No CAR-T cell therapy has been approved for acute myeloid leukemia (AML), and none is likely to reach FDA approval before 2028 at the earliest. While CAR-T therapies have transformed treatment for several other blood cancers since 2017, AML presents unique biological challenges that have slowed progress considerably. The good news: dozens of clinical trials are now producing early remission data, and newer engineering strategies are closing the gap.
Why AML Has Been So Much Harder
CAR-T therapy works by reprogramming a patient’s immune cells to recognize and destroy cancer. In cancers like B-cell lymphoma and acute lymphoblastic leukemia (ALL), the target is straightforward: a surface protein found on cancer cells but not on cells the body can’t live without. When CAR-T wipes out all B cells, patients can survive on antibody infusions. Seven CAR-T products are now FDA-approved for B-cell cancers and multiple myeloma.
AML doesn’t offer that same clean target. The core problem is that AML cells share nearly all of their surface proteins with healthy blood-forming stem cells in the bone marrow. Destroying those stem cells causes prolonged myeloablation, meaning the body loses its ability to produce white blood cells, red blood cells, and platelets. That’s not a manageable side effect; it’s life-threatening. This overlap between leukemia cells and healthy cells is the single biggest reason AML lags years behind other blood cancers in CAR-T development.
AML also tends to be a heterogeneous disease. Different leukemia cells within the same patient can express different surface proteins, so a CAR-T product targeting one marker may kill some cancer cells while leaving others untouched. On top of that, AML patients are often older and sicker, and the disease can progress rapidly, leaving a narrow window for treatment.
Targets Under Investigation
Researchers are testing CAR-T cells aimed at several AML surface proteins: CD33, CD123, FLT3, CLL-1, CD7, CD38, and others. Each has trade-offs. CD33 and CD123 are found on most AML cells, making them attractive, but they also appear on normal blood-forming cells. CLL-1 is somewhat more selective for leukemia, which has made it one of the more promising candidates. CD7 is another active area of research, though it’s also expressed on some healthy immune cells.
No single antigen has emerged as the clear winner. The field is instead moving toward combination approaches, targeting two proteins at once to improve precision while reducing the chance that leukemia escapes by losing one marker.
What Clinical Trials Have Shown So Far
Most AML CAR-T trials are in Phase 1, meaning they’re primarily testing safety in small groups of patients. The results are a mixed but increasingly encouraging picture.
CLL-1 targeted CAR-T cells have produced some of the strongest early numbers. In one trial, 75% of patients achieved complete remission, with the remaining 25% reaching a near-complete response. A larger study of CLL-1 CAR-T reported complete remission in 7 out of 10 patients. A Phase 1 trial called CARMEN is currently evaluating CLL-1 CAR-T as a bridge to stem cell transplant, with a dose expansion phase being considered to further assess efficacy.
CD123-targeted CAR-T showed a 100% overall response rate in a small three-patient study, with the engineered cells persisting for up to six months. CD7 CAR-T produced complete remission in 70% of cases, with some patients achieving long-term disease-free survival after a follow-up stem cell transplant. CD38-targeted CAR-T achieved remission in 4 out of 6 patients, though median overall survival was 7.9 months.
Not every target has panned out. A CD33 CAR-T trial using standard autologous (patient-derived) cells produced zero objective responses in one study. A separate analysis found CD33 CAR-T induced complete remission in about 42% of patients who relapsed after a prior stem cell transplant, but that still leaves more than half without a meaningful response.
These numbers are promising for early-stage research, but the patient groups are tiny, often fewer than a dozen people. Durable remissions, the kind that last years, remain difficult to demonstrate.
Smarter CAR-T Designs
Because no single target is both safe and effective enough on its own, engineers are building CAR-T cells that can read two signals before attacking. These designs borrow from computer science, using what are called logic gates.
An “OR gate” CAR-T cell activates when it detects either of two tumor proteins. This makes it harder for leukemia to escape by losing one antigen. An “AND gate” design requires both proteins to be present before the cell attacks, which adds a layer of safety. If healthy stem cells express only one of the two targets, the CAR-T cell leaves them alone. Researchers are also developing inhibitory CARs that actively shut down when they encounter a protein found only on healthy cells.
These dual-target approaches could eventually solve AML’s fundamental problem. One research group working on dual-target CAR-T for high-risk leukemia described the goal as creating a stand-alone therapy that doesn’t require a rescue bone marrow transplant afterward. If that works, it would be a major step toward making CAR-T practical for AML patients outside of specialized transplant centers.
The Rescue Transplant Problem
Right now, most AML CAR-T strategies assume the patient will need a stem cell transplant afterward. Because CAR-T cells targeting AML antigens tend to also destroy healthy blood-forming cells, patients can’t recover normal blood counts on their own. A “rescue” bone marrow transplant replaces those destroyed stem cells.
This creates a two-step process: CAR-T to eliminate leukemia, then transplant to restore blood production. It works in concept, but it limits who can receive the therapy (you need to be healthy enough for transplant), adds significant cost and recovery time, and introduces transplant-related risks like graft-versus-host disease. Any CAR-T product that could avoid this requirement would dramatically expand the eligible patient population.
CAR-NK Cells as an Alternative
Natural killer (NK) cells are a parallel track that could reach AML patients sooner or complement CAR-T approaches. NK cells are part of the body’s innate immune system and have a natural ability to recognize and kill abnormal cells. When engineered with the same chimeric antigen receptors used in CAR-T, they gain targeted precision while keeping a more favorable safety profile.
In early trials, CAR-NK cells have caused minimal cytokine release syndrome (the inflammatory storm that makes CAR-T dangerous), virtually no neurotoxicity, and no graft-versus-host disease. A Phase 1 trial treating three patients with relapsed AML using anti-CD33 CAR-NK cells reported no relevant adverse effects. Two other early trials, treating a combined 19 patients, showed the same clean safety record.
The practical advantage of CAR-NK is that these cells can come from donors, cord blood, or even cell lines rather than requiring each patient’s own cells. That opens the door to “off-the-shelf” products that could be manufactured in advance and given immediately, avoiding the weeks-long manufacturing delay that is especially problematic for fast-moving AML. Efficacy data is still early, but the safety and scalability advantages are drawing significant investment.
A Realistic Timeline
The honest answer is that AML CAR-T therapy is still in early clinical development, with most trials in Phase 1 or early Phase 2. For context, it typically takes 5 to 10 years to move from Phase 1 results to FDA approval, and even that timeline assumes everything goes well. The first CAR-T product approved for any cancer (for pediatric ALL) was based on research that began more than a decade earlier.
Several factors could accelerate or slow the timeline. On the optimistic side, the FDA has shown willingness to fast-track cell therapies through breakthrough designations, and the engineering advances in dual-target and logic-gate designs are progressing rapidly. CLL-1 and CD123 CAR-T candidates appear furthest along, with dose-expansion studies underway. On the cautious side, the small trial sizes mean we don’t yet know how durable these remissions are, and the safety challenges around healthy cell destruction haven’t been fully solved.
A reasonable estimate is that the first CAR-T or CAR-NK therapy specifically approved for AML could arrive in the late 2020s, likely around 2028 to 2030, assuming at least one candidate produces strong enough Phase 2 data to support an accelerated approval pathway. For patients with relapsed or refractory AML today, clinical trials remain the primary route to access these therapies.