1 Introduction

Neuroendocrine neoplasms (NENs) are a heterogeneous group of epithelial neoplastic proliferations arising in a large number of body organs. Irrespective of their primary site and of their grade of differentiation, neoplastic cells share features of neural and endocrine differentiation including the presence of secretory granules, synaptic-like vesicles and the ability to produce amine and/or peptide hormonal products.

The high heterogeneity of biological and clinical features of NENs represents a challenge for oncologists and pathologists, and a correct diagnostic approach is crucial for the management of patients. Indeed, NENs encompass a wide spectrum of neoplasms ranging from “benign” or very indolent tumors to highly aggressive neuroendocrine carcinomas. Accordingly, the morphological features of these neoplastic proliferations are variable and must be carefully identified by pathologists in order to produce a correct and complete histopathological report, which is the starting point for the correct treatment and follow-up of each patient [1]. The clinicopathological features are related to the morphology and immunohistochemical profile, which also depend on the site of origin. Most cases arise in the bronchopulmonary and gastroenteropancreatic (GEP) systems, while neoplasms arising in other sites, including pituitary gland, parathyroid, thyroid, adrenal glands, and paraganglia are rarer. In addition to these sites, NENs may occasionally arise in other organs, including nasal and paranasal cavities, salivary glands, skin, and urogenital and gynecological organs, showing peculiar clinicopathological features. The systematic description of all of these is beyond the scope of the present chapter, which will discuss the most frequent neoplasms arising in the digestive and bronchopulmonary systems.

Both GEP and lung NENs are relatively rare tumors, although their incidence is steadily increasing. Poorly differentiated neuroendocrine carcinomas (NECs), mainly of the small cell type, are more frequent in the lung than well-differentiated tumors (carcinoids), whereas in the GEP system well-differentiated neuroendocrine tumors (NETs) occur more commonly than NECs [2]. Although the diagnosis of GEP and lung NENs is generally simple in routine practice, there are critical aspects that need to be taken into account during the diagnostic workup. The most challenging tasks include the difficulty in achieving the correct prognostic evaluation of NETs by recognizing their metastatic/aggressive potential and the identification of NECs. To this aim, a comprehensive morphological, immunohistochemical, and molecular investigation represents the most reliable tool for pathologists.

2 Nomenclature and Classification

2.1 Nomenclature

Although the terms used to define NENs change in relation to the site of insurgence (lung versus GEP system), the WHO classifications of lung and digestive tumors separate both lung and GEP NENs into two main groups with different morphological, clinical, and molecular features: well-differentiated and poorly differentiated neoplasms [3, 4]. This distinction is crucial for the different prognoses and therapeutic approaches of the two categories, which are distinct entities despite a similar neuroendocrine phenotype [5, 6].

The terminology used to define NENs has been a matter of debate in the last 100 years creating some confusion among clinicians and pathologists. At the beginning of the twentieth century, Siegfried Oberdorfer described a series of six tumors of the small intestine, which he called “carcinoid” (i.e., carcinoma-like) on the assumption that they were similar to intestinal adenocarcinomas but showed a more bland morphology and benign/indolent behavior [7]. This term, which became very popular among clinicians and pathologists, underlined that these neoplasms were different from the more frequent exocrine carcinomas of the same site, although they showed an epithelial morphology. The neuroendocrine nature of these lesions was unknown at that time and remained obscure for several years, until it was demonstrated by Claude L. Pierre Masson, who developed the argentaffin reaction and called this tumor type “argentaffinoma” [8]. Forty-five years later, the term “APUDoma” was introduced by Antony G. E. Pearse, who proposed the so-called “APUD concept”: all neuroendocrine cells of the body are capable of amine precursor uptake and decarboxylation (APUD cells) and represent the cells from which neuroendocrine tumors derive, including those of nonintestinal sites [9]. Nevertheless, the term carcinoid has remained very popular among pathologists and clinicians and it has been widely used to define the wide spectrum of NENs arising in digestive and nondigestive sites. However, since digestive NENs originate from several different neuroendocrine cell types showing different clinical, pathological, and molecular features, the term “carcinoid” has failed to adequately convey the variety of such tumors and, in 1995, it was replaced by the term “neuroendocrine tumor” by Capella and coworkers who proposed a prognostic classification of GEP NENs [10]. Since then the use of the term “carcinoid” has been discouraged in diagnostic practice in favor of “neuroendocrine tumor (NET)”, maintaining the term carcinoid solely in the context of the “carcinoid syndrome” [4, 11, 12]. Conversely, in the lung the term carcinoid has been retained over the years because the heterogeneous spectrum of NENs, including different immunophenotypes and clinical syndromes, is extremely reduced if compared with those of GEP NENs. Based on morphological features, lung carcinoids are divided into typical and atypical (see below).

As a general principle, lung carcinoids correspond to NETs as classified in the GEP system because of their well-differentiated morphology. It is worth noting that all these tumors are now considered to be malignant and potentially metastatic with different biological aggressiveness which mainly depends on proliferative indices, presence of specific endocrine syndromes and metastases at the time of diagnosis. Conversely, NECs of both small and large cell types are high grade, very aggressive cancers with an ominous prognosis without significant difference in outcome between lung and GEP cases. Moreover, the therapeutic strategy for NEC seems to be the same, independently of their origin [13].

