Our study aimed to identify CT features of locally advanced gastric cancer (LAGC) associated with deficient mismatch repair (dMMR), explore the underlying pathological basis, and develop a morphology-based model for dMMR prediction.

This multicenter, retrospective analysis included 397 patients with LAGC treated at three institutions between January 2016 and January 2022. Mismatch repair (MMR) status was determined by immunohistochemical analysis of postoperative specimens. The patients were divided into training and external validation groups. CT features of primary tumours and lymph nodes associated with MMR status were identified, and the underlying pathological basis for these features were explored. Multivariable logistic analysis was used to identify independent CT features and to develop a morphology-based model.

Tumours with dMMR were characterized by increased thickness (p = 0.04), location in the lower stomach (p < 0.001), heterogeneous enhancement patterns (p = 0.02), a well-defined margin (p < 0.001), the presence of mushroom sign (p < 0.001), and the presence of abnormal lymph nodes (p = 0.07). Pathologically, tumours with dMMR exhibited extensive lymphocytic infiltration and solid growth with a pushing border, consistent with the distinctive CT features observed. Multivariable logistic analysis identified lower tumour location, well-defined tumour margin, and mushroom sign as independent predictors of dMMR. The morphology-based model achieved AUC values of 0.73-0.74 for dMMR prediction in the training and external validation groups.

LAGC with dMMR showed distinct CT features. A morphology-based model constructed from these CT features has the potential to predict dMMR status for LAGC.

Deficiency in DNA mismatch repair (dMMR) is a common pathway of carcinogenesis across different tumour types and confers a characteristic microsatellite instability-high (MSI-H) molecular phenotype. The prevalence of the MSI-H/dMMR phenotype is highest in endometrial and colorectal cancers, and this phenotype is associated with a distinct tumour biology, prognosis and responsiveness to various anticancer treatments. In a minority of patients, MSI-H/dMMR cancers result from an inherited pathogenic variant in the context of Lynch syndrome, which has important implications for familial genetic screening. Whether these hereditary cancers have a different biology and clinical behaviour to their sporadic counterparts remains uncertain. Interest in this tumour molecular subtype has increased following the discovery of the high sensitivity of metastatic MSI-H/dMMR cancers to immune-checkpoint inhibitors (ICIs) in a histology-agnostic manner, which reflects the genomic hypermutation resulting from dMMR that renders these tumours highly immunogenic and immune infiltrated. This vulnerability is now also being exploited in early stage disease settings. Despite this common biological foundation, different MSI-H/dMMR cancers have histotype-specific features that correspond to their particular cell or tissue of origin, which might be associated with differences in prognosis and sensitivity to ICIs. In this Review, we provide an overview of the epidemiology, biology, pathogenesis, clinical diagnosis and treatment of MSI-H/dMMR tumours as a histology-agnostic cancer phenomenon. We also highlight peculiarities associated with specific pathogenetic alterations and histologies of MSI-H/dMMR tumours.

Gastric cancer arises from complex carcinogenic factor interactions, with limited treatment options due to the lack of targetable driver gene mutations and significant tumor heterogeneity. Recent studies have provided promising novel approaches to improve our understanding of gastric cancer heterogeneity through integrated characterization, combining genomics with emerging technologies. Delineating the molecular changes and targeting specific molecular subtypes will enhance the efficacy of gastric cancer treatment and improve clinical outcomes. This review provides a comprehensive overview of current technologies used in gastric cancer research, highlighting key discoveries and treatment strategies driven by these innovations. Finally, we discuss the emerging technology-guided directions and potential breakthroughs that could enhance the understanding of gastric cancer tumor heterogeneity, ultimately improving clinical outcomes.

Determining tumor microsatellite status has significant clinical value because tumors that are microsatellite instability-high (MSI-H) or mismatch repair deficient (dMMR) respond well to immune checkpoint inhibitors (ICIs) and oftentimes not to chemotherapeutics. We propose MSI-SEER, a deep Gaussian process-based Bayesian model that analyzes H&E whole-slide images in weakly-supervised-learning to predict microsatellite status in gastric and colorectal cancers. We performed extensive validation using multiple large datasets comprised of patients from diverse racial backgrounds. MSI-SEER achieved state-of-the-art performance with MSI prediction by integrating uncertainty prediction. We achieved high accuracy for predicting ICI responsiveness by combining tumor MSI status with stroma-to-tumor ratio. Finally, MSI-SEER's tile-level predictions revealed novel insights into the role of spatial distribution of MSI-H regions in the tumor microenvironment and ICI response.

The Cancer Genome Atlas (TCGA) project defined four distinct molecular subtypes of esophagogastric adenocarcinoma: microsatellite instable (MSI), Epstein-Barr virus (EBV)-associated, genomically stable (GS), and chromosomally instable (CIN). However, an association between molecular subtypes and clinical outcomes has not been clearly demonstrated. Given few actionable biomarkers, we investigated the clinical relevance of TCGA classification system.

