Compare immune-cell recognition systems
Recognition system, cell source and engineering method are separate choices. Use this comparison early for vocabulary, then as the final lesson of the engineered-cell guide.
Before you start: A T-cell receptor (TCR) reads peptide plus human leukocyte antigen (HLA). A chimeric antigen receptor (CAR) usually reads an accessible surface antigen.
Where this step sits
The guide has separated target, delivery, behavior and safety. The family names now describe combinations of those features. They do not rank therapies for a particular patient.
What the families read
| Family | Recognition | Distinct question |
|---|---|---|
| CAR-T | An engineered receptor, commonly binding a surface antigen | Is the target safely accessible across enough tumor cells? |
| Engineered TCR-T | A selected receptor for a peptide–HLA complex | Does the tumor retain the exact presentation requirement? |
| Tumor-infiltrating lymphocyte (TIL) therapy | Expanded cells harvested from tumor, using native receptors | Which cells are tumor-reactive, and can they be expanded usefully? |
| CAR natural killer (NK) cells | Engineered recognition plus endogenous NK signals | How do target binding, inhibitory inputs and cell persistence interact? |
| Gamma-delta T cells, with or without a CAR | Diverse stress, metabolic or nonclassical recognition systems | Which recognition mechanism does the actual product use? |
Conventional alpha-beta TCR approaches depend on their presenting HLA. CAR recognition often avoids that particular requirement. NK and gamma-delta biology cannot be reduced to a single universal “HLA-independent” switch. Natural killer T cells are also a distinct family, rather than another name for NK cells.
A retained allele is different from a shared defect
Losing one HLA allele may remove targets restricted to it while leaving other alleles usable. Losing functional beta-2 microglobulin (B2M) can impair many classical class-I complexes. Class-II display has different dependencies.
A surface-target CAR can avoid the affected display route, but still requires a safe surface target and competent effector cells. Missing-self signals can influence NK activity; they do not guarantee killing. Receptor choice describes one dependency, not a complete treatment solution.
Target and cell platform are different axes
The same accessible antigen can be pursued with a CAR in different cell types, an antibody or an antibody-linked payload. Those formats differ in distribution, amplification, persistence and toxicity. A useful target does not establish which format has the best clinical evidence.
Conversely, the same cell platform can receive different receptors. Each new target introduces its own healthy-tissue distribution, recognition and validation questions. One successful product cannot validate every target on that platform.
Worked example
A fictional tumor loses HLA allele A while retaining allele B. A TCR restricted to A loses a required route. A receptor restricted to B may retain its route if the peptide is still processed and displayed. A surface CAR has a different recognition requirement, but its target and safety have not yet been established by the HLA result.
What can go wrong in this comparison
Family-wide safety rankings can hide construct, dose, conditioning and population differences. A list of plausible branches can also look like an access menu. Mechanistic suitability, human evidence and named eligibility criteria must remain separate.
Try it
A cell is harvested from a tumor. Does that prove its receptor recognizes cancer?
Answer: No. Tumor tissue can contain bystander cells. Specificity, useful function and manufacturing suitability still need evidence.
Explain it back
“The family name tells me ___, but suitability still needs ___.” One answer: “a recognition and cell strategy; target, safety, clinical evidence and access assessment.”
Takeaway
Compare dependencies and measured evidence before drawing a treatment conclusion from a family name.
Next: Return to the guide map or read the clinical evidence.
Sources and scope
Source check: October 8, 2026; expert and learner review pending. This comparison contains no case-specific ranking or approval inventory.
- NCI T-cell transfer therapy.
- FDA CAR T-cell product-development guidance.
- Gamma-delta biology lesson — specialist reading with scoped sources.