Frontiers in Oncology | 2021
Editorial: Special Issue on Innovative Multi-Disciplinary Approaches for Precision Studies in Leukemia
Abstract
By investigating the role of recurrent mutations, epigenetic modifications and aberrant activation of oncogenic signaling (1–4), our understanding of leukemogenesis has advanced to the stage that precision medicine is capable of transforming diagnostic and therapeutic approaches. This Research Topic focuses on the innovative technologies that will aid in the development of a multidimensional integrated molecular network, enabling the prediction of therapy responses and creation of novel treatments for those in most clinical need. Risk stratification scores in AML are important tools for clinicians and scientists alike, due to the relative ease with which biomarker panels can assist with diagnosis, prediction of prognosis, overall survival (OS) and guide appropriate treatment choices. Our work has identified a gene signature associated with the histone demethylase KDM4A, the KDM4A-9 score, which is associated with poor OS and independent of age, cytogenetic risk and mutation status (2). KDM4A-9 is highly correlative with the LSC17 score, while having no overlap, showing the collective power of integrating risk scoring systems (2, 5). In this Research Topic, risk stratification was addressed by Lu et al. through bioinformatic analysis of the publicly-available databases. The authors explored the relationship of CXCR family members in AML and found that over-expression of specific CXCR receptors correlates with FAB subtypes and AML risk stratification. Of note, CXCR2 expression was identified as an independent prognostic factor. Applying machine learning algorithms to gene expression datasets, Mosquera Orgueira et al. described a new model containing 123 variables (ST-123) that was capable of predicting survival of AML patients. Two genes were identified, namely KDM5B and LAPTM4B, that have an established role in the pathogenesis of myeloid malignancies. Similarly, Sanchez Corrales et al. reviewed machine learning potential to dissect those heterogeneities that render AML hard to tackle with effective therapies, by identifying specific leukemic populations and the biomarkers that define these populations. They propose to design precision immunotherapy strategies, accounting for the heterogeneous sub-populations characteristic of each disease. Van Gils et al. offered a comprehensive review of the diverse, mainly non-genetic, mechanisms of resistance to chemotherapy in AML, from survival pathways mediated by the bone marrow microenvironment to altered epigenetic profiles and metabolic reprogramming. Unlike most reviews on the same subject, it focused also on AML “persisters” highlighting that they are not