Currently, gastric cancer is ranked fifth for incidence and third for morbidity not just in the Russian Federation, but throughout the world. A significant increase in the number of young patients with this disorder has given rise to numerous studies aimed at identifying early predictors for carcinogenesis and critical risk factors for malignization [1, 2].

Among the gastric cancer pathogenesis factors, the Helicobacter pylori (HP) infection is the leading cause of the disorder. It's been proven that 80% of patients with gastric cancer have had a history of HР infection [3, 4]. There is no doubt that HР is one of the crucial factors for a cascade of carcinogenesis, which initiates the development of acute gastritis, and progression from acute to chronic gastritis with subsequent atrophy, metaplasia, and dysplasia, ultimately leading to the development of adenocarcinoma [5].

In recent years, a growing body of research has been focused on the changes in the tumor microenvironment. Studying the tumor-associated macrophages (TAMs) can be very promising in the aspect of cancer research [6]. TAMs are associated with the occurrence of tumors playing a vital part in metastasis and immune regulation of neoplasms. Today, the macrophage functional heterogeneity has been established. There are type 1 macrophages, which initiate the pro-inflammatory response, and type 2 macrophages involved in remodeling, which can promote tumor cell migration and invasion due to activation of neoangiogenesis [7–9].

It is known that HР can actively recruit macrophages, monocytes and dendritic cells to the gastric mucosa. HР forces these cells to secrete pro-inflammatory cytokines and chemokines, causing inflammation and gastric mucosal injury [10, 11]. Several studies report the influence of НР infection on the macrophage polarization due to promotion of free radicals induced oxidative processes. According to the results of a number of research studies, bacterial factors contribute to the increased production of reactive oxygen species and hypoxiainducible factor 1 α [8, 9]. Furthermore, HP is able to synthesize large amounts of inducible nitric oxide synthase (iNOS), contributing to macrophage polarization into М2 phenotype, which, in contrast to М1, exhibit a rather low phagocytic activity, thus facilitating the further pathogen persistence [12].

Currently, endoscopy with targeted biopsies is the main method for diagnosing atrophic gastritis and its complications. However, focal lesions are quite common for НР-associated atrophic gastritis, that is why it is essential to ensure obtaining adequate biopsy samples, which is often difficult to accomplish in the outpatient setting [13]. Nevertheless, early detection of atrophy and subsequent precancerous changes (intestinal metaplasia and dysplasia) is essential for gastric cancer prevention. Immunohistochemical verification of morphological changes in gastric mucosal immune cells could be part of the solution to this problem.

In view of the above, the study was aimed to assess the predictive potential of the macrophage population immunohistochemical phenotyping in the early malignization of the HP-associated chronic gastritis.

METHODS

Morphological studies were performed in the Center for the Shared Use of Scientific Equipment "Molecular Biology" of the S.I. Georgievsky Medical Academy, V.I. Vernadsky Crimean Federal University, with support from the Crimean Federal University development program. Gastic biopsy samples of male and female patients aged 48 ± 7.2 infected with HP were used as the research material. The patients were divided into three groups: non-atrophic chronic gastritis (n = 10), atrophic chronic gastritis (n = 10), G1/G2 Lauren intestinal type gastric tubular adenocarcinoma (n = 10). Inclusion criteria: positive HP test. Exclusion criteria: negative HP test or no information about this test in medical history. According to their medical history, all patients were НР-positive. НР was verified based on the serum anti-НР immunoglobulin G (IgG) quantification by the enzyme-linked immunosorbent assay and the rapid urease test.

The control group included fragments of visually intact gastric mucosa obtained from patients, who died from causes other than gastrointestinal disorders, with no histological signs of gastritis and not infected with НР.

To perform morphological assessment of the biopsy and necropsy samples, the fragments of gastric mucosa were fixed in the 10% neutral buffered formalin. Tissue processing and histological slide preparation were performed by standard methods with subsequent hematoxylin and eosin staining [14].

Immunohistochemical staining (IHC) was performed on the slices 3–4 µm thick by peroxidase-antiperoxidase (PAP) complex technique in accordance with the Thermo Scientific protocols (USA) with the use of Leica Biosystems visualization system (Novocastra; USA) and DAB Chromogen (Abcam; UK). Subsequently, cell nuclei were stained with hematoxylin. Macrophage population was visualized using the CD68 pan-macrophage/monocyte marker (clone PG-M1, dilution 1 : 50, DAKO; Denmark) and the CD163 marker of typical 2α reparative phenotype macrophages CD163 (clone EPR-19518, dilution 1 : 500, Abcam; UK). According to meta-analysis, studying the СD68 and CD163 markers is more infomative than studying the other М2 marker, CD206 [15]. Meanwhile, CD163 can be expressed as an additional surface marker of blood monocytes, which are also capable of phenotypic polarization. However, monocytes demonstrate lower expression of this marker compared to macrophages.