We identified all patients with esophagogastric adenocarcinoma whose tumors underwent prospective next-generation sequencing using the Memorial Sloan Kettering-IMPACT assay from 2014 to 2023. We classified all tumors in accordance with TCGA methodology and correlated molecular subtypes with high-quality clinicopathologic data.

Among 1,438 included patients, 941 had CIN, 344 had GS, 103 had MSI, and 50 had EBV tumors. Accounting for the clinical stage and tumor grade, molecular classification was independently associated with overall cancer-specific survival (P < 0.001) on Cox multivariable analysis. Furthermore, genomic signatures, patient demographics, pathologic responses to neoadjuvant therapy, patterns of recurrence, and metastatic organotropism differed significantly by molecular subtype. Although most distal esophageal and gastroesophageal junction tumors were CIN, up to 25% of these included GS, MSI, or EBV subtypes in contrast to TCGA. Random forest machine learning demonstrated that the molecular subtype is more influential in predicting response to treatment than tumor location.

Molecular classification is independently prognostic and may warrant inclusion in future staging and treatment guidelines. Routine molecular profiling is clinically feasible and may play a role in the management of patients to help guide appropriate treatment selection and clinical trial enrollment in the place of tumor location.

Findings from previous gastric cancer microbiome studies have been conflicting, potentially due to patient and/or tumor heterogeneity. The intratumoral gastric cancer microbiome and its relationship with clinicopathological variables have not yet been characterized in detail. We hypothesized that variation in gastric cancer microbial abundance, alpha diversity, and composition is related to clinicopathological characteristics.

Metagenomic analysis of 529 GC samples was performed, including whole exome sequencing data from The Cancer Genome Atlas (TCGA) and whole genome sequencing data from the 100,000 Genomes Project. Microbial abundance, alpha diversity, and composition were compared across patient age, sex, tumor location, geographic origin, pathological depth of invasion, pathological lymph node status, histological phenotype, microsatellite instability status, and TCGA molecular subtype.

Gastric cancer microbiomes resembled previous results, with Prevotella, Selenomonas, Stomatobaculum, Streptococcus, Lactobacillus, and Lachnospiraceae commonly seen across both cohorts. Within the TCGA cohort, microbial abundance and alpha diversity were greater in gastric cancers with microsatellite instability, lower pathological depth of invasion, intestinal-type histology, and those originating from Asia. Microsatellite instability status was associated with microbiome composition in both cohorts. Sex and pathological depth of invasion were associated with microbiome composition in the TCGA cohort.

The intratumoral gastric cancer microbiome appears to differ according to clinicopathological factors. Certain clinicopathological factors associated with favourable outcomes in gastric cancer were observed to be associated with greater microbial abundance and diversity. This highlights the need for further work to understand the underlying biological mechanisms behind the observed microbiome differences and their potential clinical and therapeutic impact.

Colorectal cancer (CRC) is a major global health concern, with genetic factors influencing its development. This study investigated the genomic profile of Amazonian indigenous populations (INDG) by analyzing five genes-APC, MLH1, MSH2, MSH6, and PMS2-associated with CRC. A total of 64 healthy individuals from 12 ethnic groups were analyzed using exome sequencing and bioinformatic tools. We identified 55 genetic variants, including three novel variants exclusive to the INDG, located in the MLH1 and MSH6 genes, which may represent genetic risks for CRC in this population. Additionally, three high-impact variants, already described in the literature, were identified in the APC and MSH2 genes. The study highlights the genetic isolation of Amazonian indigenous groups, with notable differences compared to continental populations. These findings emphasize the need for further genomic research to enhance the understanding of genetic risk factors and improve early detection and targeted therapies in vulnerable populations.

Hepatocellular carcinoma (HCC) is the third major cause of cancer death worldwide, with more than a doubling of incidence over the past two decades in the United States. Yet, the survival rate remains less than 20%, often due to late diagnosis at advanced stages. Current HCC screening approaches are serum alpha-fetoprotein (AFP) testing and ultrasound (US) of cirrhotic patients. However, these remain suboptimal, particularly in the setting of underlying obesity and metabolic dysfunction-associated steatotic liver disease/steatohepatitis (MASLD/MASH), which are also rising in incidence. Therefore, there is an urgent need for novel biomarkers that can stratify risk and predict early diagnosis of HCC, which is curable. Advances in liver cancer biology, multi-omics technologies, artificial intelligence, and precision algorithms have facilitated the development of promising candidates, with several emerging from completed phase 2 and 3 clinical trials. This review highlights the performance of these novel biomarkers and algorithms from a mechanistic perspective and provides new insight into how pathological processes can be detected through blood-based biomarkers. Through human studies compiled with animal models and mechanistic insight in pathways such as the TGF-β pathway, the biological progression from chronic liver disease to cirrhosis and HCC can be delineated. This integrated approach with new biomarkers merit further validation to refine HCC screening and improve early detection and risk stratification.

