Role of Genetic Variation in ABC Transporters in Breast Cancer Prognosis and Therapy Response

Flexible integration of spatial and expression information for precise spot embedding via ZINB-based graph-enhanced autoencoder

Domain identification is a critical problem in spatially resolved transcriptomics data analysis, which aims to identify distinct spatial domains within a tissue that maintain both spatial continuity and expression consistency. The degree of coupling between expression data and spatial information in different datasets often varies significantly. Some regions have intact and clear boundaries, while others exhibit blurred boundaries with high intra-domain expression similarity. However, most domain identification methods do not adequately integrate expression and spatial information to flexibly identify different types of domains. To address these issues, we introduce Spot2vector, a computational framework that leverages a graph-enhanced autoencoder integrating zero-inflated negative binomial distribution modeling, combining both graph convolutional networks and graph attention networks to extract the latent embeddings of spots. Spot2vector encodes and integrates spatial and expression information, enabling effective identification of domains with diverse spatial patterns across spatially resolved transcriptomics data generated by different platforms. The decoders enable us to decipher the distribution and generation mechanisms of data while improving expression quality through denoising. Extensive validation and analyses demonstrate that Spot2vector excels in enhancing domain identification accuracy, effectively reducing data dimensionality, improving expression recovery and denoising, and precisely capturing spatial gene expression patterns.

Potential role of genetic polymorphisms in neoadjuvant chemotherapy response in breast cancer

Chemoresistance leads to treatment failure, which can arise through different mechanisms including patients' characteristics. Searching for genetic profiles as a predictor for drug response and toxicity has been extensively studied in pharmacogenomics, thus contributing to personalized medicine and providing alternative treatments. Numerous studies have demonstrated significant evidence of association between genetic polymorphisms and response to neoadjuvant chemotherapy (NAC) in breast cancer. In this review, we explored the potential impact of genetic polymorphisms in NAC primary resistance through selecting a specific clinical profile. The genetic variability within pharmacokinetics, pharmacodynamics, DNA synthesis and repair, and oncogenic signaling pathways genes could be predictive or prognostic markers for NAC resistance. The clinical implication of these results can help provide individualized treatment plans in the early stages of breast cancer treatment. Further studies are needed to determine the genetic hosts of primary chemoresistance mechanisms in order to further emphasize the implementation of genotypic approaches in personalized medicine.

ADME gene-driven prognostic model for bladder cancer: a breakthrough in predicting survival and personalized treatment

BACKGROUND

Genes that participate in the absorption, distribution, metabolism, excretion (ADME) processes occupy a central role in pharmacokinetics. Meanwhile, variability in clinical outcomes and responses to treatment is notable in bladder cancer (BLCA).

METHODS

Our study utilized expansive datasets from TCGA and the GEO to explore prognostic factors in bladder cancer. Utilizing both univariate Cox regression and the lasso regression techniques, we identified ADME genes critical for patient outcomes. Utilizing genes identified in our study, a model for assessing risk was constructed. The evaluation of this model's predictive precision was conducted using Kaplan-Meier survival curves and assessments based on ROC curves. Furthermore, we devised a predictive nomogram, offering a straightforward visualization of crucial prognostic indicators. To explore the potential factors mediating the differences in outcomes between high and low risk groups, we performed comprehensive analyses including Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG)-based enrichment analyses, immune infiltration variations, somatic mutation landscapes, and pharmacological sensitivity response assessment etc. Immediately following this, we selected core genes based on the PPI network and explored the prognostic potential of the core genes as well as immune modulation, and pathway activation. And the differential expression was verified by immunohistochemistry and qRT-PCR. Finally we explored the potential of the core genes as pan-cancer biomarkers.

RESULTS

Our efforts culminated in the establishment of a validated 17-gene ADME-centered risk prediction model, displaying remarkable predictive accuracy for BLCA prognosis. Through separate cox regression analyses, the importance of the model's risk score in forecasting BLCA outcomes was substantiated. Furthermore, a novel nomogram incorporating clinical variables alongside the risk score was introduced. Comprehensive studies established a strong correlation between the risk score and several key indicators: patterns of immune cell infiltration, reactions to immunotherapy, landscape of somatic mutation and profiles of drug sensitivity. We screened the core prognostic gene CYP2C8, explored its role in tumor bioregulation and validated its upregulated expression in bladder cancer. Furthermore, we found that it can serve as a reliable biomarker for pan-cancer.

