Developing a classification of hematologic neoplasms in the era of precision medicine

MOSAIC: An Artificial Intelligence-Based Framework for Multimodal Analysis, Classification, and Personalized Prognostic Assessment in Rare Cancers

PURPOSE

Rare cancers constitute over 20% of human neoplasms, often affecting patients with unmet medical needs. The development of effective classification and prognostication systems is crucial to improve the decision-making process and drive innovative treatment strategies. We have created and implemented MOSAIC, an artificial intelligence (AI)-based framework designed for multimodal analysis, classification, and personalized prognostic assessment in rare cancers. Clinical validation was performed on myelodysplastic syndrome (MDS), a rare hematologic cancer with clinical and genomic heterogeneities.

METHODS

We analyzed 4,427 patients with MDS divided into training and validation cohorts. Deep learning methods were applied to integrate and impute clinical/genomic features. Clustering was performed by combining Uniform Manifold Approximation and Projection for Dimension Reduction + Hierarchical Density-Based Spatial Clustering of Applications with Noise (UMAP + HDBSCAN) methods, compared with the conventional Hierarchical Dirichlet Process (HDP). Linear and AI-based nonlinear approaches were compared for survival prediction. Explainable AI (Shapley Additive Explanations approach [SHAP]) and federated learning were used to improve the interpretation and the performance of the clinical models, integrating them into distributed infrastructure.

RESULTS

UMAP + HDBSCAN clustering obtained a more granular patient stratification, achieving a higher average silhouette coefficient (0.16) with respect to HDP (0.01) and higher balanced accuracy in cluster classification by Random Forest (92.7% ± 1.3% and 85.8% ± 0.8%). AI methods for survival prediction outperform conventional statistical techniques and the reference prognostic tool for MDS. Nonlinear Gradient Boosting Survival stands in the internal (Concordance-Index [C-Index], 0.77; SD, 0.01) and external validation (C-Index, 0.74; SD, 0.02). SHAP analysis revealed that similar features drove patients' subgroups and outcomes in both training and validation cohorts. Federated implementation improved the accuracy of developed models.

CONCLUSION

MOSAIC provides an explainable and robust framework to optimize classification and prognostic assessment of rare cancers. AI-based approaches demonstrated superior accuracy in capturing genomic similarities and providing individual prognostic information compared with conventional statistical methods. Its federated implementation ensures broad clinical application, guaranteeing high performance and data protection.

Advances towards genome-based acute myeloid leukemia classification: A comparative analysis of WHO-HAEM4R, WHO-HAEM5, and International Consensus Classification

Two recent guidelines, the 5th edition of the World Health Organization Classification of Haematolymphoid Tumours (WHO-HAEM5) and the International Consensus Classification (ICC), were published to refine the diagnostic criteria of acute myeloid leukemia (AML). They both consider genomic features more extensively and expand molecularly defined AML subtypes. In this study, we compared the classifications of 1135 AML cases under both criteria. According to WHO-HAEM5 and ICC, the integration of whole transcriptome sequencing, targeted gene mutation screening, and conventional cytogenetic analysis identified defining genetic abnormalities in 89% and 90% of AML patients, respectively. The classifications displayed discrepancies in 16% of AML cases after being classified using the two guidelines, respectively. Both new criteria significantly reduce the number of cases defined by morphology and differentiation. However, their clinical implementation heavily relies on comprehensive and sophisticated genomic analysis, including genome and transcriptome levels, alongside the assessment of pathogenetic somatic and germline variations. Discrepancies between WHO-HAEM5 and ICC, such as the assignment of RUNX1 mutations, the rationality of designating AML with mutated TP53 as a unique entity, and the scope of rare genetic fusions, along with the priority of concurrent AML-defining genetic abnormalities, are still pending questions requiring further research for more elucidated insights.

Artificial intelligence and classification of mature lymphoid neoplasms

Hematologists, geneticists, and clinicians came to a multidisciplinary agreement on the classification of lymphoid neoplasms that combines clinical features, histological characteristics, immunophenotype, and molecular pathology analyses. The current classification includes the World Health Organization (WHO) Classification of tumours of haematopoietic and lymphoid tissues revised 4th edition, the International Consensus Classification (ICC) of mature lymphoid neoplasms (report from the Clinical Advisory Committee 2022), and the 5th edition of the proposed WHO Classification of haematolymphoid tumours (lymphoid neoplasms, WHO-HAEM5). This article revises the recent advances in the classification of mature lymphoid neoplasms. Artificial intelligence (AI) has advanced rapidly recently, and its role in medicine is becoming more important as AI integrates computer science and datasets to make predictions or classifications based on complex input data. Summarizing previous research, it is described how several machine learning and neural networks can predict the prognosis of the patients, and classified mature B-cell neoplasms. In addition, new analysis predicted lymphoma subtypes using cell-of-origin markers that hematopathologists use in the clinical routine, including CD3, CD5, CD19, CD79A, MS4A1 (CD20), MME (CD10), BCL6, IRF4 (MUM-1), BCL2, SOX11, MNDA, and FCRL4 (IRTA1). In conclusion, although most categories are similar in both classifications, there are also conceptual differences and differences in the diagnostic criteria for some diseases. It is expected that AI will be incorporated into the lymphoma classification as another bioinformatics tool.

