Differentiation of benign and malignant parotid gland tumors based on the fusion of radiomics and deep learning features on ultrasound images.

Objective: The pathological classification and imaging manifestation of parotid gland tumors are complex, while accurate preoperative identification plays a crucial role in clinical management and prognosis assessment. This study aims to construct and compare the performance of clinical models, traditional radiomics models, deep learning (DL) models, and deep learning radiomics (DLR) models based on ultrasound (US) images in differentiating between benign parotid gland tumors (BPGTs) and malignant parotid gland tumors (MPGTs). Methods: Retrospective analysis was conducted on 526 patients with confirmed PGTs after surgery, who were randomly divided into a training set and a testing set in the ratio of 7:3. Traditional radiomics and three DL models (DenseNet121, VGG19, ResNet50) were employed to extract handcrafted radiomics (HCR) features and DL features followed by feature fusion. Seven machine learning classifiers including logistic regression (LR), support vector machine (SVM), RandomForest, ExtraTrees, XGBoost, LightGBM and multi-layer perceptron (MLP) were combined to construct predictive models. The most optimal model was integrated with clinical and US features to develop a nomogram. Receiver operating characteristic (ROC) curve was employed for assessing performance of various models while the clinical utility was assessed by decision curve analysis (DCA). Results: The DLR model based on ExtraTrees demonstrated superior performance with AUC values of 0.943 (95% CI: 0.918-0.969) and 0.916 (95% CI: 0.861-0.971) for the training and testing set, respectively. The combined model DLR nomogram (DLRN) further enhanced the performance, resulting in AUC values of 0.960 (95% CI: 0.940- 0.979) and 0.934 (95% CI: 0.876-0.991) for the training and testing sets, respectively. DCA analysis indicated that DLRN provided greater clinical benefits compared to other models. Conclusion: DLRN based on US images shows exceptional performance in distinguishing BPGTs and MPGTs, providing more reliable information for personalized diagnosis and treatment plans in clinical practice.

Emodin, a natural anthraquinone, suppresses liver cancer in vitro and in vivo by regulating VEGFR2 and miR-34a.

The pharmacokinetic (PK) and potential effects of Emodin on liver cancer were systematically evaluated in this study. Both the intragastric administration (i.g.) and hypodermic injection (i.h.) of Emodin exhibited a strong absorption (absorption rate < 1 h) and elimination capacity (t1/2 ≈ 2 h). The tissue distribution of Emodin after i.h. was rapid and wide. The stability of Emodin in three species of liver microsomes wasrat >human> beagle dog. These PK data provided the basis for the subsequent animal experiments. In liver cancer patient tissues, the expression of vascular endothelial growth factor (VEGF)-induced signaling pathways, including phosphorylated VEGF receptor 2 (VEGFR2), AKT, and ERK1/2,were simultaneously elevated, but miR-34a expression was reduced and negatively correlated with SMAD2 and SMAD4. Emodin inhibited the expression of SMAD2/4 in HepG2 cells by inducing the miR-34a level. Subsequently, BALB/c nude mice received a daily subcutaneous injection of HepG2 cells with or without Emodin treatment (1 mg/kg or 10 mg/kg), and Emodin inhibited tumorigenesis and reduced the mortality rate in a dose-dependent manner. In vivo experiments showed that cell proliferation, migration, and invasion were promoted by VEGF or miR-34a signal treatment but were inhibited when combined with Emodin treatment. All these results demonstrated that Emodin inhibited tumorigenesis in liver cancer by simultaneously inhibiting the VEGFR2-AKT-ERK1/2signaling pathway and promoting a miR-34a-mediated signaling pathway.

C/EBPα Regulates FOXC1 to Modulate Tumor Growth by Interacting with PPARγ in Hepatocellular Carcinoma.

BACKGROUND: Forkhead box C1 (FOXC1) is an important cancer-associated gene in tumor. PPAR-γ and C/EBPα are both transcriptional regulators involved in tumor development. OBJECTIVE: We aimed to clarify the function of PPAR-γ, C/EBPα in hepatocellular carcinoma (HCC) and the relationship of PPAR-γ, C/EBPα and FOXC1 in HCC. METHODS: Western blotting, immunofluorescent staining, and immunohistochemistry were used to evaluate protein expression. qRT-PCR was used to assess mRNA expression. Co-IP was performed to detect the protein interaction. And ChIP and fluorescent reporter detection were used to determine the binding between protein and FOXC1 promoter. RESULTS: C/EBPα could bind to FOXC1 promoter and PPAR-γ could strengthen C/EBPα's function. Expressions of C/EBPα and PPAR-γ were both negatively related to FOXC1 in human HCC tissue. Confocal displayed that C/EBPα was co-located with FOXC1 in HepG2 cells. C/EBPα could bind to FOXC1 promoter by ChIP. Luciferase activity detection exhibited that C/EBPα could inhibit FOXC1 promoter activity, especially FOXC1 promoter from -600 to -300 was the critical binding site. Only PPAR-γ could not influence luciferase activity but strengthen inhibited effect of C/EBPα. Further, the Co-IP displayed that PPAR-γ could bind to C/EBPα. When C/EBPα and PPAR-γ were both high expressed, cell proliferation, migration, invasion, and colony information were inhibited enormously. C/EBPα plasmid combined with or without PPAR-γ agonist MDG548 treatment exhibited a strong tumor inhibition and FOXC1 suppression in mice. CONCLUSION: Our data establish C/EBPα targeting FOXC1 as a potential determinant in the HCC, which supplies a new pathway to treat HCC. However, PPAR-γ has no effect on FOXC1 expression.