An artificial intelligence-assisted clinical framework to facilitate diagnostics and translational discovery in hematologic neoplasia.

BACKGROUND: The increasing volume and intricacy of sequencing data, along with other clinical and diagnostic data, like drug responses and measurable residual disease, creates challenges for efficient clinical comprehension and interpretation. Using paediatric B-cell precursor acute lymphoblastic leukaemia (BCP-ALL) as a use case, we present an artificial intelligence (AI)-assisted clinical framework clinALL that integrates genomic and clinical data into a user-friendly interface to support routine diagnostics and reveal translational insights for hematologic neoplasia. METHODS: We performed targeted RNA sequencing in 1365 cases with haematological neoplasms, primarily paediatric B-cell precursor acute lymphoblastic leukaemia (BCP-ALL) from the AIEOP-BFM ALL study. We carried out fluorescence in situ hybridization (FISH), karyotyping and arrayCGH as part of the routine diagnostics. The analysis results of these assays as well as additional clinical information were integrated into an interactive web interface using Bokeh, where the main graph is based on Uniform Manifold Approximation and Projection (UMAP) analysis of the gene expression data. At the backend of the clinALL, we built both shallow machine learning models and a deep neural network using Scikit-learn and PyTorch respectively. FINDINGS: By applying clinALL, 78% of undetermined patients under the current diagnostic protocol were stratified, and ambiguous cases were investigated. Translational insights were discovered, including IKZF1plus status dependent subpopulations of BCR::ABL1 positive patients, and a subpopulation within ETV6::RUNX1 positive patients that has a high relapse frequency. Our best machine learning models, LDA and PASNET-like neural network models, achieve F1 scores above 97% in predicting patients' subgroups. INTERPRETATION: An AI-assisted clinical framework that integrates both genomic and clinical data can take full advantage of the available data, improve point-of-care decision-making and reveal clinically relevant insights promptly. Such a lightweight and easily transferable framework works for both whole transcriptome data as well as the cost-effective targeted RNA-seq, enabling efficient and equitable delivery of personalized medicine in small clinics in developing countries. FUNDING: German Ministry of Education and Research (BMBF), German Research Foundation (DFG) and Foundation for Polish Science.

The IL-7R antagonist Lusvertikimab reduces leukemic burden in xenograft-ALL via antibody-dependent cellular phagocytosis.

Acute lymphoblastic leukemia (ALL) arises from the uncontrolled proliferation of precursor B or T cells (BCP- or T-ALL). Current treatment protocols obtain high cure rates in children but are based on toxic polychemotherapy. Novel therapies are urgently needed, especially in relapsed/refractory (r/r) disease, high-risk leukemias and T-ALL, where immunotherapy approaches remain scarce. While the Interleukin-7 receptor (IL-7R) plays a pivotal role in ALL development, no IL-7R-targeting immunotherapy has yet reached clinical application in ALL. The IL-7Rα chain (CD127)-targeting IgG4 antibody Lusvertikimab (formerly OSE-127) is a full antagonist of the IL-7R pathway showing a good safety profile in healthy volunteers. Here, we show that ~85% of ALL cases express surface CD127. We demonstrate significant in vivo efficacy of Lusvertikimab immunotherapy in a heterogeneous cohort of BCP- and T-ALL patient-derived xenografts (PDX) in minimal residual disease (MRD) and overt leukemia models, including r/r and high-risk leukemias. Importantly, Lusvertikimab was particularly effective when combined with polychemotherapy in a phase 2-like PDX study with CD127high samples leading to MRD-negativity in >50% of mice treated with combination therapy. Mechanistically, Lusvertikimab targeted ALL cells via a dual mode of action comprising direct IL-7R antagonistic activity and induction of macrophage-mediated antibody-dependent cellular phagocytosis (ADCP). Lusvertikimab-mediated in vitro ADCP levels significantly correlated with CD127 expression levels and the reduction of leukemia burden upon treatment of PDX animals in vivo. Altogether, through its dual mode of action and good safety profile, Lusvertikimab may represent a novel immunotherapy option for any CD127-positive ALL, particularly in combination with standard-of-care polychemotherapy.

The landscape of cytogenetic and molecular genetic methods in diagnostics for hematologic neoplasia.

