Hepatocellular carcinoma ranks as the sixth most common cancer worldwide and the fourth leading cause of cancer death, a grim combination that reflects both late diagnosis and limited options once the disease spreads. For patients with advanced disease, five-year survival remains persistently low, and the standard arsenal of tyrosine kinase inhibitors, immune checkpoint inhibitors, and locoregional procedures such as radioembolization and hepatic arterial infusion chemotherapy frequently buys only months. Adoptive cell transfer immunotherapy, in which a patient’s own lymphocytes are harvested, genetically reprogrammed, and reinfused, has transformed the treatment of several blood cancers, but solid tumors have proven far more resistant. TCR-engineered T cell therapy represents one of the most sophisticated attempts to crack that problem: rather than recognizing surface proteins through a chimeric antigen receptor, as CAR-T cells do, TCR-T cells are equipped with a natural T cell receptor that can detect peptide fragments presented inside cells by human leukocyte antigen molecules, vastly expanding the range of targetable antigens.

The choice of target in this case was elegant in its logic. Because the patient’s tumor arose in the context of chronic hepatitis B infection, many of his malignant cells carried integrated hepatitis B DNA and continued to produce the viral surface antigen, HBsAg. By engineering his T cells with a receptor specific for HBsAg-derived peptides, the therapeutic team aimed to convert a persistent viral infection into a tumor-specific bullseye, sparing healthy hepatocytes that do not present the antigen. The strategy exploits a biological quirk of hepatitis B-related liver cancer that simply does not exist in tumors of other origins, and it illustrates how viral oncogenesis can be turned against the cancer itself.

The patient’s medical history underscores how heavily pretreated he was before receiving the experimental cells. Diagnosed with hepatocellular carcinoma in April 2018 after laparoscopic resection, he had been maintained on antiviral therapy, first entecavir and later tenofovir alafenamide fumarate after renal function declined. When his alpha-fetoprotein tumor marker rose in June 2022, imaging revealed retroperitoneal lymph node metastases, which were attacked with stereotactic body radiotherapy and CT-guided iodine-125 seed brachytherapy. Systemic therapy followed in force: three different tyrosine kinase inhibitors, three immune checkpoint inhibitors including the bispecific antibody cadonilimab, and eventually FOLFOX4 hepatic arterial infusion chemotherapy for recurrent intrahepatic lesions. None of it held the line. By July 2024, PET-CT showed new metastases in cervical and supraclavicular lymph nodes, the lungs, and the pelvic bones, and the patient was enrolled in an HBsAg-targeted TCR-T clinical trial classified as stage IIIB with preserved liver function.

The manufacturing and administration protocol followed the now-standard architecture of adoptive cell therapy. Mononuclear cells were collected by apheresis on August 9, 2024, and seventeen days later the patient received a lymphodepleting preconditioning regimen of cyclophosphamide and fludarabine, the latter dose-adjusted from 25 to 20 milligrams per square meter per day because of his reduced creatinine clearance. He then received an intravenous infusion of autologous TCR-T cells at a dose of 1 times 10 to the eighth cells per kilogram, totaling 5.64 billion TCR-positive T cells. Within two hours, the inflammatory cascade began: fever spiking to 39.7 degrees Celsius, chills, falling oxygen saturation, chest tightness, nausea, and vomiting. The clinical team graded his cytokine release syndrome using the National Cancer Institute’s CTCAE v5.0 criteria and initiated oxygen, fluids, antipyretics, empiric antibiotics, and the interleukin-6 blocker tocilizumab at five and thirteen hours post-infusion.

By the first day, the syndrome had escalated to grade 3. His blood pressure dropped to 96 over 53 millimeters of mercury from a baseline of 135 to 150 over 70 to 80, fluid output fell behind intake, and serum creatinine climbed, prompting low-dose norepinephrine and a third dose of tocilizumab at thirty hours. Blood pressure recovered within two hours and the vasopressor was weaned, but the picture that followed was textbook capillary leak syndrome: purpura across the lower limbs, axillae, and chest, pitting edema, diminished breath sounds, and CT evidence of mild pulmonary edema with bilateral pleural effusions. The physicians managed the leakage phase with deliberately restrained fluid resuscitation, favoring albumin-containing solutions to maintain oncotic pressure, and then pivoted to aggressive diuresis during the reabsorption phase to prevent fluid overload. Temperature stabilized by day five, and fatigue and dyspnea resolved within two weeks. Laboratory markers told a parallel story, with C-reactive protein peaking at 21.9 milligrams per liter, ferritin hitting the detection ceiling on day one, and interleukin-6 cresting on day one before falling steadily from day three.

