Sustaining Proliferative Signaling

A core cancer-driving trait is acquisition of sustainable growth signaling. Under normal conditions, growth factors bind to specific receptors localized on the cell membrane, leading to release of signal transducers and second messengers in the cell that transmit the signal to the nucleus. As a consequence, nuclear transcription factors are activated and initiate transcription of specific proteins that induce cell division and growth. Regulation of normal cells is related either to their dependency on other cells for growth factors or by an overproduction of a specific growth receptor required to respond to the corresponding growth factor.

Through bypassing the normal regulatory mechanisms, cancer cells acquire self-sufficiency of growth by one of the following common mechanisms:

• 1.
  

  Secreting growth factor ligands themselves (autocrine)

  
  
• 2.
  

  Inducing the normal stromal cells to secrete the growth factors (paracrine)

  
  
• 3.
  

  Overexpression of growth inducing receptors

  
  
• 4.
  

  Altering the structure or type of the receptor, causing ligand-independent activation of proliferative signaling

Abnormal activation of multiple alternative pathways may confer self-sufficiency of growth signaling in breast cancer. Estrogen receptor (ER) signaling, human epidermal growth factor receptor-2 (HER2) membrane receptor signaling, and vascular endothelial growth factor (EGF) receptor family signaling are the most common known proliferation-inducing pathways in breast cancer.

Approximately 70% of breast cancers express ERα, and therefore ER signaling is a key signaling pathway of diagnostic and therapeutic interest. Estrogen could induce proliferation in ERα-mediated and ERα-independent mechanisms. The classical ERα-mediated proliferative response involves estrogen binding to ERα, followed by modulation of target gene regulation through binding to estrogen response elements (ERE) in target gene promoters. In nonclassical activity, ERα can interact with transcription factors, such as SP1 or Fos/Jun, to modulate transcription of genes that do not possess the ERE in their promoter. In addition, ER activates mitogen-activated protein kinase (MAPK) by forming a complex with it in the nucleus to mediate transcription of specific target genes. An increased rate of cell proliferation results in increased mutagenesis. ER-independent proliferative mechanisms, not widely accepted yet, propose genotoxic damage by estradiol metabolites generated by cytochrome P450.

EGF receptor family member HER2, also known as ERBB2 (avian erythroblastosis oncogene B2), is another common proliferative signaling pathway in breast cancer. HER2 gene amplification results in overexpression of the HER2 transmembrane receptor, which is involved in growth signaling. Approximately 15% to 25% of invasive breast cancers have amplification of HER2, and these cancers are associated with aggressive disease behavior and poor prognosis. The overexpressed HER2 protein forms both homodimers and heterodimers with other epidermal growth factor receptor family receptor tyrosine kinases, and these activate downstream signaling. In addition to homodimerization, HER2 can form heterodimers with other members of the EGF receptor family and activate downstream growth signaling pathways, including but not limited to mitogen-activated protein kinase and phosphatidylinositol 3-kinase (PI3K) signaling.

Other upstream growth pathways involved in proliferative signaling in breast cancer are vascular endothelial growth factor (VEGF) signaling, fibroblast growth factor receptor (FGFR) signaling, and autocrine activation of wingless and integration site growth factor (Wnt) signaling pathway signaling. Although somatic mutations are known to activate downstream signaling pathways, a recent study in breast cancer showed amplifications of PIK3CA, KRAS, BRAF, and EGFR, thereby contributing to the abnormal activation of Ras-Raf-MAPK signaling.

Inactivation of negative-feedback mechanisms known to attenuate self-sufficient proliferative signaling is an alternative way by which cancer cells enhance proliferative signaling. PTEN is an example of negative feedback, known to antagonize PI3K signaling. In addition, nuclear PTEN downregulates cyclin D1 and prevents the phosphorylation of MAPK, thus loss of function alterations in PTEN promotes proliferation because of loss of its negative feedback.

Evading Growth Suppressors

Cancer cells are capable of evading a variety of growth suppression mechanisms located in the extracellular matrix or on the surface of nearby cells.

Retinoblastoma gene (RB) is considered a gatekeeper and prototype tumor suppressor protein that has a pivotal role in cell cycle progression because of its role as transducer of extracellular growth inhibitory signals. Alterations in RB signaling can lead to persistent proliferation. Cell proliferation in breast cancer cell lines may be reversed through inhibition of G1 cyclin-dependent kinases (CDKs) with CDK 4/6 inhibitors leading to hypophosphorylation of pRb and a G1/S-phase cell cycle arrest. Studies of breast cancer cell lines show the most sensitive cell lines have elevated expression of Rb and cyclin D1 but decreased expression of CDKN2A (p16 ^INK4a ). These observations have been confirmed in clinical trials of CDK 4/6 inhibitors (palbociclib, ribociclib, and abemaciclib) in patients with metastatic ER-positive, HER2-negative breast cancer.

Transforming growth factor (TGF)β signaling, another growth suppressor pathway, has been shown to have antiproliferative activity in early breast cancers by downregulation of c-myc oncogene. Altered TGFβ receptors or mutations of components of the signaling pathway might lead to evasion of growth suppression by TGFβ. Mutation or altered expression of SMAD4 protein, which is involved in transducing the TGFβ signals could alter the growth suppression by TGFβ.

