Breast Cancer Subtypes: Morphologic and Biologic Characterization

Abstract

Advances in basic science, technology and translational research have created a revolution in breast cancer diagnosis and therapy. Researchers' discoveries of genes defining variability in response to therapy and heterogeneity in clinical presentations and tumor biology are the foundation of the path to personalized medicine. The success of personalized breast cancer care depends on access to pertinent clinical information and risk factors, optimal imaging findings, well-established morphologic features, and traditional and contemporary prognostic/predictive testing. The integration of these entities provides an opportunity to identify patients who can benefit from specific therapies, and demonstrates the link between breast cancer subtypes and their association with different tumor biology. It is critical to recognize specific types of breast cancer in individual patients and design optimal personalized therapy. This article will highlight the roles of morphologic features and established tumor biomarkers on patient outcome.

The recent decline in mortality from breast cancer in resource-rich countries is attributed to increased public awareness, advances in breast imaging and screening, and to the new innovations in breast cancer therapy. The emerging discoveries about the biology of this disease and the introduction of molecular targeted therapy are exciting, and could potentially further reduce mortality from breast cancer. The challenge, however, is the heterogeneity of breast cancer in presentation, clinical behavior and response to therapy. It is clear that there is a critical need for the delivery of personalized medicine for breast cancer patients. Planning individualized therapy for each breast cancer patient requires access to effective tools for appropriate stratification of patients. It is important to identify those who may require aggressive therapy, versus those who may not need, or may not respond to similar therapy.

Traditional clinical and pathologic factors such as age, histologic grade, tumor type, tumor size and hormone receptors have commonly been used to assign patients into risk groups to receive adjuvant hormonal, radiation therapy and/or chemotherapy. These factors accurately stratify the patients based on the long term follow-up studies [1]. However, it is recognized that traditional prognostic factors are limited in their ability to provide reliable stratification in all patients. It has been shown that up to 30% of women with node-negative breast cancer die of the disease regardless of adjuvant therapy, and 70% survive without adjuvant therapy [2]. As the heterogeneity in breast cancer cannot be captured by the traditional prognostic factors, it is therefore essential to search for factors that may supplement traditional prognostic factors in segregating patients who need adjuvant therapy, and in predicting clinical response to the available therapeutic modalities [3]. There has also been a need to develop additional forms of systemic therapy for those tumors that fail to express known targets such as estrogen receptor (ER) and progesterone hormone receptor (PR), and HER-2/neu oncogenes.

Breast cancers are diverse in their natural history and responsiveness to treatment. Differences in genetic makeup account for much of the biological diversity of breast tumors. In each cell, signal transduction and regulatory systems transduce information from the cell's identity to its environmental status that controls the level of expression of the selected gene in the genome. It is proposed that phenotypic diversity of breast tumors might be accompanied by a corresponding diversity in gene expression patterns that can be captured using cDNA microarrays [4]. It is assumed that the expression of various genes in different tumors provides an opportunity to classify tumors at a genomic level into subclasses of potential prognostic significance [5].

During the last several years, the advent of oligonucleotide and cDNA microarray and research on gene expression profiling have been used to drive molecular characterization of breast tumors and help predict the probability of local recurrence and survival. Compared with traditional prognostic factors, gene expression profiling is now regarded as a more powerful independent predictor in breast oncology [6,7]. However, until these new discoveries about the genomic characterization of breast cancer subtypes become universally accepted as standard trends in clinical practice, the value of established breast cancer tumor markers such as estrogen and progesterone receptors and Her-2/neu oncogene in relation to different breast cancer subtypes has to be fully recognized.

This article is designed to highlight a summary of the events leading to molecular subtyping of breast cancer and discuss the morphology and pattern of expression of biomarkers and their impact on the current state of follow-up and management of breast cancer patients. The detailed description of emerging genetic and genomic pathways are beyond the scope of this article and will be presented in future contributions.

