Search

Purpose: The application of deep learning methods to the intelligent diagnosis of diseases has been the focus of intelligent medical research. When dealing with image classification tasks, if the lesion area is small and uneven, the background image involved in the training will affect the ultimate accuracy in determining the extent of the lesion. We did not follow the traditional approach of building an intelligent system to assist physicians in diagnosis from the perspective of CNN models, but instead proposed a pure transformer framework that can be used for diagnostic grading of pathological images. Methods: We propose a Symmetric Mask Pre-Training vision Transformer SMiT model for grading pathological cancer images. SMiT performs a symmetrically identical high probability sparsification of the input image token sequence at the first and last encoder layer positions to pre-train visual transformers, and the parameters of the baseline model are fine-tuned after loading the pre-training weights, allowing the model to concentrate more on extracting detailed features in the lesion region, effectively getting rid of the potential feature dependency problem. Results: SMiT achieved 92.8% classification accuracy on 4500 histopathological images of colorectal cancer processed by Gaussian filter denoising. We validated the effectiveness and generalizability of this study's methodology on the publicly available diabetic retinopathy dataset APTOS2019 from Kaggle and achieved quadratic Cohen Kappa, accuracy and F1-score of 91.9%, 86.91% and 72.85%, respectively, which were 1–2% better than previous studies based on CNN models. Conclusion: SMiT uses a simpler strategy to achieve better results to assist physicians in making accurate clinical decisions, which can be an inspiration for making good use of the visual transformers in the field of medical imaging. [ABSTRACT FROM AUTHOR]

Objective: To analyze the prognostic significance of plasma total cholesterol (TC) and high-density lipoprotein (HDL) in patients with gastric cancer receiving oxaliplatin combination with S-1 (SOX) chemotherapy after radical resection and to establish models of relevant prognostic influencing factors. Method: The clinicopathologic features of 301 patients treated with SOX after radical gastrectomy were retrospectively analyzed. Univariate analysis, multivariate analysis, and the Kaplan–Meier survival curve were used to analyze the prognostic value of TC and HDL in patients undergoing adjuvant SOX chemotherapy after curative gastric surgery. Based on the results of multivariate Cox regression, we constructed nomograms to predict 1 and 3-year CSS (cancer-specific survival time) and DFS (disease-free survival time) in patients with adjuvant chemotherapy after radical gastrectomy. We assessed the model's accuracy using the consistency index (C index) and calibration curve, and the ROC curve and DCA curve were applied to compare it with TNM staging. Results: Multivariate analysis revealed that TC and HDL were independent influencing factors of CSS, while HDL was an independent factor unique to DFS. According to Kaplan–Meier curves, low TC and HDL suggested poor survival (P < 0.001). The relevant prognostic factors from the multivariate study were used to build nomograms for DFS and CSS. Both DFS and CSS had C index and AUC values that were higher than 0.71. The calibration curves demonstrated that the predicted results are consistent with the observed results. The AUC valves for DFS and CSS in our models were higher than TNM staging. The decision curve analysis indicated that net benefits were moderately positive. Significant differences in survival were seen between the high and low-risk groups according to the nomogram risk score. Conclusions: TC and HDL have a certain significance for the prognosis of patients with gastric cancer after radical resection receiving adjuvant SOX chemotherapy. Lower TC and HDL suggested poor DFS and CSS. Both prediction models for CSS and DFS demonstrated good predictive ability and had a higher predictive value than the TNM staging system. [ABSTRACT FROM AUTHOR]

Purpose: Lung adenocarcinoma (LUAD) is a malignant tumor with a high lethality rate. Immunotherapy has become a breakthrough in cancer treatment and improves patient survival and prognosis. Therefore, it is necessary to find new immune-related markers. However, the current research on immune-related markers in LUAD is not sufficient. Therefore, there is a need to find new immune-related biomarkers to help treat LUAD patients. Methods: In this study, a bioinformatics approach combined with a machine learning approach screened reliable immune-related markers to construct a prognostic model to predict the overall survival (OS) of LUAD patients, thus promoting the clinical application of immunotherapy in LUAD. The experimental data were obtained from The Cancer Genome Atlas (TCGA) database, including 535 LUAD and 59 healthy control samples. Firstly, the Hub gene was screened using a bioinformatics approach combined with the Support Vector Machine Recursive Feature Elimination algorithm; then, a multifactorial Cox regression analysis by constructing an immune prognostic model for LUAD and a nomogram to predict the OS rate of LUAD patients. Finally, the regulatory mechanism of Hub genes in LUAD was analyzed by ceRNA. Results: Five genes, ADM2, CDH17, DKK1, PTX3, and AC145343.1, were screened as potential immune-related genes in LUAD. Among them, ADM2 and AC145343.1 had a good prognosis in LUAD patients (HR < 1) and were novel markers. The remaining three genes screened were associated with poor prognosis in LUAD patients (HR > 1). In addition, the experimental results showed that patients in the low-risk group had better OS rates than those in the high-risk group (P < 0.001). Conclusion: In this paper, we propose an immune prognostic model to predict OS rate in LUAD patients and show the correlation between five immune genes and the level of immune-related cell infiltration. It provides new markers and additional ideas for immunotherapy in patients with LUAD. [ABSTRACT FROM AUTHOR]