Despite advances in metastatic colorectal cancer (mCRC) treatment, long-term survival remains poor, particularly in right-sided colorectal cancer (RCRC), which has a worse prognosis compared to left-sided CRC (LCRC). This disparity is driven by the complex biological diversity of these malignancies. RCRC and LCRC differ not only in clinical presentation and outcomes but also in their underlying molecular and genetic profiles. This article offers a detailed literature review focusing on the distinctions between RCRC and LCRC. We explore key differences across embryology, anatomy, pathology, omics, and the tumor microenvironment (TME), providing insights into how these factors contribute to prognosis and therapeutic responses. Furthermore, we examine the therapeutic implications of these differences, considering whether the conventional classification of CRC into right- and left-sided forms should be refined. Recent molecular findings suggest that this binary classification may overlook critical biological complexities. Therefore, we propose that future approaches should integrate molecular insights to better guide personalized treatments, especially anti-EGFR therapies, and improve patient outcomes.

Immunity, stemness, and DNA damage repair (DDR) are crucial for cancer development and therapy resistance. With advancements in multiomics technology, the exploration of cancers related to immunity, stemness, and the DDR has triggered interest, but the combination of these levels for analyzing multiple cancers remains insufficient.

In this study, 9906 solid tumor samples from 31 TCGA cancer types were clustered on the basis of the enrichment levels of 13 gene sets associated with stemness, immunity, and DDR. Moreover, a soft ensemble model was constructed on the basis of the enrichment levels of these 13 gene sets to predict cancer subtypes via other omics data.

We identified four pancancer subtypes, termed C1, C2, C3, and C4, which presented distinct molecular and clinical features, including the immune microenvironment, stemness, genome instability, intratumor heterogeneity, methylation levels, tumor progression, sensitivity to chemotherapy and immunotherapy, and survival prognosis. The soft ensemble model validated this subtyping method in two breast cancer datasets (gene expression level), a pancancer proteomic dataset (protein expression level), and a pancancer cell line dataset (cell line gene expression level).

Our findings indicate that immune, stemness, and DDR signature-based subtyping offers new perspectives on cancer biology and holds promise for improving the clinical management of cancers.

Colorectal carcinoma (CRC) exhibits metastatic organotropism, primarily targeting liver, lung, and rarely the brain. Here, we study chromosomal imbalances (CIs) in cohorts of primary CRCs and metastases. Brain metastases show the highest burden of CIs, including aneuploidies and focal CIs, with enrichment of +12p encoding KRAS. Compared to liver and lung metastases, brain metastases present with increased co-occurrence of KRAS mutation and amplification. CRCs with concurrent KRAS mutation and amplification display significant metabolic reprogramming with upregulation of glycolysis, alongside upregulation of cell cycle pathways, including copy number gains of MDM2 and CDK4. Evolutionary modeling suggests early acquisition of many organotropic CIs enriched in both liver and brain metastases, while brain-enriched CIs preferentially emerge later. Collectively, this study supports a model where cytogenetic events in CRCs favor site-specific metastatic colonization. These site-enriched CI patterns may serve as biomarkers for metastatic potential in precision oncology.

In this study, we propose an alternative approach for stratifying genome-scale DNA methylation profiles of gastrointestinal (GI) adenocarcinomas based on a robust analytical framework. A set of 978 GI adenocarcinomas and 120 adjacent normal tissues from public repositories was quality controlled and analyzed. Hierarchical consensus clustering of the tumors, based on differential epigenetic variability between malignant and normal samples, identified six distinct subtypes defined either by a pan-GI or a lower GI-specific phenotype. In addition to methylation levels, aberrant methylation frequencies and the degree of DNA methylation instability contributed to the characterization of each subtype. We found significant differences in the outcome of patients, with the poorest overall survival seen for those belonging to a pan-GI subtype with infrequent aberrant methylation. In conclusion, our standardized approach contributes to a refined characterization of the epigenetic heterogeneity in GI adenocarcinomas, offering insights into subtype-specific methylation with the potential to support prognostication.