Resisting Cell Death

Programmed cell death or apoptosis is an essential pathway for maintaining tissue homeostasis and elimination of cells with irreversible damage. Apoptosis mechanisms are divided into two components: First, an extrinsic pathway, which receives and processes extracellular death signals involving Fas ligand and tumor necrosis factor-α (TNFα); second, an intrinsic pathway that integrates intracellular signals, such as release of mitochondrial cytochrome C. Both pathways result in activation of downstream effector caspases. The intrinsic pathway is regulated by the B-cell lymphoma 2 (BCL2) family proteins. Sensing DNA damage, TP53, a checkpoint regulator, can induce apoptosis by activating BH3-only proteins, activating survival pathways such as PI3K-AKT/PKB pathway, increasing insulin-like growth factor (IGF)1/2 and interleukin 3. Cancer cells evade apoptosis by acquiring loss of function mutations in TP53, observed in approximately 37% of breast cancer samples, downregulating proapoptotic factors (BH3 only proteins) and inducing antiapoptotic proteins, for instance, BCL2.

Enabling Replicative Immortality

In contrast to normal cells, which are capable of passing through a finite number of growth and division cycles before becoming senescent, cancer cells must acquire replicative immortality. It has been proposed that for cancer cells to acquire this hallmark trait, they upregulate telomerase, a specialized DNA polymerase enzyme that is involved in maintaining telomere length or can add nucleotide repeats—TTAGGG oriented 5′-to-3′ toward and at the end of the chromosomes. More recently a protein subunit of telomerase, known as telomerase reverse transcriptase or TERT, has been shown to possess other functions, even after elimination of the canonical telomerase activity.

Telomerase expression observed in approximately 53% of breast cancers was associated with poor outcomes in patients treated with chemotherapy, while the opposite effect was observed with endocrine therapy. ERα can increase the telomerase activity, whereas the anti-ER tamoxifen is known to suppress telomerase activity.

Inducing Angiogenesis

In contrast to transient activity of angiogenesis in normal tissues, activation of an “angiogenic switch” in cancer cells underlies sustained angiogenesis, thereby stimulating abnormal and sustained angiogenesis in cancer that transports oxygen and nutrients to cancer cells and evacuates the waste and carbon dioxide. An imbalance between the factors that promote and oppose angiogenesis results in activation of persistent angiogenesis. In response to hypoxia or oncogenic stimuli, prototype angiogenesis initiator VEGF is expressed and binds to VEGF receptors (VEGFR1, VEGFR2, and VEGFR3) on cell membrane of endothelial cells. In addition, fibroblast growth factor 1/2 are potent proangiogenic factors, whereas TSP-1 (thrombospondin-1) counteracts the proangiogenic factors by binding the transmembrane receptors on endothelial cells. Although initial clinical trial results using humanized anti-VEGF-A monoclonal antibodies (bevacizumab) in breast cancer patients appeared promising, subsequent clinical trials failed to demonstrate a significant improvement in either progression-free (PFS) or overall (OS) survival.

Additionally, hematopoietic progenitors such as bone marrow–derived progenitor cells may potentially play a role in cancer angiogenesis. However, more comprehensive evidence is necessary to understand migration and invasion of these cells into developing breast tumors in vivo.

Activation of Invasion and Metastasis

Transformation of well-differentiated normal epithelial cells into often poorly differentiated cancer cells with migratory and invasive properties is a consequence of specific molecular alterations. The multistep process of invasion and metastasis is described as the invasion-metastasis cascade, which results in local invasion, intravasation of cancer cells, transit, extravasation at distant tissue sites, followed by colonization with a small fraction of the cancer cells, resulting in formation of disseminated cancer nodules (micrometastasis) and, lastly, some proportion of micrometastatic nodules adapt to the local microenvironment and proliferate to become mature metastases. The invasion-metastasis cascade is regulated by a multifaceted program, referred as epithelial-mesenchymal transition (EMT), which permits transformed epithelial cells to acquire mesenchymal features including motility, invasiveness, and resistance to apoptosis. The entry of cancer cells into EMT may be transient, stable, or partial.

E-cadherin is a key protein involved in formation of cell-to-cell adherens junctions that assembles adjacent epithelial cells and maintains quiescence of the cells. Loss of functional E-cadherin by transcriptional repression, mutation, chromosomal deletion, or protein degradation may confer metastatic potential and ultimately provoke EMT, in addition to loss of cell-to-cell adherens junctions.

Among other factors, specific members of the integrin family, matrix metalloproteinases, and crosstalk between tumor and stromal cells contribute to invasion and metastasis.

Genome Instability and Mutation

Unlike normal cells, which only infrequently contain mutations, cancer cells have significantly high rates of acquired mutations. Initial occurrence of mutations in gatekeeper genes result in genomic instability, and multiple genomic alterations are a key feature of cancer cells because of their role in developing other cancer hallmark capabilities. Whereas most cancers originate from a single ancestor cell due to a specific mutation(s), their progression is attributed to clonal evolution of the cells with acquisition of additional genomic alterations that confer selective growth advantages and lead to more aggressive behavior.

