Abstract Acute myeloid leukemia (AML) is a complex disease with highly individualized treatment plans for patients. Mostly these are based on genetic and phenotypic markers of the disease, but the patient’s overall condition is playing a significant role as well. To develop innovative drugs for a broad patient cohort, it is essential to understand the biological implications of these differences and to model them preclinically. In this study, we are assessing the utility of EpiCypher’s CUT&Tag-FFPE platform to characterize the epigenome of AML PDX samples. CUT&Tag-FFPE is an epigenome profiling assay that is compatible with Formalin-Fixed Paraffin-Embedded (FFPE) samples, enabling robust genome-wide profiling of histone modifications from limited clinical samples. We derived PDX model LEXFAM 2799 (NPM1 wt; FLT3 wt, t(8;21)) and LEXFAM 2966 (NPM1 mut; FLT3 wt) from treatment naïve patients and established in immunocompromised NSG mice. LEXFAM 2799 is a fast-growing model with passaging times of 33 days, whereas LEXFAM 2966 displays a passaging time of 140 days. We next treated with Standard of Care (SoC) agents Decitabine, Cytarabine and all-trans-retinoic acid (ATRA): LEXF 2966 was highly sensitive towards Decitabine leading to complete remission. Cytarabine induced a partial remission and ATRA had only minor effect on tumor growth. In contrast LEXFAM 2799 was resistant towards treatment with ATRA and showed only partial remission under treatment with Cytarabine or Decitabine. Using the CUT&Tag-FFPE platform we are analyzing FFPE slides from the above mentioned two AML PDX models to investigate differences in the genome-wide histone profile that relate to the different growth behavior as well as sensitivity towards SoC treatment. The goal of these studies is to identify tumor model specific profiles that correlated specifically with the sensitivity towards the hypomethylating agent Decitabine. In addition, we plan to analyze additional untreated as well as treated AML PD models to validate the identified signatures confirming them as predictive biomarkers for sensitivity towards current standard of care and prospectively as well innovative compounds. Citation Format: Alysha Simmons, Eva Oswald, Georg Kuales, Michael Keogh, Vishnu U. Kumary, Eva Brill, Martis W. Cowles, Bryan Venters, Julia B. Schueler. Identifying genome-wide histone profiles in patient derived models of AML for predictive biomarker development [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4413.