Improvements made during the last decades in the management of patients with hematologic neoplasia have resulted in increase of overall survival. These advancements have become possible through progress in our understanding of genetic basis of different hematologic malignancies and their role in the current risk-adapted treatment protocols. In this review, we provide an overview of current cytogenetic and molecular genetic methods, commonly used in the genetic characterization of hematologic malignancies, describe the current developments in the cytogenetic and molecular diagnostics, and give an outlook into their future development. Furthermore, we give a brief overview of the most important public databases and guidelines for sequence variant interpretation.

Cytogenetics and genomics in pediatric acute lymphoblastic leukaemia.

The last five decades have witnessed significant improvement in diagnostics, treatment and management of children with acute lymphoblastic leukaemia (ALL). These advancements have become possible through progress in our understanding of the genetic and biological background of ALL, resulting in the introduction of risk-adapted treatment and novel therapeutic targets, e.g., tyrosine kinase inhibitors for BCR::ABL1-positive ALL. Further advances in the taxonomy of ALL and the discovery of new genetic biomarkers and therapeutic targets, as well as the introduction of targeted and immunotherapies into the frontline treatment protocols, may improve management and outcome of children with ALL. In this review we describe the current developments in the (cyto)genetic diagnostics and management of children with ALL, and provide an overview of the most important advances in the genetic classification of ALL. Furthermore, we discuss perspectives resulting from the development of new techniques, including artificial intelligence (AI).

The Gene Expression Classifier ALLCatchR Identifies B-cell Precursor ALL Subtypes and Underlying Developmental Trajectories Across Age.

Current classifications (World Health Organization-HAEM5/ICC) define up to 26 molecular B-cell precursor acute lymphoblastic leukemia (BCP-ALL) disease subtypes by genomic driver aberrations and corresponding gene expression signatures. Identification of driver aberrations by transcriptome sequencing (RNA-Seq) is well established, while systematic approaches for gene expression analysis are less advanced. Therefore, we developed ALLCatchR, a machine learning-based classifier using RNA-Seq gene expression data to allocate BCP-ALL samples to all 21 gene expression-defined molecular subtypes. Trained on n = 1869 transcriptome profiles with established subtype definitions (4 cohorts; 55% pediatric / 45% adult), ALLCatchR allowed subtype allocation in 3 independent hold-out cohorts (n = 1018; 75% pediatric / 25% adult) with 95.7% accuracy (averaged sensitivity across subtypes: 91.1% / specificity: 99.8%). High-confidence predictions were achieved in 83.7% of samples with 98.9% accuracy. Only 1.2% of samples remained unclassified. ALLCatchR outperformed existing tools and identified novel driver candidates in previously unassigned samples. Additional modules provided predictions of samples blast counts, patient's sex, and immunophenotype, allowing the imputation in cases where these information are missing. We established a novel RNA-Seq reference of human B-lymphopoiesis using 7 FACS-sorted progenitor stages from healthy bone marrow donors. Implementation in ALLCatchR enabled projection of BCP-ALL samples to this trajectory. This identified shared proximity patterns of BCP-ALL subtypes to normal lymphopoiesis stages, extending immunophenotypic classifications with a novel framework for developmental comparisons of BCP-ALL. ALLCatchR enables RNA-Seq routine application for BCP-ALL diagnostics with systematic gene expression analysis for accurate subtype allocation and novel insights into underlying developmental trajectories.

Clonal dynamics in pediatric B-cell precursor acute lymphoblastic leukemia with very early relapse.

INTRODUCTION: One-quarter of the relapses in children with B-cell precursor acute lymphoblastic leukemia (BCP-ALL) occur very early (within 18 months, before completion of treatment), and prognosis in these patients is worse compared to cases that relapse after treatment has ended. METHODS: In this study, we performed a genomic analysis of diagnosis-relapse pairs of 12 children who relapsed very early, followed by a deep-sequencing validation of all identified mutations. In addition, we included one case with a good initial treatment response and on-treatment relapse at the end of upfront therapy. RESULTS: We observed a dynamic clonal evolution in all cases, with relapse almost exclusively originating from a subclone at diagnosis. We identified several driver mutations that may have influenced the outgrowth of a minor clone at diagnosis to become the major clone at relapse. For example, a minimal residual disease (MRD)-based standard-risk patient with ETV6-RUNX1-positive leukemia developed a relapse from a TP53-mutated subclone after loss of the wildtype allele. Furthermore, two patients with TCF3-PBX1-positive leukemia that developed a very early relapse carried E1099K WHSC1 mutations at diagnosis, a hotspot mutation that was recurrently encountered in other very early TCF3-PBX1-positive leukemia relapses as well. In addition to alterations in known relapse drivers, we found two cases with truncating mutations in the cohesin gene RAD21. CONCLUSION: Comprehensive genomic characterization of diagnosis-relapse pairs shows that very early relapses in BCP-ALL frequently arise from minor subclones at diagnosis. A detailed understanding of the therapeutic pressure driving these events may aid the development of improved therapies.

