Ductal carcinoma in situ (DCIS) was first described in the late 19th century, but was not recognized as a “premalignant” lesion until the 1930s. It is defined as clonal proliferation of malignant mammary ductal epithelial cells within the duct lumen that do not invade through the basement membrane. The biological features of DCIS are fairly heterogeneous. Consequently, classification and prediction of the natural history of these lesions can be difficult. DCIS is classified according to architectural pattern, tumor grade, and the presence or absence of comedo necrosis. The incidence of DCIS in the United States has increased with the increased utilization of screening mammograms, from 1.87 per 100,000 in 1973 to 1975 to 32.5 in 2004. DCIS rates significantly decreased in women age 50 to 69 years after 2002. Although hormone replacement therapy (HRT) has not been proven to be a risk factor for DCIS, this decrease is in parallel to the decline in use of HRT. Known risk factors for developing DCIS include older age and family history of breast cancer.

Recent institutional studies have reported that DCIS accounts for 14% to 44% of all new mammographically detected neoplasms. Prior to mammographic screening, DCIS was most commonly diagnosed as a palpable mass, bloody nipple discharge, or as Paget’s disease of the nipple. The most common mode of detection today is screening mammography, often at baseline screening. DCIS is identified as clustered microcalcifications in 75% to 90% of cases and as a density in 22% to 27% of cases. It is also frequently diagnosed in the background of a pathologic specimen of invasive carcinoma. It is not uncommon for DCIS to be multifocal (a second focus of disease occurring near a lesion within the same quadrant) or multicentric (disease occurring in multiple quadrants of the same breast) on mammography. Many of these cases have contiguous intraductal carcinoma that can only be identified by careful histologic evaluation. In addition, up to 20% of patients with a diagnosis of DCIS on core biopsy are found to have invasive carcinoma upon resection. Factors that correlate with increased risk of a synchronous invasive lesion are lesion size on imaging and the grade of DCIS on core biopsy.

It has been proposed that some in situ lesions are indolent and never progress to invasive cancer. Based on its clinical, molecular, and histological features, DCIS is considered by most to be a direct precursor of invasive breast cancer. Support for this theory includes similar risk factors for DCIS and invasive carcinoma and the frequent presence of DCIS adjacent to invasive disease. Untreated DCIS is reported to progress to invasive carcinoma in 14% to 75% of cases. In addition, there is up to 13% risk of invasive local recurrence (LR) following breast-conserving therapy for DCIS. Based upon these factors, aggressive local therapy has been advocated for DCIS. This treatment strategy results in excellent long-term survival rates of 96% to 98% at 10 years. Due to the variability in malignant potential of this disease, the management of DCIS continues to be debated. Future research may focus on developing risk stratification models to identify DCIS patients that can be treated less aggressively without sacrificing the excellent survival outcome. Some prognostic markers for invasive breast cancer have been studied in DCIS. Although estrogen receptor and Her-2 testing in DCIS is not considered to be standard, small studies have shown that estrogen receptor positivity correlates with decreased risk, and Her-2 overexpression is associated with increased risk of recurrence.