2.2 Classification of Neuroendocrine Neoplasms

2.2.1 GEP System

Due to the need to develop a classification scheme able to separate aggressive NECs from more indolent NETs and neoplasms with different behavior among the latter, several changes in the classification of GEP NENs have been necessary in the past few years, with the aim of providing tools to better stratify patients in different prognostic groups [10, 11, 14]. The current WHO classifications identify two main categories [4, 15] (Table 2.1): neuroendocrine tumors (NETs), broadly corresponding to “carcinoid tumors” or “well-differentiated neuroendocrine tumors/carcinomas” of previous classifications [11, 14] and neuroendocrine carcinomas (NECs), corresponding to poorly differentiated neuroendocrine carcinomas of the previous classification [11]. In the 2010 WHO classification NETs are further divided in two groups (G1 and G2) based on the mitotic count (<2 per 10HPF or 2–20 per 10HPF, respectively) and/or Ki67 index (<2% or 3–20%, respectively). NECs, including small cell and large cell subtypes, present high mitotic count (>20 per 10 HPF) and Ki67 index (>20%) and, by definition, are graded G3. Although this classification is easy to use, it has highlighted a controversial issue represented by a subset of well-differentiated neoplasms with an elevated Ki67 index, generally ranging from 20 to 55%. Indeed, the application of the 2010 WHO classification has emphasized the existence of these cases which show well-differentiated morphology and a high proliferation rate [16,17,18,19,20,21,22]. Several recent studies have shown that these cases have an overall survival rate that is worse than NET G2, but significantly better than poorly differentiated NECs. Additionally, they do not seem to benefit from platinum-based therapy [18,19,20, 23]. Furthermore, they show a molecular profile more similar to that of well-differentiated NETs rather than that of NECs [24]. These findings have confirmed the existence of a new tumor category characterized by a well-differentiated morphology associated with a high proliferation rate. Some authors have proposed defining such a group of tumors as NET G3, and this terminology has been introduced in the 2017 WHO classification of pancreatic NENs (Table 2.1) [15]. It may also be discussed in the future WHO classification of tubular NENs. Moreover, in the 2017 WHO classification of pancreatic NENs, the Ki67 cut-off to separate G1 from G2 NETs has been moved from 2 to 3%, to avoid the gap existing in the previous 2010 WHO classification. This choice has been made by considering that there are papers in the literature demonstrating that the cut-off of 3% is better in differentiating G1 from G2 tumors [25, 26]. It is worth noting, that a Ki67 cut-off of 5% has been proposed by some authors as the best one to differentiate G1 from G2 pancreatic NETs [27,28,29]. However, studies on large series have not demonstrated an improved performance of this threshold, compared to the old cut-off, sufficient to justify such a great change (from 2 to 5%) [30].

Table 2.1 Comparison of 2010 WHO classification of digestive neuroendocrine neoplasms and 2017 WHO classification of pancreatic neuroendocrine neoplasms
Full size table

In addition to pure NENs, there are cases composed of both neuroendocrine and nonneuroendocrine components. Such mixed neoplasms have been defined in the 2010 WHO classification with the term “mixed adenoneuroendocrine carcinomas (MANECs)” [4]. By definition, they are composed of both exocrine and neuroendocrine components and each must represent at least 30% of the lesion. However, this term does not adequately convey the different types of mixed neoplasms. In fact, both exocrine and neuroendocrine components can have variable morphological features, ranging from adenomas to adenocarcinomas or squamous cell carcinomas in nonneuroendocrine components and from well to poorly differentiated NENs in neuroendocrine components [31]. The different combinations of these tumor types gives rise to different prognostic categories ranging from indolent neoplasms composed of adenoma and NET G1 or G2, to very aggressive neoplasms characterized by NECs associated with a nonneuroendocrine component (adenoma, adenocarcinoma or squamous cell carcinoma). For this reason, MANEC should not be considered as a single entity but as a spectrum of different neoplastic diseases. We have recently suggested modifying the term MANEC with “mixed neuroendocrine/nonneuroendocrine neoplasms (MiNENs)” [31], which has been included in the 2017 WHO classification of pancreatic NENs [15].

2.2.2 Pulmonary System

The classification of pulmonary NENs has maintained the same general approach and terminology over the past 20 years, and in the last WHO classification published in 2015 lung NENs have been grouped in the same chapter (Table 2.2). In the previous WHO classification, small cell carcinomas were considered separately from other lung NENs, while large cell neuroendocrine carcinomas were included in the chapter on large cell carcinomas [32]. Lung NENs are divided into four categories: typical carcinoids, atypical carcinoids, large cell neuroendocrine carcinomas, and small cell carcinomas [3]. The distinction among the different categories is purely morphological and the histological criteria will be described subsequently. In the WHO classification of GEP NENs, the Ki67 index represents a crucial tool in distinguishing between the different categories, whereas in the WHO classification of lung NENs it is considered as an ancillary tool which is helpful in the diagnosis of small biopsies with crush artifacts. However, the prognostic value of the Ki67 proliferative index in lung carcinoids has been recently proved [33]. Cases composed of both neuroendocrine and nonneuroendocrine components are also present in the lung where they are defined as combined carcinomas. They are characterized by the combination of small or large cell NECs with a nonneuroendocrine component (adenocarcinomas or squamous). This term is also used to define neuroendocrine carcinomas composed of an admixture of small and large cells [3].

Table 2.2 WHO classification of lung neuroendocrine neoplasms
Full size table

3 Morphological Aspects

3.1 Cytological Features

The cytological diagnosis of NENs is based on well-defined cytomorphological features [34,35,36] although, whenever possible, it should be confirmed using immunocytochemistry. Furthermore, cytological diagnosis of NENs has added value, as several patients, mainly those with poorly differentiated neoplasms, do not undergo surgical procedures due to the extension of the disease at the time of diagnosis. Thus, cytological preparations remain the only available material for diagnosis and molecular analyses which are useful for choosing the most appropriate therapy. In general, NETs are easily diagnosed by means of the most common cytological procedures: expectorations, aspiration, brushing, lavage, and fine-needle aspiration (FNA), performed either with the assistance of computed tomography (CT) (transcutaneous) or with ultrasound (US) guidance (transcutaneous and/or endoluminal). Rapid on-site evaluation (ROSE) is becoming more commonly requested by clinicians and provided by cytopathologists. This procedure allows the increase in yield of collected material and consequently the best triage of the specimen can be achieved. Subtyping NENs in cytological specimens is strongly encouraged because of the clinical implications. In most cases, specific cytomorphological features permit the distinction between low grade (well differentiated) and high grade (poorly differentiated) neoplasms (Fig. 2.1).

