Prognostic Impact of Subclinical FLT3-ITD Microclones in Patients with Acute Myeloid Leukemia

Background

Patients with acute myeloid leukemia (AML) without FLT3 internal tandem duplication (ITD) at diagnosis often relapse with FLT3-ITD. Therefore, we evaluated the frequency of FLT3-ITD microclones in FLT3-ITD negative patients to evaluate the prognostic impact of subclinical FLT3-ITD microclones in a large cohort of AML patients.

Methods

Patients with AML who underwent allogeneic hematopoietic cell transplantation (alloHCT) at Hannover Medical School between 2000 and 2023 were included, if FLT3-ITD was negative by fragment length analysis at diagnosis (FLT3-ITD mutant to wildtype ratio below 0.05). Ultrasensitive error-corrected next-generation sequencing (NGS) using an established amplicon sequencing approach was used to identify FLT3-ITD clones with high sensitivity. Three different programs using independent methods (filt3r, getITD, custom script) were applied to ensure the reproducibility of results. DNA was sequenced on a MiSeq instrument to achieve coverage of at least 100,000 reads. Diagnostic samples were also sequenced with a panel of 48 recurrently mutated genes. All patients with FLT3-ITD microclones were evaluated for FLT3-ITD before and 2-8 months after alloHCT, and at relapse, if applicable.

Results

A total of 227 FLT3-ITD negative patients were included. The median age was 54 years (range 18-75 years), and 45% were female. 16%, 33%, and 51% were classified to ELN 2022 favorable, intermediate, and adverse risk groups. All patients received intensive induction chemotherapy; none were treated with an FLT3 inhibitor during induction. The median sequencing depth of FLT3 was 202,063 aligned reads. At least one FLT3-ITD sequence read was detected in 57 (25.1%) patients, with a VAF ranging from 0.00041% to 3.28%. In 40 (17.6%) and 17 (7.5%) patients one or more than one FLT3-ITD clones were detected, respectively. We set a threshold of ≥0.01% VAF, corresponding to at least 50 read counts of FLT3-ITD, to define the presence of a microclone for clinical correlations. Using this threshold, 23 (10%) patients had a FLT3-ITD microclone.

Clinical and genetic characteristics of patients with (n=23) and without (n=204) FLT3-ITD microclones were similar, except a higher frequency of patients with ELN 2022 favorable risk (48 vs 12%), a lower proportion of patients with adverse cytogenetic risk (4 vs 24%), and a higher proportion of patients with NPM1 (58 vs 11%) and DNMT3A mutations (37 vs 15%) in the cohort of patients with FLT3-ITD mutations.

After induction complete remission (CR) or CR with incomplete hematologic recovery (CRi) was achieved by 18 (78%) and 138 (68%) patients with and without FLT3-ITD microclones (P=0.28). FLT3-ITD was detected at this time point in one of 18 (6%) patients with microclones with a VAF of 15.4% and a different FLT3-ITD clone from diagnosis. Eight patients with microclones relapsed before undergoing alloHCT and achieved a second CR/CRi with salvage chemotherapy. FLT3-ITD was detected at second CR/CRi in two of these eight (25%) patients with VAFs of 0.32% and 0.003% of the same FLT3-ITD clone from diagnosis. Five patients did not reach CR/CRi; three of these patients showed persisting FLT3-ITD clones from diagnosis with VAF of 20%, 11.9%, and 0.08%, respectively.

Five patients with a VAF ≥0.01% at diagnosis, who relapsed after alloHCT, were further evaluated for FLT3-ITD. Three patients were FLT3-ITD positive at relapse with VAFs of 72.9%, 35.1% and 0.03%; only the latter had reached CR before alloHCT. The identical FLT3-ITD clone from diagnosis was found in the relapse samples of all three patients.

After a median follow-up of 3.2 years, relapse-free survival (RFS) was significantly shorter for patients with FLT3-ITD microclones compared to patients without microclones (HR 1.98, 95% CI 1.04-3.78, P=0.038; 2-year RFS 41% vs 62%). Overall survival (OS) was similar between patients with and without microclones (HR 1.12, 95% CI 0.65-2.05, P=0.7; 2-year OS 65% vs 69%).

Conclusions

FLT3-ITD microclones are frequently found at diagnosis in patients who are considered FLT3-ITD negative by conventional diagnostic methods. FLT3-ITD microclones are associated with NPM1 and DNMT3A mutations, a favorable ELN 2022 risk, and shorter RFS. Microclones with VAFs ≥0.1% at diagnosis may give rise to FLT3-ITD positive relapse. Detection of FLT3-ITD microclones may provide the basis to evaluate the clinical efficacy of FLT3 inhibitors in FLT3-ITD negative AML patients.