A cohort study and meta-analysis of the evidence for consideration of Lauren subtype when prescribing adjuvant or palliative chemotherapy for gastric cancer

A Retrospective Analysis of the Lauren Classification in the Choice of XELOX or SOX as an Adjuvant Chemotherapy for Gastric Cancer

BACKGROUND

We aim to retrospectively explore the guiding value of the Lauren classification for patients who have undergone D2 gastrectomy to choose oxaliplatin plus capecitabine (XELOX) or oxaliplatin plus S-1 (SOX) as a further systemic treatment after the operation.

METHODS

We collected data of 406 patients with stage III gastric cancer(GC)after radical D2 resection and regularly received XELOX or SOX adjuvant treatment after surgery and followed them for at least five years. According to the Lauren classification, we separated patients out into intestinal type (IT) GC together with non-intestinal type(NIT) GC. According to the chemotherapy regimen, we separated patients into the SOX group together with the XELOX group.

RESULTS

Among non-intestinal type patients, the 3-year DFS rates in the SOX group and the XELOX group were 72.5%, respectively; 54.5% (P=0.037); The 5-year OS rates were 66.8% and 51.8% respectively (P=0.038), both of which were statistically significant.

CONCLUSION

The patients of non-intestinal type GC may benefit from the SOX regimen. Differences were counted without being statistically significant with intestinal-type GC in the SOX or XELOX groups.

Deep learning-based subtyping of gastric cancer histology predicts clinical outcome: a multi-institutional retrospective study

INTRODUCTION

The Laurén classification is widely used for Gastric Cancer (GC) histology subtyping. However, this classification is prone to interobserver variability and its prognostic value remains controversial. Deep Learning (DL)-based assessment of hematoxylin and eosin (H&E) stained slides is a potentially useful tool to provide an additional layer of clinically relevant information, but has not been systematically assessed in GC.

OBJECTIVE

We aimed to train, test and externally validate a deep learning-based classifier for GC histology subtyping using routine H&E stained tissue sections from gastric adenocarcinomas and to assess its potential prognostic utility.

METHODS

We trained a binary classifier on intestinal and diffuse type GC whole slide images for a subset of the TCGA cohort (N = 166) using attention-based multiple instance learning. The ground truth of 166 GC was obtained by two expert pathologists. We deployed the model on two external GC patient cohorts, one from Europe (N = 322) and one from Japan (N = 243). We assessed classification performance using the Area Under the Receiver Operating Characteristic Curve (AUROC) and prognostic value (overall, cancer specific and disease free survival) of the DL-based classifier with uni- and multivariate Cox proportional hazard models and Kaplan-Meier curves with log-rank test statistics.

RESULTS

Internal validation using the TCGA GC cohort using five-fold cross-validation achieved a mean AUROC of 0.93 ± 0.07. External validation showed that the DL-based classifier can better stratify GC patients' 5-year survival compared to pathologist-based Laurén classification for all survival endpoints, despite frequently divergent model-pathologist classifications. Univariate overall survival Hazard Ratios (HRs) of pathologist-based Laurén classification (diffuse type versus intestinal type) were 1.14 (95% Confidence Interval (CI) 0.66-1.44, p-value = 0.51) and 1.23 (95% CI 0.96-1.43, p-value = 0.09) in the Japanese and European cohorts, respectively. DL-based histology classification resulted in HR of 1.46 (95% CI 1.18-1.65, p-value < 0.005) and 1.41 (95% CI 1.20-1.57, p-value < 0.005), in the Japanese and European cohorts, respectively. In diffuse type GC (as defined by the pathologist), classifying patients using the DL diffuse and intestinal classifications provided a superior survival stratification, and demonstrated statistically significant survival stratification when combined with pathologist classification for both the Asian (overall survival log-rank test p-value < 0.005, HR 1.43 (95% CI 1.05-1.66, p-value = 0.03) and European cohorts (overall survival log-rank test p-value < 0.005, HR 1.56 (95% CI 1.16-1.76, p-value < 0.005)).

