Abstract
Background:
Mucosal gastric atrophy and intestinal metaplasia (IM) increase the risk for the development of gastric cancer (GC) as they represent a field for development of dysplasia and intestinal-type gastric adenocarcinoma.
Methods:
We have investigated the expression of two dysplasia markers, CEACAM5 and TROP2, in human antral IM and gastric tumors to assess their potential as molecular markers.
Results:
In the normal antral mucosa, weak CEACAM5 and TROP2 expression was only observed in the foveolar epithelium, while inflamed antrum exhibited increased expression of both markers. Complete IM exhibited weak CEACAM5 expression at the apical surface, but no basolateral TROP2 expression. On the other hand, incomplete IM demonstrated high levels of both CEACAM5 and TROP2 expression. Notably, incomplete IM with dysplastic morphology (dysplastic incomplete IM) exhibited higher levels of CEACAM5 and TROP2 expression compared to incomplete IM without dysplastic features (simple incomplete IM). In addition, dysplastic incomplete IM showed diminished SOX2 and elevated CDX2 expression compared to simple incomplete IM. CEACAM5 and TROP2 positivity in incomplete IM was similar to that of gastric adenomas and GC. Significant association was found between CEACAM5 and TROP2 positivity and histology of GC.
Conclusions:
These findings support the concept that incomplete IM is more likely associated with GC development. Overall, our study provides evidence of the heterogeneity of gastric IM and the distinct expression profiles of CEACAM5 and TROP2 in dysplastic incomplete IM. Our findings support the potential use of CEACAM5 and TROP2 as molecular markers for identifying individuals with a higher risk of GC development in the context of incomplete IM.
Mini-Abstract:
We have evaluated the expression of dysplastic markers TROP2 and CEACAM5 in gastric antral intestinal metaplasia (IM) and cancer. TROP2 and CEACAM5 are elevated in dysplastic incomplete IM and cancers.
Introduction
Gastric intestinal metaplasia (IM) is a common precancerous lesion in which the normal gastric epithelium is replaced by intestinal-like epithelium as an adaptive response to chronic injury (1). Despite the increased risk of gastric cancer (GC) associated with IM, the overall risk of GC in asymptomatic patients with IM is low, and surveillance is not typically recommended. However, in regions with a high incidence of GC, endoscopic surveillance has been shown to effectively reduce the risk of developing GC (2). Due to the multifocal nature of the metaplastic process, extensive mapping techniques have been proposed to identify IM (3). Histologic sections stained with hematoxylin and eosin (H&E) are commonly used to recognize IM, but it is widely recognized that IM is a heterogeneous lesion, and several classification systems have been proposed to categorize its subtypes (4–6). For instance, Jass et al. proposed a classification system based on mucin phenotypes, which distinguishes three subtypes: type I (complete, sialomucin-positive), type II (incomplete, sulfomucin-negative), and type III (incomplete, sulfomucin-positive) (5). Several studies have shown a significant relationship between incomplete IM and GC (7–9), highlighting the importance of identifying the subtype of IM in assessing the risk of developing GC in patients with IM.
Numerous efforts have been made to discover molecular markers that can specifically identify pre-neoplastic lesions with a high risk of developing dysplasia in the human stomach. CEACAM5 and TROP2 are considered promising candidates. Carcinoembryonic antigen-related cell adhesion molecules (CEACAMs) are a group of mammalian immunoglobulin-related glycoproteins that function as cell adhesion proteins. CEACAM5 was first reported in 1965 as a gastrointestinal oncofetal antigen (10), but it is now known to be overexpressed in various types of cancers, including gastrointestinal, respiratory, genitourinary, and mammary cancers (11). Different expression patterns of CEACAM5 have been reported in non-neoplastic and neoplastic lesions of the stomach (12, 13). CEACAM5 is up-regulated in dysplastic lineages in the human stomach (14). We have also observed a similar pattern of CEACAM6 upregulation in dysplastic cells with single cell RNA sequencing (14). TROP2 is a transmembrane glycoprotein that was first discovered as a trophoblast surface marker (15). It is upregulated in many epithelial human cancers and is associated with tumor aggressiveness and metastasis (16). In GC, TROP2 overexpression is a prognostic marker for poor outcome (17, 18). More recently, TROP2 has been identified as a critical biomarker that is upregulated in the transition from metaplasia to dysplasia in the stomach (19, 20). However, the combined expression of CEACAM5 and TROP2 has not been explored in human gastric precancerous lesions and GC.
In this study, we aimed to examine the expression profile of CEACAM5 and TROP2 in IM and gastric tumors of the human gastric antrum. The expression of CEACAM5 and TROP2 was investigated in normal antral mucosa, inflamed antrum, complete IM and incomplete IM lesions, with a specific focus on incomplete IM that exhibits a dysplastic morphology. Additionally, we assessed CEACAM5 and TROP2 expression in gastric adenomas and carcinomas and compared their expression profile with those in IM. Our studies indicate that co-expression of CEACAM5 and TROP2 can identify a sub-class of dysplastic incomplete intestinal metaplasia.
