this had risen to 32.5/100000 in 2004 [
Screening and cancer registry data from Norway including 2.3 million women reported in 2010 showed an increase in incidence of DCIS from 4/100000 before the introduction of screening to 11/100000 postintroduction. In women of screening age, the proportion of DCIS within breast malignancy rose from 5% to 13%. Age-standardised rates of all breast cancer including DCIS increased over time in those of screening age, but a large peak at the point screening was introduced, subsequent drop in incidence (but not to prescreening levels), then a steady climb over time. Rates were also higher in prevalent as opposed to incident screens [
These studies seem to suggest that the introduction of screening is largely responsible for the apparent increased incidence of DCIS in recent times, but that the stage of the disease may be much earlier and possibly less clinically relevant [
The trend is likely to continue with further technological advances, including the transition from analogue to full-field digital mammography (FFDM) and the development of computer-aided detection (CAD) [
Although the role of MRI in the management of DCIS is yet to be fully evaluated by randomised trials, it is being used to assess disease extent and distribution and to assess the contralateral breast [
MRI can overinterpret nonmalignant incidental lesions which may result in unnecessary interventions [
Figure
MRI appearance of recurrent DCIS.
Mammary ductoscopy has been used to directly visualise DCIS. Figure
Appearances at mammary ductoscopy of DCIS.
Currently, the preoperative diagnosis of impalpable lesions suspicious of DCIS requires either stereotactic or MRI guided core biopsy. Vacuum-assisted core biopsy (VACB) has been shown to increase the diagnostic yield and upgrade atypical ductal hyperplasia (ADH) to DCIS in approximately 25% of cases [
DCIS is defined by two features: firstly, the malignant epithelial proliferation is limited by the ductal basement membrane and secondly, stromal invasion is absent. DCIS behaves as a nonobligate precursor of invasive carcinoma and does not fully express the malignant phenotype [
DCIS may be classified by grade, by architecture or morphology, by the level of differentiation, or by systems which use a combination of these factors [
Conventional histopathological types include comedo (tending to high grade cellular/nuclear features, often with central necrosis and calcification), solid, cribriform (with small holes or open spaces), and micropapillary (finger-like projections), however, lesions often demonstrate architectural and morphological heterogeneity [
Cytonuclear grade is conventionally defined as low, intermediate, or high. It may vary between pathologists [
Low-grade DCIS.
High-grade DCIS.
The “Comedo” subtype (high-grade, central confluent necrosis, and solid architectural pattern in >50% of the duct spaces) and the presence or absence of necrosis are important features and are incorporated into classifications such as the Van Nuys Index [
All of the above classification methods as well as tumour size and the presence of absence of inflammatory changes have been found to be statistically associated with the risk of local recurrence in an independent pathological review of cases from the UKCCCR/ANZ DCIS trial [
The elusive natural history of DCIS probably reflects the biodiversity of the condition. Preinvasive lesions do not invariably progress to invasive malignancy [
The natural history of small, noncomedo, low-grade DCIS treated by biopsy alone has been evaluated in long-term followup studies. After a median of thirty-one years, 39% of patients developed invasive breast cancer, all of which occurred in the index quadrant and 45% of these patients died of metastatic disease [
There is a wide body of evidence on the risk factors for breast cancer overall, but evidence on risk factors specific to DCIS is limited. However, it does appear that the same factors are involved as for invasive disease; high mammographic density, significant family history of breast-cancer, age, obesity, and high lifetime exposure to oestrogens [
Hence, it would seem that patients who receive no treatment beyond a diagnostic biopsy remain at significant risk of progression to invasive disease and that DCIS represents a precursor of invasive cancer. Increased risk has been demonstrated in lesions of all nuclear grade. On the other hand, a significant proportion of DCIS lesions do not progress. As diagnostic frequency continues to increase, there is an impetus to accurately identify clinically relevant lesions in order to rationalise management.
Women with palpable DCIS and those who present symptomatically exhibit higher rates of LR than mammographically detected cases [
One study identified a family history of invasive breast cancer as a significant predictor of LR in women with DCIS treated with BCS and RT [
Young age (<40 years) has emerged as an independent risk factor for LR after BCS with or without adjuvant RT [
A meta-analysis of 44 trials has reported significantly increased pooled risk estimates for local recurrence after treatment for DCIS if the disease is classified as “comedo” type, multifocal, if the lesion is large or highgrade. Involved margins were associated with the highest increase in risk estimates and there was limited evidence that ER- and PR-positive HER2 negative disease is less likely to recur [
In the meta-analysis, the pooled risk estimates for size were derived from 7097 women. Lesions greater than 20 mm in size were compared with lesions less than 20 mm. The risk estimate for larger lesions was 1.63 (95% CI 1.30–2.06). Accurate and reliable measurement of DCIS can be challenging and several landmark studies have been criticised for their performance in this regard [
Involved margins are associated with an increase in LR, in patients treated by BCS alone, and in those who also undergo RT [
High nuclear grade is associated with a greater risk of LR. In Wang’s meta-analysis [
The combination of nuclear grade and comedonecrosis is strongly associated with the risk of LR after BCS [
A recent population-based case-control study found that comedo-type DCIS shares a similar profile of hormonal and reproductive risk factors to IBC, including ≥10 years of oral contraceptive intake and an inverse association with ≥3 full-term pregnancies. These findings were in contrast to those for noncomedo lesions, providing some further support for the differential management of DCIS lesions [
High-grade DCIS which is oestrogen receptor (ER) and progesterone receptor (PR) negative is significantly associated with HER2 and p53 positivity [
In the Wang meta-analysis [
Various molecular markers have been studied in DCIS as possible predictive or prognostic factors for progression to invasion or for the development of invasive recurrences.
