With recent technological advances and significant progress in understanding the pathogenesis of acute myeloid leukemia (AML), the updated fifth edition WHO Classification (WHO-HAEM5) and the newly introduced International Consensus Classification (ICC), as well as the European LeukemiaNet (ELN) recommendations in 2022, require the integration of immunophenotypic, cytogenetic, and molecular data, alongside clinical and morphologic findings, for accurate diagnosis, prognostication, and guiding therapeutic strategies in AML. Flow cytometry offers rapid and sensitive immunophenotyping through a multiparametric approach and is a pivotal laboratory tool for the classification of AML, identification of therapeutic targets, and monitoring of measurable residual disease (MRD) post therapy. The association of immunophenotypic features and recurrent genetic abnormalities has been recognized and applied in informing further diagnostic evaluation and immediate therapeutic decision-making. Recently, the evolving role of machine learning models in assisting flow cytometric data analysis for the automated diagnosis and prediction of underlying genetic alterations has been illustrated.
The accurate diagnosis and classification of myelodysplastic/myeloproliferative neoplasm (MDS/MPN) are challenging due to the overlapping pathological and molecular features of myelodysplastic syndrome (MDS) and myeloproliferative neoplasm (MPN). We investigated the genomic landscape in different MDS/MPN subtypes, including chronic myelomonocytic leukemia (CMML; n = 97), atypical chronic myeloid leukemia (aCML; n = 8), MDS/MPN-unclassified (MDS/MPN-U; n = 44), and MDS/MPN with ring sideroblasts and thrombocytosis (MDS/MPN-RS-T; n = 12). Our study indicated that MDS/MPN is characterized by mutations commonly identified in myeloid neoplasms, with TET2 (52%) being the most frequently mutated gene, followed by ASXL1 (38.7%), SRSF2 (34.7%), and JAK2 (19.7%), among others. However, the distribution of recurrent mutations differs across the MDS/MPN subtypes. We confirmed that specific gene combinations correlate with specific MDS/MPN subtypes (e.g., TET2/SRSF2 in CMML, ASXL1/SETBP1 in aCML, and SF3B1/JAK2 in MDS/MPN-RS-T), with MDS/MPN-U being the most heterogeneous. Furthermore, we found that older age (≥65 years) and mutations in RUNX1 and TP53 were associated with poorer clinical outcomes in CMML (p < 0.05) by multivariate analysis. In MDS/MPN-U, CBL mutations (p < 0.05) were the sole negative prognostic factors identified in our study by multivariate analysis (p < 0.05). Overall, our study provides genetic insights into various MDS/MPN subtypes, which may aid in diagnosis and clinical decision-making for patients with MDS/MPN.