CONCLUSION

The risk assessment model formulated in our research stands as a formidable instrument for forecasting BLCA prognosis, while also providing insights into the disease's progression mechanisms and guiding clinical decision-making strategies.

Association between cancer-associated fibroblasts and prognosis of neoadjuvant chemoradiotherapy in esophageal squamous cell carcinoma: a bioinformatics analysis based on single-cell RNA sequencing

BACKGROUND

Esophageal squamous cell carcinoma (ESCC) is a prevalent and aggressive subtype of esophageal cancer, posing a significant mortality and economic burden, especially in East and Southeast Asia. Current therapeutic strategies have limitations in improving patient survival, particularly regarding disease progression and resistance. This study aimed to investigate the impact of neoadjuvant chemoradiotherapy (NCRT) on the ESCC microenvironment.

METHODS

We utilized single-cell RNA sequencing to systematically characterize the tumor and cancer-associated fibroblasts (CAFs) subtypes. Marker genes of myofibroblastic CAFs (myCAFs) were employed to establish a prognostic model and verify its application in other datasets. Other experiments were conducted on clinical samples to explore potential ESCC risk-related genes.

RESULTS

Our bioinformatics and statistical analyses revealed an increased proportion of fibroblasts and epithelial cells in NCRT and identified the Ep_c1 subtype associated with a better prognosis. Further results indicated a complex communication network between Ep_c1 and myCAFs. The top 30 marker genes of myCAFs were used to construct a prognostic signature with a significant response to immunotherapy. Finally, experiments identified Complement C1s subcomponent (C1S), Decorin (DCN), and Neuroblastoma suppression of tumorigenicity 1 (NBL1) as potential ESCC risk-related genes.

CONCLUSION

Our findings highlight the dynamic alterations in the post-NCRT ESCC microenvironment and provide a foundation for the development of personalized treatment and immunotherapeutic approaches. Future studies are warranted to further validate these findings and explore their clinical implications.

Designing of neoepitopes based vaccine against breast cancer using integrated immuno and bioinformatics approach

Cancer is characterized by genetic instability due to accumulation of somatic mutations in the genes which generate neoepitopes (mutated epitopes) for targeting by Cytotoxic T lymphocytes (CTL). Breast cancer has a high transformation rate with unique composition of mutational burden and neoepitopes load that open a platform to designing a neoepitopes-based vaccine. Neoepitopes-based therapeutic cancer vaccines designed by neoantigens have shown to be feasible, nontoxic, and immunogenic in cancer patients. Stimulation of CTL by neoepitope-based vaccine of self-antigenic proteins plays a key role in distinguishing cancer cells from normal cells and selectively targets only malignant cells. A neoepitopes-based vaccine to combat breast cancer was designed by combining immunology and bioinformatics approaches. The vaccine construct was assembled by the fusion of CTL neoepitopes, helper sequences (used for better separation of the epitopes), and adjuvant together with linkers. The neoepitopes were identified from somatic mutations in the MUC16, TP53, RYR2, F5, DNAH17, ASPM, and ABCA13 self-antigenic proteins. The vaccine construct was undertaken to study the immune simulations (IS), physiochemical characteristics (PP), molecular docking (MD) and simulations, and cloning in appropriate vector. Together, these parameters establish safety, stability, and a strong binding affinity against class I MHC molecules capable of inducing a complete immune response against breast cancer cells.Communicated by Ramaswamy H. Sarma.

The uniqueness of ABCB5 as a full transporter ABCB5FL and a half-transporter-like ABCB5β

The ABCB5 gene encodes several isoforms, including two transporters (i.e., ABCB5FL, ABCB5β) and several soluble proteins, such as ABCB5α which has been hypothesized to have a regulatory function. ABCB5FL is a full ABC transporter and is expressed in the testis and prostate, whereas ABCB5β is an atypical half-transporter with a ubiquitous expression pattern. ABCB5β has been shown to mark cancer stem cells in several cancer types. In addition, ABCB5β and ABCB5FL have been shown to play a role in tumorigenesis and multidrug resistance. However, ABCB5β shares its entire protein sequence with ABCB5FL, making them difficult to distinguish. It cannot be excluded that some biological effects described for one transporter may be mediated by the other isoform. Therefore, it is difficult to interpret the available data and some controversies remain regarding their function in cancer cells. In this review, we discuss the data collected on ABCB5 isoforms over the last 20 years and propose a common ground on which we can build further to unravel the pathophysiological roles of ABCB5 transporters.