Molecular subtypes of breast cancer

In 2000, Perou et al. [4] published a seminal article on gene expression profile-based classification of breast tumors. They identified five subgroups based on 496 genes that differentiate breast cancers into separate groups based on gene expression patterns (Figure 1). These subtypes differ markedly in prognosis and in the repertoire of therapeutic targets they express [8,9] (Figure 2). There are two types of ER positive tumors which include luminal A and luminal B [8]. The luminal subtype A and B tumors express ER, GATA3 and genes regulated by both ER and GATA3 [9]. The intrinsic subtypes include two main subtypes of ER negative tumors that include basal-like and HER-2/neu oncogene receptor positive tumors. Basal-like tumors typically show low or no expression of HER-2/neu oncogene and ER and PR, and also exhibit high expression of gene characteristics of the basal epithelial cell layer which includes expression of cytokeratin 5, 6, 17 and integrin-B4 [10]. A small percentage of tumors is regarded as unclassified/normal breast-like (Box 1) [4,11]. Please note that in the years since the publication of the article by Perou et al., which is referenced in Box 1, the ‘normal breast-like’ category has been removed from molecular subtypes of breast cancer.

Figure 1.

Hierarchical clustering of 115 tumor tissues and seven nonmalignant tissues by using the intrinsic gene set. (A) A scaled-down representation of the entire cluster of 540 genes and 122 tissue samples based on similarities in gene expression. (B) Experimental dendrogram showing the clustering of the tumors into five subgroups. Branches corresponding to tumors with low correlation to any subtype are show in gray. (C) Gene cluster showing the ERBB2 oncogene and other coexpressed genes. (D) Gene cluster associated with luminal subtype B. (E) Gene cluster associated with the basal subtype. (F) A gene cluster relevant for the normal breast-like group. (G) Cluster of genes including the ER (ESR1) highly expressed in luminal subtype A tumors. Scale bar represents – fold change for any given gene relative to the median level of expression across all samples.

Reproduced with permission from [12] © Springer (2007).

Figure 2.

Kaplan–Meier analysis of disease outcome in two patient cohorts. (A) Overall survival for 72 patients with locally advanced breast cancer in the Norway cohort. The normal-like tumor subgroups were omitted from both data sets in this analysis. (B) Time to development of distant metastasis in the 97 sporadic cases.

Reproduced with permission from [12] © Springer (2007).

Box 1.

Molecular classification of breast tumors.

Breast cancer is a family of diseases

• ER+ (luminal A)

• ER+ (luminal B)

• Her-2/neu+

• Basal-like/triple negative

• Unclassified/normal breast-like

Luminal A & B

Luminal A/B generally carry a good prognosis, and show a favorable response to endocrine therapy. Luminal A has better prognosis than luminal B. Luminal B has a moderate expression of gene expressed by the breast luminal cells, higher proliferation rates and lower PR (Figures 3 & 4) [4].

Figure 3.

Photomicrograph of a luminal type breast carcinoma (H&E stain × 200).

Figure 4.

Another view of the same case as in Figure 5 displaying expression of estrogen receptor protein as evidenced by brown nuclear staining (immunostain × 200).

Her-2/neu positive type

This type presents as the distinct form associated with either ER negative or ER positive breast tumors. It is frequently seen with associated ductal carcinoma in situ and carries a poor prognosis (Figures 5 & 6) [4].

Figure 5.

Her-2/neu oncogene positive high grade breast carcinoma (H&E stain × 400).

Figure 6.

The same tumor demonstrating prominent gene amplification by FISH in a Her-2/neu oncogene positive breast cancer subtype (FISH × 400).

Basal-like breast cancers

Basal-like breast cancers are the most extensively studied group. These groups of breast tumors have distinct genetic expression patterns and immunocytochemical characteristics. Although there is no international consensus on the precise complement of markers that define basal-like breast cancer, most authors include a lack of ER, PR and HER-2/neu oncogene, known as triple-negative breast cancer. This type of tumor expresses high molecular weight cytokeratins, such as CK5 or CK5/6, CK14, CK17, and EGFR, CKIT, P63, P-Cadherin, SMA, as the profile characteristic of this tumor. Overall, ductal carcinomas that are ER-, HER-2/negative and CK5/6 and EGFR+ are considered basal-like breast carcinoma [10,13–15]. Triple-negative breast tumors represent the majority of cancers within the basal-like subtype. Not all triple-negative breast cancers display the basal-like phenotype and vice versa. Currently, no specific targeted approach is available for triple-negative tumors. However, several clinical trials have been designed to find the most suitable molecular targeted therapy for triple-negative breast cancers (Table 1) [16].

Table 1.

Possible molecular targeted therapies for the treatment of TN breast cancer.

Molecular targets	Agents tested in clinical phase trials
EGFR	Anti-EGFR antibody: cetuximab EGFR tyrosine kinase inhibitor: erlotinib
c-kit	Multiple tyrosine kinase inhibitors: imatinib, sunitinib
Src	Multiple tyrosine kinase inhibitors: dasatinib
mTOR	mTOR inhibitor: everolimus

Reproduced with permission from [17].

