AUTHORS
Mélinda Dantec, PharmaD, Laurene Fenwarth, MD2, Lucie Rigolot, MD3, Céline Berthon, MD4, Lionel Ades, MD, PhD5, Delphine Lebon, MD, PhD6, Stéphane de Botton, MD, PhD7, Juliette Lambert8, Thomas Cluzeau, MD, PhD9, Laure Farnault10, Stephanie Nguyen Quoc11, Sylvain Chantepie, MD12, Thorsten Braun, MD, PhD13, Emmanuel Raffoux, MD14, Mael Heiblig, MD15, Alexandra COELHO Jr16, Renaud Buffet, MD17, Isabelle Luquet18, Benjamin Carpentier, MD19, Cecile Pautas20, Regis Peffault De Latour, MD, PhD21, Laetitia Largeaud22, Dominique Penther23, Pierre-Yves Dumas, MD, PhD24, Claire Bories25, Adrien Daniel26, Isabelle Plantier27, Jean-Baptiste Micol, MD, PhD28, Arnaud Pigneux, MD, PhD29, Raphael Itzykson, MD, PhD30, Christian Recher, MD, PhD31, Herve Dombret Sr, MD32, Aline Schmidt, MD33, Claude Preudhomme, PhD34, Eric Delabesse35, Nicolas Duployez, MD36
AFFILIATIONS
1 Lille University Hospital, Lille, France
2 Lille University Hospital, Lille, France
3 Toulouse University Hospital, Toulouse, France
4 Lille University Hospital, Lille, France
5 Saint-Louis Hospital, AP-HP, Paris, France
6 University Hospital of Amiens, Amiens, France
7 Gustave Roussy Cancer Center, Villejuif, France
8 Versailles Hospital Center, Le Chesnay, France
9 Nice University Hospital, Nice, France
10 La Conception Hospital, AP-HM, Marseille, France
11 Pitié-Salpétrière Hospital, AP-HP, Paris, France
12 Caen University Hospital, Caen, France
13 Avicenne Hospital, AP-HP, Paris, France
14 Saint-Louis Hospital, AP-HP, Paris, France
15 Hospital Lyon Sud, Lyon, France
16 ALFA coordination, Hôpital Saint-Louis, Université Paris Cité, Paris, France
17 ALFA coordination, Hôpital Saint-Louis, Université Paris Cité, Paris, France
18 Toulouse University Hospital, Toulouse, France
19 GHICL, Lille, France
20 Henri Mondor Hospital, AH-HP, Paris, France
21 Saint-Louis Hospital, AP-HP, Paris, France
22 Toulouse University Hospital, Toulouse, France
23 Henri Becquerel Cancer Center, Rouen, France
24 Bordeaux University Hospital, Bordeaux, France
25 Lens Hospital Center, Lens, France
26 Dunkirk Hospital Center, Dunkirk, France
27 Victor Provo Hospital Center, Roubaix, France
28 Gustave Roussy Cancer Center, Villejuif, France
29 Bordeaux University Hospital, Bordeaux, France
30 Saint-Louis Hospital, AP-HP, Paris, France
31 Toulouse University Hospital, Toulouse, France
32 ALFA coordination, Hôpital Saint-Louis, Université Paris Cité, Paris, France
33 Angers University Hospital, Angers, France
34 Lille University Hospital, Lille, France
35 Toulouse University Hospital, Toulouse, France
36 Lille University Hospital, Lille, France
ABSTRACT
Context: KMT2A partial tandem duplication (KMT2A-PTD) occurs in 3–10% of adults with acute myeloid leukemia (AML), but its prognostic significance and biological features remain incompletely defined. We performed an integrated molecular, transcriptomic, and clinical analysis of a large cohort of patients with KMT2A-PTD AML to characterize its biological landscape, identify prognostic factors, and explore therapeutic vulnerabilities.