Despite the progress of treatment in gastric cancer (GC), the overall outcomes remain poor in patients with advanced diseases, underscoring the urgency to develop more effective treatment strategies. BH3-mimetic drugs, which inhibit the pro-survival BCL2 family proteins, have demonstrated great therapeutic potential in cancer therapy. Although previous studies have implicated a role of targeting the cell survival pathway in GC, the contribution of different pro-survival BCL2 family proteins in promoting survival and mediating resistance to current standard therapies in GC remains unclear. A systematic study to elucidate the hierarchy of these proteins using clinically more relevant GC models is essential to identify the most effective therapeutic target(s) and rational combination strategies for improving GC therapy. Here, we provide evidence from both in vitro and in vivo studies using a broad panel of GC cell lines, tumoroids, and xenograft models to demonstrate that BCLXL and MCL1, but not other pro-survival BCL2 family proteins, are crucial for GC cells survival. While small molecular inhibitors of BCLXL or MCL1 exhibited some single-agent activity, their combination sufficed to cause maximum killing. However, due to the unsolved cardiotoxicity associated with direct MCL1 inhibitors, finding combinations of agents that indirectly target MCL1 and enable the reduction of doses of BCLXL inhibitors while maintaining their anti-neoplastic effects is potentially a feasible approach for the further development of these compounds. Importantly, inhibiting BCLXL synergized significantly with anti-mitotic and HER2-targeting drugs, leading to enhanced anti-tumour activity with tolerable toxicity in preclinical GC models. Mechanistically, anti-mitotic chemotherapies induced MCL1 degradation via the ubiquitin-proteasome pathway mainly through FBXW7, whereas HER2-targeting drugs suppressed MCL1 transcription via the STAT3/SRF axis. Moreover, co-targeting STAT3 and BCLXL also exhibited synergistic killing, extending beyond HER2-amplified GC. Collectively, our results provide mechanistic rationale and pre-clinical evidence for co-targeting BCLXL and MCL1 (both directly and indirectly) in GC. (i) Gastric cancer cells rely on BCLXL and, to a lesser degree, on MCL1 for survival. The dual inhibition of BCLXL and MCL1 with small molecular inhibitors acts synergistically to kill GC cells, regardless of their TCGA molecular subtypes or the presence of poor prognostic markers. While the effect of S63845 is mediated by both BAX and BAK in most cases, BAX, rather than BAK, acts as the primary mediator of BCLXLi in GC cells. (ii) Inhibiting BCLXL significantly synergizes with anti-mitotic and HER2-targeting drugs, leading to enhanced anti-tumour activity with tolerable toxicity in preclinical GC models. Mechanistically, anti-mitotic chemotherapies induce MCL1 degradation via the ubiquitin-proteasome pathway mainly through FBXW7, whereas HER2-targeting drugs suppress MCL1 transcription via the STAT3/SRF axis. The combination of the STAT3 inhibitor and BCLXL inhibitor also exhibits synergistic killing, extending beyond HER2-amplified GC.

Microsatellite instability (MSI)-high tumors represent a distinct, small-fraction subtype in esophagogastric junction cancer or gastric cancer (GC), yet their clinical significance remains poorly understood. This study aimed to investigate the prevalence and clinicopathological features of chemotherapy-naïve metastatic or recurrent MSI-high GC as a prescreening study for a phase II trial of nivolumab plus ipilimumab.

Key inclusion criteria included metastatic or recurrent adenocarcinoma of GC, ECOG performance status of 0 or 1, and no prior systemic therapy for metastatic or recurrent disease. MSI status was tested using multiplex PCR fragment analysis (MSI Testing Kit, FALCO). The primary endpoint was the prevalence of MSI-high GC.

Between October 2020 and October 2022, 930 eligible patients from 75 centers in Japan were analyzed. The prevalence of MSI-high GC was 5.6% (95% CI 4.2-7.3). MSI-high GC was more frequently observed in females than males (9.6% vs 3.8%, p < 0.001), patients aged ≥ 70 years compared to those < 70 years (8.0% vs 2.8%, p < 0.001), in the lower stomach than other locations (10.5% vs 3.2%, p < 0.001), HER2-negative tumors than HER2-positive tumors (6.5% vs 1.8%, p = 0.02), and in patients without liver metastasis than those with liver metastasis (6.9% vs 2.2%, p = 0.004).

The prevalence of MSI-high tumors among chemotherapy-naïve patients with unresectable GC was 5.6%. These tumors were associated with female sex, older age, lower stomach, HER2-negative, and absence of liver metastasis. These findings would help assuming MSI-high tumors and may have significant implications for clinical practice and studies targeting this GC subtype.

Colorectal cancers (CRCs) are traditionally divided into those with either chromosomal instability (CIN) or microsatellite instability (MSI). By utilizing TCGA data, the Laird team found a subset of CRCs, namely, genome-stable CRCs (GS CRCs), which lack both CIN and MSI. Although the molecular features of GS CRCs have been described in detail, the clinicopathological features are not well defined. A total of 437 CRCs were analyzed for copy number variation (CNV) statuses in eight genes (ARID1A, EGFR, FGFR1, KDM5B, MYBL2, MYC, SALL4, and SETDB1) using droplet-digital PCR. CRCs that showed CNV in ≤ one gene and no MSI were defined as GS-like CRCs. Clinicopathological and molecular features of GS-like CRCs were compared with those of CIN-like CRCs. GS-like CRCs comprised 4.6% of CRCs and showed a predilection toward the proximal colon, lower nuclear optical density, KRAS mutation, PIK3CA mutation, and aberrant expression of KRT7. Survival analysis showed no significant difference between the three subgroups. Through our study, the GS-like subtype was found to comprise a minor proportion of CRCs and have proclivity toward a proximal bowel location, hypochromatic tumor nuclei, aberrant KRT7 expression, and a high frequency of KRAS and PIK3CA mutations.

