This article is mentioned in:

Introduction

Neuroendocrine neoplasms (NENs) can originate from several regions of the body, but are most commonly observed in the gastrointestinal tract, and, to a lesser extent, in the lungs, breast, larynx, prostate, bladder, ovaries and cervix. Included among these, primary NEN of the breast (PNENB) is a rare subtype of breast cancer, accounting for <1% of all breast tumors (1). Current understanding of this disease is limited, as researchers have not detected neuroendocrine cells in normal breast tissue, and it has been suggested that PNENB may be derived from an early-stage differentiation of breast cancer tumor cells (2). The definition of PNENB as well as the diagnostic criteria have been under debate since this type of cancer was first reported (1,3,4). The different criteria used for the definition of cases, due to the overlap in diagnostic features between PNENB and invasive breast cancer (IBC) with neuroendocrine differentiation (5), have led to a lack of comparability between studies on the treatment and prognosis of the disease. In the present article, a case of a patient with PNENB is described and the current body of literature on PNENB is reviewed.

Case report

A 70-year-old woman was admitted to the Affiliated Hospital of Inner Mongolia Medical University (Hohhot, China) on May 22, 2014 with a left breast mass that had been detected >1 month earlier. On examination, a medium-sized (~15×20 mm) mass with unclear borders was palpable in the upper outer quadrant of the left breast. Enlarged lymph nodes were also palpable in the left axilla. An ultrasound on May 24, 2014 showed a 15×20-mm irregular hypoechoic solid mass, identified at 1 o'clock in the left breast, and two transverse fingers from the nipple, 7.6 mm below the surface of the skin, with no clear borders. Striated blood flow signals were observed at the margins of the lesion by color Doppler flow imaging, which was assessed as Breast Imaging Reporting and Data System (BI-RADS) category 4b (6) (Fig. 1). A mammography on May 25, 2014 showed a nodular hyperdense shadow in the upper quadrant of the left outer breast, ~15×20 mm in size, which was assessed as BI-RADS 4a (Fig. 2). Subsequently, left breast lumpectomy followed by intraoperative frozen section analysis was performed. The gross pathological examination revealed a piece of light-yellow soft tissue of 5×4×3 cm, with a bleeding gray-white mass inside. The maximum diameter of the mass was 2 cm. Hematoxylin and eosin staining and immunohistochemical (IHC) staining were performed to further define the tumor. For hematoxylin and eosin staining, tissues were fixed with 10% neutral formalin at 24°C for 20 h. Tissue sections were 4 µm thick. Tissues were immersed in xylene for 5 min, and 75, 85 and 95% gradient ethanol for 1 min. Tissues were stained with hematoxylin for 6 min at 37°C, differentiated with 0.5% hydrochloric acid ethanol for 10 sec, incubated with 0.2% ammonia water for 40 sec and stained with eosin staining solution for 3 min at 60°C. Staining was observed under a light microscope. For IHC staining, tissues were fixed with 10% neutral formalin at 24°C for 20 h and embedded in paraffin. The tissue sections were 4 µm thick. The tissue sections were blocked with 15% blocking serum (Fish Serum Blocking Buffer; Thermo Fisher Scientific, Inc.) for 20 min at 37°C. Tissue sections were incubated with Cyclin D1 Recombinant Rabbit Monoclonal Antibody (SP4; dilution, 1:100; cat. no. MA5-14512; Cusabio Technology, LLC) for 2 h at 37°C, followed by incubation with Goat anti-Human IgG (H+L) Cross-Adsorbed Secondary Antibody (conjugate, Alexa Fluor™ 488; dilution, 1:100; cat. no. A-11013; Cusabio Technology, LLC) for 2 h at 37°C and sealing with neutral resin. Tissue sections were observed under a light microscope. A simple left mastectomy and anterior lymph node biopsy were also performed. The postoperative pathological results suggested that the mass was a low-grade neuroendocrine tumor (NET) with negative margins.

Figure 1. Ultrasound image of the tumor. A 15×20 mm irregular hypoechoic solid mass is visible (arrows), two transverse fingers from the nipple and 7.6 mm below the surface. This was assessed as Breast Imaging Reporting and Data System 4b at 1 o'clock in the left breast.

