Chimeric antigen receptor T-cell therapy has transformed treatment for several blood cancers by genetically programming a patient’s T cells to recognize and destroy malignant cells. Yet the same approach has produced much less impressive results in solid tumors. According to a meta-analysis cited by the authors, CAR-T therapy has achieved a pooled objective response rate of approximately 9 percent across solid malignancies. The disparity reflects the fundamentally different biology of solid cancers. Unlike many blood cancers, solid tumors form dense physical structures, contain abnormal and poorly organized blood vessels, display patchy antigen expression and create local conditions that can disable incoming immune cells.

To reach and attack a solid tumor, CAR-T cells must first leave the bloodstream and cross abnormal vasculature and dense extracellular matrix. Once inside, they encounter low oxygen, limited glucose and amino acids, high concentrations of lactic acid and suppressive metabolites. Tumor-associated macrophages, regulatory T cells, myeloid-derived suppressor cells and cancer-associated fibroblasts further reinforce immune resistance. Signals such as transforming growth factor beta can inhibit T-cell activity, while persistent exposure to tumor antigen can push CAR-T cells into exhaustion. This state is marked by altered transcriptional and epigenetic programs, reduced cytokine production, impaired proliferation and declining cytotoxicity.

The review highlights a growing shift from maximizing activation toward building cellular resilience. Earlier CAR designs often focused on stronger intracellular signaling and costimulatory domains intended to produce rapid T-cell expansion. In solid tumors, however, excessive stimulation can accelerate exhaustion. Newer strategies attempt to preserve function over time by modifying the metabolic, epigenetic and signaling systems that regulate T-cell fitness. The authors discuss c-Jun overexpression as one method of restoring AP-1-dependent transcription, a pathway involved in T-cell activation and persistence. They also examine disruption of DNMT3A, an epigenetic regulator associated with the establishment of exhaustion-related cellular states.

Additional forms of “armoring” are designed to help CAR-T cells withstand suppression after they enter the tumor. Cytokine-armored cells may be engineered to produce or respond more effectively to interleukins such as IL-10, IL-15, IL-18 or IL-21, each of which can influence survival, proliferation or effector function in different ways. Other designs interfere directly with inhibitory signals. A dominant-negative TGF-β receptor, for example, can bind suppressive cues without transmitting the full inhibitory signal into the T cell. This principle is being explored in GPC3-targeted C-CAR031 for hepatocellular carcinoma, which has reportedly produced objective response rates of approximately 50–57 percent in early clinical reports. The authors caution that these findings remain based on conference abstracts pending full peer-reviewed publication.

Getting engineered T cells to the tumor is another major challenge. Many solid tumors secrete chemokines that do not match the receptors naturally expressed by circulating T cells, leaving therapeutic cells poorly recruited to the cancer site. Adding receptors such as CCR2b may improve recognition of tumor-associated chemokine gradients and increase infiltration. Researchers are also developing hypoxia-responsive CAR systems that use the low-oxygen conditions found inside tumors as a biological switch, helping restrict activation to the tumor microenvironment. These approaches could improve both delivery and safety by reducing activity in healthy tissues where the target antigen may be present at lower levels.

Synthetic biology is adding another layer of control. SynNotch systems use one receptor to detect an initial antigen and trigger production of a second CAR, creating a sequential activation process. Tmod “NOT-gate” designs are intended to activate against malignant cells while suppressing responses to healthy cells that carry a protective antigen. Drug-controlled CARs offer yet another safety mechanism, allowing clinicians to regulate T-cell activity with an externally administered compound. Together, these circuits seek to address antigen heterogeneity, one of the defining problems of solid tumors, in which not every cancer cell displays the same target and antigen loss can allow resistant clones to survive.

