Multimodal dissection of infant acute lymphoblastic leukemia

Acute leukemia arising in infants, i.e. children aged 0-12 months, is characterized by KMT2A-gene rearrangements (KMT2A-r) and a very poor prognosis with 70% of patients succumbing to disease. The cause of this poor prognosis is not known. Although over 90% of infants reach clinical remission, they are prone to rapid relapse, and our understanding of the mechanisms driving relapse remains limited. The overall goal of this proposal is to improve our mechanistic understanding of infant leukemia development and relapse and thereby impact how it is diagnosed and treated in the future.

Molecular genetic studies, in recent years accelerated by high-throughput sequencing technologies, have significantly improved our understanding of the molecular pathogenesis of acute leukemia. However, these methods sequence cells mixed together and typically do not have the resolution to identify and characterize distinct subpopulations. Some of these cell subpopulations may play a fundamental role in disease evolution and progression and in acquired resistance to treatment. Recently, single cell RNA-sequencing (scRNA-seq) and single-cell DNA sequencing have provided unique possibilities to study these processes.

Herein, we have used scDNA- and scRNA-sequencing to study lineage plasticity. One of our KMT2A-r ALL patients switched to mixed phenotype leukemia 15 days into treatment and our preliminary data have provided a unique insight into the biology of lineage switching. The data show that genetic evolution is independent of the cellular immunophenotype, suggesting an inherent lineage plasticity of the transformed cells. Further, the molecular and cellular characteristics of the cells that expanded on day 15 were shared by a small subset of diagnosticcells (manuscript in preparation).

We have also completed a study of the genetic mechanisms of relapse in infant and childhood KMT2A-r acute leukemia (In revision, Ahlgren L et al). We performed whole genome and exome sequencing of relapsed KMT2A-r cases (n=33), and longitudinal deep-seq of 258 samples from 30 patients. Somatic alterations in drug-response genes, most commonly in TP53 and IKZF1 (75%), were highly enriched in early relapse ALL (83%, 9-36 months after diagnosis), but virtually absent in very early relapse ALL (<9 months). By single-cell DNA sequencing we could determine the mutational order as well as which mutations that resided together at relapse. We could also track residual leukemia cells, clonal drug responses, and the upcoming relapse, thereby providing new insights into mechanisms of relapse in this highly lethal form of pediatric acute leukemia.

‍