Integrative study reveals the prognostic and immunotherapeutic value of CTCs/CTMs -related prognostic signature in Hepatocellular carcinoma

doi:10.21203/rs.3.rs-2222900/v1

Download PDF

Research Article

Integrative study reveals the prognostic and immunotherapeutic value of CTCs/CTMs -related prognostic signature in Hepatocellular carcinoma

https://doi.org/10.21203/rs.3.rs-2222900/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Hepatocellular carcinoma (HCC) is the second major cause of tumor death worldwide, threatening people's lives and health. Circulating tumor cells (CTCs) facilitate the progression of many cancers, including HCC.

Methods

Combining transcriptomic data from TCGA, ICGC and ctcRbase databases, we characterized differentially expressed circulating tumor cells/microemboli (CTCs/CTMs) related genes (CRGs). These differentially expressed CRGs were then analyzed with a lasso and multivariate COX, and a five-gene prognostic signature was constructed. Then, the prognostic signature was validated in the ICGC dataset and GES14520 dataset with survival analysis and receiver operating characteristic analysis. The immune infiltration, tumor mutation burden (TMB), tumor immune dysfunction and exclusion (TIDE), and somatic mutation rate were also compared between the high- and low-risk groups to further evaluate the immunotherapeutic value of the model. Molecular subtypes of HCC were characterized by the non-negative matrix factor decomposition method (NMF), and potential compounds were evaluated for different subtypes. Nomograms were constructed to forecast the prognosis of patients. Comparison with other six models to assess the superiority of the CRGS signaturte.

Result

We constructed the prognostic characteristics of five CRGs (including CDCA8, TP53I3, MYCN, HAVCR1, and TXNRD1) and identified two HCC subtypes, C1 and C2. Patients in the low-risk group had a better survival time compared to the high-risk group. And patients in the high-risk group were associated with higher TMB, immune infiltration, somatic mutation rates, and lower TIDE scores. This prognostic profile was validated in the ICGC and GSE14520 datasets and had good predictive performance.

Conclusion

We have studied the perspective of CTC-related genes and developed a new prognostic signature to predict the prognosis of HCC patients. This signature with good performance will help to assess the prognosis of patients and may provide new insights for clinical management strategies.

Bioinformatics

circulating tumor cells

hepatocellular carcinoma

prognostic signature

immune microenvironment

immune infiltration

Primary liver cancer is one of the six most widespread malignancies in the world and has the third highest mortality rate in the world, with approximately 90% of cases being hepatocellular carcinoma [1-3]. More than 90% of HCC is closely related to chronic liver disease and causes of cirrhosis are important risk factors for HCC. Alcohol consumption, diabetes, obesity-induced NASH, HBV and HCV are all major risk factors for HCC, in addition to biliary cirrhosis and hemochromatosis.[4, 5]. Currently, the main treatment options for patients with liver cancer remain surgical liver resection and liver transplantation. However, for patients with advanced, recurrent HCC or those who have lost the chance of surgery, the prognosis is still not promising. Despite some advances in the detection, diagnosis, and treatment of HCC in recent years, its prognosis remains poor due to the high rates of recurrence, vascular invasion, or metastasis [6]. Therefore, it is urgent to explore both effective and representative biomarkers and new predictive tools.

Circulating tumor cells (CTCs) are those cells that depart from original or metastatic tumors and shed them into the bloodstream, while rare in healthy individuals [7-9]. As an important member of liquid biopsy technology, CTCs play an essential role in cancer diagnosis and treatment, carrying heterogeneous information about the primary tumor and serving as an effective biomarker and modeling tool. Increasingly, researchers are finding that CTCs have a significant role in the metastatic process of tumors. Therefore, the isolation and identification of CTCs with non-invasive biopsy can be widely applied for early diagnosis, real-time efficacy monitoring, and prognosis assessment of tumors [10-14]. In recent years, an increasingly popular view shows that the higher level of CTCs, the worse prognosis of tumor patients. In the CTC-positive cohort with HCC, survival time was significantly shorter and was also correlated with poor clinical features [15, 16]. Similarly, sun et al. found that the risk of tumor recurrence was increasing when HCC patients had a preoperative CTC (7.5ml) of ≥ 2, especially at AFP levels ≤ 400 ng / mL [17]. With advances in technology, genomic, transcriptomic, and proteomic analysis of CTCs at the single-cell level, and refinement of CTC in vitro models, we can better understand the critical role of CTCs in cancer [18, 19]. Nevertheless, our understanding of the biological functions of CTC in tumors at the molecular level is still not sufficient. Therefore, this research aimed to identify the CTCs/CTMs related genes (CRGs) and explore their clinical significance in HCC.

In this study, we comprehensively analyzed the transcriptomic data and clinical information of HCC in TCGA and ICGC databases. By combining the mRNA data associated with HCC in ctcRbase, we identified 258 CTCs/CTMs related genes (CRGs). Subsequently, we constructed a prognostic model and risk subgroups for HCC patients based on the five CRGs and synthesized the correlation between different subgroups of patients and immune labels such as immune infiltration, immune checkpoints, and tumor mutation burden. Finally, we further evaluated the detection efficacy and clinical value of the signature and screened some chemotherapeutic agents with potential therapeutic significance.

Data Sources

The training gene expression of transcriptome profiles and corresponding clinical trait data of LIHC patients were downloaded from The Cancer Genome Atlas database (TCGA, https://portal.gdc.cancer.gov/); while the validation dataset of mRNA expression and clinical trait data (the Liver Cancer LIRI-JP) were downloaded from the International Cancer Genome Consortium database (ICGC, https://icgc.org/). There are 12518 CTCs/CTMs genes were downloaded from the ctcRbase(http://www.origin-gene.cn/database/ctcRbase/) [7].

Identification of Candidate Genes

To obtain the differentially expressed genes (DEGs) associated with CTCs/CTMs, we utilized the “limma” R package for processing the RNA-seq data and the cutoff point was set as false discovery rate (FDR) <0.05 and |log2 (fold change) | > 2. Then we applied the Venn diagram (http://bioinformatics.psb.ugent.be/webtools/Venn/) to illustrate the intersection among TCGA-DEGs, ICGC-DEGs, and CTCs/CTMs genes. And these intersected genes we called CTCs/CTMs related genes (CRGs).