Morphologic features of basal-like breast carcinomas

Basal-like breast carcinomas are characterized by high mitotic rate and proliferative activity (Figure 7). They often present as a large tumor with distinct, geographic, central cellular zones composed of hyaline material, necrotic/ischemic tissue and collagen (Figure 8). By immunostaining, they display myoepithelial cell differentiation [14]. These tumors often have pushing, noninfiltrative borders, and may display some degree of lymphoplasmacytic infiltrate with medullary-like features (Figures 9 & 10). Most basal-like breast carcinomas are characterized by sheets of cells with minimal tubule formation, however, a subset of them with features of ‘adenoid cystic’ change and ‘ribbon like’ architecture have also been described. There is also a small subset (8–10%) of ductal carcinoma in situ cases that have morphologic features of high nuclear grade, lymphocystic infiltrate and necrosis with immunophenotypic characteristic of basal-like breast carcinoma. Currently, it is not clear whether these cases are a representation of true precursor lesions, or if they share lines of differentiation with basal-like breast carcinoma [14].

Figure 7.

Photomicrograph of a basal-like breast carcinoma demonstrating numerous mitosis (H&E stain × 400).

Figure 8.

Basal-like breast carcinoma displaying large control necrotic area (H&E stain × 400).

Figure 9.

Pushing border seen in basal-like breast carcinoma (H&E stain × 400).

Figure 10.

Presence of lymphocytic instillation in a basal-like breast carcinoma (H&E × 400).

Biology of basal-like breast carcinoma

Genetic expression profiling of breast carcinomas is designed to identify prognostically relevant subgroups of tumors that may benefit from individualized management. Among the current molecular subtypes of breast cancers identified, tumors that are classified as basal-type by genetic expression profiling have distinct biologic behavior. Independent studies have demonstrated that basal-like breast carcinomas have a worse overall, and/or disease specific survival as compared with the other molecular subtypes. In addition, basal-like breast cancers are more commonly seen in young black women and those with germline BRCA1 mutation carriers [15]. They often metastasize to the lung and brain [18], and display different patterns of response to chemotherapy. Basal-like breast carcinomas are biologically more aggressive, with the majority of deaths occurring in the first 5 years after diagnosis [16].

Triple negative breast cancer

It is generally recognized that currently, not all basal-like breast cancers display a triple negative phenotype and not all basal-like breast cancers are stratified as basal-like tumors by gene expression profiling [4,8,19]. It appears that by immunostaining, only 71% of triple negative cancers are basal-like subtypes by gene expression profiling and 77% of basal-like subtypes by gene expression profiling are triple-negative type [20].

Triple-negative breast cancers are considered a heterogeneous group of tumors, the majority of which present as basal-like breast cancers. Aside from ‘medullary’ like morphologic features, some triple-negative tumors present with other histologic subtypes such as adenoid cystic carcinoma, metaplastic carcinoma and apocrine carcinoma. Overall, triple-negative breast cancers express a high level of Ki67 as a reflection of the high proliferation rate and achieve pathologic complete response to neoadjuvant chemotherapy.

Molecular characterization of special subtypes of breast cancer

As a heterogenous group of tumors with different clinical presentation, morphologic features and behaviors, there are several different histologic breast cancer types that are described by the WHO [21]. About 50–80% of breast cancers are called invasive ductal carcinoma, not otherwise specified. This is simply because these tumors do not show any special morphologic features that can be further classified into special types. On the other hand, there are 25% of invasive breast cancers that are recognized as ‘special type’. Special types of breast cancer include invasive lobular carcinoma, adenoid cystic carcinoma, apocrine carcinoma, infiltrating ductal carcinoma with osteoclastic giant cells, medullary carcinoma, metaplastic carcinoma, micropapillary carcinoma, mucinous carcinoma, neuroendocrine carcinoma, invasive cribriform carcinoma, tubular carcinoma, secretory carcinoma, lipid-rich carcinoma and glycogen-rich clear cell carcinoma.

Infiltrating lobular carcinoma

The current morphologic stratification of breast cancer remains to be subjective, and may not adequately reflect the biologic capacity of different breast cancer subtypes. For example, poorly differentiated infiltrating breast carcinoma can only occasionally be accurately diagnosed as lobular carcinoma when there is no evidence of expression of E-cadherin (Figure 11) [22].

Figure 11.