Different colorectal cancer (CRC) studies rarely report overlapping prognostic genes. This study aimed to investigate the effects of sample size on prognostic genes analysis in CRC.

We included 418 CRC cases detected by whole-exome sequencing (WES) in the TCGA PanCancer cohort and 931 CRC cases detected by targeted sequencing in the MSK cohort. Prognostic genes analysis was repeated 200 times at each sample size level using a random resampling method.

For WES data, the number of prognostic genes increased in a power-law with increasing sample size in CRC cases with stage III and IV. This pattern also applied to CRC patients with stage II after the removal of patients with MSI-H or POLE mutations. However, for targeted sequencing data, the number of prognostic genes increased linearly with increasing sample size in CRC cases with stage III and IV. About 550 cases were required for stage IV CRC to reach the plateau of prognostic genes. In both cohorts, the proportion of true prognostic genes relative to sample size was consistent with a binomial distribution, indicating a significant effect of sample size on the reliability of prognostic genes. At the same sample size level, the number of prognostic genes from the WES data was higher than that from the targeted sequencing data, while the reliability of prognostic genes from the WES data was lower.

This study shows the relationship between the number of prognostic genes and sample size in CRC and how mutation data affects this relationship. This will contribute to the trial design for prognostic genetic analysis in CRC.

Our study's aim was to evaluate the clinicopathological profile of colorectal cancer (CRC) patients from North-East Romania in relation to the Kirsten rat sarcoma viral oncogene homolog (KRAS). We designed a retrospective study on 108 CRC patients using the fully automated real-time PCR-based molecular testing system, IdyllaTMKRAS Mutation Test (Biocartis, Mechelen, Belgium). Of the patients, 64 (59.3%) were men and 62 (57.4%) were older than the group average, with left bowel location in 38 cases (35.2%), adenocarcinoma NOS in 102 cases (94.4%), mixed histological pattern in 65 cases (60.2%), T3 in 60 patients (55.6%), N2 in 46 patients (42.6%), and 7-12 tumour buds registered in 58 tumours (53.7%). A total of 54 tumour samples (50%) showed KRAS mutation. Statistical comparative analyses associated KRAS mutations with the histopathological pattern (p = 0.018), tumour grade (p = 0.030), depth of invasion (pT) (p < 0.001), lymph node involvement (pN) (p < 0.001), venous vascular invasion (p = 0.048), and tumour buds' number (p = 0.007). Our results demonstrate the relationship between KRAS mutation and clinicopathological features, with possible impact in clinical tumour stratification and therapeutic management.

Molecular subtypes, such as defined by The Cancer Genome Atlas (TCGA), delineate a cancer's underlying biology, bringing hope to inform a patient's prognosis and treatment plan. However, most approaches used in the discovery of subtypes are not suitable for assigning subtype labels to new cancer specimens from other studies or clinical trials. Here, we address this barrier by applying five different machine learning approaches to multi-omic data from 8,791 TCGA tumor samples comprising 106 subtypes from 26 different cancer cohorts to build models based upon small numbers of features that can classify new samples into previously defined TCGA molecular subtypes-a step toward molecular subtype application in the clinic. We validate select classifiers using external datasets. Predictive performance and classifier-selected features yield insight into the different machine-learning approaches and genomic data platforms. For each cancer and data type we provide containerized versions of the top-performing models as a public resource.

Deep Learning (DL) has emerged as a powerful tool to predict genetic biomarkers directly from digitized Hematoxylin and Eosin (H&E) slides in colorectal cancer (CRC). However, few studies have systematically investigated the predictability of biomarkers beyond routinely available alterations such as microsatellite instability (MSI), and BRAF and KRAS mutations.

Our primary dataset comprised H&E slides of CRC tumors across five cohorts totaling 1,376 patients who underwent comprehensive panel sequencing, with an additional 536 patients from two public datasets for validation. We developed a DL model using a single transformer model to predict multiple genetic alterations directly from the slides. The model's performance was compared against conventional single-target models, and potential confounders were analyzed.

The multi-target model was able to predict numerous biomarkers from pathology slides, matching and partly exceeding single-target transformers. The Area Under the Receiver Operating Characteristic curve (AUROC, mean ± std) on the primary external validation cohorts was: BRAF (0·78 ± 0·01), hypermutation (0·88 ± 0·01), MSI (0·93 ± 0·01), RNF43 (0·86 ± 0·01); this biomarker predictability was mirrored across metrics and co-occurrence analyses. However, biomarkers with high AUROCs largely correlated with MSI, with model predictions depending considerably on MSI-associated morphology upon pathological examination.

Our study demonstrates that multi-target transformers can predict the biomarker status for numerous genetic alterations in CRC directly from H&E slides. However, their predictability is mainly associated with MSI phenotype, despite indications of slight biomarker-inherent contributions to a phenotype. Our findings underscore the need to analyze confounders in AI-based oncology biomarkers. To enable this, we developed a validated model applicable to other cancers and larger, diverse datasets.

