Practical hints for the diagnosis of mixed neuroendocrine-non-neuroendocrine neoplasms of the digestive system

Abstract

In this editorial, a comment on the article by Díaz-López et al published in the recent issue of the 2024 is provided. We focus on the practical implications critical for providing a correct and complete diagnosis of mixed neuroendocrine-non-neuroendocrine neoplasm (MiNEN) in the gastrointestinal system. The diagnosis of MiNEN begins with the recognition of neuroendocrine features in one component of a biphasic tumor. The non-neuroendocrine counterpart can be virtually represented by any neoplastic type, even though the most frequent histologies are glandular and squamous. However, qualification of the neuroendocrine component requires histological and immunohistochemical confirmation. Neuroendocrine tumors are characterized by a peculiar architectural organization and bland nuclei with granular “salt and pepper” chromatin. Although neuroendocrine carcinomas have multiple and variable presentations, they typically show a solid or organoid architecture. The histological aspect needs to be confirmed by immunohistochemistry, and a diagnosis is confirmed whenever the expression of keratin and neuroendocrine markers is observed. Once both histopathological and immunohistochemical features of neuroendocrine neoplasms are identified, it is important to consider the three major pitfalls of MiNEN diagnostics: (1) Entrapment of neuroendocrine non-neoplastic cells within the tumor mass; (2) Differential diagnosis with amphicrine neoplasms; and (3) Differential diagnosis of tumors that partially express neuroendocrine markers. According to the current guidelines for diagnosing digestive MiNEN, each component must represent at least 30% of the entire neoplastic mass. Although the high-grade histopathological subtype frequently determines disease prognosis, both components can significantly affect prognosis. Thus, if one of the components, either neuroendocrine or non-neuroendocrine, does not fulfill the volumetric criteria, the guidelines still encourage reporting it. These strict criteria are essential for correctly recognizing and characterizing digestive MiNENs. This task is essential because it has prognostic relevance and substantial potential value for guiding further studies in this field. In the future, systematic analyses should be performed to validate or reconsider the current 30% cutoff value.

Key Words: Mixed neuroendocrine-non-neuroendocrine neoplasm; Digestive system; Neuroendocrine neoplasm; Immunohistochemistry

Core Tip: Mixed neuroendocrine-non-neuroendocrine neoplasms are a heterogeneous group of neoplastic diseases that share histological and immunohistochemical features. The most important factor is the uncertain presence of a neuroendocrine component. The presence of entrapped neuroendocrine cells, differential diagnosis of amphicrine neoplasms, and neuroendocrine expression in non-neuroendocrine carcinoma can lead to misdiagnosis. Current guidelines require the fulfillment of volumetric criteria, but the prognostic relevance of the current cutoff remains to be proven.

INTRODUCTION

Mixed neuroendocrine-non-neuroendocrine neoplasm (MiNENs) are mixed neoplasms consisting of two components, one neuroendocrine and the other frequently showing an epithelial nature, with features that overlap with those of their pure counterparts in the same region[1]. Neuroendocrine neoplasms represent a particular and heterogeneous group of malignancies that share histological and immunohistochemical profiles and present some site-based specificities[1]. It is crucial to identify some practical features that support the diagnosis of MiNENs in routine clinical practice. These were appropriately presented and discussed by Díaz-López et al[2] in their recent manuscript, however some considerations should be added.

MiNENs refers to a diagnostic category rather than a specific diagnosis. It refers to a neoplastic lesion with two recognizable components, one of which must present neuroendocrine differentiation[1,3], which must be assessed through an examination[1,3]. The typical histomorphology of neuroendocrine differentiation [well-differentiated neuroendocrine tumors (NETs)] includes an architecture showing organoid, nested, ribbon-like, or trabecular growth patterns; neoplastic cells with eosinophilic cytoplasm; and monomorphic nuclei with “salt and pepper” chromatin. A few mitoses and the absence of necrosis are typical features[1-6]. Conversely, high-grade neuroendocrine carcinomas typically harbor two different cellular morphologies (small cell and large cell features), sometimes within the same tumor mass. Small-cell neuroendocrine carcinoma has a distinct histomorphology characterized by a solid and/or sheet-like growth pattern, with atypical round or spindle cells that present scant cytoplasm, hyperchromatic nuclei, inconspicuous nucleoli, and nuclear molding. As suggested by the name, the cell size is small and typically does not exceed three times the dimensions of a resting lymphocyte. Large-cell neuroendocrine carcinomas differ in their larger cell size and different growth patterns; they are often nest-like with peripheral palisading tumor cells and usually exhibit eosinophilic cytoplasm, polymorphous nuclei, prominent nucleoli, and vesicular chromatin[1,3,7-11].

