Skip to main content
Premium Trial:

Request an Annual Quote

New Approach Improves Estimating Cancer Risk From Age-Related Clonal Hematopoiesis

NEW YORK — Researchers have developed a new method that may allow them to better determine which patients with clonal hematopoiesis of indeterminate potential (CHIP) are more likely to progress to diseases like leukemia.

CHIP, or the clonal expansion of hematopoietic stem and progenitor cells, affects more than 10 percent of individuals aged 60 or older and is linked to an increased risk of hematological neoplasms. Because of this, detecting CHIP early could identify patients who might benefit from more frequent clinical monitoring to detect disease early.

CHIP is typically diagnosed by uncovering somatic mutations in cancer-linked genes that are present at variant allele frequencies (VAF) of at least 2 percent and are present in more than 4 percent of all blood cells. But researchers from the University of Edinburgh hypothesized that CHIP and its associated disease risk could instead be determined by examining the clonal fitness of stem cells over time. In a study appearing Monday in Nature Medicine, they found that their approach, called LiFT (likelihood-based filter for time series data), could identify additional variants that the traditional approach did not and could help stratify patients to personalize their management.

"Our gene-specific results, now, but especially when scaled up to cohort size, can provide clinicians with a comprehensive catalog of direct fitness estimates," co-senior author Tamir Chandra said in an email, adding that "these estimates can also provide an important layer of patient stratification for clinical cross-sectional studies, where fitness cannot be directly measured."

For their analysis, the researchers used sequencing to profile CHIP variants within subsets of the Lothian Birth Cohorts of 1921 and 1936. The cohorts were followed longitudinally, with follow-ups every three years for about a dozen years beginning at the age of 70 or 79, depending on the cohort.

Using the VAF criteria approach, the researchers identified 76 CHIP mutations.

They then used the longitudinal data to examine whether variants were increasing or decreasing in number and found that 70 percent of the CHIP mutations that reached the 2 percent VAF threshold were actually growing, suggesting that the approach captures mutations that are shrinking in frequency and do not confer a fitness advantage. At the same time, the approach could miss fast-growing mutations that do provide an edge.

The researchers turned to their LiFT algorithm, which uses longitudinal data to distinguish between variants that increase fitness versus those that do not. In particular, Chandra said that it relies on the sequencing depth at each mutation to gauge how accurate the VAF call is and compares their trajectory to mathematical models of stem cell division.

In this cohort, the researchers identified 114 variant trajectories, 86 percent of which increased over time. Of these, 50 variants would not have been detected using the VAF threshold approach, including the U2AF1 Q157R and DNMT3A R882H variants.

They further noted that the VAF threshold approach did not identify any TP53 variants, but that LiFT uncovered four, all of which were growing in number. These were termination or frameshift mutations or had previously been linked to cancer.

The results could inform clinical decisions such as how frequently to reassess and monitor patients.

"One of the main arguments we make in the paper is that a 2 percent VAF threshold, which is the current definition of CHIP, is problematic," Chandra said. "For example, you could have a fast-growing clone at low VAF that could be more pathogenic and worth monitoring than a slow-growing clone at 2 percent VAF. Our approach allows the identification of fast-growing clones at much lower VAF."

Going forward, Chandra added that the team hopes to extend their study to the full Lothian Birth Cohort as well as to delve into the mechanisms that might be driving the CHIP mutations.