Clinical impact of FLT3 microclones on outcomes in patients with newly diagnosed FLT3-ITD negative AML: A subanalysis from the pethema quiwi trial

Abstract

Abstract Background: FLT3-ITD negative acute myeloid leukemia (AML) remains challenging to treat, with limited targeted therapeutic options available upfront. The PETHEMA group conducted the QUIWI trial (NCT04107727), a 2:1 randomized, double-blind, placebo-controlled (PBO) phase II study evaluating the addition of quizartinib (Quiz) to intensive chemotherapy in patients with FLT3-ITD-negative AML (i.e., allelic ratio <0.03). Median event-free survival (EFS) was 20.4 months and 9.9 months in the Quiz and PBO arms, respectively (P=0.045). Median overall survival (OS) was not reached and 29.3 months in the Quiz and PBO arms, respectively (P=0.01). Recent data suggest that small FLT3-ITD subclones (microclones), even when below conventional detection thresholds, may have clinical impact. Aim: To assess the clinical impact of FLT3 microclones in patients enrolled in the QUIWI trial by evaluating their association with survival outcomes and response to Quiz vs. PBO. Methods: Adults (18-70 years) with newly diagnosed FLT3-ITD-negative AML eligible for intensive chemotherapy were randomized 2:1 to standard induction and consolidation chemotherapy plus Quiz or PBO, followed by 12 months maintenance with Quiz or PBO. Baseline characteristics, CR/CRi, MRD-negativity after induction, EFS, OS, and duration of CR/CRi (DoR), were analyzed according to microclone status (pos/neg) and treatment arm (Quiz vs. PBO). Analyses were performed on an intent-to-treat basis. Next Generation Sequencing (NGS) data were systematically processed using the FiLT3r algorithm (v:90969f4b), which enables the detection of small FLT3 clones with an allelic ratio ≥0.0003%. A microclone was defined as an FLT3 allelic ratio <0.05. Based on this approach, patients were classified as positive microclones when at least one clone was detected by FiLT3r algorithm. Results: Of the 273 randomized patients (Quiz n=180, PBO n=93), 253 (93%) had available results for microclone analysis. FLT3 (ITD juxtamembrane domain [JMD] and/or tyrosine kinase domain [TKD]) clones were identified in 45 patients (18%), while 208 (82%) had no detectable FLT3. Among the 45/253 patients with FLT3 clones (31/165 in Quiz arm and 14/88 in PBO arm), 43 (96%) had microclones, while 2 patients (4%) with higher allelic ratios (0.11 and 0.12, also grouped as microclones for analyses). FLT3 microclones mutations were in the JMD in 29 pts (64%), TKD in 9 (20%), and both JMD/TKD in 7 (16%). The median number of microclones was 1 (range 1-5), with 12 patients (27%) harboring more than one clone. Median sum of allelic ratios was 0.0012 (range 0.0003-0.12), and median variant allele frequency was 0.0011 (range 0.0003-0.11). When compared with those without microclones, patients with FLT3 microclones showed a trend for intermediate-low cytogenetic risk (59.2% vs. 77.8%; P=0.06), favorable ELN risk (37.8% vs. 25.5%; P=0.17), and NPM1 mutations (26.7% vs. 18.3%; P=0.28). In all patients, microclone-pos and -neg patients showed similar induction CR/CRi (77.8% vs. 78.4%) and MRD negativity rates (51.4% vs. 55.2%) (P=0.99 and 0.83, respectively). Median OS was not reached (NR) in the microclone-pos group vs. 41.8 months in the microclone-neg group (HR 0.89; 95%; P=0.63), and a non-significant trend toward improved EFS was observed in microclone-pos patients (33.2 vs. 13.4 months; HR 0.77; P=0.26), with similar results for DoR (NR vs. NR months; HR 0.70; P=0.28). When analyzed by treatment arm, Quiz significantly improved OS and EFS compared to PBO in patients without FLT3 microclones (median NR vs. 20.6 months; HR 0.62; P=0.0202), and there was a non-significant trend toward improved OS among those microclone-pos (NR vs. 27.1 months; HR 0.57; P=0.22). In PBO arm, the 4-year OS, EFS, and DoR in microclone-neg vs. microclone-pos patients were: 42.7% vs. 40.0%; P=0.99; 31.1% vs. 32.1%; P=0.58; and 54.9 vs. 46.9%; P=0.95, respectively. In Quiz arm, the 4-year OS, EFS, and DoR in microclone-neg vs. microclone-pos patients were: 53.4% vs. 64.4%; P=0.59; 42.6% vs. 54.8%; P=0.34; and 58.8 vs. 73.9%; P=0.20, respectively. In microclone-pos patients, the 4-year DoR was 73.9% in Quiz arm and 46.9% in PBO arm (P=0.22). Conclusion: In the QUIWI trial, patients randomized to Quiz had improved outcomes irrespectively of the presence of FLT3 microclones. Acknowledging that the study is limited by the sample size, there was an apparent improvement in DoR among microclone-pos patients randomized to Quiz as compared with PBO.