Negative expression of E-Cadherin immunostaining of an infiltrating lobular carcinoma with extension into ductal system that shows contrasting positive immunostaining for E-cadherin.

Infiltrating lobular carcinoma (ILC) is defined as an invasive tumor that is often associated with lobular carcinoma in situ (LCIS). The loss of adhesion molecules explains the pattern of these tumor cells that have a limited ability to form a palpable mass [12]. Clinically, they present as an ill-defined lesion and have no distinct imaging findings [23–25].

There are four types of lobular carcinoma: classical, pleomorphic, alveolar and solid, each with their own unique set of mutations and genomic differences. The classical form of infiltrating lobular carcinoma is characterized by strands of single epithelial cells with insignificant atypia, and are typically lower grade and ER rich (Figure 12). Occasionally, ILC may be negative for progesterone receptors. Morphologically, pleomorphic lobular carcinomas show more nuclear abnormality and mitosis, and are more likely to have p53 and Her-2/neu oncogene mutations (Figure 13).

Figure 12.

Photomicrograph of an infiltrating lobular carcinoma.

Figure 13.

Photomicrograph of an infiltrating pleomorphic lobular breast carcinoma.

Several studies have demonstrated that ILCs are associated with older age, are larger and well-differentiated, are receptor positive, have a lower prevalence of lymphovascular invasion and are more commonly treated with mastectomy, when compared with infiltrating ductal carcinoma. Overall, ILCs show similar outcomes and response to therapy with invasive ductal carcinoma [26]. However, they have a different pattern of metastasis. They more frequently metastasize to bone, lung, visceral organs and areas such as meninges, pleura, peritoneum, stomach and ovaries. There is also more evidence of mulifocality and bilaterality associated with ILCs [27].

Naturally, because of luminal molecular patterns, endocrine therapy is mainstay for treatment of patients with ILCs and because of low proliferation rate (except for the pleomorphic variant) there is limited value in chemotherapy [28].

Adenoid cystic carcinoma

Similar to other adenoid cystic carcinoma (ADCC) in other sites such as the salivary gland, this tumor shows definitive evidence of myoepithelial call differentiation. It only accounts for 0.1% of breast carcinoma and has common ectodermal ‘sweat gland’ origins with the salivary gland [29]. ADCC occurs in adult women of the same age as other breast cancers and often presents as a discrete firm mass and rarely presents as an image-detected abnormality [30].

Grossly, ADCC are circumscribed and occasionally present as a cystic lesion. Morphologically, growth patterns include cribriform, solid, glandular, reticular and basalid in appearance with prominent epithelial and myoepithelial differentiation (Figure 14). Tumor cells are positive for S-100, P63, SMM-HC, CD117 and actins. ADCCs do not express ER and PR, and are negative for HER2/neu oncogene, so they are regarded as triple negative tumors.

Figure 14.

Photomicrograph of an adenoid cystic carcinoma of the breast.

Adenoid cystic carcinomas typically have excellent outcomes [31] and mastectomy has been curative in the vast majority of patients. Rarely, an aggressive form of ADCC may present with lymph node metastasis, and tumor cells show high proliferation rates and positive expression for P53 protein [29,32].

Apocrine carcinoma

Apocrine carcinoma constitutes 0.3–4% of breast carcinomas with similar clinical presentation and outcomes to other infiltrating ductal carcinomas. The reported age group is wide between 19 to 86 years, but is more commonly seen in postmenopausal women [18]. There is no distinct imaging or gross appearance of apocrine carcinoma and morphologically pure forms of apocrine carcinoma show apocrine differentiation in the entire tumor. Microscopically, apocrine carcinomas have abundant grandular eosinophilic cytoplasm and prominent nucleoli (Figure 15).

Figure 15.

Photomicrograph of an apocrine carcinoma of the breast.

Apocrine carcinomas are often positive for GCDFP-15 [33]. Tumor cells are often ER/PR and Her-2/neu oncogene negative as a triple negative tumor. Tumor cells are positive for basal phenotype markers such as CK5, CK5/6 and EGFR is positive in 50% of cases [34]. Treatment of apocrine carcinoma is no different from any other breast carcinoma of similar stage and receptor status. However, demonstration of androgen receptor positivity in apocrine carcinoma may provide a specific targeted therapy [35].

Infiltrating ductal carcinoma with osteoclastic giant cells

Carcinoma with osteoclast like giant cells rarely occurs in the breast. Clinically, the presentation is similar to other cancers, and the average age at diagnosis is 50 years [36]. By imaging, it appears as a round and well-circumscribed mass that may contain calcifications [37].