Pathway Enrichment and PPI Network Analysis

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis were applied to explore functional annotation and pathway analysis of the CRGs with the “clusterProfiler” R package [20]. The cut-offs were set as: p < 0.05, and FDR < 0.05. Gene Set Enrichment Analysis (GSEA) was employed to explore the common biological pathways [21] and cp.kegg.v7.1.symbols.gmt served as a reference gene set, with a threshold of p < 0.05, to screen for dominant enrichment pathways in different risk groups. Besides, to analyze the interaction between the CRGs, we employed the STRING database (https://string-db.org/) to set a protein-protein interaction (PPI) network. And an interaction score >0.7 was set as significant. Moreover, we used the Cytoscape software to visually converting the PPI network. Plug-in Molecular Complex Detection (MCODE) was utilized in Cytoscape to select the central modules of PPI network with degree cut-off = 2, node score cut-off = 0.2, k-core = 2, and max. Depth = 100 as the criterion [22]. In addition, we adopted a novel Cytoscape plugin, cytoHubba, for ranking the nodes in the network by their network functionality [23]. The "GSVA" package was used to explore the signaling pathways between the high-risk and low-risk groups [24].

Construction and Validation of the Risk Prognostic Model

Univariate Cox regression analysis was performed to identify the prognostic CRGs and the CRGs associated with survival time(p<0.01) were considered to be statistically significant. Then, the least absolute shrinkage and selection operator (LASSO) penalized Cox regression analysis was applied to further filter prognostic CRGs for overall survival in patients with HCC [25]. In addition, the risk signature was developed based on multivariate Cox proportional hazards regression analysis. As previously reported, prognostic gene signatures were constructed based on linear combinations of regression coefficients derived by multiplying the LASSO Cox regression model coefficients by their mRNA expression levels [26]. The risk score = ∑βmRNAn * expression level of mRNAs. ROC curves and Kaplan-Meier curves were used to evaluate the predictive value of the TCGA cohort in the prognostic model. Meanwhile, the ICGC database was used as external validation data to test the predictive capability of the model.

Identification of HCC subtypes

A non-negative matrix decomposition (NMF) clustering algorithm was performed using the five obtained signature genes to determine the subtypes of CRG in HCC by the "NMF" R package [27]. Using conformal, scatter and silhouette features, the optimal number of clusters with n=2 was determined. In addition, we further explored the prognostic value and the level of immune infiltration between the two subgroups.

Establishing the predictive nomogram

Researchers often use nomograms as a tool for prognostic analysis of tumor patients [28]. We constructed a simplified HCC Nomogram based on the CRGs model and evaluated its predictive performance by plotting calibration curves.

Bioinformatics analysis of the prognostic signature

We explored the correlations between different risk groups and clinical characteristics using chi-square tests, and the results were represented as heat maps. We also studied the association between the signature genes and immune cell infiltration in TIMER (https://cistrome.shinyapps.io/timer/). We used six algorithms including CIBERSORT-ABS, TIMER, QUANTISEQ, MCPCOUNTER, XCELL, and EPIC to assess the differences in immune microenvironment between the two risk groups. Waterfall plots about the high-risk and low-risk groups were produced by the R package "maftools". We also compared the differences in the expression of MHC molecules, human lymphocyte antigen (HLA) signature, chemokines and potential immune checkpoints between the two groups.

Screening potential therapeutic small molecule drugs for HCC

To better identify small molecule compound relevant to HCC treatment, we using the pRRophetic R package to calculate the half maximal inhibitory concentration (IC50) with the Genomics of Drug Sensitivity in Cancer (GDSC) database[29].

Statistical analysis

Bioinformatics analysis and mapping in our study were achieved with R software. Survival rates were compared using KM analysis, Chi-square test for comparison between categorical variables, and Student's t-test to assess differences between groups. Correlations between variables were assessed using Spearman's correlation test. For each statistical analysis, P < 0.05 was considered the threshold value for significant differences.

Differentially expressed CRGs

We downloaded the LIHC gene expression profiles from the TCGA portal and ICGC portal and screened out 1622 and 628 differentially expressed genes (DEGs) respectively using “limma” R package. The DEGs of the TCGA database and ICGC database volcano plots were shown in Figure 1A and Figure 1B. A Venn diagram was used to identify the differentially expressed CRGs (Figure1C). By taking the intersection of the DEGs of TCGA and ICGC database with the 12518 CTCs/CTMs related genes (CRGs) from ctcRbase, we evaluated a total of 258 differentially expressed CRGs (Table S1).

PPI network construction

To investigate the interrelationship of the differentially expressed CRGs and acquire hub genes, we performed PPI, module analysis and obtained the co-expression networks. Firstly, the 258 differentially expressed CRGs were analyzed in STRING, and the minimum required interaction score was set to 0.7, which indicates a strong interaction between CRGs. Then, the STRING interactions were analyzed in Cytoscape and the co-expression network was displayed in Figure 2A, which contained 155 nodes and 2731deges. Besides, we identified modules with genes >50 using the MCODE plugin and characterized 10 hub genes (TOP2A, CCNB2, CDCA8, BIRC5, AURKB, CCNB1, BUB1, BUB1B, KIF20A, TTK) in that module using the cytoHubba plugin (Figure 2B). Meanwhile, this module includes 57 nodes and 1497 edges. These potential hub genes may be instrumental in the biological progression of hepatocellular carcinoma.

GO and KEGG enrichment analyses

To explore the biological category and biological processes of differentially expressed CRGs, GO and KEGG enrichment analyses were performed using R software, and the enrichment results for each term were shown in bubble charts (Figure2C-D). GO enrichment analysis showed that the differentially expressed CRGs were principally enriched in biological process (BP) terms nuclear division, organelle fission, chromosome segregation, mitotic nuclear division, etc. Besides, the cellular component (CC) terms mainly included chromosomal region, centromeric region, spindle, kinetochore, etc. Moreover, the molecular function (MF) terms showed that the CRGs were associated with tubulin binding, microtubule-binding, ATPase activity, moto activity, etc. (Figure 2C). Regarding the KEGG analysis, the main terms were showed in Figure 2D, and these differentially expressed CRGs were particularly enriched in cell cycle, microRNAs in cancer, p53 signaling pathway, and cellular senescence.

