These results support a mechanistic link between soluble bacterial products and inflammatory signaling, but they do not establish that S. anginosus causes gastric cancer in people. The investigators combined observations in human stomach tissue with experiments involving macrophages, gastric cancer cells, and gastric epithelial cells. They examined bacterial abundance, macrophage infiltration, and pyroptosis-related marker expression in tissue from people without Helicobacter pylori infection who had chronic superficial gastritis, intestinal metaplasia, or gastric cancer.

S. anginosus abundance, macrophage infiltration, and markers associated with the NLRP3/interleukin-1 beta pathway were higher in the intestinal metaplasia and gastric cancer groups than in the chronic superficial gastritis group. These measures were positively correlated. Because the tissue comparison was observational, it cannot establish whether the bacteria, macrophage activation, and disease-associated changes cause one another or whether other factors associated with more advanced gastric disease contribute to the pattern.

RNA sequencing of macrophages treated with S. anginosus supernatant identified 1,175 differentially expressed genes, including 535 that were upregulated and 640 that were downregulated. NLRP3 and interleukin-1 beta were significantly upregulated, whereas caspase-1 and GSDMD showed no significant changes at the transcriptional level. Pathway analysis identified enrichment of NOD-like receptor, tumor necrosis factor, interleukin-17, nuclear factor-kappa B, and Toll-like receptor signaling.

Functional assays provided evidence beyond the transcript measurements by showing increased inflammatory cytokine release, ASC speck formation, and expression of cleaved caspase-1, GSDMD-N, and interleukin-1 beta. Conditioned medium collected from S. anginosus supernatant-treated macrophages increased epithelial-to-mesenchymal transition-associated markers in gastric cancer cells and intestinal metaplasia-associated markers in gastric epithelial cells. Partial attenuation of these effects by MCC950, an inhibitor of NLRP3-related signaling, supported the involvement of that pathway.

The study did not directly test whether interleukin-1 beta was required for the downstream changes, so the increased expression and release of this cytokine cannot be assigned a specific causal role. Untargeted metabolomics identified 15(S)-HPETE as a differentially abundant metabolite associated with the supernatant treatment. Functional assays found that 15(S)-HPETE enhanced malignant phenotype-associated responses in AGS gastric cancer cells, promoted pyroptosis-associated activation in macrophages, and increased intestinal metaplasia-associated marker expression in gastric epithelial cells.

This supports 15(S)-HPETE as a candidate metabolic mediator, but it does not establish the complete mechanism connecting bacterial supernatant exposure to the cellular responses. Overall, the findings are consistent with a model in which soluble products associated with S. anginosus activate inflammatory macrophage responses involving NLRP3-related signaling, while material released by the treated macrophages promotes tumor-associated and intestinal metaplasia-associated features in epithelial cells. The evidence comes from a combination of human tissue associations and in vitro mechanistic experiments, not from a longitudinal clinical study or an animal disease model.

The reported results therefore cannot demonstrate patient outcomes, quantify cancer risk, or show that blocking the pathway would prevent progression.

Subject of Research: In vitro effects of Streptococcus anginosus supernatant on macrophage activation linked to gastric inflammation-to-cancer phenotypes

Article Title: Streptococcus anginosus supernatant promotes macrophage pyroptosis-associated activation and phenotypes associated with the gastric inflammation-to-cancer transition via NLRP3-related signaling

Article References: Han, X., Ji, J., Cao, D., Zhu, L., Liu, Z., Hui, Y., Yang, M., Lin, L., Yu, Z., Yuan, J., Yao, Q., Zhong, W., Zhang, Q., & Chen, X. (2026). Streptococcus anginosus supernatant promotes macrophage pyroptosis-associated activation and phenotypes associated with the gastric inflammation-to-cancer transition via NLRP3-related signaling. Journal of Translational Medicine. https://doi.org/10.1186/s12967-026-09060-7

Image Credits: AI Generated

DOI: 10.1186/s12967-026-09060-7

Keywords: Streptococcus anginosus, gastric cancer, intestinal metaplasia, macrophages, NLRP3 signaling, pyroptosis, interleukin-1 beta, 15(S)-HPETE, in vitro research, Streptococcus, anginosus, supernatant

The research draws on a cohort of 620 subjects, including 219 patients with gastric cancer and more than 400 non-cancer individuals, among them 103 peptic ulcer patients and 299 controls without ulcers. From this group, the team defined a microbiome-based cancer risk using 47 specific bacterial taxa that they had previously shown to be significantly altered between gastric cancer patients and non-ulcer controls. A microbiome index above 0.25 classified a subject as carrying the high-risk microbial signature. In parallel, genetic risk was defined by the presence of two single nucleotide polymorphisms that genome-wide association studies in the Japanese population had linked to gastric cancer susceptibility: rs2294008 in the PSCA gene and rs4072037 in the MUC1 gene. Individuals carrying both the PSCA T/T and MUC1 A/A risk genotypes were classified as genetically predisposed.

