Sialic Acid-Modified O-GlcNAc Transferase Inhibitor Liposome Presents Antitumor Effect in Hepatocellular Carcinoma.

O-linked-N-acetylglucosaminylation (O-GlcNAcylation) plays a key role in hepatocellular carcinoma (HCC) development, and the inhibition of O-GlcNAcylation has therapeutic potential. To decrease the systemic adverse events and increase targeting, we used sialic acid (SA)-decorated liposomes loaded with OSMI-1, an inhibitor of the O-GlcNAcylation, to further improve the anti-HCC effect. Fifty pairs of HCC tissue samples and the cancer genome atlas database were used to analyze the expression of O-GlcNAc transferase (OGT) and its effects on prognosis and immune cell infiltration. OSMI-1 cells were treated with SA and liposomes. Western blotting, immunofluorescence, cell proliferation assay, flow cytometry, enzyme-linked immunosorbent assay, immunohistochemistry, and tumorigenicity assays were used to investigate the antitumor effect of SA-modified OSMI-1 liposomes in vitro and in vivo. OGT was highly expressed in HCC tissues, negatively correlated with the degree of tumor infiltration of CD8+ and CD4+T cells and prognosis, and positively correlated with the degree of Treg cell infiltration. SA-modified OSMI-1 liposome (OSMI-1-SAL) was synthesized with stable hydrodynamic size distribution. Both in vitro and in vivo, OSMI-1-SAL exhibited satisfactory biosafety and rapid uptake by HCC cells. Compared to free OSMI-1, OSMI-1-SAL had a stronger capacity for suppressing the proliferation and promoting the apoptosis of HCC cells. Moreover, OSMI-1-SAL effectively inhibited tumor initiation and development in mice. OSMI-1-SAL also promoted the release of damage-associated molecular patterns, including anticalreticulin, high-mobility-group protein B1, and adenosine triphosphate, from HCC cells and further promoted the activation and proliferation of the CD8+ and CD4+T cells. In conclusion, the OSMI-1-SAL synthesized in this study can target HCC cells, inhibit tumor proliferation, induce tumor immunogenic cell death, enhance tumor immunogenicity, and promote antitumor immune responses, which has the potential for clinical application in the future.

O-GlcNAcylation promotes tumor immune evasion by inhibiting PD-L1 lysosomal degradation.

Programmed-death ligand 1 (PD-L1) and its receptor programmed cell death 1 (PD-1) mediate T cell-dependent immunity against tumors. The abundance of cell surface PD-L1 is a key determinant of the efficacy of immune checkpoint blockade therapy targeting PD-L1. However, the regulation of cell surface PD-L1 is still poorly understood. Here, we show that lysosomal degradation of PD-L1 is regulated by O-linked N-acetylglucosamine (O-GlcNAc) during the intracellular trafficking pathway. O-GlcNAc modifies the hepatocyte growth factor-regulated tyrosine kinase substrate (HGS), a key component of the endosomal sorting machinery, and subsequently inhibits its interaction with intracellular PD-L1, leading to impaired lysosomal degradation of PD-L1. O-GlcNAc inhibition activates T cell-mediated antitumor immunity in vitro and in immune-competent mice in a manner dependent on HGS glycosylation. Combination of O-GlcNAc inhibition with PD-L1 antibody synergistically promotes antitumor immune response. We also designed a competitive peptide inhibitor of HGS glycosylation that decreases PD-L1 expression and enhances T cell-mediated immunity against tumor cells. Collectively, our study reveals a link between O-GlcNAc and tumor immune evasion, and suggests strategies for improving PD-L1-mediated immune checkpoint blockade therapy.

Deep Learning Radiomics to Predict Regional Lymph Node Staging for Hilar Cholangiocarcinoma.

BACKGROUND: Our aim was to establish a deep learning radiomics method to preoperatively evaluate regional lymph node (LN) staging for hilar cholangiocarcinoma (HC) patients. METHODS AND MATERIALS: Of the 179 enrolled HC patients, 90 were pathologically diagnosed with lymph node metastasis. Quantitative radiomic features and deep learning features were extracted. An LN metastasis status classifier was developed through integrating support vector machine, high-performance deep learning radiomics signature, and three clinical characteristics. An LN metastasis stratification classifier (N1 vs. N2) was also proposed with subgroup analysis. RESULTS: The average areas under the receiver operating characteristic curve (AUCs) of the LN metastasis status classifier reached 0.866 in the training cohort and 0.870 in the external test cohorts. Meanwhile, the LN metastasis stratification classifier performed well in predicting the risk of LN metastasis, with an average AUC of 0.946. CONCLUSIONS: Two classifiers derived from computed tomography images performed well in predicting LN staging in HC and will be reliable evaluation tools to improve decision-making.