Fig. 2.1
figure 1

Cytological features of neuroendocrine neoplasms compared with Ki67 labeling index. (a, b): Well-differentiated neuroendocrine tumor (NET) with clear background, showing cohesive groups of medium-sized cells with moderately abundant eosinophilic cytoplasm and round nuclei with mild pleomorphism; the Ki67 labeling index in this field is less than 1%. (c, d): Poorly differentiated neuroendocrine carcinoma (NEC), small cell type, with necrotic background, composed of small to medium sized cells with a high nucleocytoplasmic ratio, nuclear molding, inconspicuous nucleoli and numerous apoptotic bodies; the Ki67 labeling index is very high, approaching 90%

Full size image

3.1.1 Well-Differentiated Neuroendocrine Neoplasms (Typical Carcinoids, Atypical Carcinoids, Well-Differentiated Neuroendocrine Tumors)

These usually show a clean or hemorrhagic background and tumor necrosis is absent (necrosis is usually associated with aggressive behavior also on cytology). FNA material is highly cellular and contains predominantly monotonous and cohesive groups of medium-sized cells. Cells can also be isolated, but there are usually fewer than in poorly differentiated cases. Architecturally, trabeculae, nests, ribbons, acini, and rosettes are the most frequently encountered structures. Cells are round, columnar, or plasmacytoid in shape and contain eosinophilic cytoplasm with granular nuclei, stripped chromatin, and moderate pleomorphism. Nucleoli are small, and molding is usually absent. Mitoses are absent or very rare. Vascularity can be present on the FNA material as branching capillaries surrounded by neoplastic cells, especially in pancreatic NETs. In cases of cystic NETs, which are more frequently observed in the pancreas representing up to 15% of cases, the background can be proteinaceous and viscous, with abundant macrophages [37]. CEA and amylase levels are usually low in these cystic lesions [38]. In cases of mixed forms, nonneuroendocrine malignant cells (squamous, adenocarcinomatous or undifferentiated) may also be seen.

The differential diagnosis between typical (TC) and atypical carcinoids (AC) of the lung is not easy on cytological grounds since the diagnostic criteria including the presence of necrosis (often punctate and focal) and mitotic count are difficult to evaluate in cytological specimens [3]. However, nuclear pleomorphism with occasional molding and a more disorganized architecture can be observed in ACs. Cells can be fusiform or spindle in peripherally located carcinoids. In cases of hypocellular smears and abundant crushing artifacts, the differential diagnosis with poorly differentiated neuroendocrine carcinomas, especially of the small cell type, is a difficult task. A mitotic count of less than 10/mm2 does not allow the diagnosis of AC rather than SCLC and reaspiration or a biopsy should be requested.

3.1.2 Poorly Differentiated Neuroendocrine Carcinomas (NEC) (Large Cell and Small Cell Types)

Morphologically, the diagnosis of poorly differentiated NECs does not cause any problems for the differential diagnosis with well-differentiated forms, except in the case of paucity of material. The main challenge for the cytopathologist is to think about the possibility of a NEC rather than a lymphoma, small blue round cell tumor or poorly differentiated nonneuroendocrine carcinoma, and the use of immunocytochemical markers is mandatory. Morphologically, the slide background is generally dirty and occupied by necrotic debris and cellular ghosts. Necrosis can be very abundant. Cells are variable in size: small, twice the size of a resting lymphocyte in the case of small cell NECs and large, plasmacytoid-appearing in the case of a large cell NECs. Cells are usually isolated and discohesive. They have a high nucleocytoplasmic ratio in small cell NECs and their most distinctive features are nuclear molding, marked nuclear pleomorphism and nuclear crushing artifacts. Small cell NECs have typical salt-and-pepper chromatin with inconspicuous nucleoli and scant cytoplasm, while large cell NECs have visible nucleoli, coarsely granular chromatin and abundant cytoplasm. Nuclear molding is not so frequently observed as in small cell NECs. Mitotic count is high in both lesions (more than 10 mitoses/mm2 is necessary in case of lung small cell NECs and large cell NECs). Diffuse nuclear molding is frequently observed in cases of small cell NECs while a vaguely basaloid/rosettiform appearance is most evident in cases of large cell NECs (especially on cytoblock material) [39, 40]. Other features associated with malignancy include spindling or multinucleation of tumor cells. Combined forms of well-differentiated neoplasms exist, and should also be reported on cytological specimens.

3.2 Histological Features

GEP and lung NENs are morphologically heterogeneous (Fig. 2.2). A useful and effective framework for the diagnosis is based on the degree of differentiation of these neoplasms, which are subdivided into two broad categories: well-differentiated neuroendocrine tumors and poorly differentiated neuroendocrine carcinomas. It is worth noting that a third category composed of a neuroendocrine and a nonneuroendocrine component (usually adenoma, adenocarcinoma, or squamous cell carcinoma) represents a distinct group with peculiar morphological and clinical features [4, 31].

Fig. 2.2
figure 2figure 2

Histological features of neuroendocrine neoplasms compared with Ki67 labeling index. (a, b): Well-differentiated neuroendocrine tumor (NET) with trabecular pattern of growth, composed of uniform medium sized cells with moderately abundant eosinophilic cytoplasm and round nuclei with only bland atypia; the Ki67 labeling index in this field is less than 1%. (c, d): NET G2 growing in solid nests. The cytological features are similar to those seen in NET G1. The Ki67 labeling index in this field is 5%. (e, f): NET G3 of the pancreas with solid and pseudoglandular pattern of growth, composed of cells with moderate pleomorphism, but with still differentiated nuclear and cytoplasmic features; the Ki67 labeling index in this field is higher than 30%. (g, h): Large cell neuroendocrine carcinoma (NEC) growing in solid sheets of large pleomorphic cells with vesicular nuclei showing prominent nucleoli and abundant eosinophilic cytoplasm: mitotic figures are evident; the Ki67 index is higher than 40%. (i, j): Small cell neuroendocrine carcinoma (NEC) growing in ribbons and solid sheets, composed of small sized cells with high nucleocytoplasmic ratio, dark nuclei with inconspicuous nucleoli and high mitotic count; the Ki67 labeling index is very high, approaching 80%

Full size image

Well-differentiated neuroendocrine tumors (NETs) are characterized by an organoid proliferation of uniform cells, with moderately abundant granular and eosinophilic cytoplasm containing numerous secretory granules. Nuclei are generally round, with clumped or finely granular (“salt and pepper”) chromatin and small nucleoli. Nuclear atypia may be moderate in some cases, and pleomorphic cells with large atypical nuclei may be present. It is worth noting that their presence is not related to an increased biological aggressiveness [41]. NETs can show different architectural features, including small to medium size solid nests, trabecular, pseudo-glandular, or diffuse patterns of growth. These different architectural aspects may be related to the site of origin: ileal and appendiceal NETs are mostly characterized by nests, rectal NETs are frequently trabecular, while ampullary NETs show a characteristic pseudoglandular architecture. As NETs may behave in a malignant fashion, it is important to look for morphological clues that can be associated with tumor aggressiveness, such as the invasion of blood and lymphatic vessels and of perineural spaces, which represent histological signs of malignancy. These morphological features need to be carefully searched and indicated in the pathology report. Other important histological features to be included in the pathology report are the presence of necrosis and the mitotic count, which have prognostic value for GEP-NETs and are the crucial histological features in the differential diagnosis between typical and atypical carcinoids of the lung.

