Molecular Patterns Identify Distinct Subclasses of Myeloid Neoplasia

Abstract

BackgroundGenomic mutations drive the pathogenesis of myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML). While morphological and clinical features, complemented by cytogenetics, have dominated the classical criteria for diagnosis and classi cation, incorporation of molecular mutational data can illuminate functional pathobiology. MethodsWe combined cytogenetic and molecular features from a multicenter cohort of 3588 MDS, MDS/ myeloproliferative neoplasm (including chronic myelomonocytic leukemia [CMML]), and secondary AML patients to generate a molecular-based scheme using machine learning methods and then externally validated the model on 412 patients. Molecular signatures driving each cluster were identi ed and used for genomic subclassi cation. FindingsUnsupervised analyses identi ed 14 distinctive and clinically heterogenous molecular clusters (MCs) with unique pathobiological associations, treatment responses, and prognosis. Normal karyotype (NK) was enriched in MC2, MC4, MC6, MC9, MC10, and MC12 with different distributions of TET2, SF3B1, ASXL1, DNMT3A, and RAS mutations. Complex karyotype and trisomy 8 were enriched in MC13 and MC1, respectively. We then identi ed ve risk groups to re ect the biological differences between clusters. Our clustering model was able to highlight the signi cant survival differences among patients assigned to the similar IPSS-R risk group but with heterogenous molecular con gurations. Different response rates to hypomethylating agents (e.g., MC9 and MC13 [OR: 2.2 and 0.6, respectively]) re ected the biological differences between the clusters. Interestingly, our clusters continued to show survival differences regardless of the bone marrow blast percentage. InterpretationDespite the complexity of the molecular alterations in myeloid neoplasia, our model recognized functional objective clusters, irrespective of anamnestic clinico-morphological features, that re ected disease evolution and informed classi cation, prognostication, and molecular interactions. Our subclassi cation model is available via a web-based open-access resource as well (https://drmz.shinyapps.io/mds_latent).

Classification and prognostic stratification based on genomic features in myelodysplastic neoplasms, myeloproliferative neoplasms, and their overlapping conditions

“…Recent genomic studies have introduced an alternative paradigm for the classi cation of myeloid neoplasms (MNs), focusing on shared and distinct genomic patterns (1)(2)(3). Landmark studies in myeloproliferative neoplasms (MPN) (4), myelodysplastic neoplasms (MDS) (5,6), and their overlapping diseases (7) demonstrate the potential of genomic classi cation systems for disease subclassi cation and personalized prognostication. This has led to a fundamental shift towards genetics-based approaches in characterizing these diseases, with recent classi cation systems increasingly incorporating genomic attributes to align more closely with their underlying biology (2,8).…”

Network-based clustering unveils interconnected landscapes of genomic and clinical features across myeloid malignancies

“…Although guided by threshold values (e.g., blast values being below or above a certain threshold) of these clinical variables, the classifications have changed significantly over the years and are increasingly driven by mutational and cytogenetic features 5–9 . The presence and interaction of these genetic changes can influence the prognosis and treatment strategy 10–14 .…”

“…Leveraging a data-driven approach, previous studies have clustered the mutational and cytogenetic profiles of myeloid malignancies to derive risk scores in AML 10,13,15,16 , MDS 14,17,18 , CMML 19 , and MPS 20,21 . Outside of myeloid malignancies, network-based approaches have been employed to effectively condense heterogeneous genomic profiles for prediction and patient stratification 22–25 .…”

“…However, most previous clustering strategies did not integrate clinical covariates, such as blood and bone marrow values and demographic factors, into their clustering approaches, despite their established relevance for prognosis and clinical decision-making 26–29 . For instance, a previous study focused exclusively on genomic profiles to cluster MDS patients with secondary AML patients 14 . While insightful, as the authors acknowledged, this approach was unable to integrate clinical data for more nuanced patient stratification.…”

Using machine learning to unravel the intricacy of acute myeloid leukemia

“…Having said that, this study complements other key results achieved through ML in the field of AML in recent years. The interconnection of several variables in large cohorts of patients has allowed researchers to explore, through ML, different patient stratifications 2,3 and integrated prog-nostic algorithms, 4 to identify biomarkers, 5 and support cytomorphological diagnosis. 6 This huge amount of data has offered insights into different aspects of AML management, respecting the granularities of disease features, and suggesting the possibilities of adding new factors or classifiers to consider in the tailoring of treatment strategy.…”