The hematologic toxicity was, if anything, more severe than the inflammatory syndrome. Lymphocyte counts predictably collapsed after lymphodepletion, but white cells and neutrophils also plunged to grade 4 and grade 3 respectively by day one. Encouragingly, recovery began without granulocyte colony-stimulating factor, with leukocytes rebounding on day three and neutrophils on day seven, both reaching grade 2 severity by day fourteen, a pattern the authors classify as rapid recovery and one that suggests preserved bone marrow reserve. Hemoglobin, already low before infusion, bottomed out at 71 grams per liter on day four, and platelets dipped to grade 2 between days three and six before returning to baseline by day 87. Transaminases followed an inverted V-shaped curve, with grade 4 ALT and AST elevations downgrading within seven days, while cholestatic enzymes remained stable. Coagulation studies revealed a consumptive pattern, with D-dimer surging to 22.89 milligrams per liter alongside falling fibrinogen and mildly prolonged partial thromboplastin time, consistent with cytokine-driven endothelial activation and microthrombosis; the D-dimer fell by half within four days and anticoagulation was withheld.

The therapeutic signal, though transient, was genuine. At six weeks, imaging showed slight shrinkage of the cervical lymph node and lung metastases, although bone lesions continued to grow and a partial response was not achieved under modified RECIST 1.1 criteria. Alpha-fetoprotein, which had climbed to 1,575 nanograms per milliliter before infusion, fell to 413 by day eleven before rebounding. TCR-T cells were detectable in peripheral blood from day four, when they peaked, until week six. Most strikingly, HBsAg levels dropped to nearly undetectable concentrations by day seven, a direct pharmacodynamic readout confirming that the engineered cells were striking their viral target, while hepatitis B viral DNA, initially suppressed below the detection limit, rose again at week six in synchrony with tumor progression. By day 87, every lesion had progressed. The patient later suffered lacunar infarcts in the pons and left posterior ventricular horn, a rare complication the authors attribute cautiously to cytokine-mediated hypercoagulability and cerebral hypoperfusion, though his pre-existing hypertension and diabetes confound any causal attribution. Palliative radiotherapy and chemotherapy followed, and after malignant obstruction of the inferior vena cava required stenting, he entered hospice care in late February 2025 and died two weeks later.

The authors draw several sobering lessons from the trajectory. Solid tumors present barriers that hematologic malignancies do not: engineered T cells accumulate first in the lungs and lymphoid organs and migrate inefficiently into tumor masses, which may explain the transient regression of lymph node lesions, while the relentless progression of bone metastases suggests the cells never meaningfully infiltrated osseous tissue, a compartment for which no tracking data yet exist. On the safety side, the case demonstrates that TCR-T toxicity mirrors the well-mapped CAR-T experience, with grade 3 to 4 cytopenias occurring at rates comparable to published CAR-T series, where neutropenia affects 59 to 95 percent of patients, and that capillary leak syndrome demands a disciplined, phase-specific fluid strategy. The report concludes that HBsAg-targeted TCR-T therapy for advanced hepatocellular carcinoma carries a tolerable, controllable safety profile when grade 3 cytokine release syndrome and hematologic toxicity are monitored closely, even as the modest and short-lived antitumor effect makes clear that improving cell persistence, trafficking, and bone infiltration remains the decisive challenge for this promising but still immature modality.