Neuroendocrine differentiation must be confirmed by immunohistochemical analysis. The most sensitive markers are synaptophysin, insulinoma-associated protein 1, and chromogranin A, which sometimes present diminished intensity in poorly differentiated neuroendocrine neoplasia (NEN)[1,3,12,13]. Such markers are beneficial and have reliable staining patterns in a biopsy setting[14]. The epithelial origin of neoplastic cells should also be proven by staining for keratin to rule out the presence of paragangliomas[1]. To complete the final pathology report of NETs, a study of proliferative activity using the Ki-67 index (MIB-1 clone) must be conducted. The mitotic count is now considered poorly reproducible and, thus, is less meaningful than Ki-67 for diagnosis of NENs of the digestive system[1,3]. Ki-67 can be estimated using digital pathology-based tools to improve standardization, even in a biopsy setting[15].

One could ask why immunohistochemical analysis of neuroendocrine differentiation is not widely performed, considering that NENs have a heterogeneous appearance that could be misinterpreted. Immunohistochemical analysis has low specificity in unselected cases, with a risk of overdiagnosis of neuroendocrine neoplasms[11,12,16]. The three main examples of this phenomenon are as follows: (1) Since neuroendocrine cells are typically present in different tissues, normal neuroendocrine cells can be misinterpreted as part of a neoplastic process. A classic example is pancreatic insulae entrapped in pancreatic ductal adenocarcinoma or enterochromaffin-like cells intermingled with neoplastic elements in gastric cancer[17,18]; (2) The second possible misinterpretation in the neuroendocrine context is amphicrine neoplasia. An amphicrine tumor is defined as a neoplastic process in which tumor cells present both exocrine and neuroendocrine phenotypes[19-21]. An example is appendiceal goblet cell adenocarcinoma, in which the cells are both mucinous and positive for neuroendocrine markers[22,23]. Immunohistochemical, genetic, and transcriptome analyses revealed that this histological type significantly differs from adenocarcinoma, neuroendocrine neoplasms, and MiNENs in the same site. Therefore, special attention should be given to identifying this complex entity[24]; and (3) The last example in this context is neoplasms showing aberrant or partial expression of neuroendocrine markers. For example, a solid pseudopapillary neoplasm of the pancreas is a well-known entity that can express synaptophysin and CD56; however, it is not listed as NEN[25-28]. Another example is colorectal adenocarcinoma with synaptophysin expression, a subgroup of BRAF-mutated colorectal adenocarcinomas that does not fulfill the histological criteria for NEN diagnosis[29]. Differentiation of this subgroup is important due to its poor prognosis, as confirmed by other investigators[30].