Grossly, these well-circumscribed lesions are spongey and firm, and display dark brown color [38]. Microscopically, neoplastic cells are immersed in a hypervascular stroma with multinucleated giant cells and red blood cell extravasation (Figure 16). Immunostaining shows positive reactions with CD68 for the giant cells. Tumor cells are usually hormone receptor positive and HER-2/neu oncogene negative [39]. Therefore, these tumors are classified as luminal molecular subtypes [40]. Axillary lymph node involvement is common in this tumor, with 5-year survival rates of approximately 70%, similar to the other types of infiltrating ductal carcinoma [41].

Figure 16.

Photomicrograph of an infiltrating ductal carcinoma of the breast with osteoclastic giant cells.

Medullary carcinoma

Medullary carcinoma is a distinctive subtype of invasive breast carcinoma that accounts for less than 1% of all invasive breast carcinoma and is associated with a favorable prognosis despite aggressive morphologic features [19]. It occurs in patients in their 30s and 40s who present with a breast mass, and by imaging, they are round, oval or lobulated [42]. Grossly, most tumors are more than 2 cm and microscopically, tumor cells have a syncytial growth pattern with pushing border surrounded by lymphoplasmacystic infiltrate (Figure 17) [43].

Figure 17.

Photomicrograph of medullary carcinoma of the breast.

The immunoprofile of medullary carcinoma is similar to those seen in triple-negative tumors with no immunoreactivity for ER, PR and HER-2/neu oncogene. P53 is often positive. Medullary carcinoma are frequently associated with BRCA1 mutations [44]. Medullary carcinoma are associated with better outcomes and lower rates of axillary lymph node involvement [45]. Treatment is no different from other breast carcinomas.

Metaplastic carcinoma

Metaplastic carcinomas account for less than 1% of breast carcinomas and represent a heterogenous group that can display squamous, adenocarcinoma, spindle cells and/or heterogenous mesenclymal elements in various combinations [46]. Metaplastic carcinomas, regardless of the association with heterologous elements, originate from carcinomas that undergo sarcomatous neometaplasia, as evidenced by the presence of monoclonal histogenesis of various components of metaplastic carcinomas [47]. Immunohistochemical studies have shown evidence of myoepithelial cell differentiation demonstrated by expression of basal cell type cytokeratins and established myoepithelial markers CD10, P63, smooth muscle actin and S-100 [48].

Clinically, metaplastic carcinomas are no different from other types of breast cancer. Microcalcification is a common finding by breast imaging. Grossly, they present as firm, solid and well-defined lesions. Morphologically, metaplastic carcinomas are biphasic, with epithelial and spindle cell components (Figure 18). Occasionally, spindle cell components may be dominant. Histologic grade varies and there is a spectrum of features seen that collectively stain from epithelial markers. In the absence of positive immunostaining for cytokeratin, another diagnosis should be entertained [49].

Figure 18.

Metaplastic carcinoma of the breast.

(A) Squamous cell carcinoma areas; (B) osseous differentiation.

Molecular distinction of metaplastic carcinoma includes triple-negative features with no expression of ER, PR and HER-2/neu oncogene. Biologic behaviors of metaplastic carcinoma varies from lower to higher grade, and prognostic and therapeutic decision making are based on the type of tumor. Clearly, high-grade metaplastic carcinomas have an aggressive biologic behavior with similar therapy as any other breast carcinoma. Triple-negative status of metaplastic carcinomas remain to be the limiting factor in the use of targeted therapy [46,50].

Micropapillary carcinoma

Pure micropapillary carcinoma accounts for 2% of all breast carcinomas. The age range, and clinical presentation of micropapillary carcinomas are almost the same as other types, and most frequently occur in the upper outer quadrant of the breast [51]. Imaging characteristics of micropapillary carcinoma are often suggestive of malignancies, and show a high-density irregular mass with speculated margins and areas of microcalcifications. Often, there is evidence of axillary lymphadenopathy by breast imaging [52].

Grossly, tumors range from a few millimeters to a few centimeters. However, these tumors are often larger than other breast carcinomas of no special type [39]. Microscopically, tumor cells are arranged in micropapillary, tubuloalveolar or morular clusters with sponge-like appearance and lack the fibrovascular core commonly seen in papillary carcinoma (Figure 19) [39]. Micropapillary carcinomas are often high-grade in nature. This tumor is associated with poor prognosis, with a 5-year disease-free survival rate of 50%. Lymph node metastasis occurs in about 80% of patients with micropapillary carcinoma, and the treatment is the same as other breast cancers [39,53].