The German Federal Ministry of Health, the Max-Eder-Programme of German Cancer Aid, the German Federal Ministry of Education and Research, the German Academic Exchange Service, and the EU.

Treatment approaches for colorectal cancer (CRC) are highly dependent on the molecular subtype, as immunotherapy has shown efficacy in cases with microsatellite instability (MSI) but is ineffective for the microsatellite stable (MSS) subtype. There is promising potential in utilizing deep neural networks (DNNs) to automate the differentiation of CRC subtypes by analyzing hematoxylin and eosin (H&E) stained whole-slide images (WSIs). Due to the extensive size of WSIs, multiple instance learning (MIL) techniques are typically explored. However, existing MIL methods focus on identifying the most representative image patches for classification, which may result in the loss of critical information. Additionally, these methods often overlook clinically relevant information, like the tendency for MSI class tumors to predominantly occur on the proximal (right side) colon.

We introduce 'CIMIL-CRC', a DNN framework that: (1) solves the MSI/MSS MIL problem by efficiently combining a pre-trained feature extraction model with principal component analysis (PCA) to aggregate information from all patches, and (2) integrates clinical priors, particularly the tumor location within the colon, into the model to enhance patient-level classification accuracy. We assessed our CIMIL-CRC method using the average area under the receiver operating characteristic curve (AUROC) from a 5-fold cross-validation experimental setup for model development on the TCGA-CRC-DX cohort, contrasting it with a baseline patch-level classification, a MIL-only approach, and a clinically-informed patch-level classification approach.

Our CIMIL-CRC outperformed all methods (AUROC: 0.92±0.002 (95% CI 0.91-0.92), vs. 0.79±0.02 (95% CI 0.76-0.82), 0.86±0.01 (95% CI 0.85-0.88), and 0.87±0.01 (95% CI 0.86-0.88), respectively). The improvement was statistically significant. To the best of our knowledge, this is the best result achieved for MSI/MSS classification on this dataset.

Our CIMIL-CRC method holds promise for offering insights into the key representations of histopathological images and suggests a straightforward implementation.

Epstein-Barr virus (EBV), the first human oncovirus discovered in 1964, has become a focal point in virology, immunology, and oncology because of its unique biological characteristics and significant role in human diseases. As we commemorate the 60th anniversary of EBV's discovery, it is an opportune moment to reflect on the major advancements in our understanding of this complex virus. In this review, we highlight key milestones in EBV research, including its virion structure and life cycle, interactions with the host immune system, association with EBV-associated diseases, and targeted intervention strategies.

Colorectal cancer (CRC) incidence and mortality before 50 have been rising alarmingly in the recent decades.

Using a cohort of 10,000 patients, this study investigates the clinical, mutational, and co-mutational features of CRC in early-onset (EOCRC, < 50 years) compared to late-onset (LOCRC, ≥ 50 years).

EOCRC was associated with a higher prevalence of Asian and Hispanic patients, rectal or left-sided tumors (72% vs. 59%), and advanced-stage disease. Molecular analyses revealed differences in mutation patterns, with EOCRC having higher frequencies of TP53 (74% vs. 68%, p < 0.01) and SMAD4 (17% vs. 14%, p = 0.015), while BRAF (5% vs. 11%, p < 0.001) and NOTCH1 (2.7% vs. 4.1%, p = 0.01) mutations were more prevalent in LOCRC. Stratification by tumor site and MSI status highlighted significant location- and age-specific molecular differences, such as increased KRAS and CTNNB1 mutations in right-sided EOCRC and higher BRAF prevalence in MSI-H LOCRC (47% vs. 6.7%, p < 0.001). Additionally, co-occurrence analysis revealed unique mutational networks in EOCRC MSS, including significant co-occurrences of FBXW7 with NOTCH3, RB1, and PIK3R1.

This study highlights the significance of age-specific molecular profiling, offering insights into the unique biology of EOCRC and potential clinical applications.

Intratumoral heterogeneity emerges from accumulating genetic and epigenetic changes during tumorigenesis, which may contribute to therapeutic failure and drug resistance. However, the lack of a quick and convenient approach to determine the intratumoral epigenetic heterogeneity (eITH) limit the application of eITH in clinical settings. Here, we aimed to develop a tool that can evaluate the eITH using the DNA methylation profiles from bulk tumors.

Genomic DNA of three laser micro-dissected tumor regions, including digestive tract surface, central bulk, and invasive front, was extracted from formalin-fixed paraffin-embedded sections of colorectal cancer patients. The genome-wide methylation profiles were generated with methylation array. The most variable methylated probes were selected to construct a DNA methylation-based heterogeneity (MeHEG) estimation tool that can deconvolve the proportion of each reference tumor region with the support vector machine model-based method. A PCR-based assay for quantitative analysis of DNA methylation (QASM) was developed to specifically determine the methylation status of each CpG in MeHEG assay at single-base resolution to realize fast evaluation of epigenetic heterogeneity.