Mutations in DNA repair genes can lead to a mutator phenotype that subsequently gives rise to more mutations and eventually genomic instability. Genomic instability is a primary event in hereditary cancers with germline mutations in genes, such as BRCA1 and BRCA2, that are linked to DNA repair. In sporadic cancers, oncogene-induced collapse of DNA replication forks may lead to DNA double strand breaks and genomic instability.

Occurrence of mutations may have diverse, often overlapping or linked functional contributions in conferring hallmark traits including genomic instability, resulting in cancer progression ( Fig. 22.2 ). Tumor suppressor genes are of prominent interest in development of cancer hallmarks. The gatekeeper mutations occur in early cancers and contribute to loss of growth regulation and poor differentiation of cancer cells. TP53 is a well-studied gene involved in direct regulation of cell cycle, apoptosis, and DNA repair, thereby maintaining genomic integrity. Germline or early occurring somatic mutations of TP53 drive cancer by loss of their gatekeeper function. Mutation in E-cadherin is another such example. Caretaker genes are involved in DNA repair and checkpoint control; however, they cannot initiate carcinogenesis as a direct consequence of mutations. Mutations in BRCA1 and BRCA2 in breast cancer represent good examples of caretaker mutations. The landscaper mutations modulate the microenvironment and can promote cell proliferation. No comprehensive evidence is available that identifies any mutations in the role of a landscape mutation. However, among hereditary BRCA1/2 mutant cases, alterations in stromal cells have been proposed to contribute as landscape mutations. In addition, mutations may be broadly categorized as driver and passenger mutations. The driver mutations are capable of conferring selective advantage to cancer cells promoting carcinogenesis and survival of the cancerous cells. Mutations in TP53, PIK3CA, MAP3K1, GATA3, CDH1, BRCA1, PTEN, PIK3R1, AKT1, and RB1 are among the prominent driver mutations in breast cancer. Passenger mutations are other nonsynonymous mutations that do not play a direct role in carcinogenesis or progression. Mutations could also cause partial or complete loss of function that applies to tumor suppressor functions. Specific mutations of TP53 cause gain of novel functions. For instance, interference with the p53 function preventing DNA damage and promoting genetic instability, is the opposite of its wild-type role.

Classifications of mutations in cancer: mutations with diverse functional roles in acquisition of cancer hallmark traits by the cancer cells.

In addition to mutations, loss of telomeric DNA and specific copy number alterations are other factors that may contribute to genomic instability.

Tumor Promoting Inflammation

Cancers induce immune response, resulting in immune infiltration and inflammation. Inflammation could contribute to cancer progression by expression of growth factors, survival factors, and proangiogenic factors to the local microenvironment to promote tumor formation.

Reprogramming Energetics

Normal cells favor glycolysis under anaerobic stress. In contrast, cancer cells, even under aerobic conditions, favor metabolism of glucose to lactate, a state referred as anaerobic glycolysis. This means that the usual preference for oxidative phosphorylation (OXPHOS) in normal cells is shifted to metabolically inefficient glycolysis despite the presence of oxygen, mainly attributed to mitochondrial malfunction and a hypoxic microenvironment. Cancer cells face two challenges: first, the need to continue support for proliferation and growth by reprogramming metabolism and, second, achieving metabolic adaptation to survive metabolic stress. Regulation of adenosine triphosphate (ATP) synthesis by regulating substrate uptake and glycolytic enzymes is a possible mechanism that enables cells to adapt to the microenvironment.

Mutations in cancer cells also play a role in reprogramming of cancer cell energetics. Mutations in PIK3CA are common in breast cancer. Mutant PIK3CA or loss of its regulator PTEN can promote PI3K/Akt signaling and ultimately changes in metabolism, by increased expression of nutrient transporters, stimulation of hexo-/phosphor-fructokinase to drive glycolysis, enhanced transcription of genes involved in glycolysis, and/or by an enhanced rate of protein synthesis through Akt-dependent mTOR activation.

Considering reprogramming of metabolism as a consequence of oncogenic activity brings up the question of whether it is a core hallmark of cancer. Regardless, it remains a potential target for novel therapeutic strategies.

Evading Immune Destruction

Cancer cells evade the immunosurveillance and thereby continue to grow locally, metastasize to distant tissues and ultimately result in a macroscopic tumor. There are two potential mechanisms underlying dissemination of cancers: First, mutations accumulated either in a subpopulation of primary tumor or in disseminated cells to confer metastatic potential to a fraction of cells and second, expression of immune mediators (i.e., TGFβ, TP53, HIF, VEGF) by the tumor cells could modulate immune response. T regulatory cells (Tregs) that normally suppress T-cell response and natural killer (NK)-cell activity are observed to be significantly associated with larger tumor size and trend toward an association with estrogen receptor negativity. In addition, the same study observed an association with ER/progesterone receptors (PR)/HER2-negative or triple-negative breast cancer (TNBC) and Nottingham grade III cases that were significantly associated with high number of FOXP3 (fork-head/winged-helix transcription factor, which is involved in the development in Treg cells).