Molecular characterization, clinical value, and cancer-immune interactions of genes related to disulfidptosis and ferroptosis in colorectal cancer

BACKGROUND

This research strived to construct a new signature utilizing disulfidptosis-related ferroptosis (SRF) genes to anticipate response to immunotherapy, prognosis, and drug sensitivity in individuals with colorectal cancer (CRC).

METHODS

The data for RNA sequencing as well as corresponding clinical information of individuals with CRC, were extracted from The Cancer Genome Atlas (TCGA) dataset. SRF were constructed with the help of the random forest (RF), least absolute shrinkage and selection operator (LASSO), and stepwise regression algorithms. To validate the SRF model, we applied it to an external cohort, GSE38832. Prognosis, immunotherapy response, drug sensitivity, molecular functions of genes, and somatic mutations of genes were compared across the high- and low-risk groups (categories). Following this, all statistical analyses were conducted with the aid of the R (version 4.23) software and various packages of the Cytoscape (version 3.8.0) tool.

RESULTS

SRF was developed based on five genes (ATG7, USP7, MMD, PLIN4, and THDC2). Both univariate and multivariate Cox regression analyses established SRF as an independent, prognosis-related risk factor. Individuals from the high-risk category had a more unfavorable prognosis, elevated tumor mutational burden (TMB), and significant immunosuppressive status. Hence, they might have better outcomes post-immunotherapy and might benefit from the administration of pazopanib, lapatinib, and sunitinib.

CONCLUSION

In conclusion, SRF can act as a new biomarker for prognosis assessment. Moreover, it is also a good predictor of drug sensitivity and immunotherapy response in CRC but should undergo optimization before implementation in clinical settings.

ABCC4 suppresses glioblastoma progression and recurrence by restraining cGMP-PKG signalling

BACKGROUND

Cyclic nucleotides are critical mediators of cellular signalling in glioblastoma. However, the clinical relevance and mechanisms of regulating cyclic nucleotides in glioblastoma progression and recurrence have yet to be thoroughly explored.

METHODS

In silico, mRNA, and protein level analyses identified the primary regulator of cyclic nucleotides in recurrent human glioblastoma. Lentiviral and pharmacological manipulations examined the functional impact of cyclic nucleotide signalling in human glioma cell lines and primary glioblastoma cells. An orthotopic xenograft mice model coupled with aspirin hydrogels verified the in vivo outcome of targeting cyclic nucleotide signalling.

RESULTS

Elevated intracellular levels of cGMP, instead of cAMP, due to a lower substrate efflux from ATP-binding cassette sub-family C member 4 (ABCC4) is engaged in the recurrence of glioblastoma. ABCC4 gene expression is negatively associated with recurrence and overall survival outcomes in glioblastoma specimens. ABCC4 loss-of-function activates cGMP-PKG signalling, promoting malignancy in glioblastoma cells and xenografts. Hydrogels loaded with aspirin, inhibiting glioblastoma progression partly by upregulating ABCC4 expressions, augment the efficacy of standard-of-care therapies in orthotopic glioblastoma xenografts.

CONCLUSION

ABCC4, repressing the cGMP-PKG signalling pathway, is a tumour suppressor in glioblastoma progression and recurrence. Aspirin hydrogels impede glioblastoma progression through ABCC4 restoration and constitute a viable translational approach.

Factors Influencing Mortality in Children with Central Nervous System Tumors: A Cohort Study on Clinical Characteristics and Genetic Markers