Diagnosis and classification of myelodysplastic syndromes

ABSTRACT

Myelodysplastic syndromes (MDSs) are neoplastic myeloid proliferations characterized by ineffective hematopoiesis resulting in peripheral blood cytopenias. MDS is distinguished from nonneoplastic clonal myeloid proliferations by the presence of morphologic dysplasia and from acute myeloid leukemia by a blast threshold of 20%. The diagnosis of MDS can be challenging because of the myriad other causes of cytopenias: accurate diagnosis requires the integration of clinical features with bone marrow and peripheral blood morphology, immunophenotyping, and genetic testing. MDS has historically been subdivided into several subtypes by classification schemes, the most recent of which are the International Consensus Classification and World Health Organization Classification (fifth edition), both published in 2022. The aim of MDS classification is to identify entities with shared genetic underpinnings and molecular pathogenesis, and the specific subtype can inform clinical decision-making alongside prognostic risk categorization. The current MDS classification schemes incorporate morphologic features (bone marrow and blood blast percentage, degree of dysplasia, ring sideroblasts, bone marrow fibrosis, and bone marrow hypocellularity) and also recognize 3 entities defined by genetics: isolated del(5q) cytogenetic abnormality, SF3B1 mutation, and TP53 mutation. It is anticipated that with advancing understanding of the genetic basis of MDS pathogenesis, future MDS classification will be based increasingly on genetic classes. Nevertheless, morphologic features in MDS reflect the phenotypic expression of the underlying abnormal genetic pathways and will undoubtedly retain importance to inform prognosis and guide treatment.

WHO, what, when, where, and why: New classification systems for acute myeloid leukemia and their impact on clinical practice

The goal of a disease classification system is (or should be) to provide a tool for researchers and clinicians to study and treat the disease. The last decade has seen a markedly improved understanding of the pathophysiology of acute myeloid leukemia (AML), the development of new methods to measure the disease, and approval by the Food and Drug Administration (FDA) of at least ten new therapies targeted to its treatment. In response, in 2022 one updated and one new AML classification system were published. In the same year, the European LeukemiaNet updated their recommendations about how to incorporate the advances in diagnosis and treatment into the risk stratification of AML and its treatment. The following discussion summarizes the highlights of these changes and offers an opinion of how well these changes meet the goal of aiding researchers and clinicians in the study and treatment of AML.

The pathobiology of follicular lymphoma

Follicular lymphoma is one of the most frequent lymphomas. Histologically, it is characterized by a follicular (nodular) growth pattern of centrocytes and centroblasts; mixed with variable immune microenvironment cells. Clinically, it is characterized by diffuse lymphadenopathy, bone marrow involvement, and splenomegaly. It is biologically and clinically heterogeneous. In most patients it is indolent, but others have a more aggressive evolution with relapses; and transformation to diffuse large B-cell lymphoma. Tumorigenesis includes an asymptomatic preclinical phase in which premalignant B-lymphocytes with the t(14;18) chromosomal translocation acquire additional genetic alterations in the germinal centers, and clonal evolution occurs, although not all the cells progress to the tumor stage. This manuscript reviews the pathobiology and clinicopathological characteristics of follicular lymphoma. It includes a description of the physiology of the germinal center, the genetic alterations of BCL2 and BCL6, the mutational profile, the immune checkpoint, precision medicine, and highlights in the lymphoma classification. In addition, a comment and review on artificial intelligence and machine (deep) learning are made.

The Times, They Are A-Changing: The Impact of Next-Generation Sequencing on Diagnosis, Classification, and Prognostication of Myeloid Malignancies With Focus on Myelodysplastic Syndrome, AML, and Germline Predisposition

Myeloid malignancies are a manifestation of clonal expansion of hematopoietic cells driven by somatic genetic alterations that may arise in a potential background of deleterious germline variants. As next-generation sequencing technology has become more accessible, real-world experience has allowed integration of molecular genomic data with morphology, immunophenotype, and conventional cytogenetics to refine our understanding of myeloid malignancies. This has prompted revisions in the classification and the prognostication schema of myeloid malignancies and germline predisposition to hematologic malignancies. This review provides an overview of significant changes in the recently published classifications of AML and myelodysplastic syndrome, emerging prognostic scoring, and the role of germline deleterious variants in predisposing to MDS and AML.

Risk stratifying MDS in the time of precision medicine

Myelodysplastic syndromes (MDS) are myeloid neoplasms characterized by morphologic dysplasia, persistent cytopenia, and a variable risk of evolution to acute myeloid leukemia (AML). Risk stratification is crucial in a patient-centered approach to the treatment of MDS. Based on hematologic parameters and cytogenetic abnormalities, the Revised International Prognostic Scoring System is currently used for this purpose. In the past years, the use of massively parallel DNA sequencing has clarified the genetic basis of MDS and has enabled development of novel diagnostic and prognostic approaches. When conventional cytogenetics is combined with gene sequencing, more than 90% of patients are found to carry a somatic genetic lesion. In addition, a portion of patients has germline variants that predispose them to myeloid neoplasms. The recently developed International Consensus Classification of MDS includes new entities that are molecularly defined-namely, SF3B1-mutant and TP53-mutant MDS. The International Working Group for Prognosis in MDS has just developed the International Prognostic Scoring System-Molecular (IPSS-M) for MDS, which considers hematologic parameters, cytogenetic abnormalities, and somatic gene mutations. The IPSS-M score is personalized and can be obtained using a web-based calculator that returns not only the individual score but also the expected leukemia-free survival, overall survival, and risk of AML transformation. Providing an efficient risk stratification of patients with MDS, the IPSS-M represents a valuable tool for individual risk assessment and treatment decisions.