Deep learning holds significant promise for improving clear cell renal cell carcinoma (ccRCC) diagnosis and treatment. However, studies concerning tumor microenvironment (TME) based on pathological images and artificial intelligence are very limited. In this multi-center cohort study, we developed a deep learning-based CD8+ T cell inflammation signature (DL-CD8T) for ccRCC from hematoxylin and eosin (H&E)-stained whole slide images through a clustering-constrained attention multiple-instance learning method. We developed the DL-CD8T for ccRCC in the training cohort, which could distinguish patients with higher level of CD8+ T cell inflammation through conventionally used H&E-stained whole slide image, with area under the curve (AUC) of 0.781 (0.741-0.817, p < 0.001). Verification of the DL-CD8T was also performed well in the independent CPTAC cohort, with AUC of 0.741 (0.599-0.853, p < 0.001). Further verification indicated that DL-CD8T(+) was significantly associated with high infiltration abundance of CD8+ T cell, higher expression levels of PD-1, PD-2, PD-L1, CTLA-4, TIM-3 and higher tumor mutation burden, which could distinguish ccRCC patients with high survival risk. In addition, DL-CD8T might predict combined targeted and immunotherapy benefit from digital pathology in ccRCC with hazard ratio of 0.27 (p = 0.029). Pending prospective validation, DL-CD8T might be used to improve risk stratification and inform personalized targeted therapy.

The tumor microenvironment (TME) is increasingly recognized as a complex ecosystem shaped by dynamic interactions among tumor cells, immune cells, and microbial components. While growing evidence has established the microbiota as a key regulator of antitumor immunity and immunotherapy response, the contribution of bacteriophages, the most abundant biological entities within microbial communities, has remained largely overlooked. Recent studies suggest that bacteriophages are not merely passive regulators of bacterial populations but can actively modulate host immune responses and influence tumor-associated immune landscapes. In this review, we summarize emerging evidence suggesting that bacteriophages may influence antitumor immunity through both direct and indirect mechanisms. Evidence from immune-cell and non-cancer experimental systems indicates that phage nucleic acids can engage TLR9-dependent sensing and, for selected phages, STING-associated inflammatory signaling; however, the relevance of these pathways within human tumors remains to be established. Indirectly, phages may alter microbial community structure and metabolic outputs, which could influence systemic immune tone and the composition of immune infiltrates within the TME. We further discuss accumulating data linking phageome features with tumor progression and responses to immune checkpoint blockade and other cancer therapies. However, much of the available evidence remains preclinical, indirect, or correlative, and causal roles for endogenous phages in human tumor immunity still require further validation. Distinct from the putative ecological and immunological roles of naturally occurring phages, engineered bacteriophages are being developed as therapeutic platforms for cancer immunotherapy, including tumor-antigen display, targeted delivery of immune agonists, cytokines or nucleic acids, and combination strategies with existing treatments. Finally, we address key methodological, mechanistic, and safety challenges that must be overcome to translate phage-based immunomodulation into clinical applications. Collectively, this review highlights the phageome as an emerging regulatory layer of tumor immunity and a promising, yet underexplored, target for therapeutic intervention.

Acquired endocrine resistance in ER+ breast cancer (BC) involves metabolic reprogramming, yet key drivers are unclear. Multi-omics of endocrine-resistant BC revealed upregulated oxidative phosphorylation (OXPHOS) and identified intercellular adhesion molecule 2 (ICAM2) as a biomarker of high-OXPHOS cells. ICAM2-positive cells were significantly enriched in resistant tumors and predicted poor survival, and were functionally essential for maintaining resistance and promoting metastasis in vivo. Mechanistically, ICAM2 binds dynein light chain DYNLT3 and the mitochondrial complex I subunit MT-ND2, thereby facilitating dynein-mediated mitochondrial trafficking and further modulating the assembly of mitochondrial complex I. Disrupting this interaction through ICAM2 knockdown or dynein inhibition (Ciliobrevin D) effectively suppressed OXPHOS activity. Importantly, ERα inhibition alleviates the transcriptional repression of ICAM2 by ERα. Therapeutically, combining the complex I inhibitor IACS-10759 with fulvestrant potently inhibited both tumor growth and metastasis. Collectively, these findings reveal that ICAM2 drives endocrine resistance via dynein-dependent OXPHOS activation, revealing a targetable axis in refractory ER+ BC. In summary, we establish ICAM2 as a novel biomarker and driver of endocrine resistance in ER⁺ breast cancer. ICAM2⁺ cancer cells-enriched in treatment-resistant tumors-maintain elevated OXPHOS by assembling a functional complex with dynein and mitochondrial Complex I, thereby promoting mitochondrial trafficking. Disruption of this axis, either through ICAM2 depletion or Complex I inhibition, re-sensitizes tumors to therapy, revealing a targetable metabolic dependency in resistant disease.

