Metastatic Breast Cancer: Cytology Diagnosis with Implications for Treatment

: Breast cancer is among the most frequent malignancies in women worldwide. While early detection and effective treatment provide many women with a cure and prevent their cancer from spreading, metastases to distant sites still occur in around 20% of women suffering from breast cancer. These relapses occur in many forms and locations and are as varied as the primary breast tumors. Metastatic spread makes a cancer incurable and potentially lethal, but new, targeted treatments can offer control of the cancer cells if the features of new targets are unlocked by advanced diagnostic testing. The article offers an overview of the pathomechanisms of metastatic progression and describes the types of metastases, such as hormone-receptor-positive and -negative breast cancers, and HER2 -overexpressing or triple-negative types. Once distant metastatic spread occurs, cytology allows a precise diagnosis to conﬁrm the breast origin. Other molecular targets include ESR1 and PIK3CA mutations, MSI , NTRK fusion, PD-L1 expression and others, which can be obtained also from cytology material and used to determine eligibility for emerging targeted therapeutic options. We outline the diagnostic features of metastatic breast cancer in cytology samples, together with validated and emergent biomarkers that may provide new, targeted treatment options


Introduction and Overview
More than 2.3 million patients with breast cancer are newly diagnosed globally, and an estimated 685,000 women died of the disease in 2020 [1,2]. Approximately 3.5 million women in the United States are in treatment or living with a history of breast cancer. While the overall survival rate is favorable with early detection and advanced treatment options, 15-25% of women eventually develop distant metastatic disease, and around 40,000 women die from breast cancer each year in the U.S. [1,2].
These overall numbers of breast cancer incidence, obtained through public databases, include a portion of around 25% that are ductal carcinoma in situ, without demonstrating stromal invasion and (theoretically) with no possibility of metastasis. The term "breast cancer" in this article generally refers to invasive breast cancer [3,4].
With increased awareness and breast cancer screening programs in place for decades, only approximately 5% of western women present with breast cancer at an advanced stage-that is, with distant metastatic disease (or stage IV cancer) [5]. Regional lymph nodes are usually the first site of cancer spread outside the breast; these lymph node metastases, typically in the axillary lymph nodes, are considered a local disease for staging and treatment purposes. In this article, emphasis is placed on distant metastatic spread to other organs away from the breast.
Globally, there are significant regional differences in morbidity and mortality. While the rate among women in low-resource countries might be lower than in the western world, their likelihood of dying from this disease is also higher [2]. Estimates prognosticate a steep increase in both frequency and risk of death in the next 20 years. With this worrisome perspective of a higher disease burden on a global scale, effective methods for diagnosis in every layer of the medical resource spectrum are necessary [1,2,6].
In most clinical studies of primary breast cancer treatment, distant metastatic disease represents an endpoint and treatment failure. Although much progress has been made with the early detection of breast cancer, advanced hormonal therapy, chemotherapy regimens and targeted treatments to improve the prognosis of breast cancer patients, only modest improvements are found once the cancer shows a distant metastatic relapse. These patients are excluded from primary treatment schemes, and their psychological, social and economic needs are often not realized [6]. New treatment options are needed to improve the outlook for patients with metastatic breast cancer [7,8].
With the majority of breast cancers being hormone-receptor-positive, hormonal treatment has played an important role for many years [9,10]. As it became better known that hormonally responsive breast cancers may show a pattern of late relapses beyond the first five years after the primary cancer diagnosis and initial treatment, new anti-hormonal agents were discovered and treatment regimens refined [11,12]. These may include agents such as aromatase inhibitors, which subsequently may cause resistance through mutations, or selective estrogen receptor degraders, with which treatment can be continued and also escalated in the metastatic setting [13][14][15].
The discovery of human epidermal growth factor receptor 2 (HER2/c-erbB2) gene amplification and protein overexpression as a predictive factor and treatment target for humanized monoclonal antibodies was a breakthrough in breast cancer treatment. Currently, several treatment regimens target HER2-positive tumors in the neoadjuvant and adjuvant setting and have greatly improved the survival of patients with this type of breast cancer [16,17].
In addition to established predictive and prognostic markers for primary breast cancer, new molecular targets are being identified, adding new treatment options and improving the diagnosis for breast cancer patients with metastatic disease [7,10,18]. One such example is immune checkpoint or cyclin dependent kinase (CDK 4/6) inhibitors, which may allow further treatment options with agents such as pembrolizumab; primarily, this applies to the triple-negative subset of breast cancers that are found to express programmed death ligand (PD-L1) on immune cells in the tumor [19,20]. A new and promising agent to expand the treatment options for triple-negative breast cancers (TNBC), especially in BRCA 1/2 gene mutation carriers, is poly-ADP ribose or PARP inhibitors [21].
Genomic testing of metastatic lesions is necessary to identify actionable targets, and tissue sampling is at the core of this diagnostic process to determine the presence of therapeutic targets [23]. Cytology is often the method of choice to sample metastatic lesions because of the location, accessibility and choice of a minimally invasive sampling technique [24][25][26]. The role of liquid-based biopsy from the peripheral blood or other body fluids is considered later in this article.