Poorly differentiated neuroendocrine carcinomas (NECs) are highly aggressive neoplasms. Macroscopically, they are poorly circumscribed, may show large areas of necrosis and hemorrhage, and are frequently metastatic when diagnosed. Microscopically, NECs are characterized by a solid proliferation of cells, in large nests or in sheets with large areas of “geographic chart” necrosis. They are divided into small and large cell subtypes, based on the morphological features of the neoplastic cells, but this distinction does not have a relevant prognostic impact. Small cell carcinomas are composed of small to medium-sized (2–4 times the size of a small lymphocyte), round to oval cells with scant cytoplasm, indistinct cell borders and hyperchromatic nuclei with inconspicuous nucleoli. Large cell subtypes are composed of large cells with vesicular nuclei showing prominent nucleoli and abundant eosinophilic cytoplasm. Although tumor cells grow forming sheets or large nests, in the large cell subtype more structured organoid architecture can be observed. Mitotic figures are extremely frequent and the mitotic count >10 mitoses/10HPF is the cut-off for the distinction between atypical carcinoids and large cell neuroendocrine carcinoma of the lung (Table 2.2). The neuroendocrine nature of the neoplastic proliferation has to be confirmed by immunohistochemical analyses, as the differential diagnosis may be a challenging task and includes a number of poorly differentiated nonendocrine epithelial neoplasms, as well as nonepithelial tumors such as PNET, Ewing sarcoma, desmoplastic small round cell tumors, and myeloid and lymphoid leukemia.

Mixed neuroendocrine/nonneuroendocrine neoplasms (MiNENs) are neoplasms with both a neuroendocrine and a nonneuroendocrine component, each representing at least 30% of the tumor mass (Fig. 2.3). The spectrum of mixed neoplasms is wide and encompasses all the possible combinations between neuroendocrine neoplasms (NETs and NECs) and other epithelial tumors (adenomas, adenocarcinomas, and squamous cell carcinomas). Consequently, their biological and clinical behavior is variable and depends on the grade of malignancy of each component [31].

Fig. 2.3
figure 3

Mixed neuroendocrine/nonneuroendocrine neoplasm (MiNEN) of the colon, composed of a moderately differentiated adenocarcinoma and a large cell neuroendocrine carcinoma (a). The immunostaining with anti-chromogranin A antibody highlights the neuroendocrine component and a number of neuroendocrine cells interspersed in the adenocarcinomatous component (b)

Full size image

4 Immunohistochemical Markers

Immunohistochemistry is a cornerstone for both the diagnosis and the prognostic classification of neuroendocrine neoplasms (NENs). In addition, immunohistochemical markers provide important clues to the possible primary site of origin of NENs presenting as metastatic lesions.

The diagnosis of NENs relies on the confirmation of their neuroendocrine nature, by means of general neuroendocrine markers. In addition, the demonstration of the epithelial nature of the neoplastic proliferation, using cytokeratins, is important to rule out the possibility of neural-derived and other neoplasms that can express general neuroendocrine markers. The histopathological classification of NENs is the result of a careful cytomorphological analysis, including mitotic count. However, the use of an immunohistochemical marker of cell proliferation is highly advisable to improve the prognostic stratification. In fact, the Ki67-related proliferative index is incorporated in the WHO classification of GEP NENs and has a pivotal role in distinguishing different prognostic categories among well-differentiated NETs [4]. In addition to morphological parameters and the proliferative index, a number of potential prognostic markers have been proposed, mostly related to specific sites. Finally, in NENs clinically presenting as metastatic lesions, the use of site-specific markers may help in identifying the primary site of origin, with important prognostic and therapeutic implications.

4.1 General Neuroendocrine Markers

4.1.1 Chromogranin A

Chromogranin A, chromogranin B and secretogranin II (chromogranin C) are the best characterized members of the granin family; a group of glycoproteins which represent the major constituents of neuroendocrine secretory granules [42, 43]. While anti-chromogranin A commercial antibodies are widely used in routine diagnostics (Fig. 2.4a, b), antibodies against chromogranin B and secretogranin II are available but are not generally used. Chromogranin A is strongly expressed in normal neuroendocrine cells and, together with synaptophysin (see later in the text) it is the first-choice marker to confirm the neuroendocrine nature of a neoplasm. It is a very specific neuroendocrine marker, however the immunostaining, which is strong and diffuse in NETs, may be weak, focal or even absent in NECs, whose cells contain few secretory granules. Furthermore, a subset of NETs, mainly including L-cell NETs of the hindgut, may be negative for chromogranin A [44]. For these reasons, using a panel of neuroendocrine markers, including at least synaptophysin, is always advisable.

Fig. 2.4
figure 4

General neuroendocrine markers expression in neuroendocrine neoplasms. Dot-like paranuclear immunostaining for chromogranin A (a) in a cytological preparation of a poorly differentiated carcinoma, which is also positive for synaptophysin (c). Intense and diffuse immunostainings for chromogranin A (b) and synaptophysin (d) in histological slides of a well-differentiated neuroendocrine tumor

Full size image

4.1.2 Synaptophysin

Synaptophysin is an integral membrane calcium-binding glycoprotein (38,000 kDa), which is the main constituent of synaptic vesicles of neurons [45]. In normal and neoplastic neuroendocrine cells, synaptophysin is present in cytoplasmic microvesicles, and not in secretory granules [45], and it represents the most sensitive general neuroendocrine marker, being expressed both in well differentiated and in poorly differentiated NENs (Fig. 2.4c, d). However, it is not a specific neuroendocrine marker, as it is also expressed in nonneuroendocrine tissues and neoplasms, such as adrenal cortical carcinomas, neuroblastomas, olfactory neuroblastomas, and Ewing sarcomas/PNETs [45, 46]. Again, it is evident that a panel of antibodies is mandatory for a correct diagnosis of the NEN.