The presence of immune cells in breast cancers may also reflect the potential for harnessing the patients’ own immune system to eliminate tumor cells. The presence of tumor-infiltrating lymphocytes (TILs), especially in lymphocyte-predominant breast cancers, has been associated with significant improvements in pathologic complete response to chemotherapy. The potential protective role of lymphocytes has been appreciated in cancers with microsatellite instability. The inability of these cancers to efficiently repair errors in exons leads to expression of proteins containing neo-antigens that may be identified by a patient’s own immune system ostensibly leading to a high density of TILs. These neo-antigens expressed in proteins on the cell surface are speculated to be most likely to have this effect. This concept has led to the development of novel immunomodulatory therapies, such as programmed cell death 1 inhibitors for the treatment of metastatic cancers, including breast cancers.

Endogenous Hormones and Growth Factors

Understanding breast carcinogenesis requires an understanding of the normal development of the breast. Breast tissue is composed of epithelial and mesenchymal components. Opening onto the skin surface and lined by squamous cells, breast ducts are lined by a double-layered epithelium composed of luminal and myoepithelial cells for the vast majority of the ductal system throughout the breast. The luminal epithelial cells are most prone to malignant transformation. During development steroid hormone receptors, especially ER and PR, facilitate cellular responses to hormones that, among other activities, lead to proliferation of the luminal cells.

Although all mammals are born with a rudimentary mammary gland, after birth, through a process of cyclical growth and involution coupled to body hormone status ( Fig. 22.3 ), the primitive breast undergoes extensive morphogenesis through invasion of the surrounding fat pad and progressive branching to the outer limits of the mesenchymal fat pad to create an extensive treelike structure, the terminal duct lobule unit (TDLU), specialized for milk production during pregnancy.

Multistage development from fetal to menopausal stages. Appearance of alveoli, increase in their number with increased degree of branching followed by the decrease in both at involution. GH, Growth hormone; MFP, mammary fat pad; MMP, matrix metalloproteinase; TEB, terminal end bud.

(Modified from Inman JL, Robertson C, Mott JD, Bissell MJ. Mammary gland development: cell fate specification, stem cells and the microenvironment. Development. 2015;142:1028-1042.)

The branching of large ducts eventually leads to the TDLUs that are composed of terminal ducts and blind-ended ductules. TDLUs in adult women branch into small acini to form the lobules. During adulthood, TDLUs demonstrate proliferative activity, predominantly during the second half of each menstrual cycle, and subsequent epithelial apoptosis at the conclusion of each cycle. Presumably as a result of this repetitive proliferative activity with each menstrual cycle, it is the TDLUs that are responsible for the development of the vast majority of breast cancers.

Estradiol and other estrogens interact with two specific intracellular (nuclear) receptors, which are both inducible transcription factors, ERα and ERβ. Some normal luminal epithelial cells express the ERα and are considered relatively mature cells that infrequently cycle in the adult breast. ERα is expressed in infant, pubertal, and cycling adult breast epithelial cells. ERβ is expressed in most luminal, myoepithelial cells, fibroblasts, and lymphocytes. PR is expressed in the breast as two isoforms, PR-A and PR-B, based on alternative initial transcription start sites from a single PGR gene. PR is expressed in a minority of cells in the luminal epithelium concentrated in the terminal bud cells. Expression of PR is induced by estrogen action at the transcriptional level and is decreased by progestins at both transcriptional and translational levels. In addition to ER/PR-mediated proliferation, a paracrine or juxtacrine mode of regulation of proliferation has been proposed.

Multiple stages of mammary development are depicted in Fig. 22.3 . Here we briefly present an overview of the process, the factors involved in it, and the role of stem cells in breast morphogenesis, the details of which are described in other chapters.

Clinical Perspective on Carcinogenesis and Progression

Initiated as premalignant or preinvasive lesions, such as atypical ductal or lobular hyperplasia, ductal carcinoma in situ (DCIS), and lobular carcinoma in situ (LCIS), breast carcinomas acquire multiple hallmarks of cancer in multiple stages and evolve into invasive carcinomas, as demonstrated by the modified Wellings-Jensen-Marcum model. This model provides a context for identification of the underlying molecular alterations in the cancer hallmarks theory and, thereby, the study of cancer progression is viewed as a consequence of acquisition of driver and passenger mutations, copy number aberrations, and other genomic and epigenomic changes in the terminal duct lobular unit (TDLU). In addition, the intertumor, intratumoral spatial and temporal heterogeneity, as explained by the clonal evolution model ( Fig. 22.4 ), provides a broad overview of clinical cancer progression as a phenotypic outcome of interactions within multiple competing clones within the stromal microenvironment under selective pressure, including the dominant and outcome-determinant subclones in combination with minor subclones, in the context of treatment. From a clinical perspective, this implies the need for multimodal diagnostic and therapeutic approaches to achieve effective disease management.

Clonal evolution model: multistage progression of cancer originating from a founder cell evolves into localized tumor (niche) and ultimately a mature primary tumor. Under the selective pressure, dominant clone(s) grow and metastasize to distant organs. The minor subclone may potentially drive clinical progression.

(Modified from Kleppe M, Levine RL. Tumor heterogeneity confounds and illuminates: assessing the implications. Nat Med. 2014;20:342-344.)

