Low-grade DCIS is characterized by the growth of small, fairly monotonous cells respecting each other’s cell borders. The nuclei are often hyperchromatic and uniform, with inconspicuous nucleoli. The cytoplasm is usually scant, with a slight increase in the nuclear–cytoplasmic ratio. Mitoses are infrequent, but when present, aid in the diagnosis.

Low-grade DCIS may exhibit a variety of architectural patterns of which the most frequent are micropapillary, cribriform, and solid. Micropapillary low-grade DCIS is composed of rounded tufts of neoplastic cells that project into the duct lumen without fibrovascular cores. The cells forming the micropapillae are typically evenly distributed and are uniform with dark nuclei and a monotonous appearance (Fig. 4.5). The cribriform pattern of low-grade DCIS is characterized by neoplastic cells forming round and rigid secondary lumens. These spaces are frequently calcified. While necrosis is rare in low-grade DCIS, it may be observed in some cases .

Micropapillary low-grade DCIS is frequently multicentric, and most of these patients accordingly have extensive microcalcifications on imaging. Not surprisingly, despite the low-grade features, micropapillary DCIS is nevertheless a risk factor for recurrence in patients treated with lumpectomy. The ducts with tumor are often admixed with benign or atypical ducts. Hence, diagnosis can be a challenge, especially in core biopsies.

Stringent criteria need to be used to diagnose micropapillary low-grade DCIS. These include: (a) micropapillae are club like and involve the entire duct circumference, extend at least one third of the duct diameter, and show a tendency to detach, (b) the nuclei are not overlapping, and (c) at least three adjacent ducts must be fully involved .

While the architectural patterns of intermediate grade DCIS are similar to low-grade lesions, intermediate-grade DCIS is characterized by slightly increased cellular pleomorphism, more frequent prominent nucleoli, and coarser chromatin than low-grade DCIS (Fig. 4.2) . Mitoses and necrosis are also more frequently encountered in intermediate-grade DCIS compared to low-grade lesions.

High-grade DCIS is defined by marked nuclear enlargement and pleomorphism . The nuclei have coarse chromatin and prominent nucleoli. There is loss of cell polarity and mitoses are frequent and often atypical. The comedo-type necrosis architectural pattern is very common, but solid, cribriform, micropapillary, and clinging patterns are also often seen. Central comedonecrosis is often accompanied by calcifications, which may be large and amorphous (Fig. 4.6). High-grade DCIS is also frequently associated with periductal chronic inflammation (Fig. 4.7) and angiogenesis, and a desmoplastic or sclerotic stromal response, with both concentric and distorting patterns (Fig. 4.8a–c).

High-grade DCIS may sometimes mimic microinvasive carcinoma , especially when DCIS extends into small glands of lobules that are distorted by marked sclerosis or secondarily involve sclerosing adenosis, radial scars, or complex sclerosing lesions. In such cases, deeper sections and immunohistochemical (IHC) stains that highlight myoepithelium, such as p63, calponin, muscle-specific or smooth muscle actin, among others, may be helpful in assessing the presence of microinvasion. The use of myoepithelial stains is particularly helpful if DCIS extends into sclerosing lesions such as sclerosing adenosis, which is rich in myoepithelial cells (Fig. 4.9a, b). However, staining for myoepithelial markers may not always aid in the differentiation between DCIS and microinvasive carcinoma, as staining may be patchy and difficult to interpret or the small focus may be mostly or completely lost in deeper immunostained slides. In such cases, the suspicion for a focus of microinvasion should be stated in the pathology report .

Consonant with the histological heterogeneity, DCIS exhibits significant variability in biomarker profiles and genetic aberrations . Low nuclear grade DCIS almost universally has diffuse, strong expression of estrogen receptor (ER) and progesterone receptor (PR) and lacks HER-2/neu protein overexpression or gene amplification. In contrast, high nuclear grade DCIS can be ER/PR positive or negative and is HER-2/neu positive in 60–80 % of cases (Fig. 4.10a–h) [33–36]. Unlike low-grade DCIS, high-grade DCIS also has high proliferative rates and is often p53 positive by IHC and mutational analyses [29, 37]. Molecular studies have shown that low-grade DCIS is characterized by chromosomal losses at 16q and 17p and gains at 1q, whereas high-grade DCIS shows more numerous and variable alterations, including losses at 11q, 14q, 8p, and 13q and gains at 17q, 8q, and 5p [38, 39] .