A PRACTICAL DIAGNOSTIC ALGORITHM FOR MiNENs

One of the most important topics in the digestive MiNEN landscape is identification of the correct diagnostic algorithm (Figure 1), as highlighted by Díaz-López et al[2]. The first step is to histologically identify a biphasic neoplasm with at least one component that presents both epithelial aspects and a neuroendocrine phenotype[1,3]. The second step concerns the immunohistochemical characterization of both neoplastic components using different staining, including: (1) Keratins to support the epithelial origin[31]; (2) Synaptophysin, chromogranin A, and preferably also INSM-1. In addition, CD56 is recognized as a neuroendocrine marker; nevertheless, the aforementioned markers are used more often for immunohistochemical staining because of their major sensitivity and specificity[1,3,32]; (3) Ki-67 (MIB-1 clone) is also used to investigate the proliferation index of the neoplasm[5,13,33,34]; and (4) Other potentially useful markers are used to identify the site of origin, such as CDX-2 for small bowel NENs[12,31], SATB2 for large bowel NENs and Merkel cell carcinomas[35], and Islet-1 for pancreatic neoplasms[3]. The neuroendocrine component may present challenging features that could hamper the immediate distinction between NET G3 and NEC. In those cases, immunohistochemical assessment of Rb and SSTR2, which are typically lost in NECs and retained in G3 NETs, and p53, which is more frequently altered in NECs than in G3 NETs, is helpful[4,36,37]. Moreover, pancreatic G3 NETs are enriched in DAXX/ATRX mutations, corresponding to a lack of immunohistochemical expression of the homonymous protein, while the expression of these proteins is conserved in NECs[37-40]. In addition, the alternative lengthening of telomeres, ALT, a vital biological mechanism important in pancreatic NETs, is never activated in NEC[39,41]. The presence of necrosis favors the diagnosis of poorly differentiated carcinomas, but it is not listed among the grading parameters of tumors in the gastrointestinal and pancreaticobiliary tracts[1,3]. The diagnosis of a MiNENs is supported whenever neuroendocrine marker expression is identified in one component of the biphasic mass. Notably, NEC can develop a glandular phenotype, but the diagnosis of MiNEN cannot be made in these patients because of the diffuse staining of neuroendocrine markers[42]. The last consideration is tumor staging for MiNEN, which is classified as a site-specific non-neuroendocrine component according to the AJCC/UICC TNM classification[43]. This information is crucial for stratifying patients according to prognosis and guiding clinical management.

Figure 1 Diagnostic algorithm for the distinction of mixed neuroendocrine and non-neuroendocrine neoplasm. Mixed neuroendocrine and non-neuroendocrine neoplasm needs to be distinguished from neoplasms with focal neuroendocrine aspects like paraganglioma, non-neuroendocrine neoplasm with single neuroendocrine marker expression, neoplasm with interspersed neuroendocrine cells, or amphicrine neoplasm. ¹The definition of scattered cells includes, but is not exclusive of, other appearances like sheets, ribbons, micronodules, or insulae. MiNEN: Mixed neuroendocrine and non-neuroendocrine neoplasm; NE: Neuroendocrine; NEN: Neuroendocrine neoplasm.

PROGNOSIS AND FUTURE PERSPECTIVE OF MiNENs

The recognition of MiNENs is a critical task because there is much evidence that the two components do not behave like their isolated counterparts[41,44,45]. Generally, the high-grade histological subtype guides the prognosis of the disease; however, both components can progress and metastasize, significantly impacting the clinical history of patients with MiNENs[46].

In most districts, there is no minimum percentage of the two tumor components for establishing a diagnosis of MiNEN. For the gastrointestinal tract, the discussion is less straightforward, and the guidelines indicate that the smaller component must occupy at least 30% of the entire tumor mass[47]. Although this cutoff was arbitrarily determined when there was limited knowledge about these lesions, it has been maintained over time[2,48]. Systematic studies have not yet investigated the possibility of changing the cutoff value. Recent findings suggest that the presence of a high-grade component significantly impacts prognosis, even in cases where the high-grade component represents less than 30% of the entire mass[49]. For gastrointestinal tumors that do not satisfy the current diagnostic criteria for MiNENs, reports of the presence of smaller components (< 30%), both neuroendocrine and non-neuroendocrine, are highly recommended[46,49].

CONCLUSION

Although first described in 1924[50], MiNENs remain poorly understood. This is partly due to the rarity of these lesions and the different clinical behaviors of MiNENs at distinct sites. In addition, the misdiagnosis or underrecognition of this rare entity could have affected its identification and the overall literature on mixed neoplasms. In this editorial, inspired by the elegant manuscript by Díaz-López et al[2], some controversial lesions that are not real MiNENs have been highlighted, including: (1) Tumors with (scattered) neuroendocrine non-neoplastic cells; (2) Amphicrine neoplasms; and (3) Adenocarcinomas that partially or aberrantly express neuroendocrine markers.

Current guidelines/diagnostic criteria for MiNENs require that the smallest tumor component comprises at least 30% of the entire neoplastic mass. Nevertheless, smaller components (< 30%) should be reported since they can affect prognosis and may represent an additional point of inspiration for further research on MiNEN.