The following are considered priorities for multimodal clinical management:

• 1.
  

  Large studies to identify specific genomic alterations that underlie altered activity of specific biological pathways that can drive tumor progression and clinical outcome.

  
  
• 2.
  

  Development of high-resolution, robust, and reproducible tools to identify molecular drivers of breast cancers.

  
  
• 3.
  

  Availability of decision-making tools to expand the use of effective targeted therapeutics for particular cancer-driver pathways.

Estrogen Receptor

ERα, a member of the steroid hormone receptor superfamily and a ligand-dependent transcription factor, is a regulator of various essential developmental and physiologic processes. Estradiol, the predominant form of biologically active estrogen, interacts with ER for physiologic development and regulation of the reproductive system, bone metabolism, and effects on the cardiovascular and central nervous systems. Given its core involvement in growth and proliferation of breast tissue, a dysregulation of the estrogen hormonal axis in developmental aberrations associated with breast carcinogenesis is not surprising. The link between estrogen and breast cancer was observed as early as 1896 by George Beatson, who reported regression of advanced breast cancers in patients after oophorectomy. Since then a number of clinical studies have shown that both endogenous estrogen levels and the duration of hormone replacement therapy, especially estrogen plus progestin, are correlated with an increased risk of breast cancer and disease-related mortality, which has evolved into the current view of ER and PR signaling as a core determinant of breast cancer biology and patient outcome.

The effects of estrogen are mediated through two receptor proteins, ERα and ERβ, encoded by corresponding genes on chromosome 6q24 and 14q22, respectively. Both receptors share about 97% homology in the DNA binding domain, with relatively low homology in other domains ( Fig. 22.5 ), indicating that they may share specificity for similar estrogen-responsive elements (EREs) on target gene promoters or enhancers. Although the role of ERβ in normal tissue is not known, loss of ERβ expression in breast cancer is associated with increased proliferation. Nevertheless, ERα, but not ERβ, is currently considered to be relevant for patient management and treatment decisions, and therefore only ERα is discussed further.

The structure of the estrogen receptors, their regulation, and functions. (A) The estrogen receptor contains three main functional domains: the N-terminal activating function 1 (AF1); a DNA-binding domain, a hinge region; and a ligand-binding domain, consisting of activating function 2 (AF2). The AFs bind to the coregulators. (B) Typical domain structure of nuclear receptor with two major functional domains: the protein-protein interactions domain, which binds the coactivators in the coactivator complex, and the enzyme activity domain, with either intrinsic enzymatic activity or an activity by binding the protein.

(From Jordan VC, O’Malley BW. Selective estrogen-receptor modulators and antihormonal resistance in breast cancer. J Clin Oncol. 2007;25:5815-5824.)

Upon binding of 17β estradiol, nuclear ER monomers dimerize, bind to EREs, and, in cooperation with coregulators (coactivators and corepressors), facilitate transcription of target genes, referred as classical genomic activity of ER (see Fig. 22.5 ). The nuclear coregulators modulate transcription through multiple mechanisms, particularly chromosome remodeling and posttranslational modification. The indirect genomic activity involves ER binding to other transcription factors, such as SP1, AP1, and NFκB, and thereby regulation of transcription through their binding to the target gene promoters. Cyclin D1 is a direct ER target that can activate CDK4 and CDK6, which phosphorylate RB with release of E2F, leading to induction of the cell cycle proliferative activity.

In recent years, cytoplasmic/membrane-bound ER cross-talk with HER2, EGFR, IGF1R, and FGFR has been proposed as a trigger contributing to the activation of AKT and downstream signaling components of the PI3K pathway. However, the mechanism for this cross-talk is not clear, and comprehensive evidence is needed to establish the presence of significant quantities of cytoplasmic ER, as well as its functional impact in patients.

Breast cancers that express ER are potentially sensitive to endocrine therapy in both primary and metastatic disease settings. ER-positive breast cancer patients have relatively better 5-year survival compared with ER-negative breast cancer patients, although substantially reduced difference in 10-year survival because of the delayed recurrence observed in ER-positive disease. ER expression status is a core predictor of disease aggressiveness, therapeutic response, and patient outcome. The objective of endocrine therapy is to target the key growth-driving pathway—ER signaling—by one of three alternative approaches:

• 1.
  

  Reduction of circulating estrogen levels through surgical ovarian ablation or, more recently, through inhibition of peripheral conversion of steroids to estrogens using drug-mediated inhibition of aromatase with aromatase inhibitor (AI) therapies.

  
  
• 2.
  

  Competitive inhibition of estrogen binding to ER using selective estrogen receptor modulators (SERM) such as tamoxifen to inhibit recruitment of activated coregulators through conformational displacement of ER helix 12.

  
  
• 3.
  

  Downregulation of ER through irreversible binding of small molecules, such as fulvestrant, to ER with 100 times higher affinity to ER compared with tamoxifen, leading to selective degradation of the ER in breast cancer cells and thereby preventing estrogen-mediated transcriptional activation.