CONCLUSION

Our study shows that gastric adenocarcinoma subtyping using pathologist's Laurén classification as ground truth can be performed using current state of the art DL techniques. Patient survival stratification seems to be better by DL-based histology typing compared with expert pathologist histology typing. DL-based GC histology typing has potential as an aid in subtyping. Further investigations are warranted to fully understand the underlying biological mechanisms for the improved survival stratification despite apparent imperfect classification by the DL algorithm.

Molecular mechanism and potential therapeutic targets of liver metastasis from gastric cancer

Gastric cancer (GC) is characterized by high invasion and poor prognosis. The occurrence of liver metastasis seriously affects advanced GC prognosis. In recent years, great progress has been made in the field of GC liver metastasis. The abnormal expression of related genes leads to the occurrence of GC liver metastasis through metastasis cascades. The changes in the liver microenvironment provide a pre-metastasis condition for GC cells to colonize and grow. The development of several potential therapeutic targets might provide new therapeutic strategies for its treatment. Therefore, we reviewed the regulatory mechanism of abnormal genes mediating liver metastasis, the effect of liver resident cells on liver metastasis, and potential therapeutic targets, hoping to provide a novel therapeutic option to improve the quality of life and prognosis of GC patients with liver metastasis.

Adversarial attacks and adversarial robustness in computational pathology

Artificial Intelligence (AI) can support diagnostic workflows in oncology by aiding diagnosis and providing biomarkers directly from routine pathology slides. However, AI applications are vulnerable to adversarial attacks. Hence, it is essential to quantify and mitigate this risk before widespread clinical use. Here, we show that convolutional neural networks (CNNs) are highly susceptible to white- and black-box adversarial attacks in clinically relevant weakly-supervised classification tasks. Adversarially robust training and dual batch normalization (DBN) are possible mitigation strategies but require precise knowledge of the type of attack used in the inference. We demonstrate that vision transformers (ViTs) perform equally well compared to CNNs at baseline, but are orders of magnitude more robust to white- and black-box attacks. At a mechanistic level, we show that this is associated with a more robust latent representation of clinically relevant categories in ViTs compared to CNNs. Our results are in line with previous theoretical studies and provide empirical evidence that ViTs are robust learners in computational pathology. This implies that large-scale rollout of AI models in computational pathology should rely on ViTs rather than CNN-based classifiers to provide inherent protection against perturbation of the input data, especially adversarial attacks.

Artificial Intelligence can support diagnostic workflows in oncology, but they are vulnerable to adversarial attacks. Here, the authors show that convolutional neural networks are highly susceptible to white- and black-box adversarial attacks in clinically relevant classification tasks.

Receptor Tyrosine Kinases Amplified in Diffuse-Type Gastric Carcinoma: Potential Targeted Therapies and Novel Downstream Effectors

Simple Summary

Diffuse-type gastric carcinoma (DGC) is an aggressive subtype of gastric carcinoma with an extremely poor prognosis due to frequent peritoneal metastasis and high probability of recurrence. Its pathogenesis is poorly understood, and consequently, no effective molecular targeted therapy is available. The importance of oncogenic receptor tyrosine kinase (RTK) signaling has been recently demonstrated in the malignant progression of DGC. In particular, RTK gene amplification appears to accelerate peritoneal metastasis. In this review, we provide an overview of RTK gene amplification in DGC and the potential of related targeted therapies.

Abstract

Gastric cancer (GC) is a major cause of cancer-related death worldwide. Patients with an aggressive subtype of GC, known as diffuse-type gastric carcinoma (DGC), have extremely poor prognoses. DGC is characterized by rapid infiltrative growth, massive desmoplastic stroma, frequent peritoneal metastasis, and high probability of recurrence. These clinical features and progression patterns of DGC substantially differ from those of other GC subtypes, suggesting the existence of specific oncogenic signals. The importance of gene amplification and the resulting aberrant activation of receptor tyrosine kinase (RTK) signaling in the malignant progression of DGC is becoming apparent. Here, we review the characteristics of RTK gene amplification in DGC and its importance in peritoneal metastasis. These insights may potentially lead to new targeted therapeutics.