Material and Methods
Subjects
Formalin-fixed and paraffin-embedded (FFPE) samples of antral mucosa were obtained from the Gastric Cancer Epidemiology Programme (GCEP) National University Hospital, Singapore. Clinicopathological details of the Singapore cohort are shown in Supplementary Table 1. FFPE samples of early gastric cancers (EGC, n=59) and gastric adenomas (GA, n=19) were collected from the patients who underwent surgical resection or endoscopic submucosal dissection at Jeju National University Hospital (JNUH), Jeju, Korea, from 2008 to 2016. Submucosal invasion was observed in 42 cases (71%). Lymph node dissection was performed in 47 cases, among which metastasis was detected in 10 cases (21%). The clinicopathological features are detailed in Supplementary Table 1. The informed consent from the patients was waived due to the retrospective nature of our study. This study was approved by the Institutional Review Board of Jeju National University Hospital (IRB No. 2020-01-009). All procedures were in accordance with the Helsinki Declaration of 1964 and later versions.
Histology
The diagnosis of gastric IM was independently performed by two pathologists (B.J. and S.L.) based on histologic evaluation of sections stained with hematoxylin and eosin (H&E). The diagnostic criteria for complete and incomplete IM were as follows: complete IM is characterized by the presence of eosinophilic enterocytes with a well-defined brush border and well-formed goblet cells, with or without Paneth cells. In contrast, incomplete IM is characterized by irregular mucin droplets of variable size in the cytoplasm, without a brush border, and the absence of Paneth cells (6). When IM glands display increased cellularity, crowded glands, and disorganized structure, they were considered to have dysplastic features. The diagnosis of dysplastic IM was established when both pathologists reached an agreement on the presence of dysplastic characteristics. It should be noted that dysplastic IM does not meet the criteria for definite dysplasia.
Tissue microarray construction
In total, four tissue microarrays (TMAs) containing 59 EGC, 19 GA, and normal (n=3), inflamed (n=11) or metaplastic (n=51) gastric lesions were generated from the samples of Koran cohort. Briefly, through histologic examination, the representative normal, IM or tumor area was marked in each case. In the case of tumor, the representative area makes up at least 70 % of the tumor cell population. Tumor core (4 mm in diameter) was extracted from individual FFPE gastric tumors (donor blocks) and placed in a new recipient paraffin block (tissue array block) using a trephine apparatus (SuperBioChips Laboratories).
Immunohistochemistry including Multiplexed immunofluorescence and Interpretation.
Paraffin-embedded tissues were sectioned at a thickness of five-micrometer and stained with H&E. For immunostaining, unstained paraffin-embedded tissue sections were deparaffinized in Histoclear and rehydrated through a serial dilution of ethanol. Antigen retrieval was performed in target retrieval solution (Dako) using a pressure cooker for 15 minutes. Sections were incubated with Serum-free Protein Block Solution (Dako) for 1.5 hours at room temperature (RT). Incubation with primary antibody for CD44v9 (Cosmo Bio, LKG-M001; 1:15000), CEACAM5 (ABclonal, A18131; 1:1000), SOX2 (Thermofisher, 14-9811-82;1:500), CDX1 (Thermofisher, PA5–23056; 1:500), CDX2 (Cell Signaling, 12306S; 1:500), and/or TROP2 (R&D Systems, AF650; 1:1000) was performed at 4°C overnight. For immunohistochemistry, sections were incubated with HRP-conjugated secondary antibodies (Vector Laboratories) for 15 min at RT. Then, sections were visualized using the Dako Envision+ Detection System Peroxidase/DAB substrate (Dako) according to the manufacturer’s instructions and counterstained with Mayer’s Hematoxylin (Sigma-Aldrich). After dehydration, sections were scanned on an SCN400 Slide Scanner (Leica Biosystems) at 20x magnification. For multiplexed immunofluorescence, Alexa-conjugated secondary antibodies were incubated at RT for 1 hour. For nuclear staining, sections were incubated with Hoechst 33342 (Thermo Fisher Scientific) for 5 min at RT. Images were captured on a Zeiss Axio Imager M2 microscope using Axiovision digital imaging system at 20x magnification. The expression of CEACAM5 or TROP2 was determined by evaluating the cell cytoplasm and/or membranes, and SOX2, CDX1, and CDX2 expression was by evaluating the nucleus. The intensity and percentage of cells expressing markers were assessed. Histo-scores (H-scores) were calculated by multiplying the intensity (0 = negative; 1 = weak; 2 = moderate; 3 = strong) and the percentage of positive cells (range = 0–100), ranging from 0 to 300. For statistical analyses, we used a cutoff of 61 and 56 for CEACAM5 and TROP2 based on the average H-scores of non-neoplastic gastric lesions. Gastric lesions with H-scores higher than 61 or 56 were considered positive for CEACAM5 and TROP2, respectively.