The results were unambiguous. People whose gastric microbiomes carried the high-risk signature alone showed an odds ratio of 11.16 for gastric cancer, while those with the high-risk genotype alone showed an odds ratio of 2.76. But when both factors were present, the odds ratio climbed to 23.55—a synergistic leap that suggests the two risk factors interact biologically rather than acting independently. Crucially, the researchers found no association between the prevalence of the high-risk genotypes and the microbiome index itself, which argues against the possibility that one factor simply confounds the other. Instead, the data point to a genuine functional interplay between host genetics and the microbial ecosystem lining the stomach.

The synergy extended beyond cancer diagnosis to specific disease subtypes. The combination of risk genotypes and high-risk microbiome was associated with an odds ratio of 29.67 for intestinal-type gastric cancer and an even higher odds ratio of 30.72 for advanced-stage disease, compared with non-cancer controls. These are among the strongest risk signals ever reported for combined host-microbiome profiling in gastric carcinogenesis, and they suggest that the interaction between inherited susceptibility and bacterial community structure may be particularly consequential for the slower, inflammation-driven pathway that produces the intestinal type of the disease.

To probe the biology behind these statistics, the researchers examined somatic changes in the gastric mucosa—alterations that accumulate in stomach tissue itself over years of injury and repair. Three markers stood out: the histological severity of intestinal metaplasia, a precancerous transformation of the stomach lining; the accumulation of DNA methylation, an epigenetic change long associated with H. pylori infection and cancer risk; and telomere shortening, an indicator of rapid cell turnover and oxidative stress that has been implicated in tumor development. Across all three measures, the team observed a stepwise increase in severity according to the combined burden of genetic and microbial risk. Individuals with both risk factors showed the highest degrees of intestinal metaplasia and DNA methylation and the shortest telomeres, painting a molecular portrait of a stomach pushed progressively toward malignancy.

These findings suggest a plausible mechanistic narrative. The PSCA rs2294008 T allele and the MUC1 rs4072037 A allele both alter upstream promoter activity of their respective genes—PSCA encodes a glycosylphosphatidylinositol-anchored cell surface antigen, while MUC1 encodes a membrane-bound mucin protein—and both variants have been shown to confer susceptibility to gastric cancer, particularly the diffuse type. The new work implies that when these genetic variants are present, the cancer-predisposing microbial community may sustain or amplify the genotoxic changes that H. pylori infection initiates, promoting long-term tumorigenesis even as the initial infection evolves. In other words, the bacteria may not merely be passengers in an inflamed stomach; they may be active collaborators with the host genome in driving cells toward cancer.

Support for this idea comes from the team’s earlier work, which characterized gastric microbiome signatures across a large Japanese cohort and identified community-wide bacterial networks tied to cancer risk and progression. That study found that the cancer-predisposing microbiome signature is significantly associated with decreased levels of superoxide reductase and superoxide dismutase, enzymes that protect cells from oxidative damage. The suppression of these antioxidant defenses would leave the gastric epithelium exposed to heightened oxidative stress, creating exactly the kind of environment in which DNA methylation errors accumulate and telomeres erode. The authors caution, however, that the precise mechanisms by which the predisposing microbiome promotes tumorigenesis in the presence of PSCA and MUC1 risk genotypes remain to be fully elucidated, and they call for well-designed in vivo and ex vivo studies to clarify these host-microbe interactions.

The clinical implications are considerable, particularly for East Asian populations. More than half of the Japanese population carries both risk genotypes, which means that genetic testing alone would flag a large swath of people without meaningfully narrowing down who is truly at risk. The new findings suggest that layering microbiome profiling on top of the two SNPs could dramatically improve risk stratification, identifying a much smaller subpopulation in which the two risk factors converge and whose members might benefit from intensified endoscopic surveillance or targeted eradication strategies. Because gastric cancer is frequently diagnosed at advanced stages, when survival rates plummet, any tool that reliably identifies high-risk individuals before symptoms appear could translate into earlier detection and saved lives.

The study also reframes the role of the non-H. pylori gastric microbiota in cancer biology. While H. pylori infection remains the dominant established risk factor for gastric tumorigenesis, metagenomic studies in recent years have revealed that a diverse community of other bacteria resides in the gastric mucosa and that dysbiotic, cancer-associated microbial communities exist independently of the famous pathogen. What has been missing is evidence that this broader ecosystem interacts with host genetics to shape disease outcomes. The Japanese cohort now provides that evidence in quantitative form, demonstrating that the microbiome is not a passive bystander but a partner in a synergistic relationship that can multiply cancer risk more than twentyfold.

Limitations and open questions remain. The study was conducted in a Japanese cohort, and the two SNPs examined were identified through genome-wide association studies in that population, so generalization to other ethnic groups will require replication. The microbiome index was built from 47 bacterial taxa whose functional roles in carcinogenesis are only beginning to be understood, and the cross-sectional design cannot fully disentangle whether the predisposing microbiome is a cause or a consequence of the mucosal changes it accompanies. Still, the convergence of microbial signatures, host genotypes, and somatic molecular markers—methylation, telomere attrition, and metaplasia—into a single coherent risk model marks a significant advance. If larger prospective studies confirm the synergy reported here, the combination of a simple genetic test with gastric microbiome profiling could become a practical diagnostic tool, ushering in an era in which stomach cancer prevention is tailored not just to who patients are, but to what lives inside them.

