The idea sounds counterintuitive.
Why would you edit healthy donor stem cells to remove something?
Because the current treatments are too blunt. They smash cancer cells. They also smash healthy cells.
For patients with aggressive blood cancers like AML and MDS, a stem cell transplant is often the only real shot at a cure. But the cancer often comes back. The immunotherapies used to kill those lingering cancer cells — things like CAR-T — are powerful. They’re also messy.
John F. DiPersio and his team at Washington University in St. Louis wanted to fix that. They published their findings in Nature Medicine.
They used CRISPR to remove a protein called CD33 from donor stem cells before transplanting them.
CD33 is a marker found on both cancer cells and healthy myeloid blood-forming cells. Current therapies target it. They don’t know the difference between a cancer cell and a healthy stem cell from the donor. Result? The therapy attacks both. The patient loses healthy blood production. The cancer wins by default because the body’s defense is neutered.
Remove CD33 from the donor cells, and you create a safe house. The immune therapy still hunts down the cancer cells that still have CD33. The donor stem cells? They’re invisible to the treatment. They survive. They repopulate the marrow.
It’s a workaround. A genetic stealth mode.
Why CAR-T struggles with blood cancers
Here’s the snag with acute myeloid leukemia.
CAR-T cells are engineered T-cells. You wire them to seek out a specific protein. Problem is. That protein exists on the good stuff too.
When the CAR-T cells go after the protein on healthy cells, the patient gets severe inflammation. Their blood counts crash. The therapy stops working effectively because it’s too busy fighting healthy tissue to fight the tumor.
Miriam Y. Kim first proposed this strategy. She was a postdoc at UPenn. Now she’s an assistant professor at WashU. She joined DiPersio’s lab to build the case for editing donor cells.
The logic was simple. If you delete CD33 from the transplant, the only things left in the body that have CD33 are the cancer cells.
Any remaining healthy blood-forming cells should lack the marker.
A case study that worked
The team reported a phase 1/2 trial. Thirty adults with high-risk AML or MDS.
Donor stem cells were edited. The product, called tremtelectogene empogedite (trem-cel). Funded by Vor Biopharma.
Patients got the edited cells. Then they got gemtuzumab ozogemicin.
It’s an antibody linked to a poison. It targets CD33. It’s already approved for some AML cases. It’s being studied for MDS. The issue? It’s hard on the liver. It wipes out white blood cells, red cells, platelets. Toxicity is a major limit.
With edited cells, that toxicity should theoretically bypass the healthy bone marrow.
All thirty patients had the new stem cells take hold in 28 days. Some bounced back faster. Platelets returned in an average of 16. Days.
That’s standard. Even fast. For a transplant.
Nineteen patients got maintenance therapy. They kept their blood counts stable. No crashes. The edited cells protected them.
DiPersio called it encouraging.
“In the future, we are hopeful we can combine this with CD33-target immunotherapies and improve treatment options.”
Then there’s the one-off case.
A patient with super aggressive AML. Got the CD33-less transplant. Cancer came back. They gave him CD33-targeting CAR-T. Using T-cells from the same donor.
The CAR-T wiped out the cancer. The edited donor cells remained untouched. The patient achieved full remission. Been clean for over a year. His blood now makes cells without CD33. The edit stuck. The cure held.
Published in JCO Precision Oncology.
It’s not perfect yet
Side effects were real.
Anemia. Fevers. Infections. Graft-versus-host disease — where the donor cells attack the body anyway.
Seven patients died.
Four died of their cancer.
Three died of transplant complications. Kidney failure. Liver toxicity. Sepsis.
It’s heavy medicine. Editing doesn’t erase the risk of dying.
But the survival average hit over 14 months. That’s better than many expected for these specific high-risk profiles.
CD33 isn’t vital. Some people are born without it. They live normal lives. So removing it from donor cells is likely safe long-term.
The trial was just a proof of concept. Feasible? Yes. Safe enough? Yes.
Does it solve the relapse problem entirely? Maybe not yet. But it changes the battlefield. You can finally unleash the immune therapy without destroying the infrastructure needed to run the human body.
The study opens the door for precise attacks on cancer without collateral damage.
Vor Biopharma funded it. Several authors work there. Transparency is part of the report.
What’s next? Combining these edits with more therapies. Maybe beyond CD33. Maybe beyond AML.
It’s early. The numbers are small. The deaths still happen. But the direction feels right. Stop fighting the host while you fight the guest.
The science catches up. Then maybe the outcomes will too.