In the discovery set with 79 patients, the differentially methylated CpGs among the three tumor regions were found. The 7 most representative CpGs were identified and subsequently selected to develop the MeHEG algorithm. We validated its performance of deconvolution of tumor regions in an independent cohort. In addition, we showed the significant association of MeHEG-based epigenetic heterogeneity with the genomic heterogeneity in mutation and copy number variation in our in-house and TCGA cohorts. Besides, we found that the patients with higher MeHEG score had worse disease-free and overall survival outcomes. Finally, we found dynamic change of epigenetic heterogeneity based on MeHEG score in cancer cells under the treatment of therapeutic drugs.

By developing a 7-loci panel using a machine learning approach combined with the QASM assay for PCR-based application, we present a valuable method for evaluating intratumoral heterogeneity. The MeHEG algorithm offers novel insights into tumor heterogeneity from an epigenetic perspective, potentially enriching current knowledge of tumor complexity and providing a new tool for clinical and research applications in cancer biology.

Bile tract cancer (BTC) is a malignant tumor with a poor prognosis. Recent studies have reported the heterogeneity of the genomic background and gene alterations in BTC, but its genetic heterogeneity and molecular profiles remain poorly understood. Whole-genome sequencing may enable the identification of novel actionable gene mutations involved in BTC carcinogenesis, malignant progression, and treatment resistance.

We performed whole-genome sequencing of six BTC samples to elucidate its genetic heterogeneity and identify novel actionable gene mutations. Somatic mutations, structural variations, copy number alterations, and their associations with clinical factors were analyzed.

The average number of somatic mutations detected in each case was 53,705, with SNVs accounting for most of these mutations (85.02%). None of the 331 mutations related to BTC in The Cancer Genome Atlas (TCGA) database were found in the mutations identified in our study. A higher prevalence of gene mutations was observed in samples without vascular invasion than in those with vascular invasion. Several genes with differences in mutation accumulation between groups were identified, including ADAMTS7, AHNAK2, and CAPN10.

Our study provides novel insights into the genomic landscape of BTC and highlights the potential of whole-genome sequencing analysis to identify actionable gene mutations and understand the molecular mechanisms underlying this malignancy. The high mutational burden, structural variations, and copy number alterations observed in BTC samples in this study underscore the genetic complexity and heterogeneity of this disease.

Despite similarities in microsatellite instability (MSI) between colon and endometrial cancer, there are many clinically important organ-specific features. The molecular differences between these 2 MSI cancers are underexplored because the usual differentially expressed gene analysis yields too many noncancer-specific normally expressed genes. We aimed to identify cancer-specific genes in MSI colorectal adenocarcinoma (CRC) and MSI endometrial carcinoma (ECs) using a modified partial least squares discriminant analysis. We obtained a list of cancer-specific genes in MSI CRC and EC by taking the intersection of the genes obtained from tumor samples and normal samples. Specifically, we obtained publically available 1319 RNA sequencing data consisting of MSI CRCs, MSI ECs, normal colon including the rectum, and normal endometrium from The Cancer Genome Atlas and genome-tissue expression sites. To reduce gene-centric dimensions, we retained only 3924 genes from the original data by performing the usual differentially expressed gene screening for tumor samples using DESeq2. The usual partial least squares discriminant analysis was performed for tumor samples, producing 625 genes, whereas for normal samples, projection vectors with zero covariance were sampled, their weights were square-summed, and genes with sufficiently high values were selected. Gene ontology (GO) term enrichment, protein-protein interaction, and survival analyses were performed for functional and clinical validation. We identified 30 cancer-specific normal-invariant genes, including Zic family members (ZIC1, ZIC4, and ZIC5), DPPA2, PRSS56, ELF5, and FGF18, most of which were cancer-associated genes. Although no statistically significant GO terms were identified in the GO term enrichment analysis, cell differentiation was observed as potentially significant. In the protein-protein interaction analysis, 17 of the 30 genes had at least one connection, and when first-degree neighbors were added to the network, many cancer-related pathways, including MAPK, Ras, and PI3K-Akt, were enriched. In the survival analysis, 16 genes showed statistically significant differences between the lower and higher expression groups (3 in CRCs and 15 ECs). We developed a novel approach for selecting cancer-specific normal-invariant genes from relevant gene expression data. Although we believe that tissue-specific reactivation of embryonic genes might explain the cancer-specific differences of MSI CRC and EC, further studies are needed for validation.

N6-Methyladenosine (m6A) is the most abundant internal modification in mRNAs. Despite accumulating evidence for the profound impact of m6A on cancer biology, there are conflicting reports that alterations in genes encoding the m6A machinery proteins can either promote or suppress cancer, even in the same tumor type. Using data from The Cancer Genome Atlas, we performed a pan-cancer investigation of 15 m6A core factors in nearly 10000 samples from 31 tumor types to reveal underlying cross-tumor patterns. Altered expression, largely driven by copy number variations at the chromosome arm level, results in the most common mode of dysregulation of these factors. YTHDF1, YTHDF2, YTHDF3 and VIRMA are the most frequently altered factors and the only ones to be uniquely altered when tumors are grouped according to the expression pattern of the m6A factors. These genes are also the only ones with coherent, pan-cancer predictive power for progression-free survival. On the contrary, METTL3, the most intensively studied m6A factor as a cancer target, shows much lower levels of alteration and no predictive power for patient survival. Therefore, we propose the non-enzymatic YTHDF and VIRMA genes as preferred subjects to dissect the role of m6A in cancer and as priority cancer targets.