Statistical analysis
Statistical analyses were performed using the PASW 18.0 statistical software program (IBM SPSS Statistics) and Prism version 5.0 (GraphPad Software, Inc., https://www.graphpad.com/scientific-software/prism). Between-group comparisons were performed using Tukey’s multiple comparison test. The correlation between TROP2/CEACAM5 positivity and clinicopathological characteristics were tested using Pearson’s chi-square test. Correlations between TROP2 and CEACAM5 H-scores were evaluated using the Spearman correlation test. P-value < 0.05 was considered statistically significant.
Results
Expression of TROP2 and CEACAM5 in antral intestinal metaplasia.
Our previous investigations have focused on biomarkers of IM in the corpus mucosa. In the corpus, TROP2 and CEACAM5 were upregulated in incomplete IM, but not in complete IM (14, 19). We therefore examined the expression of TROP2 and CEACAM5 in IM located in the antrum of a cohort comprising 12 patients from Singapore. Similar to the findings in the corpus mucosa,(19) TROP2 was expressed in incomplete IM, but not in complete IM (Figure 1A and Supplementary Figure S1). CEACAM5 was weakly expressed in complete IM, but was prominently expressed in a subset of TROP2 positive incomplete IM (Figure 1A and Supplementary Figure S1). In addition, we identified tissues containing incomplete IM with lineages demonstrating more dysplastic features. These IM glands demonstrated increases in CEACAM5 staining (Figure 1A, B). These findings suggested that TROP2 and CEACAM5 were significant markers of incomplete IM and up-regulation of CEACAM5 may be associated with dysplastic transition in the antrum.
Figure 1.
Expression of TROP2 and CEACM5 in intestinal metaplasia (IM) from patients from Singapore. Antral mucosal sections from 12 patients with IM were stained for TROP2, CEACAM5 and CD44v9. A Regions of normal antral mucosa showed staining for CD44v9 in deep antral glands, but there was no significant expression of either TROP2 or CEACAM5. In sections with mixed IM containing both complete and incomplete IM, TROP2 staining was observed in incomplete, but not in complete IM. In incomplete IM with dysplastic features, many glands showed staining with both TROP2 and CEACAM5 or CEACAM5 alone. Scale bar=100μm. B Quantitation of glands in mixed and dysplastic IM determined for TROP2 only, CEACAM5 only, both TROP2 and CEACAM5 or negative for either. Mean ± standard deviation.
Expression of CEACAM5 and TROP2 in complete and incomplete intestinal metaplasia in a Korean cohort.
To provide greater analysis of these markers, we next investigated the expression of CEACAM5 and TROP2 in complete and incomplete IM from a cohort of Korean patients. Normal antral mucosa exhibited a weak expression of CEACAM on the surface of foveolar epithelium, while minimal TROP2 expression was observed on the basolateral side of foveolar epithelium (Figure 2A). Inflammation induced an increased expression of both CEACAM5 and TROP2 in the foveolar epithelium (Figure 2B). In the normal small intestine, CEACAM5 was weakly expressed on the luminal side of villi with a slight accentuation in goblet cells, while no TROP expression was observed (Supplementary Figure S2A). Conversely, colonic mucosa displayed a diffuse and strong cytoplasmic CEACAM5 expression, but no TROP2 expression was observed (Supplementary Figure S2B).
Figure 2.
Expression of CEACAM5 and TROP2 in complete and incomplete intestinal metaplasia (IM) in a Korean cohort. A Immunostaining for CEACAM5 or TROP2 was performed in in normal antrum, chronic gastritis, complete and incomplete IM. Complete IM exhibits a weak expression of CEACAM5 on the apical surface of epithelium, but no TROP2 expression. Incomplete IM shows moderate to strong CEACAM5 expression and weak to moderate TROP2 expression. Red dotted boxes denote enlarged area. B Co-immunostaining for CEACAM5 and TROP2 was performed in each gastric lesion. C, D Histo-scores (H-scores) of CEACAM5 and TROP2 in various gastric lesions including normal antrum (n=3), inflamed antrum (n = 11), complete IM (n= 14), and Inc IM (n=37). Mean ± standard deviation. One-way ANOVA with Tukey’s multiple comparisons. **P < .01, ***P < .001, ****P < .0001. Scale bar=100μm.
Complete IM showed an expression pattern of CEACAM5 and TROP2 almost identical to that of the small intestine, with weak expression of CEACAM5 on the apical surface of the epithelium and no TROP2 expression (Figure 2A). In contrast, incomplete IM exhibited diffuse and high levels of CEACAM5 expression similar to colonic epithelium, while also displaying a significant amount of TROP2 expression (Figure 2A). To evaluate further the co-localization of CEACAM5 and TROP2, we performed dual immunofluorescence staining in antral mucosa with or without IM. CEACAM5 and TROP2 tended to be weakly co-expressed on the surface foveolar epithelium in normal antral mucosa without IM (Figure 2B). While complete IM only showed weak luminal CEACAM5 expression (Figure 2B), incomplete IM exhibited strong co-expression of both markers (Figure 2B). Histo-scores (H-scores) of CEACAM5 and TROP2 demonstrated that incomplete IM exhibited higher expression levels of both markers (90.5±38.9, 86.9±49.5) than complete IM (8.2±7.6, 2.0±3.0) (Figure 2C, D).
