GMILT: A Novel Transformer Network That Can Noninvasively Predict EGFR Mutation Status.

Noninvasively and accurately predicting the epidermal growth factor receptor (EGFR) mutation status is a clinically vital problem. Moreover, further identifying the most suspicious area related to the EGFR mutation status can guide the biopsy to avoid false negatives. Deep learning methods based on computed tomography (CT) images may improve the noninvasive prediction of EGFR mutation status and potentially help clinicians guide biopsies by visual methods. Inspired by the potential inherent links between EGFR mutation status and invasiveness information, we hypothesized that the predictive performance of a deep learning network can be improved through extra utilization of the invasiveness information. Here, we created a novel explainable transformer network for EGFR classification named gated multiple instance learning transformer (GMILT) by integrating multi-instance learning and discriminative weakly supervised feature learning. Pathological invasiveness information was first introduced into the multitask model as embeddings. GMILT was trained and validated on a total of 512 patients with adenocarcinoma and tested on three datasets (the internal test dataset, the external test dataset, and The Cancer Imaging Archive (TCIA) public dataset). The performance (area under the curve (AUC) = 0.772 on the internal test dataset) of GMILT exceeded that of previously published methods and radiomics-based methods (i.e., random forest and support vector machine) and attained a preferable generalization ability (AUC = 0.856 in the TCIA test dataset and AUC = 0.756 in the external dataset). A diameter-based subgroup analysis further verified the efficiency of our model (most of the AUCs exceeded 0.772) to noninvasively predict EGFR mutation status from computed tomography (CT) images. In addition, because our method also identified the "core area" of the most suspicious area related to the EGFR mutation status, it has the potential ability to guide biopsies.

The Better Survival of MSI Subtype Is Associated With the Oxidative Stress Related Pathways in Gastric Cancer.

Background: Gastric cancer (GC) is the third leading fatal cancer in the world and its incidence ranked second among all malignant tumors in China. The molecular classification of GC, proposed by the The Cancer Genome Atlas (TCGA), was added to the updated edition (2019) of WHO classification for digestive system tumor. Although MSI and EBV subtypes appeared as ever-increasingly significant roles in immune checkpoint inhibitor therapy, the underlying mechanisms are still unclear. Methods: We systematically summarized the relationship between EBV, d-MMR/MSI-H subtypes and clinicopathological parameters in 271 GC cases. Furthermore, GSE62254/ACRG and TCGA-STAD datasets, originated from Gene Expression Omnibus (GEO) and TCGA respectively, were analyzed to figure out the prognosis related molecular characteristics by bioinformatics methods. Results: Patients with MSI subtype had better prognosis than the MSS subtype (P = 0.013) and considered as an independent biomarker by the univariate analysis (P = 0.017) and multivariate analysis (P = 0.050). While there was no significant difference between EBV positive and negative tissues (P = 0.533). The positive prognostic value conferred by MSI in different cohorts was revalidated via the clinical analysis of GSE62254/ACRG and TCGA-STAD datasets regardless of race. Then key gene module that tightly associated with better status and longer OS time for MSI cases was obtained from weighted gene co-expression network analysis(WGCNA). NUBP2 and ENDOG were screened from the gene cluster and oxidative phosphorylation, reactive oxygen species(ROS) and glutathione metabolism were analyzed to be the differential pathways in their highly expressed groups. Conclusions: Our results manifested the significant prognostic value of MSI in Chinese GC cohort and comparisons with other populations. More opportunities to induce apoptosis of cancer cells, led by the unbalance between antioxidant system and ROS accumulation, lay foundations for unveiling the better prognosis in MSI phenotype through the bioinformatics analysis.

Integrated assessment of PD-L1 expression and molecular classification facilitates therapy selection and prognosis prediction in gastric cancer.

PURPOSE: Targeting the PD-1/PD-L1 pathway has emerged as a novel therapy for cancer. To identify rational candidates for anti-PD-1/PD-L1 immunotherapy in gastric cancer (GC), the abundance of PD-L1 expression was evaluated on a kind of biomarker-based molecular classification for shaping prognosis and treatment planning. METHODS: One hundred and sixty-five GCs were classified into five subgroups using immunohistochemistry (IHC) and in situ hybridization (ISH) methods, based on a panel of seven markers (MLH1, PMS2, MSH2, MSH6, E-cadherin, P53, and Epstein-Barr virus mRNA). The expression of PD-L1 in GC tissues was analyzed immunohistochemically. RESULTS: The five categories (Epstein-Barr virus positivity, microsatellite instability, aberrant E-cadherin, aberrant P53 expression, and normal P53 expression) correspond to the reported molecular subgroups for similar proportions and clinicopathologic characteristics. Survival analysis indicated that subgroups with aberrant E-cadherin expression independently predicted a worse prognosis in GC patients (HR=2.51, P=0.010). The clinical and prognostic profiles produced by this stratification in nonintestinal-type GC were distinguishable from those in intestinal-type. Although PD-L1 was not a significant prognostic factor, that more frequent presence of PD-L1-positive in microsatellite instability tumors than other subtypes (P<0.010) hinted at a prolonged clinical course. Moreover, the lowest level of PD-L1 but the highest of Her2 was observed in the group of aberrant P53, namely it was suggested that there was a negative correlation between PD-L1 and Her2 overexpression. CONCLUSION: Different molecular subtypes in GC may have a tendency to react differently to anti-PD-L1/PD-1 immunotherapy or anti-Her2 therapy. A combination of PD-L1 expression and this cost-effective classification strategy would be helpful for predicting prognosis and promoting personalized therapy in clinical practice.