In invasive breast cancer, classifications based on biological profile (derived from gene profiling and correlated with immunohistochemical profile) rather than morphology have been developed and shown to correlate with prognosis. In order; Luminal A, Luminal B, Triple negative, and Basal Type invasive breast cancers are associated with a worsening prognosis [
Chromosome-wide comparative genomic hybridization has shown DCIS to be a genetically advanced lesion with alterations corresponding to adjacent invasive disease and independent pathways of genetic evolution [
One such study has identified a gene expression classifier of 35 genes which differ between DCIS and IBC and a further 43 genes distinguishing well-from poorly differentiated DCIS [
Balleine et al. recently reported on a binary molecular grading scheme for DCIS, based on expression at 173 oligo-nucleotide probes. Two conventional parameters amenable to routine evaluation (nuclear grade and Ki67 score) were capable of accurately assigning lesions into low or high molecular grade [
Proteomics analysis of DCIS and normal breast tissue has also identified differential expression patterns, distinct from previous nucleic-acid-based studies [
Intuitively, molecules such as matrix-metalloproteinases (MMPs) and tissue inhibitors of matrix-metalloproteinases (TIMPs) that influence the invasion of stroma and basement membrane should be important in the progression of DCIS to invasive breast cancer. Significantly different expression profiles of MMPs and TIMPs have been noted in DCIS, admixed DCIS, and invasive breast cancer [
It is possible that not all DCIS needs to be treated aggressively as not all DCIS will become invasive. In particular, small, low-grade lesions detected by screening may fit into this group. Management strategies need to consider the breast and axilla, the need for adjuvant RT, and the role of systemic adjuvant therapy. Treatment of the breast can involve BCS (with or without RT) or mastectomy (Mx). Axillary surgery, even SLNB, warrants particular caution in view of their low yield and potential for harm. Adjuvant systemic treatments have mainly involved oestrogen blockade with Tamoxifen. The optimal management of DCIS remains controversial [
Complete excision of DCIS with clear margins is the most important factor in reducing the risk of LR. Mx is indicated for large tumours (>4 cm depending on breast size), multicentric lesions, inadequate margins after BCS, local recurrence after BCS (particularly with prior RT), and patient preference. Mx affords excellent local control, approximately 98% at 7 years, with an overall recurrence rate of 1.5% [
In England and Wales between 1990 and 2001, the absolute number of mastectomies for
If patients do require Mx for DCIS, an immediate breast reconstruction is relatively uncomplicated as postmastectomy radiotherapy and lymph node dissection will not be required [
BCS combined with RT is an acceptable treatment option for smaller, unifocal areas of DCIS. There is probably not enough evidence to justify BCS without RT routinely. Significant numbers of patients undergoing BCS alone develop LR, of which approximately half are invasive and up to one fifth ultimately metastatic. The literature reveals an overall LR rate of approximately 28% at 7 years, around 45% of which are invasive [
The benefit of adjuvant RT, in terms of reduced LR in those undergoing BCS, has been demonstrated by several large randomized controlled trials. However, clear margins are necessary even if RT is given to obtain acceptable rates of LR [
The National Surgical Breast and Bowel Project (NSABP B-17) trail randomized 818 patients after BCS surgery for DCIS, to either whole breast RT or no further treatment [
An analysis of long-term data from the NSABP B-17 and NSABP B-24 trials [
The European Organisation for Research and Treatment of Cancer (EORTC) conducted a similar study recruiting 1010 patients [
The UK/ANZ DCIS trial involved 1701 patients treated by BCS, with subsequent randomisation to RT and/or Tamoxifen [
The Cochrane Collaboration has recently published a systematic review of four adjuvant RT trials: NSABP 2001 [
A further meta-analysis also concluded that adjuvant RT significantly reduces the risk of LR after BCS—by approximately 60%, with most benefit to patients with high-grade lesions and positive margins. RT did not significantly alter the rate of distant metastases or overall survival [
Overall, LR rates have been reported to range from 2.7% to 18.9%, averaging 10% at 7 years, with invasive LR accounting for approximately 60% [
Although it has been long been proven that radiotherapy after mastectomy for invasive breast-cancer reduces local recurrence [
Strategies such as a boost of RT to the tumour bed are used in IBC. There is no evidence that this reduces LR in DCIS. A study of 75 patients treated by BCS+RT, including 20 women receiving an additional 10 Gy boost to the tumour bed, identified no improvement in LR reduction after a median followup of 81 months [
The ECOG group (Eastern Cooperative Oncology Group) prospectively studied 565 nonrandomised patients with single areas of DCIS less than 2.5 cm in size treated by breast conserving surgery alone, with margins of greater than 3 mm and split these patients into a low and intermediate group versus a high-grade group [
The Radiation Therapy Oncology Group trial (98-04) was a randomised trial designed to assess the need for radiotherapy for DCIS in patients with “low-risk” but unfortunately closed due to nonaccrual. A recent study attempted to account for the nonrandomisation in the ECOG DCIS study by comparing two groups of patients (low and intermediate or high grade DCIS) that were treated with breast-conserving surgery and radiotherapy with the two groups in the ECOG study [
More evidence is needed to confirm if there is a subgroup of patients with DCIS that do not need radiotherapy after breast conservation.