How Does Metastatic Spread Occur?
Most of the diagnostic parameters and initial treatments in early breast cancer focus on the primary tumor. Features that are associated with spread outside of the breast are lymphatic vessel invasion and lymph node metastases. Involvement of axillary lymph nodes, which provide the first site of lymphatic drainage to the breast, is still indicative of local disease but indicates a higher risk of distant relapse [5]. Regional lymph node involvement is unfavorably correlated with diagnosis as a staging parameter.
Pathogenetically, several models help to explain how tumor cells disseminate into other sites of the body: tumor cells may gain the capacity to metastasize within the primary tumor in a linear progression model to later spread through the genetic evolution of metastasis-capable founder cells. Another possibility assumes the early dissemination of tumor cells with the acquisition of new mutations at the new site that would allow this. In a historical hypothesis formulated by Stephen Paget in 1889, cancer cells ("seed") may settle preferentially in the selected microenvironment ("soil").
Following this model, tumor cells encounter a pre-metastatic niche, which they colonize; after various lengths of dormancy, and mediated by tumor-initiated soluble factors and by possibly creating an immune-suppressed field, these tumor cells eventually cause a metastatic outgrowth at the new site. In the reality of metastatic spread, various pathways might be used by traveling tumor cells to remain dormant or establish new growth sites [27]. Eventually, a metastatic cascade gives rise to innumerable foci of tumor cells that become resistant to treatment and lead to the patient's demise. The genetic underpinnings of why, when and where tumor cells precisely cause distant spread are the subject of ongoing investigation [28].
As to the preferred sites of breast cancer metastases, the molecular subgroups based on hormone receptors and HER2 expression are associated with certain metastatic patterns and tropisms. Estrogen-receptor-positive types (luminal A/B type carcinomas following the intrinsic subtype model) more frequently cause bone metastases. As a morphological subtype, lobular carcinomas, which are mostly ER-positive, have different metastatic patterns to breast cancers of a ductal/no special type, with more frequent visceral, serosal and gynecological metastases (40). Triple-negative (basal-like) and HER2-positive carcinomas were shown to cause brain metastases more frequently, posing particular challenges to treatment [22,29,30].

Diagnosis of Breast Cancer Metastases
Different types of breast cancer based on morphology, histologic grade, hormone receptor and HER2 expression show different patterns in relapse and metastatic behavior. While some changes can occur in morphology and immunohistochemical profile, most metastatic lesions resemble the primary tumor, and key morphologic features in metastases often lead to a further diagnostic investigation [31,32]. For the diagnosis of primary breast cancers and their respective metastases, the current classification for tumor diagnosis is followed [33].
Morphological comparison with the primary may be most helpful when metastasis from a known breast primary is suspected. In conjunction with clinical correlation, a selected immunohistochemical profile of the metastatic focus will help to narrow down the differential diagnosis. A panel of immunohistochemical markers, including estrogen and progesterone receptors, cytokeratins and GATA3, may allow us to confirm the primary site in the breast. Other successfully employed markers are mammaglobin and GCDFP-15, but now often the nuclear marker GATA3 is used. GATA3 is not specific to the breast but, in the appropriate clinical context, has proven very useful as a diagnostic marker for metastatic breast cancer [34]. In combination with SOX10 (SRY-box transcription factor 10), it has been shown to be helpful to diagnose triple-negative breast cancer [35]. More recently, the novel marker TRPS-1 (trichorhinophalangeal syndrome type 1) was used to confirm the breast origin [36]. The hormone receptor and HER2 profile of a known breast primary can be utilized to confirm the metastatic site. It is important to note at this point is that most metastatic foci resemble the primary tumor immunomorphologically, but exceptions and "switches" from hormone-receptor-or HER2-positive to HER2-negative, and vice versa, may occur in a considerable proportion of cases [22,26].
The breast ©tself is not a frequent site of metastasis. Nonetheless, some tumors can metastasize to the breast, such as disseminated lymphomas, melanomas and small-cell lung cancers. However, this topic is beyond the scope of this discussion [31].
Estrogen-receptor-positive breast cancers can recur late (after 5, 10 or even 20 years), often causing bone metastases. Other visceral sites, such as multiple liver or lung metastasis, are seen as part of the metastatic cascade.
Triple-negative breast cancers tend to be of high histologic grade and typically recur in the first three to five years after the primary diagnosis. Both triple-negative and HER2positive breast cancers appear to cause brain metastases more frequently than hormonereceptor-positive tumors [29].
With the various organs involved, cytology is often the most accessible means of tissue sampling for breast cancer diagnosis. Depending on the practice patterns, predictive markers such as hormone receptors and HER2 studies may be determined on the metastatic lesions as a standard of care and following international practice guidelines [37,38]. In addition, molecular profiles are often requested to determine patients eligible for novel targeted treatments such as checkpoint inhibitors, PIK3CA inhibitors or immunotherapy [22,39]. Individual protocols vary and depend on departmental preferences and practice patterns: some departments routinely apply molecular tests on liquid cytology remnants, while other use cell block material or send samples for molecular testing.
The subsequent paragraphs illustrate cytological samples of breast cancer subtypes from metastatic lesions, including examples of breast cancer of ductal/no special type, special types such as lobular carcinoma or others and the groups based on hormone receptor and HER2 expression.

