| Background:Primary liver cancer(PLC)is a malignant tumor with high morbidity and mortality,including hepatocellular carcinoma(HCC),cholangiocarcinoma(ICC)and HCC-ICC,of which HCC is the most common type.The current treatments for PLC include resection,liver transplantation,ablation,transcatheter arterial chemoembolization(TACE),radiotherapy and chemotherapy,targeted drugs and immunotherapy.However,the survival rate and recurrence rate are still not ideal.Traditional Chinese medicine(TCM)has a long history in the treatment for PLC in China.Aim:Fufang Banmao capsules(FFBM)has the effect of attacking poison and erosion ulcer and can be used in treating cancer,such as PLC,lung cancer,colorectal cancer.However,the research on the mechanism of FFBM for treating for PLC is very scarce.Our aim is to predict the possible target of FFBM for treating PLC based on network pharmacology,validate the effectiveness with clinical data,and provide evidence for the treatment of PLC.Methods:Part one: Target prediction of FFBM for treating PLC baesd on network pharmacologyAll the components of FFBM were obtained from HERB database and CNKI,removing the duplications,and then used the Swiss ADME database to screen active ingredients.Finded the Canonical SMILES or the sdf file of the active ingredient by using the Pubchem database and predicted the targets by using the Swiss Target Prediction database,and finally geted the targets of FFBM after removing the duplications.Three studies related to PLC based on the GPL570 platform in the NCBI GEO database were selected,namely GSE45267,GSE62232 and GSE101685,and GEO2 R was used to identify the differential genes between liver cancer tissue and normal liver tissue.We regarded the genes as differentially expressed genes in PLC when the Log|Fold Change,FC|>1,and adjusted P value<0.01.Common target genes were obtained by intersection and performed GO,KEGG and PPI analysis.Obtained the hub genes through the cyto Hubba and MCODE plugins of the cytoscape software.To validate whether there was a difference in the expression of the hub genes in PLC and normal liver by using CCLE,HPA,and TCGA database.To verify the efficacy of the hub genes in the diagnosis and prognosis of PLC by using the TCGA database and to mine the transcription factors of the hub genes by using the TRRUST database.Finally,the molecular regulation mechanism of drug-active ingredients-target genes and transcription factors-target genes-signaling pathway were constructed.Part two: Validate the effectiveness of FFBM in the treatment of PLC after TACEPatients who were diagnosed with Hepatitis B Virus(HBV)related PLC and underwent TACE were selected retrospectively from January 2008 to January 2020 by using the data platform in our hospital.Patients were recruited based on the inclusion and exclusion criteria.The follow-up endpoint was 2 years after TACE or death.Patients were divided into FFBM group and control group according to whether the patients took FFBM or not.Age,gender,number and diameter of tumors,liver cirrhosis and ascites,number of TACE,HBV-DNA,antiviral therapy,alpha fetoprotein(AFP),albumin(ALB),total bilirubin(TBIL)and prothrombin time(PT)were compared between the two groups.Child-Pugh score and BCLC staging were calculated.The KM survival curve was drawn by univariate Cox regression analysis,and the value of FFBM in the treatment of PLC was obtained after eliminating the interference of confounding factors by multivariate Cox regression analysis.Results:Part one: Target prediction of FFBM for treating PLC baesd on network pharmacology1.FFBM contained 1760 ingredients and 764 active active ingredients.The three most important active ingredients were quercetin,kaempferol and isorhamnetin,and a total of 1397 targets were predicted.There were 2690 differential expressed genes of PLC based on the GEO database.Finally,293 common target genes were obtained,of which 116 were up-regulated genes and 177 were down-regulated genes.2.KEGG with up-regulated genes were enriched in signal pathways such as cell cycle,cell senescence,micro RNA in cancer,viral carcinogenesis,hepatitis B,PI3K-Akt pathway,p53 pathway,MAPK pathway,etc.3.The hub genes included CDK1,CCNA2,CCNB1,CCNB2,AURKA,TOP2 A,KIF11,PBK,KIF20,AURKB,TTK,NEK2 and MELK.The hub genes except NEK2 were over-expressed in PLC based on the CCLE,HPA,and TCGA database(NEK2was not detected in liver cancer tissues and normal liver tissues according to the HPA database).4.The area under the ROC curve of 13 Hub genes for the diagnosis of PLC was0.951,0.897,0.956,0.948,0.918,0.951,0.937,0.928,0.934,0.927,0.929,0.951 and0.938,respectively.KM survival analysis showed that patients with higher gene expressions had a poor prognosis compared with those patients with lower lower gene expressions,whether the Overall Survival(OS)or the Progress Free Interval(PFI).5.13 Hub genes are related to a variety of immune cells,especially Th2 cells.The correlation coefficients of each hub gene and Th2 cells were 0.749,0.743,0.745,0.747,0.633,0.726,0.717,0.748,0.710,0.706,0.731 and 0.716,respectively(P<0.001).6.It suggested that transcription factors of the hub gene are mainly E2F3,E2F1,E2F4 and TP53 based on the TRRUST database,and related to the cell cycle,cell senescence and P53 signaling pathway.Part two: Validate the effectiveness of FFBM in the treatment of PLC after TACE1.A total of 301 patients with PLC after TACE were enrolled,including 101 patients in the FFBM group and 200 patients in the control group.The age,gender,number and diameter of tumors,liver cirrhosis and ascites,number of TACE,antiviral therapy,AFP,TBIL,ALB,Child-Pugh score,and BCLC staging were no statistical difference in the two groups(P>0.05).While,HBV-DNA,PT and follow-up time were statistically different between the two groups(P<0.05).2.The KM survival curve suggested that the prognosis of patients in the control group was worse than those in the FFBM group [HR 1.44,95% confidence interval(CI): 1.02-2.03,P=0.04].Multivariate Cox regression analysis indicated that the independent risk factors affecting the 2-year survival prognosis of patients with PLC after TACE were age [(≥65 years),HR 1.6,P=0.04)],liver cirrhosis(HR 1.43,P=0.04),BCLC staging C(HR 2.72,P<0.001),number of TACE [(>4 times),HR6.19,P=0.002)] and high AFP [(≥400ng/ml),HR 1.79,P=0.001)],and taking FFBM was an independent protective factor(HR 0.69,P=0.04).Conclusion:1.The three most important active ingredients of FFBM in the treatment of PLC are quercetin,kaempferol and isorhamnetin.2.The targets of FFBM in the treatment of PLC may be CDK1,CCNA2,CCNB1,CCNB2,AURKA,TOP2 A,KIF11,PBK,KIF20,AURKB,TTK,NEK2 and MELK.Signal pathways may involve cell cycle,cell senescence,micro RNA in cancer,viral carcinogenesis,hepatitis B,PI3K-Akt pathawy and p53 pathway.E2F3,E2F1,E2F4 and TP53 may be involved in the regulation of transcription,and may be related to immune infiltration.3.It is effective for FFBM in treating patients with PLC after TACE,and can prolong the OS. |