Construction of prognostic model and validation of the model in ICGC cohort

A total of 88 CRGs were found to be strongly associated with survival in HCC patients by univariate Cox regression analysis (p<0.01), of which all of the 88 CRGs were prognostic risk factors (Figure 3A). Then, the 88 CRGs were regression penalized using LASSO Cox regression to exclude relatively insignificant parameters (Figure 3B-4C). Besides, the stepwise multivariate Cox regression was employed to construct a predictive signature for HCC patients in the TCGA cohort (Figure 3D). The five genes characterized were cell division cycle associated 8 (CDCA8), tumor protein p53 inducible protein 3 (TP53I3), hepatitis A virus cellular receptor 1 (HAVCR1), MYCN proto-oncogene (MYCN), thioredoxin reductase 1 (TXNRD1). The risk score = (0.0826 x expression level of CDCA8) + (0.0112 x expression level of TP53I3) +(0.0824 x expression level of MYCN) + (0.0376 x expression level of HAVCR1) + (0.0120 x expression level of TXNRD1) (Table S2). Patients in the TCGA cohort were divided into low-risk and high-risk groups with median predictive index as the cut-off point. As figure 4A shown, we founded that low-risk HCC patients were significantly associated with better clinical outcomes. Moreover, to better judge the predictive ability of this prognostic signature, we conducted the ROC analysis with the risk score. The data showed that the area under the curve (AUC) predicted using the ROC curve for 1-year, 3-year, and 5-year overall survival was 0.804, 0.736, and 0.707, respectively (Figure 4B). According to figure 4C, the upper panel shows the expression heat map of 5 prognostic model genes in the high- and low subgroup, the middle panel reveals that the risk of HCC patients increases with the increasing risk score, the lower panel demonstrates the poor survival time of the patients in the high-risk group compared to those in the low-risk group. For validating the predictive power of the signature, we employed the same formula to analyze the risk score of each patient in the ICGC data portal (LIRI-JP) as an independent external validation. Similarly, the Kaplan–Meier curves show a poor prognosis for the high-risk subgroup of HCC patients (Figure 4D). In addition, the ROC curve also showed the good ability of the signature to predict the prognosis of the patients, and the AUCs for risk score predicting OS at 1-year, 3-year, 5-year was 0.714, 0.730, 0.726 (Figure 4E). Likewise, the expression of the 5 CRGs and mortality in HCC patients increased in parallel with the rise in risk scores (Figure 4F).

Independent prognostic role of the gene signature

Furthermore, to investigate whether the CTCs/CTMs related 5-gene signature could be an independent prognostic factor in HCC patients, we compared the prognostic value of this signature with several clinicopathological factors such as age, gender, grade and AJCC stage for both cohorts by using univariate and multivariate Cox regression analyses. For the TCGA cohort, univariate Cox analysis of risk score (p < 0.001, HR = 1.258, 95%CI = 1.183 – 1.338), AJCC stage (p < 0.001, HR = 1.658, 95%CI = 1.340 – 2.053) and T status (p < 0.001, HR = 1.633, 95%CI = 1.332 – 2.003) were candidate factors. Further multivariate Cox regression analysis emphasized that risk score was an independent factor for HCC patients (p < 0.001, HR = 1.227 95%CI = 1.142 – 1.318) (Figure 4G). For the ICGC cohort, univariate Cox regression analysis demonstrated that risk score (p = 0.010, HR = 1.076 95%CI = 1.017- 1.137) and tumor stage (p < 0.001, HR = 2.252, 95%CI = 1.543 – 3.287) were potential risk factors. Multivariate Cox regression analysis also indicated that risk score (p = 0.030, HR = 1.070 95%CI = 1.006 – 1.138) and tumor stage (p < 0.001, HR = 2.330, 95%CI = 1.601 – 3.391) were independent predictor for ICGC patients (Figure 4H). In conclusion, these findings indicated that our 5-CRGs risk signature was closely associated with clinical characteristics, which also had a fine predictive capacity and was expected to act as a potential prognostic indicator for HCC patients.

Validation of the signature in GEO cohort

To further validate the prognostic signature, we employed the GES14520 dataset to evaluate the predictive power of this model. As the results illustrated in figure 5A, patients in the low-risk subgroup had better survival outcomes compared to the high-risk subgroup. In the GEO cohort, the AUC for 5-year OS reached 0.684 (Figure 5B). The distribution of life status, time, and expression of model genes for each patient in the validation cohort is shown in figure 5C. As indicated by the results of independent prognostic analysis, risk score (p < 0.001, HR = 1.280 95% CI = 1.145 - 1.432), as well as AJCC staging (p < 0.001, HR = 2.017 95% CI = 1.518 - 2.681) (Figure 5D-E) were also demonstrated to be independent hazard factors. From the analysis of the above results, we found that the prognostic model is promising for predictor signature.

Identification of CTCs/CTMs related molecular subtypes

Patients were clustered into different subtypes based on the expression levels of the five prognostic signature genes using the NMF algorithm. To ensure the robustness of the clustering results, the co-efficient of correlation was used to determine the optimal number of clusters, and when k = 2, we observed clear boundaries for both subtypes (C1 and C2), indicating stable and reliable clustering of HCC samples (Figure 6A). Survival curves showed that the overall survival of cluster 1 was significantly better than that of cluster 2 (p = 0.002) (Figure 6B). Subsequently, we found that most of the immune checkpoints were more highly expressed in the C2 group, compared with the C1 group (Figure 6C). In addition, the level of immune infiltration in the tumor microenvironment was also significantly different, with the immune score, stromal score, and ESTIMATE score of the C1 group being significantly lower than those of C2 (Figure 6D). This suggests that type C tends to present "cold tumors", whereas type C2 tends to present "hot tumors". We also noted that the level of neutrophils infiltration was higher in the C1 group than in the C2 group, whereas the level of monocyitc lineage and fibroblasts infiltration was lower than in the C2 group (Figure 6E).

Identification potential therapeutic small molecule drugs

We compared the sensitivity of high and low risk groups to chemotherapeutic agents to evaluate potential drugs that could be used in HCC. Our findings indicate that the low-risk group is associated with a higher IC50 for chemotherapeutic compounds such as ABT.888 (Veliparib), AS601245 (an ATP-competitive JNK inhibitor), AG.014699 (Rucaparib), A.443654 (pan-Akt inhibitor), ATRA (Tretinoin), and AUY922 (Luminespib). Additionally, Axitinib, A.770041 (An LCK inhibitor), AZD.0530 (Saracatinib), AMG.706 (Motesanib), AKT.inhibitor.VIII, and AICAR (Acadesine) had higher IC50 in the high-risk group, which means that the low-risk group patients could benefit more from treatment with these compounds (Figure 7B). Besides, we also analyzed the sensitivity of different HCC subtypes to chemotherapeutic drugs. Group C1 might be more sensitive to Metformin, Lapatinib, Elesclomol, Docetaxel, Camptothecin, Bosutinib, Axitinib and Vinblastine, and patients in group C2 would likely benefit from survival with Cisplatin, Bortezomib, Bleomycin, Bicalutamide, Mitomycin.C, Imatinib, Etoposide, and Gemcitabine treatment (Figure S1).

