PAIRWISE: Deep Learning-based Prediction of Effective Personalized Drug Combinations in Cancer.

Combination therapies offer promise for improving cancer treatment efficacy and preventing recurrence. However, identifying optimal drug combinations tailored to specific cancer subtypes and individual patients is extremely challenging due to the vast number of possible combinations and tumor heterogeneity. To address this gap, we take a machine learning approach combining deep learning with transfer learning to incorporate prior scientific knowledge and predict drug synergy based on tumor-specific transcriptome profiles.

SETD1B mutations confer apoptosis resistance and BCL2 independence in B cell lymphoma.

The translocation t(14;18) activates BCL2 and is considered the initiating genetic lesion in most follicular lymphomas (FL). Surprisingly, FL patients fail to respond to the BCL2 inhibitor, Venetoclax. We show that mutations and deletions affecting the histone lysine methyltransferase SETD1B (KMT2G) occur in 7% of FLs and 16% of diffuse large B cell lymphomas (DLBCL).

Profiling of Copy Number Alterations Using Low-Coverage Whole-Genome Sequencing Informs Differential Diagnosis and Prognosis in Primary Cutaneous Follicle Center Lymphoma.

Primary cutaneous follicle center lymphoma (PCFCL) has an excellent prognosis using local treatment, whereas nodal follicular lymphoma (nFL), occasionally presenting with cutaneous spread, often requires systemic therapy. Distinction of the 2 diseases based on histopathology alone might be challenging. Copy number alterations (CNAs) have scarcely been explored on a genome-wide scale in PCFCL; however, they might serve as potential biomarkers during differential diagnosis and risk stratification.

Defining an Optimized Workflow for Enriching and Analyzing Residual Tumor Populations Using Intracellular Markers.

Tumor relapse is well recognized to arise from treatment-resistant residual populations. Strategies enriching such populations for in-depth downstream analyses focus on tumor-specific surface markers; however, enrichment using intracellular biomarkers remains challenging. Using B-cell lymphoma as an exemplar, we demonstrate feasibility to enrich B-cell lymphoma 2 (BCL2) populations, a surrogate marker for t(14;18)+ lymphomas, for use in downstream applications.

Integrative phosphoproteomics defines two biologically distinct groups of KMT2A rearranged acute myeloid leukaemia with different drug response phenotypes.

Acute myeloid leukaemia (AML) patients harbouring certain chromosome abnormalities have particularly adverse prognosis. For these patients, targeted therapies have not yet made a significant clinical impact. To understand the molecular landscape of poor prognosis AML we profiled 74 patients from two different centres (in UK and Finland) at the proteomic, phosphoproteomic and drug response phenotypic levels.

European standard clinical practice - Key issues for the medical care of individuals with familial leukemia.

Although hematologic malignancies (HM) are no longer considered exclusively sporadic, additional awareness of familial cases has yet to be created. Individuals carrying a (likely) pathogenic germline variant (e.g.

Germline predisposition to haematological malignancies: Best practice consensus guidelines from the UK Cancer Genetics Group (UKCGG), CanGene-CanVar and the NHS England Haematological Oncology Working Group.

The implementation of whole genome sequencing and large somatic gene panels in haematological malignancies is identifying an increasing number of individuals with either potential or confirmed germline predisposition to haematological malignancy. There are currently no national or international best practice guidelines with respect to management of carriers of such variants or of their at-risk relatives. To address this gap, the UK Cancer Genetics Group (UKCGG), CanGene-CanVar and the NHS England Haematological Oncology Working Group held a workshop over two days on 28-29th April 2022, with the aim of establishing consensus guidelines on relevant clinical and laboratory pathways.

Biallelic TET2 mutations confer sensitivity to 5'-azacitidine in acute myeloid leukemia.

Precision medicine can significantly improve outcomes for patients with cancer, but implementation requires comprehensive characterization of tumor cells to identify therapeutically exploitable vulnerabilities. Here, we describe somatic biallelic TET2 mutations in an elderly patient with acute myeloid leukemia (AML) that was chemoresistant to anthracycline and cytarabine but acutely sensitive to 5'-azacitidine (5'-Aza) hypomethylating monotherapy, resulting in long-term morphological remission. Given the role of TET2 as a regulator of genomic methylation, we hypothesized that mutant TET2 allele dosage affects response to 5'-Aza.