Subject of Research: HBsAg-targeted TCR-engineered T cell therapy for advanced hepatocellular carcinoma

Article Title: Treatment of advanced hepatocellular carcinoma with HBsAg-targeted TCR engineered -T Cells: a case report

Article References: Wei, S., Gu, L., Shi, J., & Wang, W. (2025). Treatment of advanced hepatocellular carcinoma with HBsAg-targeted TCR engineered -T Cells: a case report. Clinical Cancer Bulletin, 4(1), Article 19. https://doi.org/10.1007/s44272-025-00048-w

Image Credits: AI Generated

DOI: 10.1007/s44272-025-00048-w

Keywords: hepatocellular carcinoma, TCR-T cell therapy, HBsAg, cytokine release syndrome, capillary leak syndrome, adoptive cell transfer, hematologic toxicity, hepatitis B, immunotherapy, case report, solid tumors, tocilizumab

The most mature of these currents is adoptive cell therapy, particularly chimeric antigen receptor T cell therapy, known as CAR-T. The technique involves harvesting a patient’s own T lymphocytes, genetically engineering them to express an artificial receptor that recognizes a specific tumor antigen, expanding the modified cells in culture, and reinfusing them into the patient. The chimeric antigen receptor combines an antibody-derived binding domain that locks onto the target antigen with intracellular signaling domains that activate the T cell upon binding. Since the first CAR-T products were approved for certain B cell leukemias and lymphomas, the field has expanded toward solid tumors, where it faces formidable obstacles: the physical barriers of tumor stroma, an immunosuppressive tumor microenvironment rich in molecules such as TGF-beta and adenosine, antigen heterogeneity that allows tumor cells lacking the target antigen to escape, and the risk of severe cytokine release syndrome, a systemic inflammatory reaction caused by massive immune activation. Combination approaches — pairing CAR-T cells with checkpoint inhibitors, oncolytic viruses, or stroma-modulating agents — are among the strategies under active investigation to overcome these barriers.

A closely related but technically distinct platform is TCR-T cell therapy, which the editorial singles out as an area of anticipated growth. Rather than using an artificial antibody-like receptor, TCR-T therapy engineers T cells to express a natural T cell receptor that recognizes peptide fragments of intracellular proteins presented on the cell surface by human leukocyte antigen molecules. This is a crucial advantage: most cancer-specific antigens are not displayed on the cell membrane where antibodies can reach them, but are instead processed inside the cell and presented on HLA molecules. TCR-T cells can therefore target the vast intracellular reservoir of mutated and aberrantly expressed proteins, including drivers of pancreatic, ovarian, and colorectal cancers that have remained largely untouchable by antibody-based approaches. The technical challenges include identifying HLA-restricted neoantigens that are truly tumor-specific, ensuring sufficient receptor affinity without cross-reactivity against healthy tissue, and overcoming the downregulation of HLA presentation that tumors use as an immune evasion mechanism.

Cancer vaccines, another pillar highlighted in the editorial’s recent special issue on cancer cytotherapy, are being transformed by genomic sequencing. Personalized neoantigen vaccines are designed from the unique mutation profile of an individual patient’s tumor. Computational algorithms predict which mutant peptides will bind the patient’s HLA molecules and be recognized as foreign by the immune system, and these epitopes are synthesized as mRNA or peptide-based vaccines. Combined with checkpoint blockade, which releases the brakes on T cell responses, such vaccines have shown promise in early trials for melanoma and pancreatic cancer, aiming to amplify a targeted immune attack against the molecular fingerprints unique to each malignancy.

Antibody-drug conjugates, often described as guided-missile therapeutics, represent another frontier that is becoming steadily more sophisticated. An ADC links a monoclonal antibody that targets a tumor-associated antigen to a highly potent cytotoxic payload via a chemical linker designed to remain stable in the bloodstream and release the drug after the conjugate is internalized by the cancer cell. The new generation of multi-specific ADCs discussed in the editorial goes further, engaging two or more distinct antigens simultaneously. This bispecific recognition addresses one of the biggest causes of treatment failure: tumors that downregulate a single target antigen can no longer escape simply by losing one molecular flag. Beyond ADCs, bispecific and trispecific antibodies are also being engineered to physically bridge T cells to tumor cells or to block multiple signaling pathways at once, blurring the line between antibody therapy and cellular immunotherapy.

Underpinning all of these advances is a quiet revolution in how cancer models are built and tested. Cancer organoids — three-dimensional cultures grown from patient tumor cells that self-organize into structures recapitulating the architecture and genetics of the original malignancy — are emerging as a bridge between traditional two-dimensional cell lines, which often fail to reflect tumor biology, and patient-derived xenografts in mice, which take months to establish. Because organoids can be generated within weeks and maintained in large panels, they allow drug sensitivity testing on a scale and speed previously impossible. In principle, a patient’s own tumor organoids can serve as an avatar, screening candidate regimens in the laboratory before the patient receives them in the clinic. Organoid co-cultures with immune cells and stromal components are extending this approach to predict responses to immunotherapy, one of the hardest response categories to model outside the human body.