Expression of SOX2, CDX1, and CDX2 in complete and incomplete intestinal metaplasia.
SOX2 is a transcription factor involved in gastric differentiation, while CDX1 and CDX2 are responsible for intestinal differentiation (Supplementary Figure S3C). We evaluated the expression of SOX2, CDX1, and CDX2 in complete and incomplete IM. In normal antral mucosa, SOX2 was primarily localized in the isthmus/neck areas, and inflammation led to an increase in SOX expression (Figure 3A). Conversely, complete IM exhibited no detectable SOX2 expression, while incomplete IM prominently featured SOX2-positive cells (Figure 3A, B), indicative of the persistence of a gastric phenotype in incomplete IM. This observation aligns with previous findings suggesting that incomplete IM represents a gastric-intestinal mixed type (21). CDX1 and CDX2 were robustly expressed in both complete and incomplete IM (Figure 3A, C, D). Interestingly, low expression of CDX2 was observed in the inflamed condition, but was significantly more pronounced in complete IM compared to incomplete IM (Figure 3A, D). In mixed lesions, we confirmed the presence of CDX1 and CDX2 expression in both complete and incomplete IM, along with SOX2 expression in incomplete IM (Figure 3E, F). Given the consistent expression of SOX2 and CDX2 in incomplete IM, we conducted a comparative analysis of their expression with CEACAM5 or TROP2. However, we did not observe any significant correlations (Supplementary Figure S3).
Figure 3.
Expression of SOX2, CDX1, and CDX2 in the intestinal metaplasia (IM). A Immunostaining for SOX2, CDX1 and CDX2 was performed in normal gastric antral tissue, inflamed antrum (inf ant), and complete (com) and incomplete (incom) IM. Red arrows indicate SOX2-positive cells. SOX2 expression was observed in normal antral tissue with or without inflammation, and incomplete IM, but not in complete IM. B-D Histo-scores (H-scores) of SOX2, CDX1, and CDX2 expression in various gastric lesions including normal antrum (n=3), inf ant (n = 4), com IM (n= 9), and Inc IM (n=23). E, F Co-immunostaining for CDX1 or CDX2/SOX2 along with CEACAM5 and TROP2 in mixed IM. White and yellow dotted boxes indicate com IM and incom IM areas, respectively. Mean ± standard deviation. One-way ANOVA with Tukey’s multiple comparisons. **P < .01, ***P < .001, ****P < .0001. Scale bar=100μm.
Expression of CEACAM5 and TROP2 in intestinal metaplasia with dysplastic features.
IM often exhibits varying degrees of dysplastic features, including cellular pleomorphism, stratification, loss of polarity, and structural complexity. For the Korean cohort, we classified incomplete IM into two categories based on those histologic findings: simple incomplete IM, which lacks dysplastic features, and dysplastic incomplete IM, which exhibits dysplastic morphology (Figure 4A). Notably, simple incomplete IM showed relatively weak expression of CEACAM5, while TROP2 expression remained persistent (Figure 4B and Supplementary Figure S4A). In contrast, dysplastic incomplete IM exhibited moderate to strong expression of both CEACAM5 and TROP2, and CEACAM5 expression tended to extend into cytoplasm (Figure 4Band Supplementary Figure S4B). H-scores CEACAM5 and TROP2 were significantly higher in dysplastic incomplete IM (133.3±28.0, 108±33.7) compared to simple incomplete IM (72.5±19.1,60.0±41.4) (Figure 4C,D). Combined H-scores of CEACAM5 and TROP2 were also higher in the dysplastic incomplete IM (Figure 4E). These findings suggest that increased CEACAM5 and TROP2 expression may serve as a marker for dysplastic progression of incomplete IM. We further compared SOX2, CDX1, and CDX2 expression between simple and dysplastic incomplete IM and observed a lower SOX2 and higher CDX2 in dysplastic incomplete IM than in simple incomplete IM, suggesting a more pronounced intestinal phenotype in the context of dysplastic incomplete IM (Figure 4F, G).
Figure 4.
Increased expression of CEACAM5 and TROP2 in simple and dysplastic incomplete intestinal metaplasia (Inc IM). A Representative hematoxylin and eosin (H&E) images of simple and dysplastic Inc IM. Black dotted boxes denote enlarged area. B Co-immunostaining for CEACAM5 and TROP2 was performed in simple and dysplastic Inc IM. White dotted boxes denote enlarged area. C-E Histo-scores (H-scores) of CEACAM5, TROP2, and combined CEACAM5 and TROP2 expression in simple (n=15) and dysplastic (n=22) Inc IM. F-I Co-immunostaining for SOX2, CDX2, and TROP2 and their H-scores in simple a (n=15) and dysplastic (n=22) Inc IM. Mean ± standard deviation. Unpaired student t-test. *P < .05, **P < .01, ****P < .0001. ns, not significant. Scale bar = 100μM.
