Damaged mitochondrial DNA marked the leukemia cells that answered imatinib better

Why do some people with chronic myeloid leukemia (CML) respond quickly and deeply to the drug imatinib, while others respond slowly or barely at all? A study published in HemaSphere in October 2026 (Pagani and colleagues) offers a mitochondrial answer: the state of a leukemia cell's mitochondrial DNA — and how hard its mitochondria are working — tracks closely with how that cell fares under treatment.

CML is usually treated with tyrosine kinase inhibitors such as imatinib, which have brought survival close to that of the general population. The remaining puzzle is the minority of patients whose disease responds poorly. This study suggests part of that puzzle sits inside the cell's energy system.

What did the researchers actually measure?

The team studied 120 patients with chronic-phase CML, examining mitochondrial DNA mutations alongside measures of how their leukemia cells produced energy. They also drew on a shared single-cell dataset from collaborators in Singapore, which let them screen mitochondrial mutations across thousands of individual cells, and on laboratory protocols developed with colleagues at the University of Glasgow and the South Australian Health and Medical Research Institute (SAHMRI).

Two patterns emerged. Leukemia cells from poorer responders leaned more heavily on oxidative phosphorylation — the mitochondrial pathway that makes most of a cell's ATP. And mitochondrial DNA mutations were linked to impaired oxidative phosphorylation: in other words, cells whose mitochondrial genomes carried more damage were less able to run that energy pathway at full power.

Why would weaker mitochondria mean a better drug response?

The logic is about backup power. A leukemia cell with fit, active mitochondria has a flexible energy supply it can draw on while imatinib blocks the cancer's main growth signal. A cell whose mitochondria are already compromised has less reserve — so the same drug pressure is harder to survive. That would explain why patients whose cells showed more mitochondrial impairment tended to respond better to imatinib.

The researchers then tested the flip side in the laboratory: treating CML stem and progenitor cells with a mitochondrial inhibitor reduced their ability to grow and survive, and combining the inhibitor with imatinib nearly wiped out colony formation in preclinical experiments. It is early, dish-level evidence — but it supports the idea that mitochondrial energy production is a real vulnerability in treatment-resistant CML cells, not just a marker.

What could this mean for patients?

If the findings hold up, mitochondrial genetics could one day help clinicians predict, at diagnosis, which patients are less likely to respond well to imatinib — and consider closer monitoring or different strategies for them. The authors also raise the longer-term possibility of pairing existing CML drugs with medicines that disrupt how cancer cells generate energy.

Both ideas are firmly in the 'promising, not proven' category. Mitochondrial inhibitors are not currently used to treat CML, the combination work is preclinical, and the patient findings are associations from a single study of 120 people. Replication in larger cohorts — and proof that acting on the signal improves outcomes — would need to come first.

Where the evidence stands

• Established: Tyrosine kinase inhibitors such as imatinib are frontline CML treatment and have brought survival close to general-population levels; a minority of patients respond slowly or poorly.
• Preliminary: In 120 chronic-phase patients, mitochondrial DNA mutations were associated with impaired oxidative phosphorylation and better imatinib response, while fitter mitochondria marked poorer responders; a mitochondrial inhibitor plus imatinib nearly eliminated colony formation in preclinical cell experiments.
• Absent: Any clinical trial of mitochondrial inhibitors in CML, validation of mtDNA testing as a routine predictor of imatinib response, and evidence that acting on this signal changes patient outcomes.