Subject of Research: Interaction between gastric microbiome composition and host PSCA/MUC1 genetic variants in predisposition to gastric cancer

Article Title: Synergistic effect of gastric microbiome and host genetic factor predispose to gastric cancer

Article References: Shimogama, T., Tahara, T., Shijimaya, T., Yamazaki, J., Yano, N., Ohtsu, T., Yasuoka, R., Kobayashi, S., Nakamura, N., Takahashi, Y., Tomiyama, T., Honzawa, Y., Fukui, T., & Naganuma, M. (2026). Synergistic effect of gastric microbiome and host genetic factor predispose to gastric cancer. Journal of Translational Medicine, 24(1), Article 1300. https://doi.org/10.1186/s12967-026-09009-w

Image Credits: AI Generated

DOI: 10.1186/s12967-026-09009-w

Keywords: gastric cancer, gastric microbiome, Helicobacter pylori, PSCA, MUC1, single nucleotide polymorphisms, DNA methylation, telomere shortening, intestinal metaplasia, risk stratification, host-microbe interaction, oxidative stress

The researchers, led by investigators at Nanjing Medical University, began by building a single-cell transcriptomic atlas spanning the disease continuum. They collected tissue from five patients undergoing gastrectomy for intestinal-type gastric cancer, sampling not only the tumors themselves but also matched intestinal metaplasia lesions near the tumor margin and non-atrophic gastritis tissue farther away. After rigorous quality control, the dataset comprised 55,785 individual cells from 13 specimens, which the team sorted into nine major cell types, including epithelial cells, malignant cells, T cells, B cells, plasma cells, myeloid cells, mast cells, fibroblasts, and endothelial cells. The composition shifted tellingly across disease stages: as lesions progressed, the fractions of myeloid cells, endothelial cells, and fibroblasts expanded, while malignant epithelial cells came to dominate the tumor samples.

Within the epithelial compartment, the team re-clustered more than 34,000 cells into eight subtypes, ranging from chief and parietal cells, which dwindled as disease advanced, to tumor cells, which appeared only in cancerous tissue. The pivotal move came from pseudotime trajectory analysis, a computational technique that orders cells along an inferred developmental path based on their gene expression. The trajectory originated in neck cells and flowed toward enterocytes, the absorptive cells characteristic of intestinal metaplasia. One enterocyte population, which the researchers designated Enterocytes-S3, sat squarely between earlier epithelial states and the malignant cells along this trajectory, and tumor cells were heavily enriched in the same transcriptional state. To corroborate this intermediate identity, the team scored each cell against a published signature of incomplete intestinal metaplasia. The score climbed steadily from neck cells through successive enterocyte states, peaked in Enterocytes-S3, and then fell in cancer cells, exactly the pattern expected of a waystation on the road to malignancy.

What makes Enterocytes-S3 dangerous, the data suggest, is its cell-cycle program. Weighted gene co-expression network analysis identified 25 gene modules across the epithelial subtypes, and one of them, Module 4, was strongly upregulated in Enterocytes-S3 and enriched for genes governing mitotic nuclear division and mitotic cell-cycle phase transitions. The subpopulation also showed elevated activity of the G2/M checkpoint and E2F target gene sets, hallmarks of cells primed to proliferate. In short, this transitional population was not merely drifting toward a cancer-like identity; it was rewiring its division machinery, acquiring the proliferative momentum that tumors require.

To find the conductor of that rewiring, the researchers scored bulk transcriptomes from The Cancer Genome Atlas stomach adenocarcinoma cohort against the Enterocytes-S3 signature. PTTG1 emerged as the most significantly upregulated gene in tumors with high signature scores. The finding held up across multiple independent datasets, including the Asian Cancer Research Group cohort and the team’s own in-house transcriptomic data, with paired analyses confirming that PTTG1 expression was elevated in intestinal-type gastric cancer relative to matched normal mucosa. Pseudotime analysis showed the gene’s expression rising progressively from enterocytes toward malignant cells, and immunohistochemistry on a tissue microarray of 88 evaluable paired samples told the same story at the protein level: strong nuclear PTTG1 staining in tumors, weak or absent staining in adjacent normal tissue, and significantly higher levels in intestinal-type than in diffuse-type gastric cancer.

PTTG1 is no stranger to oncology. Originally discovered in rat pituitary tumor cells, the gene encodes securin, an inhibitor that prevents sister chromatids from separating until mitosis is properly timed, thereby safeguarding genomic integrity. It is barely expressed in most normal tissues apart from the testis, yet it is markedly overexpressed in hepatocellular, breast, lung, colon, testicular, and ovarian cancers, where it has been linked to proliferation, invasion, and metastasis through pathways ranging from PI3K/AKT to Wnt/beta-catenin and MAPK signaling. What the new study adds is a specific, mechanistic account of how PTTG1 reprograms the cell cycle in gastric epithelium, and evidence that its activity tracks the premalignant-to-malignant transition in a subtype-specific manner.