Metastatic Ductal Breast Cancer (No Special Type)
Invasive ductal carcinomas/carcinoma of no special type constitute approximately 80% of breast carcinomas ( Figure 1) and range in histologic grade from well (grade 1) to moderately (grade 2) to poorly differentiated (grade 3). Usually, metastatic lesions are not independently graded. Nuclear pleomorphism, necrosis or other features can of course be used to describe, for example, a poorly differentiated tumor on cytology.

Metastatic Lobular Carcinoma
Lobular carcinomas are typically low-grade (grade 2 or 1) and represent approximately 15% of breast carcinomas ( Figure 2). Metastatic lobular carcinoma shows more frequent involvement of visceral sites, the intestinal or gynecological tract or pleural membranes [40,41]. Characteristically, in effusions, there are medium-sized dyscohesive tumor cells against a background of reactive mesothelial cells.

Metastatic Breast Cancer, Other Special Types
Invasive breast cancers occur in several other special types which can be distinguished by morphological features; only a few can be shown here as examples (33). Some of these special types are associated with a more favorable prognosis, for example mucinous carcinomas of the breast (Figure 3), medullary carcinomas or adenoid cystic carcinomas. On the other hand, matrix producing breast cancers are typically high-grade breast cancers with an aggressive behavior. Micropapillary carcinomas are characterized by a high rate of lymph node metastasis and poor prognosis.

ER-Positive Metastatic Breast Cancer
Most breast carcinomas are positive for estrogen receptors (Figure 4). Hormone receptors are reported in a semi-quantitative or quantitative way following international guidelines [37]. While it is feasible to perform immunocytochemical stains for nuclear receptors on cytology preparations (direct smears as well as liquid-based preparations), staining protocols need to be separately validated for cytology. Cell block preparations can be prepared following several techniques and are feasible for immunohistochemical as well as for molecular testing [24,25,42]. . Of note is that the cell block material here is markedly cellular; the cytology sample appears suitable for diagnostic testing as well as advanced molecular testing.

Metastatic Lobular Carcinoma
Lobular carcinomas are typically low-grade (grade 2 or 1) and represent approximately 15% of breast carcinomas ( Figure 2). Metastatic lobular carcinoma shows more frequent involvement of visceral sites, the intestinal or gynecological tract or pleural membranes [40,41]. Characteristically, in effusions, there are medium-sized dyscohesive tumor cells against a background of reactive mesothelial cells.  shows metastatic lobular carcinoma that is positive for estrogen receptor protein (immunohistochemical stain).

Metastatic Breast Cancer, Other Special Types
Invasive breast cancers occur in several other special types which can be distinguished by morphological features; only a few can be shown here as examples (33). Some of these special types are associated with a more favorable prognosis, for example mucinous carcinomas of the breast (Figure 3), medullary carcinomas or adenoid cystic carcinomas. On the other hand, matrix producing breast cancers are typically high-grade breast cancers with an aggressive behavior. Micropapillary carcinomas are characterized by a high rate of lymph node metastasis and poor prognosis.

ER-Positive Metastatic Breast Cancer
Most breast carcinomas are positive for estrogen receptors (Figure 4). Hormone receptors are reported in a semi-quantitative or quantitative way following international guidelines [37]. While it is feasible to perform immunocytochemical stains for nuclear receptors on cytology preparations (direct smears as well as liquid-based preparations), staining protocols need to be separately validated for cytology. Cell block preparations can be prepared following several techniques and are feasible for immunohistochemical as well as for molecular testing [24,25,42].   Figure 4a) shows strong staining for estrogen receptor protein by immunohistochemical stain.

