**1. Introduction**

Gastric carcinoma is a rare form of cancer in domestic animals, and in dogs accounts for <1% of all reported neoplasms [1], with adenocarcinoma the most frequent (50–90%) [1]. The

**Citation:** Hardas, A.; Suárez-Bonnet, A.; Beck, S.; Becker, W.E.; Ramírez, G.A.; Priestnall, S.L. Canine Gastric Carcinomas: A Histopathological and Immunohistochemical Study and Similarities with the Human Counterpart. *Animals* **2021**, *11*, 1409. https://doi.org/10.3390/ ani11051409

Academic Editor: Adelina Gama

Received: 10 March 2021 Accepted: 10 May 2021 Published: 14 May 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

median age of gastric carcinoma development is 10 years, with rough collies, Staffordshire bull terriers, chow-chows, Belgian shepherds, Norwegian Lundehunds, Cairn terriers and West Highland white terriers being the breeds most likely to be affected [2].

Compared to gastric carcinoma in humans, the incidence of canine gastric carcinoma (CGC) is relatively low; however, in recent years the disease has been more frequently diagnosed [3,4]. Considering the proposed causal effect of diet on gastric neoplasia in humans [5], this increased frequency in dogs over the last 30 years could be similarly attributed. Breed predisposition has also contributed [6], with those breeds at increased risk becoming more popular. In addition, increased longevity and advances in veterinary diagnostic techniques, such as gastroscopy, may have also contributed to increased diagnosis.

Clinical signs of CGC include vomiting that may progress to hematemesis, melaena, anemia, lethargy, ptyalism, polydipsia, abdominal distension, and abdominal discomfort. Prognosis is generally poor, with a median survival time of 35 days, and confirmed metastasis in about 70–90% of cases at the time of diagnosis or death [7]. Common sites of metastasis include the gastric lymph nodes, omentum, liver, duodenum, pancreas, spleen, esophagus, adrenal glands and lungs [2].

Classification of gastric carcinomas in dogs follows the World Health Organization (WHO) [8] scheme, adapted from humans, which is based on the predominant histological features and the main patterns of cells within the neoplasm: papillary, tubular, mucinous, signet-ring cell (SRC) and undifferentiated types [2]. An alternative scheme, again adapted from human medicine, the Lauren classification, divides tumors into intestinal—cohesive masses and tubular structures; diffuse—individual or scattered nests of neoplastic cells; and mixed—incorporating features of both intestinal and diffuse types [9]. Both schemes have been applied in previous studies [1,3].

The association between chronic inflammation, caused by a variety of factors (bacterial, viral, and parasitic infections, chemical irritants, and nondigestible particles), and carcinogenesis is now well established in humans and animals [10]. The risk of carcinogenesis is higher, the longer the inflammation persists [11]. In humans, it has been shown that several risk factors, such as *Helicobacter pylori* infection, diet, and smoking, are involved in the precancerous cascade of events that lead to gastric adenocarcinomas [12].

The pathogenesis of CGC remains elusive, albeit a high prevalence in certain breeds (e.g., Staffordshire bull terrier, Norwegian Lundehund and Belgian shepherd dog) suggests an underlying genetic etiology [6]. A clear role for *Helicobacter* spp., similar to *H. pylori* in humans, has not been reported in domestic species to date [13]. Whether other *Helicobacter* spp. are involved with carcinogenesis is unclear. *H. pylori* has occasionally been recognized in the canine stomach [2,12], however, the predominant species in dogs are *H. felis*, *H. bizzozeronii* and *H. heilmannii* [14,15]. A clear association between gastric inflammation and *Helicobacter* spp. presence has not been made in previous studies, and in addition, an association with gastric carcinoma has also not been investigated [16].

The role of cell cycle regulators and cell adhesion molecules in cancer is complex and paradoxical, varying by cell type and stage of tumorigenesis [17]. In this study, we aimed to examine the involvement of important cell cycle regulators and cell adhesion molecules, previously studied in human gastric carcinoma, in dogs.

E-cadherin is a calcium-dependent cell–cell adhesion molecule that preserves epithelial integrity and can act both as a tumor-suppressor and as an oncoprotein [18,19]. A recent large-scale study separating subtypes according to their growth pattern (polypoid or nonpolypoid, i.e., signet cell, mucinous and undifferentiated carcinoma) showed that there is a complete loss of E-cadherin in non-polypoid and undifferentiated carcinomas, and reduced expression in polypoid, with no evidence of malignant alteration or invasion in canine gastrointestinal tumors [20].