The functional experiments were decisive. When the team silenced PTTG1 in MKN1 and MKN7 gastric cancer cell lines using small interfering RNAs, proliferation measured by CCK-8 and EdU assays dropped significantly, and Transwell assays showed reduced migration and invasion. RNA sequencing of the knockdown cells revealed widespread downregulation of cell-cycle genes, with gene set enrichment analysis pinpointing the G2/M checkpoint pathway. Flow cytometry then delivered a subtle but important insight: PTTG1 depletion reduced the proportion of cells in G2 phase and increased the G1 fraction, indicating impaired progression through the G2/M transition rather than a classical arrest at the checkpoint. Western blotting confirmed the molecular signature of stalled mitotic entry, with Cyclin B1 and phosphorylated CDK1 both diminished, while levels of gamma-H2AX, a marker of DNA double-strand breaks, rose sharply, and immunofluorescence showed increased gamma-H2AX foci in the nuclei of PTTG1-deficient cells.

Digging deeper, the researchers traced a regulatory axis from PTTG1 to the machinery of mitosis. Knockdown reduced the expression of E2F1, a transcription factor known to bind directly the promoter of the CCNB1 gene, which encodes Cyclin B1. Using CUT&RUN-qPCR, a technique that maps where a protein sits on chromatin, the team showed that E2F1 occupied the CCNB1 promoter in control cells but that this occupancy fell markedly after PTTG1 silencing, accompanied by reduced CCNB1 messenger RNA. The model that emerges is one in which PTTG1 sustains E2F1-dependent transcription of Cyclin B1, thereby licensing the G2-to-M transition, while simultaneously helping to coordinate cell-cycle progression with DNA damage responses. The precise mechanism by which PTTG1 controls E2F1 remains an open question the authors acknowledge.

Two model systems extended the story beyond cell lines. In human paracancerous gastric organoids derived from intestinal-type gastric cancer patients, lentiviral overexpression of PTTG1 produced organoids that grew significantly larger than controls by days 6, 10, and 12 of culture, with more Ki67-positive proliferating cells and elevated Cyclin B1, much of it in the nucleus, a location associated with mitotic commitment. Strikingly, the overexpressing organoids also stained more intensely with Alcian blue and produced more MUC2, a goblet-cell mucin characteristic of intestinal metaplasia, suggesting that PTTG1 can push gastric epithelium toward an intestinal, mucin-producing identity. In the opposite direction, stable PTTG1 knockdown in MKN1 cells injected subcutaneously into nude mice markedly suppressed tumor growth, with smaller volumes and lower weights at the experimental endpoint, reduced Ki67 and Cyclin B1 staining, and altered AB-PAS mucin staining that, however, was not accompanied by a change in MUC2 expression, hinting that PTTG1’s effects on mucin phenotypes depend on the biological context.

The study is candid about its limits. Tissue dissociation erases the spatial geography of the gastric gland, the cross-sectional sampling of five patients cannot fully capture the temporal dynamics of a cascade that unfolds over years, and the cohort was too small to assess sex-associated differences. The tumor-initiating potential of Enterocytes-S3 remains to be proven, and more physiologically relevant culture systems will be needed to settle the mucin question. Notably, PTTG1 expression showed no significant association with overall survival in the TCGA or ACRG cohorts, suggesting the gene functions as an engine of tumor-intrinsic biology rather than an independent prognostic marker. Even so, by placing a chromosome-segregation guardian at the hinge between intestinal metaplasia and carcinoma, the work offers the clearest single-cell-resolved view yet of how the Correa cascade accelerates, and it flags PTTG1, and the G2/M checkpoint it commands, as a target worth pursuing in the fight against intestinal-type gastric cancer.

Subject of Research: The role of PTTG1 in G2/M checkpoint-mediated cell-cycle reprogramming during intestinal-type gastric cancer progression

Article Title: PTTG1 drives intestinal-type gastric cancer progression via G2/M checkpoint-mediated cell-cycle reprogramming

Article References: Hu, C., Bian, L., Luo, L., Zheng, Z., Chen, Y., Gao, X., Li, F., Yang, D., Liu, Y., Li, G., Xu, H., Yan, C., & Jin, G. (2026). PTTG1 drives intestinal-type gastric cancer progression via G2/M checkpoint-mediated cell-cycle reprogramming. iScience, 29(11), Article 117742. https://doi.org/10.1016/j.isci.2026.117742

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.117742

Keywords: gastric cancer, PTTG1, securin, G2/M checkpoint, single-cell RNA sequencing, intestinal metaplasia, Correa cascade, Cyclin B1, E2F1, organoids, xenograft, cell cycle

The stakes of this distinction are considerable. Intestinal metaplasia, or IM, is a well-recognized precursor state in the Correa cascade of gastric carcinogenesis, in which chronic inflammation drives the stomach’s normal epithelium to take on intestinal characteristics. Most patients with IM will never develop cancer, but their risk is elevated enough that international guidelines, including the 2025 MAPS III update from the European Society of Gastrointestinal Endoscopy, recommend structured endoscopic surveillance. Gastric cancer, by contrast, demands immediate oncological workup and treatment. When a patient with psychological symptoms arrives complaining of vague abdominal discomfort, weight loss, or appetite changes, deciding how urgently to pursue endoscopy and biopsy can determine whether a tumor is caught at a curable stage or discovered only after it has advanced.

