Comparative Assessment of FLT3 Variant Allele Frequency By Capillary Electrophoresis and Next-Generation Sequencing in FLT3^mut+ patients with Relapsed/Refractory (R/R) Acute Myeloid Leukemia (AML) Who Received Gilteritinib Therapy

Background:

The novel fms-like tyrosine kinase 3 (FLT3)/AXL inhibitor, gilteritinib, showed robust antileukemic activity at doses ≥80 mg/day in FLT3 mutation-positive (FLT3^mut+) patients with R/R AML in the CHRYSALIS Phase 1/2 study (NCT02014558). We analyzed variant allele frequency (VAF) in a CHRYSALIS study FLT3^mut+ R/R AML subpopulation with baseline FLT3 internal tandem duplication (ITD) and tyrosine kinase domain (TKD) mutations using two different assays: capillary electrophoresis (CE) and next-generation sequencing (NGS). We assessed analytical concordance and survival outcomes at two FLT3 loci: ITD and D835 mutations in the TKD.

Methods:

Assays were performed on bone marrow and blood samples. The CE (LeukoStrat® CDx) FLT3 mutation assay detected both FLT3 -ITD and D835/I836 FLT3 -TKD mutations; these loci were amplified using polymerase chain reaction. Amplicon size differences distinguished samples with FLT3-ITD mutations from those with wild-type FLT3 . Undigested amplicons following restriction enzyme digestion identified the presence of FLT3-TKD mutations. The capture based NGS assay targeted potential mutation loci in all FLT3 exons. Whole genome libraries were generated and hybridized with a custom probe to capture target fragments, which were sequenced on an Illumina® MiSeq platform. The CE FLT3 signal ratio outputs were converted to estimated VAF values for comparison with NGS outputs. For FLT3-ITD analysis, both assays were optimized to detect large ITD mutations. Median overall survival (OS) was determined for patients who received ≥80 mg/day gilteritinib, specifically among those with FLT3-ITD with/without concomitant TKD mutations, and those with TKD mutations only. Median VAF values were calculated; patient VAFs were stratified according to whether they were greater than/equal to, or less than the median value. Lastly, OS of patients with a FLT3-ITD mutation detected by NGS was stratified using a ≥5% VAF cut point.

Results:

Bone marrow and blood samples from 241 patients were analyzed. The two assays demonstrated strong concordance for both qualitative (mutation detection) and quantitative (VAF) measures at both FLT3-ITD and FLT3-TKD loci. Using a VAF threshold of ≥0.05 and the most prominent ITD in each sample, FLT3-ITD detection by CE or NGS showed a strong concordance in 98.8% (238 of 241) of samples (155 and 156 positives by CE and NGS, respectively). All 3 discordant samples had VAFs that were near (±0.02) the 0.05 VAF threshold and within the 95% CI values for both assays. Strong concordance for FLT3-ITD detection was maintained when multiple ITDs were observed within a sample; a strong linear correlation between both assays was observed for mutation frequencies (R²=0.987) (Figure 1). At a VAF threshold of 0.05, FLT3-TKD mutation concordance was observed in 98.3% (236/240) of samples (34 and 30 positives by CE and NGS, respectively). The 4 discordant samples were near (±0.04) the 0.05 VAF threshold and there was a strong a linear correlation between observed mutation frequencies (R²=0.906).

At a VAF threshold of ≥5% for FLT3 mutation positivity, there was no apparent difference in OS in the ≥80 mg/day gilteritinib subgroup among those with FLT3-ITD with/without concomitant TKD mutations who had VAF values that were greater than/equal to or less than the median values (Figure 2A and 2B), regardless of the assay used. Patients who had a TKD mutation only with VAF values greater than/equal to the median value had longer OS than patients with VAF values less than the median value. Patients who had a TKD mutation only with VAF values greater than/equal to the median value had OS durations similar to those in patients with ITD mutations. By NGS, patients with a FLT3 -ITD VAF <5% had similar OS duration as patients with a FLT 3-ITD VAF ≥5% (Figure 3).

Conclusions:

We observed high concordance using both CE and NGS methods to measure FLT3 mutational burden in patients with R/R AML who received gilteritinib in the CHRYSALIS study. Analytical concordance between the 2 assays was high and survival outcomes with stratification based on FLT3 mutational burden were similar with both assays. Our results suggest that FLT3-ITD^mut+ R/R AML patients may benefit from gilteritinib therapy, regardless of FLT3 mutational burden. In patients with FLT3-TKD mutations only, a higher mutational burden may be associated with gilteritinib benefit; this finding will be further explored in future studies.

View largeDownload slide

View largeDownload slide

Close modal