CAR T Solid Tumors: Breakthrough Strategies for Immune Targeting

CAR T-cell therapy has transformed treatment for certain blood cancers, but applying the same approach to solid tumors has proven far more difficult. The obstacles are layered: solid tumors hide behind physical barriers, suppress immune cells that manage to arrive, and present a patchwork of surface markers that make clean targeting unreliable. Researchers are now attacking each of these problems with increasingly creative engineering strategies, from rewiring the chemical signals that guide T cells toward tumors to building genetic logic gates that let engineered cells distinguish cancer from healthy tissue. No single fix has cracked the problem, but the convergence of several approaches is producing the most promising results the field has seen.

Why Solid Tumors Are So Much Harder

When CAR T cells go after leukemia or lymphoma, they benefit from a relatively straightforward setup. The cancer cells float freely in blood and bone marrow, they are physically accessible, and they carry well-defined surface proteins that healthy tissue can afford to lose. Solid tumors present a fundamentally different challenge. The tumor mass sits behind layers of dense connective tissue, bathed in an environment actively hostile to immune cells. The cells within a single tumor can differ dramatically from one another, meaning a target that appears on some cancer cells may be absent on others.

These barriers break down into a few distinct categories. First, there is no solid-tumor equivalent of CD19, the near-universal marker on B-cell cancers that made the first CAR T therapies possible. Most candidate targets on solid tumors also appear on some normal tissues, creating a risk of damaging healthy organs. Second, the tumor microenvironment suppresses immune cells through chemical signals, nutrient starvation, and recruitment of regulatory immune cells that dial down the attack. Third, even when CAR T cells are infused into the bloodstream, relatively few manage to physically reach and penetrate the tumor mass.

¹ These interlocking problems explain why, despite dozens of clinical trials, solid-tumor CAR T therapy has not yet produced the dramatic remission rates seen in blood cancers.²

Getting CAR T Cells to the Tumor

One of the most basic problems is transportation. After an intravenous infusion, CAR T cells enter the bloodstream, but they lack the molecular “GPS” to efficiently home in on a solid tumor buried in, say, the pancreas or the brain. Researchers are addressing this by engineering CAR T cells to express chemokine receptors, which are surface proteins that detect chemical trails tumors inadvertently leave behind. By matching the receptor on the T cell to the chemokine the tumor secretes, the engineered cells migrate toward the cancer more reliably.

A well-studied example involves the CCL2/CCR2 axis. Many solid tumors secrete CCL2, a signaling molecule that normally attracts certain immune cells. T cells, however, express very little of the matching receptor, CCR2. When researchers engineered CAR T cells to overexpress CCR2b (the more stable of the two receptor forms), those cells showed significantly increased migration toward CCL2 gradients produced by non-small cell lung cancer cells, including in brain metastases.³ Similar work on mesothelin-targeted CAR T cells equipped with either CCR2 or CXCR6 showed improved migration, invasion, and sustained killing compared to conventional versions.⁴ The broader concept is to treat the chemokine system as a tunable navigation layer that can be matched to whatever signals a particular tumor type produces.⁵

Even once a T cell reaches the tumor’s neighborhood, it may encounter a dense physical barrier. Solid tumors are often surrounded by extracellular matrix rich in molecules like heparan sulfate proteoglycans, which act as a structural wall. One approach engineers CAR T cells to produce heparanase, an enzyme that degrades this matrix. In preclinical work, heparanase-expressing CAR T cells broke through the barrier more effectively, improving both infiltration and anti-tumor activity.⁶

Attacking the Tumor’s Blood Supply Instead

An alternative to trying to penetrate the tumor mass directly is targeting its blood vessels. Tumor blood vessels express specific markers that normal vasculature largely does not, and these markers are accessible from within the bloodstream, which sidesteps the infiltration problem entirely. CLEC14A, a glycoprotein overexpressed on the vasculature of many solid cancers, has been tested as a CAR T target. In mouse models of pancreatic cancer and lung carcinoma, CLEC14A-targeting CAR T cells significantly inhibited tumor growth and reduced vascular density within the tumor without causing toxicity in healthy mice.⁷ The appeal of this approach is its breadth: because many different solid tumors rely on similar vascular growth patterns, a single vascular target could potentially apply across cancer types.⁸

Overcoming Antigen Heterogeneity

A CAR T cell recognizes one specific protein on the tumor surface. If some cancer cells in the mass do not display that protein, or if they lose it under selective pressure from the therapy, those cells survive and the tumor can regrow. This phenomenon, called antigen escape, is one of the most persistent obstacles in solid-tumor CAR T work.⁹

Several strategies aim to reduce this risk. One involves designing CAR T cells that recognize two or more antigens simultaneously, so a tumor cell would need to lose multiple surface markers to escape detection. Another uses tandem CARs, where a single receptor binds two different targets. More sophisticated approaches use synthetic biology to create logic-gated circuits. In an AND-gate design, a synthetic receptor for one antigen triggers the expression of a CAR for a second antigen. The T cell only activates when both targets are present on the same tumor, which makes it far less likely that a rogue cell can slip through by shedding a single marker. In preclinical mouse models, these dual-receptor T cells killed tumors expressing both antigens while sparing nearby tissues expressing only one.¹⁰

