Subclassification and Detection of New Markers for the Discrimination of Primary Liver Tumors by Gene Expression Analysis Using Oligonucleotide Arrays

Identification of BRIP1, NSMCE2, ANAPC7, RAD18 and TTL from chromosome segregation gene set associated with hepatocellular carcinoma

INTRODUCTION

Hepatocellular carcinoma is one of the most frequent cancers with high mortality rate worldwide.

METHODS

TCGA LIHC HTseq counts were analyzed. GSEA was performed with GO BP gene sets. GO analysis was performed with differentially expressed genes. The subset of genes contributing most of the enrichment result of GO_BP_CHROMOSOME_SEGREGATION of GSEA were identified. Five genes have been selected in this subset of genes for further analysis. A microarray data set, GSE112790, was analyzed as a validation data set. Survival analysis was performed.

RESULTS

According to GSEA and GO analysis several gene sets and processes related to chromosome segregation were enriched in LIHC. GO_BP_CHROMOSOME_SEGREGATION gene set from GSEA had the highest size of the genes contributing most of the enrichment. Five genes in this gene set; BRIP1, NSMCE2, ANAPC7, RAD18 and TTL, whose expressions and prognostic values have not been studied in hepatocellular carcinoma in detail, have been selected for further analyses. Expression of these five genes were identified as significantly upregulated in LIHC RNA-seq and HCC microarray data set. Survival analysis showed that high expression of the five genes was associated with poor overall survival in HCC patients.

CONCLUSION

Selected genes were upregulated and had prognostic value in HCC.

Panel of potential lncRNA biomarkers can distinguish various types of liver malignant and benign tumors

PURPOSE

Liver cancers are among the deadliest malignancies due to a limited efficacy of early diagnostics, the lack of appropriate biomarkers and insufficient discrimination of different types of tumors by classic and molecular methods. In this study, we searched for novel long non-coding RNA (lncRNA) as well as validated several known candidates suitable as probable biomarkers for primary liver tumors of various etiology.

METHODS

We described a novel lncRNA HELIS (aka "HEalthy LIver Specific") and estimated its expression by RT-qPCR in 82 paired tissue samples from patients with hepatocellular carcinoma (HCC), cholangiocarcinoma (CCA), combined HCC-CCA, pediatric hepatoblastoma (HBL) and non-malignant hepatocellular adenoma (HCA) and focal nodular hyperplasia (FNH). Additionally, we examined expression of cancer-associated lncRNAs HULC, MALAT1, UCA1, CYTOR, LINC01093 and H19, which were previously studied mainly in HCC.

RESULTS

We demonstrated that down-regulation of HELIS strongly correlates with carcinogenesis; whereas in tumors with non-hepatocyte origin (HBL, CCA) or in a number of poorly differentiated HCC, this lncRNA is not expressed. We showed that recently discovered LINC01093 is dramatically down-regulated in all malignant liver cancers; while in benign tumors LINC01093 expression is just twice decreased in comparison to adjacent samples.

CONCLUSION

Our study revealed that among all measured biomarkers only down-regulated HELIS and LINC01093, up-regulated CYTOR and dysregulated HULC are perspective for differential diagnostics of liver cancers; whereas others demonstrated discordant results and cannot be considered as potential universal biomarkers for this purpose.

The role of HSP90 molecular chaperones in hepatocellular carcinoma

Misfolded proteins have enhanced formation of toxic oligomers and nonfunctional protein copies lead to recruiting wild-type protein types. Heat shock protein 90 (HSP90) is a molecular chaperone generated by cells that are involved in many cellular functions through regulation of folding and/or localization of large multi-protein complexes as well as client proteins. HSP90 can regulate a number of different cellular processes including cell proliferation, motility, angiogenesis, signal transduction, and adaptation to stress. HSP90 makes the mutated oncoproteins able to avoid misfolding and degradation and permits the malignant transformation. As a result, HSP90 is an important factor in several signaling pathways associated with tumorigenicity, therapy resistance, and inhibiting apoptosis. Clinically, the upregulation of HSP90 expression in hepatocellular carcinoma (HCC) is linked with advanced stages and inappropriate survival in cases suffering from this kind of cancer. The present review comprehensively assesses HSP90 functions and its possible usefulness as a potential diagnostic biomarker and therapeutic option for HCC.

Impact of a deep learning assistant on the histopathologic classification of liver cancer

Artificial intelligence (AI) algorithms continue to rival human performance on a variety of clinical tasks, while their actual impact on human diagnosticians, when incorporated into clinical workflows, remains relatively unexplored. In this study, we developed a deep learning-based assistant to help pathologists differentiate between two subtypes of primary liver cancer, hepatocellular carcinoma and cholangiocarcinoma, on hematoxylin and eosin-stained whole-slide images (WSI), and evaluated its effect on the diagnostic performance of 11 pathologists with varying levels of expertise. Our model achieved accuracies of 0.885 on a validation set of 26 WSI, and 0.842 on an independent test set of 80 WSI. Although use of the assistant did not change the mean accuracy of the 11 pathologists (p = 0.184, OR = 1.281), it significantly improved the accuracy (p = 0.045, OR = 1.499) of a subset of nine pathologists who fell within well-defined experience levels (GI subspecialists, non-GI subspecialists, and trainees). In the assisted state, model accuracy significantly impacted the diagnostic decisions of all 11 pathologists. As expected, when the model's prediction was correct, assistance significantly improved accuracy (p = 0.000, OR = 4.289), whereas when the model's prediction was incorrect, assistance significantly decreased accuracy (p = 0.000, OR = 0.253), with both effects holding across all pathologist experience levels and case difficulty levels. Our results highlight the challenges of translating AI models into the clinical setting, and emphasize the importance of taking into account potential unintended negative consequences of model assistance when designing and testing medical AI-assistance tools.