Genomic Sequencing Could Individualize Diagnosis, Treatment of Myeloid Cancers

Extensive genomic testing of people with acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) has revealed numerous disease subtypes, many previously unknown and each with its own prognosis and possibilities for treatment. Knowledge about these subtypes and the genetic alterations that drive them could leave doctors more able to individualize treatment for their patients with myeloid malignancies. The information could also help doctors diagnose these diseases and determine expected prognoses.

The testing was done as part of a study whose results were presented Dec. 10 during the 61^st Annual Meeting and Exposition of the American Society of Hematology (ASH) in Orlando, Florida.

AML and MDS are blood cancers that undermine the body’s ability to make healthy blood cells.

Ilaria Iacobucci, a pathologist at St. Jude Children’s Research Hospital in Memphis, Tennessee, embarked on the study with colleagues because previous genomic sequencing studies had analyzed only certain subtypes or looked for specific gene alterations. This study was different in that it involved analysis of all the DNA and messenger RNA molecules in the blood and bone marrow of adult patients with AML and MDS. That information was studied along with data on health outcomes and the physical features of each patient’s cancer. This uncovered a wide range of both inherited and acquired genetic alterations and illuminated the links between genes, gene expression, the physical structure of cancer cells and patient outcomes.

“Treating these diseases is a challenge because subtypes present different features,” Iacobucci said. “Our study provides a much richer understanding of these subtypes, akin to a dictionary of all the genomic alterations. It also underscores the value of having comprehensive genomic information at the start of treatment to remove uncertainty and help clinicians better understand a patient’s outlook.”

The study looked at the blood and bone marrow of 598 adults with AML and 706 adults with MDS. It found previously unknown connections between genetic alterations and the behavior of the cancers. The researchers also discovered that some myeloid cancers with very different structures shared genetic similarities. In all, the researchers found more than 7,000 variants in 839 genes.

In looking at 87% of the cases, the researchers found that TET2 and DNMT3A were the most frequently mutated genes. A subgroup of this category also had NPM1 mutations (27.4% of AML cases and 1% of MDS cases). Within that subgroup, researchers found four different gene expression signatures with different combinations of cooperating mutations in cohesion and signaling genes. These combinations led to different prognoses. For instance, patients with NPM1 and FLT3 mutations had a worse outcome than patients with only the NPM1 mutation. But co-occurring mutations in NPM1 and cohesion genes predicted a better health outcome, even if FLT3 mutations were also present.

A separate 11% of patients with AML had genetic alterations known as recurrent, meaning they are associated with specific chromosomal changes. In these cases, which carry a good prognosis, the researchers found a distinct gene expression pattern; they divided these into three subtypes, each driven by its own series of mutated genes. Categorized according to the presence of specific altered protein coding genes, the subtypes are: RUNX1-RUNXT1 leukemia; PML-RARA promyelocytic leukemia; and CBFB-rearranged leukemia.

Other predictors of poor outcome are RUNX1 alterations and mutations to the MN1, KMT2A and TP53 genes, the researchers reported.

While a variety of mutations led to many different gene expression patterns in AML cases, expression was less variable in MDS samples, even in those with varying patterns of gene mutation. Common mutations in MDS included SF3B1, SFRS2, U2AF1, TP53 and RUNX1.