María Isabel Prieto‐Conde,Cristina Jiménez,María García‐Álvarez,Fernando Ramos,Alejandro Medina,Rebeca Cuello,Ana Balanzategui,José M. Alonso,Maria Eugenia Sarasquete,José Antonio Queizán,Miguel Alcoceba,Abelardo Bárez,Noemí Puig,Alberto Cantalapiedra,Norma C. Gutiérrez,Ramón García‐Sanz,Marcos González‐Díaz,María Carmen Chillón

Abstract:Recommended genetic categorization of acute myeloid leukaemias (AML) includes a favourable‐risk category, but not all these patients have good prognosis. Here, we used next‐generation sequencing to evaluate the mutational profile of 166 low‐risk AML patients: 30 core‐binding factor (CBF)‐AMLs, 33 nucleophosmin (NPM1)‐AMLs, 4 biCEBPα‐AMLs and 101 acute promyelocytic leukaemias (APLs). Functional categories of mutated genes differed among subgroups. NPM1‐AMLs showed frequent variations in DNA‐methylation genes (DNMT3A, TET2, IDH1/2) (79%), although without prognostic impact. Within this group, splicing‐gene mutations were an independent factor for relapse‐free (RFS) and overall survival (OS). In CBF‐AML, poor independent factors for RFS and OS were mutations in RAS pathway and cohesin genes, respectively. In APL, the mutational profile differed according to the risk groups. High‐risk APLs showed a high mutation rate in cell‐signalling genes (P = 0·002), highlighting an increased incidence of FLT3 internal tandem duplication (ITD) (65%, P < 0·0001). Remarkably, in low‐risk APLs (n = 28), NRAS mutations were strongly correlated with a shorter five‐year RFS (25% vs. 100%, P < 0·0001). Overall, a high number of mutations (≥3) was the worst prognostic factor RFS (HR = 2·6, P = 0·003). These results suggest that gene mutations may identify conventional low‐risk AML patients with poor prognosis and might be useful for better risk stratification and treatment decisions.

Identification of relapse‐associated gene mutations by next‐generation sequencing in low‐risk acute myeloid leukaemia patients