The research team designed a retrospective, single-center, cross-sectional study that enrolled 302 patients with psychological symptoms, of whom 185 had intestinal metaplasia and 117 had gastric cancer confirmed by standard diagnostic pathways. The cohort was randomly split into a training set of 212 patients and an independently retained validation set of 90 patients, a design that guards against the optimistic bias that plagues many machine learning studies in medicine. Rather than feeding every available variable into their algorithms, the investigators first pruned the feature space using three complementary selection methods: least absolute shrinkage and selection operator regression, known as LASSO, the Boruta algorithm, and recursive feature elimination. This consensus approach converged on seven features that carried the most diagnostic signal: age, serum albumin, sex, the type of psychological symptom, total bilirubin, smoking status, and monocyte count.

With those seven variables in hand, the team compared seven different machine learning algorithms using nested cross-validation, a rigorous technique in which an inner loop tunes the model’s hyperparameters while an outer loop estimates how the tuned model will perform on unseen data. Nested cross-validation is widely regarded as the gold standard for honest performance estimation, yet it remains underused in clinical prediction research. Among the contenders, XGBoost, a gradient-boosted decision tree ensemble celebrated for its performance on tabular medical data, emerged as the winner with a mean outer-fold area under the receiver operating characteristic curve of 0.839, plus or minus 0.029. The AUC, a measure of how well a model ranks diseased patients above healthy ones across all thresholds, ranges from 0.5, equivalent to a coin flip, to 1.0, perfect discrimination.

When the final XGBoost model was unleashed on the 90-patient validation cohort it had never seen, it achieved an AUC of 0.793, with a 95 percent confidence interval spanning 0.670 to 0.895. Its overall accuracy reached 0.800, with a specificity of 0.891, meaning it correctly identified most patients with intestinal metaplasia, and a sensitivity of 0.657, meaning it caught roughly two-thirds of the actual cancers. The F1 score, which balances precision and recall, came in at 0.719, and the Brier score, a measure of calibration that penalizes both wrong predictions and misplaced confidence, was 0.167. Decision curve analysis, a method that quantifies the net clinical benefit of acting on a model’s predictions across a range of threshold probabilities, suggested that using the model could yield meaningful benefit compared with treating all patients or none.

What elevates this study above the crowded field of medical machine learning papers is its commitment to interpretability. Black-box models have earned well-deserved skepticism from clinicians who cannot act on predictions they cannot understand. The Qingdao team therefore applied SHAP, or Shapley Additive Explanations, a technique borrowed from cooperative game theory that assigns each feature a contribution value for every individual prediction. The SHAP analysis revealed that age and the type of psychological symptom were the leading drivers of the model’s classifications, followed by the laboratory markers. This transparency allows physicians to see not just what the model predicts but why, and to judge whether those reasons align with clinical plausibility.

The feature list itself tells an intriguing biological story. Serum albumin, a protein synthesized by the liver, tends to fall in the context of both malignancy-associated inflammation and malnutrition, and low albumin has repeatedly been linked to poor outcomes in gastric cancer. Total bilirubin, another hepatic marker, has been associated with the clinical characteristics of gastric cancer patients in prior hematological studies. Monocytes, the innate immune cells that circulate in blood and seed tumors as macrophages, are increasingly recognized as players in the tumor microenvironment, and an elevated monocyte count can reflect the systemic inflammatory milieu of an active cancer. Smoking status and sex are established epidemiological risk factors, while age simply reflects the cumulative probability of neoplastic progression. The prominence of psychological symptom type is perhaps the most novel finding, hinting that the pattern of a patient’s distress may carry diagnostic information that has been largely overlooked.

The connection between psychological symptoms and gastric pathology is not merely coincidental. Previous research has documented a high prevalence of psychological distress among gastric cancer patients, and experimental work has shown that chronic stress can accelerate gastric cancer progression, with recent studies implicating gut microbial metabolites in stress-driven tumor growth. Conversely, malnutrition and systemic inflammation associated with advanced malignancy can themselves produce depressive and anxious symptoms, creating a bidirectional loop. Patients with functional gastrointestinal disorders and precancerous conditions also experience elevated rates of anxiety and depression, which is precisely why the distinction between IM and cancer in this population is so difficult and so consequential.

The authors are careful, appropriately so, about the limits of their work. The model showed only moderate discrimination, and the study was retrospective and single-center, drawing on patients from one Chinese hospital. The team explicitly states that the model should not replace endoscopic or histopathological diagnosis, which remain the definitive standards, and that external multicenter validation is required before any broader clinical implementation. Psychological symptom classification in the study relied on established rating instruments, including the Hamilton scales for depression and anxiety, but symptom type as a predictor will need careful operationalization in other healthcare settings and languages.

Even with those caveats, the study points toward a genuinely useful clinical role. Imagine a triage tool embedded in the electronic health record that, for a patient with psychological symptoms and dyspeptic complaints, computes a probability of current gastric cancer from seven routinely collected variables. A high score could accelerate the endoscopy queue; a low score could support a more measured surveillance pathway consistent with guidelines for metaplasia. In health systems where endoscopic resources are scarce and waiting lists are long, such prioritization could translate directly into earlier cancer detection. The Qingdao model is not yet that tool, but it is a credible, honestly evaluated step along the way, and its emphasis on interpretability sets a standard that clinical machine learning studies would do well to follow. As gastric cancer continues to impose a heavy global burden, with incidence and mortality projected to remain substantial through 2035, every legitimate advance in distinguishing the benign from the malignant deserves attention.