Figure 19.

Photomicrograph of a micropapillary carcinoma of the breast.

The aggressive nature of micropapillary carcinoma is evident by its large size at the time of presentation, marked lymphotropism, extensive axillary lymph node involvement, frequent recurrence and distant metastasis. Biologically, micropapillary carcinomas show variable expression of ER, PR and are frequently HER-2/neu oncogene positive, have high proliferation rates with evidence of P53 expression [54]. Genomic hybridization has also shown evidence of chromosome 8 abnormalities [55].

Mucinous carcinomas

Mucinous carcinomas of the breast, also known as colloid carcinoma, is rare and accounts for approximately 2% of all primary breast carcinomas. Pure mucinous carcinoma of the breast is characterized by the presence of extracellular mucin and is associated with a better prognosis compared with the mix from when there is less than 90% mucinous differentiation [56].

Clinically, mucinous carcinoma occurs at older age groups, and often presents as a palpable mass, however it may also be discovered as an image-detected abnormality [57]. Mucinous carcinoma presents as a well-defined lesion by mammography, similar to its gross appearance. Microscopically, the characteristic features of mucinous carcinoma are the presence of abundant amounts of extracellular mucin surrounding isolated and clusters of tumor cells that show different growth patterns (Figure 20) [58].

Figure 20.

Photomicrograph of a mucinous (colloid) carcinoma of the breast.

Pure mucinous carcinoma of the breast is associated with favorable biologic markers, and often are positive for ER and PR, and negative for Her-2/neu oncogene [59]. Concurrent Iq gain and IqG loss, a hallmark feature of low-grade infiltrating ductal cell carcinoma, has been identified in about 15% of pure mucinous carcinoma by comparative genomic hybridization [59]. Patients with pure mucinous carcinoma of the breast are the most suitable candidates for breast conservation therapy [60].

Invasive cribriform carcinoma

Invasive cribriform carcinoma is a well-differentiated variant of breast carcinoma that is characterized by cribriform patterns of growth similar to what is seen in low-grade ductal carcinoma in situ, cribriform type (Figure 21). This tumor often occurs in patients in their 50s, presenting as a mass or an image-detected abnormality such as a spiculated mass with and without microcalcification [61,62]. Invasive cribriform carcinoma experiences a better overall survival compared with other types of breast carcinoma, with almost no evidence of regional or distant metastasis. Tumor cells are positive for PR and ER, and negative for Her-2/neu oncogene [63].

Figure 21.

Photomicrograph of Cribriform carcinoma of the breast.

Neuroendocrine carcinoma

Recently, the WHO recognized neuroendocrine carcinoma of the breast as a special histologic type of breast cancer [41]. This tumor is characterized by morphologic features similar to those neuroendocrine tumors of the gastrointestinal tract and lung that express neuroendocrine markers in more than 50% of neoplastic cells. Neuroendocrine carcinomas account for 0.5 to 5% of breast carcinoma [41,64] and occur at older age compared with other types of breast cancer [65]. Most patients present with a palpable mass and experience rapid growth and advanced stage breast cancer [65].

Breast imaging and gross findings of neuroendocrine carcinomas demonstrate circumscribed tumors. Microscopically, the WHO recognizes different variants of this tumor that include solid, large cell and small/oat cell carcinomas [41]. An alveolar variant is similar to those of infiltrating lobular carcinoma. Occasionally, carcinoid-like patterns with rosette-like structures and peripheral palisading can be seen. Tumor cells are large with granular and eosinophilic cytoplasm. The small/oat cell variant is similar to those seen in other organs and must be accurately recognized since this tumor is associated with poor prognosis and is responsive to chemotherapy (Figure 22) [65–67].

Figure 22.

Photomicrograph of small cell neuroendocrine carcinoma of the breast.

Neuroendocrine carcinomas express neuroendocrine markers and by electron microscopy, they show presence of intracytoplasmic dense-core secreting granules and clear vesicles of synoptic type [41]. Tumor cells are typically of luminal type with ER and PR positive immunostaining and negative HER-2/neu oncogene protein [68]. The presence of an in situ component and positive expression for ER may serve as the best indication to distinguish primary neuroendocrine tumors versus a metastasis from another site [69]. There is no consensus yet as to what is the best approach to therapy for neuroendocrine carcinoma. However, response to conventional therapy without disease progression at a follow-up to 48 months has been reported [70]. Others recorded the use of VP6 and cisplatin as for other small-cell carcinomas [71].