Figure 2. Mammogram of the tumor. A 15×20 mm nodular hyperdense shadow was observed, which was assessed as Breast Imaging Reporting and Data System 4a.

Gross pathological examination confirmed that the mass had a maximum diameter of ~2 cm. Hematoxylin and eosin staining showed that it comprised tumor cells that were uniform in size and rounded, with well-defined nuclei (Fig. 3A) and CD56 negative (Fig. 3B). IHC staining for chromogranin A (CgA) (Fig. 3C) and synaptophysin (Syn) (Fig. 3D) was positive, and detected in >50% of the tissue. Postoperative pathological diagnosis suggested a low-grade NET of the left breast. IHC staining results also revealed the presence of estrogen receptors (ERs, >90%; Fig. 4A), progesterone receptors (>90%; Fig. 4B) and Ki67 (~10%; data not shown), while IHC staining for human epidermal growth factor receptor 2 (HER-2) was negative (Fig. 4C).

Figure 3. Pathology and immunohistochemistry of neuroendocrine markers. (A) Hematoxylin and eosin staining of the tumor tissue. The tumor cell sizes were consistent, and the cells were circular, with the nucleolus clearly visible (magnification, ×400). Immunohistochemical staining of the tumor tisse showed that the tumor was (B) CD56 negative, (C) chromogranin A positive and (D) synaptophysin positive (magnification, ×200).

Figure 4. Immunohistochemistry of hormone receptors and HER-2 in the tumor. Immunohistochemical staining of (A) estrogen receptors (magnification, ×100) and (B) progesterone receptors was positive (magnification, ×100), while that of (C) HER-2 was negative (magnification, ×400). HER-2, human epidermal growth factor receptor 2.

Because anterior lymph node metastasis was not found, the patient did not receive any chemotherapy or other treatment after surgery. The patient has been followed up for 12 months, with a monthly telephone follow-up, with no axillary lymph node metastases, and no tumors at other sites have been detected.

Discussion

PNENB can be distinguished from other types of breast tumors based on morphological features as well as neuroendocrine markers. The neuroendocrine differentiation of breast cancer was first observed in 1963 by Feyrter and Hartmann (7) in mucinous carcinoma of the breast. This was followed by the discovery of breast tumors that were morphologically similar to carcinoid tumors of other organs by Cubilla and Woodruff (8) in 1977, who named them primary carcinoid tumors of the breast. In 1985, Bussolati et al (9) detected CgA in breast tissue, which provided evidence that some cells in the breast can exhibit neuroendocrine features. In 2002, Sapino and Bussolati (10) proposed the first diagnostic criteria for PNENB. In 2003, the World Health Organization (WHO) officially recognized PNENB as a distinct type of breast cancer, and defined PNENB as a tumor of epithelial origin (11). Following the diagnostic criteria of Sapino and Bussolati, breast tumors with morphological features similar to those of NENs originating in the lungs and gastrointestinal system in which >50% of the tumor cells expressed neuroendocrine markers were referred to as PNENB, with CgA and Syn considered as the most sensitive and specific neuroendocrine markers (11). These cancers were subdivided into large-cell neuroendocrine carcinomas (LCNEC), small-cell neuroendocrine carcinomas (SCNEC), and solid carcinomas based on their morphological features. In 2012, the WHO Classification of Breast Tumors defined PNENB as a carcinoma with morphological features similar to those of NEN originating in the lungs and gastrointestinal system, irrespective of the percentage of tumor cells expressing neuroendocrine biomarkers. In addition, it suggested that based on morphology, PNENB can be classified into three subgroups: Well-differentiated NET of the breast (NETB), poorly differentiated neuroendocrine carcinomas of the breast (NECB)/SCNEC, and IBCs with neuroendocrine differentiation, including solid papillary and mucinous carcinomas (12).

