Internal tandem duplications (ITD) of the FMS-like Tyrosinkinase 3 ( FLT3 ) are detected in up to 30% of all newly diagnosed acute myeloid leukemia (AML) cases, with a decreasing prevalence in elderly patients [ 1 ]. FLT3 -ITD positivity in AML patients predicts poor prognosis and shorter median overall survival (OS), corresponding to an intermediate genetic risk classification per the ELN 2024 less-intensive score [ 2 ]. There are promising new studies examining triplet therapies combining hypomethylating agents (HMA), venetoclax and targeted FLT3 -inhibitors for unfit FLT3 -ITD positive patients [ 3 ]. Interestingly, FLT3-ITD microclones (defined by an allelic ratio (AR) < 0.05) have recently been detected in up to 25% of young AML patients considered fit for intensive chemotherapy who were classified as FLT3 -ITD negative by conventional fragment length analysis [ 4 , 5 , 6 ]. FLT3 -ITD microclones co-occur with NPM1 and DNMT3A mutations and have been linked to inferior relapse-free survival (RFS) as well as higher measurable residual disease (MRD) levels in younger patients with NPM1-mutated AML [ 4 , 5 , 6 , 7 ]. Currently, the incidence and prognostic impact of FLT3 -ITD microclones in elderly, unfit patients remain unclear. Given the prognostic relevance of FLT3 -ITD microclones in younger cohorts and its potential therapeutic implications, we sought to evaluate the impact of FLT3 -ITD microclones in elderly individuals. 201 AML patients aged ≥65 years at diagnosis were included in this study: patients received first-line less-intensive therapy or best supportive care (BSC) at our institution between 2006 and 2026 and were FLT3 -ITD negative in fragment length analysis at initial diagnosis with an AR < 0.05 (Supplementary Fig. S1 ). The median age at diagnosis was 76 years (range 65–90), and 125 patients were male (62%). 95 patients (47.3%) were treated with HMA-monotherapy, 67 patients (33.3%) with HMA plus combination therapy (+venetoclax [ n = 62]/ivosidenib [ n = 5]), 38 patients (18.9%) with best supportive care (BSC) only, and complete treatment data were not available for one patient (0.5%). Targeted ultrahigh-sensitivity error-corrected next-generation sequencing (NGS) and fragment length analysis by capillary electrophoresis was performed for all patients at initial diagnosis, as previously described [ 8 , 9 ]. The presence of FLT3 -ITD by NGS was analyzed with three different methods to enhance sensitivity and specificity, the open-source algorithms filt3r, getITD and a custom script [ 10 , 11 ]. NGS-based myeloid panel and cytogenetic analysis was conducted as described before [ 8 , 12 ]. The median sequencing depth of our ultrasensitive NGS-assay was 161.392 reads per sample (range 51.629–474.239). FLT3 -ITD microclones were detected in 21 out of 201 (10.4%) patients at initial diagnosis with a median VAF of 0.035% (range 0.007–1.153%) (Supplementary Fig. S2 ). 16 out of 21 patients (76.2%) harbored a single FLT3 -ITD insertion, two patients (9.5%) had two FLT3 -ITD insertions, and three patients (14.3%) had three to five FLT3 -ITD insertions. Insertion length ranged from 6 to 102 nucleotides, with a median insertion length of 33 nucleotides. When comparing the ultrahigh-sensitivity NGS results with fragment length analysis, only 6 of 21 (28.5%) FLT3 -ITD microclones identified by NGS were detected by fragment length analysis with an AR below 0.05. Notably, the FLT3 -ITD microclones detected by both methods showed a higher VAF with a median of 0.617% (range 0.023–1.153%), whereas the microclones missed by fragment analysis were smaller with a median VAF of 0.031% (range 0.007–0.293). Patients with FLT3 -ITD microclones demonstrated a significantly higher white blood cell (WBC) count at diagnosis compared to patients without FLT3 -ITD microclones (23.9 vs. 6.95 10^9/l, p = 0.027), whereas no difference was observed for hemoglobin (9 vs. 8.9 10^9/l, p = 0.517) and platelets (67 vs. 47.5 10^9/l, p = 0.502) (Table 1 ). We found no significant difference for age, sex and AML type. According to the ELN 2024 classification for patients receiving less-intensive therapy, patients with FLT3 -ITD microclones were more frequently classified as intermediate risk (38.1% vs. 17.2%) and less frequently as adverse risk (0% vs 25%) compared to patients without FLT3 -ITD microclones ( p = 0.019). Cytogenetic analysis revealed a trend toward a lower incidence of a complex aberrant karyotype in patients with FLT3 -ITD microclones compared to those without FLT3 -ITD microclones, although the difference was not statistically significant (4.8% vs 27.8%, p = 0.068). Evaluation of molecular co-mutations showed a trend toward a lower frequency of TP53 mutations in patients with FLT3 -ITD microclones (0% vs. 25.6%, p = 0.11). Furthermore, we did not detect a higher frequency of NPM1 mutations in the FLT3 -ITD microclone cohort (10% vs. 13.6%, p = 0.817). However, patients with FLT3 -ITD microclones showed a significantly higher prevalence of DNMT3A ( p = 0.02), SMC3 ( p = 0.02) and NRAS ( p = 0.03) mutations, as well as numerical increase in BCOR mutations ( n = 0.16) (Table 1 , Fig. 1A ) . Fig. 1: Genetic landscape and prognostic impact of FLT3 -ITD microclones. Full size image A Comparison of the mutational landscape between FLT3 -ITD microclone positive and negative patients. Myeloid panel data were missing for one FLT3 -ITD microclone-positive and four FLT3 -ITD microclone-negative patients. B Overall survival (OS) between FLT3 -ITD microclone positive and negative patients. Table 1 Baseline characteristics of elderly AML patients with or without FLT3 -ITD microclones. Full size table We next investigated the prognostic impact of FLT3 -ITD microclones in elderly AML patients receiving non-intensive therapy, particularly given that the current ELN risk classification for less-intensively treated patients considers FLT3 -ITD positivity as an adverse classifier. In the whole group, there was no difference in overall survival (OS) between patients with vs without FLT3 -ITD microclones ( p = 0.97) (Fig. 1B ). However, among patients receiving BSC only, those with FLT3- ITD microclones had a significantly shorter OS than FLT3 -ITD negative patients (HR 2.8, CI 1.006–7.808, p = 0.04) (Supplementary Fig. S3A ). This difference in OS was not observed in patients who received HMA-monotherapy ( p = 0.64) or HMA-based combination therapy ( p = 0.68) (Supplementary Fig. S3B –D). To characterize the clonal evolution of FLT3 -ITD microclones, available follow-up samples from 7 FLT3 -ITD microclone positive patents at different timepoints during treatment were analyzed (Supplemental Figs. S2 , S4 ). One Patient (patient 20) achieved complete remission after four cycles of azacitidine and venetoclax and lost the FLT3 -ITD microclone upon remission. Another patient (patient 15) showed re-emergence of peripheral blasts after two years of decitabine treatment and the FLT3 -ITD microclone clone expanded from 1.15 to 9.52%. Two patients harboring small FLT3 -ITD microclones of 0.029% (patient 8) and 0.0074% (patient 6) lost their microclones following refractory disease after one and 8 cycles of HMA monotherapy, whereas three other patients (patient 17, 16 and 14) with refractory disease showed an expansion of their FLT3 -ITD microclone after ≤4 cycles of HMA-based therapy. We next used ultrahigh-sensitivity NGS to evaluate whether patients initially assessed as FLT3 -ITD microclone negative acquired a FLT3 -ITD mutation at the time of relapsed or refractory (R/R) disease. Among 19 initially FLT3 -ITD microclone negative patients with available R/R samples, no detectable FLT3 -ITD mutations emerged during R/R disease. Taken together, our analysis using ultrahigh-sensitivity NGS detected FLT3 -ITD microclones in 10.4% of elderly AML patients receiving less-intensive therapy who were previously classified as FLT3 -ITD negative by conventional