Expression of CEACAM5 and TROP2 in gastric adenomas and gastric cancers
We also investigated CEACAM5 and TROP2 expression in gastric tumors including GA (n=19) and early GC (n=59). CEACAM5 and TROP2 expression was found to be variable in GA (Figure 5A), with H-scores of 69.4±73.8 and 83.6±59.6, respectively, which were slightly lower than those in dysplastic incomplete IM, although not statistically significant (Figure 5B,C). Moreover, no association was observed between CEACAM5 and TROP2 expression in GA (Figure 5D). GCs also showed a heterogeneous expression of CEACAM5 and TROP2 (Figure 5E), and it was not uncommon to observe incomplete IM adjacent to GCs expressing CEACAM5 and TROP2 (Figure 5B). H-scores of CEACAM5 and TROP2 in GCs were 133.4±92.9 and 59.1±57.7, respectively, with no significant difference between intestinal type and diffuse type GC (Figure 5G). Additionally, no correlation was found between CEACAM5 and TROP2 H-scores (Figure 5H). The average H-scores for non-neoplastic lesions for CEACAM5, TROP2, and their combination were 61, 56, and 117, respectively. We considered a gastric lesion positive for CEACAM5, TROP2, and CEACAM5+TROP2 when the H-score was over 61, 56, or 117, respectively. The positive rates for CEACAM5 and TROP2 in each gastric lesion are shown in Table 1. CEACAM5 and TROP2 were negative in all normal antral tissues and complete IM, but they were positive in 9.1% of inflamed antral tissues. While simple incomplete IM showed positivity rates of 40.0% and 46.7% for CEACAM5 and TROP2, dysplastic incomplete IM exhibited 95.4 and 86.4 positivity. CEACAM5 was positive in 31.6% and 66.1% of GA and GC, while TROP2 was positive in 57.9% and 49.2% of GA and GC. When combining CEACAM5 and TROP2 expression, the positive rate of GA increased from 31.6% to 63.2%. In GC, CEACAM5 or TROP2 positivity had no significant associations with clinicopathological features such as age, gender, depth of invasion, lymph node metastasis, and H. pylori infection except that TROP2 positivity was higher in moderately differentiated GC (P=0.016) (Table 2). Additionally, we classified GCs into four classes based on combined CEACAM5 and TROP2 positivity and observed no significant association except with histology (P=0.036) (Supplementary Table S2).
Figure 5.
Expression of CEACAM5 and TROP2 in gastric adenomas (GA) and gastric cancers (GC). Co-Immunostaining for CEACAM5 or TROP2 was performed in GA (n=19) and GC (n=59). A Representative case of a GA showing both CEACAM5 and TROP2 expression.(B, C Histo-scores (H-scores) of CEACAM5 and TROP2 in simple and dysplastic incomplete intestinal metaplasia (S_Inc IM and D_Inc IM) and GA. D Correlation between H-scores of CEACAM5 and TROP2 in GA. E Representative case of a GA expressing both CEACAM5 and TROP2. F A GC (indicated by yellow arrows) as well as incomplete intestinal metaplasia lesion (indicated by red arrows) adjacent to GC displays positivity for CEACAM5 and TROP2. G H-scores of CEACAM5 and TROP2 in GC according to Lauren classification. H Correlation between H-scores of CEACAM5 and TROP2 in GC. Mean ± standard deviation. One-way ANOVA with Tukey’s multiple comparisons. ns, not significant. Scale bar=100μm.
Table 1.
Positive rates of CEACAM5 and/or TROP2 expression in gastric pre-neoplastic lesions and gastric tumors
| Diagnosis | Normal antrum (n = 3) | Inflamed antrum (n = 11) | Complete IM (n = 14) | Simple Inc IM (n = 15) | Dysplastic Inc IM (n = 22) | GA (n = 19) | GC (n = 59) |
|---|---|---|---|---|---|---|---|
| CEACAM5 (%) | 0/3 (0.0) | 1/11 (9.1) | 0/14 (0.0) | 6/15 (40.0) | 21/22 (95.4) | 6/19 (31.6) | 39/59 (66.1) |
| TROP2 (%) | 0/3 (0.0) | 1/11 (9.1) | 0/14 (0.0) | 7/15 (46.7) | 19/22 (86.4) | 11/19 (57.9) | 25/59 (42.4) |
| CEACAM5 + TROP2 (%) | 0/3 (0.0) | 1/11 (9.1) | 0/14 (0.0) | 8/15 (53.3) | 22/22 (100.0) | 12/19 (63.2) | 41/59 (69.5) |
IM, intestinal metaplasia; Inc, incomplete; GA, gastric adenoma; GC, gastric carcinoma
Table 2.