The 5th edition of the WHO Classification of Breast Tumors, published in 2019, uses the same classification criteria for PNENB, gastroenterology-pancreas NETs and lung NETs. It also divides breast tumors in which >90% of tumor cells have neuroendocrine features into two main categories based on the degree of differentiation: Well-differentiated NETB and poorly differentiated NECB, including SCNEC and LCNEC. LCNEC was added to the classification, and special tissue-type breast cancers, including solid breast cancer and mucinous carcinoma, were removed from the PNENB category (13). Breast tumors in which 10–90% of the cells exhibit neuroendocrine differentiation are referred to as mixed IBCs or NETs/neuroendocrine carcinomas (NECs), and tumors in which <10% of the cells exhibit neuroendocrine differentiation are classified as either IBC of no special type (IBC-NST) or IBCs of other special types. The Nottingham grading system can be used to grade PNENB tumors based on their specific characteristics, independent of the parameters used in neuroendocrine tumors originating from other parts of the body (13). It is noteworthy that although NETB has been classified in the framework of NETs, it has no defined morphological features, and its identification among other breast tumors relies on the staining of neuroendocrine markers, including Syn and CgA, which are also expressed in non-PNENB tumors. Based on a review of previous reports, the molecular and genetic characteristics of NETB are not similar to those of well-differentiated NENs from other sites, but resemble those of luminal A breast cancer (14). In contrast to NETB, NECB represent a well-defined entity, showing morphological and clinical analogies with pulmonary and extra-pulmonary NECs (15).

PNENB occurs predominantly in >60-year-old postmenopausal women, is very rare in men, and does not differ markedly from other types of breast cancer in terms of clinical presentation (16). The majority of patients present with isolated breast lumps as the primary symptom, and may also have carcinoid syndrome and clinical manifestations associated with ectopic hormone secretion, such as paraneoplastic thrombocytosis and hyperprolactinemia, as well as rare elevations of α-fetoprotein (17–19). NECB is a high-grade tumor, most often detected at an advanced stage, and metastasized at the time of initial diagnosis (20). The imaging features of PNENB are not specific. Previous studies have reported that mammography reveals round, ovoid, lobulated or irregularly shaped masses, most of which have poorly defined borders with surrounding tissues and burr-free margins, whereas IBC-NST tends to have an irregular shape with burr-like margins, and is associated with microcalcifications (21,22). The majority of cases present on breast ultrasound as an irregular mass that is hypoechoic or poorly defined, with absent or enhanced posterior acoustic features (23). Park et al (22) suggested that neuroendocrine differentiation in tumors may influence the imaging presentation, and that the absence of burr edges on mammograms and of posterior echo enhancement on ultrasound images may be indicative of a tumor with neuroendocrine features. When morphological features of PNENB are not evident, imaging is necessary to rule out metastatic malignancy in the breast since ≥97% of NECs originate from the gastrointestinal tract or lungs (24).

