NEW YORK (GenomeWeb) – A new genomic analysis of acute erythroid leukemia (AEL) has uncovered recurrent tumor gene mutation and expression profiles, including genomic features that appear to coincide with outcomes for patients affected by the rare, difficult-to-treat form of acute myeloid leukemia (AML).
"These results mark a new era in understanding and treatment of AEL, an aggressive leukemia that has been plagued by diagnostic controversy and poor outcomes," senior author Charles Mullighan, a pathology researcher and co-leader of the St. Jude Children's Research Hospital's hematological malignancies program, said in a statement.
As they reported online today in Nature Genetics, Mullighan and colleagues performed whole-genome, exome, targeted, and transcriptome sequencing on samples from 159 pediatric or adult AEL patients treated at sites around the world, comparing the somatic mutations and gene expression patterns they found to those in samples from more than 1,900 individuals with non-AEL conditions — from other forms of AML to myelodysplastic syndrome.
In the process, they saw distinct alterations in the AEL tumors and identified half a dozen AEL subgroups that clustered with respect to patient age at diagnosis, tumor mutations, tumor gene expression, and patient outcomes.
"We found that patients with AEL, AML, and MDS had many of the same mutations, but the frequency and patterns of the mutations were very different," first author Ilaria Iacobucci, a staff scientist in Mullighan's St. Jude lab, said in a statement.
The analyses also highlighted a significant proportion of AEL patients with recurrent ALK or NTRK1 kinase signaling gene mutations that may be susceptible to targeted treatment. In follow-up experiments on a mouse model of AEL, for example, the investigators demonstrated that treatment with the tropomyosin receptor kinase-inhibitor larotrectinib could significantly delay tumor growth in six treated mice compared with five untreated mice.
Mullighan noted that such findings " highlight a potential role for such inhibitors in future clinical trials."
For their initial genomic analysis, the researchers focused on 159 AEL patients treated in the US, Europe, Singapore, Japan, and Australia. Of those, 35 cases involved patients 20 years old or younger and eight AEL cases affecting individuals between the ages of 21 and 39. Another 32 patients were between 40 and 59 years old, while the remaining 84 patients were at least 60, they reported.
The team sequenced the exomes of matched tumor-normal samples from 20 individuals with AEL and the exomes of 122 tumor-only samples and two cell lines. They also generated Complete Genomics whole-genome sequencing data on paired tumor-normal samples from five pediatric AEL cases and tumor-only genome data from one adult tumor.
To that data, the researchers added RNA sequence data for 139 AEL tumors and two cell lines, along with targeted sequencing data from a dozen AEL tumors and array-based copy number profiles for 137 cases. They also considered data for 1,903 controls with non-AEL AML, MDS, or other myeloid conditions.
Along with 11 recurrently mutated pathways, the team documented five subgroups with shared age, transcription, or tumor mutation patterns: a group of adult cases with mutations in the TP53 tumor suppressor gene, another adult subgroup with DDX4-mutated tumors, a set of pediatric cases involving NUP98 rearrangements, a subgroup marked by KMT2A mutations or rearrangements, and a subgroup containing NPM1. A sixth AEL subgroup lacked such recurrent gene mutations.
For example, the researchers noted that the TP53-mutated subgroup — which included nearly one-third of the AEL tumors tested — typically occurred in older adult patients with particularly poor outcomes and a lack of long-term survival. On the other hand, the vast majority of patients with NPM1-mutated tumors or increased expression of HOXB9 were long-term AEL survivors. The latter subgroup, which had a good prognosis, encompassed roughly 12 percent of the adult patients profiled.
"Genomic alterations and gene expression profiles were the strongest predictors of outcome in patients with AEL, which suggests they should be incorporated into the diagnostic and prognostic criteria," Mullighan said.