When Naveen Pemmaraju first treated Maria, a patient with an aggressive blood cancer called BPDCN, the drug tagraxofusp worked beautifully — at first. Then the cancer stopped responding. Now, Pemmaraju and a team at MD Anderson Cancer Center in Houston, Texas, think they know why.
In a study published in the journal Leukemia, researchers identified two biological markers that predict when patients will resist tagraxofusp, the first FDA-approved treatment for blastic plasmacytoid dendritic cell neoplasm (BPDCN). BPDCN is a rare and aggressive leukemia that develops from immune cells in the bone marrow. Patients with BPDCN have few treatment options and face a poor prognosis.
The team, co-led by Hannah Beird, a senior research scientist, and Naveen Pemmaraju, a professor of leukemia, analyzed nearly 100,000 individual cancer cells from 30 bone marrow samples. They found that patients with severe mutations in a gene called TET2 were much more likely to stop responding to tagraxofusp. They also discovered that resistant cancer cells consistently had low levels of an enzyme called TXNRD1 — and that enzyme is exactly what tagraxofusp needs to release its cancer-killing toxin inside cells.
"Our findings show that specific cancer cells can effectively escape destruction by dialing down key enzymes that tagraxofusp needs in order to work," Beird said.
Without enough TXNRD1, the drug's toxin gets trapped inside cancer cells instead of activating, allowing the disease to survive and spread. About 10 to 25 percent of newly diagnosed BPDCN patients do not initially respond to tagraxofusp, leaving doctors searching for explanations.
But this discovery opens new doors. Because TET2 mutations appear to predict who will benefit from tagraxofusp, doctors could test for this marker before treatment begins. Monitoring TXNRD1 levels during therapy could also alert clinicians early when resistance is developing, giving them time to adjust the treatment plan.
The researchers also tested a potential fix: combining tagraxofusp with another drug called azacitidine restored the pathways that the cancer cells had shut down, improving outcomes in laboratory models. If this combination works in patients, it could prevent relapse in people like Maria.
"Armed with this information, we can begin to predict which patients are less likely to respond, and we can design smarter, more personalized treatments to help improve outcomes," Beird said.
For patients with a cancer that had no approved therapies just a few years ago, the ability to predict resistance — and potentially overcome it — represents real progress.