Other proteins associated with synaptic vesicles are present in normal and neoplastic neuroendocrine cells and may be used as neuroendocrine markers. Among these, synaptic vesicle protein 2 (SVP-2) [47], vesicular monoamine transporters 1 and 2 (VMAT1 and VMAT2) (Fig. 2.5a, b) [48, 49], and L-type amino acid transporter (LAT) 1 and 2 [50] have been investigated the most.

Fig. 2.5
figure 5

Additional general neuroendocrine markers in neuroendocrine neoplasms. v-MAT2 is intensely expressed in a gastric ECL-cell NET (a). v-MAT1 immunostaining is strong and diffuse in an ileal EC-cell NET (b). Immunoreactivities for histidine decarboxylase (c) and ASH1 (d) are present in poorly differentiated neuroendocrine carcinomas

Full size image

4.1.3 Other Markers

A number of antigens have been proposed and used in immunohistochemistry as general neuroendocrine markers. As a whole, most of them are not as sensitive and specific as synaptophysin and chromogranin A. However, they may be of use when the immunostaining for one of these markers, like chromogranin A, is weak or absent which may happen in NECs or in specific types of NETs (i.e., hindgut NETs).

4.1.3.1 Neuron-Specific Enolase and Other Enzymes Involved in Hormone Synthesis and Metabolism

Historically, neuron-specific enolase (NSE), the gamma-gamma isoform of enolase, has been widely used as a general neuroendocrine marker. However its specificity has been debated and it has been demonstrated that anti-NSE antibodies cross-react with other dimeric isoforms of enolase, expressed in nonneuronal and nonneuroendocrine cells [51]. This marker has currently been removed from the diagnostic immunohistochemical panels for NENs. Other enzymes, such as the protein gene product 9.5 (PGP9.5)/ubiquitin-C-terminal hydrolase 1 (UCHL-1) [52], L-DOPA decarboxylase (L-aromatic amino acid decarboxylase) [49, 53], tyrosine-hydroxylase, dopamine β-hydroxylase, phenylethanolamine N-methyltransferase, and histidine decarboxylase (Fig. 2.5c) [49, 54] have been variably used as general neuroendocrine markers, although their use is not currently recommended in daily diagnostic practice.

4.1.3.2 Surface Antigens

Two cell surface proteins, classified as a cluster of differentiation (CD), have also been suggested as general neuroendocrine markers. CD56, which is the neural cell adhesion molecule (N-CAM), has high sensitivity in identifying the neuroendocrine phenotype of a neoplasm. However, it lacks specificity, as anti-CD56 antibodies also stain other neoplasms, such as well-differentiated thyroid carcinomas, hepatocellular carcinoma, cholangiocarcinoma, renal cell carcinoma, ovarian carcinoma, endometrial carcinoma, Wilms’ tumor, neuroblastoma, and plasma cell myeloma [55,56,57,58,59,60].

The second cell surface protein employed as a general neuroendocrine marker is CD57, which, again, is not specific for neuroendocrine differentiation, as it is expressed in a variety of other normal and neoplastic cell types, including oligodendroglial cells, Schwann cells and epithelial cells [61, 62].

4.1.3.3 Achaete-Scute Homologue 1

Achaete-scute complex-like 1 (ASCL1), or Achaete-scute homologue1, termed mASH1 in rodents and hASH1 in humans, is a member of the basic helix-loop-helix family. It is a crucial transcription factor for neuroendocrine cell differentiation and it has been shown to be involved in the development of neuroendocrine cells of the thyroid, adrenal medulla, and foregut [63, 64]. The high value of ASH1 in the immunohistochemical workup of NENs relies on the fact that it seems to be expressed exclusively in poorly differentiated NECs (Fig. 2.5d), whereas in NETs its expression is lacking, except for a subset of lung carcinoids [65, 66]. In small biopsies, ASH1 may thus be helpful, together with Ki67, in the differential diagnosis between high grade and low grade NENs [26, 27]. In addition, as ASH1 expression seems to be restricted to NECs, it can also be used as a general neuroendocrine marker in the differential diagnosis of high grade neoplasms.

4.2 Epithelial Markers

The demonstration of the epithelial nature of neoplastic cells is important in the diagnosis of NEN, as both NETs and NECs have nonepithelial mimickers, and the differential diagnosis is important for the correct management of patients. Although a fraction of NENs may not express epithelial markers, cytokeratin-negative NETs should be differentiated from paragangliomas [67], as well as cytokeratin-negative NECs, particularly in small biopsies, which may be confused with other neuroectodermal neoplasms and with neuroendocrine markers-expressing sarcomas [68, 69].

The most useful epithelial markers are cytokeratins, which are intermediate filaments, present in the cytoskeleton of epithelial cells. Antibodies directed against low- and high-weight cytokeratins (CK AE1/AE3) and anti-cytokeratin 8 (CAM 5.2, which cross-reacts with CK18) are most commonly used in routine immunohistochemical practice to detect the epithelial nature of a neoplasm. Antibodies directed against specific cytokeratins may be used in the differential diagnosis with nonneuroendocrine neoplasms (e.g., high weight cytokeratins are absent in lung neoplasms, while they are expressed in squamous cell lung cancer), in the search for an unknown primary origin (e.g., CK20 is expressed in Merkel cell carcinoma and not in lung NENs, whereas CK7 may give a clue to pulmonary or pancreatic origin), or in assessing prognosis (e.g., CK19 expression in pancreatic NENs) [70,71,72].

4.3 Markers for Proliferation: Ki67

Alterations in cell growth and proliferation are key events in neoplastic transformation and in cancer progression. The proliferative fraction of a neoplastic population is correlated with tumor grade and to its biological aggressiveness and has important clinical implications in terms of patients’ outcome and management [73]. As mitotic count represents only one aspect of the proliferating cell, in order to better assess the proportion of neoplastic cells in all of the phases of the cell cycle, immunohistochemical markers of proliferation have been optimized.

Ki67 antigen is a cell proliferation marker expressed in the nuclei of normal and neoplastic proliferating cells, along all cell cycle phases (G1, S, G2, and M), while it is absent from resting cells in the G0 phase. These properties make the Ki67 labeling index (i.e., the percentage of immunoreactive neoplastic nuclei of the total neoplastic nuclei) a good proliferation marker, with a close correlation to the real growth fraction of the neoplastic population [74]. The important prognostic value of the Ki67 labeling index in NENs has gained almost universal consensus, and the neuroendocrine proliferations of the GEP tract are currently classified by the World Health Organization (WHO) on the bases of morphological differentiation, mitotic count, and Ki67 index [4, 15]. In the recent new edition of the WHO classification of pulmonary NENs, Ki67 index is not integrated in the definition of the different entities. However, the diagnostic and prognostic value of the Ki67 index is recognized, and the range of values are inserted among the diagnostic criteria for the first time: 50–100% for small cell carcinomas, 40–80% for large cell carcinomas, up to 20% for atypical carcinoids and up to 5% for typical carcinoids [3].