Clinicopathological differences, risk factors and prognostic scores for western patients with intestinal and diffuse-type gastric cancer

BACKGROUND

In the molecular era, the Laurén system is still a cost-effective and widely implemented classification for gastric cancer (GC) and it has been recently associated with clinical, histological and molecular features of these tumors. Despite recent advances in the understanding of the molecular biology of GC, there is a need to develop new prognostic tools for patient stratification in clinical practice. Thus, the identification of easily available prognostic factors in patients with intestinal and diffuse-type tumors can significantly improve risk assessment and patient stratification in GC.

AIM

To identify clinicopathological differences, risk factors, and to develop cost-effective prognostic scores for patients with intestinal and diffuse-type GC.

METHODS

Retrospective study of all patients undergoing surgery for GC at a tertiary referral center from 2001 to 2019. 286 cases met inclusion criteria (intestinal: 190, diffuse: 96). Clinical data and gross findings were collected. All specimens were reviewed by two independent pathologists and a detailed protocol for histologic evaluation was followed. Five tissue microarrays (TMAs) were constructed and sections of the TMA block were immunostained for HERCEPTEST, MSH2, MSH6, MLH1 and PMS2. Statistical analyses were performed and prognostic scores were developed based on hazard ratios.

RESULTS

Intestinal and diffuse-type GC showed different epidemiological, clinicopathological and prognostic features. Diffuse tumors were significantly associated with younger age, less symptomatology, flat morphology, deeper invasion, perineural infiltration, advanced stage at diagnosis, administration of adjuvant therapy and poorer prognosis. Intestinal lesions were fungoid or polypoid, showed necrosis, desmoplasia, microsatellite instability and HERCEPTEST positivity and were diagnosed at earlier stages. Tumor depth, desmoplasia, macroscopic type and lymph node involvement were independently related to the Laurén subtype. Furthermore, intestinal and diffuse GC were associated with different risk factors for progression and death. Vascular invasion, perineural infiltration and growth pattern were important prognostic factors in intestinal-type GC. On the contrary, tumor size and necrosis were significant prognosticators in diffuse-type GC. Our recurrence and cancer-specific death scores for patients with intestinal and diffuse-type GC showed an excellent patient stratification into three (diffuse GC) or four (intestinal) prognostic groups.

CONCLUSION

Our findings support that Laurén subtypes represent different clinicopathological and biological entities. The development of specific prognostic scores is a useful and cost-effective strategy to improve risk assessment in GC.

Costunolide-Induced Apoptosis via Promoting the Reactive Oxygen Species and Inhibiting AKT/GSK3β Pathway and Activating Autophagy in Gastric Cancer

Objective: Costunolide (Cos) is a sesquiterpene lactone extracted from chicory. Although it possesses anti-tumor effects, the underlying molecular mechanism against gastric cancer cells remains unclear. This study aimed to explore the effect and potential mechanism of Cos on gastric cancer.

Methods: The effect of Cos on HGC-27 and SNU-1 proliferation was detected by CCK-8 and clone formation assay. The changes in cell apoptosis were determined using Hoechst 33258 and tunel staining. The morphology of autophagy was analyzed by autophagosomes with the electron microscope and LC3-immunofluorescence with the confocal microscope. The related protein levels of the cell cycle, apoptosis, autophagy and AKT/GSK3β pathway were determined by Western blot. The anti-tumor activity of Cos was evaluated by subcutaneously xenotransplanting HGC-27 into Balb/c nude mice. The Ki67 and P-AKT levels were examined by immunohistochemistry.

