Hepatocellular carcinoma (HCC) accounts for a substantial proportion of cancer mortality worldwide, and clinical outcomes remain poor for patients with advanced disease. T cell receptor-engineered T cells (TCR-T) represent an attractive immunotherapeutic modality because they can recognize peptide epitopes derived from intracellular tumor proteins and displayed on the cell surface by major histocompatibility complex molecules. This review critically summarizes recent progress in TCR-T therapy for HCC, with particular emphasis on innovations in TCR engineering. We survey the current landscape of targetable antigens in HCC and current approaches for identifying therapeutic TCR candidates, followed by a discussion of TCR-T and review their major safety and toxicity concerns. Importantly, we discuss the key challenges that limit the clinical efficacy, particularly the immunosuppressive tumor microenvironment, antigen heterogeneity, HLA restriction, and limited T cell persistence. We further highlight emerging strategies to enhance therapeutic efficacy and safety, including genetic engineering, metabolic and functional optimization, combinational approaches, and controllable safety mechanisms. Special considerations regarding TCR-T therapy for HCC recurrence after liver transplantation are also discussed. Finally, we outline future directions centered on refined target selection, artificial intelligence-assisted TCR discovery and design, scalable manufacturing, biomarker-driven patient selection, and rigorous safety evaluation. These insights provide a structured perspective on the current status of TCR-T therapy and the key steps required to facilitate its clinical translation in HCC.
Abstract Mitochondria are metabolic hubs that house their own genomes (mitochondrial DNA [mtDNA]), which encode components of the oxidative phosphorylation (OXPHOS) machinery. The mitochondrial central dogma not only governs compartmentalised metabolism but also intensively intertwines with multiple biological processes, and its dysregulation is a hallmark of cancer and metabolic diseases. In this review, we highlight recent advances in mitochondrial biogenesis from a metabolic perspective, with a particular emphasis on cancer. Metabolites act as donors for diverse chemical modifications, which have been systematically identified on mtDNA, rRNA, and tRNA. Besides, post-translational modifications of proteins involved in mtDNA replication, transcription, and translation has been revealed to connect metabolic signals with mitochondrial biogenesis. A comprehensive landscape of the mitochondrial central dogma has deepened our understanding of how mitochondria coordinate OXPHOS with other organelle-specific processes to obtain a flexible metabolic network, which potentiates tumor growth. Notably, non-canonical products and biological functions of the mitochondrial central dogma further reshape our concepts of cancer initiation and progression. Given that dysregulated mitochondrial biogenesis is found in multiple human disorders, including cancer, targeting this pathway offers new therapeutic opportunities. Genome-wide studies and drug screens have identified metabolic nodes and small molecules with potential to correct mitochondrial dysfunction in cancer, while emerging tools such as mtDNA editing enable precise intervention. Despite a maturing picture of mitochondrial biogenesis, many hidden players and functions remain to be uncovered to fully decipher mitochondrial biology in cancer.
The tumor microenvironment (TME) is a dynamic and immunosuppressive niche that critically governs cancer progression, immune evasion, and therapeutic resistance. Accurately modeling tumor–immune interactions in vitro is therefore essential for advancing our understanding of cancer biology and accelerating the development of effective immunotherapies. Co-culture systems have emerged as powerful platforms for this purpose, enabling the controlled study of cellular crosstalk between tumor cells and immune, stromal, or endothelial components under defined experimental conditions. Among available co-culture approaches, patient-derived tumor organoid–peripheral blood mononuclear cell (PDTO–PBMC) co-culture systems represent the most physiologically faithful platform currently available, preserving the histological, genetic, and phenotypic heterogeneity of the original tumor while enabling reconstitution of autologous immune responses. These systems allow the expansion of tumor-reactive T cells in an antigen–agnostic manner, providing a clinically relevant platform for evaluating immune checkpoint inhibitors (ICI), adoptive cell therapies (ACT), chimeric antigen receptor (CAR) T cell constructs, and combination immunotherapy regimens across multiple cancer types. This review comprehensively examines the spectrum of tumor–immune co-culture systems, from conventional 2D monolayer models to 3D spheroid and organoid-based platforms, critically comparing their biological fidelity, technical requirements, and translational utility. Particular emphasis is placed on the PDTO–PBMC co-culture model, its applications across diverse immune cell populations, including T cells, NK cells, macrophages, dendritic cells, and CAR T cells, and its potential as a functional companion diagnostic tool for precision immunotherapy. Current limitations, standardization challenges, and future directions including microfluidics, spatial transcriptomics, and artificial intelligence integration are also discussed.
Abstract Radiotherapy (RT) can initiate antitumor immunity through immunogenic cell death, tumor-associated antigen release, and innate immune engagement, but these responses are often too weak and transient to achieve durable systemic tumor control. This perspective identifies insufficient immune amplification as a central bottleneck in radioimmunotherapy and positions nano-immunoadjuvant (NIA) as programmable systems that enhance RT-induced immune priming, remodel the suppressive tumor microenvironment, and support adaptive antitumor immunity. Unlike conventional approaches focused mainly on radiosensitization, drug delivery, or immune initiation, we propose a “trigger–amplify–sustain” framework in which RT initiates immunity, NIA amplifies it, and immune checkpoint blockade maintains it. We also discuss key translational challenges, including radiation fractionation, scalable manufacturing, biomarker-guided patient selection, and regulatory evaluation, supporting a shift toward programmable radioimmunotherapy.
Accurate breast cancer prognosis remains a major challenge in precision oncology due to tumor heterogeneity and the complexity of integrating high-dimensional multi-omics data. Although multimodal learning approaches have improved predictive performance by combining clinical and molecular information, many existing methods rely on a single ensemble strategy that remains susceptible to prediction variance and limited robustness in high-dimensional, low-sample-size biomedical datasets. This study investigated whether integrating complementary ensemble strategies within a unified multimodal framework could improve the robustness and predictive performance of breast cancer prognosis. A heterogeneous multimodal ensemble framework was developed in which stacking was used to integrate complementary information from clinical, gene expression, and copy number variation (CNV) data through meta-learning, while bagging was incorporated to stabilize the meta-learning process via bootstrap aggregation. The outputs of the stacking and bagging branches were combined using weighted probability fusion. The framework was evaluated on the METABRIC breast cancer cohort and compared with unimodal models and a conventional stacking ensemble using an independent test set and stratified tenfold cross-validation. The proposed hybrid framework achieved a ROC-AUC of 0.936, outperforming unimodal clinical and molecular models (ROC-AUC = 0.8140.885) and the conventional stacking ensemble (ROC-AUC = 0.898). Stratified tenfold cross-validation further demonstrated consistent improvements in mean ROC-AUC, recall, F1-score, balanced accuracy, and Matthews correlation coefficient, indicating improved robustness and stable performance across the internal validation folds. On the independent test set, the hybrid framework reduced false-negative predictions and increased sensitivity relative to the stacking ensemble, demonstrating a more favorable balance between identifying high-risk patients and maintaining overall predictive performance. Rather than introducing a new ensemble algorithm, this study demonstrates that assigning complementary roles to stacking multimodal information integration and bagging for prediction stabilization provides an effective and robust framework for multi-omics breast cancer prognosis. The proposed hybrid strategy consistently improved predictive performance and robustness compared with conventional stacking while demonstrating stable performance across internal validation, supporting the use of complementary ensemble paradigms for multimodal prediction in precision oncology.