Intensity of neoangiogenesis was defined using the CD31 endothelial marker (clone 1A10, Leica Bond; USA) by assessing the total cross-sectional area of blood vessels.

The samples obtained were analyzed with the CX41 light microscope (Olympus; Japan). Morphometric assessment was performed using the Image J software (NIH; USA) on the standardized area of the histology section image. Cells with positive cytoplasmic staining were counted in 10 fields of view using the 20х/0.50 UPlanFL N objective (Olympus; Japan) with subsequent counting the average number of cells in the biopsy sample.

Statistical analysis was carried out using the STATISTICA 10.0 software for Windows (StatSoft Inc.; USA). Testing for normal distribution was performed by the Kolmogorov– Smirnov test. The following statistical parameters were defined: mean (M), error of the mean (m). The inter-group comparisons of morphological parameters were carried out by one-way analysis of variance (One-Way ANOVA) with the use of the Dunnett test (Q) for comparison of the small samples with the control group. The intra-group differences were considered significant when p ≤ 0.05.

RESULTS

According to serological study, the antibody titer of 1 : 5 was observed in 33% of patients with NACG; in 67% of patients the antibody titer exceeded 1 : 10. In the group with ACG, the antibody titer exceeding 1 : 5 was observed only in 10% of patients; 49% of the examined patients had the antibody titer exceeding 1 : 10, and another 41% of patients demonstrated the high antibody titer (in the range of 1 : 20–1 : 80). In 77% of patients with verified tubular adenocarcinoma (GAC), the antibody titer was 1 : 20 or higher, and 23% of patients had the antibody titer of 1 : 5–1 : 10.

According to morphological assessment, gastric mucosa of the controls had a typical pattern, represented by the lining of regularly shaped columnar epithelium, forming the gastric pits with no signs of gastritis (fig. 1А). In the submicosal layer, there were solitary resident macrophages of both 1^st and 2^nd phenotype (fig. 1B, C). The moderately prominent vascular pattern was clearly visible. Blood vessels were represented mainly by capillaries and venules (fig. 1D).

When performing the overview staining, it was found that gastric biopsy specimens of patients with NCG were characterized by lymphoplasmacytic infiltration of various degree (mild to moderate), as well as by formation of solitary hyperplastic lymphoid aggregates within the stroma, increased mucoidization of the columnar epithelium of the integument and pits, and foveolar hyperplasia. In some zones of biopsy specimens there were foci of fundic gland hyperplasia (fig. 2А).

According to IHC study of biopsies obtained from patients with NACG, CD68⁺ cells were distributed homogeneously in the lamina propria of the gastric mucosa; type 2 monocytes/ macrophages made up about a half of these cells. The CD163 : CD68 ratio of 0.67 ± 0.02 (fig. 2B, C) was significantly larger compared to values obtained for the control group (р < 0.05). In our view, the monocyte/macrophage population ratio index shift compared to the control group is indicative of the proinflammatory response cascade initiation.

The total cross-sectional area of blood vessels in patients with NACG was 3590.92 ± 356.27 µm², which was indicative of insignificant intensification of angiogenesis compared with the control group specimens (р > 0.05). The cross-sectional area of blood vessels in patients with this type of gastritis was minimal in comparison with the groups of patients with ACG and GAC (fig. 2D; table).

Biopsies of patients with ACG also demonstrated polymorphic histological changes. Glandular atrophy resulted from two fundamentally different processes: fibrous stroma transformation or intestinal metaplasia (small/large intestine or the combination of these metaplasia types). Atrophic gastritis with the lamina propria fibrosis was characterized by the marked lymphoplasmacytic infiltration, often with lymphoid follicles consisting of expanded germinal centers (fig. 3А).

The monocyte/macrophage population was also distributed homogeneously throughout the whole biopsy specimens, both in foveolar zones and in the vicinity of the lymphoid accumulation. The majority of cells had the type 2 monocyte/ macrophage immunophenotype. The CD163 : CD68 ratio was 0.81 ± 0.04 (fig. 3B, C). Furthermore, microvascular area in biopsies of this subgroup was significantly increased compared to values obtained both for NACG group and control group (fig. 3D).

Morphological and immunohistochemical changes in biopsies of patients with verified tubular adenocarcinoma were particularly illustrative in terms of the macrophage population functional re-profiling and the intensity of neoangiogenesis. Among tubular structures with signs of cellular and nuclear atypia, and areas of solid structure, the desmoplastic stroma with moderate and marked dense inflammatory cell infiltrate of lymphoid cells, histiocytes, neutrophils were observed (fig. 4А).