Associations of TROP2 and CEACAM5 with clinicopathological features in early gastric cancers (n =59)
| CEACAM5 | P#-value | TROP2 | P#-value | ||||
|---|---|---|---|---|---|---|---|
| Low | High | Low | High | ||||
| Case number (%) | 20 (34)) | 39 (66) | 34 (58) | 25 (42) | |||
| Age, Mean ± SD | 68 ± 10 | 62 ± 10 | 64 ± 11 | 75 ± 10 | |||
| Gender (%) | 0.952 | 0.432 | |||||
| Female | 6 (33) | 12 (67) | 9 (50) | 9 (50) | |||
| Location (%) | 0.142 | 0.501 | |||||
| Cardia | 1 (100) | 0 (0) | 1 (100) | 0 (0) | |||
| Depth of invasion (%) | 0.209 | 0.133 | |||||
| Submucosa | 16 (39) | 25 (61) | 21 (51) | 20 (49) | |||
| Histology (%) | 0.389 | 0.016 | |||||
| PCC (SRCC) | 1 (14) | 6 (86) | 6 (86) | 1 (14) | |||
| LN metastasis* (%) | 0.359 | 0.244 | |||||
| Present | 5 (50) | 5 (50) | 4 (40) | 6 (60) | |||
| H. pylori infection** | 0.678 | 0.077 | |||||
| Present | 3 (43) | 4 (57) | 6 (86) | 1 (14) | |||
Lymph node dissection was performed in 48 gastric cancer cases.
H. pylori infection status was available in 53 cases.
Pearson Chi-Square test.
no. number; SD, standard deviation; WD, well differentiated; MD, moderately differentiated; PD, poorly differentiated; PCC, poorly cohesive carcinoma; SRCC, signet ring cell carcinoma.
Discussion
Gastric IM subsequent to atrophic gastritis is a well-established precancerous lesion for the development of intestinal-type gastric adenocarcinoma (22). It has been reported that the pooled incidence rate for the progression of IM to GC was 12.4 per 10,000 person-years (23, 24). However, currently, there is no way to predict who will develop neoplastic progression. A recent study by Tan et al. has suggested that IM patients with shortened telomeres and chromosomal alterations are more likely to develop dysplasia or GC, indicating that genomic testing may be useful for risk stratification (25). However, such genomic and epigenomic profiling may be inaccessible in countries with high GC incidence rates. Thus, there is a need for cost-effective biomarkers for predicting neoplastic progression. This study aimed to evaluate the potential of CEACAM5 and TROP2, two dysplastic markers, as molecular biomarkers by analyzing their expression profiles in gastric precancerous lesions and tumors (22).
Complete IM is characterized by the expression of most or all of small intestinal digestive enzymes (4), and it closely resembles the small intestine with eosinophilic enterocytes displaying a well-defined brush border and goblet cells (6). Our findings indicate that complete IM and small intestine have a similar expression pattern of CEACAM5 and TROP2, with minimal expression of CECAM5 on the apical surface epithelium and no TROP2 expression, supporting the notion that complete IM is almost identical to small intestine. In contrast, incomplete IM is thought to resemble colonic epithelium, with multiple irregular mucin droplets of variable size in the cytoplasm and an absence of a brush border (6). In this study, we observed diffuse CEACAM5 expression in incomplete IM, similar to colonic epithelium. However, while incomplete IM exhibited significant expression of TROP2, colonic epithelium was negative for TROP2, indicating that incomplete IM does not represent colonic epithelium to the same degree that complete IM represents small intestinal epithelium. Histochemical mucin stains have shown that incomplete IM is a hybrid form expressing a mixture of gastric and intestinal lineages (5, 20, 26). This lineage confusion or complexity observed in incomplete IM may contribute to its progression to dysplasia and cancer.
We found that incomplete IM exhibited higher levels of CEACAM5 and TROP2 expression compared to normal antrum or complete IM. This increased expression of dysplastic markers in incomplete IM supports the concept that incomplete IM is more likely associated with GC. We further subdivided incomplete IM into simple and dysplastic incomplete IM based on histologic characteristics, such as thickened epithelium and structural complexity despite of some degree of subjectivity in the evaluation. Importantly, dysplastic incomplete IM showed further increases in CEACAM5 and TROP2 expression compared to simple incomplete IM, suggesting that dysplastic incomplete IM may represent a distinct type of IM. Moreover, dysplastic incomplete IM displayed heightened CDX2 expression and decreased SOX2 expression when compared to simple incomplete IM, implying that dysplastic progression in gastric IM might be associated with a stronger transcriptional direction toward the intestinal phenotype. Previous studies have proposed an association between type III incomplete IM and GC (7, 27–29). However, High iron diamine stain use has been discontinued due to its potentially toxic diamine reagents, and the association of mucin-based IM subtypes with GC risk are controversial (30). Therefore, our sub-classification of simple versus dysplastic incomplete IM, supported by the expression of dysplasia markers CEACAM5 and TROP2, may provide a valuable framework for investigating the association between incomplete IM subtypes and GC development.
