Abstract 1264: Enhanced genomic characterization and clonal profiling in acute myeloid leukemia

Abstract Introduction: Acute myeloid leukemia (AML) is an aggressive blood cancer characterized by serial acquisition of somatic mutations in hematopoietic stem and progenitor cells which lead to their clonal expansion and differentiation arrest. Current methods for genomically profiling AML are limited by (1) the need for serial invasive bone marrow biopsies and (2) paucity of recurrently shared mutations across AML patients using standard NGS sequencing approaches. Methods: We integrate an AML-specific CAPP-Seq (Cancer Personalized Profiling by Deep Sequencing) hybrid capture panel targeting 58 recurrently mutated, ‘canonical’ AML genes, with personalized panels derived from whole exome sequencing (WES) of flow-sorted AML blasts and T cells (germline) in 56 AML patients. We quantify WES-derived single nucleotide variants (SNVs), indels, and copy number alterations (CNAs) in both canonical and ‘noncanonical’ genes (non-recurrently mutated) across a median of 5 follow up timepoints from cellular (bone marrow mononuclear cell/BMMC, peripheral blood mononuclear cell/PBMC) and cell-free (PB-ctDNA, BM-ctDNA) compartments. Variant pathogenicity was assessed using CADD, an ensemble model; ESM-1b, a protein language model; PolyPhen2, SIFT and a novel graphical association method. Mutation signatures were extracted using the SigProfiler suite. Clonal hierarchy was modeled using SciClone and latent class analysis. Results: We observed a median of 32.5 (IQR 21-42) tumor SNVs per patient compared to only 2.5 (IQR 1-4, p < 0.0001) if using the canonical AML panel alone. Overall variant detection was highly concordant across compartments: median of 94% (BMMC vs PBMC), 85% (PB-ctDNA vs BMMC), 94% (PB-ctDNA vs PBMC) and 98% (PB-ctDNA vs BM-ctDNA) respectively. Noncanonical mutations were 99.9% unique across patients and less pathogenic across our ensemble of measures (p < 0.0001). C-to-T transitions were enriched in both canonical and noncanonical mutation sets, but other signatures suggest distinct etiology. Moreover, personalized patient-specific AML-CAPP-Seq enabled better inference of clonal dynamics compared to our institutional standard-of-care bulk NGS assay. Our approach predicted clonal hierarchies supported by single-cell genotyping (Tapestri) and helped predict potential mechanisms of therapy resistance. Finally, we show AML-CAPP-Seq permits monitoring of donor chimerism after allogeneic hematopoietic stem cell transplantation. Conclusions: AML-CAPP-seq enables ultrasensitive, WES-scale, non-invasive detection of patient-specific mutations, revealing novel biology in AML pathogenesis and allowing improved clonal profiling for in-depth characterization of therapy resistance mechanisms. Citation Format: Sanjeeth Rajaram, Ruwan Gunaratne, Crystal Zhou, Emily Yang, Charu Tiwari, Grace Zhang, Ravindra Majeti, Asiri Ediriwickrema, David Kurtz, Tian Yi Zhang. Enhanced genomic characterization and clonal profiling in acute myeloid leukemia [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 1264.