Evaluation of hypereosinophilia in a case of FLT3-mutant acute myeloid leukemia treated with gilteritinib.

Acute myeloid leukemias (AMLs) frequently harbor activating mutations in Fms-like tyrosine kinase 3 (FLT3). The use of FLT3 inhibitors (FLT3i) is the standard of care for treatment of newly diagnosed and relapsed patients with AML. Differentiation responses including clinical differentiation syndrome have been previously reported with FLT3i when used as single agents in relapsed disease. We present a case of hypereosinophilia in a patient on FLT3i therapy with persistent FLT3 polymerase chain reaction (PCR) positivity in peripheral blood. We sorted mature leukocytes by lineage to determine if the eosinophils were leukemia-derived. FLT3 PCR and next-generation sequencing analysis demonstrated monocytic differentiation of the FLT3-ITD leukemic clone with reactive hypereosinophilia that was derived from a preleukemic SF3B1, FLT3 wild-type clone. Our case is the first to definitively demonstrate the emergence of clonal FLT3-ITD monocytes with FLT3i and the first to demonstrate a differentiation response following decitabine, venetoclax, and gilteritinib triplet therapy.

Development and implementation of an automated and highly accurate reporting process for NGS-based clonality testing.

B and T cells undergo random recombination of the VH/DH/JH portions of the immunoglobulin loci (B cell) and T-cell receptors before becoming functional cells. When one V-J rearrangement is over-represented in a population of B or T cells indicating an origin from a single cell, this indicates a clonal process. Clonality aids in the diagnosis and monitoring of lymphoproliferative disorders and evaluation of disease recurrence. This study aimed to develop objective criteria, which can be automated, to classify B and T cell clonality results as positive (clonal), No evidence of clonality, or invalid (failed). Using clinical samples with "gold standard" clonality data obtained using PCR/CE testing, we ran NGS-based amplicon clonality assays and developed our own model for clonality reporting. To assess the performance of our model, we analyzed the NGS results across other published models. Our model for clonality calling using NGS-based technology increases the assay's sensitivity, more accurately detecting clonality. In addition, we have built a computational pipeline to use our model to objectively call clonality in an automated fashion. Collectively the results outlined below will have a direct clinical impact by expediting the review and sign-out process for concise clonality reporting.