4.4 Site-Specific Markers

NENs are frequently metastatic at clinical presentation and in up to one third of the cases the site of origin of the tumor is unknown [75]. The identification of the site of the primary neoplasm is important in making the correct treatment decision, especially in the case of well-differentiated NETs, in which therapeutic protocols may vary also according to the site of origin. In poorly differentiated NECs, the major clinical problem is represented by cutaneous neoplasms, in which the distinction of Merkel cell carcinomas from visceral NEC is crucial for correct management. Imaging techniques, including positron emission tomography, are able to identify the primary NEN in a consistent proportion of cases, but in more than 15% of patients it remains occult [72]. The pathologist is therefore asked to give clues to the possible primary site, and the use of a correct panel of immunohistochemical markers is a powerful tool to answer this question.

4.4.1 Well-Differentiated NETs

4.4.1.1 Transcription Factors
4.4.1.1.1 Caudal Type Homeobox 2 (CDX2)

CDX2 is a homeobox domain-containing transcription factor, which is involved in gut development and the maintenance of the intestinal phenotype in epithelial cells. It is expressed in the epithelium of the small and large intestine [76]. CDX2-expressing cells are present not only in pancreatic centroacinar, intercalated and intralobular duct cells, but also in scattered ductal cells [77]. CDX2 immunostaining is widely used in diagnostic pathology to assess the intestinal differentiation of adenocarcinomas. It is expressed in the vast majority of intestinal and appendiceal adenocarcinomas, but also in intestinal type adenocarcinomas of the stomach, esophagus, pancreas, gallbladder and extrahepatic biliary tract, ovary, uterine cervix, urinary bladder, and nasal cavity [78,79,80].

CDX2 immunostaining (Fig. 2.6a, b) in well-differentiated NETs is highly sensitive and fairly specific for a midgut origin [81,82,83]. A meta-analysis of 14 papers assessing CDX2 expression in NETs of different sites (lung, stomach, duodenum, pancreas, jejunum/ileum, cecum, colon, and rectum) has revealed that a strong and diffuse CDX2 immunostaining is present in more than 90% jejunoileal and appendicular NETs. By contrast, CDX2 immunoreactivity was detected only in about 30% of duodenal and rectal primaries, and in about 15% of gastric and pancreatic tumors, with a faint and patchy staining. As little as 3% of lung carcinoids have been found to show CDX2 expression [71].

Fig. 2.6
figure 6

Transcription factors are useful in the identification of the site of origin of well-differentiated neuroendocrine tumors (NETs). CDX2 is intensely expressed in an ileal NET (a), which is negative for TTF-1 (c). By contrast, pulmonary carcinoids are CDX2-negative (b), and TTF-1-positive (d)

Full size image
4.4.1.1.2 Thyroid Transcription Factor-1 (TTF-1)

TTF-1 is another homeodomain-containing transcription factor, and it is involved in the development of the thyroid, of the lung and of the diencephalon. This marker is widely used in the diagnostic pathology of lung and thyroid neoplasms. TTF-1 expression in well-differentiated NET is a very specific marker of pulmonary origin (Fig. 2.6c, d), but its sensitivity is not high and it has been reported as very variable in different papers. Interestingly, peripheral spindle cell carcinoids seem to express TTF-1 more frequently than central carcinoids [84]. In the NENs context, one should also bear in mind that TTF-1 expression is nearly always present in medullary carcinoma of the thyroid and immunostaining for calcitonin and CEA may be of help in defining the diagnosis [67].

4.4.1.1.3 Paired Box Gene 8 (PAX8)

PAX8, a member of the paired box transcription factors family, is involved in thyroid and kidney development and is expressed in carcinomas arising in these organs. It has also been described as a good marker for Müllerian duct-derived neoplasms [72]. PAX8 immunostaining has also been detected in pancreatic islet cells and it has been demonstrated that it can be used as a marker of a pancreatic origin in well-differentiated NETs [67, 72]. Of note, it has been reported that duodenal and rectal NETs express PAX8, whereas ileal NETs are not immunoreactive [72].

4.4.1.1.4 Insulin Gene Enhancer Binding Protein Isl-1 (Islet 1)

Islet 1 is another homeodomain-containing transcription factor, which is important in the embryonal development of neuroendocrine and neural cells and is highly expressed in Langerhans’ islet cells [85]. It is a good marker of pancreatic origin in well-differentiated NETs and its sensitivity is superior to PAX8, with which it shares the immunostaining of ileal and rectal NETs [72].

4.4.1.1.5 Pancreatic and Duodenal Homeobox 1 (PDX1)

This transcription factor is crucial in the development of the pancreas and of the duodenum. In the adult, its expression is restricted to pancreatic islet cells, whereas it is absent in acinar and ductal structures [86]. Among well-differentiated NETs, PDX1 expression is neither a specific nor a sensitive marker for primary pancreatic neoplasms. However, as it has been detected in a subset of pancreatic, duodenal and gastric NETs, whereas it is absent in ileal and pulmonary carcinoids, the main utility of positive immunostaining for PDX-1 seems to be the exclusion of an ileal or pulmonary neoplasm [87].

4.4.1.2 Amine and Peptide Hormones

Commercial antibodies to a wide variety of peptide hormones are available, including serotonin, substance P, calcitonin, gastrin, pancreatic hormones (insulin, glucagon, somatostatin, and pancreatic polypeptide) and intestinal hormone peptides (gastric inhibitory peptide, motilin, secretin, cholecystokinin, vasoactive intestinal polypeptide, glicentin, and peptide YY). Apart from calcitonin, which is, along with CEA, a very sensitive and specific marker for medullary carcinoma of the thyroid, the use of these antibodies is of limited clinical utility.