Pathway Analysis by GSEA and GSVA

To further investigate the potential molecular mechanism for the signature genes and prognostic module, we utilized the Gens Set Enrichment Analysis (GSEA) in the HCC cohort. Figure 8A-E reveals the KEGG pathways of five signature genes (CDCA8, HAVCR1, MYCN, TP53I3, TXNRD1), which contains the five up-regulated and down-regulated pathways, respectively. The signature genes are mainly enriched in pathways including the KEGG cell cycle, KEGG p53 signaling pathway, KEGG complement, and coagulation cascades, KEGG drug metabolism cytochrome p450, etc. In addition, GSEA compared high and low-risk groups according to risk scores. KEGG pathways enriched in the high-risk group involved KEGG cell cycle, KEGG WNT signaling pathway, KEGG ubiquitin-mediated proteolysis, KEGG snare interactions in vesicular transport, and KEGG p53 signaling pathway. KEGG pathways enriched in the low-risk group included KEGG complement and coagulation cascades, KEGG fatty acid metabolism, KEGG primary bile acid biosynthesis, KEGG tryptophan metabolism, and KEGG valine leucine and isoleucine degradation (Figure 8F). In addition, in the present study we also utilized GSVA to analyze the differences in biological behavior between high and low risk groups. According to our results, we demonstrated that pathways related to tumor progression, such as cell cycle, DNA replication, RNA degradation, mTOR signaling pathway, and P53 signaling pathway, were mainly enriched in the high-risk group of HCC patients. In contrast, metabolism-related pathways, such as fatty acid metabolism, propanoate metabolism, butanoate metabolism, and tyrosine metabolism, were mainly present in the low-risk group of patients (Figure 7A).

Differentiation of immune infiltration between two risk subgroups

In view of the important role of immune checkpoints (ICIs) in tumor immunotherapy, we analyzed the differential expression of these genes between the two risk subgroups and displayed that common immune check genes including CTLA4, CD274, PDCD1, and TIGTI were highly expressed in the high-risk group (Figure 9A). Chemokines and their receptors are necessary for the targeted migration of immune cells and the occurrence and completion of the body's immune response. Therefore, we detected differential expression of chemokines and their receptors in the two risk subgroups, with higher levels of expression especially in patients in the high-risk group (Figure 9B-C). We also demonstrated a correlation between risk score and HLA-associated gene expression. As shown in figure 9D the abundance of HLA-related genes was lower in the low-risk group than in the high-risk group. We visualized different algorithms using heat maps, including the assessment of immune cell infiltration between the two risk subgroups, and the results suggest that the high-risk group appears to have a higher abundance of immune cell infiltration (Figure 9E). In addition, further exploration of the correlation between risk score and immune pathway activity revealed that cytolytic activity, type 1 IFN response and type 2 IFN response were higher in the low-risk group than in the high-risk group, and conversely, MHC class 1 was higher in the high-risk group (Figure 10F). Obviously, the above results demonstrate that patients in the high-risk group are more likely to benefit from immunotherapy.

Clinicopathological Parameters Correlation Analysis

To investigate the prognostic value of CRGs signature in patients with different clinical features, we drew a heat map to reveal whether there was a potential correlation between features and clinicopathological features in the high-risk and low-risk subgroups (Figure 10A). The expression of CDCA8 and TXNRD1 was higher in the high-risk group. In addition, the results illustrated that the high-risk score was closely correlated with worse T stage (p < 0.001), grade (p < 0.001), tumor stage (p < 0.001) and patient survival status (p < 0.001).

Analysis of the immunological value of CRGs signature

Since tumor mutation burden (TMB) and tumor immune dysfunction and exclusion (TIDE) are better indicators of response to immunotherapy, we calculated sample scores for each HCC patient and assessed variability between high and low risk subgroups. Consistently, our findings revealed that the high-risk group had a higher TMB and a lower TIDE index, which once again demonstrates that patients in the high-risk group have a better responsiveness to immunotherapy (Figure 10B-C). The mRNA expression-based RNAss score found a positive correlation between HCC tumor stemness and risk score, indicating that tumors in the high-risk group are more likely to undergo malignant progression and thus lose their differentiated phenotype (Figure 10D). In addition, we used the "maftools" R package to visualize the differences in somatic mutation distribution between high and low risk groups and found that the high-risk group (91.38%) had a higher mutation rate compared to the low-risk group (77.84%). The most mutated gene in the low-risk group was CTNNB1 (23%) and the most mutated gene in the high-risk group was TP53 (43%) (Figure 10E).

Analysis of risk signature genes and immune correlation

Due to the important role of immune checkpoints in immune regulation and the potential application of immune checkpoint inhibitors in cancer therapy. Therefore, we evaluated the association between the signature genes and expression of immune checkpoints (including PD-1, PD-L1, CTLA4). The results in the Figure 11A indicated that the expression of CDCA8 was positive associated with the three immune checkpoints, PD-1 (R = 0.3, p = 4.9e-09), PD-L1 (R = 0.32, p = 3.8e-10), CTLA4 (R = 0.3, p = 4.9e-09). Meanwhile, the expression of HAVCR1 was positively correlated with CTLA4 (R = 0.32, p = 1.3e-10). In addition, we evaluated the expression of immune checkpoints expression in high-risk and low-risk group. The expression of PD-1, PD-L1 and CTLA4 were remarkably higher in the high-risk group than in the low-risk HCC group (Figure 6C). This suggests that the poor prognosis manifested by high-risk HCC patients is at least partially caused by the immunosuppressive microenvironment. TISIDB (http://cis.hku.hk/TISIDB/index.php) was used to analyze the expression of signature genes in different immune subtypes, C1 (wound healing); C2 (IFN-gamma dominant); C3 (inflammatory); C4 (lymphocyte depleted); C5 (immunologically quiet); C6 (TGF-b dominant) [30]. According to the results, we noticed that these five genes have different roles in different immune subtypes, and CACA8, MYCN and TXNRD1 are differentially expressed in these immune subtypes. Among them, CDCA8 was highly expressed in C1 and C2 types, MYCN was highly expressed in C1 types, while TXNRD1 was mainly expressed in C2 and C4 types (Figure 11B). In addition, we further used the IMvigor dataset to predict the responsiveness of five labeled genes to atelelizumab treatment. Interestingly, consistent with the previous results, patients with high expression of CDCA8 and TXNRD1 could obtain a better treatment outcome (Figure 11C). We also investigated the association between abundant tumor immune infiltration (CD4+ T cells, CD8+ T cells, B cells, neutrophils, macrophages and dendritic cells) and the expression of five signature genes (Figure S2). Meanwhile, correlation coefficients >0.3 and p<0.05 were considered significant. As the result shown in the figure, we identified that CDCA8 was positive associated with the six immune cells, B cells (partial.cor = 0.441, p = 9.08e-18), CD8+ T cells (partial.cor = 0.303, p = 1.03e-08), CD4+ T cells (partial.cor = 0.359, p = 6.74e-12), macrophages (partial.cor = 0.439, p = 1.70e-17), neutrophils (partial.cor = 0.368, p = 1.63e-12), and dendritic cells (partial.cor = 0.465, p = 1.22e-19). Similarly, HAVCR1 had positive correlations with B cells (partial.cor = 0.302, p = 1.14e-08), macrophages (partial.cor = 0.302, p = 1.34e-08), neutrophils (partial.cor = 0.392, p = 4.00e-14), and dendritic cells (partial.cor = 0.317, p = 2.18e-09). And TXNRD1 was also positive associated with neutrophils (partial.cor = 0.322, p = 8.67e-10).

