Highlights d AI system that can diagnose COVID-19 pneumonia using CT scans d Prediction of progression to critical illness d Potential to improve performance of junior radiologists to the senior level d Can assist evaluation of drug treatment effects with CT quantification

IMPORTANCE Sorafenib is the first-line treatment for hepatocellular carcinoma with portal vein invasion; however, it has shown unsatisfactory survival benefit. Sorafenib plus hepatic arterial infusion chemotherapy (HAIC) of oxaliplatin, fluorouracil, and leucovorin (FOLFOX) has shown promising results for these patients in a previous phase 2 study.OBJECTIVE To investigate the efficacy and safety of sorafenib plus HAIC compared with sorafenib for hepatocellular carcinoma with portal vein invasion. DESIGN, SETTING, AND PARTICIPANTSThis randomized, open-label clinical trial enrolled 818 screened patients. Of the 818 participants, 247 with hepatocellular carcinoma and portal vein invasion were randomly assigned (1:1) via a computer-generated sequence to receive sorafenib plus HAIC or sorafenib. This trial was conducted at 5 hospitals in China and enrolled patients from April 1, 2016, to October 10, 2017, with a follow-up period of 10 months.INTERVENTIONS Randomization to receive 400 mg sorafenib twice daily (sorafenib group) or 400 mg sorafenib twice daily plus HAIC (SoraHAIC group) (oxaliplatin 85 mg/m 2 , leucovorin 400 mg/m 2 , fluorouracil bolus 400 mg/m 2 on day 1, and fluorouracil infusion 2400 mg/m 2 for 46 hours, every 3 weeks). MAIN OUTCOMES AND MEASURESThe primary endpoint was overall survival by intention-to-treat analysis. Safety was assessed in patients who received at least 1 dose of study treatment.RESULTS For 247 patients (median age, 49 years; range, 18-75 years; 223 men and 24 women), median overall survival was 13.37 months (95% CI, 10.27-16.46) in the SoraHAIC group vs 7.13 months (95% CI, 6.28-7.98) in the sorafenib group (hazard ratio [HR], 0.35; 95% CI, 0.26-0.48; P < .001). The SoraHAIC group showed a higher response rate than the sorafenib group (51 [40.8%] vs 3 [2.46%]; P < .001), and a longer median progression-free survival (7.03 [95% CI, 6.05-8.02] vs 2.6 [95% CI, 2.15-3.05] months; P < .001). Grade 3/4 adverse events that were more frequent in the SoraHAIC group than in the sorafenib group included neutropenia (12 [9.68%] vs 3 [2.48%]), thrombocytopenia (16 [12.9%] vs 6 [4.96%]), and vomiting (8 [6.45%] vs 1 [0.83%]).CONCLUSIONS AND RELEVANCE Sorafenib plus HAIC of FOLFOX improved overall survival and had acceptable toxic effects compared with sorafenib in patients with hepatocellular carcinoma and portal vein invasion.

Our study provided a novel methodology to evaluate PD-L1 expression in the tumour microenvironment, which might help to select patients who would benefit from anti-PD-1/PD-L1 immunotherapies.

BackgroundAccumulating studies suggest that targeting epigenetic modifications could improve the efficacy of tumor immunotherapy; however, the mechanisms underlying this phenomenon remain largely unknown. Here, we investigated the ability of the epigenetic modifier, enhancer of zeste 2 polycomb repressive complex 2 subunit (EZH2), to regulate the expression of immune checkpoint inhibitor, programmed death-1 ligand 1 (PD-L1) in hepatocellular carcinoma (HCC).MethodsImmunohistochemistry and multiplex immunofluorescence staining were performed to analyze the expression and correlation of EZH2 and PD-L1 in HCC tissues. Immunoblotting, quantitative real-time PCR, flow cytometry, chromatin immunoprecipitation, and dual-luciferase reporter gene assays were performed to evaluate the regulatory roles of EZH2 on PD-L1 expression.ResultsIn vitro cell experiments revealed that EZH2 negatively regulated the PD-L1 expression of hepatoma cell lines in IFNγ-dependent manner. Mechanistic studies demonstrated that EZH2 could suppress PD-L1 expression by upregulating the H3K27me3 levels on the promoters of CD274 (encoding PD-L1) and interferon regulatory factor 1 (IRF1), an essential transcription factor for PD-L1 expression, without affecting the activation of the IFNγ-signal transducer and activator of transcription 1 (STAT1) pathway. Clinical samples from HCC patients with immune-activated microenvironments showed negative correlations between EZH2 and PD-L1 expression in hepatoma cells. Multivariate Cox analysis demonstrated that the combination of EZH2 and PD-L1 was an independent prognostic factor for both OS and RFS for patients with HCC.ConclusionsThe epigenetic modificator EZH2 can suppress the expression of immune checkpoint inhibitor PD-L1 by directly upregulating the promoter H3K27me3 levels of CD274 and IRF1 in hepatoma cells, and might serve as a potential therapeutic target for combination of immunotherapy for immune-activated HCC.