Hormonal therapies (mainly Tamoxifen) are the main stay of systemic adjuvant therapy in DCIS.
The NSABP B-24 trial was designed to assess the benefit of Tamoxifen for 5 years versus placebo after BCS and RT for DCIS [
The UK/ANZ DCIS trial also assessed the effect of adjuvant treatment with Tamoxifen after BCS and RT for DCIS. The results were originally reported after a median followup of 4.4 years [
There is, therefore, good data that Tamoxifen reduces local recurrence and the risk of contra-lateral tumours in DCIS treated by BCS and RT. Some DCIS is probably low-risk enough to omit it, but clear evidence on this is lacking.
There is currently only limited data on the use of aromatase inhibitors in DCIS.
Trials are ongoing to determine if Aromatase inhibitors are superior to Tamoxifen in the adjuvant setting after breast conserving surgery for DCIS (NSABP B-35 and IBIS II).
Recently, inhibition of cyclo-oxygenase 2 (COX-2), implicated in epithelial-stromal interactions and promoting the progression of DCIS, has been evaluated using nonsteroidal anti-inflammatory drugs (NSAIDS). Results from experimental studies were encouraging [
Pure DCIS does not exhibit lymphatic or vascular invasion so surgical staging of the axilla is not necessary [
Retrospective analyses from the NSABP B-17 and B-24 trials support the strategy of avoiding routine axillary surgery in DCIS due to low yield and risk of morbidity [
If LR occurs after DCIS, it may be
Completion Mx is indicated following LR within the breast when reexcision would be cosmetically unacceptable, or when LR is confirmed to be invasive and for those with an absolute or relative contraindication to RT (i.e., previous adjuvant RT). In the NSABP B-17 trial, the Mx rate for LR was 48% in the BCS group and 62% in the BCS + RT group [
Management of the elderly DCIS patient (particularly those over 70 years) is not strongly evidence based as this group has often been excluded from important trials and screening programs [
Women exposed to thoracic radiation, including prior treatment for haematological malignancies, are at risk of developing secondary tumours, with breast cancer representing the most common solid lesion and DCIS accounting for 11–17.7%. The risk is significantly increased at adolescence and young adulthood with a median onset interval of 16 years. In one study, the majority of these patients were treated with Mx, however, 29% underwent BCS
Male DCIS has been reported in approximately 300 cases, however, the incidence of DCIS within IBC ranges from 0% to 17% with an average of 7% [
Minimally invasive interventions for breast cancer seek to redress the balance between benefit and risk and may, therefore, be of particular use in asymptomatic patients with low-risk lesions or patients deemed unfit for conventional management. Image-guided radiofrequency ablation therapy (RFA) has been demonstrated in pilot studies to be effective with few complications and a favourable safety profile. However, complete ablation may not achievable in all patients and exhaustive histological specimen analysis is not possible. Furthermore, current imaging modalities are relatively imprecise at delineating the extent of DCIS and predicting/confirming complete ablation [
DCIS should be managed within the multidisciplinary team and management tailored to patient and tumour factors. Local control depends upon adequate surgical clearance, and in order to reduce the risk of LR, surgical margins of at least 2 mm should be achieved. SLNB can be considered in patients with a high-risk of occult invasive disease. RT following BCS significantly reduces LR, particularly in those at high-risk. There remains a lack of level-1 evidence supporting the omission of adjuvant RT in selected low-risk cases. Large, multicentric, or recurrent lesions (particularly in cases of prior RT) should be treated by Mx and immediate reconstruction should be discussed. Adjuvant Tamoxifen may reduce the risk of LR in patients with hormone sensitive disease. Further research is required to determine the role of contemporary RT regimes and endocrine therapies. Biological profiling and molecular analysis represent an opportunity to improve our understanding of the tumour biology of this condition and rationalise its treatment. Reliable identification of low-risk lesions could allow treatment to be less radical or safely omitted.
Articles were identified by searches of Medline, PubMed, Embase, and Cochrane databases up to September 2011 using the terms: “DCIS” or “ductal carcinoma
The authors declared no conflict of interests.