As a research hotspot in the field of molecular biology, N6-methyladenosine (m6A) modification has made progress in the treatment of colorectal cancer (CRC), leukemia and other cancers. Numerous studies have demonstrated that the tumour microenvironment (TME) regulates the level of m6A modification in the host and activates a series of complex epigenetic signalling pathways through interactions with CRC cells, thus affecting the progression and prognosis of CRC. However, with the diversity in the composition of TME factors, this action is reciprocal and complex. Encouragingly, some studies have experimentally revealed that the intestinal flora can alter CRC cell proliferation by directly acting on m6A and thereby altering CRC cell proliferation. This review summarizes the data, supporting the idea that the intestinal flora can influence host m6A levels through pathways such as methyl donor metabolism and thus affect the progression of CRC. We also review the role of m6A modification in the diagnosis, treatment, and prognostic assessment of CRC and discuss the current status, limitations, and potential clinical value of m6A modification in this field. We propose that additional in-depth research on m6A alterations in CRC patients and their TME-related targeted therapeutic issues will lead to better therapeutic outcomes for CRC patients.

Tumours exhibit significant heterogeneity in their molecular profiles across patients, largely influenced by the tissue of origin, where certain driver gene mutations are predominantly associated with specific cancer types. Here, we unveil an additional layer of complexity: some cancer types display anatomic location-specific mutation profiles akin to tissue-specificity. To better understand this phenomenon, we concentrate on colon cancer. While prior studies have noted changes of the frequency of molecular alterations along the colon, the underlying reasons and whether those changes occur rather gradual or are distinct between the left and right colon, remain unclear. Developing and leveraging stringent statistical models on molecular data from 522 colorectal tumours from The Cancer Genome Atlas, we reveal disparities in molecular properties between the left and right colon affecting many genes. Interestingly, alterations in genes responsive to environmental cues and properties of the tumour ecosystem, including metabolites which we quantify in a cohort of 27 colorectal cancer patients, exhibit continuous trends along the colon. Employing network methodologies, we uncover close interactions between metabolites and genes, including drivers of colon cancer, showing continuous abundance or alteration profiles. This underscores how anatomic biases in the composition and interactions within the tumour ecosystem help explaining gradients of carcinogenesis along the colon.

The success of immunotherapy for cancer treatment is limited by the presence of an immunosuppressive tumor microenvironment (TME); Therefore, identifying novel targets to that can reverse this immunosuppressive TME and enhance immunotherapy efficacy is essential. In this study, enrichment analysis based on publicly available single-cell and bulk RNA sequencing data from gastric cancer patients are conducted, and found that tumor-intrinsic interferon (IFN) plays a central role in TME regulation. The results shows that KDM3A over-expression suppresses the tumor-intrinsic IFN response and inhibits KDM3A, either genomically or pharmacologically, which effectively promotes IFN responses by activating endogenous retroviruses (ERVs). KDM3A ablation reconfigures the dsRNA-MAVS-IFN axis by modulating H3K4me2, enhancing the infiltration and function of CD8 T cells, and simultaneously reducing the presence of regulatory T cells, resulting in a reshaped TME in vivo. In addition, combining anti-PD1 therapy with KDM3A inhibition effectively inhibited tumor growth. In conclusions, this study highlights KDM3A as a potential target for TME remodeling and the enhancement of antitumor immunity in gastric cancer through the regulation of the ERV-MAVS-IFN axis.

Temporal ordering of cellular events offers fundamental insights into biological phenomena. Although this is traditionally achieved through continuous direct observations1,2, an alternative solution leverages irreversible genetic changes, such as naturally occurring mutations, to create indelible marks that enables retrospective temporal ordering3-5. Using a multipurpose, single-cell CRISPR platform, we developed a molecular clock approach to record the timing of cellular events and clonality in vivo, with incorporation of cell state and lineage information. Using this approach, we uncovered precise timing of tissue-specific cell expansion during mouse embryonic development, unconventional developmental relationships between cell types and new epithelial progenitor states by their unique genetic histories. Analysis of mouse adenomas, coupled to multiomic and single-cell profiling of human precancers, with clonal analysis of 418 human polyps, demonstrated the occurrence of polyclonal initiation in 15-30% of colonic precancers, showing their origins from multiple normal founders. Our study presents a multimodal framework that lays the foundation for in vivo recording, integrating synthetic or natural indelible genetic changes with single-cell analyses, to explore the origins and timing of development and tumorigenesis in mammalian systems.