Background: Artificial intelligence-derived parameters hold substantial promise as indicators for tumor prognosis prediction and treatment guidance. However, existing studies have not sufficiently addressed the application of these parameters across different cancer types. Methods: We employed a deep learning algorithm to conduct a multi-cancer analysis for disease-free survival (MC-DFS) prediction using 8856 cases with associated whole-slide images and clinical data. The training cohort consisted of 7392 cases from the TCGA set (24 cancer types), and the independent external validation cohort comprised 1464 cases from the CPTAC and General Hospital sets (9 cancer types). A nomogram prediction signature for disease-free survival (NOMO) was developed by integrating the MC-DFS, tumor stage, and patient age. The prognostic model’s performance was validated in an independent cohort. Results: In the training and validation cohorts, the MC-DFS model achieved area under the curve (AUC) values of 0.750 and 0.682, respectively. It effectively differentiated patients with poorer disease-free survival, with hazard ratios of 4.823 (95% CI: 4.343–5.356, p < 0.0001) in the training cohort and 2.092 (95% CI: 1.472–2.971, p < 0.0001) in the validation cohort. Each cancer subtype’s analysis confirmed the model’s robust performance. Additionally, using nomogram analysis, we developed a multi-model prediction signature for disease-free survival across multiple cancer types based on MC-DFS and the clinicopathologic features in the training cohort. This enhanced model offers more precise risk stratification for stage I malignancies and complements the existing tumor staging systems by identifying high-risk patients. Conclusions: The newly developed MC-DFS shows marked improvements in prognostic predictions across multiple cancer types. With further validations across multiple centers, this nomogram prediction system could become a valuable practical tool for managing various cancers.

The CT image segmentation task of non-small cell lung cancer (NSCLC) suffers from the challenges of small target and severe background noise interference, etc. Current research incorporates a variety of neural network architectures to improve the overall segmentation performance, but most of the frameworks still have difficulty in balancing local detail preservation and global background modeling, especially the lack of ability to perceive small nodules. Therefore, this paper proposes dynamic multi-scale vision Mamba UNet (DMS-Vmunet). HDVSS modules for enhancing the segmentation performance on small targets are proposed in this framework, including Multi-SS2D module and IDConv module. The Multi-SS2D module enhances the local sensing bias and perception of small targets through multi-scale parallel branching design, while the IDConv module focuses the convolution operation on the important regions of small targets through adaptive dynamic convolution. In addition, in order to ensure the effective fusion of local and global features in parallel branches, a dual-path feature fusion (DFC) module is introduced to fuse the multi-scale feature information from multiple branches. Experimental results show that the proposed framework can effectively improve the performance of CT image segmentation for non-small cell lung cancer, especially for small target segmentation.