Purpose: Skin cancer is one of the ten most common cancer types in the world. Early diagnosis and treatment can effectively reduce the mortality of patients. Therefore, it is of great significance to develop an intelligent diagnosis system for skin cancer. According to the survey, at present, most intelligent diagnosis systems of skin cancer only use skin image data, but the multi-modal cross-fusion analysis using image data and patient clinical data is limited. Therefore, to further explore the complementary relationship between image data and patient clinical data, we propose multimode data fusion diagnosis network (MDFNet), a framework for skin cancer based on data fusion strategy. Methods: MDFNet establishes an effective mapping among heterogeneous data features, effectively fuses clinical skin images and patient clinical data, and effectively solves the problems of feature paucity and insufficient feature richness that only use single-mode data. Results: The experimental results present that our proposed smart skin cancer diagnosis model has an accuracy of 80.42%, which is an improvement of about 9% compared with the model accuracy using only medical images, thus effectively confirming the unique fusion advantages exhibited by MDFNet. Conclusions: This illustrates that MDFNet can not only be applied as an effective auxiliary diagnostic tool for skin cancer diagnosis, help physicians improve clinical decision-making ability and effectively improve the efficiency of clinical medicine diagnosis, but also its proposed data fusion method fully exerts the advantage of information convergence and has a certain reference value for the intelligent diagnosis of numerous clinical diseases. [ABSTRACT FROM AUTHOR]

Purpose: The rise of immune checkpoint inhibitors (ICIs) in recent years has coincided with unusual clinical response patterns. Modification of the sum of longest diameters (SLD)-based threshold that reflecting dynamic change of the tumor burden and predicting response to ICIs, may markedly improve current treatment regimens. Methods: The baseline and post-treatment SLD of target lesion was recorded and the maximum percent change of the SLD from baseline was designated as SLD-change score. The optimal cut-off value was obtained using the X-tile program. The relationship between SLD-change score and survival outcome (PFS, OS) was evaluated. Results: 10% cut-off value of SLD-change score was found to be most distinctive for PFS. Responders defined according to this cut-off value showed a significant improvement for PFS and OS. Bone metastasis and brain metastasis were also two independent prognostic factors of PFS and OS, respectively. Conclusions: 10% SLD change score could discriminate for ICIs treatment response, which holds great promise in promoting the development of precise immunotherapeutic strategy. [ABSTRACT FROM AUTHOR]

Purpose: Tyrosine kinase inhibitors (TKIs) targeting epidermal growth factor receptor (EGFR) improve the prognosis of lung adenocarcinoma (LUAD). However, the factors affecting its clinical efficacy remain unclear. This study aimed to determine the correlation between Osteopontin (OPN) and EGFR, and explore the inhibitory effect of first-generation TKI gefitinib on LUAD cells. Methods: The correlation between OPN and EGFR was determined through bioinformatics technology, and the clinical information as well as samples of related patients were collected to verify the relationship between them. Using three different NSCLC cell lines A549, H1299 and PC9, we studied the effects of OPN expression and EGFR phosphorylation on the first-generation TKI's efficacy in vitro. Results: Our data revealed that OPN staining positively linked to a more advanced clinical stage. Compared with the control group, LUAD cells with elevated OPN levels are more sensitive to the growth inhibitory effect of TKI. Knocking down of OPN decreased the response of cells to gefitinib. Besides, OPN also upregulated the phosphorylation of EGFR, thereby affecting the effect of TKI. Conclusion: OPN enhanced the sensitivity of LUAD cells to gefitinib by promoting EGFR phosphorylation. OPN may be a potential target for evaluating TKI efficacy and a potential target for molecular therapy. [ABSTRACT FROM AUTHOR]

Background: Targeted cancer therapy has shed light on the treatment of tumor, especially for patients with non-small cell lung cancer. However, only a limited portion of NSCLC patients carrying specific mutations showed an ideal drug response. In addition, DNA methylation status showed a great potential for cancer detection and prognosis prediction. Methods: Bisulfite sequencing was performed to analyze the DNA methylation of WIF1 promoter in cfDNA and tumor tissue samples collected from NSCLC patients. PFS and OS analyses were carried out to evaluate the prognosis of gefitinib treatment in patients with differential levels of WIF1 DNA methylation. Quantitative real-time PCR was used to analyze the expression of WIF1 mRNA, while immunohistochemistry was performed to assess the expression of WIF1 protein. Furthermore, ELISA was carried out to evaluate the WIF1 activity in plasma. Results: The DNA methylation level of WIF1 promoter was lower in the cfDNA of NSCLC patients with a complete or partial response to gefitinib, and NSCLC patients with hypomethylated WIF1 showed better PFS and OS. The DNA methylation of WIF1 promoter in the resected tumor tissues was consistent with WIF1 DNA methylation in cfDNA, indicating that cfDNA was mainly derived from lung cancer tissues. As a result, the expression of WIF1 in tissue samples and the WIF1 activity in plasma was inhibited in patients with hypermethylated WIF1. Moreover, the cell viability of gefitinib-resistant cells was decreased by the suppressed WIF1 methylation in vitro. And the expression level of WIF1 mRNA was higher in gefitinib-resistant cells overexpressing ALKBH5, a known suppressor of WIF1 methylation. Conclusion: In summary, the findings of this study demonstrated that the level of WIF1 methylation in cfDNA was associated with the insusceptibility of gefitinib in the treatment of lung cancer. [ABSTRACT FROM AUTHOR]