Downregulation of miRNA-214 in cancer-associated fibroblasts contributes to migration and invasion of gastric cancer cells through targeting FGF9 and inducing EMT.

BACKGROUND: Cancer-associated fibroblasts (CAFs), one of the principal constituents of the tumor microenvironment, have a pivotal role in tumor progression. Dysregulation of microRNAs (miRNAs) in CAFs contributes to the tumor-promoting ability of CAFs. However, the mechanism underlying the involvement of miRNAs in CAFs of gastric cancer (GC) is not fully understood. This study aimed to explore the effects of miRNA-214 in CAFs on GC migration and invasion. METHODS: The primary CAFs and corresponding normal fibroblasts (NFs) were isolated. Cell counting kit-8, EdU cell proliferation staining and Transwell assays were used to determine the role of miRNA-214 in GC progression. Real-time polymerase chain reaction, Western blot analysis, and dual-luciferase reporter assay were performed to verify the target genes of miRNA-214. Immunofluorescence and Western blot analysis were applied to detect the expression of epithelial-mesenchymal transition (EMT) markers. Immunohistochemistry and in situ hybridization were implemented to analyze the fibroblast growth factor 9 (FGF9) and miRNA-214 expression in human GC tissues, respectively. Finally, to assess its prognostic relevance, Kaplan-Meier survival analysis was conducted. RESULTS: MiRNA-214 was significantly downregulated in CAFs of GC compared with NFs. The upregulation of miRNA-214 in CAFs inhibited GC cell migration and invasion in vitro but failed to affect proliferation. Moreover, GC cells cultured with conditioned medium from CAFs transfected with miR-214 mimic showed increased expression of E-cadherin and decreased expression of Vimentin, N-cadherin and Snail, indicating the suppression of EMT of GC cells. Furthermore, FGF9 was proved to be a direct target gene of miR-214. The expression of FGF9 was higher in CAFs than that in tumor cells not only in primary tumor but also in lymph node metastatic sites (30.0% vs 11.9%, P < 0.01 and 32.1% vs 12.3%, P < 0.01, respectively). Abnormal expression of FGF9 in CAFs of lymph node metastatic sites was significantly associated with poor prognosis in patients with GC (P < 0.05). CONCLUSIONS: This study showed that miR-214 inhibited the tumor-promoting effect of CAFs on GC through targeting FGF9 in CAFs and regulating the EMT process in GC cells, suggesting miRNA-214/FGF9 in CAFs as a potential target for therapeutic approaches in GC.

Cancer associated fibroblasts tailored tumor microenvironment of therapy resistance in gastrointestinal cancers.

Gastrointestinal cancers (GI), are a group of highly aggressive malignancies with heavy cancer-related mortalities. Even if continued development of therapy methods, therapy resistance has been a great obstruction for cancer treatment and thereby inevitably leads to depressed final mortality. Peritumoral cancer associated fibroblasts (CAFs), a versatile population assisting cancer cells to build a facilitated tumor microenvironment (TME), has been demonstrated exerting a promotion influence on cancer proliferation, migration, invasion, metastasis, and also therapy resistance. In this review, we provide an update progress in describing how CAFs mediate therapy resistance in GI by various means, meanwhile highlight the crosstalk between CAFs and cancer cells and present some vital signaling pathways activated by CAFs in this resistant process. Furthermore, we discuss the current advances in adopting novel drugs against CAFs and how the knowledge contributing to improved therapy efficacy in clinical practice. In sum, CAFs create a therapy-resistant TME in several aspects of GI progression, although some key problems about distinguishing CAFs subpopulations and controversial issues on pleiotropic CAFs in medication need to be solved for subsequent clinical application. Predictably, targeting therapy-resistant CAFs is a promising adjunctive treatment to benefit GI patients.

Influence of the CYP4F2 polymorphism on the risk of hemorrhagic complications in coumarin-treated patients.

OBJECTIVES: To evaluate the impact of the CYP4F2 polymorphism on bleeding complications and over-anticoagulation due to coumarin. METHODS: A comprehensive literature search was performed to look for eligible studies published prior to February 2015 in EMBASE and PubMed. References were strictly identified by inclusion and exclusion criteria, and authors of primary studies were consulted for additional information and data. Revman 5.3 software was used to analyze the impact of the CYP4F2 polymorphism on hemorrhagic complications and over-anticoagulation events (international normalized ratio greater than 4). RESULTS: Eight studies involving 3,101 samples met the specified inclusion criteria. Compared with wild-type homozygotes (CYP4F2*1*1), carriers of the CYP4F2*3 variant had no significant effects on total bleeding events (odds ratio [OR]: 0.86; 95% confidence interval [CI]: 0.71-1.05; p=0.15), major hemorrhage complications in coumarin users (OR: 0.80; 95% CI: 0.64-1.01; p=0.06). Patients carried CYP4F2*3 also had nonsignificant associations with the risk of over-anticoagulation (relative risk [RR]: 079; 95% CI: 0.59-1.06; p=0.12). We found a lower risk in patients with homozygotes for CYP4F2*3, but there was no statistical significance (RR: 0.66; 95% CI: 0.43-1.01; p=0.05). CONCLUSION: This meta-analysis indicated the impact of the CYP4F2 polymorphism on bleeding complications and over-anticoagulation in coumarin-treated patients failed to reach the level of statistical significance. However, large-scale and well designed studies are necessary to determine conclusively the association between the CYP4F2 polymorphism and hemorrhage risk.