Background:As a common gynecological malignancy, cervical cancer has a rising incidence rate and mortality, which has brought huge pressure to global public health. Although immunotherapy has been applied in clinical practice, its therapeutic effect is still far from satisfactory. Methods:InferCNV was used to calculate the CNV score and the ssGSE, which is an algorithm to calculate the abundance of samples. CellChat analysis and pseudotime analysis were used to observe the evolution and interaction relationships between different clusters. Establish a prognostic model for CC patients using univariate, LASSO, and Cox analysis, and evaluate copy number variation and TME in low-risk groups. Finally, ssGSEA was applied to calculate the relationship between the hallmark gene sets and immune cycle steps and to calculate drug sensitivity in different risk groups using "oncopredict." A series of experiments including CCK-8 assay, clone formation, EdU assay, and Transwell assay were performed to detect the role of COL4A1 in CC. Results:The epithelial cells were divided into nine clusters. Among them, Cluster 8 has a lower CNV score, a lower degree of variation, and a better prognosis. After that, Cluster 8 sends a signal to fibroblasts through the PTN signaling pathway. A cervical cancer-related model (CCM) was constructed based on the marker genes of Cluster 8, and it can effectively distinguish the prognosis. There is a great difference in standardized TMB, immune cell infiltration, and ESTIMATE scores between the groups. Nine drugs were identified which may achieve better therapeutic effects when applied to low-risk patients. Finally, knockdown of COL4A1 inhibits the proliferation and metastatic ability of CC cells. Conclusion:Our study revealed different interactions between subgroups in the tumor microenvironment of CC epithelial cells. We established an effective prognostic model. Ultimately, through a series of in vitro function experiments, COL4A1 was recognized as a new potential target for the therapeutic intervention of CC.

Sphingosine-1-phosphate receptor-1 (S1PR1), a G protein-coupled receptor, has been reported to be involved in lymphangiogenesis. Degradations of extracellular matrix (ECM) are recognized as dynamic modulators in regulating the formation of new lymphatic vessels. However, little research has studied the ECM on S1PR1 in the regulation of lymphatic endothelial cells (LECs) in tumor lymphangiogenesis. Here we attempt to investigate hyaluronan fragments abundant in tumor microenvironment (TME) on S1PR1 in new lymphatic vessel formation. First, we verified that low molecular weight hyaluronan (LMW-HA) derived from tumor cells could promote LECs migration and capillary-like tube formation. Then, we demonstrated that S1PR1 on LECs underwent internalization into the endoplasmic reticulum in response to LMW-HA treatments. Notably, the S1PR1 endocytosis could upregulate lymphangiogenesis. Next, we found that the ablation of lymphatic vessel endothelial hyaluronan receptor 1 (LYVE-1) could attenuate the S1PR1 endocytosis, implying a novel role of LMW-HA/LYVE-1 in the S1PR1 cycling pathway. Furthermore, we identified that LMW-HA/LYVE-1 interaction could activate Src kinase which in turn upregulates S1PR1 tyrosine phosphorylation, resulting in S1PR1 endocytosis. Collectively, our findings suggested that hyaluronan fragments in TME could induce S1PR1 internalization in LECs, leading to lymphangiogenesis promotion.

Luminal type breast cancer (BrCa) cells invade into the surrounding tissues as collective strands, making them more metastatic than single cells. We have previously reported that the leading subpopulation of collective cells expressed high levels of CD44, which was associated with enhanced migratory and invasive potential of BrCa. It is crucial to elucidate how CD44 becomes enriched in leader cells and contributes to collective migration. In this study, we aimed to uncover the mechanisms responsible for CD44 upregulation in this context. First, we demonstrated that CD44 could facilitate dynamic lamellipodia formation by interacting with cytoskeletal proteins through its cytoplasmic domain. Then, we identified that a transcriptional regulator, Sin3a, was remarkably upregulated at the front edge of collectively migrating cells, exhibiting a correlation with enhanced CD44 expression. Notably, the knockdown of Sin3a effectively suppressed CD44 enrichment and lamellipodia outgrowth in leader cells, resulting in a significantly decreased cohesive movement of BrCa cells in vitro and in vivo. Our findings suggested that Sin3a was a novel regulator in CD44-facilitated lamellipodia formation and subsequent collective migration. This study elucidated the molecular mechanism underlying CD44 upregulation during collective migration of luminal-type BrCa cells, providing potential therapeutic targets to prevent cancer metastasis.