Morphologic, immunophenotypic, and molecular evidences support that DCIS represents a heterogeneous group of diseases that differ in biologic potential, ranging from little-to-no to very high risk of progression to invasive carcinoma. Due to its diversity and in attempt to avoid the anxiety-producing word “carcinoma” in its name, some support renaming DCIS as ductal intraepithelial neoplasia (DIN) as proposed by Tavassoli et al. [30], which is similar nomenclature to that used to describe dysplasias of the genital tract and perineum. Further study is necessary to determine whether DCIS should be further subclassified and/or redefined based on grade, hormone receptor status or other remarkable histological features .

Atypical ductal hyperplasia . The main differential diagnosis of low-grade DCIS is with atypical ductal hyperplasia (ADH). ADH is characterized by one or two moderately distended ducts, filled by cells that are evenly spaced, having uniform nuclei, and some regular secondary lumens (Fig. 4.11). Importantly, ADH lacks pleomorphism, individual cell necrosis, frequent mitoses, or prominence of nucleoli, which characterize DCIS. The presence of cells that overlap and stream, features more in keeping with benign usual ductal hyperplasia, mixed with atypical cells is a helpful feature to distinguish ADH from DCIS.

It is common to observe a spectrum of disease in the same breast that ranges from ADH to DCIS. Thus, borderline cases may be seen at either the low end or high ends of the spectrum. At times, the distinction between ADH and low-grade DCIS is one of the most challenging diagnostic difficulties in breast pathology. Most experts would assign these challenging cases to the less severe diagnostic category. Interestingly, most ADH is detected in the vicinity of—or within—a sclerosing lesion, a papilloma and—especially—within or close to a worrisome columnar altered lobule. ADH is often near the targeted calcifications in a biopsy.

Pleomorphic lobular carcinoma in situ . Pleomorphic lobular carcinoma in situ (pLCIS) has morphologic features of both classic LCIS and high-grade DCIS. Like classic LCIS, the cells are discohesive and have prominent plasmacytoid morphology. However, the cells of pLCIS are also enlarged and have marked nuclear pleomorphism and readily identifiable mitotic activity as seen in high-grade DCIS. pLCIS may also have comedo-type necrosis, a common architectural pattern of high-grade DCIS. pLCIS has been further divided into non-apocrine and apocrine types, the latter of which has apocrine features including prominent nucleoli and abundant eosinophilic cytoplasm (Fig. 4.12) .

pLCIS demonstrates loss of E-cadherin expression and chromosomal losses at 16q and 17p and gains of 1q, which are hallmark features of classic invasive and in situ lobular carcinomas [40, 41]. However, compared to classic LCIS which is invariably ER/PR positive and HER-2/neu negative, pLCIS has lower levels and is less frequently ER positive and may be HER-2/neu positive in approximately 25 % of cases [42]. pLCIS also has been shown to have additional genetic aberrations including deletion of 8p and 13q and gains of 8q, which were also identified in matched pleomorphic invasive lobular carcinoma (pILC) [40, 41, 43]. However, Chen et al. [44] found that apocrine pLCIS had greater numbers and diversity of genetic abnormalities than non-apocrine pLCIS, which had a similar profile to classic LCIS. Specifically, they identified amplification at 17p11.2–17q12 and 11.q.13.3, gain of 16p and losses of 3q, 11q, 13q, and 17p in the apocrine pLCIS group .

Classic LCIS is generally regarded as a marker of bilateral future cancer risk. In contrast, pLCIS has higher-grade morphology, an often unfavorable biomarker profile and greater extent of genetic aberrations, and thus is thought to be a non-obligate precursor of pILC [45]. Therefore, pLCIS is treated similar to DCIS, with excision to histologically negative margins and eradication of all suspicious mammographic calcifications. However, use of adjuvant radiation for pLCIS is controversial with many advocating it but with little data to support its use.

pLCIS is a morphologically and genetically distinct entity but may be misinterpreted as high-grade DCIS, especially if it is the apocrine subtype. Current practice is, treatment in a similar manner to DCIS, thus misclassification may have little detrimental effect to the patient. However, correct classification may influence decision for adjuvant radiation therapy. Furthermore, correct classification will aid in identifying cases for future study to better understand how pLCIS differs from classic LCIS and high-grade DCIS .

Microinvasive carcinoma often occurs in the setting of abundant DCIS. Therefore, it is extremely important to thoroughly sample excision specimens for microscopic evaluation if there is a prior biopsy diagnosis of DCIS only, especially, if there is radiologic or gross evidence of multifocal or mass-forming disease. Microinvasive carcinoma more often arises in the setting of high-grade DCIS but may also be seen in association with low-grade DCIS (Fig. 4.13a–b). Coexisting microinvasive carcinoma and DCIS components very frequently have similar morphologic features and ER, PR, and HER-2/neu staining patterns.