Characterization of fusion transcripts in AML without recurrent genetic abnormalities unravels new putative fusion genes.

In the last few years, whole-transcriptome sequencing has shown a high number of low-frequency fusion transcripts (FTs) involved in acute myeloid leukemia (AML) pathogenesis. Most of them are not identifiable through conventional diagnostic techniques. In this research, using RNA sequencing, we have investigated FTs in 109 cases of AML without recurrent genetic abnormalities (as defined by the fourth edition of the World Health Organization Classification of Hematolymphoid Tumours). We identified and validated 6 well-known AML-causing FTs (Tier-1), 9 FTs in which recurrently affected genes in AML were involved (Tier-2), and 4 Tier-3 FTs, along with other FTs found in healthy tissue databases (Tier-4). We highlighted 2 previously unknown FTs (ARHGAP11A::NUTM1 and RAP1B::GPC3) that constitute putative driver fusion genes in AML after performing a thorough analysis of their intrinsic properties, expression pattern, and clinical data correlation. Altogether, 15 patients from our cohort (14%) presented at least 1 validated FT, half of which had diagnostic and/or therapeutic implications. Furthermore, we were able to monitor 8 FTs during disease evolution, finding a good correlation with tumor burden. Nevertheless, the significance of many FTs remains unknown, which makes it necessary to enlarge curated FT databases to implement whole-transcriptome sequencing in clinical practice.