Current Chinese Science - Volume 3, Issue 6, 2023

Significant advances in liver imaging technology have facilitated the early detection of subcentimeter hepatocellular carcinoma (HCC). Contrast-enhanced ultrasound, computed tomography, and magnetic resonance imaging (MRI) can be used to diagnose subcentimeter HCC based on the typical imaging features of HCC. Ancillary imaging features such as T2 weightedimaging mild-moderate hyperintensity, restricted diffusion, and hepatobiliary phase hypointensity may improve the diagnostic accuracy of gadoxetic acid-enhanced MRI for subcentimeter HCC. More information is needed to choose between immediate treatment or watchful waiting in subcentimeter HCC. Surgical resection, ablation, and transarterial chemoembolization are effective and safe methods for the management of subcentimeter HCC.

Liquid biopsies in cancer have received considerable attention as new biomarkers and subjects of clinical translational research. Liquid biopsy-derived biomarkers, including circulating tumor cells, cell-free nucleic acids, exosomes, tumor-educated platelets, proteins, and metabolites, have shown enormous potential to improve the sensitivity and specificity in early diagnosis of multiple cancer types, including hepatocellular carcinoma (HCC). Several analytes are already used in HCC clinical trials and are currently under investigation with encouraging preliminary results. This article will review the latest studies on liquid biopsy for early diagnosis and screening of HCC and discuss the current advantages and challenges of liquid biopsies in tumor biomarkers and clinical applications. In the future, the validation of large prospective clinical studies, the development of standardized operating procedures, and the integration of other diagnostic tools will open new avenues for applying liquid biopsy in early cancer detection.

Macrovascular tumor thrombosis (MVTT) is common among patients with advanced hepatocellular carcinoma (HCC), and it contains two main types, portal vein tumor thrombus (PVTT) and hepatic vein tumor thrombus (HVTT). MVTT is a significant prognostic factor, and the prognosis of HCC with MVTT remains unsatisfactory. The classification of MVTT varies in different studies, and the management of HCC and MVTT differs among Eastern and Western countries. Over the past years, a multidisciplinary approach, including surgical resection, radiotherapy, and transcatheter arterial chemoembolization, especially the advent of new targeted drugs and immune checkpoint inhibitors, has brought considerable survival benefits for HCC patients with MVTT. Recently, multiple novel treatment strategies, such as neoadjuvant therapy or tumor downstaging based on radiotherapy, immunotherapy, and molecular-targeted therapy, have broadened the treatment landscape of MVTT. In this review, we conclude the consensus and progress regarding diagnosis and treatment for patients with HCC and MVTT.

In the past 20 years, with the emergence and update of digital intelligent technology, the diagnosis and treatment of hepatocellular carcinoma (HCC) have undergone profound changes. Three-dimensional visualization technology has revolutionized the traditional two-dimensional diagnosis and treatment model of HCC and realized preoperative visualization of tumors and complex liver anatomy. The emergence of ICG fluorescence imaging has realized intraoperative tumor boundary visualization from the molecular and cellular levels. Augmented reality (AR) and mixed reality (MR) technology can realize the three-dimensional visualization of anatomical structures in surgical navigation. Traditional experiential surgery has been transformed into modern intelligent navigation surgery, and surgery has stepped into a new era of digital intelligent technology. In addition, artificial intelligence, molecular imaging and nanoprobes are also expected to achieve early diagnosis and treatment of HCC and improve the prognosis of patients. This article reviews the latest application of digital intelligent diagnosis and treatment technology related to diagnosing and treating HCC, hoping to help achieve accurate diagnosis and treatment of HCC.

Hepatocellular carcinoma (HCC) is highly prevalent, which contributes to the high disease burden worldwide. It was recognized as the sixth most common cancer and the fourth leading cause of mortality in cancer patients in the Global Cancer Statistics 2018. The management of HCC between eastern and western countries varies from region to region due to the different epidemiology of HCC and the approach of clinicians. Over the past decades, the multidisciplinary team approach has taken a leading role in the therapeutic paradigm of HCC, and the main treatment options include, but are not limited to 1) surgical resection; 2) transarterial chemoembolization (TACE); and 3) systemic therapies. Among these, surgical resection and TACE are applicable in earlyintermediate stage HCC with curative intent, while systemic therapies are mostly applied to advanced unresectable HCC. An individualized approach should be encouraged and tailored for each patient, taking into account the different patient population, staging, and treatment responses to achieve the best efficacy. This review discusses current advances and compares the eastern and western treatment options in the management of HCC.

Background: The task of relation extraction is a crucial component in the construction of a knowledge graph. However, it often necessitates a significant amount of manual annotation, which can be time-consuming and expensive. Distant supervision, as a technique, seeks to mitigate this challenge by generating a large volume of pseudo-training data at a minimal cost, achieved by mapping triple facts onto the raw text.Objective: The aim of this study is to explore the novelty and potential of the distant supervisionbased relation extraction approach. By leveraging this innovative method, we aim to enhance knowledge reliability and facilitate new knowledge discovery, establishing associations between knowledge from specific biomedical data or existing knowledge graphs and literature.Methods: This study presents a methodology to construct a biomedical knowledge graph employing distant supervision techniques. Through establishing links between knowledge entities and relevant literature sources, we methodically extract and integrate information, thereby expanding and enriching the knowledge graph. This study identified five types of biomedical entities (e.g., diseases, symptoms and genes) and four kinds of relationships. These were linked to PubMed literature and divided into training and testing datasets. To mitigate data noise, the training set underwent preprocessing, while the testing set was manually curated.Results: In our research, we successfully associated 230,698 triples from the existing knowledge graph with relevant literature. Furthermore, we identified additional 205,148 new triples directly sourced from these studies.Conclusion: Our study markedly advances the field of biomedical knowledge graph enrichment, particularly in the context of Traditional Chinese Medicine (TCM). By validating a substantial number of triples through literature associations and uncovering over 200,000 new triples, we have made a significant stride in promoting the development of evidence-based medicine in TCM. The results underscore the potential of using a distant supervision-based relation extraction approach to both validate and expand knowledge bases, contributing to the broader progression of evidence-based practices in the realm of TCM.