MRI Features of Hepatic Sarcomatoid Carcinoma Different From Hepatocellular Carcinoma and Intrahepatic Cholangiocarcinoma.

INTRODUCTION: Hepatic sarcomatoid carcinoma (HSC) is a rare type of liver cancer with a high malignant grade and poor prognosis. This study compared the clinical characteristics and magnetic resonance imaging (MRI) features of HSCs with those of hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC), aiming to identify valuable features for HSC diagnosis. METHODS: In total, 17 pathologically confirmed HSC cases, 50 HCC cases and 50 common ICC cases were enrolled from two hospitals. The clinical characteristics and MRI features of all cases were summarized and statistically analyzed. RESULTS: On the one hand, the incidence rates of elevated carbohydrate antigen (CA) 19-9 and elevated carcinoembryonic antigen (CEA) were significantly higher in the HSC cases than in the HCC cases (29.4% vs. 0%; 17.6% vs. 0%). The HSC enhancement patterns, primarily including progressive enhancement, were also significantly different from HCC cases. The incidence rates of heterogeneous signals on T2-weighted imaging and during the arterial phase were significantly higher in the HSC cases than in the HCC cases (94.1% vs. 66.0%; 100.0% vs. 72.0%). The diameter of HSCs was significantly larger than that in the HCC cases (6.12 cm vs. 4.21 cm), and the incidence rates of adjacent cholangiectasis, intrahepatic metastasis and lymph node enlargement were considerably higher in the HSC cases than in the HCC cases (52.9% vs. 6.0%; 47.1% vs. 12.0%; 41.2% vs. 2.0%). On the other hand, the incidence rate of elevated CA199 was significantly lower in the HSC cases than in the ICC cases (29.4% vs. 60.0%). The incidence rates of intratumoral necrosis and pseudocapsules were significantly higher in the HSC cases than in the HCC cases (35.3% vs. 8.0%; 47.1% vs. 12.0%). However, the incidence rates of target signs were significantly lower in the HSC cases than in the HCC cases (11.8% vs. 42.0%). In addition, there was no significant difference in the enhancement patterns between HSC cases and ICC cases. CONCLUSIONS: HSCs were frequently seen in elderly men with clinical symptoms and elevated CA199 levels. The MRI features, including large size, obvious heterogeneity, hemorrhage, progressive enhancement, pseudocapsule and lymph node enlargement, contributed to the diagnosis of HSC.

DeepCOVID-XR: An Artificial Intelligence Algorithm to Detect COVID-19 on Chest Radiographs Trained and Tested on a Large U.S. Clinical Data Set.

Background There are characteristic findings of coronavirus disease 2019 (COVID-19) on chest images. An artificial intelligence (AI) algorithm to detect COVID-19 on chest radiographs might be useful for triage or infection control within a hospital setting, but prior reports have been limited by small data sets, poor data quality, or both. Purpose To present DeepCOVID-XR, a deep learning AI algorithm to detect COVID-19 on chest radiographs, that was trained and tested on a large clinical data set. Materials and Methods DeepCOVID-XR is an ensemble of convolutional neural networks developed to detect COVID-19 on frontal chest radiographs, with reverse-transcription polymerase chain reaction test results as the reference standard. The algorithm was trained and validated on 14 788 images (4253 positive for COVID-19) from sites across the Northwestern Memorial Health Care System from February 2020 to April 2020 and was then tested on 2214 images (1192 positive for COVID-19) from a single hold-out institution. Performance of the algorithm was compared with interpretations from five experienced thoracic radiologists on 300 random test images using the McNemar test for sensitivity and specificity and the DeLong test for the area under the receiver operating characteristic curve (AUC). Results A total of 5853 patients (mean age, 58 years ± 19 [standard deviation]; 3101 women) were evaluated across data sets. For the entire test set, accuracy of DeepCOVID-XR was 83%, with an AUC of 0.90. For 300 random test images (134 positive for COVID-19), accuracy of DeepCOVID-XR was 82%, compared with that of individual radiologists (range, 76%-81%) and the consensus of all five radiologists (81%). DeepCOVID-XR had a significantly higher sensitivity (71%) than one radiologist (60%, P < .001) and significantly higher specificity (92%) than two radiologists (75%, P < .001; 84%, P = .009). AUC of DeepCOVID-XR was 0.88 compared with the consensus AUC of 0.85 (P = .13 for comparison). With consensus interpretation as the reference standard, the AUC of DeepCOVID-XR was 0.95 (95% CI: 0.92, 0.98). Conclusion DeepCOVID-XR, an artificial intelligence algorithm, detected coronavirus disease 2019 on chest radiographs with a performance similar to that of experienced thoracic radiologists in consensus. © RSNA, 2020 Supplemental material is available for this article. See also the editorial by van Ginneken in this issue.