Multidrug resistance (MDR) commonly leads to cancer treatment failure because cancer cells often expel chemotherapeutic drugs using ATP-binding cassette (ABC) transporters, which reduce drug levels within the cells. This study investigated the clinical characteristics and single nucleotide variant (SNV) in ABCB1, ABCC1, ABCC2, ABCC4, and ABCG2, and their association with mortality in pediatric patients with central nervous system tumors (CNST). Using TaqMan probes, a real-time polymerase chain reaction genotyped 15 SNPs in 111 samples. Patients were followed up until death or the last follow-up day using the Cox proportional hazards model. An association was found between the rs1045642 (ABCB1) in the recessive model (HR = 2.433, 95% CI 1.098-5.392, p = 0.029), and the ICE scheme in the codominant model (HR = 9.810, 95% CI 2.74-35.06, p ≤ 0.001), dominant model (HR = 6.807, 95% CI 2.87-16.103, p ≤ 0.001), and recessive model (HR = 6.903, 95% CI 2.915-16.544, p = 0.038) significantly increased mortality in this cohort of patients. An association was also observed between the variant rs3114020 (ABCG2) and mortality in the codominant model (HR = 5.35, 95% CI 1.83-15.39, p = 0.002) and the dominant model (HR = 4.421, 95% CI 1.747-11.185, p = 0.002). A significant association between the ICE treatment schedule and increased mortality risk in the codominant model (HR = 6.351, 95% CI 1.831-22.02, p = 0.004, HR = 9.571, 95% CI 2.856-32.07, p ≤ 0.001), dominant model (HR = 6.592, 95% CI 2.669-16.280, p ≤ 0.001), and recessive model (HR = 5.798, 95% CI 2.411-13.940, p ≤ 0.001). The genetic variants rs3114020 in the ABCG2 gene and rs1045642 in the ABCB1 gene and the ICE chemotherapy schedule were associated with an increased mortality risk in this cohort of pediatric patients with CNST.

Metastasis-Related Signature for Clinically Predicting Prognosis and Tumor Immune Microenvironment of Osteosarcoma Patients

Osteosarcoma is the most prevalent clinical malignant bone tumor in adolescents. The prognosis of metastatic osteosarcoma is still very poor. The aim of our study was to investigate the clinical diagnosis and prognostic significance of metastasis related genes (MRGs) in patients with osteosarcoma. Clinical information and RNA sequencing data with osteosarcoma patients were obtained and set as the training set from UCSC databases. GSE21257 were downloaded and chosen as the verification cohort. An eight gene metastasis related risk signature including MYC, TAC4, ABCA4, GADD45GIP1, TNFRSF21, HERC5, MAGEA11, and PDE1B was built to predict the overall survival of osteosarcoma patients. Based on risk assessments, patients were classified into high- and low-risk groups. The high-risk patients had higher risk score and shorter survival time. ROC curves revealed that this risk signature can accurately predict survival times of osteosarcoma patients at the 1-, 2-, 3-, 4- and 5- year. GSEA revealed that MYC targets, E2F targets, mTORC1 signaling, Wnt /β-catenin signaling and cell cycle were upregulated, and cell adhesion molecules, and primary immunodeficiency were decreased in high-risk group. MRGs were highly linked with the tumor immune microenvironment and ICB response. These results identified that MRGs as a novel prognostic and diagnostic biomarker in osteosarcoma.

Metabolic characterisation of transglutaminase 2 inhibitor effects in breast cancer cell lines

Transglutaminase 2 (TG2), which mediates post-translational modifications of multiple intracellular enzymes, is involved in the pathogenesis and progression of cancer. We used 1 H-NMR metabolomics to study the effects of AA9, a novel TG2 inhibitor, on two breast cancer cell lines with distinct phenotypes, MCF-7 and MDA-MB-231. AA9 can promote apoptosis in both cell lines, but it is particularly effective in MD-MB-231, inhibiting transamidation reactions and decreasing cell migration and invasiveness. This metabolomics study provides evidence of a major effect of AA9 on MDA-MB-231 cells, impacting glutamate and aspartate metabolism, rather than on MCF-7 cells, characterised by choline and O-phosphocholine decrease. Interestingly, AA9 treatment induces myo-inositol alteration in both cell lines, indicating action on phosphatidylinositol metabolism, likely modulated by the G protein activity of TG2 on phospholipase C. Considering the metabolic deregulations that characterise various breast cancer subtypes, the existence of a metabolic pathway affected by AA9 further points to TG2 as a promising hot spot. The metabolomics approach provides a powerful tool to monitor the effectiveness of inhibitors and better understand the role of TG2 in cancer.

