Cytokeratin 19-positive (CK19+) hepatocellular carcinoma (HCC) is an aggressive subtype characterized by early recurrence and chemotherapy tolerance. However, there is no specific therapeutic option for CK19+ HCC. The correlation between tumor recurrence and expression status of CK19 were studied in 206 patients undergoing liver transplantation for HCC. CK19−/+ HCC cells were isolated to screen effective antitumor drugs. The therapeutic effects of regorafenib were evaluated in patient-derived xenograft (PDX) models from 10 HCC patients. The mechanism of regorafenib on CK19+ HCC was investigated. CK19 positiveness indicated aggressiveness of tumor and higher recurrence risk of HCC after liver transplantation. The isolated CK19+ HCC cells had more aggressive behaviors than CK19− cells. Regorafenib preferentially increased the growth inhibition and apoptosis of CK19+ cells in vitro, whereas sorafenib, apatinib, and 5-fluorouracil did not. In PDX models from CK19−/+ HCC patients, the tumor control rate of regorafenib achieved 80% for CK19+ HCCs, whereas 0% for CK19− HCCs. RNA-sequencing revealed that CK19+ cells had elevated expression of mitochondrial ribosomal proteins, which are essential for mitochondrial function. Further experiments confirmed that regorafenib attenuated the mitochondrial respiratory capacity in CK19+ cells. However, the mitochondrial respiration in CK19− cells were faint and hardly repressed by regorafenib. The mitochondrial respiration was regulated by the phosphorylation of signal transducer and activator of transcription 3 (STAT3), which was inhibited by regorafenib in CK19+ cells. Hence, CK19 could be a potential marker of the therapeutic benefit of regorafenib, which facilitates the individualized therapy for HCC. STAT3/mitochondria axis determines the distinct response of CK19+ cells to regorafenib treatment.
Deep Neural Networks (DNNs) usually work in an end-to-end manner. This makes the trained DNNs easy to use, but they remain an ambiguous decision process for every test case. Unfortunately, the interpretability of decisions is crucial in some scenarios, such as medical or financial data mining and decision-making. In this paper, we propose a Tree-Network-Tree (TNT) learning framework for explainable decision-making, where the knowledge is alternately transferred between the tree model and DNNs. Specifically, the proposed TNT learning framework exerts the advantages of different models at different stages: (1) a novel James–Stein Decision Tree (JSDT) is proposed to generate better knowledge representations for DNNs, especially when the input data are in low-frequency or low-quality; (2) the DNNs output high-performing prediction result from the knowledge embedding inputs and behave as a teacher model for the following tree model; and (3) a novel distillable Gradient Boosted Decision Tree (dGBDT) is proposed to learn interpretable trees from the soft labels and make a comparable prediction as DNNs do. Extensive experiments on various machine learning tasks demonstrated the effectiveness of the proposed method.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.