Although AIs are superior to tamoxifen for prevention of contralateral breast cancers as well as prevention of recurrence and metastases in postmenopausal women, tamoxifen is more suitable for adjuvant therapy in premenopausal women. Tamoxifen is also used for prolonged antihormonal therapy beyond the initial 5 years of treatment. A recent meta-analysis demonstrated a 30% reduction in recurrences associated with 5-year therapy with AIs compared with tamoxifen in postmenopausal women with early breast cancers as well as a 15% reduction in mortality.

Approximately 30% of women treated by antihormone therapy may have a recurrence within 15 years. Among the mechanisms of acquired resistance, deregulation of estrogen receptor signaling, either by loss of ERα expression or mutation in the ESR1 gene encoding ERα, are the most common associations. Loss of ER positivity is a common mode of resistance to endocrine therapy. Mutations clustering in the ligand-binding domain (LBD) are observed in advanced metastatic ER-positive samples ( Fig. 22.6A ) but are seldom observed in the primary carcinomas, suggesting a role for acquisition of ESR1 point mutations in advanced-stage disease, leading to resistance to endocrine therapy due to constitutive ligand-independent activation of ER. Recently a plasma circulating DNA-based prospective-retrospective study focused on patients with ER-positive metastatic cancers who developed disease progression after aromatase inhibitor therapy demonstrated that ESR1 mutations were predictive of resistance to Exemestane and sensitivity to fulvestrant. Notably, this study observed that a combination of palbociclib (CDK4/6 inhibitor) and fulvestrant was equally effective in patients with or without ESR1 mutations. Larger studies are required to further establish the predictive role of ER mutations and their relationship to response to various endocrine treatment options.

Estrogen-independent activation of ERα as a potential mechanism of resistance to endocrine therapy: (A) frequencies of ERα mutations displayed by functional domains ERα in ER-positive metastatic breast cancer after endocrine therapy. (B) The WW protein-protein interaction domains on the ERα-YAP1 fusion protein derived by balanced translocation between 6q and 11q that substituted the ligand-binding domain (LBD) of ERα by the carboxyl terminus of YAP1.

(Reproduced with permission from Ma CX, Reinert T, Chmielewska I, Ellis MJ. Mechanisms of aromatase inhibitor resistance. Nat Rev Cancer. 2015;15:261–275.)

In addition, ESR1 chromosomal translocation, leading to fusion genes with ESR1, for instance, ESR1 (exons 1–6)- YAP1 (as illustrated in Fig. 22.6B ), has been proposed to be another potential mechanism underlying resistance to endocrine therapy.

Progesterone Receptor

The PR is a nuclear receptor closely associated to the ER, being a direct target of ERα as well as its involvement in ERα chromatin binding. Expressed as two highly studied isoforms, PRA (94 kDa) and PRB (116 kDa), and coded from a single gene, PR forms upon hormone binding to form homo/hetero-dimers, binding to the PR response elements (PREs) on target gene promoter regions via the DNA binding domain and thereby modulate transcriptional activation of target genes. PRB has been shown to be the predominant regulator for several of the target genes and uniquely possesses an activation function 3 (AF3) domain in addition to common AF1 and AF2 domains on both isoforms, suggesting both common as well as unique targets. PRA may suppress PRB activation, although genes regulated uniquely by PRA are known and PRA expression is higher. Both isoforms are equally expressed in normal breast tissues but altered in breast cancer, with higher PRA in many cases.

PR positivity is slightly less common than ER positivity in invasive cancers, although PR positivity indicates active ER signaling. The independent role of PR status in breast cancer phenotypic outcome has been a matter of debate. ER-positive tumors without PR expression are shown to have increased occurrence of copy number changes compared with ER-positive/PR-positive tumors, was well as enrichment of growth factor signaling via PI3K/Akt/mTOR. Loss of PR expression during tamoxifen therapy tends to be dramatic compared with loss of ER and marks potential resistance to therapy. Loss of PR expression is explained by transcriptional inhibition of PR by other growth pathways that can also modify ER functions, thus supporting a potentially beneficial role of SERDs that can effectively block ER signaling or of simultaneous inhibition of other growth pathways. However, a predictive role of PR has not been observed in early breast cancers.

HER2 Membrane Receptor

HER2 or epidermal growth factor receptor 2, a member of the human EGFR family of tyrosine kinases (RTKs), through homodimerization or heterodimerization with other family members, induces tyrosine kinase activity and consequential activation of pathways involved in cell growth, differentiation, and survival. The expression of HER2 is normally found at low levels in all epithelial cells of both fetal and adult tissues, but not in cells of hematopoietic origin. Amplification and/or overexpression of HER2 is found in a number of solid tumors including breast cancer. HER2 protein is a key driver of growth, an aggressive phenotype in cancers and, therefore, a therapeutic target in the subset of breast and gastric cancers that have alterations in the gene encoding this receptor. Amplification of the HER2 gene (also known as ERBB2 ) is observed in preinvasive breast carcinomas (DCIS) as well as 20% to 30% of invasive breast carcinomas. A role for HER2 gene amplification as a driver alteration that determines biological behavior including progression, resistance to apoptosis, invasion, and metastasis is well established for invasive carcinomas but not DCIS.