Subject of Research: Machine learning for distinguishing gastric cancer from intestinal metaplasia in patients with psychological symptoms

Article Title: A machine learning model for distinguishing gastric cancer from intestinal metaplasia in patients with psychological symptoms

Article References: Wang, S.-R., Mao, Y.-Q., & Hu, G.-J. (2026). A machine learning model for distinguishing gastric cancer from intestinal metaplasia in patients with psychological symptoms. Cancer Causes & Control, 37(10), Article 174. https://doi.org/10.1007/s10552-026-02251-z

Image Credits: AI Generated

DOI: 10.1007/s10552-026-02251-z

Keywords: gastric cancer, intestinal metaplasia, machine learning, XGBoost, SHAP, psychological symptoms, endoscopy, predictive markers, nested cross-validation, clinical decision support, serum albumin, monocytes

The research, published in the journal Gut Pathogens, tracked patients with Helicobacter pylori–associated precancerous stomach lesions over two years, repeatedly measuring levels of interleukin-8, a pro-inflammatory chemokine long implicated in the biology of gastric carcinogenesis. The findings indicate that the trajectory of IL-8 over time, rather than a single snapshot, carries meaningful predictive information about which patients will progress toward malignancy. In an era when precision oncology increasingly depends on molecular signals rather than gross symptoms, the study adds a potentially valuable piece to the biomarker puzzle.

The scientific rationale behind the work rests on the well-characterized inflammatory cascade that H. pylori initiates in the gastric mucosa. This bacterium colonizes the stomach lining of roughly half the global population, and while most carriers never fall ill, a subset develops a stepwise sequence of pathology: chronic gastritis, atrophic gastritis, intestinal metaplasia, and eventually adenocarcinoma. Interleukin-8 sits at a critical junction in this process. The chemokine is secreted by gastric epithelial cells and macrophages in response to H. pylori infection and acts as a potent recruiter of neutrophils, sustaining the chronic inflammatory milieu that drives DNA damage, cell turnover, and, ultimately, neoplastic transformation. Tumor cells themselves also produce IL-8, where it promotes angiogenesis, proliferation, and metastatic behavior, which makes the molecule interesting both as a consequence of inflammation and as a possible engine of cancer progression.

To test whether circulating IL-8 could serve as a practical surveillance tool, the investigators designed a multicenter, prospective longitudinal cohort study spanning five Egyptian centers. They enrolled 200 participants distributed across five groups: 50 patients with atrophic gastritis, 50 with intestinal metaplasia, 25 with H. pylori–positive gastric cancer, 25 with H. pylori–negative gastric cancer, and 50 healthy controls. Every participant underwent baseline clinical assessment, laboratory evaluation including the established tumor markers CEA and CA19-9, H. pylori testing by stool antigen assay and urea breath test performed according to the Maastricht VI consensus guidelines, and endoscopy with histopathology interpreted under the Updated Sydney System. Critically, the 100 patients with precancerous lesions were then followed for 24 months, with serum IL-8 measured at 0, 6, 12, 18, and 24 months and a repeat endoscopy performed at the end of the observation period.

The baseline cross-sectional results alone were striking. Mean serum IL-8 climbed in a stepwise fashion across the disease spectrum: healthy controls averaged 9.7 picograms per milliliter, patients with atrophic gastritis 42.1, those with intestinal metaplasia 64.8, and those with established gastric cancer 107.4, a gradient that was highly statistically significant. Notably, IL-8 elevation in cancer patients appeared independent of H. pylori status, with comparable levels in H. pylori–positive and H. pylori–negative tumors. The authors caution, however, that these stage-stratified subgroup comparisons involved small numbers and should be regarded as exploratory. Nevertheless, the pattern is biologically coherent: as the gastric mucosa marches through the precancerous cascade, the inflammatory chemokine burden in the circulation appears to rise in parallel, reflecting the escalating molecular chaos within the stomach wall.

The longitudinal data proved even more informative. Of the 100 patients with precancerous lesions, 94 completed the full follow-up, and 9 of them progressed to gastric cancer during the study, 7 from intestinal metaplasia and 2 from atrophic gastritis. Using a cutoff of 52.3 picograms per milliliter at baseline, elevated IL-8 was associated with subsequent progression, achieving an area under the receiver operating characteristic curve of 0.84, with sensitivity of 88.9 percent, specificity of 78.0 percent, and a negative predictive value of 98.5 percent. That last figure deserves emphasis: a normal IL-8 level at a given visit was an extraordinarily reassuring finding, correctly identifying nearly 99 percent of patients who did not go on to develop cancer within the follow-up window. The established markers fared far worse, with CEA yielding an AUC of 0.62 and CA19-9 an AUC of 0.58, both barely better than chance in this setting.

