Introduction:

The current prognostic stratification of acute myeloid leukemia (AML) patients recognizes several recurrently mutated genes in AML as prognostic markers at the time of diagnosis. However, previous studies indicate that additional prognostic information can be further obtained by detection of pre-leukemic mutations that persist in clinical remission and represent clonal hematopoiesis.

Aim:

To analyze persistence of pre-leukemic mutations in AML patients at the time of disease remission and assess their prognostic relevance.

Methods:

Paired diagnosis and remission samples were analyzed for 114 AML patients; all carried at least 1 pre-treatment mutation at the time of diagnosis. Remission samples were collected 1 - 7 months from the start of conventional induction therapy (median 3 months, none of the patients received transplantation therapy prior to the sampling). All patients achieved at least hematological remission or deeper molecular remission that was identified by standard minimal residual disease monitoring for NPM1 or FLT3-ITD mutations, KMT2A-rearragements, or RUNX1-RUNX1T1 and CBFB-MYH11 fusion genes. Sequencing of 47 exons in 19 genes was performed by targeted amplicon sequencing with ClearSeq AML Haloplex panel (Agilent) on MiSeq and NextSeq instruments (Illumina); MPL gene was excluded for inadequate coverage. Sequencing data were processed using VarDict, VarScan and Pindel algorithms. Mutation detection threshold was set to 2% variant allele frequency (VAF). Detection of CEBPA mutations was confirmed by sanger sequencing.

Results:

Pre-leukemic mutations persisting in remission were detected in 38/114 patients (33%). In 31/38 (82%) patients, the persistence was accompanied by loss of other mutation/s. Persisting mutations were most frequently detected in genes: DNMT3A (mutation persistence in 24 of 41 patients with pre-treatment DNMT3A mutation, 58%), IDH2 (5/19 patients, 26%), TET2 (4/12 patients, 33%) and ASXL1 (5/11 patients, 45%). On the contrary, complete clearance of all mutations in remission below 2% VAF was observed for the following genes: NPM1, FLT3, NRAS, CEBPA and IDH1, except for a single persisting FLT3-TKD mutation. The median VAF of persisting mutations was 18% (range 2 - 94%). Moreover, in almost half of the patients (17/38, 45%), VAF of persisting mutation exceeded 25% - thus more than half of the hematopoiesis was clonal at remission.

Persistence of any mutation in remission was associated with impaired overall survival (OS) and event-free survival (EFS), compared to patients with no persisting mutations (2-year OS: 50% vs 81%, p=0.001; 2-year EFS: 28% vs 57%, p=0.0007). Persistence of 2 or more mutations at remission did not show further effect on OS or EFS compared to patients with exactly one persisting mutation.

DNMT3A was the single persisting gene in remission in 20/38 (53%) patients, and was associated with shorter EFS but not OS compared to patients without any persisting mutations (2-year EFS: 26% vs 57%, p=0.03). Persistence of mutations in other genes than DNMT3A was detected in 14/38 (37%) patients and was associated with shorter OS but not EFS compared to patients without any persisting mutations (2-year OS: 37% vs 81%, p=0.001).

Conclusions:

One third of AML patients in our cohort carried pre-leukemic mutations in remission. These persisting mutations were retained at high VAFs and were limited to specific genes, while at the same time mutations in other genes were cleared by AML therapy. This supports the evidence that persisting mutations represented surviving clonal hematopoiesis, not minimal residual disease. The persistence of mutations in remission showed prognostic impact, however analysis of a larger cohort of patients is required to assess how to combine results from the mutational screening at diagnosis and remission to provide the best prognostic stratification.

Supported by Ministry of Health of the Czech Republic, grant nr. 15-25809A. All rights reserved. This report was written with the support of the Specific University Research (nr. MUNI/A/0968/2017) provided by MEYS.

Disclosures

Mayer:Novartis: Research Funding; Eisai: Research Funding; Johnson & Johnson: Research Funding.

Author notes

*

Asterisk with author names denotes non-ASH members.

This icon denotes a clinically relevant abstract

Sign in via your Institution