Triple-negative breast cancer (TNBC) cells are rich in glycocalyx (GCX) that is closely correlated with the reorganization of cytoskeletal filaments. Most studies have focused on cell membrane glycoproteins in this context, but rarely on the significance of glycosaminoglycans, particularly the hyaluronan (HA)-associated GCX. Here, we reported that removal of GCX HA could significantly increase breast cancer cells (BCCs) stiffness, leading to impaired cell growth and decreased stem-like properties. Furthermore, we found that the delay of TNBC cells progression could be restored after the cells were re-softened. Meanwhile, in vivo studies revealed that hyaluronidase (HAase)-pretreated BCCs displayed reduced tumor growth and migration. Intriguingly, we identified that ZC3H12A, a zinc-finger RNA binding protein encoded gene, was significantly upregulated after the GCX HA impairment. Of note, knockdown of ZC3H12A could soften the HAase-treated TNBC cells, implying a GCX HA-ZC3H12A regulation on cell stiffening. Taken together, our findings suggested that the breakdown of pericellular HA coat could influence TNBC cells mechanical properties which might be helpful to the future breast cancer research.

Macropinocytosis (MP) in cancer cells is an endocytic process for nutrient extraction initiated by membrane ruffling that supports tumor progression. Hyaluronan (HA), a major component of glycocalyx coating the surface of living cells, is reported to influence the membrane morphology. However, whether HA-related glycocalyx (HA-GCX) regulates MP through membrane reconstruction remains unclear. Here, we demonstrated that the HA-GCX thickness was positively correlated with MP activity in cancer cells. Intriguingly, the disruption of HA-GCX could suppress MP by decreasing membrane ruffle formation. Furthermore, the knockdown of cyclin-dependent kinase inhibitor 1 (CDKN1A/p21) and its downstream matrix metalloproteinase (MMP) 1/9, which were identified to be significantly upregulated following HA-GCX removal, could rescue MP. Importantly, the breakdown of HA-GCX could impede cancer cell proliferation through inhibition of MP. Taken together, our study revealed a novel regulatory role of HA-GCX in MP, which might provide a promising therapeutic avenue in cancer treatment.

IntroductionThe extracellular matrix (ECM) stiffness serves as a critical biomechanical regulator of cellular behavior. However, its specific roles on lymphatic endothelial cells (LECs) remains poorly characterized, particularly in the context of tumorigenesis where progressive matrix stiffening is a hallmark of the tumor microenvironments (TME).MethodsThe effects of ECM stiffness on LEC proliferation and migration were assessed using a tunable polyacrylamide hydrogel system. Differential gene expression in LECs on soft versus stiff substrates was identified by RNA-seq. To evaluate the stiffness-dependent regulation of FAT1 and its downstream mechanisms, we performed RT-qPCR, Western blot, immunofluorescence, wound healing, and spheroid assays. Furthermore, immunofluorescence was also used to compare FAT1 expression in tumor-associated versus normal lymphatic vessels.ResultsIn this study, we demonstrated that ECM stiffening significantly promotes LEC proliferation and migration. Notably, we observed marked downregulation of FAT1 expression in LECs cultured on tumor stiffness-mimicking matrix, a finding validated in clinical breast cancer specimens and murine models of breast cancer and melanoma. Mechanistic investigations identified FAT1 as a pivotal mechanotransducer that orchestrates LECs functional responses to biomechanical cues. Specifically, the knockdown of FAT1 facilitated β-catenin nuclear translocation, activating transcription of cell cycle regulators Myc and Cyclin D1 to coordinately promote LEC proliferation. Furthermore, FAT1 deficiency increased LEC mechanosensitivity by modulating focal adhesion formation, inducing cytoskeleton reorganization and consequent enhancement of migratory potentials.DiscussionTogether, our study uncovers FAT1 as a pivotal mechanosensor in LECs and highlight its significance in the biomechanical regulation. Targeting the FAT1-mediated signaling pathways may serve as a novel therapeutic strategy to inhibit tumor lymphatic metastasis.