Comparison of CRGs signature with external prognostic models

To better assess the predictive efficacy of the CRGs prognostic model, we compared the model with 6 published HCC prognostic models. Du signature was a m6A based gene signature [31]; Fu signature was a pyroptosis-related gene signature for HCC [32]; Guo signature was a gene signature containing nine genes [33]; Lei signature was a starvation-based 9-mRNA signature [34]; Tian signature was a five gene based prognostic signature for HCC [35]; Zheng signature was a five pyroptosis-related gene signature [36]. Obviously, when comparing the accuracy of these models, we find that the C-index and RSM of our constructed CRGs signature are better than the other six models, which indicates that our model is the more optimal one (Figure 12A-B). In addition, the 1-year, 3-year, and 5-year AUCs of the CRGs model were 0.807, 0.711, and 0.667, respectively, which outperformed the other models, and these results validated our previous conclusions (Figure 12C).

Establishing and validating a predictive nomogram

To better predict the survivability of HCC patients, a nomogram including factors such as age, gender, stage, and risk score was created to predict the probability of OS at 1, 3, and 5 years in the TCGA cohort. In addition, calibration plots were drawn to assess the predictive power of the nomogram. (Figure 12D). Similarly, we also constructed two nomograms in ICGC and GSE14520 cohorts (Figure S3). All of these findings indicate the good predictive power of the prediction model.

Hepatocellular carcinoma remains a significant challenge to human health, with a continuing increase in incidence and a high rate of recurrence even after surgical resection. In recent years, many studies have demonstrated that CTCs are closely related to the metastasis, epithelial-mesenchymal transition (EMT), and recurrence of malignant tumors, including HCC [37]. Therefore, it is critical to screen molecules associated with CTCs to identify validated biomarkers for predicting HCC. Here, we constructed a reliable prognostic signature based on CTCs-related genes and explored its clinical application in HCC patients, showing that the CRGs prognostic model can accurately predict the prognosis and immunotherapy sensitivity of HCC patients.

In this study, we systematically identified 258 CRGs by analyzing the DEGs in TCGA and ICGC databases and CTC expression profiles of hepatocellular carcinoma. Then, these genes were screened to construct a five-CRGs signature in the TCGA cohort. We used KM survival analysis and ROC analysis to confirm the prognostic value of the signature, and similar results were validated in the ICGC cohort and the GEO cohort. Univariate and multifactorial Cox analyses confirmed that the signature was an independent prognostic factor. In addition, nomograms for all three cohorts showed the good predictive power of the model. Our prognostic signature genes include CDCA8, HAVCR1, TP53I3, MYCN, and TXNRD1, all of which have the potential to be HCC prognostic risk genes. Previous studies have demonstrated the ability of CDCA8 to promote cancer cell proliferation and migration in several tumors, including esophageal squamous cell carcinoma [38], thyroid Cancer [39], malignant glioma (malignant glioma) and cutaneous melanoma [40]. More importantly, Jeon et al. found that silencing CDCA8 could effectively inhibit HCC growth and stemness, suggesting the possibility that CDCA8 may be a CTCS-related gene [41]. HAVCR1 is highly expressed in a variety of tumors, including colorectal cancer, non-small-cell lung cancer, clear cell renal cell carcinoma, and hepatocellular carcinoma; and can be an independent prognostic factor [42-45]. Moreover, Ye et al. found that TIM-1⁺(HAVCR1⁺) regulatory B cell infiltration was significantly higher in tumor tissues of HCC patients compared with paraneoplastic tissues and could promote the immune escape of hepatocellular carcinoma cells, implying that it could be used as a new immune therapeutic target [45]. The p53-inducible gene 3 (PIG3), also known as TP53I3, is involved in the process of apoptosis and DNA damage response. Previous studies have revealed that TP53I3 promotes invasion and metastasis of lung cancer cells and that silencing TP53I3 enhances the chemosensitivity of non-small cell lung cancer to docetaxel [46, 47]. These findings may provide us with a new strategy for the treatment of liver cancer. MYCN, a member of the MYC proto-oncogene family, Qin et al. highlighted that MYCN may be a stem cell-like marker for HCC and could be used as a therapeutic target of acyclic retinoid (ACR) for HCC [48, 49]. TXNRD1 is an antioxidant enzyme that has been reported to be overexpressed in hepatocellular carcinoma. Lee et al. observed that inhibition of TXNRD1 suppressed HCC cell proliferation, promoted apoptosis, and could induce oxidative stress, which implies that it could be used as a therapeutic target for HCC [50]. In conclusion, these historical studies suggest that these five signature genes have an important role in the development of HCC and could all be the next therapeutic targets.

As indicated by KEGG, CRGs promote development, metastasis, and recurrence of hepatocellular carcinoma probably through the cell cycle and p53 signaling pathway. GSEA analysis of five signature genes and high-risk groups in the prognostic model identified various oncogenesis-related features, including cell cycle, p53 signaling pathway, WNT signaling pathway, DNA replication, TGF-beta signaling pathway, and notch signaling pathway. In addition, GSVA results showed that tumor progression-related pathways, such as cell cycle, DNA replication, mTOR signaling pathway and P53 signaling pathway, were mainly enriched in the high-risk group of HCC patients, which was generally consistent with GSEA results. In light of this, we also evaluated a number of potential therapeutic agents, with Veliparib [51], ATRA [52, 53], and AUY922 [54] having high drug sensitivity in the high-risk group, suggesting that these agents are likely to be therapeutic candidates.