The treatment objective for DCIS is prevention of LR of in situ and invasive carcinoma. Treatment options for DCIS are total mastectomy or breast-conserving surgery with partial mastectomy and RT. Sentinel lymph node biopsy (SLNB) is considered in selected patients, particularly those with large or high-grade lesions and/or those undergoing total mastectomy. Prior to the advent of breast-conserving surgery for invasive carcinoma, mastectomy was the treatment of choice. Total mastectomy is highly effective in controlling DCIS, with a long-term LR rate of only 1% to 2%. The rationale for performing total mastectomy for DCIS while less-extensive procedures were being performed for invasive cancer was questioned. This led to the application of breast-conserving surgery for patients with DCIS. The decision to proceed with breast-conserving therapy or mastectomy should be based on the extent of disease, the risk of LR after partial mastectomy, the patient’s medical history, and personal preference. The maximal size of DCIS that can be treated with breast conservation varies with the location of the lesion and the patient’s breast
volume. It is generally difficult to excise lesions greater than 5 cm in size with tumor-free margins and an acceptable aesthetic result. LR rates are high for lesions of this size, so mastectomy is the preferred option for such lesions. Mastectomy with or without breast reconstruction is also the optimal approach in patients with multicentric (but not multifocal) disease and women with contraindications to RT, such as collagen vascular disease (scleroderma and systemic lupus), prior breast RT, or pregnancy. For women who do not have one of these contraindications, breast-conserving surgery is usually the treatment of choice. Exceptions may be those at increased risk of LR after partial mastectomy, including younger women (age <50) with larger (3 to 5 cm), high-grade lesions, close margins (less than 1 to 2 mm) that cannot be reexcised, micropapillary histology, and those with a family history that places them at high risk of developing a second primary breast malignancy. These factors should be considered relative contraindications to breast conservation. A balanced discussion of risks and benefits of each approach should be undertaken to determine the best treatment option for each patient. The technical aspects of partial and total mastectomies are described in prior chapters and do not vary significantly when the procedure is performed for DCIS rather than invasive carcinoma.

Local excision by partial mastectomy or “lumpectomy” is generally followed by adjuvant RT to decrease LR. The routine use of RT is largely based on the results of NSABP B-17 (National Surgical Adjuvant Breast Project protocol B-17). This prospective randomized trial compared total mastectomy with lumpectomy alone and lumpectomy followed by RT in women with DCIS. After 7.5 years of follow-up, the LR rate of invasive cancer was 13.4% in patients who underwent surgery alone compared with 3.9% in those who received RT. The recurrence rates for in situ disease were 13.4% versus 8.2%, respectively. However, overall survival was equivalent for the mastectomy and local excision with or without RT cohorts. Similar results were seen in the EORTC-10853 Trial (European Organization for Research and Treatment of Cancer Randomized Phase III Trial 10853), which reported a LR rate of 16% after local excision versus 9% after local excision and RT with 4 years of follow-up. A 10-year follow-up report analyzed the effect of RT on the risk of LR overall in subgroups of patients with different clinical, histologic, and treatment factors. The risk of in situ and invasive recurrence was reduced by 48% and 42% respectively. Factors associated with increased LR were young age, symptomatic detection, intermediate or high-grade DCIS, cribriform or solid growth pattern, close or unspecified margins, and treatment by local excision alone. However, the beneficial effect of RT was seen in patients with all the risk factors that were assessed.

Although RT reduces the risk of local failure in patients undergoing partial mastectomy for DCIS, there may be a subset of women with low-risk DCIS for whom excision alone is sufficient. This hypothesis is supported by retrospective studies reporting that women with “low-risk” DCIS have a low probability of LR after margin-negative excision. The criteria used to define risk varied among studies, but include patient age, tumor size, grade, and margin width. A prospective trial of local excision alone for low-risk DCIS of the Eastern Cooperative Oncology Group (ECOG) E5194 has been reported. With a median follow-up of 6 years, rigorously selected patients with low- to intermediate-grade DCIS with margins 3 mm or wider had an acceptably low rate (6.1%) of LR 5 years after excision without irradiation. However, patients with high-grade lesions had a much higher LR rate (15.3%).

Controversy exists regarding the degree to which the width of the resection margin affects outcomes and its relationship to the benefit of RT. In large, multiinstitutional studies, it has been difficult to identify a subgroup of patients with widely clear margins that have such a low LR rate that they may be managed with local excision without RT. This led to the hypothesis that conventional pathology specimen analysis is unable to determine completeness of excision of DCIS. Silverstein and colleagues have shown that excellent long-term local control can be achieved without RT if complete excision with widely tumor-free margins is confirmed by meticulous sequential and complete embedding of the entire specimen, with three-dimensional reconstruction. A recent update of their data provides recommendations for management based on the USC/Van Nuys Prognostic Index. This index takes into consideration tumor size, margin width, grade/necrosis, and age. Treatment recommendations are based upon what is required to achieve a LR rate of less than 20% at 12 years, with recommendations ranging from excision alone, excision with radiation, and mastectomy.

