Skip to main content
Premium Trial:

Request an Annual Quote

Sequencing Study Suggests Mutation Selection Behind Glitches in Therapy-related AML


NEW YORK (GenomeWeb) – The risk of developing tricky-to-treat leukemia in the years following treatment for other types of cancer appears to reflect selection for cells containing treatment resistance-related mutations that are already circulating in the body, according to a new study.

As they reported online today in Nature, researchers from Washington University in St. Louis and elsewhere did whole-genome sequencing on samples from 22 individuals with therapy-related AML (t-AML) — a complication that occurs in a fraction of those receiving cytotoxic chemotherapy and/or radiotherapy in the first one to five years after their cancer treatment.

The mutational burden in t-AML tumors resembled those in non-treatment-related forms of AML, the team found, arguing against the notion that t-AML stems from rampant, treatment-related DNA damage.

Instead, treatment seemed to select for circulating clones that had acquired resistance-related mutations as part of the normal aging process — an idea supported by the presence of TP53 gene mutations in a fraction of circulating blood cells from t-AML patients years prior to diagnosis and in healthy individuals who'd never received chemotherapy.

"The expectation of most scientists, us included, was that chemo induces widespread genomic DNA damage and that explains the complex chromosomal abnormalities and so on," corresponding author Daniel Link, an oncology researcher with Washington University's Siteman Cancer Center, told GenomeWeb.

"But when we sequenced these cases, what we noticed was that the mutation burden was no different than de novo AML," he explained. "So that really got us to re-think things."

As one of the first cancer types to be tackled by whole-genome sequencing, the mutational repertoire of AML has been increasingly well defined. For instance, past studies have looked at mutations present in de novo AML and in secondary AMLs that arise in individuals with a history of blood conditions such as myelodysplastic syndrome (MDS) or myeloproliferative neoplasms (MPNs).

Targeted testing of AML and/or MDS cases that crop up following exposure to chemotherapy and/or radiation suggests that the tumors tend to contain more-frequent-than-usual TP53 mutations, along with alterations involving chromosomes 5 and 7, Link explained.

But the reason for the enhanced risk of AML and MDS following chemotherapy was unclear, he noted. "One of our opening questions for this [study] was, 'How does this exposure to chemotherapy years prior shape this disease to make it a much nastier type of AML and one with genetic changes that are different?'"

To look into that process in more detail, Link and his colleagues used Illumina sequencing to do whole-genome sequencing on samples from 22 individuals with t-AML who had undergone prior treatment for a range of other cancer types.

The team found complicated cytogenetic profiles in almost one-quarter of the t-AML tumors, suggesting chromosomal alterations can contribute to t-AML, as proposed in the past.

Even so, more than one-third of the fully sequenced samples had typical cytogenetic characteristics. And when the researchers compared the full mutational burden in the t-AML genomes to those found in de novo AML or secondary AML tumors, they saw comparable mutational frequencies rather than an avalanche of mutations in the t-AML tumors.

Since the overall numbers of single nucleotide changes, small insertions and deletions, and transversions were similar in all three AML types, the team concluded that t-AML is not simply a consequence of rampant chemotherapy-induced DNA damage.

The researchers then looked at another 52 individuals with t-AML and 59 individuals with t-MDS to assess the sorts of genes that are mutated in t-AML and t-MDS, sequencing a panel of 150 genes that seemed suspicious in the t-AML genomes or had been implicated in past studies of de novo AML and MDS.

Compared to existing exome or genome data for hundreds of de novo AMLs and MDS samples, the t-AML tumors and t-MDS samples particularly prone to TP53 mutations and chromosomal rearrangements were linked to poor outcomes, as reported previously.

Since past research — including work by members of the same team — has hinted that mutations can crop up in a subset of hematopoietic stem cells as part of the normal aging process, the investigators speculated that the rise in TP53 mutations in t-AML might reflect an expansion of pre-existing clones after chemotherapy.

Indeed, the researchers identified TP53 mutations that preceded AML diagnoses when they tested blood samples collected from four individuals with t-AML or t-MDS between three and six years before they developed the treatment-associated blood conditions. In two of the four individuals, these mutations could be detected in a fraction of circulating cells even before their initial chemotherapy treatments.

Such glitches were detected in healthy individuals, too. The team used a sensitive adaptor sequencing assay to search for low-frequency TP53 mutations in 19 individuals between the ages of 68 and 89 years with no history of cancer or chemotherapy. Almost half of the individuals carried sub-populations of cells in which one copy of the TP53 gene was mutated.

"People are mosaics because of aging," Link said. "We have, maybe, 20,000 stem cells and each one of those stem cells has its own set of aging-related mutations."

In mouse models, meanwhile, the investigators determined that chemotherapy is capable of selecting for subsets of blood cells that contain mutations affecting one copy of TP53, leading to expansions of a clonal population that's expected to show enhanced resistance to subsequent cancer treatments.

Given their findings so far, Link and his colleagues are keen to start looking for TP53 mutations in samples collected prospectively from individuals undergoing cancer treatments. In particular, they plan to focus on cancer types that most often progress to t-AML such as relapsed lymphoma, which typically requires large doses of chemotherapy over relatively long time periods.

By profiling the presence and prevalence of TP53 mutations in the blood over time in these individuals, the team hopes to find clues for predicting which cancer patients are at greatest risk of developing t-AML.

The researchers are also in the process of doing mouse experiments to try to determine whether drugs that modulate the activity of TP53 gene products could dial back the selective advantage that mutated TP53 sub-populations seem to gain during chemotherapy.

"If you apply those drugs while you're giving chemo, will you mitigate this advantage of the mutant cells?" Link said. "Those are pure mouse studies at this point, but if those are true, they could be a chemo prevention strategy."