The first and most fundamental barrier is antigen heterogeneity. CAR-T cells work by recognizing a single molecular target on the surface of cancer cells, then killing anything that displays it. In blood cancers, that target tends to be expressed uniformly, so a single receptor can wipe out the malignant population. Breast tumors are different. Within a single patient, cancer cells can display wildly different combinations of surface proteins, and the cells that lack the chosen target simply survive, multiply and drive relapse. Tumors can also shed or downregulate the target altogether, a process known as antigen escape. Compounding the problem is on-target, off-tumor toxicity: many candidate antigens appear at low levels on healthy tissues, meaning a receptor designed to attack the tumor can also attack vital organs.

The second barrier is subtype adaptation. Breast cancer is not one disease but a family of molecularly distinct diseases, including luminal A and luminal B tumors driven by hormone receptors, HER2-positive tumors amplifying a growth-factor receptor, and triple-negative breast cancer, the aggressive subtype that lacks all three standard markers. Each subtype carries its own immune landscape, its own candidate antigens and its own vulnerabilities, yet the review notes that research effort has been unevenly distributed, with triple-negative and HER2-positive disease attracting far more CAR-T attention than the luminal subtypes that make up the majority of cases. A receptor strategy optimized for one subtype cannot simply be transplanted to another.

The third barrier is the tumor microenvironment, the dense, hostile territory that solid tumors build around themselves. Breast tumors recruit immunosuppressive cells such as regulatory T cells, tumor-associated macrophages and myeloid-derived suppressor cells, which actively disable incoming immune attacks. They erect physical barriers, including abnormal vasculature, stiff extracellular matrix and high interstitial pressure, that prevent CAR-T cells from infiltrating deeply. They also impose metabolic constraints, depleting the nutrients and oxygen that T cells need to sustain their killing function, and they drench the surroundings in checkpoint molecules such as PD-L1 that switch activated T cells off. A CAR-T cell that reaches a breast tumor, the review emphasizes, often arrives exhausted, starved and surrounded by enemies.

Against these three barriers, the authors lay out three interconnected optimization strategies. The first is subtype-specific precision CAR design: selecting targets matched to each molecular subtype, such as HER2 for HER2-positive disease or antigens like mesothelin, MUC1 and others for triple-negative tumors, and validating that expression is strong and uniform on tumor cells while minimal on healthy tissue. This precision approach acknowledges that the future of breast cancer CAR-T lies not in a universal receptor but in a portfolio of subtype-tailored constructs, each grounded in the biology of the disease it is meant to treat.

The second strategy is dual-target CAR engineering, a family of techniques designed to defeat heterogeneity and escape. Tandem CARs carry two recognition domains on a single receptor, allowing one cell to sense two antigens at once. Parallel CARs express two independent receptors, so the cell can attack through either target. The most sophisticated variant is logic-gated control using SynNotch receptors, synthetic circuitry that requires one antigen to prime the cell before a second receptor can trigger killing. This AND-gate logic means the T cell fires only when it encounters the precise combination of markers found on tumor cells, dramatically reducing the risk of destroying healthy tissue that carries just one of the targets. By demanding multiple molecular confirmations before attacking, these designs make it far harder for tumors to escape by losing a single antigen.

The third strategy is multi-dimensional remodeling of the tumor microenvironment. The review describes approaches that range from clearing immunosuppressive cells, for example by depleting regulatory T cells or reprogramming suppressive macrophages, to physically and metabolically reshaping the tumor so that engineered cells can enter and function. CAR-T cells can themselves be armored with cytokines or engineered to resist exhaustion, and they can be combined with immune checkpoint inhibitors that neutralize the PD-L1 shield tumors deploy. The unifying idea is that CAR-T cells should not be sent into battle alone; they need the battlefield altered in their favor, whether through drugs, radiation, engineered secretion of local agents or genetic modifications that make the cells self-sustaining in hostile conditions.

Where does the clinic stand? The review summarizes early clinical trial experience showing manageable safety and preliminary efficacy in patients with HER2-positive and triple-negative breast cancer, the two subtypes where trials have concentrated. These results fall short of the spectacular complete-remission rates seen in blood cancers, but they demonstrate that CAR-T cells can be manufactured, infused and tolerated in solid tumor patients, and that some patients derive meaningful benefit. The authors identify three key translation challenges that now dominate the field: safety management, since solid-tumor targets raise the stakes of off-tumor toxicity; manufacturing scalability, since producing personalized cell products for each patient is slow and expensive; and efficacy durability, since responses in solid tumors have so far tended to be shorter-lived than those in hematologic disease. Each challenge, they argue, has targeted coping strategies, from safer receptor designs and kill-switch safeguards to streamlined manufacturing platforms and combination regimens that extend cell persistence.