Background Metastasis renal cell carcinoma (RCC) patients have extremely high mortality rate. A predictive model for RCC micrometastasis based on pathomics could be beneficial for clinicians to make treatment decisions.Methods A total of 895 formalin-fixed and paraffin-embedded whole slide images (WSIs) derived from three cohorts, including Shanghai General Hospital (SGH), Clinical Proteomic Tumor Analysis Consortium (CPTAC) and Cancer Genome Atlas (TCGA) cohorts, and another 588 frozen section WSIs from TCGA dataset were involved in the study. The deep learning-based strategy for predicting lymphatic metastasis was developed based on WSIs through clustering-constrained-attention multiple-instance learning method and verified among the three cohorts. The performance of the model was further verified in frozen-pathological sections. In addition, the model was also tested the prognosis prediction of patients with RCC in multi-source patient cohorts.Results The AUC of the lymphatic metastasis prediction performance was 0.836, 0.865 and 0.812 in TCGA, SGH and CPTAC cohorts, respectively. The performance on frozen section WSIs was with the AUC of 0.801. Patients with high deep learning-based prediction of lymph node metastasis values showed worse prognosis.Conclusions In this study, we developed and verified a deep learning-based strategy for predicting lymphatic metastasis from primary RCC WSIs, which could be applied in frozen-pathological sections and act as a prognostic factor for RCC to distinguished patients with worse survival outcomes.

BACKGROUND:Although separate analysis of individual factor can somewhat improve the prognostic performance, integration of multimodal information into a single signature is necessary to stratify patients with clear cell renal cell carcinoma (ccRCC) for adjuvant therapy after surgery. METHODS:A total of 414 patients with whole slide images, computed tomography images, and clinical data from three patient cohorts were retrospectively analyzed. The authors performed deep learning and machine learning algorithm to construct three single-modality prediction models for disease-free survival of ccRCC based on whole slide images, cell segmentation, and computed tomography images, respectively. A multimodel prediction signature (MMPS) for disease-free survival were further developed by combining three single-modality prediction models and tumor stage/grade system. Prognostic performance of the prognostic model was also verified in two independent validation cohorts. RESULTS:Single-modality prediction models performed well in predicting the disease-free survival status of ccRCC. The MMPS achieved higher area under the curve value of 0.742, 0.917, and 0.900 in three independent patient cohorts, respectively. MMPS could distinguish patients with worse disease-free survival, with HR of 12.90 (95% CI: 2.443-68.120, P <0.0001), 11.10 (95% CI: 5.467-22.520, P <0.0001), and 8.27 (95% CI: 1.482-46.130, P <0.0001) in three different patient cohorts. In addition, MMPS outperformed single-modality prediction models and current clinical prognostic factors, which could also provide complements to current risk stratification for adjuvant therapy of ccRCC. CONCLUSION:Our novel multimodel prediction analysis for disease-free survival exhibited significant improvements in prognostic prediction for patients with ccRCC. After further validation in multiple centers and regions, the multimodal system could be a potential practical tool for clinicians in the treatment for ccRCC patients.

When a curling rock slides on an ice sheet with an initial rotation, a lateral movement occurs, which is known as the curling phenomenon. The force of friction between the curling rock and the ice sheet changes continually with changes in the environment; thus, the sport of curling requires great skill and experience. The throwing of the curling rock is a great challenge in robot design and control, and existing curling robots usually adopt a combination scheme of a wheel chassis and gripper that differs significantly from human throwing movements. A hexapod curling robot that imitates human kicking, sliding, pushing, and curling rock rotating was designed and manufactured by our group, and completed a perfect show during the Beijing 2022 Winter Olympics Games. Smooth switching between the walking and throwing tasks is realized by the robot’s morphology transformation based on leg configuration switching. The robot’s controlling parameters, which include the kicking velocity vk, pushing velocity vp, orientation angle θc, and rotation velocity ω, are determined by aiming and sliding models according to the estimated equivalent friction coefficient μequ and ratio e of the front and back frictions. The stable errors between the target and actual stopping points converge to 0.2 and 1.105 m in the simulations and experiments, respectively, and the error shown in the experiments is close to that of a well-trained wheelchair curling athlete. This robot holds promise for helping ice-makers rectify ice sheet friction or assisting in athlete training.

