Impact of gene mutations on treatment response and prognosis of acute myeloid leukemia secondary to myeloproliferative neoplasms. Issue 3 (6th December 2017)

• Record Type: Journal Article
• Authors: Venton, Geoffroy; Courtier, Frédéric; Charbonnier, Aude; D'incan, Evelyne; Saillard, Colombe; Mohty, Bilal; Mozziconacci, Marie‐Joëlle; Birnbaum, Daniel; Murati, Anne; Vey, Norbert; Rey, Jérôme
• Is Part Of: American journal of hematology. Volume 93:Issue 3(2018:Mar.)
• Page Start: 330
• Page End: 338
• Publication Date: 2017-12-06
• Subjects: Hematology -- Periodicals; 616.15
• DOI: 10.1002/ajh.24973
• Languages: English
• ISSNs: 0361-8609

Abstract

Abstract: Acute myeloid leukemias secondary (sAML) to myeloproliferative neoplasms (MPN) have variable clinical courses and outcomes, but remain almost always fatal. Large cohorts of sAML to MPN are difficult to obtain and there is very little scientific literature or prospective trials for determining robust prognostic markers and efficient treatments. We analyzed event‐free survival (EFS) and overall survival (OS) of 73 patients with MPN who progressed to sAML, based on their epidemiological characteristics, the preexisting MPN, the different treatments received, the different prognostic groups and the responses achieved according to the ELN, and their mutational status determined by next‐generation DNA sequencing (NGS). For 24 patients, we were able to do a comparative NGS analysis at both MPN and sAML phase. After acute transformation EFS and OS were respectively of 2.9 months (range: 0‐48.1) and 4.7 months (range: 0.1‐58.8). No difference in EFS or OS regarding the previous MPN, the ELN2017 prognostic classification, the first‐line therapy or the response was found. After univariate analysis, three genes, TP53, SRSF2 and TET2, impacted pejoratively sAML prognosis at sAML time. In multivariate analysis, TP53 ( P = .0001), TET2 ( P = .011) and SRSF2 ( P = .018) remained independent prognostic factors. Time to sAML transformation was shorter in SRSF2 ‐mutated patients (51.2 months, range: 14.7‐98) than in SRSF2‐ unmutated patients (133.8 months, range: 12.6‐411.2) ( P < .001). Conventional clinical factors (age, karyotype, ELN2017 prognostic classification, treatments received, treatments response, Allo‐SCT…) failed to predict the patients' outcome. Only the mutational status appeared relevant to predict patients' prognosis at sAML phase.