The absence of specific features in the routine examination of PNENB underscores the importance of accurately diagnosing this condition, as the diagnosis directly influences the subsequent treatment and prognosis. Diagnosis relies on morphological features, IHC staining and genetic analyses to determine the type of cancer and identify therapeutic targets. Morphologically, NETB consists of dense nests of cells and cellular trabeculae, and the cells may show spindle, plasma cell-like or polygonal features, with eosinophilic or thylakoid cytoplasm separated by a fibrovascular stroma, with little morphological resemblance to NETs at other sites (14,25). SCNEC tumors exhibit infiltrative growth, darkly stained nuclei, inconspicuous nucleoli, high nucleoplasmic ratios, sparse cytoplasm and ill-defined cytoplasmic boundaries, while LCNEC tumors have darkly stained nuclei, diverse nuclear morphology and more cytoplasm than SCNEC tumors (26). The IHC markers commonly used to identify the presence of neuroendocrine differentiation are Syn, CgA, neuron-specific enolase (NSE) and CD56. Syn and CgA are diffusely positive in NECB and are specific markers, whereas NSE and CD56 are less sensitive and specific. The transcription factor insulinoma-associated protein 1 (INSM1) is a relatively novel marker that is detected in the cytosol and differs from other neuroendocrine markers, which are detected in the cytoplasm; however, its sensitivity and specificity are not significantly different from those of Syn and CgA (27). Zhong et al (28) investigated the expressed pattern of ‘second generation’ neuroendocrine markers INSM1, achaete-scute homolog 1 (ASCL1) and POU class 2 homeobox 3 (POU2F3) in breast cancers with neuroendocrine morphology. The study found that INSM1 was more specific than Syn and more sensitive and specific than CgA, but the positivity rates of ASCL1 (4/35) and POU2F3 (1/35) were too low to support routine testing of their expression. Recent studies have demonstrated that syntaxin-1 (STX1) is a sensitive and specific marker of neuroendocrine cells (29–31). For example, in a study of NECB in which STX1 and INSM1 were compared with neuroendocrine markers such as Syn, CgA and CD56, a sensitivity of 84.7% (50/59) and specificity of 98.1% were recorded for STX1, suggesting that STX1 has potential as an NE marker (31). However, to the best of our knowledge, no studies have assessed the expression of these newer NE markers in the rare PNENB subtype. Juhlin et al (32) suggested that NE markers such as INSM1, secretagogin, and ISL LIM homeobox 1 are tissue-specific, which may aid in the detection of the primary tumor site. These new NE markers may provide novel avenues for research into the clinical implications of NE differentiation in PNENB. Another study found that the expression of transcription factors GATA-binding protein 3 (GATA3) and gross cystic disease fluid protein 15 was positive in tumors originating from the breast and negative in metastatic tumors, while the expression of caudal-type homeobox 2 and thyroid transcription factor-1 (TTF-1) was suggestive of metastatic tumors (33). Notably, TTF-1 also exhibited strong positivity in high-grade NECB. In addition, the expression of hormone receptors (HRs) has been shown to contribute to the diagnosis of PNENB. Specifically, strongly positive HR expression and negative HER-2 expression is observed in the majority of PNENBs, which, according to their molecular subtypes, exhibit tubulointerstitial characteristics (34,35). However, the proportion of cases of PNENB that are negative for HR and HER-2 expression is greater than that for invasive ductal carcinoma (IDC). Several studies have reported ER positivity in NEN of the lung, pancreas, small intestine and ovary (36–38). Although ER positivity is not specific, it can be used as a target for therapeutic agents. For example, Zhang et al (39) found that the tumor in a patient with poorly differentiated NEC of the breast shrank by 78.87% after 3 months of neoadjuvant endocrine therapy.

PNENB is genetically heterogeneous. It has a different mutational spectrum from other ER-positive and HER 2-negative IBC-NSTs. In one study, lysine-specific methyltransferase 2C was found to be the most commonly mutated gene in PNENB (3/17, 17.6%) and predicted to serve as a driver gene that may play an important role in the neuroendocrine differentiation of breast cancer (40). PNENB was also indicated to have common copy number variants (CNVs) such as 8q, 11q and 17q amplification, with variants in 8q being high-frequency CNVs and potential targets for tumor therapy (40). Wei et al (40) found that the mutation rates of members of the PI3K and MAPK signaling pathway in NETB were higher than those in NECB, suggesting that breast NET and NEC may have different genetic phenotypes and prognoses, and thus require different therapeutic strategies. A study by Marchiò et al (41) showed that the most commonly mutated genes were GATA3, forkhead box protein A1 (FOXA1), T-box transcription factor 3 (TBX3) and AT-rich interactive domain-containing protein 1A in 3/18 cases (17%), and PI3K catalytic subunit α (PIK3CA), AKT1 and cadherin 1 in 2/18 cases (11%). In addition, the study indicated that NECB is characterized by a lower frequency of tumor protein P53 (TP53) and PIK3CA mutations, and enrichment of FOXA1 and TBX3 mutations compared with common forms of ductal breast cancer. In another study, co-mutations in TP53 and RB transcriptional corepressor 1 (RB1) were found in 6/7 cases of SCNEC (86%), 2/2 cases ambiguous for small cell vs. large cell morphology (100%) and 2/4 LCNEC cases (50%). In addition, one case of wild-type TP53/RB1 SCNEC had other p53 pathway aberrations, specifically amplification of mouse double minute 2 and 4 homologs, and was found to be RB-negative by IHC analysis (42). It was suggested that co-inactivation of TP53 and RB1 may be important for the pathogenesis of the NEC phenotype in the mammary gland. Vranic et al (43) detected the expression of folate receptor 1 (FOLR1), trimethylated Lys-36 of histone 3 (H3K36me3) and tumor-associated calcium signal transducer 2 (TROP-2) in a subset of NECBs; however, they did not detect any biomarkers indicating that immune checkpoint inhibitors could be used, including programmed death ligand 1 expression, tumor mutational burden and microsatellite instability. However, based on the detection of TROP-2, FOLR1 and H3K36Me3, they suggested that antibody-drug conjugates or inhibitors of these proteins may be potential therapeutic options for the treatment of NECB. Mutations in the epidermal growth factor receptor (EGFR) gene serve as a suitable target for most cancers. In one study a rare EGFR p.Thr790Met (T790M) mutation was found in a patient with poorly differentiated NEBC, which was suggested to be responsible for the resistance of the tumor to tyrosine kinase receptor inhibitors (44).

