Research Signal – Source date 2021-01-19
By January 2021, two CAR T cell therapies had US approval for certain B cell cancers, yet the same treatment approach had not entered clinical practice for solid tumors. The contrast reflects more than a difference in cancer location. In a solid tumor, engineered immune cells must find a suitable target, reach the cancer cells, and keep working amid local conditions that can suppress them. A January 2021 review describes design ideas aimed at those distinct problems, alongside the safety challenges seen with CAR T treatment.
CAR T cells are T cells genetically engineered to carry a chimeric antigen receptor, or CAR. The receptor gives a T cell a new way to recognize a chosen molecule on a cancer cell. Its outer portion binds the target; inner signaling components activate the T cell after that recognition. Changing those components can affect how strongly the cell responds and how long it persists, so design involves balancing activity with durability and safety.
Persistence became part of the design problem
The earliest CARs paired an antibody-derived recognition component with a single internal activation signal called CD3ζ. They could prompt a response to the target, but the modified cells did not persist well in patients, limiting clinical benefit. Second-generation CARs added a costimulatory signal, commonly from CD28 or 4-1BB, alongside CD3ζ. The added signal improved cell function and persistence in the body compared with first-generation designs.
Third-generation receptors combined two costimulatory signals with CD3ζ, with the aim of strengthening the response further. The review describes a clinical study in which second- and third-generation CD19-targeting cells were given to patients with B cell non-Hodgkin lymphoma; the third-generation cells showed greater expansion and persistence. But the clinical advantage of third-generation designs over second-generation CARs remained unclear. Adding another signal was not, by itself, proof of better outcomes.
More activity can bring serious toxicity
When CAR T cells engage cancer, immune signals called cytokines can spill into the body. A strong systemic reaction can cause cytokine release syndrome, which may include fever, low blood oxygen, and low blood pressure, and can progress to organ dysfunction. Neurological toxicity, known as immune effector cell-associated neurotoxicity syndrome, can affect speech, awareness, thinking, movement, or cause seizures and brain swelling. Both complications can be serious and, in some cases, fatal.
One possible way to alter the balance was to change the receptor’s hinge and membrane-spanning portion as well as its target-binding component. In a phase I study of 20 people with large B cell lymphoma, a humanized anti-CD19 CAR using hinge and transmembrane regions from CD8α was associated with ongoing complete remission in 40% of participants and severe neurological toxicity in 5%. The review compared that toxicity figure with 32% in the separate ZUMA-1 trial, which used an earlier CAR construct. The studies were not a direct controlled comparison, so those percentages alone cannot show that one design was safer. They do illustrate how receptor choices were being investigated alongside efforts to preserve antitumor activity.
Choosing a target means weighing healthy tissue
B cell cancers offered a relatively workable target: the CD19 molecule on B cells. Solid tumors present a harder selection problem. A marker may appear on cancer cells but also at low levels on healthy tissue, creating the risk that CAR T cells damage normal cells while attacking the tumor. At the same time, not every cancer cell in a solid tumor necessarily displays the target. Cells that do not may escape recognition.
The review describes a preclinical design that divided recognition across two markers, PSMA and PSCA. One CAR supplied a costimulatory signal and the other supplied the main activation signal. In the reported model, the cells responded when both markers were present, while reducing reactivity toward healthy tissue expressing either marker alone. This approach makes the design goal concrete: combine signals so that recognition depends on a more selective pattern. It remains a strategy tested in a model, not an established way to prevent harm in patients.
Uneven target expression also creates an escape route. One proposed response was to engineer CAR T cells to release a bispecific T cell engager, or BiTE: a molecule that can connect a T cell with a second tumor marker. In preclinical models, CAR T cells secreting BiTEs helped address antigen escape in B cell leukemia and marker variation in a glioblastoma model. The approach aims to broaden which cancer cells can be engaged even when they do not all carry the CAR’s original target.
Reaching the tumor takes more than finding its marker
A solid tumor can be physically difficult for immune cells to enter. Chemokines are local signals that help guide cells toward tissue. In xenograft models—mice carrying implanted tumors—adding CCR2 improved CAR T cell trafficking and tumor infiltration, while adding CXCR2 improved tumor control compared with conventional CAR T cells. These experiments suggest that chemokine receptors can help address access, but tumor control alone does not show which step improved.
The tissue surrounding cancer cells can add another barrier. In a neuroblastoma xenograft model, CAR T cells engineered to produce heparanase, an enzyme that breaks down part of the surrounding extracellular matrix, showed improved tumor infiltration and antitumor activity. This is a different design problem from target selection: even a cell that recognizes cancer may not reach it in sufficient numbers if the route through surrounding tissue is difficult.
The tumor environment can switch down a cell’s response
After entering a tumor, CAR T cells may encounter inhibitory signals and immune-suppressing cells. One such signal involves PD-1, a receptor that can dampen T cell activity when engaged. In a mesothelioma xenograft model, CAR T cells engineered with a shortened, signaling-disabled form of PD-1 had improved antitumor activity. The modification was intended to keep the inhibitory signal from being passed into the engineered cell.
Other experiments tried to change the local immune environment. In preclinical pancreatic and lung tumor models, CAR T cells engineered to release the immune messenger IL-18 showed greater antitumor activity. The tumors also contained more immune cells associated with antitumor responses and fewer cells associated with suppression. That suggests a way to change the conditions around a tumor, though these models do not show whether doing so would help patients safely.
What could make solid tumors treatable with CAR T cells
If researchers can improve target selectivity, tumor entry, and resistance to local suppression without adding unacceptable toxicity, CAR T cells could become useful against some solid tumors as well as blood cancers. That would extend a treatment approach already producing clinical benefit in certain B cell malignancies to cancers that pose different physical and biological obstacles. In January 2021, however, the review’s solid-tumor examples were preclinical, while its clinical evidence came from blood-cancer settings and a small phase I construct study. Animal-model tumor control does not establish patient benefit, and the reviewed designs do not show that one receptor change can solve targeting, access, suppression, and safety together.
Key Sources
- Primary source: Insight into next-generation CAR therapeutics: designing CAR T cells to improve clinical outcomes.
DOI or Journal Page · Review or Meta-Analysis · 2021-01-19
Keep exploring