Because the true power of a surveillance biomarker lies in its performance over time, the researchers applied time-dependent ROC analysis using a cumulative and dynamic approach. The discriminatory capacity of serial IL-8 strengthened as follow-up accumulated: the time-dependent AUC rose to 0.91 at 12 months and reached 0.94 at 18 months. In other words, the longer the team watched a patient’s IL-8 curve, the more accurately they could distinguish the individuals on a path to malignancy from those whose lesions would remain stable. This dynamic behavior is precisely what one hopes for from a monitoring strategy, since clinical decisions about escalating endoscopic surveillance depend on recognizing change rather than on static risk estimates made once at enrollment.

To capture that change formally, the team employed group-based trajectory modeling, executed while blinded to outcome status, which stratified the 100 patients into three distinct patterns of IL-8 behavior: a Stable-Low group of 42 patients, a Moderate-Rising group of 49, and a High-Accelerating group of 9. Every single patient in the High-Accelerating group progressed to gastric cancer, and the acceleration point in their IL-8 curves preceded the clinical diagnosis by a median of six months, with a range of three to twelve months. The researchers are appropriately measured about this observation, describing it as exploratory and hypothesis-generating and stressing that external validation is required before any causal or firm clinical interpretation. Still, the notion that a molecular inflection point in the bloodstream can herald cancer’s arrival by half a year is a tantalizing clue for early detection.

The statistical analysis also grappled with a challenge familiar to anyone studying rare events: only 9 progression events occurred, which limits the power of conventional regression. The investigators therefore used multivariable logistic regression with Firth penalization, a technique designed to reduce bias in small-sample, low-event settings. In that model, baseline IL-8 emerged as an independent predictor of progression, with an adjusted odds ratio of 2.31 for every 10-picogram-per-milliliter increase (95 percent confidence interval 1.48 to 3.61), and the presence of intestinal metaplasia carried an adjusted odds ratio of 4.22 (95 percent confidence interval 1.31 to 13.61). The study’s conclusions additionally report an adjusted hazard ratio of 6.94 for a baseline IL-8 above 52.3 picograms per milliliter, reinforcing the same message: the chemokine’s concentration and the histological stage together define a high-risk profile that standard markers miss.

What would it take to translate these numbers into clinical practice? First, the authors and independent observers alike will want confirmation in larger, geographically diverse cohorts, since the event count here was small and Egypt’s specific epidemiology of gastric cancer may not generalize everywhere. Second, the operational logistics matter: the required IL-8 sampling schedule, assay standardization across laboratories, and cost-effectiveness relative to simply intensifying endoscopic surveillance all remain open questions. Third, a biomarker with 88.9 percent sensitivity and 78.0 percent specificity would likely function best as a triage tool, selecting which patients with atrophic gastritis or intestinal metaplasia need closer endoscopic follow-up rather than replacing endoscopy outright. Yet if the six- to twelve-month lead time observed in the High-Accelerating trajectory is reproduced, the payoff could be substantial, catching tumors at a stage when endoscopic resection can be curative. In a disease that claims hundreds of thousands of lives annually and disproportionately burdens regions where advanced endoscopic infrastructure is scarce, a cheap, repeatable blood test that watches the gastric precancerous cascade in real time would represent a genuinely meaningful advance, and this study offers the most rigorous longitudinal evidence to date that interleukin-8 might be exactly that signal.

Subject of Research: Serum interleukin-8 trajectories as biomarkers of progression from Helicobacter pylori–associated precancerous gastric lesions to gastric cancer

Article Title: Serial serum interleukin-8 trajectories are associated with progression from Helicobacter pylori–associated precancerous lesions to gastric cancer: a prospective longitudinal cohort study

Article References: Semeya, A. A., Abdel Hafez, R. S. A., Elgamal, R., Badr, A., Orabi, M. I., Younis, S. R., Morgan, M. M., Ahmed Hassan, M., Alawy, M. R., & Othman, A. A. A. (2026). Serial serum interleukin-8 trajectories are associated with progression from Helicobacter pylori–associated precancerous lesions to gastric cancer: a prospective longitudinal cohort study. Gut Pathogens. https://doi.org/10.1186/s13099-026-00876-8

Image Credits: AI Generated

DOI: 10.1186/s13099-026-00876-8

Keywords: gastric cancer, interleukin-8, Helicobacter pylori, atrophic gastritis, intestinal metaplasia, biomarker, early detection, prospective cohort study, inflammation, tumor markers, risk stratification, Gut Pathogens

The study, published in Medical & Biological Engineering & Computing, describes a deep learning pipeline built around a hybrid Transformer-convolutional neural network architecture called UDTransNet. The system ingests digitized whole slide images of gastric mucosa biopsies, identifies metaplastic glands at the pixel level, and then translates those segmentation results into a clinical severity grade. Crucially, the grading step follows the Sydney system, the internationally recognized standard that pathologists use to classify gastritis and intestinal metaplasia by estimating the proportion of tissue occupied by metaplastic glands.

The technical challenge the researchers faced is formidable. A single whole slide image can contain tens of thousands of glands scanned at resolutions exceeding a hundred thousand pixels per side, far beyond what any neural network can process in one pass. The team therefore employed multi-scale patch cropping, slicing each slide into overlapping tiles at several magnifications so the network can see both fine cellular texture and broader tissue architecture. UDTransNet then fuses the strengths of two complementary families of deep learning models: Transformers, whose self-attention mechanisms capture long-range spatial relationships across a tissue section, and convolutional neural networks, which excel at extracting precise local edge and texture features. The architecture’s learnable skip connections shuttle these multi-level features back into the decoder, allowing sharp, anatomically faithful segmentation boundaries.

