Abstract
Acute myelogenous leukemia (AML) is a fatal disease with dismal outcomes. Despite aggressive chemotherapy regimens, most patients relapse and ultimately succumb to their disease. Relapse is thought to be driven by a surviving residual population of leukemia stem cells (LSCs) that remain quiescent and survive chemotherapy. Moreover, relapsed disease has been shown to be frequently more genetically complex, acquiring new mutations not seen in de novo disease. Since the acquisition of new mutations and survival of LSCs are both evident with relapse, we hypothesized that genetic diversity of LSCs is a likely contributor to disease outcomes. To assess diversity, we performed ultra-deep targeted next generation sequencing at up to 15000x non-duplicate depth of coverage, constructed local haplotypes, and performed subclonal variant analysis on AML samples across varying molecular and cytogenetic risk strata. The resulting high resolution molecular portrait enabled us to examine the clonal composition of both lymphocytes and leukemia, including LSCs.
Consistent with previous reports, we detected established pre-leukemic mutations in DNMT3A(R882H) evident in both patient-matched sort-purified lymphocytes and leukemic cells. In this patient, the pre-leukemic acquisition of DNMT3A(R882H) observed in lymphocytes was followed by additional acquisition of NRAS(G13D) seen exclusively in the LSC and non-LSC compartments, but not the sorted lymphocyte compartment. Interestingly, we found that LSC populations frequently exhibited more complex clonal architecture and nucleotide diversity (p < 10-18) compared to non-stem matched patient counterparts. Because LSCs inherently resist chemotherapy, genetic lesions occurring in the surviving LSC pool are capable of supporting sustainable outgrowths of new drug resistant clones. Indeed, within this complex stem cell architecture, we identified deeply subclonal lesions in clinically relevant AML genes including FLT3. For example, in one high risk patient, the drug-resistant tyrosine kinase domain mutant FLT3(D835Y) was evident at below 0.25% allele frequency in the LSC compartment. This FLT3 allele demonstrated penetrance into the bulk tumor, but not into normal lymphocytes, indicating the AML-specific context of this subclonal variant.
The data suggest that, in addition to surviving chemotherapy, LSCs represent a diverse reservoir of surviving residual cells that can support emergence of drug resistant mutants and refractory relapse. Moreover, ultra-deep sequencing of LSCs and bulk disease will enable improved precision medicine approaches to enhance AML outcomes via early detection of emergent and clinically impactful mutant alleles.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.