The review points to several clinical developments suggesting that these principles are beginning to translate into meaningful patient outcomes. In H3K27M-mutated diffuse midline glioma, intracerebroventricular administration of GD2-targeted CAR-T cells produced substantial tumor reductions, including a complete response that was sustained beyond 30 months. In advanced gastric cancer, the CLDN18.2-targeted therapy satricabtagene autoleucel, also known as satri-cel, was reported to outperform physician’s choice in a randomized Phase 2 trial. The treatment produced a progression-free survival hazard ratio of 0.37 and an overall survival hazard ratio of 0.69, results the authors describe as the first randomized evidence of CAR-T superiority over standard treatment in a solid malignancy.

The authors stress that the engineered cell is only one part of the therapeutic system. Conditioning chemotherapy, the phenotype of T cells at infusion, manufacturing time, the patient’s gut microbiome and systemic neuroendocrine signals may all influence whether CAR-T cells persist and remain functional. Manufacturing is also undergoing rapid change, with next-day production methods and experimental technologies designed to generate CAR-T cells directly inside the patient using targeted lipid nanoparticles or receptor-targeted lentiviral particles. These approaches could reduce production delays, infrastructure requirements and treatment costs, although their safety, regulatory status and clinical feasibility remain under investigation. The review ultimately calls for an integrated development model that combines delivery, resilience, logic and whole-patient biology. It also warns that long-term genomic stability after multiplex gene editing, the safety of sustained cytokine production and the complexity of increasingly sophisticated cell products must be established through larger studies, longer follow-up and prospective biomarker-guided trials. The central message is that solid-tumor CAR-T therapy may advance not through a single breakthrough, but through the coordinated solution of many biological problems that once appeared inseparable.

Subject of Research: Not applicable

Article Title: Engineering CAR-T cells for solid tumors: Overcoming the microenvironment through integrated design and clinical translation

News Publication Date: August 11, 2026

Web References: https://doi.org/10.18632/oncoscience.666

References: Oncoscience, Volume 13; Figure 6: https://www.oncoscience.us/article/666/text/#F6

Image Credits: Copyright © 2026 Onyekweli et al.; distributed under the Creative Commons Attribution License (CC BY 4.0).

Keywords: CAR-T cell therapy, solid tumors, tumor microenvironment, cancer immunotherapy, immunotherapy engineering, synthetic biology, T-cell exhaustion, clinical translation, oncology, cellular therapy

The work addresses a long-standing challenge in cancer therapy: how to make powerful treatments concentrate inside tumors while limiting damage to healthy tissues. Attenuated strains of Salmonella have attracted interest because they can preferentially accumulate in the abnormal environment of solid tumors. Tumors often contain regions with poor oxygen levels, disorganized blood vessels and local immune suppression, conditions that can support bacterial growth. Once inside these sites, therapeutic Salmonella can act as a biological delivery platform and stimulate immune responses against malignant cells.

The researchers focused on calreticulin, a protein normally found inside the endoplasmic reticulum, where it helps regulate calcium storage and protein folding. Under cellular stress, including stress caused by chemotherapy, radiation or other anticancer treatments, calreticulin can move to the outer surface of a cancer cell. There, it functions as an “eat-me” signal, alerting immune cells that the damaged cell should be engulfed. Because surface-exposed calreticulin is associated with immunogenic forms of cell death, it provides a potential molecular address for directing therapeutic agents toward stressed tumor cells.

The experimental construct combines this targeting concept with L-asparaginase, an enzyme already used in clinical oncology, especially in the treatment of acute lymphoblastic leukemia. L-asparaginase breaks down circulating L-asparagine into aspartic acid and ammonia. Some cancer cells, particularly those with limited capacity to synthesize their own asparagine, depend heavily on the amino acid supplied through the bloodstream. Depleting extracellular asparagine can therefore interrupt protein production, trigger metabolic stress and promote cancer-cell death. The enzyme’s effectiveness, however, can be limited by immune reactions, pharmacological instability and toxicity, making targeted delivery an important goal.