Glioblastoma (GBM) is the most aggressive form of glioma, characterized by high mortality and poor prognosis. Dysregulation of microRNAs (miRNAs) plays a critical role in the progression and metastasis of GBM. This study aimed to investigate the role and molecular mechanism of miR-124-3p in GBM. Levels of miR-124-3p, EPHA2, and ALKBH5 were measured using quantitative real-time polymerase chain reaction (qRT-PCR). Cell proliferation, migration, invasion, and stemness were assessed using the Cell Counting Kit-8 (CCK-8), colony formation, Transwell, and sphere formation assays, respectively. Bioinformatics prediction, dual-luciferase reporter assays, and RNA pull-down experiments were employed to validate the target of miR-124-3p. RNA binding protein immunoprecipitation (RIP) and methylated RNA immunoprecipitation (Me-RIP) were utilized to evaluate the regulation of miR-124-3p maturation by ALKBH5. The results indicated that overexpression of miR-124-3p inhibited the proliferation, migration, invasion, and stemness of GBM cells. EPHA2 was identified as a direct downstream target of miR-124-3p, and its overexpression reversed the inhibitory effects of miR-124-3p on cellular functions. Furthermore, miR-124-3p targeted EPHA2 to inactivate the Wnt/β-catenin pathway. Additionally, ALKBH5 negatively regulated miR-124-3p by impeding its processing. In conclusion, knockdown of ALKBH5 promoted the processing of pri-miR-124-3p, increasing mature miR-124-3p levels, which inhibited the malignant behaviors of GBM cells by targeting EPHA2. These findings highlight the importance of the ALKBH5/miR-124-3p/EPHA2 axis in GBM.

Objective: This study aimed to analyze the functional roles and molecular mechanism of Wilms’ tumor 1-associating protein (WTAP) in the tumorigenesis of nonsmall-cell lung cancer (NSCLC). Methods: Retrospective analysis was used. Tumor tissues and surrounding nontumor tissues of 150 patients with NSCLS who were surgically resected in the Fourth Hospital of Hebei Medical University from January 2016 to January 2018 were selected. The expression of WTAP in NSCLC tissues was detected by immunohistochemistry. Clinicopathologic parameters were then subjected to univariate and multivariate Cox regression analysis in purpose of uncovering the independent risk factors for overall survival time. MTS (3-[4,5-dimethylthiazol-zyl]-5-[3-carboxymethoxyphenyl]-2-[4-sulfophenyl]-2H-tetrazoliuzolium, inner salt) assay, colony formation assay, and transwell assays were performed to estimate cell proliferation, migration, and invasion. Meanwhile, the relationship between WTAP and the cell migration and invasion marker-related proteins were evaluated by Western blot analysis and RT-qPCR. WTAP expression was knocked-down in cell lines by shRNA, and RNA-Seq was performed to investigate the pathways regulated by WTAP. Results: In NSCLC patients, WTAP was highly expressed in tumor tissues and the higher expression was significantly associated with poor overall survival (OS) (P<0.01). Compared with the control group in vitro, the overexpression of WTAP could significantly promote cell proliferation, migration, and invasion (P<0.01), while knock-down WTAP significantly reduces the above effects (P<0.01). In a mouse orthotopic implantation model, higher WTAP abundance could significantly promote tumor enlargement compared with the control group (P<0.01). Compared with the control group, the knock-down of WTAP significantly inhibit the expression of carcinoembryonic antigen-related cell adhesion molecule 5 (CEACAM5) in cell lines (P<0.01). Besides, in NSCLC, knocked-down CEACAM5 significantly reduced the impact of WTAP on cell proliferation, migration, and invasion compared with the control group (P<0.05). Conclusions: This study suggests that high expression of WTAP was associated with poor clinical outcomes. CEACAM5 may play a synergistic role with WTAP to jointly promote NSCLC progression by enhancing cell proliferation, invasion, and migration.