4.4.2 Poorly Differentiated NECs

As previously mentioned, the differential diagnosis between a Merkel cell carcinoma (MCC) and a cutaneous metastasis of a poorly differentiated visceral NEC represents the single most relevant situation in the management of metastatic NECs. It has been demonstrated that the use of an immunohistochemical panel including TTF-1, cytokeratin 20 and, more recently, MCC polyomavirus (MCPyV), represents an effective approach to this problem. The immunoreactivity for cytokeratin 20 and MCPyV, in the absence of TTF-1 immunostaining is diagnostic for MCC [88, 89]. As for TTF-1 and other transcription factors, the pathologist should be well aware that their expression may be not site-specific (Fig. 2.7) and should not be used in to search for an occult primary [90].

Fig. 2.7
figure 7

Transcription factors are not useful in the identification of the site of origin of poorly differentiated neuroendocrine carcinomas (NECs). TTF-1 may be expressed in primary colonic NECs (a) and not in NECs of the lung (b) (arrowhead indicates positive internal control, represented by normal pneumocytes). On the other hand, CDX2 immunoreactivity may be absent in colonic NECs (c) (positive control in the overlying colonic mucosa, in the left side of the picture), whereas it may be expressed by scattered cells in a pulmonary NEC (d) (arrowheads indicate two positive nuclei. With permission from La Rosa et al. Virchows Arch 2004;445:248–54, reference #82)

Full size image

4.5 Prognostic and Predictive (Theranostic) Markers

4.5.1 Somatostatin Receptors

Somatostatin is a peptide hormone produced in regions of the central nervous system and by D cells in the GEP tract, where it suppresses the release of several hormones. The cell sensitivity to somatostatin is mediated through members of the somatostatin receptors family (SSTRs), composed of at least five subtypes (SSTR1, 2, 3, 4, and 5). SSTRs are frequently expressed by NENs, both in NETs and in NECs, and this is the rationale of the OctreoScan, in which the somatostatin analog octreotide is coupled to 111In to allow the identification of NENs with nuclear medicine imaging. In addition, somatostatin analogs are used as antisecretory drugs in functioning tumors (including patients with carcinoid syndrome) and seem to have a tumoristatic activity in NETs [91]. Somatostatin analogs used in diagnostic and therapeutic settings have the highest affinity for the type 2A receptor (SSTR2A), for this reason its detection in tumor tissues with immunohistochemistry has been implemented (Fig. 2.8). The availability of a monoclonal anti-SSTR2A antibody has improved the specificity and sensitivity of the immunostaining on formalin-fixed and paraffin-embedded samples [92]. Volante and coworkers have proposed a three-tiered scoring system for the evaluation of SSTR2A immunoreactivity in neuroendocrine tumors, taking into consideration both the subcellular localization and the extent of the staining. Pure cytoplasmic immunoreactivity without membranous staining corresponded to score 1, whereas score 2 and 3 were attributed to cases with membranous immunoreactivity in less or more than 50% of cells, respectively. Importantly, only membranous immunoreactivity had a good correlation with positivity to somatostatin receptor scintigraphy and a good response to cytostatic therapy with somatostatin analogues [93]. Interestingly, unrelated groups have recently demonstrated an independent prognostic role of SSTR2A immunohistochemistry in GEP and pulmonary NETs. In fact, SSTR2A membranous immunoreactivity has been reported to be associated to a longer overall and progression-free survival, both in NETs and in NECs [94,95,96].

Fig. 2.8
figure 8

Intense and complete membranous immunoreactivity for SSTR2A in a duodenal gastrinoma

Full size image

4.5.2 Cytokeratin 19

Cytokeratin 19 (CK19) is an acidic cytokeratin highly expressed in the exocrine component of the human adult pancreas, including duct and centroacinar cells, whereas it is absent in normal islets of Langerhans [97, 98]. Aberrant CK19 expression was found in a subset of pancreatic well-differentiated NETs and it has been reported to be an independent prognostic marker in these tumors [99, 100]. However, a subsequent study from our group did not confirm the independent adverse prognostic role of CK19 immunoreactivity, which, although more frequently observed in aggressive NETs, failed to reach statistical significance in multivariate analysis. In addition, we reported that the sensitivity and specificity of the anti-CK19 antibodies in detecting aggressive pancreatic well-differentiated NETs, depend on the clone employed [101].

4.5.3 CD117

CD117, also named c-kit, is a type 3 tyrosine kinase receptor of the platelet-derived growth factor subfamily. It is expressed in multiple cell types and it has been shown to be a marker of progenitor cells in the human pancreas. In pancreatic NETs, CD117-immunoreactivity was found to be an independent prognostic marker, being preferentially expressed in aggressive tumors [102, 103]. A number of studies have demonstrated significant overexpression of CD117 in high-grade NECs, including pulmonary, GEP and cutaneous (Merkel cell carcinoma) carcinomas [104,105,106], and an adverse prognostic significance of this marker’s expression has been reported [90, 102, 107]. However, CD117 immunoreactivity in NENs does not seem to be related to an underlying activating mutation of the c-Kit gene, and this could be an explanation of the poor effectiveness of imatinib mesylate (Gleevec) therapy in these patients [108, 109].

4.5.4 Mismatch Repair Proteins

The coexistence of microsatellite instability (MSI) and widespread gene methylation is a predictor of a better outcome in patients with gastrointestinal NECs [107, 110]. Since there is a good correlation between MSI and the immunohistochemical loss of mismatch repairs proteins, immunohistochemistry for MLH1, MSH2, MSH6, and PMS2 may be included in the diagnostic panel to identify a lower risk class among these very aggressive neoplasms.

5 Practical Applications of Immunohistochemistry in Routine Diagnosis

5.1 Cytology

Ancillary techniques should be employed all the time to confirm the morphological diagnosis [67]. In general, the same immunohistochemical markers used in histology can be applied on the cytoblock preparation or on smears (Fig. 2.4). The most frequently used markers include, in order of importance and in case of limited material, CD45 and pan cytokeratin, and then the panel of neuroendocrine markers such chromogranin A, synaptophysin, CD56 and NSE. Ki-67 can be used to gain an understanding of the proliferative rate in the case of particularly crashed material, but it is not mandatory in cases of pulmonary NETs and NECs. On the contrary, for pancreatic neuroendocrine lesions or metastatic pancreatic NECs, Ki67 is used for grading purposes and has also been validated for its use on cytological material (Fig. 2.1) [111, 112]. Hormones and predictive markers can also be applied in cytological specimens, if needed.