Transcriptional and Mutational Profiling of B-Other Acute Lymphoblastic Leukemia for Improved Diagnostics.

B-cell precursor acute lymphoblastic leukemia (BCP-ALL) is the most common cancer in children, and significant progress has been made in diagnostics and the treatment of this disease based on the subtypes of BCP-ALL. However, in a large proportion of cases (B-other), recurrent BCP-ALL-associated genomic alterations remain unidentifiable by current diagnostic procedures. In this study, we performed RNA sequencing and analyzed gene fusions, expression profiles, and mutations in diagnostic samples of 185 children with BCP-ALL. Gene expression clustering showed that a subset of B-other samples partially clusters with some of the known subgroups, particularly DUX4-positive. Mutation analysis coupled with gene expression profiling revealed the presence of distinctive BCP-ALL subgroups, characterized by the presence of mutations in known ALL driver genes, e.g., PAX5 and IKZF1. Moreover, we identified novel fusion partners of lymphoid lineage transcriptional factors ETV6, IKZF1 and PAX5. In addition, we report on low blast count detection thresholds and show that the use of EDTA tubes for sample collection does not have adverse effects on sequencing and downstream analysis. Taken together, our findings demonstrate the applicability of whole-transcriptome sequencing for personalized diagnostics in pediatric ALL, including tentative classification of the B-other cases that are difficult to diagnose using conventional methods.

Unravelling the Sequential Interplay of Mutational Mechanisms during Clonal Evolution in Relapsed Pediatric Acute Lymphoblastic Leukemia.

Pediatric acute lymphoblastic leukemia (ALL) is the most common pediatric malignancy and is characterized by clonal heterogeneity. Genomic mutations can increase proliferative potential of leukemic cells and cause treatment resistance. However, mechanisms driving mutagenesis and clonal diversification in ALL are not fully understood. In this proof of principle study, we performed whole genome sequencing of two cases with multiple relapses in order to investigate whether groups of mutations separated in time show distinct mutational signatures. Based on mutation allele frequencies at diagnosis and subsequent relapses, we clustered mutations into groups and performed cluster-specific mutational profile analysis and de novo signature extraction. In patient 1, who experienced two relapses, the analysis unraveled a continuous interplay of aberrant activation induced cytidine deaminase (AID)/apolipoprotein B editing complex (APOBEC) activity. The associated signatures SBS2 and SBS13 were present already at diagnosis, and although emerging mutations were lost in later relapses, the process remained active throughout disease evolution. Patient 2 had three relapses. We identified episodic mutational processes at diagnosis and first relapse leading to mutations resembling ultraviolet light-driven DNA damage, and thiopurine-associated damage at first relapse. In conclusion, our data shows that investigation of mutational processes in clusters separated in time may aid in understanding the mutational mechanisms and discovery of underlying causes.

Multiclonal complexity of pediatric acute lymphoblastic leukemia and the prognostic relevance of subclonal mutations

Genomic studies of pediatric acute lymphoblastic leukemia (ALL) have shown remarkable heterogeneity in initial diagnosis, with multiple (sub)clones harboring lesions in relapse-associated genes. However, the clinical relevance of these subclonal alterations remains unclear. We assessed the clinical relevance and prognostic value of subclonal alterations in the relapse-associated genes IKZF1, CREBBP, KRAS, NRAS, PTPN11, TP53, NT5C2, and WHSC1 in 503 ALL cases. Using Molecular Inversion Probe sequencing and breakpoint-spanning PCR we reliably detected alterations below 1% allele frequency. We identified 660 genomic alterations in 285 diagnosis samples of which 495 (75%) were subclonal. RAS pathway mutations were common, particularly in minor subclones, and comparisons between RAS hotspot mutations revealed differences in their capacity to drive clonal expansion in ALL. We did not find an association of subclonal alterations with unfavorable outcome. Particularly for IKZF1, an established prognostic marker in ALL, all clonal but none of the subclonal alterations were preserved at relapse. We conclude that, for the genes tested, there is no basis to consider subclonal alterations detected at diagnosis for risk group stratification of ALL treatment.