Peripubertal Nutritional Prevention of Cancer-Associated Gene Expression and Phenotypes

Breast cancer (BC) is a nearly ubiquitous malignancy that effects the lives of millions worldwide. Recently, nutritional prevention of BC has received increased attention due to its efficacy and ease of application. Chief among chemopreventive compounds are plant-based substances known as dietary phytochemicals. Sulforaphane (SFN), an epigenetically active phytochemical found in cruciferous vegetables, has shown promise in BC prevention. In addition, observational studies suggest that the life stage of phytochemical consumption may influence its anticancer properties. These life stages, called critical periods (CPs), are associated with rapid development and increased susceptibility to cellular damage. Puberty, a CP in which female breast tissue undergoes proliferation and differentiation, is of particular interest for later-life BC development. However, little is known about the importance of nutritional chemoprevention to CPs. We sought to address this by utilizing two estrogen receptor-negative [ER(-)] transgenic mouse models fed SFN-containing broccoli sprout extract during the critical period of puberty. We found that this treatment resulted in a significant decrease in tumor incidence and weight, as well as an increase in tumor latency. Further, we found significant alterations in the long-term expression of cancer-associated genes, including p21, p53, and BRCA2. Additionally, our transcriptomic analyses identified expressional changes in many cancer-associated genes, and bisulfite sequencing revealed that the antiproliferation-associated gene Erich4 was both hypomethylated and overexpressed in our experimental group. Our study indicates that dietary interventions during the CP of puberty may be important for later-life ER(-) BC prevention and highlights potential important genetic and epigenetic targets for treatment and study of the more deadly variants of BC.

ABC transporters in breast cancer: their roles in multidrug resistance and beyond

ATP-binding cassette (ABC) transporters are membrane-spanning proteins involved in cholesterol homeostasis, transport of various molecules in and out of cells and organelles, oxidative stress, immune recognition, and drug efflux. They are long implicated in the development of multidrug resistance in cancer chemotherapy. Existing clinical and molecular evidence has also linked ABC transporters with cancer pathogenesis, prognostics, and therapy. In this review, we aim to provide a comprehensive update on all ABC transporters and their roles in drug resistance in breast cancer (BC). For solid tumours such as BC, various ABC transporters are highly expressed in less differentiated subtypes and metastases. ABCA1, ABCB1 and ABCG2 are key players in BC chemoresistance. Restraining these transporters has evolved as a possible mechanism to reverse this phenomenon. Further, ABCB1 and ABCC1 are important in BC prognosis. Newer therapeutic approaches have been developed to target all these molecules to dysregulate their effect, reduce cell viability, induce apoptosis, and increase drug sensitivity. In the future, targeted therapy for specific genetic variations and upstream or downstream molecules can help improve patient prognosis.

DNA Copy Number Aberrations and Expression of ABC Transporter Genes in Breast Tumour: Correlation with the Effect of Neoadjuvant Chemotherapy and Prognosis of the Disease

One of the important reasons for the ineffectiveness of chemotherapy in breast cancer (BC) is considered to be the formation of a multidrug resistance phenotype in tumour cells, which is caused by the expression of energy-dependent ABC transporters. The aim of this work was to assess chromosomal aberrations and the level of transcripts of all 49 known ABC transporter genes in breast tumours. Materials and Methods. The study included 129 patients with breast cancer. A microarray study of all tumour samples was carried out on microchips. Results. This study established that the presence of a deletion in genes ABCB1, ABCB4, ABCB8, ABCC7, ABCC11, ABCC12, ABCF2, and ABCG4 is associated with an objective response to treatment (p ≤ 0.05). A decrease in the expression of genes was associated with a good response to chemotherapy, whereas an increase in expression caused the progression and stabilization of the tumour. Analysis of metastatic-free survival rates showed that the presence of ABCB1/4 and ABCC1/6 deletions was associated with 100% survival (log-rank test p = 0.01 and p = 0.03). Conclusions. The study showed that the aberrant state of ABC transporter genes, as well as a decrease in the expression of these genes, is a predictor of the effectiveness of therapeutic treatment and a potential prognostic marker of metastatic survival.

MDN1 Mutation Is Associated With High Tumor Mutation Burden and Unfavorable Prognosis in Breast Cancer

Background: Breast cancer (BRCA) is the most common cancer worldwide and a serious threat to human health. MDN1 mutations have been observed in several cancers. However, the associations of MDN1 mutation with tumor mutation burden (TMB) and prognosis of BRCA have not been investigated.