RT remains the standard treatment for most women with DCIS who undergo breast-conserving surgery. In selected patients with low-risk disease (e.g., women over 70 years of age with low- to intermediate-grade lesions less than 15 mm in size and tumor-free margins), excision alone may be considered. Balanced discussion of the risks and benefits of RT and close clinical and mammographic follow-up are essential if this approach is chosen.

Accelerated partial breast irradiation (APBI) has been studied as an alternative to conventional whole-breast RT for patients with DCIS. Results from the American Society of Breast Surgeons MammoSite Breast Brachytherapy Registry Trial have been reported with a median follow-up was 53.7 months. In patients with low to intermediate grade, pathologic size >0.3 but <2.5 cm and margins ≥3 mm, the 5-year ipsilateral breast tumor recurrence (IBTR) was 0%, compared with 6.1% at 5 years in the local excision-only group with similar eligibility criteria in the E5194 study. For patients with high-grade lesions, pathological size <1 cm and margins ≥3 mm, 5-year IBTR was 5.3%, compared with 15.3% in E5194. Although these results are early and are not from a randomized study, they do suggest that partial breast irradiation may be an acceptable option for at least a subset of patients. Confirmation by a randomized, controlled clinical trial would be optimal. Other approaches to partial breast irradiation including conformal radiotherapy and interstitial brachytherapy have been proposed, but are not well studied. Intraoperative RT is also being evaluated for DCIS and early-stage invasive breast cancer. The initial studies suggest that it is a feasible and well-tolerated alternative to postsurgical APBI, but larger studies with longer follow-up are needed to confirm the efficacy of this treatment.

Because of the limitations in pathological margin assessment for DCIS, more extensive preoperative imaging to determine the extent of disease has been proposed. The goals of enhanced imaging are to identify patients with extensive disease that is not amenable to breast-conserving surgery, to define the anatomic extent of disease more accurately, and to facilitate complete excision at the initial operative procedure. Breast magnetic resonance imaging (MRI) has been proposed for the presurgical evaluation of patients with DCIS. Studies suggest that the sensitivity for detecting multicentric disease is higher with MRI compared with mammography, although the specificity is low and the incidence of false-positive findings requiring additional workup is high. MRI may provide more accurate assessment of extent and size of disease. It is not clear if this information can be translated into more complete excisions, fewer reexcisions, and decreased LR rates. Furthermore, it is not clear that all of the occult disease identified by breast MRI is biologically significant. This has raised concerns
that MRI may lead to overdiagnosis and overtreatment through increased use of mastectomy in a disease that is almost uniformly curable with current therapy. Based upon these factors, breast MRI is considered to be a useful adjunct in selected cases, but is not a standard part of the evaluation of all patients with DCIS in our practice. Large-scale, prospective, randomized trials would be helpful to evaluate this issue, although they may be difficult to complete due to the time and expense required for such a study.

Axillary staging is another area of controversy. Based on the definition of in situ carcinoma as a noninvasive disease, it has not routinely been a part of the management of DCIS. Sentinel node biopsy may be useful for the staging of patients with “high-risk” DCIS, in which the suspicion of occult invasive carcinoma is high, or those with microinvasive disease. Axillary nodal metastases are found in less than 1% of cases of pure DCIS. However, regional lymph node metastases have been reported in 6% to 12% of patients with high-risk DCIS or microinvasion. The definition of high risk varies among studies, but includes the presence of a palpable mass, multicentric disease, large, high-grade lesions, or the presence of necrosis. These features are associated with more extensive, biologically aggressive disease that is more likely to harbor areas of occult invasive carcinoma. Axillary staging may be required if invasive disease is identified in the excision specimen. It has been established that SLNB is feasible after excision, but not after mastectomy. Therefore, in patients that will be treated with mastectomy, consideration should be given to SLNB. Models have been developed for predicting the risk of invasive cancer in women diagnosed with DCIS on core needle biopsy. A model based on the diameter of microcalcifications on imaging and grade can identify women at high risk of having invasive cancer. With an imaging lesion >50 mm and a high-grade core biopsy, the model’s positive predictive value is over 50%. These models may be used as a guide in the decision-making process of whether to proceed with sentinel node biopsy. These factors should be balanced with the well-established short- and long-term morbidity of the procedure. The technical aspects of SLNB for breast cancer are described in Chapter 46 and do not vary when the procedure is performed for DCIS rather than invasive carcinoma.

