Sequential tracking of FLT3-ITD architecture reveals length-dependant impact on clonal clearance and clinical outcome

Abstract

Abstract Introduction FLT3-ITD mutations are associated with adverse outcomes in acute myeloid leukemia (AML). However, a single baseline measurement of the clone and its characteristics offers only a static snapshot view of an intrinsically dynamic mutation, concealing the clone's clinical trajectory and sensitivity to targeted therapy. This emphasizes the need for longitudinal tracking as well as to study the role of ITD length since it may influence clinical behavior and treatment response in AML. So far, the prognostic role of that data remains unclear due to inconsistent methods and conflicting evidence from largely cross-sectional studies. Importantly, prior studies lacked serial sequencing data, limiting assessment of how ITD length influences mutational trajectories following treatment. Critically, we will test whether longer ITDs—postulated to confer stronger kinase signaling and thus possibly elicit deeper FLT3-inhibitor responses would confer superior survival relative to shorter ITDs, a question that remains unresolved. Methods We screened 1,372 cases from KCI, CCF, and UTSW (2016–2025) and identified 163 patients with FLT3-mutated AML. Targeted NGS (54-gene TruSight Myeloid Panel, Illumina MiSeqDx) was used to assess FLT3 mutation type, ITD length and domain (JMD vs TKD1), and co-mutations. FLT3 inhibitor use, transplant status, and clonal trajectories were collected from clinical records. ITD length was classified as short (S) if <70 bp or long (L) if ≥70 bp. Chi-square tests were used for categorical comparisons and Kaplan-Meier for survival analysis. Results Among 163 FLT3MT cases, 147 had ITD, 13 had TKD, and 3 had both. Among 81 patients with known ITD length, 56 had S-ITDs and 25 had L-ITDs. Median age was similar between S-ITD and L-ITD groups (62 vs 64 yrs, p=0.95), as were gender distribution (45% vs 56% male, p=0.34) and baseline hematologic parameters: WBC (42 vs 39×10³/μL, p=0.75), ANC (3.4 vs 4.0×10³/μL, p=0.63), hemoglobin (8.3 vs 8.3 g/dL, p=0.50), and platelets (52 vs 51×10³/μL, p=0.58). There was no significant difference in overall race distribution between S and L-ITD groups (p=0.79). However, within the cohort, White than Black patients were significantly more enriched in both groups: among S-ITDs and L-ITDs, 71% vs 11% and 68% vs 16% were White and Black (p=0.0001 & p=0.02). Abnormal karyotypes were less frequent than normal in both groups: 25% vs 73% in S-ITD and 20% vs 72% in L-ITD ((p=0.0004 & p=0.00006), with no overall difference between groups (p=0.12). S-ITD patients were more frequently transplanted (54% vs 32%, p=0.002). FLT3i exposure was numerically higher in L-ITD (80% vs 71%), but not statistically significant (p=0.58). Analyzing associated mutations between S-ITD and L-ITD, NPM1MT (59% vs 52%, p=0.6), DNMT3AMT (48% vs 36%, p=0.2) and IDH1MT (13% vs 4%, p=0.27) were not statistically different so as between L-ITD and S-ITD, TET2MT (20% vs 13%, p=0.5), WT1 (16% vs 7%, p=0.2) and NRASMT (12% vs 9%, p=0.6). Domain localization did not differ significantly between groups, with JMD localization seen in 64% of S-ITDs vs 67% of L-ITDs (p=1), and TKD1 in 18% vs 33%, respectively (p=0.11). Post-treatment loss was more common in S-ITDs (71% vs 48%, p=0.03), while persistence favored L-ITD (52% vs 29%, p=0.03). At relapse, L-ITDs were more frequently detected, with gain in 16% vs 11% (p=0.73) and persistence in 20% vs 11% (p=0.33) compared to S-ITDs. L-ITD was linked to longer time to relapse (9 vs 6 months) and significantly improved OS (72 vs 28 months, P=0.0005) vs S-ITD. Among FLT3i – treated patients, L-ITD (n=20) had significantly better OS (60 vs 30 months, P=0.02). L-ITD also showed a trend toward improved OS in both JMD (108 vs 42 months, P=0.06) and TKD1 (84 vs 35 mos, P=0.47) domains. In transplant recipients, L-ITD (n=8) was associated with better OS (64 vs 49 months; P=0.02). Conclusion With mixed results from prior studies on FLT3–ITD length, our findings support that L-ITDs are associated with improved outcomes (OS and RFS) vs S-ITDs, with consistent benefit across FLT3i-treated, post-transplant, and domain-specific (JMD, TKD1) subgroups. Longitudinal data showed greater clonal stability in long ITDs, while short ITDs were more often cleared but re-emerged, suggesting plasticity and potential persistence.