Methods: Genomic, transcriptomic, and clinical data of 973 patients with BRCA from The Cancer Genome Atlas (TCGA) database were analyzed. The clinical attributes of BRCA based on the MDN1 mutation status were assessed by comparing TMB and tumor infiltrating immune cells. Gene ontology analysis and gene set enrichment analysis (GSEA) were conducted to identify the key signaling pathways associated with MDN1 mutation. Moreover, univariate and multivariate Cox regression analyses were performed to assess the association between prognostic factors and survival outcomes. Finally, nomograms were used to determine the predictive value of MDN1 mutation on clinical outcomes in patients with BRCA.

Results: MDN1 was found to have a relatively high mutation rate (2.77%). Compared to the MDN1 wild-type patients, the TMB value was significantly higher in MDN1 mutant patients (p < 0.001). Prognostic analysis revealed that MDN1 mutant patients had a worse survival probability than MDN1 wild-type patients (hazard ratio = 2.91; 95% CI:1.07–7.92; p = 0.036). GSEA revealed samples with MDN1 mutation enriched in retinol metabolism, drug metabolism cytochrome P450, glucuronidation, miscellaneous transport, and binding event pathways.

Conclusion: MDN1 mutation was found to be associated with high TMB and inferior prognosis, suggesting that MDN1 mutation may play a potential role in prognosis prediction and immunotherapy guidance in BRCA.

Use of ATP-Binding Cassette Subfamily A Member 13 (ABCA13) for Sensitive Detection of Focal Pathological Forms of Subclinical Bovine Paratuberculosis

Bovine paratuberculosis (PTB) is a chronic enteritis caused by Mycobacterium avium subspecies paratuberculosis (Map) that causes a heavy economic impact worldwide. Map infected animals can remain asymptomatic for years while transmitting the mycobacteria to other members of the herd. Therefore, accurate detection of subclinically infected animals is crucial for disease control. In a previous RNA-Seq study, we identified several mRNAs that were overexpressed in whole blood of cows with different PTB-associated histological lesions compared with control animals without detected lesions. The proteins encoded by two of these mRNAs, ATP binding cassette subfamily A member 13 (ABCA13) and Matrix Metallopeptidase 8 (MMP8) were significantly overexpressed in whole blood of animals with focal histological lesions, the most frequent pathological form in the subclinical stages of the disease. In the current study, the potential of sensitive early diagnostic tools of commercial ELISAs, based on the detection of these two biomarkers, was evaluated in serum samples of 704 Holstein Friesian cows (566 infected animals and 138 control animals from PTB-free farms). For this evaluation, infected animals were classified into three groups, according to the type of histological lesions present in their gut tissues: focal (n = 447), multifocal (n = 59), and diffuse (n = 60). The ELISA based on the detection of ABCA13 was successfully validated showing good discriminatory power between animals with focal lesions and control animals (sensitivity 82.99% and specificity 80.43%). Conversely, the MMP8-based ELISA showed a poor discriminatory power between the different histological groups and non-infected controls. The ABCA13-based ELISA showed a higher diagnostic value (0.822) than the IDEXX ELISA (0.517), the fecal bacterial isolation (0.523) and the real-time PCR (0.531) for the detection of animals with focal lesions. Overall, our results indicate that this ABCA13 ELISA greatly improves the identification of subclinically infected animals with focal lesions that are undetectable using current diagnostic methods.

Mast Cells Resting-Related Prognostic Signature in Hepatocellular Carcinoma

The immune microenvironment of liver cancer is of great significance for the treatment of liver cancer. After evaluating the content of mast cells resting in the transcriptome data of The Cancer Genome Atlas database by CIBERSORT analysis, this study aimed to group the samples according to the content of mast cells resting in different samples to find the differentially expressed genes in the two groups. Significant prognostic differences were found between high and low mast cells resting infiltration groups. The prognostic model was constructed according to the differentially expressed genes. The model was validated using external independent datasets. The results revealed that the constructed model was reliable. It could well distinguish the prognostic differences of patients in different characteristic groups. The high-risk group was mainly concentrated in metabolic pathways. The risk score of this model was closely related to some immune cells, immune function, and immune checkpoints. Therefore, this model may provide new ideas for immunotherapy of hepatocellular carcinoma.