The performance figures reported on the internal test set are striking. The segmentation model achieved a Dice coefficient of 0.9698, meaning that its automatically outlined metaplastic regions overlapped with expert annotations almost perfectly. When those segmentations were converted into Sydney-system severity grades, the model reached an accuracy of 0.8879 with a Kappa statistic of 0.85, a level of agreement conventionally interpreted as near-perfect concordance with expert pathologists.

Perhaps the most provocative comparison in the study is the one against human observers at different career stages. The model’s Kappa of 0.85 exceeded the diagnostic consistency measured between junior and intermediate pathologists, which ranged from 0.67 to 0.82. In practical terms, the algorithm agreed with expert diagnoses more reliably than less experienced human specialists agreed with one another. For clinical workflows in which biopsy volume is high and specialist time is scarce, that gap matters: it suggests a machine assistant could meaningfully reduce the inter-observer variability that has long plagued gastric precancerous lesion grading.

The framework also proved valuable at the other end of the diagnostic spectrum, in ruling disease out. When identifying completely normal tissue, the model achieved a recall of 92.1 percent and an F1 score of 95.9 percent. This ‘negative exclusion’ capability is clinically significant because patients whose biopsies are genuinely free of metaplasia can be triaged toward routine surveillance rather than intensive follow-up, freeing endoscopy and pathology resources for those at genuine risk. An automated system that reliably clears normal slides lets human experts concentrate their attention on the borderline and positive cases where judgment matters most.

What distinguishes this work from many prior AI pathology studies is its interpretability and its direct mapping to a clinical standard. Rather than outputting an opaque risk score, the system produces digital maps of metaplastic gland distribution across the slide and quantifies their area proportion in accordance with Sydney system thresholds. The authors describe this as the first establishment of a digital mapping between pathological morphological features and the Sydney grading criteria. Explanatory visualization techniques such as gradient-based localization, referenced in the paper, allow the model’s attention to be checked against recognizable histological landmarks, addressing a common barrier to clinical adoption of deep learning in medicine.

The clinical context underscores the urgency. Gastric cancer incidence and mortality remain high in East Asia, and the well-characterized Correa cascade of gastric carcinogenesis, from chronic inflammation through atrophy and intestinal metaplasia to invasive cancer, offers a window in which detection and intervention can change outcomes. Global meta-analyses cited by the authors indicate that gastric intestinal metaplasia is widespread, yet its histological grading still depends on subjective visual assessment, a process the study characterizes as inefficient and poorly reproducible. Prior computational efforts have largely focused on endoscopic images; this work moves the diagnostic burden onto the histopathology slide itself, where grading criteria are formally defined.

The implications extend beyond the pathology laboratory. An automated, standardized grading system could harmonize diagnoses across hospitals and countries, enabling more consistent risk stratification for endoscopic surveillance programs and cleaner data for epidemiological research and clinical trials. The authors note that their framework provides an efficient and interpretable solution for intelligent screening of gastric precancerous lesions, and the reported ethics approval from Fujian Provincial Hospital indicates the work is grounded in real clinical specimens. The research was supported by the National Natural Science Foundation of China and provincial science funds of Fujian Province, with contributions from institutions including South China Normal University, Tsinghua University Shenzhen International Graduate School, and Fuzhou University Affiliated Provincial Hospital.

Challenges remain before such systems become routine. External validation on slides from different scanners, staining protocols and patient populations will be essential, and regulatory pathways for AI-assisted diagnosis are still maturing. Yet the core result stands: a Transformer-CNN fusion network, trained on whole slide images and constrained by an internationally accepted grading standard, can match expert pathologists in segmenting and grading one of gastric cancer’s most important precursors. As digital pathology continues to replace glass slides with gigapixel images, tools like this one point toward a future in which every biopsy receives an instant, consistent, and explainable second opinion, one that never tires and never varies from one reading to the next.

Subject of Research: Automated deep learning segmentation and grading of gastric intestinal metaplasia in whole slide pathology images

Article Title: Gastric mucosa intestinal metaplasia segmentation and grading via transformer-CNN fusion architecture: an interpretable digital pathology diagnostic framework

Article References: Jin, Y., Zhu, L., Yan, X., Lin, L., Zhu, M., Yang, L., Han, S., Li, T., Zeng, Y., Ji, Y., & Li, H. (2026). Gastric mucosa intestinal metaplasia segmentation and grading via transformer-CNN fusion architecture: an interpretable digital pathology diagnostic framework. Medical & Biological Engineering & Computing. https://doi.org/10.1007/s11517-026-03658-1

Image Credits: AI Generated

DOI: 10.1007/s11517-026-03658-1

Keywords: gastric cancer, intestinal metaplasia, deep learning, whole slide imaging, digital pathology, semantic segmentation, Sydney system, Transformer, convolutional neural network, precancerous lesions, computer-aided diagnosis, UDTransNet