The editorial also emphasizes the integration of nanotechnology with cancer diagnosis and management. Engineered nanoparticles, typically in the range of ten to a few hundred nanometers, can be decorated with targeting ligands, loaded with chemotherapeutics or nucleic acid payloads, and designed to accumulate preferentially in tumors through both passive mechanisms such as the enhanced permeability of tumor vasculature and active mechanisms such as antibody fragment targeting. Beyond drug delivery, nanoparticles serve as contrast and sensing platforms for imaging, as vehicles for delivering mRNA vaccines and gene-editing tools, and as systems that modulate the tumor microenvironment itself. The persistent translational challenge — often called the gap between bench and bedside — lies in controlling how nanoparticles behave once they encounter blood proteins, immune cells, and the dense extracellular matrix of human tumors, an area where careful pharmacokinetic and safety evaluation remains essential.

Perhaps the most transformative enabler of the coming decade is the fusion of multi-omics data with artificial intelligence. Genomics, transcriptomics, proteomics, metabolomics, and epigenomic profiling each capture a different molecular dimension of a tumor, and integrated analysis of these layers reveals dependencies and vulnerabilities invisible to any single data type. Machine learning models are increasingly used to predict biomarker status directly from routine histopathology slides, to forecast which patients will respond to immunotherapy, to detect radiographic patterns in CT and MRI scans that correlate with molecular features — a field known as radiomics — and to design optimal combination treatment sequences. Fan identifies AI-driven innovation as a priority topic for the coming years, alongside biomarker discovery and validation, real-world evidence studies, and the study of therapeutic resistance and its optimization, all areas where large, well-curated clinical datasets are the essential raw material.

None of these technologies will succeed in isolation, and this is precisely the point the editorial underscores with its call for multidisciplinary integration. A patient receiving CAR-T therapy may also need an ADC for bridging cytoreduction before cell infusion, an organoid-derived drug screen to select a second-line regimen after relapse, genomic surveillance to track emerging resistance mutations, and AI-assisted imaging to monitor response. Precision oncology in this sense is less a single breakthrough than an operating system: a framework in which molecular profiling, laboratory avatars, engineered therapeutics, and computational decision support are woven into a continuous loop of diagnosis, treatment, measurement, and adaptation. Fan writes that the journal’s mission going forward is to serve as an international platform for exactly this kind of cross-disciplinary exchange, covering cancer prevention, drug development, innovative therapies, and the optimization of clinical standards and protocols.

The editorial closes with the aspiration that these collective efforts will ultimately contribute to alleviating the global burden of cancer, and the early signals are encouraging. Journal submissions doubling in a single year reflects a field producing results at remarkable speed, and the concrete scientific agenda it lays out — TCR-T cells reaching intracellular targets, multi-specific ADCs defeating antigen escape, organoids personalizing drug selection, nanotechnology delivering payloads with precision, and AI knitting the data together — describes a coherent pathway from laboratory insight to clinical benefit. The breakthroughs are no longer hypothetical; they are in trials, in journals, and increasingly, in clinics. What remains is the hard, unglamorous work of validation, standardization, and equitable access that turns promising biology into dependable medicine for every patient who needs it.

Subject of Research: Emerging technologies in clinical oncology, including cell therapies, antibody-drug conjugates, cancer organoids, nanotechnology, and AI-driven precision medicine

Article Title: The coming breakthroughs in clinical oncology

Article References: Fan, J. (2026). The coming breakthroughs in clinical oncology. Clinical Cancer Bulletin, 5(1), Article 3. https://doi.org/10.1007/s44272-026-00056-4

Image Credits: AI Generated

DOI: 10.1007/s44272-026-00056-4

Keywords: clinical oncology, CAR-T cell therapy, TCR-T cell therapy, cancer vaccines, antibody-drug conjugates, cancer organoids, cancer nanotechnology, multi-omics, artificial intelligence, biomarkers, immunotherapy, precision medicine