Amplification of HER2 is associated with high histologic grade, high proliferative rate, reduced expression of steroid hormone receptors, and lymph node metastases and is a key predictor of shorter patient survival and time to relapse. Besides being a prognostic marker, HER2 amplification, and the associated directly related protein overexpression, predicts responsiveness to trastuzumab, a humanized anti-HER2 monoclonal antibody, with either anthracycline-based chemotherapy or non–anthracycline-based chemotherapy. Another humanized anti-HER2 monoclonal antibody, pertuzumab, as well as an antibody-drug cognate, ado-trastuzumab emtansine (T-DM1), have also been approved for treatment of HER2-positive breast cancer patients. In addition, small molecule inhibitors of the HER2 tyrosine kinase have shown promise in the treatment of HER2-positive breast cancer patients with lapatinib, a dual targeting agent that inhibits both HER2 and HER1 receptors and is Food and Drugs Administration (FDA)-approved for treatment of women with metastatic disease. Other small molecule inhibitors of HER2 are currently in clinical trials.

The downstream consequences of HER2 amplification and its messenger RNA (mRNA) overexpression include formation of active hetero-/homo-dimers of HER2 protein with other RTKs, activation of intracellular tyrosine kinase domain and phosphorylation of the intracellular domain of HER2 and recruitment of adaptor proteins that forms a complex, leading to activation of intracellular signaling pathways, including PI3-kinase/AKT, and RAS/MAPK signaling. HER2 and other downstream signaling pathways increase cell proliferation, cell survival, and loss of apoptosis. Production of VEGF and induction of angiogenesis is one of the downstream consequences of HER2 activation ( Fig. 22.7 ).

HER signaling and core mechanism of action of trastuzumab. Homo/hetero-dimerization of HER family members is followed by the phosphorylation of their intracellular tyrosine kinase domains. The activated domains activate the lipid kinase phosphoinositide 3-kinase (PI3K), which phosphorylates a phosphatidylinositol that in turn binds and phosphorylates the enzyme Ak transforming factor (Akt), driving cell survival. In parallel, a guanine nucleotide exchange factor, the mammalian homologue of the son of sevenless (SOS), activates the rat sarcoma (RAS) enzyme that, in turn, activates receptor activation factor (RAF) and then the mitogen-activated protein kinase (MAPK) and mitogen extracellular signal kinase (MEK). MEK phosphorylates, among others, the MAPK, driving cellular proliferation. Expression of vascular endothelial growth factor (VEGF) is one of the downstream effects of HER signaling, which induces angiogenesis. Trastuzumab binds to the juxtamembrane domain of HER2. It thus inhibits the dimerization involving HER2. Trastuzumab may also recruit Fc-competent immune effector cells and the other components of antibody-dependent cell-mediated cytotoxicity, leading to tumor-cell death (not shown in the figure). The binding reduces shedding of the extracellular domain (proteolytic cleavage), thereby reduces the availability of p95. In addition, potent microtubule inhibitor emtansine (DM-1) in conjugation with trastuzumab provides antimitotic activity on HER2 overexpressing cells.

Additional therapeutic strategies target this signaling pathway by simultaneously targeting HER2 and HER3 with trastuzumab and pertuzumab, a second humanized anti-HER2 antibody that prevents HER2 dimerization, with concurrent cytotoxic chemotherapy. An antibody-drug conjugate, T-DM1, directed at HER2 has also been developed for treatment of HER2-positive breast cancer patients.

Trastuzumab, a humanized monoclonal antibody was FDA-approved for the treatment of HER2-amplified metastatic breast cancers in 1998 and subsequently for early-stage HER2-positive, node-positive breast cancer in 2006. The antigen-specific site of trastuzumab binds the juxtamembrane portion of the extracellular subdomain 4 of the HER2 receptor (see Fig. 22.7 ). The mechanism of action of trastuzumab is not completely understood but multiple possible mechanisms include inhibition of the intracellular activation of the HER2 tyrosine kinase domain, antibody-dependent cellular cytotoxicity (ADCC), through binding of trastuzumab with induction of natural killer cell responses against the HER2-overexpressing cells, inhibition of receptor dimerization, receptor degradation through endocytosis, inhibition of shedding of the extracellular domain, and inhibition of ligand-independent, HER2-mediated mitogenic signaling. Trastuzumab activity has been shown in early as well as metastatic breast cancers with HER2 amplification. HER2 activity is ligand independent. Among the possible heterodimers, the HER2-HER3 pair has been shown to be the most potent in activation of cell proliferation. HER3, upon binding to its ligand, dimerizes with HER2 and activates the phosphatidylinositol-4,5-bisphosphate 3-kinase (PI3-K) signaling pathway. The clinical benefit of trastuzumab therapy can be improved by simultaneously blocking both ligand-independent and ligand-dependent activity. Accordingly, therapy with trastuzumab and pertuzumab could achieve more comprehensive control of HER2 signaling. The CLEOPATRA trial showed significantly reduced risk of disease progression (improved PFS) or death with trastuzumab, pertuzumab, and a docetaxel regimen compared with placebo plus trastuzumab with the same docetaxel therapy in metastatic breast cancers. Resistance to trastuzumab remains an important issue. One or more mechanisms could activate alternate growth signaling pathways and, thereby, overcome the HER2 or HER2-HER3 blockade. Factors such as activation of other members of HER family, activation of PI3K/AKT/mTOR by alternative pathway, IGF1 activation, increased activity of VEGF or src, c-MET overexpression or loss of PTEN are some of the potential mechanisms considered possible underlying mechanisms of resistance. Conjugation of trastuzumab with maytansine (DM1), a tubulin inhibitor, also referred as T-DM1, has demonstrated activity against the HER2-overexpressing cells that were resistant to trastuzumab and lapatinib. The combination of T-DM1 with pertuzumab provided superior pathologic complete response rate compared with the paclitaxel with trastuzumab in the I-SPY 2 (Investigation of Serial Studies to Predict Your Therapeutic Response With Imaging And moLecular Analysis 2) trial.