Sorafenib is the most recommended first‐line systemic therapy for advanced hepatocellular carcinoma (HCC). Yet there is no clinically applied biomarker for predicting sorafenib response. We have demonstrated that a vascular pattern, named VETC (Vessels that Encapsulate Tumor Clusters), facilitates the release of whole tumor clusters into the bloodstream; VETC‐mediated metastasis relies on vascular pattern, but not on migration and invasion of cancer cells. In this study, we aimed to explore whether vascular pattern could predict sorafenib benefit. Two cohorts of patients were recruited from four academic hospitals. The survival benefit of sorafenib treatment for patients with or without the VETC pattern (VETC+/VETC–) was investigated. Kaplan‐Meier analyses revealed that sorafenib treatment significantly reduced death risk and prolonged overall survival (OS; in cohort 1/2, P = 0.004/0.005; hazard ratio [HR] = 0.567/0.408) and postrecurrence survival (PRS; in cohort 1/2, P = 0.001/0.002; HR = 0.506/0.384) in VETC+ patients. However, sorafenib therapy was not beneficial for VETC‐ patients (OS in cohort 1/2, P = 0.204/0.549; HR = 0.761/1.221; PRS in cohort 1/2, P = 0.121/0.644; HR = 0.728/1.161). Univariate and multivariate analyses confirmed that sorafenib treatment significantly improved OS/PRS in VETC+, but not VETC–, patients. Further mechanistic investigations showed that VETC+ and VETC– HCCs displayed similar levels of light chain 3 (LC3) and phosphorylated extracellular signal‐regulated kinase (ERK) in tumor tissues (pERK) or endothelial cells (EC‐pERK), and greater sorafenib benefit was consistently observed in VETC+ HCC patients than VETC– irrespective of levels of pERK/EC‐pERK/LC3, suggesting that the different sorafenib benefit between VETC+ and VETC– HCCs may not result from activation of Raf/mitogen‐activated protein kinase kinase (MEK)/ERK and vascular endothelial growth factor (VEGF)A/VEGF receptor 2 (VEGFR2)/ERK signaling or induction of autophagy. Conclusion: Sorafenib is effective in prolonging the survival of VETC+, but not VETC–, patients. VETC pattern may act as a predictor of sorafenib benefit for HCC.

Common lung diseases are first diagnosed via chest X-rays. Here, we show that a fully automated deep-learning pipeline for chest-X-ray-image standardization, lesion visualization and disease diagnosis can identify viral pneumonia caused by Coronavirus disease 2019 (COVID-19), assess its severity, and discriminate it from other types of pneumonia. The deep-learning system was developed by using a heterogeneous multicentre dataset of 145,202 images, and tested retrospectively and prospectively with thousands of additional images across four patient cohorts and multiple countries. The system generalized across settings, discriminating between viral pneumonia, other types of pneumonia and absence of disease with areas under the receiver operating characteristic curve (AUCs) of 0.88–0.99, between severe and non-severe COVID-19 with an AUC of 0.87, and between severe or non-severe COVID-19 pneumonia and other viral and non-viral pneumonia with AUCs of 0.82–0.98. In an independent set of 440 chest X-rays, the system performed comparably to senior radiologists, and improved the performance of junior radiologists. Automated deep-learning systems for the assessment of pneumonia could facilitate early intervention and provide clinical-decision support.

Summary Cullin-RING ubiquitin ligases (CRL) are critical in ubiquitinating Myc, while COP9 signalosome (CSN) controls neddylation of Cullin in CRL. The mechanistic link between Cullin neddylation and Myc ubiquitination/degradation is unclear. Here we show that Myc is a target of the CSN subunit 6 (CSN6)–Cullin signaling axis and that CSN6 is a positive regulator of Myc. CSN6 enhanced neddylation of Cullin-1 and facilitated auto-ubiquitination/degradation of Fbxw7, a component of CRL involved in Myc ubiquitination, thereby stabilizing Myc. Csn6 haplo-insufficiency decreased Cullin-1 neddylation but increased Fbxw7 stability to compromise Myc stability and activity in an Eµ-Myc mouse model, resulting in decelerated lymphomagenesis. We found that CSN6 overexpression, which leads to aberrant expression of Myc target genes, is frequent in human cancers. Together, these results define a mechanism for the regulation of Myc stability through the CSN-Cullin-Fbxw7 axis and provide insights into the correlation of CSN6 overexpression with Myc stabilization/activation during tumorigenesis.

Copper and zinc are essential micronutrients, whose imbalance may be involved in the development and progression of cancer. However, the role of copper and/or zinc imbalance in the prognosis of hepatocellular carcinoma (HCC) is currently unclear. Our objective was to investigate the association between serum levels of copper, zinc and their ratio (copper/zinc) at diagnosis with HCC survival. We included 989 patients with incident HCC in this prospective cohort study, who were enrolled in the Guangdong Liver Cancer Cohort (GLCC) study within 30 days of diagnosis between September 2013 and February 2017. Serum copper and zinc were measured using inductively coupled plasma mass spectrometry. Primary outcomes were liver cancer‐specific survival (LCSS) and overall survival (OS). Cox proportional hazards models were used to calculate the multivariable hazard ratios (HRs) and 95% confidence interval (CI). Higher serum copper levels were strongly associated with worse LCSS (Q4 vs. Q1: HR = 1.87, 95% CI: 1.22–2.86; p < 0.01 for trend) and OS (Q4 vs. Q1: HR = 2.06, 95% CI: 1.36–3.11; p < 0.01 for trend). The calculated copper/zinc ratio was positively associated with LCSS (Q4 vs. Q1: HR = 1.31, 95% CI: 0.89–1.92; P = 0.04 for trend) and OS (Q4 vs. Q1: HR = 1.43, 95% CI: 0.99–2.08; P = 0.01 for trend). No overall associations were observed between serum zinc levels and LCSS or OS in the entire cohort. The results suggest that higher serum copper and copper in relation to zinc levels (i.e., higher copper/zinc ratio) may be associated with worse HCC survival, but serum zinc levels may be not associated with HCC survival.