Bioinformatic Analysis Of Portal Vein Tumor Thrombus With Different Clonal Origin In Hepatocellularc Arcinoma

BACKGROUND&AIMS:Hepatocellular carcinoma(HCC)is a malignant tumor with more than 400,000new cases and 340,000 deaths every year in China.Most HCC patients were diagnosed the advanced stage accompanied by portal vein tumor thrombosis(PVTT).PVTT leads to poor prognosis of HCC patients.HCC patients with PVTT have shorter disease free survival time and overall survival time than those without PVTT.Nowadays,there is no effective therapy for HCC with PVTT except surgical treatment due to the fact that the underlying mechanism of PVTT development is still unclear.It has been reported that the initiation of PVTT is strongly correlated with cancer stem cells(CSCs)and tumor microenvironment.Our earlier study has reported that the noncoding RNAs including ICR and miR-135a,could bind to their target genes to promote metastasis and self-renewal of CSCs in HCC.Additionally,14-3-3-?could inhibit the degradation of HIF-1?through hypoxic tumor microenvironment and promote the epithelial-mesenchymal transition(EMT),which suggests that PVTT may be induced by metastasis.Besides,we have reported that the PVTT were also found in patients with small HCC even without primary tumor,and the origin of PVTT in these patients is still unclear.Some studies showed that there are different clonal origins and tumor antigens in HCC,which may results in the failure of therapy.In the present study,we intend to:1.explore the original differentiation between PVTTs and corresponding primary tumors;2.explain the special gene expression signature of independent PVTTs;3.find a gene signature including 24 genes used to identify pairs of PVTTs and primary tumors without clonal relationship and provide some theoretical basis for the adjuvant treatment of HCC with PVTT.METHODS:The methods used in this study is as following:1.Analyze the copy number variations(CNV)of PVTT and compared primary tumor by CNV chip(Affymetrix Cytoscan HD).Bioinformatics analysis was used to analyze the CNA similarity and specific chromosome structure variations between PVTT and corresponding primary tumor from the same patient.Then analysis the copy numbers variations of high frequency mutation genes(driver genes)and the molecular type of PVTT and compared primary tumor to reveal whether PVTT originated from primary tumor.2.EBSeq was used to analysis RNA-seq data of PVTT for different expression gene between PVTT13 and its corresponding primary tumors.GO enrichment analysis and KEGG pathway analysis were used to reveal the unique molecular mechanism of PVTT origination based on expression profile data.3.Analyze RNA-seq data of PVTTs and compared primary tumor.ssGSEA was used to calculate the enrichment score of bile duct cells gene signature,mature liver cells gene signature,hepatic progenitor cells gene signature and hepatic stellate cells gene signature in tissues,which is used to explore the potential origin cell types of PVTT.Then ssGSEA analysis was also used to calculate the enrichment score of hematopoietic stem cells gene signature and liver cancer stem cells gene signature in PVTT and compared HCC tissues to explore the role of stem cells in the initiation of PVTT.4.Protein-protein interaction was used to analyze the core proteins of independent original PVTT based on expression profile data,and to search for a gene signature which could diagnose the clonal origin of PVTT.5.ssGSEA analysis were used to explore the application value of immunotherapy on HCC patients with PVTT by enrichment of immunocyte marker gene in portal vein tumor thrombus and liver cancer tissue.RESULTS:From all 19 patients,DNA copy number variation(CNV)showed that LR2 value of patient13 was only 0.0051(P>0.05)which of other patients were more than10~9(P<0.001).Further studies revealed more remarkable difference in chromosome structure variations between PVTT13 and T13.Some chromosome regions were deleted in PVTT13,which were duplicated in T13.However,there was no remarkable difference of chromosome structure variations in the others.Moreover,PVTT13 and T13 exhibited specific significantly mutation of driver genes,respectively.The copy number of ADCY2 was 5 in PVTT13 and only 2 in T13,however the copy number of STNNB1 was 3 in PVTT13 while 5 in T13.There was no these variations in the other patients.Molecular subtyping studies demonstrated that PVTT13 belonged to well-differentiated type called S3 type,however T13 belonged to moderately differentiated or poorly differentiated type called S1 type.These data suggested that PVTT13 did not originate from T13.Based on different expression genes between PVTT13 and T13,more than 2000different expression genes were observed in pair of PVTT13 and T13.GO and KEGG pathway analysis showed the abnormal expression genes in PVTT13 are mainly involved in leukocyte activation involved in immune response,NF-?B signaling pathway and so on.The abnormal expression genes of T13 mainly involves in DNA replication,cellular response to DNA damage stimulus,PPAR signal pathway and so on.SsGSEA of potential originated cell types revealed that there was a high enrichment score of the mature hepatocytes gene signature in PVTT13 while a high enrichment score of the hepatic stellate cells gene signature in T13.Besides,ssGSEA of stem cell gene signature revealed that the enrichment score of the hematopoietic stem cells gene signature in PVTT13 was 3114 which was 1031 in T13,and the enrichment score of EpCAM positive cancer stem cells gene signature in T13 was1401 which was-315 in PVTT13.Analysis of CNV chip and RNA-seq data showed there were 24 core genes in protein-protein interaction network.Cluster analysis of this 24 genes'expression in 19pair samples showed that PVTT13 and T13 belonged to different classes.These genes were defined as PVTT-OD-24GS(PVTT origin Discriminant gene signature).PVTT-OD-24GS was used to analysis GSE69164 and GSE74656 from GEO datasheet.One case in GSE69164 showed different origin of PVTT compared its primary tumor,and no case was found in GSE74656.RT-PCR data of gene expression in PVTT-OD-24GS of 8 pairs sample showed different origins between PVTT and compared primary tumor of the patient 2,which was corroborate to the CNV data of this patient.SsGSEA of immune cell gene signature showed significantly enrichment in 14cases which contained 4 PVTT cases and 10 primary tumor cases.NTP analysis of immune class showed that 8 cases including 2 PVTT cases belonged to immune class,1 primary tumor case belonged to rest class,while the others could not be identified as a class(FDR>0.05).CONCLUSIONS:PVTTs can be divided into two types according to clonal origin relationship between PVTTs and their corresponding primary tumors,one type is HCC dependent PVTT,which results from compared primary tumor of HCC and has the similar gene expression profiles with the origination tumor tissue;while the other type is HCC independent PVTT,which does not result from compared primary tumor and has different gene expression profile with the corresponding tumors.Patients with HCC independent PVTT is minority of all HCC patients.By integrated analyzing of CNV and gene expression profiles,a gene signature including 24 genes is used to identify pairs of PVTTs and primary tumors without clonal relationship based on the unique gene expression profiles.Validation in three datasets showed that these types of pairs of PVTTs and tumors can be identified by the 24-gene signature.Additionally,ssGESA of immune cell gene signature shows immune cells infiltration in some of PVTT.Immunotherapy might be a good adjuvant treatment of HCC with PVTT.