Tubular carcinomas

Tubular carcinoma accounts for less than 2% of breast cancers [71]. Higher incidence of tubular carcinoma (7.7–27%) is reported in mammographically-screened populations [72]. Clinically, it occurs in women in their 50s and 60s [71]. Tubular carcinoma presents as a palpable mass or as an image-detected abnormality [73,74]. The incidence of bilaterality in tubular carcinoma ranges from 15 to 26%. Mammographically, the lesion appears as a mass lesion with central density and microcalcifications [75,76]. By ultrasound, tubular carcinoma shows features suggestive of malignancy [76].

Grossly, tubular carcinomas demonstrate a gray-white ill-defined, and firm, spiculated mass. Microscopically, tubular carcinoma is characterized by a disorganized population for glandular structure or tubules distributed in a satellite fashion. Tumor cells are round, regular with angulated tubular configuration with no necrosis or mitosis (Figure 23). The majority of tubular carcinomas are associated with a wide range of atypical proliferative lesions, flat atypia, ductal carcinoma in situ and/or lobular neoplasia [77]. Tumor cells are ER and PR positive and HER-2/neu oncogene negative, typical of luminal type breast cancers. Because of well-differentiated nature of tubular carcinoma and low incidence of lymph node metastasis, conservative therapy is the current recommended approach [78,79].

Figure 23.

Photomicrograph of tubular carcinoma of the breast.

Other rare breast carcinomas

Secretory carcinoma, Lipid-rich carcinoma and Glycogen-Rich clear cell carcinoma are other subtypes of breast carcinoma that rarely occur and represent only a few percent of breast carcinomas. Clinical and imaging features are similar to other types of breast cancer, however morphologically, they demonstrate a spectrum of tumor cells with specific characteristics [80].

Secretory carcinomas often display a ‘triple-negative’ phenotype [80] and show presence of recurrent chromosome translocations [81]. Interestingly, this tumor is associated with a relatively good prognosis despite its molecular phenotype [82]. Lipid-rich carcinomas typically have vacuolated or clear cell cytoplasm that share positive reactions for neutral lipid. These types are high grade with aggressive behavior and tend to be hormone-receptor negative as well as HER-2/neu oncogene negative [83]. Glycogen-rich clear cell carcinoma are types with abundant clear cytoplasm that contain glycogen [84]. Expression of ER and PR and HER-2/neu oncogene is variable and prognosis is similar to that of a high grade breast carcinoma [85].

Gene-expression profiles to predict tumor recurrence & metastasis

During the last several years, advances in molecular testing and increasing interest in providing personalized treatment options have led to significant numbers of publications on genomic profiling of breast cancer. Some of the markers are still under investigation, and few are well-established, such as the multigene assay EndoPredict. For the purposes of this paper, this paragraph touches briefly on ‘Oncotype DX 21 Gene Assay’ and ‘Mammaprint 70-Gene Signature’, as extensive information on this subject is beyond the scope of this manuscript [86,87].

Currently, there are two distinct gene expression profiles, the 21-gene recurrence risk (RS) assay and the 70-gene assay of MammaPrint that have complemented the landmark work of the intrinsic subtypes of breast cancer.

Oncotype DX 21 gene assay

Oncotype DX 21 Gene Assay is the product of a quantitative PCR assay for fixed, paraffin-embedded tissue samples that is clinically validated to assess the risks of 10-year distant recurrence and the magnitude of chemotherapy benefit. This test is currently validated for women with early-stage ER positive breast cancer who will be treated with 5 years of hormonal therapy. Oncotype DX was developed using a candidate gene approach, in which 250 genes were selected from published literature, genomic databases and experimental microarray data for breast cancer. The test result is repeated into three risk groups: A low-risk score correlating with a risk of distant recurrence (DR) less than 10%, low recurrence score (RS) 0–18; an intermediate-risk score, correlating with a risk of DR between 10 and 20%, RS between 18 and 31; and a high-risk score correlating with a risk of DR greater than 20%, RS 31 or greater [6].

Based on the result of this assay, patients with low RS are spared from the use of unnecessary chemotherapy, while those with high RS tumors have up to 28% absolute benefit from chemotherapy [6].