Rewiring the Tumor Microenvironment

Even when CAR T cells reach the tumor in adequate numbers and find their target, the local environment works to shut them down. One of the most potent weapons tumors use is TGF-β, a signaling molecule that suppresses T-cell activity. Researchers have countered this by equipping CAR T cells with a decoy version of the TGF-β receptor. This dominant-negative receptor absorbs the suppressive signal without passing it along, essentially making the T cell deaf to one of the tumor’s primary shutdown commands. In mouse models of aggressive prostate cancer, CAR T cells carrying this decoy receptor showed increased proliferation, resistance to exhaustion, and the ability to eradicate tumors.¹¹

A more recent twist goes beyond merely blocking TGF-β and actually converts the suppressive signal into a stimulatory one. A “switch receptor” links a truncated TGF-β receptor to a stimulatory signaling domain. When TGF-β binds, instead of suppressing the T cell, the receptor triggers immune activation and increased energy production within the cell. This approach preserved the normal housekeeping functions of TGF-β signaling while flipping its immunosuppressive role into a weapon against the tumor.¹²

Beyond TGF-β, so-called “armored” CAR T cells can be engineered to secrete pro-inflammatory cytokines like IL-12 or IL-18 directly into the tumor, reshaping the local immune environment to favor attack rather than tolerance.¹³ And checkpoint blockade, already used as a standalone cancer therapy, can be built directly into the CAR T cell itself. Engineering T cells with a dominant-negative PD-1 receptor prevents the tumor from flipping the “off switch” that PD-L1 normally triggers, resulting in improved tumor clearance in preclinical solid-tumor models.¹⁴

The Metabolic Battle Inside the Tumor

Tumors are metabolically greedy. They consume enormous amounts of glucose, strip the local environment of amino acids, and generate waste products that poison immune cells. CAR T cells entering this environment are essentially being asked to fight while starving. Competition for glucose is a major factor: tumors hoard it, leaving T cells too energy-depleted to sustain their attack.¹⁵

Metabolic reprogramming strategies aim to make CAR T cells more adaptable to this hostile environment. These include genetic modifications that boost glucose uptake, pharmacological treatments applied during manufacturing to prime T cells for low-nutrient conditions, and engineering changes that optimize how T cells burn fatty acids, which become a critical alternative fuel when glucose is scarce. Other approaches target the immunosuppressive metabolites themselves, for example by enabling T cells to resist the effects of kynurenine, a byproduct of amino acid breakdown that accumulates in many tumors and dampens immune responses.¹⁶

Fighting Exhaustion

T-cell exhaustion is what happens when immune cells face prolonged stimulation without rest. They gradually lose their killing ability, stop dividing, and enter a state of functional shutdown. In the solid-tumor setting, where the fight is slow and the microenvironment constantly suppressive, exhaustion is a major reason CAR T cells peter out before finishing the job.

One of the more striking interventions targets c-Jun, a transcription factor involved in T-cell activation. Researchers found that overexpressing c-Jun in CAR T cells made them resistant to exhaustion across five different mouse tumor models, with enhanced expansion, improved killing, and reduced terminal differentiation. The work suggests that exhaustion in CAR T cells partly stems from a functional shortage of c-Jun, and that artificially boosting it can override the shutdown program.¹⁷

Safety and Precision Engineering

When a CAR T cell attacks healthy tissue that shares the tumor’s target protein, the result is called on-target, off-tumor toxicity. The severity ranges widely. In blood cancers, the loss of normal B cells (which share CD19 with the cancer) is manageable with antibody replacement therapy. In solid tumors, hitting the wrong tissue can be far more dangerous: skin reactions, gastrointestinal damage, and lung toxicity have all been reported.¹⁸ A first-in-human trial of Nectin-4-targeting CAR T cells, for example, documented skin, oral, and gastrointestinal toxicities that underscored the need for careful risk assessment even with promising targets.¹⁹

The synthetic Notch (synNotch) receptor system described earlier is one answer to this problem, but there are others. Affinity tuning adjusts how tightly the CAR grips its target, with the idea that a lower-affinity CAR may preferentially engage cancer cells that overexpress the target while ignoring normal cells that express it at lower levels.²⁰ SynNotch-based circuits have shown particular promise in preclinical models of glioma and pancreatic cancer, where they controlled both CAR expression and local cytokine release to stay precise within immunosuppressive tumor niches.²¹

As a failsafe, some designs include a kill switch. One well-tested system uses an inducible suicide gene: when a dimerizing drug is administered, it triggers the engineered T cells to self-destruct. In a clinical study of patients who developed graft-versus-host disease, a single dose of the drug eliminated more than 90% of the modified T cells within 30 minutes and resolved the condition without recurrence.²²