Interestingly, dysplastic incomplete IM exhibited even higher positivity for CEACAM5 (100%) and TROP2 (100%) compared to both GA (31.6%, 57.9%) or GC (66.1%, 49.2%). In particular, CEACAM5 was only positive in 31.6% of GA. These results suggest that upregulation of CEACAM5 and TROP2 may be an early event in the process of gastric carcinogenesis. When comparing between GA and GC, CEACAM5 positivity increased from 31.6% to 66.1%, whereas TROP2 positivity decreased from 57.9% to 49.2%. This inconsistent patterns between GA and GC regarding CECAM5 and TROP2 positivity may be in part explained by the fact that GA rarely develop into GC (31). Furthermore, the combined use of CEACAM5 and TROP2 resulted in a slight increase in positivity for GA and GC although still lower compared to dysplastic incomplete IM.
In summary, our study highlights the heterogeneity of gastric IM and identifies dysplastic incomplete IM as a potential precursor lesion for GC. The differential expression patterns of CEACAM5 and TROP2 in different types of IM suggest their utility as potential biomarkers for risk stratification. Future studies are necessary to accurately assess the risk of CEACAM5/TROP2-high dysplastic incomplete IM for GC development.
Supplementary Material
Acknowledgements:
This work was supported by funding to JRG from grants from a Department of Veterans Affairs Merit Review Award IBX000930, DOD CA190172, NIH R01 DK101332 and NCI R01 CA272687, and to BJ from grants from National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT, 2021R1C1C1011172 and 2020R1I1A1A01069168). This research is also supported by the National Research Foundation, Singapore, and Singapore Ministry of Health’s National Medical Research Council under its Open Fund-Large Collaborative Grant (“OF-LCG”) (MOH-OFLCG18May-0003). The Vanderbilt Digital Histology Shared Resource is supported by VA Shared Equipment grant 1IS1BX003097 and the Vanderbilt Digestive Disease Research Center (P30 DK058404).
References
- 1.Shah SC, Gawron AJ, Mustafa RA, Piazuelo MB. Histologic subtyping of gastric intestinal metaplasia: overview and considerations for clinical practice. Gastroenterology. 2020;158(3):745–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Fennerty MB. Gastric intestinal metaplasia on routine endoscopic biopsy. Gastroenterology. 2003;125(2):586–90. [DOI] [PubMed] [Google Scholar]
- 3.de Vries AC, Haringsma J, de Vries RA, ter Borg F, Nagtzaam NM, Steyerberg EW, et al. The use of clinical, histologic, and serologic parameters to predict the intragastric extent of intestinal metaplasia: a recommendation for routine practice. Gastrointestinal endoscopy. 2009;70(1):18–25. [DOI] [PubMed] [Google Scholar]
- 4.Matsukura N, Suzuki K, Kawachi T, Aoyagi M, Sugimura T, Kitaoka H, et al. Distribution of marker enzymes and mucin in intestinal metaplasia in human stomach and relation of complete and incomplete types of intestinal metaplasia to minute gastric carcinomas. Journal of the National Cancer Institute. 1980;65(2):231–40. [PubMed] [Google Scholar]
- 5.Jass J, Filipe M. The mucin profiles of normal gastric mucosa, intestinal metaplasia and its variants and gastric carcinoma. The Histochemical journal. 1981;13:931–9. [DOI] [PubMed] [Google Scholar]
- 6.Correa P, Piazuelo MB, Wilson KT. Pathology of gastric intestinal metaplasia: clinical implications. The American journal of gastroenterology. 2010;105(3):493. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Filipe MI, Muñoz N, Matko I, Kato I, Pompe‐Kirn V, Jutersek A, et al. Intestinal metaplasia types and the risk of gastric cancer: a cohort study in Slovenia. International journal of cancer. 1994;57(3):324–9. [DOI] [PubMed] [Google Scholar]
- 8.Rokkas T, Filipe M, Sladen G. Detection of an increased incidence of early gastric cancer in patients with intestinal metaplasia type III who are closely followed up. Gut. 1991;32(10):1110–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Silva S, Filipe M. Intestinal metaplasia and its variants in the gastric mucosa of Portuguese subjects: a comparative analysis of biopsy and gastrectomy material. Human pathology. 1986;17(10):988–95. [DOI] [PubMed] [Google Scholar]
- 10.Gold P, Freedman SO. Specific carcinoembryonic antigens of the human digestive system. The Journal of experimental medicine. 1965;122(3):467–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Hammarström S, editor The carcinoembryonic antigen (CEA) family: structures, suggested functions and expression in normal and malignant tissues. Seminars in cancer biology; 1999: Elsevier. [DOI] [PubMed] [Google Scholar]
- 12.Liu J-N, Wang H-B, Zhou C-C, Hu S-Y. CEACAM5 has different expression patterns in gastric non-neoplastic and neoplastic lesions and cytoplasmic staining is a marker for evaluation of tumor progression in gastric adenocarcinoma. Pathology-Research and Practice. 2014;210(10):686–93. [DOI] [PubMed] [Google Scholar]
