Panel-based RNA fusion sequencing improves diagnostics of pediatric acute myeloid leukemia

Abstract

New methods like panel-based RNA fusion sequencing (RNA-FS) promise improved diagnostics in various malignancies. We here analyzed the impact of RNA-FS on the initial diagnostics of 241 cases with pediatric acute myeloid leukemia (AML). We show that, compared to classical cytogenetics (CCG), RNA-FS reliably detected risk-relevant fusion genes in pediatric AML. In addition, RNA-FS strongly improved the detection of cryptic fusion genes like NUP98::NSD1, KMT2A::MLLT10 and CBFA2T3::GLIS2 and thereby resulted in an improved risk stratification in 25 patients (10.4%). Validation of additionally detected non-risk-relevant high confidence fusion calls identified PIM3::BRD1, C22orf34::BRD1, PSPC1::ZMYM2 and ARHGAP26::NR3C1 as common genetic variants and MYB::GATA1 as recurrent aberration, which we here describe in AML subtypes M0 and M7 for the first time. However, it failed to detect rare cytogenetically confirmed fusion events like MNX1::ETV6 and other chromosome 12p-abnormalities. As add-on benefit, the proportion of patients for whom measurable residual disease (MRD) monitoring became possible was increased by RNA-FS from 44.4 to 75.5% as the information on the fusion transcripts' sequence allowed the design of new MRD assays.

Fig. 1. Distribution of risk-relevant fusion genes (RRFGs) (based on AML-BFM registry 2017) in pediatric AML patients (n = 241) regarding detection by RNA-FS, CCG or both.

In 45.6% of patients (n = 110) RRFGs were detected by RNA-FS and CCG, in 8.3% (n = 20) by RNA-FS only and in 3.7% (n = 9) by CCG only. RNA-FS was able to describe KMT2A-r in more detail in 12 samples (5.0%). In 37.3% (n = 90) no RRFG was detected by either of the methods. CCG classical cytogenetics, KMT2A-r KMT2A rearrangement, RNA-FS panel-based RNA fusion sequencing, RRFGs risk-relevant fusion genes (excluding MECOM::RPN1).

Fig. 2. Frequency of risk-relevant fusion genes (RRFGs) (n = 152) of pediatric AML (based on AML-BFM registry 2017) detected by RNA-FS, CCG or both.

CCG classical cytogenetics, RNA-FS panel-based RNA fusion sequencing, RRFGs risk-relevant fusion genes (excluding MECOM::RPN1).

Fig. 3. RNA-FS identified common genetic variants and recurrent non-RRFGs in a cohort of 241 pediatric AML patients.

RNA-FS panel-based RNA fusion sequencing, non-RRFGs non-risk-relevant fusion genes, * reciprocal fusion transcript in same patient detected.

Fig. 4. Impact of RNA-FS on MRD monitoring assay design.

CCG, NPM1-mutational status and sequence information obtained by RNA-FS, enabled patient-specific MRD monitoring assay design in 75.5% of pediatric AML patients (n = 241). Without RNA-FS, MRD monitoring was possible for 44.4% of patients due to previously published assays. MRD measurable residual disease, RNA-FS panel-based RNA fusion sequencing, w/o without.