NEW YORK – An international team led by investigators at St. Jude Children's Research Hospital, Children's Hospital of Philadelphia, and Children's Oncology Group has spelled out the genomic alterations involved in a high-risk form of leukemia in children.
The team's analysis, published in Nature on Wednesday, looked at genome, exome, and transcriptome sequence data in tumor and matched normal samples from 1,309 cases of childhood T-lineage acute lymphoblastic leukemia (T-ALL) treated uniformly through an international Phase III Children's Oncology Group trial.
The investigators also incorporated ATAC-seq, Hi-C plus chromatin immunoprecipitation, single-cell RNA sequencing, and other data for a subset of T-ALL tumor samples to assess chromatin state, chromatin accessibility, and enhancer effects in half a dozen T-ALL subtypes.
Compared to the more common form of ALL involving B-lineage cells, T-ALL has not been as fully characterized genomically in the past, co-senior authors David Teachey, a researcher affiliated with CHOP and the University of Pennsylvania, and Charles Mullighan, an investigator with St. Jude Children's Research Hospital, said in a joint email comment.
But while T-ALL accounts for around 15 percent to one-quarter of ALL cases, it has been linked to relatively poor outcomes, prompting the team to take a closer look at the potential subtypes, drivers, targetable alterations, and prognostic features that may be gleaned from a set of childhood T-ALL genomes.
"T-ALL is less advanced in this regard, in part due to the suspected genomic complexity and involvement of the noncoding genome as a mechanism of oncogene deregulation, and the need for genomic analysis incorporating whole-genome sequencing," Mullighan and Teachey said, noting that prior studies did not uncover genomic features in T-ALL tumors with ties to patient outcomes.
With their new datasets, the researchers identified 15 gene expression- or genomic driver-based subtypes of T-ALL while providing a more nuanced view of childhood T-ALL developmental trajectories and subtypes.
"The subtypes spanned the continuum of T-cell development," Mullighan and Teachey explained, adding that "some exhibited unexpected cell type similarities."
The team's analyses also unearthed 163 genes with recurrent pathogenic or likely pathogenic changes, along with a slew of noncoding alterations that appeared to alter enhancer sequences or introduce enhancers that did not previously exist.
"The patterns of enhancer deregulation were tightly associated with activity of these enhancers in normal T-cell development, highlighting the stage-specific origins of the disease," Mullighan and Teachey said.
The new data also made it possible for the team to establish two outcome prediction models, based on tumor subtypes and the characteristic alterations they contained as well as an approach that brought together a broader range of genomic changes.
"Using multivariable outcome models, we show that genetic subtypes, driver, and concomitant genetic alterations independently predict treatment failure and survival," the researchers reported, suggesting that the findings "provide a roadmap for the classification, risk stratification, and mechanistic understanding of this disease."
Following on from their results, the investigators plan to validate their outcome prediction models in an upcoming T-ALL Children's Oncology Group study, Mullighan and Teachey said, noting that they are also exploring "the least complex genomic profiling approach that can capture this prognostic heterogeneity most accurately."