There is an urgent clinical demand to explore novel diagnostic and prognostic biomarkers for renal cell carcinoma (RCC). We proposed deep learning-based artificial intelligence strategies. The study included 1752 whole slide images from multiple centres. Based on the pixel-level of RCC segmentation, the diagnosis diagnostic model achieved an area under the receiver operating characteristic curve (AUC) of 0.977 (95% CI 0.969-0.984) in the external validation cohort. In addition, our diagnostic model exhibited excellent performance in the differential diagnosis of RCC from renal oncocytoma, which achieved an AUC of 0.951 (0.922-0.972). The graderisk for the recognition of high-grade tumour achieved AUCs of 0.840 (0.805-0.871) in the Cancer Genome Atlas (TCGA) cohort, 0.857 (0.813-0.894) in the Shanghai General Hospital (General) cohort, and 0.894 (0.842-0.933) in the Clinical Proteomic Tumor Analysis Consortium (CPTAC) cohort, for the recognition of high-grade tumour. The OSrisk for predicting 5-year survival status achieved an AUC of 0.784 (0.746-0.819) in the TCGA cohort, which was further verified in the independent general cohort and the CPTAC cohort, with AUCs of 0.774 (0.723-0.820) and 0.702 (0.632-0.765), respectively. Moreover, the competing-risk nomogram (CRN) showed its potential to be a prognostic indicator, with a hazard ratio (HR) of 5.664 (3.893-8.239, p<0.0001), outperforming other traditional clinical prognostic indicators. Kaplan-Meier survival analysis further illustrated that our CRN could significantly distinguish patients with high survival risk. Deep learning-based artificial intelligence could be a useful tool for clinicians to diagnose and predict the prognosis of RCC patients, thus improving the process of individualised treatment.

Trastuzumab is widely used in human epidermal growth factor receptor 2 (HER2)-positive gastric cancer (GC) therapy, but ubiquitous resistance limits its clinical application. In this study, we first showed that CD44 antigen is a significant predictor of overall survival for patients with HER2-positive GC. Next, we found that CD44 could be co-immunoprecipitated and co-localized with HER2 on the membrane of GC cells. By analyzing the interaction between CD44 and HER2, we identified that CD44 could upregulate HER2 protein by inhibiting its proteasome degradation. Notably, the overexpression of CD44 could decrease the sensitivity of HER2-positive GC cells to trastuzumab. Further mechanistic study showed that CD44 upregulation could induce its ligand, hyaluronan (HA), to deposit on the cancer cell surface, resulting in covering up the binding sites of trastuzumab to HER2. Removing the HA glycocalyx restored sensitivity of the cells to trastuzumab. Collectively, our findings suggested a role for CD44 in regulating trastuzumab sensitivity and provided novel insights into HER2-targeted therapy.

Background: Clear cell renal cell carcinoma (ccRCC) is the most common renal-related tumor with high heterogeneity. There is still an urgent need for novel diagnostic and prognostic biomarkers for ccRCC. Methods: We proposed a weakly-supervised deep learning strategy using conventional histology of 1752 whole slide images from multiple centers. Our study was demonstrated through internal cross-validation and external validations for the deep learning-based models. Results: Automatic diagnosis for ccRCC through intelligent subtyping of renal cell carcinoma was proved in this study. Our graderisk achieved aera the curve (AUC) of 0.840 (95% confidence interval: 0.805-0.871) in the TCGA cohort, 0.840 (0.805-0.871) in the General cohort, and 0.840 (0.805-0.871) in the CPTAC cohort for the recognition of high-grade tumor. The OSrisk for the prediction of 5-year survival status achieved AUC of 0.784 (0.746-0.819) in the TCGA cohort, which was further verified in the independent General cohort and the CPTAC cohort, with AUC of 0.774 (0.723-0.820) and 0.702 (0.632-0.765), respectively. Cox regression analysis indicated that graderisk, OSrisk, tumor grade, and tumor stage were found to be independent prognostic factors, which were further incorporated into the competing-risk nomogram (CRN). Kaplan-Meier survival analyses further illustrated that our CRN could significantly distinguish patients with high survival risk, with hazard ratio of 5.664 (3.893-8.239, p < 0.0001) in the TCGA cohort, 35.740 (5.889-216.900, p < 0.0001) in the General cohort and 6.107 (1.815 to 20.540, p < 0.0001) in the CPTAC cohort. Comparison analyses conformed that our CRN outperformed current prognosis indicators in the prediction of survival status, with higher concordance index for clinical prognosis.