CD44 is a cell surface receptor for hyaluronic acid and binds to collagen, fibronectin and chondroitin sulfate [21]. Its role in tumorigenesis and metastasis is thought to be through signaling pathways that regulate cell adhesion, migration, proliferation, differentiation and survival [22,23]. Histopathological studies of human gastric carcinoma have

associated high CD44 expression with tumor invasion, lymph node metastasis and patient survival [24–27], although the expression of CD44 in CGC has so far not been investigated.

p16 protein inhibits cyclinD-CDK4/6, and previous studies of human gastric carcinoma have shown that loss of p16 expression has been associated with increased measures of malignancy and poor clinical outcome [28]. One previous study showed loss of expression of p16 in seventeen cases of CGC [29].

14-3-3σ is the focus of much research in human medicine, including gastric carcinomas [30,31]. 14-3-3σ protein regulates the G1/S and G2/M cell cycle checkpoints through sequestration of CDK4, CDK2, and CDK1, and thus prevents mitosis and allows DNA repair. It may act as a tumor suppressor [32] or it may serve as an oncoprotein. Previously, in veterinary species, 14-3-3σ has been reported as an oncoprotein in canine mammary and urinary bladder carcinomas [33,34]. The association and implication of 14-3-3σ with CGC will be examined later in this study.

The aims of this study were to provide an update on the signalment data and histopathological classification of a large case series of CGC, and to further investigate the potential association of chronic inflammation and the presence of *Helicobacter* spp. with cancer. Furthermore, using a subset of cases, the expression patterns of four proteins (E-cadherin, p16, 14-3-3σ and CD44) were studied to determine their potential involvement in CGC development.

#### **2. Materials and Methods**

#### *2.1. Case Selection*

The surgical biopsy databases at VPG Histology, United Kingdom, and the SIDAVE-University of Lleida, Spain were searched for cases of CGC from 2009 to 2019. All cases included relevant medical records (signalment, clinical history, gross description, microscopic description and original diagnosis), and tissues were received at the Royal Veterinary College (RVC) as formalin-fixed paraffin-embedded wax blocks. Individually identifiable owner information was redacted by both supplying institutions. Cases with insufficient tissue or where the diagnosis was not certain on review were excluded.

A control group of canine gastric biopsy samples was established from cases provided by VPG and from the RVC pathology archive. Control group samples were defined based on sample quality, absence of gastric carcinoma, and no previous history of gastric carcinoma. Controls were animals presenting with typical gastrointestinal clinical signs including vomiting, diarrhea and weight loss; however, no tumor was present on histopathological examination.

## *2.2. Histopathological Evaluation*

Histological sections, cut at 4 μm and stained with hematoxylin and eosin, were produced from the provided paraffin blocks and were evaluated microscopically by three veterinary pathologists (A.H., A.S.-B. and S.L.P.). Each case was confirmed as gastric carcinoma and further classified according to the WHO standard into five categories: tubular, papillary, mucinous, SRC, or undifferentiated carcinoma [35,36]. The most frequent histological pattern served as the main classification criterion for each case.

Mucosal inflammation present within histological sections was assessed using modified World Small Animal Veterinary Association standards [8,37,38]. For the purposes of this study, only those parameters consistent with chronic inflammation (intraepithelial lymphocytes, lamina propria lymphocytes and plasma cells, and gastric lympho-follicular hyperplasia) were scored as normal—0, mild—1, moderate—2, and marked—3, following modified WSAVA criteria [37]. A total chronic inflammation score (TCIS, ranging from 0 to 9) was recorded as the sum of the three parameters. Regions of surface ulceration were avoided when assessing chronic inflammatory parameters.

Sections were stained with Warthin–Starry for *Helicobacter* spp. identification. Quantification of *Helicobacter* spp. was performed based on the presence or absence of bacteria in the mucosa using the following system: 0—no helicobacter found in the sample; 1—low

numbers of helicobacter found (<15 bacteria per field of highest helicobacter density); and 2—high numbers of helicobacter found (>15 bacteria per field).