MAMMAPRINT 70-gene signature

This assay is the result of efforts of The Netherlands Group [7] that identified a 70-gene classifier using expression microarrays using 25,000 human genes from fresh tissues obtained from 78 young patients. They reported those patients with poor prognosis as those with distant metastasis and those with good prognosis with no metastasis after 5 years. The result of this 70-gene analysis identified those two distinct groups of breast cancer patients: those with good signature, and those with poor signature (Figure 24). Mammaprint is in use for patients younger than 61, early stage disease, have tumor size of less than 5 cm and have no evidence of regional or distant metastasis.

Figure 24.

Validation of the 70-gene classifier. (A) and (B) Kaplan–Meier analysis of the probability that a patient would remain free of distant metastases and the probability of overall survival among all patients.

Reproduced with permission from [88].

Conclusion

The growing interest in betterment of the quality of care among breast cancer patients, coupled with advances in science and technology has provided the scientific community with a unique opportunity for offering personalized therapy for each individual patient. An important component of this approach is the new knowledge about the biology of breast cancer and the understanding that breast cancer is not a single cancer, and therefore cannot be treated the same. Molecular subtyping of breast cancer with different expression of biomarkers as the result of different DNA genetic make-up plays a significant role in precise stratification of breast cancer patients for selection of appropriate therapy for individualized breast cancer care.

Future perspective

Considering the above-stated information, it is clear that breast cancer is not regarded as a single disease. Breast cancer includes at least five discrete, molecularly defined subgroups with distinct natural histories, drug sensitivities and specific molecular therapeutic targets overall. Patients with the luminal type of breast cancer that are ER positive, and those with Her-2/neu oncogene positivity, will continue to benefit from endocrine and Herceptin therapy respectively. The association of basal-like breast carcinomas with triple negativity for ER, PR and HER-2/neu oncogene is a major challenge in selection of chemotherapy. Naturally, these patients will not benefit from endocrine and/or Herceptin therapy. The presence of other biomarkers, such as epidermal growth factor receptors in this subtype of breast carcinoma, may open new therapeutic modalities for patients with basal-like breast carcinomas. Additional studies are required to further identify subgroups of primary breast cancer with different genetic profiles, and to compare the presentation and outcome of breast cancer patients. Attempts are underway to search for new biological insights that may eventually lead to better therapies that are directed toward specific molecular subtypes. Additionally, the role of emerging discoveries in epigenetics, methylation, miRNAs and various different pathways and genetics of hereditary cancers will impact clinical decision-making and patient response to treatment in the future. However, these discoveries are outside the scope of this paper, therefore readers may direct themselves to recent studies by Kasinski et al. [89], Prat et al. [90], and Hon et al. [91].

In the future, more emphasis will be placed in the translation of research technology into routine clinical practice, and there will be more established quality control measures assuring the accuracy of reporting the characteristics of individual breast tumors. With the emergence of more clinically-relevant tumor characteristics and more precise prognostic and predictive information, there will be more opportunity for identification of key therapeutic targets resulting in improved patient outcomes. It is hoped that further study of the biology of more aggressive phenotypes like triple-negative basal-like breast carcinomas will improve the quality of life of breast cancer patients. With new advances in technologies, and more emphasis in further study of molecular biology of breast cancer, it is critically important to underscore the importance of standardization of the technology as well as the interpretation of the results. There is no doubt that uniformity of the use of technology and the development of scientific guidelines for recognition of new entities will result in better stratification of patients for therapy. As more advances are made in molecular genetics and more molecular targeted therapies become available, the responsibility of physicians involved in the diagnosis and management of breast cancer to find the right answers for the right patients will become greater. This approach will form the foundation of the delivery of quality, personalized breast health care.

Executive summary

• There is no doubt that the clinical decision making and response to therapy are not primarily determined by the morphologic features alone.

• Breast cancer subtyping that is the reflection of the molecular profile of each individual tumor is designed to provide optimal treatment planning.

• The luminal subtypes A and B tumors are regulated predominantly by estrogen receptors and respond to endocrine therapy.

• Her-2/neu oncogene positive breast cancer patients respond to Herceptin (R) therapy.

• Basal-like tumors are typically negative for estrogen and progesterone receptors and Her-2/neu oncogene, and exhibit high expression for basal epithelial cell markers. Basal-like tumors are heterogeneous in nature and are associated with poor clinical outcomes.

• Current and emerging technologies are evolving around more specific markers for genetic profiling and better characterization of breast cancer subtypes.