Due to the rarity of PNENB, no standard treatment guideline currently exists. Surgery is the first-line treatment option, as it is for other IBC-NSTs (16,45). Other treatment modalities include adjuvant radiotherapy, chemotherapy, endocrine therapy, targeted therapy and neoadjuvant chemotherapy (16,46). NECB has been typically treated with chemotherapy in previous cases in the relevant literature due to its high histological grade and metastasis at the time of detection. Adjuvant chemotherapy can be used in patients at a high risk of recurrence and neoadjuvant chemotherapy can be used in patients with preoperative downstaging or locally advanced disease before surgery (16,45,47). Platinum-based drugs and etoposide chemotherapeutic agents are the most widely used chemotherapeutic agents for small cell lung cancer and extrapulmonary poorly differentiated LCNEC or SCNEC, while combinations of anthracycline- and paclitaxel-containing chemotherapeutic agents are generally used for other types of PNENB (48). Somatostatin (SS) therapy is mediated through the SS receptor (SSTR) on the surface of tumor cells. SSTR2A, SSTR2B, SSTR3 and SSTR5 have been found to be expressed in breast NET and NEC (49,50), of which SSTR2A and SSTR5 exhibit a high affinity for growth inhibitor analogs. Positron emission tomography using gallium-68-labelled SS analogs has potential for use in the diagnosis of PNENB, while SS analogs combined with a β-emitting lutetium-177 radioisotope are now approved for use in the treatment of NET (50), and may also serve as a therapeutic option in the management of PNENB.

To date, the prognosis of PNENB is unclear, which may be attributed to the different diagnostic criteria used for the inclusion of cases in different studies and the small sample sizes. However, most studies have shown that PNENB has a poorer prognosis than NST-IDC (51,52). Tumor staging and histological grading remain the primary prognostic factors. An analysis of PNENB using data from the Surveillance, Epidemiology, and End Results database indicated that age, marital status, the place of registration, surgery, American Joint Committee on Cancer staging and breast subtype are independent prognostic factors (53). Although race was found to be significantly associated with PNENB in univariate Cox and Kaplan-Meier analyses, no significant association was observed in multivariate Cox analysis, which may be attributed to the elimination of spurious associations when confounding variables were accounted for in the multivariate analysis (53). The analysis also showed that, compared with non-surgical treatment, surgical intervention for excision of the primary tumor or for non-cancer-related purposes was effective in improving prognosis. It was suggested that in the latter scenario, the surgery may improve the prognosis by improving the quality of life of the patient (53).

In conclusion, PNENBs are a rare, heterogeneous subtype of breast tumors, which are yet to be fully understood. Although previous studies have identified several driver genes and mutated genes with high mutation rates, the exact diagnosis continues to rely on the examination of traditional neuroendocrine markers, namely Syn and CgA. Moreover, there are no standard treatment guidelines for this disease, with surgery being the first-line approach, as it is for other IBC-NSTs.

Acknowledgements

Not applicable.

Funding

Funding: No funding was received.

Availability of data and materials

The data generated in the present study may be requested from the corresponding author.

Authors' contributions

HJ collected the clinical, imaging and pathological data of the patient, and wrote the manuscript. ML conceived and designed the study, and revised the manuscript. HJ and ML confirm the authenticity of all the raw data. Both authors read and approved the final version of the manuscript.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Written consent for publication was obtained from the patient. All identifying information has been removed or anonymized to ensure confidentiality.

Competing interests

The authors declare that they have no competing interests.

References