HER2-Positive Metastatic Breast Cancer
Human epidermal growth factor receptor (HER2) positive tumors represent around 10-20% of breast cancers ( Figure 5). The scoring of tumor tissue is performed by immunohistochemical and/or in situ hybridization techniques, following international guidelines [38]. Fluorescent in situ hybridization can also be achieved on cytology preparations with adapted protocols. Often preferred, however, is HER2 testing on formalin-fixed, paraffinembedded cell block and core biopsy specimens, as routine protocols for both immunohistochemistry and in situ hybridization procedures can be followed, as per guidelines.

HER2-Positive Metastatic Breast Cancer
Human epidermal growth factor receptor (HER2) positive tumors represent around 10-20% of breast cancers ( Figure 5). The scoring of tumor tissue is performed by immunohistochemical and/or in situ hybridization techniques, following international guidelines [38]. Fluorescent in situ hybridization can also be achieved on cytology preparations with adapted protocols. Often preferred, however, is HER2 testing on formalin-fixed, paraffin-embedded cell block and core biopsy specimens, as routine protocols for both immunohistochemistry and in situ hybridization procedures can be followed, as per guidelines.

Triple-Negative Metastatic Breast Cancer
Triple-negative breast cancers comprise around 10% of overall breast cancers, which are often of ductal type (no special type) and poorly differentiated. Some special-type breast cancers such as matrix-producing breast cancers, spindle cell carcinomas or adenoid cystic carcinoma (33). As therapeutic targets such as hormonal treatment or HER2 overexpression do not apply, PD-L1 can be used for possible immune checkpoint inhibitor treatment. PD-L1 testing is performed on primary tumors and is feasible also on cytology samples, to test for immunotherapy eligibility [26].

Molecular Profiles in Cytology Samples and Liquid Biopsies
Circulating tumor cells or cell-free DNA (ctDNA) are in various stages of development to follow treatment responses and for clinical applications [43]. For example, estrogen receptor mutations (ESR1) following hormonal treatment with aromatase inhibitors or PIK3CA mutations have been tested by blood samples to follow treatment [10,14,44,45]. Another active area of investigation is liquid biopsies, which are proposed to be used for staging and molecular characterization in blood and in certain body compartments, such as cerebrospinal fluid disease (although, for many places, these methods are still beyond the standard diagnostic repertory) [46][47][48][49][50].

Novel Therapeutic Targets in Metastatic Breast Cancer
Multiple novel mutations have been validated and are promising for future personalized, targeted treatment. The phosphatidylinositol 3-kinase gene (PI3KCA) is involved in the cell cycle and cell proliferation, encoding for the class I catalytic isoform p110α. It is mutated in around 40% of hormone-receptor (HR) positive breast cancer cases [51,52]. Microsatellite instability (MSI) occurs in less than 1% of cases. Neurotrophin receptor tyrosine kinase (NTRK) translocation is found exclusively in secretory carcinoma [53]. The activating mutation of the estrogen receptor α (ESR1) gene results in the constitutional activity of the estrogen receptor independently of the ligand; it is commonly found as a result of aromatase inhibitor in HR+/HER2− metastatic breast cancer [10,54,55]. Clinical trials already provide levels of evidence, and molecular testing in tissue and liquid biopsies is being established.
Cyclin-dependent kinase 4 and 6 (CDK4/6) inhibitors, palbociclib, ribociclib and abemaciclib, are already being used in hormone-receptor-positive metastatic breast cancer [56]. They interrupt cell cycle progression, leading to the inhibition of tumor growth. Adding them to endocrine therapy has shown improved progression-free survival with minimal toxicity. The agent pembrolizumab, also an immune checkpoint inhibitor, is being used for triple-negative breast cancers, for which there are often few treatment options [20]. For metastatic triple-negative breast cancers in BRCA 1/2 mutation carriers, a patient population with often aggressive tumors and few treatment options, PARP inhibitors are now used, with some success [21].
HER2-positive breast cancers are defined by HER2 overexpression and/or amplification, which drives the growth of cancer cells. Therapies that target HER2 include humanized monoclonal antibodies such as trastuzumab, pertuzumab or other conjugates that are used in neoadjuvant and adjuvant treatment settings [16,17]. Adding these agents to the neoadjuvant chemotherapy reduces the tumor size and increases the breast-conserving rate.

Conclusions and Outlook
This article describes the current practice in sampling metastatic breast cancer lesions and determining their molecular profiles to better apply targeted treatments. This ranges from routine ancillary breast markers such as estrogen receptor proteins to molecular profiles to assess the response and resistance to novel hormonal treatments. HER2-targeted treatments have expanded from a single anti-HER2 agent to a range of treatment possibilities, also in the metastatic setting. Other actionable treatments include PIK3CA mutations and immune checkpoint inhibitors for hormone-receptor-positive and triple-negative breast cancer subtypes. Current oncology guidelines include circulating tumor DNA as a novel technology to determine the mutational profiles of tumor cells in metastatic breast cancer.