It is well known that immunotherapy plays an increasingly important role in HCC, so we also analyzed the relevance of our model to immune infiltration and immunotherapy. Immune cell infiltration analysis demonstrated that CDCA8 and HAVCR1 correlated with the infiltration abundance of several immune cells, such as B cells, CD8+ T cells, macrophages, neutrophils, and dendritic cells. As mentioned previously, ICIs expression, TMB scores and immune cell infiltration levels were strongly associated with patients in the high-risk subgroup. Analysis of somatic mutation rates also illustrated that patients in the high-risk group had a higher frequency of mutations and a greater occurrence of TP53 mutations. The TIDE score will better assist oncologists in selecting patients more suitable for immune checkpoint inhibition therapy, and the effect of clinical application in immune checkpoint inhibition therapy decisions will be meaningful [55]. Consistently, we found that patients in the high-risk group had lower TIDE scores, while patients in the low-risk group had higher TIDE scores, suggesting that the high-risk group patients may benefit more from immunotherapy. All these results reconfirm that our model has better risk stratification capabilities and is more suitable for selecting the right patients for immunotherapy.

Notably, we also used this five-risk signature gene to identify HCC subgroups C1 and C2, with C2 as a high-risk subgroup showing a worse prognosis. Compared to group C1, group C2 had higher ICIs expression and higher stromal score, immune score and ESTIMATE score for each sample, which also suggested that patients in group C2 were more suitable for immunotherapy. More importantly, we also identified several sensitive chemotherapeutic agents for C2 patients, Cisplatin [56], Bortezomib [57], Bleomycin, Bicalutamide, Mitomycin.C, Imatinib, Etoposide, and Gemcitabine [58], from which patients in the C2 group could improve their prognosis. In conclusion, our analysis is undoubtedly helpful, but its application in future studies is necessary. However, there are some limitations of our study, such as we did not further investigate the regulatory role of these five CRGs in HCC experimentally, and other external validation of our model is lacking and still needs to be conducted in clinical samples in the future. Therefore, more studies and evidence are required to refine our model in the future.

In summary, our study provided further insight into the function of CTCs related genes in HCC and proposed a five CRGs signature that could be personalized to predict patient prognosis. In addition, we highlighted the potential value of this signature in HCC risk stratification and immunotherapy.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

All datasets and materials involved in this study are provided in the article/Additional Material. And the scripts utilized during this work are available from the corresponding author on reasonable request.

Authors' contributions

XL and ZD designed the study. XL, YS and PQ analyzed the data. KZ and YL participated in software analysis. XL and KZ wrote the manuscript and were responsible for language revisions. W.Y. and J.W. supervised the project. All authors reviewed the manuscript.

Competing interests

The authors declare no conflict of interest.

Funding

This study was funded by the National Natural Science Foundation of China (No. 81874062; No.82072730) and Youth Program of National Natural Science Foundation of China (N0.81902439). We sincerely thank the public database for using their data.

Acknowledgments

Thanks for the support from National Natural Science Foundation of China (No.81874062, 82072730), and Youth Program of National Natural Science Foundation of China (81902439). We also thank all the authors for their enthusiastic contributions and solidarity.

1. Fitzmaurice C, Allen C, Barber RM, Barregard L, Bhutta ZA, Brenner H, Dicker DJ, Chimed-Orchir O, Dandona R, Dandona L, et al: Global, Regional, and National Cancer Incidence, Mortality, Years of Life Lost, Years Lived With Disability, and Disability-Adjusted Life-years for 32 Cancer Groups, 1990 to 2015: A Systematic Analysis for the Global Burden of Disease Study. JAMA Oncol 2017, 3:524-548.

2. Villanueva A: Hepatocellular Carcinoma. N Engl J Med 2019, 380:1450-1462.

3. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F: Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021, 71:209-249.

4. Llovet JM, Kelley RK, Villanueva A, Singal AG, Pikarsky E, Roayaie S, Lencioni R, Koike K, Zucman-Rossi J, Finn RS: Hepatocellular carcinoma. Nat Rev Dis Primers 2021, 7:6.

5. Marrero JA, Kulik LM, Sirlin CB, Zhu AX, Finn RS, Abecassis MM, Roberts LR, Heimbach JK: Diagnosis, Staging, and Management of Hepatocellular Carcinoma: 2018 Practice Guidance by the American Association for the Study of Liver Diseases. Hepatology 2018, 68:723-750.

6. Kluger MD, Salceda JA, Laurent A, Tayar C, Duvoux C, Decaens T, Luciani A, Van Nhieu JT, Azoulay D, Cherqui D: Liver resection for hepatocellular carcinoma in 313 Western patients: tumor biology and underlying liver rather than tumor size drive prognosis. J Hepatol 2015, 62:1131-1140.

7. Zhao L, Wu X, Li T, Luo J, Dong D: ctcRbase: the gene expression database of circulating tumor cells and microemboli. Database (Oxford) 2020, 2020.

8. Alix-Panabières C, Pantel K: Circulating tumor cells: liquid biopsy of cancer. Clin Chem 2013, 59:110-118.

9. Allard WJ, Matera J, Miller MC, Repollet M, Connelly MC, Rao C, Tibbe AG, Uhr JW, Terstappen LW: Tumor cells circulate in the peripheral blood of all major carcinomas but not in healthy subjects or patients with nonmalignant diseases. Clin Cancer Res 2004, 10:6897-6904.

10. Cristofanilli M, Budd GT, Ellis MJ, Stopeck A, Matera J, Miller MC, Reuben JM, Doyle GV, Allard WJ, Terstappen LW, Hayes DF: Circulating tumor cells, disease progression, and survival in metastatic breast cancer. N Engl J Med 2004, 351:781-791.

11. Aceto N, Bardia A, Miyamoto DT, Donaldson MC, Wittner BS, Spencer JA, Yu M, Pely A, Engstrom A, Zhu H, et al: Circulating tumor cell clusters are oligoclonal precursors of breast cancer metastasis. Cell 2014, 158:1110-1122.

12. Hou JM, Krebs MG, Lancashire L, Sloane R, Backen A, Swain RK, Priest LJ, Greystoke A, Zhou C, Morris K, et al: Clinical significance and molecular characteristics of circulating tumor cells and circulating tumor microemboli in patients with small-cell lung cancer. J Clin Oncol 2012, 30:525-532.

13. Ye Q, Ling S, Zheng S, Xu X: Liquid biopsy in hepatocellular carcinoma: circulating tumor cells and circulating tumor DNA. Mol Cancer 2019, 18:114.

14. De Rubis G, Rajeev Krishnan S, Bebawy M: Liquid Biopsies in Cancer Diagnosis, Monitoring, and Prognosis. Trends Pharmacol Sci 2019, 40:172-186.