The use of adjuvant endocrine therapy for DCIS is also debated. The recommendation for postoperative tamoxifen following breast-conserving treatment for DCIS originates from the NSABP B-24 trial. In this study, tamoxifen reduced the incidence of invasive breast cancer in the ipsilateral breast from 4.2% to 2.1%, and contralateral invasive or in situ carcinoma from 0.8% to 0.4% per year. Tamoxifen did not affect overall survival, and it was associated with many adverse events, including menopausal symptoms, endometrial cancer, thromboembolic disease, and uterine cancer. These factors have discouraged the routine use of this agent in patients with DCIS. Premenopausal women have a higher rate of recurrence after breast-conservation therapy and a lower risk of adverse events with tamoxifen use. In this patient group, the benefit of tamoxifen may outweigh its risks. Long-term results from the UK/ANZ DCIS study of tamoxifen and radiotherapy in women with locally excised DCIS have been recently published. With a median follow-up of 12.7 years, this report confirms a benefit for tamoxifen in reducing LR of DCIS and contralateral breast malignancies for women treated with local excision defined as no DCIS at the margin. There was no reduction in ipsilateral invasive carcinoma for this group of patients, who were older than the patients in the NSABP B-24 study. As opposed to the results of the NSABP B-24 study, the benefit of tamoxifen in the UK/ANZ DCIS study was only apparent in patients that did not receive RT. Based upon these data, we currently use adjuvant tamoxifen only selectively in younger women with DCIS who are considered to be at higher risk of recurrence.

Atypical ductal hyperplasia (ADH) is defined as a lesion that has some, but not all, of the histological features of DCIS, a lesion that has all features of DCIS but is less than 2 mm in greatest dimension, or a lesion with all features of DCIS but involves less than two duct spaces. ADH is found in 31% of biopsies for mammographic calcifications and in 4% of all benign breast biopsies. Women with ADH have a relative risk of developing invasive breast cancer that is four to five times that of an age-matched woman of average risk. This risk is almost doubled for women with a family history of breast cancer in a first-degree relative.

In many centers, core needle biopsy is the preferred biopsy technique for suspicious lesions noted on mammogram. A pathologic diagnosis of ADH on core needle biopsy may underrepresent the lesion because of sampling error. Up to 20% to 50% of patients will have in situ or invasive carcinoma on subsequent excisional biopsy. A diagnosis of ADH by needle biopsy mandates excisional biopsy to ensure that neither DCIS nor invasive carcinoma is present. Efforts are under way to identify low-risk subgroups for which reexcision may not be necessary. The factors being studied include the size of biopsy sample, extent of ADH within the sample, histologic features of the DCIS, and the presence of a residual mammographic lesion after image-guided core biopsy. Because of the fourfold to eightfold increased risk of breast cancer, a diagnosis of ADH warrants close surveillance and consideration of tamoxifen or raloxifene therapy in appropriately selected women.

Lobular carcinoma in situ (LCIS) is a proliferation of the terminal ductal–lobular epithelium without invasion of the basement membrane. The acini are distended and distorted by neoplastic cells with characteristic lobular features. There are no clinical or mammographic findings routinely associated with LCIS, and it is frequently an incidental finding in breast biopsies performed for other benign or malignant breast pathology. The incidence is reported to be 0.5% to 3.6%, but this figure is likely underestimated because of the absence of clinical or mammographic findings. It is detected in 0.5% to 8.0% of all breast biopsies. The lesion is most common in premenopausal women, and it is bilateral in 50% to 90% of cases. In up to 6% of cases a synchronous invasive cancer is found, the majority of which are of ductal histology. Women with a diagnosis of LCIS have a relative risk of developing an invasive carcinoma that is 8 to 10 times that of age-matched women of average risk.

The observation that all quadrants of both breasts are nearly equally at risk suggested that LCIS is not a true premalignant lesion, but simply a marker for increased risk of breast cancer. However, newer data indicate that the majority of invasive carcinomas in women with LCIS occur in the ipsilateral breast. Analysis of data from the Surveillance, Epidemiology and End Results (SEER) demonstrated that LCIS patients were 5.3-fold more likely than DCIS patients to develop invasive lobular carcinomas. These data suggest that LCIS may also be a direct precursor to invasive carcinoma in some cases. Given the propensity for LCIS to be multicentric, bilateral, and mammographically occult, it is very difficult to determine its exact role as a precursor lesion. Recent studies demonstrating that LCIS and invasive cancers share many genetic and molecular features also support the precursor hypothesis. Until more evidence for this hypothesis is provided, surgical management of LCIS continues to be biopsy and close surveillance rather than partial or total mastectomy.

Another area of controversy is the management of LCIS diagnosed at core needle biopsy. Studies of patients with LCIS on core biopsy who also underwent excision show a rate of upgrade to invasive carcinoma or DCIS of 25% to 38%. A higher rate has been observed when there is an associated mass, histologic features of pleomorphic LCIS (which may have biological behavior similar to that of DCIS), or extensive LCIS. Based on these data, most authors recommend surgical excision for patients diagnosed with LCIS on core biopsy. However, reexcision of LCIS diagnosed by excisional biopsy to obtain lesion-free margins has not been routinely advocated.

Tamoxifen has been shown to reduce the risk of developing invasive carcinoma in women with LCIS. Because the magnitude of risk is relatively high in this group, tamoxifen therapy should be considered in women with LCIS. This is particularly true if they are young and at relatively low risk of adverse effects from tamoxifen. Raloxifene, a drug with lower risk for thromboembolic events and uterine cancer, has been evaluated as an option for risk reduction in postmenopausal women with LCIS. The STAR trial confirmed that raloxifene is 76% and 78% as effective as tamoxifen in preventing invasive and noninvasive breast cancer, respectively. This study also suggests that raloxifene may only remain therapeutically effective as long as it is taken, in contrast to tamoxifen, whose risk-reduction benefits persist for over 5 years after discontinuing therapy.

For those with other risk factors for the development of an invasive carcinoma (e.g., a strong family history) or severe anxiety secondary to their risk, bilateral total mastectomy may be a reasonable option. This is considered only in highly selected patients who have undergone genetic counseling and carefully considered the advantages and disadvantages of all other risk-reduction strategies. For most women with LCIS, close follow-up with or without endocrine therapy is the preferred approach.

As with DCIS and ADH, atypical lobular hyperplasia (ALH) is on a histological continuum with LCIS. ALH has pathologic features that are similar to but less developed than LCIS, with only half the acini of the lobular unit involved. The criteria for diagnosis vary among pathologists, and the difficulty in accurately distinguishing LCIS from ALH has led some to classify both entities under the term lobular neoplasia. Like LCIS, ALH is an incidental finding on breast biopsy in most cases, and its clinical significance is as a marker for increased risk of developing invasive carcinoma. Identification of ALH is associated with a 10% to 20% risk of developing an invasive carcinoma during the next 10 years and a twofold to fourfold increase in relative risk compared with age-matched controls. This risk appears to be age-dependent, with minimal (if any) significance in women more than 70 years of age. Invasive carcinomas are three times more likely to arise in the ipsilateral breast as the contralateral breast. In addition, there appears to be a higher incidence of developing invasive lobular carcinoma than invasive ductal carcinoma in the ipsilateral breast. Over the past few years, several studies have documented an increased risk of occult in situ or invasive carcinoma in women diagnosed with ALH on core biopsy. The frequency of this has varied between 0% and 50%, but is clearly higher than previously reported. The largest study confirmed a 13% to 20% frequency of “upstaging” of ALH or LCIS on core biopsy to in situ or invasive cancer after surgical excision. Based upon those studies we recommend excisional biopsy for all cases of ALH diagnosed by core biopsy.