Methods: Adult patients with KMT2A-PTD AML diagnosed between 2008 and 2026 were identified from the prospective clinical trials BIG-1 (NCT02416388), ALFA-0701 (NCT00927498), and ALFA-0702 (NCT00932412), as well as the real-world AML observational cohorts ALFA-1200 (NCT01966497), ALFA-PPP (NCT04777916) and Hauts-de-France AML Observatory (CSTMT351). Targeted DNA-sequencing was performed using a capture-based NGS panel (Twist Bioscience® library preparation) and sequenced on an NovaSeq6000 (Illumina®). Transcriptomic profiling was performed in a subset of patients by RNA sequencing (Illumina® library prep kit) and sequenced on an NovaSeq6000. Clonal evolution was assessed in paired diagnosis-relapse samples when available. Survival analyses were restricted to patients aged 18–60 years treated with intensive chemotherapy (ICT).
Results: A total of 442 adult patients with KMT2A-PTD AML were included. Median age at diagnosis was 64 years. Most patients had a normal karyotype (41%). The most common cytogenetic abnormality was trisomy 11 (8%). No patient harbored an isolated KMT2A-PTD mutation. Median number of co-mutations per patient was three and the most frequent co-mutations involved IDH1/2 (42.7%), DNMT3A (35%), FLT3-ITD (27%), RUNX1 (23%) and STAG2 (19%). Myelodysplasia-related gene mutations were present in 66% of patients. KMT2A-PTD was mutually exclusive with NPM1 mutations and recurrent gene fusions. According to ELN 2022 classification, 71.2% and 28.8% of patients with KMT2A-PTD AML were classified as adverse- (adv-) and intermediate- (int-) risk respectively. Among 43 paired diagnosis-relapse samples, KMT2A-PTD remained stable in 90%, whereas WT1 and FLT3-ITD were frequently acquired at relapse (18.6% and 13.9%, respectively).
In patients aged 18-60y treated by CTI (n=141), complete remission rate after one course (CR1) was 82.9% and 5-year overall survival (OS) was 53.5% (95% CI, 45.2–63.5). Despite enrichment for adv-risk features, survival was comparable to that of patients classified as int-risk according to ELN 2022 (ELN int-risk vs. KMT2A-PTD AML: HR=0.89, 95% CI, 0.65–1.22, p=ns and ELN adv-risk vs. KMT2A-PTD AML: HR=1.67, 95% CI, 1.24–2.26). Allogeneic hematopoietic stem cell transplantation (allo-HSCT) in CR1 was performed in 71.6% of patients.
In multivariable analyses adjusted for age, genetic alterations, and allo-HSCT in CR1, FLT3-ITD, FLT3-TKD, and IDH1 mutations independently predicted inferior relapse-free survival (HR=4.00, 95% CI, 1.84–8.67; HR=3.54, 95% CI, 1.44–8.67; and HR=7.44, 95% CI, 2.82–19.62 respectively) and inferior OS (HR=2.22, 95% CI, 1.04–4.77; HR=2.52, 95% CI, 1.11–5.70; HR=2.56, 95% CI, 1.15–5.71) whereas allo-HSCT in CR1 significantly improved OS (HR=0.19, 95%CI, 0.10–0.39).
RNAseq analysis performed in patients from the BIG-1 study showed high HOXA/MEIS1 genes expression comparable to KMT2A fusions or NPM1 AML (median expression HOXA/ABL1=82%, MEIS1/ABL1=69%). Among KMT2A-PTD AML, FLT3-ITD-positive patients displayed increased expression of HOXB genes and HOXA9/MEIS1 (p<0.05), increased PRDM16 expression and higher LSC17 scores (p<0.001) compared to FLT3-ITD-negative patients.
Discussion: KMT2A-PTD AML represents a biologically heterogeneous entity characterized by a stable founder alteration, recurrent cooperating mutations, and activation of a HOXA/MEIS1 transcriptional program. Despite its frequent association with adverse-risk molecular features, KMT2A-PTD AML exhibited outcomes consistent with an intermediate-risk disease in younger adults intensivey treated. Interestingly, FLT3-ITD defined a distinct molecular subgroup associated with an increased risk of relapse and death and distinct transcriptional profile characterized by enhanced HOXA/B, MEIS1, and PRDM16 expression, suggesting that FLT3-ITD may act as a disease-modifying lesion in KMT2A-PTD AML. This highlights potential therapeutic vulnerabilities including FLT3 and IDH1 mutations, as well as HOXA/MEIS1 pathway activation, supporting the investigation of FLT3, IDH1 and menin inhibitors.
DISCLOSURES
Mélinda Dantec, PharmaD: No relevant conflicts of interest to declare.