Mutational landscape and patterns of clonal evolution in relapsed pediatric acute lymphoblastic leukemia.

Relapse of acute lymphoblastic leukemia (ALL) remains a leading cause of childhood death. Prior studies have shown clonal mutations at relapse often arise from relapse-fated subclones that exist at diagnosis. However, the genomic landscape, evolutionary trajectories and mutational mechanisms driving relapse are incompletely understood. In an analysis of 92 cases of relapsed childhood ALL, incorporating multimodal DNA and RNA sequencing, deep digital mutational tracking and xenografting to formally define clonal structure, we identify 50 significant targets of mutation with distinct patterns of mutational acquisition or enrichment. CREBBP, NOTCH1, and Ras signaling mutations rose from diagnosis subclones, whereas variants in NCOR2, USH2A and NT5C2 were exclusively observed at relapse. Evolutionary modeling and xenografting demonstrated that relapse-fated clones were minor (50%), major (27%) or multiclonal (18%) at diagnosis. Putative second leukemias, including those with lineage shift, were shown to most commonly represent relapse from an ancestral clone rather than a truly independent second primary leukemia. A subset of leukemias prone to repeated relapse exhibited hypermutation driven by at least three distinct mutational processes, resulting in heightened neoepitope burden and potential vulnerability to immunotherapy. Finally, relapse-driving sequence mutations were detected prior to relapse using deep digital PCR at levels comparable to orthogonal approaches to monitor levels of measurable residual disease. These results provide a genomic framework to anticipate and circumvent relapse by earlier detection and targeting of relapse-fated clones.

Upfront Treatment Influences the Composition of Genetic Alterations in Relapsed Pediatric B-Cell Precursor Acute Lymphoblastic Leukemia.

Genomic alterations in relapsed B-cell precursor acute lymphoblastic leukemia (BCP-ALL) may provide insight into the role of specific genomic events in relapse development. Along this line, comparisons between the spectrum of alterations in relapses that arise in different upfront treatment protocols may provide valuable information on the association between the tumor genome, protocol components and outcome. Here, we performed a comprehensive characterization of relapsed BCP-ALL cases that developed in the context of 3 completed Dutch upfront studies, ALL8, ALL9, and ALL10. In total, 123 pediatric BCP-ALL relapses and 77 paired samples from primary diagnosis were analyzed for alterations in 22 recurrently affected genes. We found pronounced differences in relapse alterations between the 3 studies. Specifically, CREBBP mutations were observed predominantly in relapses after treatment with ALL8 and ALL10 which, in the latter group, were all detected in medium risk-treated patients. IKZF1 alterations were enriched 2.2-fold (p = 0.01) and 2.9-fold (p < 0.001) in ALL8 and ALL9 relapses compared to diagnosis, respectively, whereas no significant enrichment was found for relapses that were observed after treatment with ALL10. Furthermore, IKZF1 deletions were more frequently preserved from a major clone at diagnosis in relapses after ALL9 compared to relapses after ALL8 and ALL10 (p = 0.03). These data are in line with previous studies showing that the prognostic value of IKZF1 deletions differs between upfront protocols and is particularly strong in the ALL9 regimen. In conclusion, our data reveal a correlation between upfront treatment and the genetic composition of relapsed BCP-ALL.

Accurate detection of low-level mosaic mutations in pediatric acute lymphoblastic leukemia using single molecule tagging and deep-sequencing.

Pathogenic mutations in relapse-associated genes in pediatric acute lymphoblastic leukemia may improve risk stratification when detected at subclonal levels at primary diagnosis. However, to detect subclonal mutations upfront, a deep-sequencing approach with high specificity and sensitivity is required. Here, we performed a proof-of-principle study to detect low-level mosaic RAS pathway mutations by deep sequencing using random tagging-based single molecule Molecular Inversion Probes (smMIPs). The smMIP-based approach could sensitively detect variants with allele frequency as low as 0.4%, which could all be confirmed by other techniques. In comparison, with standard deep-sequencing techniques we reached a detection threshold of only 2.5%, which hampered detection of seven low-level mosaic mutations representing 24% of all detected mutations. We conclude that smMIP-based deep-sequencing outperforms standard deep-sequencing techniques by showing lower background noise and high specificity, and is the preferred technology for detecting mutations upfront, particularly in genes in which mutations show limited clustering in hotspots.