fragment length analysis. The proportion of patients with FLT3 -ITD microclones was lower than reported in younger, intensively treated patients, consistent with a lower prevalence of FLT3 -ITD mutations in elderly AML patients [ 1 ]. Fragment length analysis detected only 28.5% of these microclones, all at an AR < 0.05, reinforcing that NGS provides higher sensitivity for identifying small FLT3 -ITD clones. Compared with FLT3 -ITD-negative patients, elderly AML patients with FLT3 -ITD microclones tended to have a lower incidence of TP53 mutations and were less frequently classified as ELN adverse risk. Additionally, the lower prevalence of complex karyotypes in microclone-positive patients suggests a distinct biological background with a weaker association with secondary, pre-existing myeloid conditions than typically seen in elderly patients. Interestingly, we did not find an association between FLT3 -ITD microclones and NPM1 -mutations, although this association has been reported for FLT3 -ITD microclones in younger AML patients treated with intensive chemotherapy [ 5 , 6 ]. Nevertheless, the previously described co-occurrence with DNMT3A was also observed in our cohort of elderly FLT3 -ITD microclone-positive patients. In addition, we identified an association between FLT3 -ITD microclones and mutations in NRAS , another gene involved in signal transduction, that has not been described in younger AML patients. The higher WBC count at initial diagnosis and inferior OS in patients only receiving BSC suggest a more proliferative phenotype in elderly AML patients with FLT3 -ITD microclones, highlighting the proliferation-enhancing potential of even small FLT3 -ITD clones. However, HMA-based therapy appears to mitigate these OS disparities, suggesting that FLT3 -ITD microclones have less prognostic impact than macroclones, as reflected by the ELN2024 recommendations. Most FLT3 -ITD microclones persisted during R/R disease, with some showing clonal expansion, whereas microclones with a VAF < 0.1% were occasionally lost, even in refractory disease. None of the patients assessed as FLT3 -ITD negative by NGS at diagnosis acquired a FLT3 -ITD mutation at R/R disease, suggesting that FLT3 -ITD mutations at relapse mainly originate from pre-existing subclones. Our study is limited by the small number of patients with FLT3 -ITD microclones, reflecting the rarity of this subgroup, as well as by the limited number of patients treated with HMA plus venetoclax. Further studies in larger cohorts are required to better understand mechanisms of resistance and clonal evolution and to evaluate the potential role of FLT3 -ITD monitoring for early detection of treatment resistance in non-intensively treated patients with FLT3 -ITD microclones detected at diagnosis. The clinical implications of FLT3 -inhibitor therapy in patients with FLT3 -ITD microclones remain unclear. In younger patients with FLT3 -ITD mutations, the addition of FLT3 -inhibitors to intensive chemotherapy has been shown to improve outcomes regardless of AR and clone size [ 9 , 13 ]. Furthermore, the QUIWI-trial demonstrated superior outcomes with quizartinib treatment even among patients considered FLT3 -ITD negative [ 14 ]. A subsequent subgroup analysis revealed a non-significant trend toward longer remission duration with quizartinib in FLT3 -ITD microclone-positive patients [ 15 ]. In summary, our data shows that FLT3 -ITD microclones occur in a substantial proportion of elderly, unfit AML patients, represent a more proliferative type of AML and show a distinct mutational pattern. In our study, FLT3 -ITD microclones have no prognostic implications in elderly patients receiving standard therapy. However, larger studies are required to validate this finding and to clarify whether targeted combination therapies play a role in this patient population.
No takes yet. Share an insight, caveat, or question.
Hupe et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: