Genetic Tracing Finds Mutations for Blood Cancer Appear Decades Before Diagnosis

A late-breaking study presented at the American Society of Hematology annual meeting raises the specter of whether interventions can occur decades before a blood cancer would appear.

If you had a crystal ball and knew you would develop a blood cancer in the future, would you do something to stop it? That’s the question sparked by research presented Tuesday on the final day of the 62nd annual meeting of the American Society of Hematology (ASH).

The findings, highlighted at a session featuring late-breaking studies, showed how scientists at the Wellcome Sanger Institute and the University of Cambridge, in the United Kingdom, used bone marrow and blood samples to trace the genetic beginnings of a blood cancer in 10 people. To their surprise, the researchers found that cancer driver mutations appeared in childhood, decades before a diagnosis—and some mutations were present before patients were born.

The study’s senior author said during a press briefing that the opportunity exists to repeat this research strategy in patients with other blood cancers, to see if the driver mutations show up just as early. And the question of whether next-generation sequencing (NGS) could be used to sniff out potential cancers in healthy people years ahead of time is not theoretical.

When does cancer start? Senior study author Jyoti Nangalia, MD, who presented the findings, said the results answer a common question among patients: when did my cancer start growing? In some cases, she said, “We were able to study how these particular cancers developed over the entire lifetime of individual patients.”

The Wellcome Sanger study involved Philadelphia-negative myeloproliferative neoplasms (MPNs), which are driven by JAK2V617F mutations in most patients. For the 10 patients with MPNs, who ranged in age from 20 to 76 years, the researchers first performed whole-genome sequencing on single-cell–derived hematopoietic colonies of the blood cancer from each patient to isolate the driver mutation. Then they resequenced older blood samples available for each patient, tracing the timing of the cancer-causing mutation alongside hundreds of thousands of other somatic mutations that occurred along the way.

“We identified 448,553 somatic mutations which were used to reconstruct phylogenetic trees of hematopoiesis, tracing blood cell lineages back to embryogenesis” the authors wrote. “We timed driver mutation acquisition, characterized the dynamics of tumor evolution, and measured clonal expansion rates over the lifetime of patients.”

When the JAK2 mutation that was the focus of the study acted alone to drive a patient’s cancer, the researchers found it was acquired early, often at the dawn of life. For this group, they estimate the mutation appeared a few weeks after conception, with upper range estimates between 4.1 months and 11.4 years. For these patients, the mean length of time between acquiring a mutation and cancer appearing was 34 years.

Even when the JAK2 mutation was a second driver, acting with another mutation, there was still a latency period of 12 to 27 years before cancer appeared.

What about other cancers? In response to a question, Nangalia said the research methods used in the study are generally applicable for other cancers, including solid tumors.

“We already know in myeloma that some of the chromosomal aberrations occur early compared to others,” she said. “In myeloma, we know the relative timing of one coming before the other or after, but what we don't know is the absolute timing in terms of whether these mutations or chromosomal aberrations occurred in childhood, in utero, or indeed, much later in life closer to when the patient presented.”

Based on the surprise of the MPN study, she wouldn’t make any guesses about multiple myeloma or any other disease. For one patient in the MPN study, the JAK2 mutation appeared more than 50 years before cancer was diagnosed. Given these results, Nangalia said, “I really don't know what to predict about other cancers.”


When asked by The American Journal of Managed Care® if the findings have implications for broader use of NGS, Nangalia said they clearly do. NGS, she said, is being used more often in both cancer diagnostics and prognostics, and to find therapeutic targets. “However, I think our study opens a whole new application for next-generation sequencing in terms of potentially identifying patients or individuals that do not yet have a diagnosis earlier, to see which patients are at risk of a future blood cancer.

“That will require further work to get there and validate this approach,” she said. But, given that patients in the study with MPNs had mutations for 10 to 40 years before blood cancer appeared, “We would have been able to detect these mutations with next-generation sequencing decades before their diagnosis. And we think we would have also been able to understand and predict which patients were on a path to future disease by estimating their rate of growth.”

The opportunity exists, she said, for researchers to use NGS to study healthy individuals to understand which patients are at risk of cancer.

Reference

Williams N, Moore L, Baxter JE, et al. Driver mutation acquisition in utero and childhood followed by lifelong clonal evolution underlie myeloproliferative neoplasms. Presented at: the 62nd American Society of Hematology Annual Meeting and Exposition; December 5-8, 2020. Abstract LB-1. https://ash.confex.com/ash/2020/webprogram/Paper143813.html