Blood Cancer’s Genetic Warning Signs Can Appear Years Early, Study Shows

The genetic changes that decide whether a slow-growing blood cancer will one day turn aggressive can show up years before any symptoms, a long-term study has found.

Researchers at the Wellcome Trust Sanger Institute, with Cambridge University Hospitals, studied myeloproliferative neoplasms (MPNs) — rare chronic blood cancers beginning in the bone marrow. They followed 30 patients, some for up to 25 years, analysing more than 450 blood samples and nearly 8,000 test results. Published July 31, 2026 in Cancer Discovery.

What MPNs are

These conditions arise when a blood stem cell acquires a mutation causing it to produce too many mature blood cells — red cells, platelets, or both — without the normal signals telling it to stop.

Most cases involve mutations in one of a small set of genes, most commonly JAK2, that lock a growth-signalling pathway permanently active.

The immediate risk is not the cancer growing uncontrollably but the blood becoming thicker or clotting more readily, so much of the management addresses thrombosis rather than the malignancy.

The longer-term risk is transformation. Some patients progress to myelofibrosis, where the marrow scars and fails, or to acute leukaemia — and predicting who has been the central unsolved problem.

What the team saw

Patients whose disease later progressed steadily accumulated new DNA mutations, while those who stayed stable picked up few additional changes.

“Mutations linked to future progression may be detectable long before symptoms worsen or standard clinical tests reveal a major change,” the researchers said.

The trajectory is the finding rather than any single mutation. It is not that progressing patients carried one identifiable marker at baseline, but that the rate at which their genomes changed differed — which is information only repeated sampling can produce.

Why 25 years of follow-up was necessary

The duration is what makes the study unusual and is dictated by the biology.

MPNs progress over decades. A study following patients for five years would capture few transformations and could not distinguish patients heading toward progression from those who never will — because on that timescale they look identical.

One participant, diagnosed in 1992 with essential thrombocythemia, progressed to myelofibrosis only after more than three decades of care.

The trade-off is sample size. Thirty patients is small, and it reflects what is achievable when each contributes decades of banked samples and records — a study of this design cannot be scaled up quickly, only started earlier.

Why blood makes this feasible

Blood cancers permit longitudinal genomic tracking in a way solid tumours do not.

The malignant cells circulate, so a routine blood draw samples the disease directly — no biopsy, no procedure, no risk. That is what allows 450 samples across 30 patients to exist at all.

Equivalent tracking of a solid tumour would require repeated biopsies, which are invasive, sample only one part of a heterogeneous mass, and cannot reasonably be performed dozens of times.

What it could change

Regular genomic monitoring could in principle identify which patients are heading toward more dangerous disease — allowing earlier intervention for those at risk while sparing stable patients unnecessary treatment.

Both halves matter, and the second is easy to overlook. Many MPN patients live for decades with well-controlled disease, and treating them more aggressively on the basis of a diagnosis alone would expose them to toxicity for no benefit.

A test separating the groups improves care in both directions, which is a stronger proposition than one that only identifies risk.

What stands between this and practice

The approach is not yet routine clinical practice, and the gap is substantive.

A prognostic signal must be turned into a defined test with thresholds — how many new mutations over what interval constitutes concerning accumulation. That requires validation in a larger independent cohort.

What clonal haematopoiesis added to this picture

The study sits within a shift in how the earliest stages of blood cancer are understood.

Sequencing blood from large numbers of healthy older adults revealed that many carry expanded populations of blood cells descended from a single mutated stem cell — a condition termed clonal haematopoiesis. It is common with age, causes no symptoms, and most people who have it never develop blood cancer.

It does carry elevated risk, both of haematological malignancy and, unexpectedly, of cardiovascular disease. But the absolute risk in any individual is low, which makes it a poor basis for intervention on its own.

What that established is that the boundary between normal aging blood and early malignancy is not sharp. Mutated clones exist in most older people, and disease is the far end of a continuum rather than a separate state.

Which is exactly why trajectory matters more than a snapshot. A single measurement showing mutated cells says little when mutated cells are near-universal; the rate at which a clone acquires further changes distinguishes one that is drifting from one that is progressing — and only repeated sampling reveals the difference.

It also requires an action. Identifying high-risk patients earlier helps only if something useful can be done, and whether earlier intervention actually improves outcomes in MPN has not been established. Research news, not medical advice.