The second component, flagellin, is the structural protein that forms the filament of bacterial flagella. It is also a potent molecular signal for the innate immune system. Immune cells recognize flagellin primarily through Toll-like receptor 5, while intracellular sensing pathways can activate inflammasome components such as NLRC4. These signals can promote the release of inflammatory mediators, stimulate antigen-presenting cells and help convert an immunologically “cold” tumor into one more visible to the immune system. By incorporating flagellin into the therapeutic design, the researchers sought to make the treatment not only directly toxic to tumor cells but also capable of amplifying antitumor immunity.

The study’s central finding was that the calreticulin-targeting L-asparaginase–flagellin conjugate strengthened the antitumor effects of Salmonella-based therapy. Rather than relying on a single mechanism, the treatment brings together several forms of pressure on the tumor. Salmonella can concentrate within the tumor microenvironment, the targeting component can help associate the conjugate with calreticulin-exposing cancer cells, L-asparaginase can deprive vulnerable cells of an essential nutrient, and flagellin can activate immune surveillance. The resulting combination is designed to produce a chain reaction in which metabolic stress and immune stimulation reinforce one another.

This type of combination may be particularly valuable because tumors frequently adapt when exposed to one therapeutic pressure. A cancer cell that survives nutrient deprivation may still be eliminated if immune recognition is intensified. Likewise, an immune response that is too weak to control a tumor may become more effective when bacterial localization and enzyme-mediated damage increase the number of abnormal antigens and danger signals released by dying cells. The researchers’ findings suggest that coordinating these mechanisms can improve the performance of bacteria-assisted cancer treatment in experimental settings.

The approach also reflects a broader shift in cancer research toward programmable biological medicines. Instead of treating bacteria only as infectious threats, scientists are redesigning them as localized delivery vehicles capable of carrying enzymes, immune activators or molecular probes. The advantage is spatial: a therapeutic payload can be produced or concentrated near the tumor rather than distributed uniformly throughout the body. The challenge is equally significant. Any clinical version would need precise control over bacterial attenuation, immune activation, enzyme exposure and potential inflammation, while also demonstrating reliable performance across genetically diverse tumors.

Calreticulin targeting may provide a useful way to address some of that complexity because the protein’s appearance on the cell surface is linked to cellular stress and treatment response. However, the extent and duration of calreticulin exposure can vary between tumor types and individual patients. Future studies will need to determine which cancers are most suitable for this strategy, how calreticulin levels predict treatment response and whether the conjugate can be combined safely with established immunotherapies such as immune-checkpoint inhibitors. Researchers will also need to assess pharmacology, manufacturing consistency and the possibility of immune reactions against the bacterial or enzymatic components.

The findings position the engineered conjugate as a promising experimental platform rather than an immediately available therapy. By merging tumor-homing bacteria with a calreticulin-directed enzyme and an innate immune stimulant, the study illustrates how cancer treatments can be designed to attack malignant cells on multiple biological fronts. If the results are confirmed in further preclinical testing and eventually in carefully controlled clinical trials, this strategy could help turn Salmonella from a passive carrier into an active, multifunctional partner in cancer immunotherapy.

Subject of Research: Calreticulin-targeted L-asparaginase–flagellin conjugate used with Salmonella-mediated cancer therapy.

Article Title: Calreticulin-targeting L-asparaginase-flagellin conjugate enhances Salmonella-mediated antitumor efficacy.

Article References: Nguyen, DH., Afzal, A.R., Nguyen, P.TM. et al. Calreticulin-targeting L-asparaginase-flagellin conjugate enhances Salmonella-mediated antitumor efficacy. Cell Death Discov. (2026). https://doi.org/10.1038/s41420-026-03300-x

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s41420-026-03300-x

Keywords: cancer immunotherapy, Salmonella, calreticulin, L-asparaginase, flagellin, tumor targeting, bacterial therapy, antitumor efficacy, immunogenic cell death