Timing of Response with Gilteritinib Monotherapy in Relapsed or Refractory FLT3-Mutated Acute Myeloid Leukemia

Introduction: Gilteritinib (GILT), an FMS-like tyrosine kinase 3 (FLT3) inhibitor, is approved for the treatment of relapsed/refractory (R/R) FLT3-mutated (FLT3^mut+) acute myeloid leukemia (AML) globally. Although rapid clearance of circulating blasts is uniform, time to marrow blast clearance and count recovery is variable. Many patients treated with GILT have marrow blast clearance after 1–2 months, but slower response is not uncommon and full count recovery can take up to 6 months. Data are needed on the timing of response of GILT, a targeted treatment, compared to salvage chemotherapy, a non-targeted treatment, which typically elicits a response within 2 cycles. We aimed to investigate the timing of response for patients with R/R FLT3^mut+ AML on GILT monotherapy, measured by composite complete remission (CRc), through a post-hoc analysis of the GILT arms of two phase 3, randomized controlled trials – ADMIRAL and COMMODORE.

Methods: All GILT-treated patients from the full analysis sets and safety analysis sets of the ADMIRAL and COMMODORE trials were included. A pooled analysis for both trials was performed. The GILT cycle length in both studies was 28 days. CRc was defined as the sum of complete remission (CR), CR with incomplete hematologic recovery (CRi), and CR with incomplete platelet recovery (CRp). Patients were defined as early responders (ERs, who achieved CRc by the end of cycle 2) and late responders (LRs, who achieved CRc after initiation of cycle 3). Variables analyzed included the proportion of GILT-treated patients who achieved CRc at each treatment cycle, type of CRc, baseline and clinical characteristics, overall survival (OS), and safety profiles by responder status. The hazard ratio (HR) was derived from a univariate analysis.

Results: The pooled analysis included 205 GILT-treated CRc responders consisting of ERs (29.3% [60/205]) and LRs (70.7% [145/205]). Of these, the proportions of CR responders were 21.7% [13/60] and 48.3% [70/145] in the ER and LR groups, respectively. Baseline demographics and clinical characteristics were similar between the two groups. Most CRc responders (82.0% [168/205]) achieved remission within 6 GILT cycles. More LRs (40.7% [59/145]) received on-study hematopoietic stem cell transplantation (HSCT) after GILT treatment compared with ERs (23.3% [14/60]). Among CRc responders who received on-study HSCT, more LRs (69.5% [41/59]) underwent HSCT in CR compared with ERs (21.4% [3/14]). More ERs (78.6% [11/14]) underwent on-study HSCT in CRi or CRp compared with LRs (30.5% [18/59]). The proportion of patients who maintained CRc status and resumed GILT after on-study HSCT was higher in LRs (81.4% [48/59]) than ERs (42.9% [6/14]).

To avoid survivor bias, the median OS was indexed at CRc date and censored at the date of HSCT. The median OS (95% confidence intervals [CI]) was 9 (7, 10) months for ERs and 11 (9, 16) months for LRs. The OS rates (95% CI) at 12 and 24 months were 27.9% (17.1, 39.8) and 20.6% (11.2, 31.9) for ERs and 48.0% (39.5, 56.1) and 32.3% (24.6, 40.2) for LRs (HR [95% CI], 0.704 [0.496, 0.998]). Median (95% CI) treatment duration was 4.7 (3.6, 5.1) and 6.4 (5.6, 7.5) months for ERs and LRs, respectively.

Grade 3 or higher treatment-emergent adverse events (TEAEs), adjusted by the total duration of exposure time, were 22.2 (907 events [E]/40.8 patient-years [PY]) E/PY and 13.5 (2527 E/187.6 PY) E/PY in ERs and LRs, respectively. Drug-related grade 3 or higher TEAEs were 10.6 (434 E/40.8 PY) E/PY in ERs, and 6.5 (1219 E/187.6 PY) E/PY in LRs. Treatment discontinuation due to disease relapse (ERs, 41.7% [25/60]; LRs, 27.6% [40/145]) and progression (ERs, 25.0% [15/60]; LRs, 19.3% [28/145]) was higher in ERs than LRs.

Conclusion: Findings from this pooled analysis indicated that most patients with R/R FLT3^mut+ AML achieved CRc within 6 cycles of GILT monotherapy; LRs were more likely to achieve full count recovery, undergo HSCT in CR and maintain remission status post-HSCT compared with ERs. These findings suggest that continuing GILT monotherapy in patients with R/R FLT3^mut+ AML, especially in those who tolerate the drug, should be considered given the possibility of achieving late positive responses. There is a need to differentiate the timing of response assessment and treatment continuation decisions for GILT from that of salvage chemotherapy.