Delivering CAR T Cells Directly to the Tumor

Rather than relying on intravenous infusion and hoping enough cells reach the tumor, some clinical programs are experimenting with regional delivery: injecting CAR T cells directly into or near the tumor site. This approach concentrates more cells where they are needed while reducing the risk of systemic side effects.²³ Early clinical data suggest that regional delivery is safe and feasible, and that it promotes not just local infiltration and proliferation but also a broader systemic immune response, as if lighting a match at the tumor site can ignite immunity throughout the body.²⁴ Local interventions such as surgery, radiation, or ablation can also be paired with CAR T infusion to make the tumor more permeable and reduce the immune-suppressive environment before the cells are introduced.²⁵

Beyond T Cells: CAR-NK Cells and CAR Macrophages

T cells are not the only immune cells that can be given a chimeric antigen receptor. Natural killer (NK) cells, which are part of the innate immune system, have attracted significant interest as an alternative platform. CAR-NK cells carry a lower risk of graft-versus-host disease and cytokine release syndrome compared to CAR T cells, which are two of the most serious side effects associated with T-cell-based therapies. Because NK cells from a healthy donor can be used without needing to match each patient individually, CAR-NK products can potentially be manufactured as “off-the-shelf” therapies, dramatically reducing cost and production time.²⁶ This advantage is especially relevant for solid tumors, where the complex multi-step manufacturing of autologous CAR T cells can mean weeks of delay while a patient’s cancer progresses.²⁷

CAR macrophages take a different angle entirely. Macrophages naturally infiltrate solid tumors in large numbers, a property T cells lack. Researchers have engineered macrophages with CARs targeting the extracellular matrix surrounding breast tumors. These CAR macrophages degraded the dense collagen barrier around the tumor, likely through the action of matrix metalloproteinases, though they did not directly inhibit tumor-cell growth in lab dishes.²⁸ The idea is that CAR macrophages could serve as a pathfinder, stripping away the physical barriers to make the tumor vulnerable to subsequent therapies.

Tracking CAR T Cells in Real Time

One challenge that gets less public attention but matters enormously for clinical development is figuring out where the engineered cells actually go after infusion. In blood cancers, you can sample the blood. In solid tumors, you need imaging. Researchers have developed PET-based reporter gene systems that let them visualize CAR T cells non-invasively in living subjects. In one approach, CAR T cells were engineered to express a modified form of PSMA (the same protein used as a cancer marker in prostate cancer, but here repurposed as a tracking tag). Sequential PET/CT scans with a standard PSMA-targeting radiotracer were then able to detect the engineered cells accumulating at the tumor site, with signal intensity increasing over time.²⁹

Other PET reporter systems use the human sodium iodide symporter or viral thymidine kinase variants to achieve similar goals. In triple-negative breast cancer models, PET imaging using the sodium iodide symporter reporter revealed unexpected differences in how well CAR T cells persisted at different tumor sites, information that would have been invisible without the imaging platform.³⁰ PET reporter genes can also double as a safety mechanism: when paired with a prodrug-activating enzyme, the same gene that makes the T cells visible on a scan can be used to selectively destroy them if toxicity develops.³¹

Early Clinical Signals

Most of the strategies described above remain in preclinical or early-phase clinical testing, and the honest assessment is that no solid-tumor CAR T therapy has yet produced the kind of sustained complete remissions seen in blood cancers. But there are signals worth watching. In pediatric neuroblastoma, a GD2-targeting CAR T product led to stable disease at six months in six out of ten patients, with four remaining stable at one year and alive after three to four years of follow-up. Median overall survival was about 25 months.³² Claudin18.2-targeting CAR T cells for gastrointestinal cancers have also reached phase 1 final results, though the field is watching to see how durable these responses turn out to be.³³

The move toward universal (allogeneic) CAR T products could also accelerate solid-tumor programs. Gene-editing tools like CRISPR are being used to knock out the T-cell receptor and certain immune markers on donor T cells, preventing rejection and allowing a single manufacturing run to serve many patients.³⁴ If successful, this would slash the time between diagnosis and treatment, a particularly important advantage for aggressive solid cancers where weeks of manufacturing delay can be clinically meaningful.

Why Combination Is Likely the Path Forward

The field has largely moved past the idea that any single modification will unlock solid tumors the way CD19-targeting unlocked B-cell cancers. The emerging consensus is that effective solid-tumor CAR T therapy will require stacking multiple engineering solutions: chemokine receptors for homing, armoring against the microenvironment, multi-antigen targeting to prevent escape, metabolic reprogramming for persistence, and either regional delivery or vascular targeting to get enough cells into the right place. Each of these strategies addresses a distinct biological barrier, and failing to address any one of them can be enough to undermine the whole effort. The engineering complexity is daunting, but the pace of innovation across synthetic biology, gene editing, and cell manufacturing is genuinely compressing timelines that would have seemed unrealistic a decade ago.