15. Yu JJ, Xiao W, Dong SL, Liang HF, Zhang ZW, Zhang BX, Huang ZY, Chen YF, Zhang WG, Luo HP, et al: Effect of surgical liver resection on circulating tumor cells in patients with hepatocellular carcinoma. BMC Cancer 2018, 18:835.

16. Kelley RK, Magbanua MJ, Butler TM, Collisson EA, Hwang J, Sidiropoulos N, Evason K, McWhirter RM, Hameed B, Wayne EM, et al: Circulating tumor cells in hepatocellular carcinoma: a pilot study of detection, enumeration, and next-generation sequencing in cases and controls. BMC Cancer 2015, 15:206.

17. Sun YF, Xu Y, Yang XR, Guo W, Zhang X, Qiu SJ, Shi RY, Hu B, Zhou J, Fan J: Circulating stem cell-like epithelial cell adhesion molecule-positive tumor cells indicate poor prognosis of hepatocellular carcinoma after curative resection. Hepatology 2013, 57:1458-1468.

18. Wang Y, Navin NE: Advances and applications of single-cell sequencing technologies. Mol Cell 2015, 58:598-609.

19. Abouleila Y, Onidani K, Ali A, Shoji H, Kawai T, Lim CT, Kumar V, Okaya S, Kato K, Hiyama E, et al: Live single cell mass spectrometry reveals cancer-specific metabolic profiles of circulating tumor cells. Cancer Sci 2019, 110:697-706.

20. Yu G, Wang LG, Han Y, He QY: clusterProfiler: an R package for comparing biological themes among gene clusters. Omics 2012, 16:284-287.

21. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Paulovich A, Pomeroy SL, Golub TR, Lander ES, Mesirov JP: Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 2005, 102:15545-15550.

22. Bader GD, Hogue CW: An automated method for finding molecular complexes in large protein interaction networks. BMC Bioinformatics 2003, 4:2.

23. Chin CH, Chen SH, Wu HH, Ho CW, Ko MT, Lin CY: cytoHubba: identifying hub objects and sub-networks from complex interactome. BMC Syst Biol 2014, 8 Suppl 4:S11.

24. Hänzelmann S, Castelo R, Guinney J: GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics 2013, 14:7.

25. Tibshirani R: The lasso method for variable selection in the Cox model. Stat Med 1997, 16:385-395.

26. Lossos IS, Czerwinski DK, Alizadeh AA, Wechser MA, Tibshirani R, Botstein D, Levy R: Prediction of survival in diffuse large-B-cell lymphoma based on the expression of six genes. N Engl J Med 2004, 350:1828-1837.

27. Gaujoux R, Seoighe C: A flexible R package for nonnegative matrix factorization. BMC Bioinformatics 2010, 11:367.

28. Iasonos A, Schrag D, Raj GV, Panageas KS: How to build and interpret a nomogram for cancer prognosis. J Clin Oncol 2008, 26:1364-1370.

29. Yang W, Soares J, Greninger P, Edelman EJ, Lightfoot H, Forbes S, Bindal N, Beare D, Smith JA, Thompson IR, et al: Genomics of Drug Sensitivity in Cancer (GDSC): a resource for therapeutic biomarker discovery in cancer cells. Nucleic Acids Res 2013, 41:D955-961.

30. Ru B, Wong CN, Tong Y, Zhong JY, Zhong SSW, Wu WC, Chu KC, Wong CY, Lau CY, Chen I, et al: TISIDB: an integrated repository portal for tumor-immune system interactions. Bioinformatics 2019, 35:4200-4202.

31. Du Y, Ma Y, Zhu Q, Liu T, Jiao Y, Yuan P, Wang X: An m6A-Related Prognostic Biomarker Associated With the Hepatocellular Carcinoma Immune Microenvironment. Front Pharmacol 2021, 12:707930.

32. Fu XW, Song CQ: Identification and Validation of Pyroptosis-Related Gene Signature to Predict Prognosis and Reveal Immune Infiltration in Hepatocellular Carcinoma. Front Cell Dev Biol 2021, 9:748039.

33. Guo DZ, Huang A, Wang YP, Cao Y, Fan J, Yang XR, Zhou J: Development of an Eight-gene Prognostic Model for Overall Survival Prediction in Patients with Hepatocellular Carcinoma. J Clin Transl Hepatol 2021, 9:898-908.

34. Lei D, Chen Y, Zhou Y, Hu G, Luo F: A Starvation-Based 9-mRNA Signature Correlates With Prognosis in Patients With Hepatocellular Carcinoma. Front Oncol 2021, 11:716757.

35. Tian D, Yu Y, Zhang L, Sun J, Jiang W: A Five-Gene-Based Prognostic Signature for Hepatocellular Carcinoma. Front Med (Lausanne) 2021, 8:681388.

36. Zheng S, Xie X, Guo X, Wu Y, Chen G, Chen X, Wang M, Xue T, Zhang B: Identification of a Pyroptosis-Related Gene Signature for Predicting Overall Survival and Response to Immunotherapy in Hepatocellular Carcinoma. Front Genet 2021, 12:789296.

37. Qi LN, Xiang BD, Wu FX, Ye JZ, Zhong JH, Wang YY, Chen YY, Chen ZS, Ma L, Chen J, et al: Circulating Tumor Cells Undergoing EMT Provide a Metric for Diagnosis and Prognosis of Patients with Hepatocellular Carcinoma. Cancer Res 2018, 78:4731-4744.

38. Xie J, Wang B, Luo W, Li C, Jia X: Upregulation of KIF18B facilitates malignant phenotype of esophageal squamous cell carcinoma by activating CDCA8/mTORC1 pathway. J Clin Lab Anal 2022, 36:e24633.

39. Xiang C, Sun WH, Ke Y, Yu X, Wang Y: CDCA8 Contributes to the Development and Progression of Thyroid Cancer through Regulating CDK1. J Cancer 2022, 13:2322-2335.

40. Ci C, Tang B, Lyu D, Liu W, Qiang D, Ji X, Qiu X, Chen L, Ding W: Overexpression of CDCA8 promotes the malignant progression of cutaneous melanoma and leads to poor prognosis. Int J Mol Med 2019, 43:404-412.

41. Jeon T, Ko MJ, Seo YR, Jung SJ, Seo D, Park SY, Park KU, Kim KS, Kim M, Seo JH, et al: Silencing CDCA8 Suppresses Hepatocellular Carcinoma Growth and Stemness via Restoration of ATF3 Tumor Suppressor and Inactivation of AKT/β-Catenin Signaling. Cancers (Basel) 2021, 13.

42. Wang Y, Martin TA, Jiang WG: HAVcR-1 expression in human colorectal cancer and its effects on colorectal cancer cells in vitro. Anticancer Res 2013, 33:207-214.