5.2 Histology

The minimal immunohistochemical tests recommended by different guidelines are: chromogranin A, synaptophysin, and Ki67 [4, 113, 114]. While chromogranin A and synaptophysin can be regarded as diagnostic tests (Fig. 2.4), Ki67 has to be considered as both a diagnostic and prognostic marker. The usefulness of chromogranin A and synaptophysin for defining the neuroendocrine nature of a tumor has been discussed above in the specific paragraph. The possible pitfalls of the use of chromogranin A in the diagnostic workup of NECs, where it may be focal or absent reflecting the reduction or absence of secretory granules, which depends on the deficient differentiation of neoplastic cells, have also been underlined. In these cases the use of other general neuroendocrine markers such as the cytosolic components NSE and PGP 9.5, or the membrane molecule CD56 may be necessary, though this depends on the experience of the pathologist. It is worth noting that chromogranin A immunoreaction can also be negative in some NETs, in particular in those of the rectum, even in the presence of abundant intracellular neuroendocrine secretory granules and heavy immunoreactivity for the hormones glicentin, PP, and PYY. As a general rule, at least two positive general neuroendocrine markers are needed to substantiate the neuroendocrine differentiation of a tumor.

The minimal number of hormonal markers necessary for the routine diagnostic workup includes antibodies against insulin, gastrin, and serotonin, either by a clinician’s request and/or to provide information for a better evaluation of the clinical profile and for the patient’s follow-up. Similarly, on specific clinical request, the assessment of SSTR2A in tumor tissue may be necessary. To this end, the use of the score recently proposed by Volante et al. [93] is recommended.

The use of a minimal immunohistochemical panel including transcription factors CDX2, TTF1 and the hormones serotonin, gastrin and insulin is also recommended for the workup of liver or lymph node metastases from occult NETs [75].

6 Circulating Molecular Markers

In the last 10 years, several attempts have been made to elucidate the molecular mechanisms associated with the development and progression of NENs and a large amount of information is available in the literature to date. It is now clear that poorly differentiated NECs show distinct molecular alterations compared to well-differentiated NETs. NECs are mainly characterized by p53 and Rb1 alterations, independently of their site of origin [5]. Conversely, gene alterations found in NETs are more heterogeneous and can be associated with the site of origin; ATRX, and DAXX gene mutations are more frequently observed in pancreatic NETs [115], while MEN1 mutations and/or losses are found in variable percentages of lung and GEP NETs [116]. It is worth noting that, in addition to gene mutations, epigenetic mechanisms are involved in NET development and progression and gene methylation has been recently demonstrated to play a developmental and prognostic role in pancreatic NET [117]. In general, the analysis of molecular alterations is not needed for tumor diagnosis and classification, which is generally easily achieved using morphology and immunohistochemistry. However, the detection of specific molecular features may be useful for the prognostic evaluation and prediction of therapy effectiveness. The systematic review of all the molecular alterations involved in GEP and lung NENs is beyond the scope of the present chapter, where only new information regarding a practical or potential role of new molecular plasma markers in clinical workup is discussed.

Traditional biomarkers that can be identified in the blood stream, including chromogranins and various hormones, have been proved to present several limitations in terms of assay reproducibility, sensitivity, and specificity with the consequent need to find more efficient biomarkers [118, 119]. Detection of circulating transcripts, microRNA, and circulating tumor cells is a new intriguing and promising approach to the diagnosis and management of patients with NENs. Currently, the most widely investigated biomarker tool is the blood-based multianalyte transcript analysis [120, 121]. The multianalyte-derived NET gene signature encompasses the expression of 51 genes which are assessed by four different prediction algorithms and seems to give information on tumor state and evolution, from stability to progression [120]. This approach defines the circulating fingerprint of the tumor showing a higher sensitivity and specificity than traditional secretory markers. The gene expression profile is mathematically analyzed using specific algorithms, which define tumor activity. Interestingly, recent data have suggested that Circulating Transcript Analysis (NETest) may identify tumor categories with a different prognosis and response to somatostatin analogues and peptide receptor radionuclide (PRRT) therapy [122,123,124]. However, this new approach shows some problematic issues including technical complexity, which restricts the analysis to specific laboratories. However, a Delphic consensus assessment has considered that circulating RNA detection seems better than traditionally employed general NEN biomarkers. It has been decided that circulating multianalyte mRNA (NETest) may have clinical utility in both the diagnosis and monitoring of therapeutic efficacy. Overall, it has been concluded that a combination of tumor spatial and functional imaging with circulating transcripts (mRNA) would represent the future strategy for real-time monitoring of disease progress and therapeutic efficacy [125].

In addition to RNA multigene analysis, miRNAs have been considered as potentially useful circulating biomarkers; miRNA are a class of small noncoding RNAs functioning as post-transcriptional regulators. They can be deregulated in neoplasia and may have a potential role as biomarkers. Global miRNA profiles have been evaluated in GEP and lung NETs and have shown nonoverlapping expression among different NET types [126, 127]. Upregulation of miR-103 and miR-107 and downregulation of miR-584, miR-1285, miR-550-002410, and miR-1825 were found in pancreatic NETs [128, 129] and, interestingly, downregulation of serum miR-1290 was able to differentiate pancreatic NETs from adenocarcinomas [129]. In small intestine NETs, other miRNAs seem to be involved and some of them were found to be upregulated (miR-96, miR-182, miR-183, miR-196, and miR-200) or downregulated (miR-31, miR-129-5p, miR-133a, and miR-215) [130]. In the lung, NENs have different miRNA expression profiles that correlate with different tumor categories [127]. However, weak correlations between miRNA expression levels in both tumor tissue and serum have been reported and the fact that some miRNAs are upregulated while others are downregulated suggests that the use of this marker is a complex task which needs to be considered with caution. The American College for Clinical Chemistry underlined several problems related to use of miRNAs in NETs including the low reproducibly and accuracy of the tests used, so additional clinical information is needed before using miRNAs in clinical practice.

Detection of circulating tumor cells, already approved for monitoring breast, prostate, and colon cancers is a novel and interesting approach to the study of NENs. In a recently published study, the prognostic value of circulating tumor cells was demonstrated [131], but further investigations are needed to corroborate the prognostic role of this marker. Recently reported guidelines established that circulating tumor cell analysis is not a sensitive and specific diagnostic tool for NETs [119].