Results: Cos significantly inhibited HGC-27 and SNU-1 growth and induced cell cycle arrest in the G2/M phase. Cos activated intrinsic apoptosis and autophagy through promoting cellular reactive oxygen species (ROS) levels and inhibiting the ROS-AKT/GSK3β signaling pathway. Moreover, preincubating gastric carcinoma cells with 3-methyladenine (3-MA), a cell-autophagy inhibitor, significantly alleviated the effects of Cos in inducing cell apoptosis.

Conclusion: Cos induced apoptosis of gastric carcinoma cells via promoting ROS and inhibiting AKT/GSK3β pathway and activating pro-death cell autophagy, which may be an effective strategy to treat gastric cancer.

Is there still a place for conventional histopathology in the age of molecular medicine? Laurén classification, inflammatory infiltration and other current topics in gastric cancer diagnosis and prognosis

Gastric cancer (GC) is the fifth most common cancer and the third cause of cancer-related deaths worldwide. In western countries, more than half of GC patients are diagnosed at advanced stages and 5-year survival rates range between 20-30%. The only curative treatment is surgery, and despite recent advances in oncological therapies, GC prognosis is still poor. The main prognostic tool for patient categorization and treatment selection is the TNM classification, but its limitations are being increasingly recognized. Early recurrences may occur in early-stage disease, and patients at the same stage show heterogeneous outcomes. Thus, there is a need to improve GC stratification and to identify new prognostic factors, which may allow us to select drug-susceptible populations, refine patient grouping for clinical trials and discover new therapeutic targets. Molecular classifications have been developed, but they have not been translated to the clinical practice. On the other hand, histological assessment is cheap and widely available, and it is still a mainstay in the era of molecular medicine. Furthermore, histological features are acquiring new roles as reflectors of the genotype-phenotype correlation, and their potential impact on patient management is currently being analyzed. The aim of this literature review is to provide a modern overview of the histological assessment of GC. In this study, we discuss recent topics on the histological diagnosis of GC, focusing on the current role of Laurén classification and the potential value of new histological features in GC, such as inflammatory infiltration and tumor budding.

Competing Endogenous RNA Networks in the Epithelial to Mesenchymal Transition in Diffuse-Type of Gastric Cancer

Simple Summary

The diffuse-type of gastric cancer is associated with epithelial to mesenchymal transition. Loss of E-cadherin expression is the hallmark of this process and is largely due to the upregulation of the transcription factors ZEB1/2, Snail, Slug, and Twist1/2. However, miRNA and lncRNAs can also participate through these transcription factors which directly target E-cadherin. The competing endogenous RNA (ceRNA) network hypothesis state that lncRNA can sponge the miRNA pool that targets these transcripts. Based on the lack of said networks in the epithelial to mesenchymal transition, we performed a prediction analysis that resulted in novel ceRNA networks which will expand our knowledge of the molecular basis of the diffuse-type of gastric cancer.

Abstract

The diffuse-type of gastric cancer (DGC), molecularly associated with epithelial to mesenchymal transition (EMT), is increasing in incidence. Loss of E-cadherin expression is the hallmark of the EMT process and is largely due to the upregulation of the EMT-inducing transcription factors ZEB1/2, Snail, Slug, and Twist1/2. However, ncRNA, such as miRNA and lncRNAs, can also participate in the EMT process through the direct targeting of E-cadherin and other EMT-inducing transcription factors. Additionally, lncRNA can sponge the miRNA pool that targets these transcripts through competing endogenous RNA (ceRNA) networks. In this review, we focus on the role of ncRNA in the direct deregulation of E-cadherin, as well as EMT-inducing transcription factors. Based on the relevance of the ceRNA network hypothesis, and the lack of said networks in EMT, we performed a prediction analysis for all miRNAs and lncRNAs that target E-cadherin, as well as EMT-inducing transcription factors. This analysis resulted in novel predicted ceRNA networks for E-cadherin and EMT-inducing transcription factors (EMT-TFs), as well as the expansion of the molecular basis of the DGC.