43. Zheng X, Xu K, Chen L, Zhou Y, Jiang J: Prognostic value of TIM-1 expression in human non-small-cell lung cancer. J Transl Med 2019, 17:178.

44. Cuadros T, Trilla E, Sarró E, Vilà MR, Vilardell J, de Torres I, Salcedo M, López-Hellin J, Sánchez A, Ramón y Cajal S, et al: HAVCR/KIM-1 activates the IL-6/STAT-3 pathway in clear cell renal cell carcinoma and determines tumor progression and patient outcome. Cancer Res 2014, 74:1416-1428.

45. Ye L, Zhang Q, Cheng Y, Chen X, Wang G, Shi M, Zhang T, Cao Y, Pan H, Zhang L, et al: Tumor-derived exosomal HMGB1 fosters hepatocellular carcinoma immune evasion by promoting TIM-1(+) regulatory B cell expansion. J Immunother Cancer 2018, 6:145.

46. Gu MM, Gao D, Yao PA, Yu L, Yang XD, Xing CG, Zhou J, Shang ZF, Li M: p53-inducible gene 3 promotes cell migration and invasion by activating the FAK/Src pathway in lung adenocarcinoma. Cancer Sci 2018, 109:3783-3793.

47. Li M, Li S, Liu B, Gu MM, Zou S, Xiao BB, Yu L, Ding WQ, Zhou PK, Zhou J, Shang ZF: PIG3 promotes NSCLC cell mitotic progression and is associated with poor prognosis of NSCLC patients. J Exp Clin Cancer Res 2017, 36:39.

48. Qin XY, Suzuki H, Honda M, Okada H, Kaneko S, Inoue I, Ebisui E, Hashimoto K, Carninci P, Kanki K, et al: Prevention of hepatocellular carcinoma by targeting MYCN-positive liver cancer stem cells with acyclic retinoid. Proc Natl Acad Sci U S A 2018, 115:4969-4974.

49. Qin XY, Su T, Yu W, Kojima S: Lipid desaturation-associated endoplasmic reticulum stress regulates MYCN gene expression in hepatocellular carcinoma cells. Cell Death Dis 2020, 11:66.

50. Lee D, Xu IM, Chiu DK, Leibold J, Tse AP, Bao MH, Yuen VW, Chan CY, Lai RK, Chin DW, et al: Induction of Oxidative Stress Through Inhibition of Thioredoxin Reductase 1 Is an Effective Therapeutic Approach for Hepatocellular Carcinoma. Hepatology 2019, 69:1768-1786.

51. Muñoz-Gámez JA, López Viota J, Barrientos A, Carazo Á, Sanjuán-Nuñez L, Quiles-Perez R, Muñoz-de-Rueda P, Delgado Á, Ruiz-Extremera Á, Salmerón J: Synergistic cytotoxicity of the poly (ADP-ribose) polymerase inhibitor ABT-888 and temozolomide in dual-drug targeted magnetic nanoparticles. Liver Int 2015, 35:1430-1441.

52. Wang S, Liu J, Wu H, Jiang A, Zhao K, Yan K, Wu W, Han H, Zhang Y, Yang W: All-trans retinoic acid (ATRA) inhibits insufficient radiofrequency ablation (IRFA)-induced enrichment of tumor-initiating cells in hepatocellular carcinoma. Chin J Cancer Res 2021, 33:694-707.

53. Sun J, Liu C, Shi J, Wang N, Jiang D, Mao F, Gu J, Zhou L, Shen L, Lau WY, Cheng S: A novel chemotherapy strategy for advanced hepatocellular carcinoma: a multicenter retrospective study. Chin Med J (Engl) 2022.

54. Augello G, Emma MR, Cusimano A, Azzolina A, Mongiovì S, Puleio R, Cassata G, Gulino A, Belmonte B, Gramignoli R, et al: Targeting HSP90 with the small molecule inhibitor AUY922 (luminespib) as a treatment strategy against hepatocellular carcinoma. Int J Cancer 2019, 144:2613-2624.

55. Jiang P, Gu S, Pan D, Fu J, Sahu A, Hu X, Li Z, Traugh N, Bu X, Li B, et al: Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. Nat Med 2018, 24:1550-1558.

56. Lee JO, Lee KW, Oh DY, Kim JH, Im SA, Kim TY, Bang YJ: Combination chemotherapy with capecitabine and cisplatin for patients with metastatic hepatocellular carcinoma. Ann Oncol 2009, 20:1402-1407.

57. Li L, Zhang Y, Zhou Y, Hu H, Hu Y, Georgiades C, Mao HQ, Selaru FM: Quaternary nanoparticles enable sustained release of bortezomib for hepatocellular carcinoma. Hepatology 2022.

58. Zhu AX, Blaszkowsky LS, Ryan DP, Clark JW, Muzikansky A, Horgan K, Sheehan S, Hale KE, Enzinger PC, Bhargava P, Stuart K: Phase II study of gemcitabine and oxaliplatin in combination with bevacizumab in patients with advanced hepatocellular carcinoma. J Clin Oncol 2006, 24:1898-1903.

CTCs/CTMs, circulating tumor cells/microemboli; CRGs, CTCs/CTMs related genes; DEGs, differentially expressed genes; GO, Gene Ontology; GSEA, Gene Set Enrichment Analysis; GSVA, Gene set variation analysis; HCC, hepatocellular carcinoma; ICGC, The International Cancer Genome Consortium database; KEGG, Kyoto Encyclopedia of Genes and Genomes; LASSO, the least absolute shrinkage and selection operator; NMF, nonnegative matrix factorization; PPI, protein–protein interaction; ROC, receiver operating characteristic; TCGA, The Cancer Genome Atlas; TMB, tumor mutation burden; TIDE, tumor immune dysfunction and exclusion.

SupplementaryFigure.S1.tif
IC50 analysis of cytotoxic chemotherapeutic agents between two C1 and C2.
SupplementaryFigure.S2.tif
The correlation between the five CRGs and six immune cells in TIMER.
SupplementaryFigure.S3.tif
Nomogram and calibration for external validation dataset.
SupplementaryTable.docx
Table1. 258 CTCs/CTMs related genes (CRGs); Table2. CRGs extracted via Lasso regression analysis

Download PDF

Version 1

posted

You are reading this latest preprint version

Integrative study reveals the prognostic and immunotherapeutic value of CTCs/CTMs -related prognostic signature in Hepatocellular carcinoma

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Results

Discussion

Conclusion

Declarations

References

Abbreviations

Supplementary Files

Status:

Version 1