Molecular Subtypes of Breast Cancer

In addition to ER, PR, and HER2, which are the key predictive biomarkers, advances in high-throughput technologies have shown breast cancer to be a molecularly, highly heterogeneous disease. The initial aim of classifying breast cancers had been for prognostication and prediction of response to chemotherapy, which with technological advances, has evolved into a broader aim of defining molecular classes by causal or driver alterations that determine patient outcome, disease aggressiveness, metastatic potential, or drug responsiveness.

Molecular expression subtypes were initially based on microarray-based differential mRNA expression patterns across the spectrum of breast cancer samples. A set of 496 genes, defining the subtypes, was composed of genes that showed significantly greater variation in expression between different cancers than between paired samples from the same tumor. These genes also represented the biological characteristics of the cancers and were referred as an intrinsic gene set. The expression clusters were separated into five subtypes: luminal A, luminal B, basal, ERBB2-enriched (HER2-positive) and normal-like. Subsequently, using a 50-gene predictor gene set (PAM50) in a study of 710 node-negative cancers, the molecular subtypes were shown to be associated with differential relapse-free survival and were associated with response to chemotherapy. A randomized trial (MA.12 study) of 672 premenopausal, stage I–III breast cancer patients receiving adjuvant tamoxifen showed the PAM50 subtype to be prognostic for both disease-free survival (DFS; p = .0003) and overall survival (OS; p = .0002). This study also showed the luminal subtype to be predictive of response to tamoxifen therapy compared with the nonluminal cases. Using a large sample set of 10,000 pooled samples from multiple studies, immunohistochemistry-based expression of ER, PR, HER2, and basal markers (cytokeratins CK5/6, and EGFR) was applied to identify six subtypes: luminal 1 (HER2-negative) with or without expression of basal markers, luminal 2 (HER2-positive) with or without expression of the basal markers, nonluminal HER2-positive, and TNBC. This classification distinguished breast cancers by their short and long-term survival. Overlapping largely with TNBC with high frequency of metaplastic and medullary differentiation, a novel class of tumors, the claudin-low subtype, was identified that are characterized by low or absent expression of luminal differentiation markers, high expression of EMT markers, immune response genes, and CSC-like features.

Although expression subtypes have improved our understanding of molecular heterogeneity, the known expression subtypes do not represent the complete spectrum of the biological variation. Subtype characterization has been based on mRNA expression using microarrays in small sample sets, but its clinical utility is limited, given the lack of reproducibility and robustness, technological difficulties, and insufficient consideration to tumor heterogeneity. These limitations in clinical utility are discussed in greater detail elsewhere.

The concept of molecular subtyping has continued to evolve with advances in identifying the underlying somatic and germline mutations, copy number alterations and other alterations. Five types of data, including mRNA expression, DNA methylation, single nucleotide polymorphism (SNP) arrays, microRNA (miRNA) sequencing, whole-exome sequencing, and protein expression (reverse phase protein array or RPPA) from tumor and germline DNA showed consensus patterns between samples by unsupervised consensus clustering analysis. The four major clusters obtained from these analyses have been correlated with the mRNA expression-subtypes as well as clinical characteristics. Frequent mutation of PIK3CA (45%), MAP3K1, GATA3, TP53, CDH1, and MAP2K4 has been observed in luminal A subtype breast cancers, whereas the luminal B subtype most frequently showed mutations in TP53 and PIK3CA (29% each). The highest occurrence (80% samples) of TP53 mutations was observed in the basal-like cancers. The HER2-positive subtype showed frequent HER2 amplification (80%) and a high rate of mutations of TP53 (72%) and PIK3CA (39%) ( Fig. 22.8 ).

Consensus of clusters representing the breast cancer subtypes using different genomic and proteomic data. (A) Sample consensus results in four major clusters (shown as blue biclusters) for breast tumors (samples, n = 348) by unsupervised consensus clustering of the data from multiple platforms. (B) Concordance patterns of miRNA expression, DNA methylation, copy number (CN), PAM50 mRNA expression, and RPPA (protein) expression is displayed in red bar representing the cluster membership. (C) Molecular and clinical features that show significant association (by chi-square p values) with the clusters representing the molecular subtypes.

(Reproduced with permission from the Cancer Genome Atlas Network. Comprehensive molecular portraits of human breast tumors. Nature. 2012;490:61-70.)