Looking further ahead, the review surveys a wave of next-generation technologies that could reshape the field. CAR-NK cells, built from natural killer cells, offer the prospect of off-the-shelf products with a lower risk of graft-versus-host complications. CAR-macrophages, or CAR-M cells, are designed to penetrate solid tumors more effectively and to remodel the microenvironment from within. Nanobody-mediated CAR-T cells use single-domain antibodies, smaller and more stable than conventional antibody fragments, to build receptors with novel recognition properties. Perhaps most transformative is in vivo CAR engineering, which would skip the laboratory manufacturing step entirely by delivering genetic instructions directly into a patient’s own T cells inside the body. Alongside these biological advances, artificial intelligence is emerging as a design tool, with computational models used to predict optimal antigen targets, engineer receptor sequences and tailor treatments to individual tumor profiles, moving the field toward genuinely personalized immunotherapy.

The overall conclusion is cautiously ambitious. CAR-T therapy has evolved, in the authors’ assessment, into a clinically viable immunotherapeutic strategy for breast cancer, with significant progress in addressing its core bottlenecks through precision engineering, combination therapies and technological innovation. The convergence of synthetic biology, computational immunology and precision oncology, they write, lays a solid foundation for further optimization, and the priorities ahead are clear: deepen subtype-targeted therapy, upgrade CAR engineering and manufacturing platforms, optimize combination regimens and advance clinical translation so that access widens beyond specialized centers. Remaining hurdles in manufacturing, durability and affordability are real, and the authors are candid that the therapy is not yet standard care. But with coordinated effort across academia, industry and regulatory bodies, they argue, CAR-T therapy could eventually become a standard option for breast cancer, particularly for aggressive subtypes such as triple-negative disease where conventional options remain limited, and could deliver durable remissions and improved survival for patients worldwide. For a field that has spent a decade chasing the solid-tumor breakthrough, the review reads less like a eulogy for failed trials and more like an engineering roadmap, one that treats each biological barrier not as a dead end but as a design problem waiting for the right solution.

Subject of Research: Chimeric antigen receptor T-cell therapy for breast cancer

Article Title: Chimeric antigen receptor T-Cell therapy for breast cancer: current status, challenges and future perspectives

Article References: Chimeric antigen receptor T-Cell therapy for breast cancer: current status, challenges and future perspectives. (n.d.). https://doi.org/10.1186/s12967-026-08881-w

Image Credits: AI Generated

DOI: 10.1186/s12967-026-08881-w

Keywords: CAR-T cell therapy, breast cancer, tumor microenvironment, triple-negative breast cancer, HER2, antigen heterogeneity, dual-target CAR engineering, SynNotch, immunotherapy, CAR-NK, clinical translation, precision oncology

The rationale begins with the specific barriers that defeat CAR cells in solid malignancies. Target antigens on solid tumors display marked spatial and temporal heterogeneity, meaning that different regions of the same tumor, and different stages of its evolution, present different molecular surfaces. When CAR cells exert continuous selective pressure, low-antigen or antigen-negative subpopulations can become enriched, a phenomenon known as antigen escape. Physical barriers compound the problem: abnormal vasculature, dense extracellular matrix, and complex tissue architecture impede the infiltration of transferred cells. Even after CAR cells gain entry, the tumor microenvironment confronts them with hypoxia, nutrient deprivation, immunosuppressive cytokines such as TGF-β, immune checkpoints, and tumor-associated myeloid cells, all of which drive functional exhaustion. The review notes that these challenges are particularly pronounced in glioblastoma (GBM), one of the most immunologically hostile solid tumors.

Oncolytic viruses offer a way to attack these barriers from within. These engineered viruses selectively infect and replicate inside tumor cells, triggering immunogenic cell death. The lytic cycle releases a potent array of tumor-associated antigens along with pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs), including ATP, HMGB1, calreticulin, and double-stranded DNA and RNA. These danger signals activate local antigen-presenting cells and induce robust inflammatory cytokine and chemokine gradients, recruiting dendritic cells, macrophages, natural killer cells, and endogenous T cells. In effect, viral oncolysis converts a cold, suppressive tumor microenvironment into an active site of immune amplification. The review emphasizes that this immune-remodeling function, rather than direct tumor cell killing alone, is where the true value of OVs lies for combination design.

Central to the authors’ argument is the concept of a self-amplifying circuit that links viral lysis to CAR-mediated killing. As they describe it, OV infection leads to tumor lysis and the release of antigens and danger signals, which drive immune activation and CAR cell recruitment and infiltration, followed by CAR-mediated killing and further antigen release, creating a positive feedback loop. Each round of destruction supplies fresh targets and inflammatory context for the next. This dynamic, multistep engineered circuit is what distinguishes the OV–CAR platform from simple additive drug combinations, in which two agents act through independent mechanisms without reinforcing one another. The tumor becomes, in the authors’ framing, a site where the therapy continuously regenerates its own momentum.

One of the most striking engineering strategies addresses the problem of missing or heterogeneous CAR targets directly. OVs can be engineered to deliver payloads that force infected tumor cells to express new surface antigens, such as truncated CD19, a molecule normally associated with B-cell malignancies. This temporarily converts CD19-negative solid tumor cells into CD19-positive targets recognizable by CAR-T cells. The elegance of the approach lies in how it separates two distinct problems: the virus determines where to install the target, while the CAR determines how to kill the marked cell. Antigen installation becomes a programmable function of the viral module rather than a fixed property of the tumor, fundamentally changing the calculus of target selection in solid tumors.

Building on this concept, the City of Hope team constructed an engineered OV, designated OVDual, capable of delivering both CD19 and EGFRvIII to GBM cells, and designed bispecific CAR-T and CAR-NK cells that recognize both targets simultaneously. This dual-module design tackles two issues at once: the virus expands the visible range of the tumor by installing antigens on infected cells, while the bispecific CAR reduces dependence on any single antigen, blunting the risk of escape through loss of one target. For a tumor as heterogeneous as glioblastoma, where antigen expression varies dramatically across regions and over time, this layered redundancy is particularly relevant. The strategy illustrates how viral and cellular modules can be matched to one another rather than selected independently.

Target recognition, however, is only half the battle; maintaining CAR cell function inside the tumor is another critical challenge. OVs can be engineered to express immunomodulatory factors locally, including IL-7, IL-12, IL-15, IL-21, CCL5, CXCL9, CXCL10, and CXCL11, converting the tumor itself into a local support factory for transferred cells. In the team’s multimodal GBM study, an OV expressing membrane-bound IL-15 and IL-21, termed OV mIL15/21, provided localized cytokine support to sustain CAR-T and CAR-NK survival, expansion, and cytotoxicity. The complete therapeutic system can therefore be understood as a division of labor: OVDual addresses who cannot be seen, the bispecific CAR addresses who is recognized, and OV mIL15/21 addresses how to maintain function under hostile conditions.

The review also highlights CAR-NK cells as an emerging partner for viral platforms. Unlike CAR-T cells, CAR-NK cells can kill through both CAR-mediated targeting and natural NK receptor-mediated recognition, providing an additional, antigen-independent killing mechanism. Off-the-shelf CAR-NK products derived from induced pluripotent stem cell platforms offer standardized, scalable manufacturing and rapid administration, advantages that align well with the logistics of viral combination therapy. In this configuration, OVs remodel the tumor, CAR-NK cells execute killing through parallel pathways, and locally delivered cytokines sustain the cellular component over time. The convergence of scalable NK manufacturing and programmable viral payloads may prove one of the most practically translatable directions in the field.

Conceptually, the authors describe the field as undergoing a transition from combination therapy to systems engineering. Early approaches treated OV plus CAR-T as the pairing of two mechanistically distinct treatments. Next-generation designs instead resemble a multiplicative architecture of virus, CAR-T or CAR-NK cells, immune modulation, and delivery, with different modules addressing distinct bottlenecks. This implies that future OV–CAR therapy may never settle on a single universal standard combination. Instead, clinicians would identify the dominant limiting factor for a given tumor, whether antigen heterogeneity, poor infiltration, or functional exhaustion, and select the virus and cell modules that specifically address that limitation. Therapy design becomes diagnostic and modular rather than one-size-fits-all.

Significant questions remain before broad clinical application. Viral delivery and in vivo dissemination, treatment sequencing, and safety all require rigorous answers. Future trials, the authors stress, must go beyond measuring whether a treatment works and ask whether the virus truly infects the tumor, whether the payload is successfully expressed, whether CAR cells genuinely enter and persist within the lesion, and which module is actually contributing to the effect. Answering these questions will require multidimensional biomarkers, including longitudinal biopsies, circulating tumor DNA, viral DNA and RNA detection, cytokine profiling, single-cell sequencing, and spatial transcriptomics. Looking further ahead, the platform may integrate multi-target CARs, synthetic Notch receptors, logic-gated CARs, off-the-shelf CAR-NK cells, and more precise viral engineering, while non-replicating technologies such as mRNA and lipid nanoparticles may complement viral approaches as pseudo-viral immune modulators. The authors’ conclusion is that the future of solid tumor immunotherapy may lie not in a single therapy that solves every problem, but in programmable systems in which viruses reshape the battlefield and CAR cells execute precise, self-reinforcing attacks. A patent on a multimodal oncolytic virus cancer therapy has been filed by City of Hope, and the work was supported in part by the National Cancer Institute under award number P30CA33572.

Subject of Research: Combining oncolytic viruses with CAR-T and CAR-NK cell therapy to overcome resistance barriers in solid tumors

Article Title: Oncolytic viruses and car cells: A modular platform to overcome solid tumor barriers

Article References: Oncolytic viruses and car cells: A modular platform to overcome solid tumor barriers. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: oncolytic viruses, CAR-T cells, CAR-NK cells, solid tumors, glioblastoma, tumor microenvironment, antigen heterogeneity, immunotherapy, immunogenic cell death, cytokines, bispecific CAR, systems engineering