The number of cells in monocyte/macrophage population increased twice compared to values obtained for ACG and four times compared to value obtained for NACG. The CD163 : CD68 expression ratio was 0.88 ± 0.03 (fig. 4B, C). The maximum microvascular area, significantly exceeding the control values, was also observed (fig. 4D), which could be indicative of the intensified neoangiogenesis aimed at forming the comfortable microenvironment, and contributed to metastatic tumor cell dissemination [16].

DISCUSSION

The known molecular biomarkers of monocytes/macrophages demonstrate the considerable diagnostic capacity and could serve as malignancy risk predictors due to specific identification of particular immune cell subpopulations, allowing one to predict the course of a number of disorders. There is evidence that macrophages can play a dual role in inflammation depending on the signals from their microenvironment due to different phenotypes [17].

Chronic НР-associated gastritis is accompanied by the dynamic increase in the size of the general monocyte/ macrophage population, since it is known that НР actively triggers the cytokine cascade playing a vital part in realization of chronic inflammatory and destructive processes in the gastric mucosa [4]. During the acute inflammatory response, macrophages are usually polarized into М1 phenotype by interferon-γ and microbial products, such as lipopolysaccharide, and exhibit strong antimocrobial activity due to production of bactericidal components (nitric oxide and oxygen radicals) [9]. However, this bacterium can inhibit the effects of active substances produced by macrophages due to neutralizstion of these substances with catalase and superoxide dismutase, which provides survival of the bacteria. However, the reactive oxygen species, abundantly secreted by macrophages, cause the death of the gastric mucosal cells, which contributes to the risk of atrophy [18].

According to our study, atrophic gastritis is characterized by vector redistribution of monocytes/macrophages into the 2nd functional type. М2 macrophages are induced by type 2 T-helper cells or anti-inflammatory cytokines and growth factors, including IL4, IL10 [6]. Apparently, monocyte/ macrophage polarization shift towards М2 with the progression of inflammation should ensure tissue regeneration and enhance humoral response. However, the long-term increased production of the НР pathogenicity factors may lower the ability of macrophages to eliminate bacterial antigens, which results in the long-term persistence of the pathogen and chronic inflammation [19, 20]. At the same time, chronic inflammation with persistent М2 monocyte/macrophage polarization is associated with increased risk of developing gastric cancer [1, 13, 21].

Thus, the dynamic increase in the pan-macrophage population size and progressive re-profiling of macrophages into the 2^nd functional phenotype with the transition from non-atrophic gastritis to atrophic gastritis, and peak values in biopsies with gastric adenocarcinoma, were observed during our study. In our opinion, these changes are related to remodeling functional activity of the alternatively activated М2 monocytes/macrophages due to increased expression of proangiogenic growth factors, which on the one hand contributes to fibrotic changes in the gastric stroma in the conditions of chronic progressive inflammation and tissue hypoxia, and on the other hand creates a comfortable microenvironment for the dysplastic and neoplastic changes in the glandular epithelium [11, 22]. Furthermore, it is well-known that certain М2 macrophage populations exhibit the vivid tumor-stimulating activity, since these populations are involved in the local immunosuppression of immunocompetent cells, which results in the tumor cell escape from the immune surveillance [23].

The diversity of monocyte/macrophage phenotypes requires further study and standardization of their predictive potential. For this purpose, the studies of the larger number of patient samples should be carried out, as well as the assessment of the other macrophage markers, typical for monocytes and macrophages. In particular, monocytes are able to express the CD163 and CD204 surface markers in addition to the CD1 and CD16 surface markers, which could be defined by double staining immunohistochemistry [24]. Furthermore, regardless of the generally accepted division of macrophages into classical and alternative macrophages, the other macrophage subtypes have been described, including the pro-fibrotic M2-like macrophages, which produce growth factors, and ischemia-induced fibrolytic M2-like macrophages, which secrete proteases [12].

 In general, identification of the TAM functional programming mechanisms would make it possible to identify terapeutic targets, and to develop new approaches to treatment of cancer patients based on the combination of chemotherapy and immunomodulatory agents in order to induce the immune response to the tumor [13]. Synergistic effects of these modalities may significantly increase the efficiency of antitumor therapy.

CONCLUSIONS

Chronic H. pylori-associated gastritis is accompanied by the dynamic increase in the size of the general macrophage population with the progression of atrophic and metaplastic processes. Atrophic gastritis and adenocarcinoma are characterized by vector redistribution of macrophages into the 2nd functional type. Increased expression of CD163 marker of monocytes/macrophages can serve as a predictor of chronic gastritis malignancy together with histological study of the glandular epithelium atrophy and metaplasia degree.