Purpose: Maintaining telomeres by recruiting telomerase-to-chromosome ends is essential for cancer cell survival. Inhibiting telomerase recruitment to telomeres represents a novel strategy for telomere-based lung cancer therapy. However, approaches for interrupting telomerase recruitment for cancer therapy still need to be explored. Methods: The telomere-binding protein TPP1 is responsible for recruiting telomerase to telomeres and synthesizing telomeres through the association between the oligosaccharide/oligonucleotide-binding (OB)-fold domain of TPP1 and telomerase reverse transcriptase. We overexpressed the TPP1 OB domain (TPP1-OB) by lentivirus infection in lung cancer cells. Telomere length was examined by Southern blot analysis of terminal restriction fragments. The effects of TPP1-OB on cell proliferation, the cell cycle, apoptosis, chemosensitivity, and tumor growth were evaluated in vitro and in vivo. Result: TPP1-OB inhibited the recruitment of telomerase to telomeres and shortened telomere length by acting as a dominant-negative mutant of TPP1. TPP1-OB resulted in reduced cell proliferation, G1 cell cycle arrest, and increased cell apoptosis in lung cancer cells. Cell apoptosis occurred mainly through the caspase-3-dependent signaling pathway. TPP1-OB also suppressed anchorage-independent growth and tumor growth in vivo. Moreover, we demonstrated that TPP1-OB enhances the sensitivity of lung cancer cells to the chemotherapeutic drug paclitaxel. Conclusion: Our results suggest that inhibiting TPP1-mediated telomerase recruitment by expressing the TPP1-OB domain is a potential novel strategy for telomere-targeted lung cancer therapy. [ABSTRACT FROM AUTHOR]

Purpose: Traditional classification of melanoma is widely utilized with little apparent results making the development of robust classifiers that can guide therapies an urgency. Successful seminal research on classification has provided a wider understanding of cancer from multiple molecular profiles, respectively. However, it may ignore the complementary nature of the information provided by different types of data, which motivated us to subtype melanoma by aggregating multiple genomic platform data.Methods: Aggregating three omics data of 328 melanoma samples, melanoma subtyping was performed by three clustering methods. Differences across subtypes were extracted by functional enrichment, epigenetically silencing, gene mutations and clinical features. Subtypes were further distinguished by putative biomarkers.Results: Functional enrichment of the subtype-specific differential expression genes endowed subtypes new designation: immune, melanin and ion, in which the first subtype was enriched for immune system, the second was characterized by melanin and pigmentation, and the third was enriched for ion-involved transmission process. Subtypes also differed in age, Breslow thickness, tumor site, mutation frequency of BRAF, PTGS2, CDKN2A, CDKN2B and incidence of epigenetically silencing for IL15RA, EPSTI1, LXN, CDKN1B genes.Conclusions: Skin cutaneous melanoma can be robustly divided into three subtypes by SNFCC+. Compared with the TCGA classification derived from gene expression, the subtypes we presented share concordance, but new traits are excavated. Such a genomic classification offers insights to further personalize therapeutic decision-making and melanoma management. [ABSTRACT FROM AUTHOR]

Purpose: Although desmocollins have an important position in cancer-related research, there are little reports about the relations between cancers and desmocollin 1 (DSC1). The present study was designed to investigate the correlations between DSC1 and head and neck squamous cell carcinoma (HNSCC). Methods: First we analyzed the GEO database; then, HNSCC and pericarcinous tissues were collected to verify the results. DSC1 expression was detected by western blot and real-time PCR. The co-expression genes of DSC1 were extracted from Cancer Cell Line Encyclopedia database (CCLE database), and their correlation was analyzed in The Cancer Genome Atlas HNSCC database (TCGA HNSCC database). Next the gene ontology analysis (GO) was carried out. Moreover, we suppressed DSC1 in FaDu cell to investigate the internal mechanism. Results: GEO database showed that DSC1 was higher in HNSCC and patients with higher DSC1 had unfavorable prognosis. The results of the samples showed that DSC1 was significantly higher in HNSCC than in normal tissue, which was consistent with the results of GEO database. The co-expression genes of DSC1 were extracted from CCLE database and verified in TCGA HNSCC database. It revealed that DSC1 was related to cell signal transduction. In FaDu/siDSC1 cells, the proliferation and migration were decreased compared to FaDu cells, and the expression levels of β-catenin, c-myc and cyclin D1 down-regulated significantly. Conclusions: The increased expression of DSC1 can promote the occurrence of HNSCC and is associated with tumor. The increased expression of DSC1 also indicates a poor prognosis of the patients with HNSCC. [ABSTRACT FROM AUTHOR]