- 13.Zhou J, Fan X, Chen N, Zhou F, Dong J, Nie Y, Fan D. Identification of CEACAM5 as a biomarker for prewarning and prognosis in gastric cancer. Journal of Histochemistry & Cytochemistry. 2015;63(12):922–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Lee SH, Contreras Panta EW, Gibbs D, Won Y, Min J, Zhang C, et al. Apposition of Fibroblasts With Metaplastic Gastric Cells Promotes Dysplastic Transition. Gastroenterology. 2023;165(2):374–90. Epub 20230515. doi: 10.1053/j.gastro.2023.04.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lipinski M, Parks DR, Rouse RV, Herzenberg LA. Human trophoblast cell-surface antigens defined by monoclonal antibodies. Proceedings of the National Academy of Sciences. 1981;78(8):5147–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Lenárt S, Lenárt P, Šmarda J, Remšík J, Souček K, Beneš P. Trop2: Jack of all trades, master of none. Cancers. 2020;12(11):3328. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Mühlmann G, Spizzo G, Gostner J, Zitt M, Maier H, Moser P, et al. TROP2 expression as prognostic marker for gastric carcinoma. Journal of clinical pathology. 2009;62(2):152–8. [DOI] [PubMed] [Google Scholar]
- 18.Zhao W, Zhu H, Zhang S, Yong H, Wang W, Zhou Y, et al. Trop2 is overexpressed in gastric cancer and predicts poor prognosis. Oncotarget. 2016;7(5):6136. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Riera KM, Jang B, Min J, Roland JT, Yang Q, Fesmire WT, et al. Trop2 is upregulated in the transition to dysplasia in the metaplastic gastric mucosa. The Journal of pathology. 2020;251(3):336–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Lee S, Jang B, Min J, Contreras-Panta EW, Presentation KS, Delgado AG, et al. Upregulation of AQP5 defines spasmolytic polypeptide-expressing metaplasia (SPEM) and progression to incomplete intestinal metaplasia. Cell Molec Gastroenterol Hepataol. 2022;13:199–217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Tsukamoto T, Mizoshita T, Tatematsu M. Gastric-and-intestinal mixed-type intestinal metaplasia: aberrant expression of transcription factors and stem cell intestinalization. Gastric Cancer. 2006;9:156–66. [DOI] [PubMed] [Google Scholar]
- 22.Park YH, Kim N. Review of atrophic gastritis and intestinal metaplasia as a premalignant lesion of gastric cancer. Journal of cancer prevention. 2015;20(1):25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Gawron AJ, Shah SC, Altayar O, Davitkov P, Morgan D, Turner K, Mustafa RA. AGA technical review on gastric intestinal metaplasia—natural history and clinical outcomes. Gastroenterology. 2020;158(3):705–31. e5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Altayar O, Davitkov P, Shah SC, Gawron AJ, Morgan DR, Turner K, Mustafa RA. AGA technical review on gastric intestinal metaplasia—epidemiology and risk factors. Gastroenterology. 2020;158(3):732–44. e16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Huang KK, Ramnarayanan K, Zhu F, Srivastava S, Xu C, Tan ALK, et al. Genomic and epigenomic profiling of high-risk intestinal metaplasia reveals molecular determinants of progression to gastric cancer. Cancer cell. 2018;33(1):137–50. e5. [DOI] [PubMed] [Google Scholar]
- 26.Riera KM, Jang B, Min J, Roland JT, Yang Q, Fesmire WT, et al. Trop2 is upregulated in the transition to dysplasia in the metaplastic gastric mucosa. Journal of Pathology. 2020;251(3):336–47. doi: 10.1002/path.5469. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Filipe M, Potet F, Bogomoletz W, Dawson P, Fabiani B, Chauveinc P, et al. Incomplete sulphomucin-secreting intestinal metaplasia for gastric cancer. Preliminary data from a prospective study from three centres. Gut. 1985;26(12):1319–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Conchillo J, Houben G, De Bruine A, Stockbrügger R. Is type III intestinal metaplasia an obligatory precancerous lesion in intestinal-type gastric carcinoma? European journal of cancer prevention. 2001:307–12. [DOI] [PubMed] [Google Scholar]
- 29.Piazuelo MB, Camargo MC, Mera RM, Delgado AG, Bravo JC, Peek RM, et al. Gastric intestinal metaplasia type III and prospective risk of gastric cancer in colombia. Gastroenterology. 2017;152(5):S473. [Google Scholar]
- 30.Dinis-Ribeiro M, Areia M, De Vries A, Marcos-Pinto R, Monteiro-Soares M, O’connor A, et al. Management of precancerous conditions and lesions in the stomach (MAPS): guideline from the European Society of Gastrointestinal Endoscopy (ESGE), European Helicobacter Study Group (EHSG), European Society of Pathology (ESP), and the Sociedade Portuguesa de Endoscopia Digestiva (SPED). Endoscopy. 2012;44(01):74–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Yamada H, Ikegami M, Shimoda T, Takagi N, Maruyama M. Long-term follow-up study of gastric adenoma/dysplasia. Endoscopy. 2004;36(05):390–6. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.