Germline ERCC excision repair 6 like 2 (ERCC6L2) mutations lead to impaired erythropoiesis and reshaping of the bone marrow microenvironment.

Despite the inclusion of inherited myeloid malignancies as a separate entity in the World Health Organization Classification, many established predisposing loci continue to lack functional characterization. While germline mutations in the DNA repair factor ERCC excision repair 6 like 2 (ERCC6L2) give rise to bone marrow failure and acute myeloid leukaemia, their consequences on normal haematopoiesis remain unclear. To functionally characterise the dual impact of germline ERCC6L2 loss on human primary haematopoietic stem/progenitor cells (HSPCs) and mesenchymal stromal cells (MSCs), we challenged ERCC6L2-silenced and patient-derived cells ex vivo.

Longitudinal expression profiling identifies a poor risk subset of patients with ABC-type diffuse large B-cell lymphoma.

Despite the effectiveness of immuno-chemotherapy, 40% of patients with diffuse large B-cell lymphoma (DLBCL) experience relapse or refractory disease. Longitudinal studies have previously focused on the mutational landscape of relapse but fell short of providing a consistent relapse-specific genetic signature. In our study, we have focused attention on the changes in GEP accompanying DLBCL relapse using archival paired diagnostic/relapse specimens from 38 de novo patients with DLBCL.

A dual role for the RNA helicase DHX34 in NMD and pre-mRNA splicing and its function in hematopoietic differentiation.

The DExD/H-box RNA helicase DHX34 is a nonsense-mediated decay (NMD) factor that together with core NMD factors coregulates NMD targets in nematodes and in vertebrates. Here, we show that DHX34 is also associated with the human spliceosomal catalytic C complex. Mapping of DHX34 endogenous binding sites using cross-linking immunoprecipitation (CLIP) revealed that DHX34 is preferentially associated with pre-mRNAs and locates at exon-intron boundaries.

CKS1 inhibition depletes leukemic stem cells and protects healthy hematopoietic stem cells in acute myeloid leukemia.

Acute myeloid leukemia (AML) is an aggressive hematological disorder comprising a hierarchy of quiescent leukemic stem cells (LSCs) and proliferating blasts with limited self-renewal ability. AML has a dismal prognosis, with extremely low 2-year survival rates in the poorest cytogenetic risk patients, primarily due to the failure of intensive chemotherapy protocols to deplete LSCs and toxicity of therapy toward healthy hematopoietic cells. We studied the role of cyclin-dependent kinase regulatory subunit 1 (CKS1)-dependent protein degradation in primary human AML and healthy hematopoiesis xenograft models in vivo.

Generation and Surgical Analysis of Genetic Mouse Models to Study NF-κB-Driven Pathogenesis of Diffuse Large B Cell Lymphoma.

Enforced activation of NF-κB signaling can be achieved by constitutive NF-κB-inducing kinases, IKK2 and NIK, or via lymphoma-associated mutants of MYD88, CARD11, and CD79B. In order to model Diffuse Large B Cell Lymphoma (DLBCL) in mice, conditional alleles for these proteins are combined with alleles targeting Cre recombinase expression in mature B cells. However, unopposed NF-κB signaling promotes plasmablast differentiation, and as a consequence the model system must be complemented with further mutations that block differentiation, such as Prdm1/BLIMP1 inactivation or overexpression of BCL6.

KDM5 inhibition offers a novel therapeutic strategy for the treatment of KMT2D mutant lymphomas.

Loss-of-function mutations in KMT2D are a striking feature of germinal center (GC) lymphomas, resulting in decreased histone 3 lysine 4 (H3K4) methylation and altered gene expression. We hypothesized that inhibition of the KDM5 family, which demethylates H3K4me3/me2, would reestablish H3K4 methylation and restore the expression of genes repressed on loss of KMT2D. KDM5 inhibition increased H3K4me3 levels and caused an antiproliferative response in vitro, which was markedly greater in both endogenous and gene-edited KMT2D mutant diffuse large B-cell lymphoma cell lines, whereas tumor growth was inhibited in KMT2D mutant xenografts in vivo.

Drug ranking using machine learning systematically predicts the efficacy of anti-cancer drugs.

Artificial intelligence and machine learning (ML) promise to transform cancer therapies by accurately predicting the most appropriate therapies to treat individual patients. Here, we present an approach, named Drug Ranking Using ML (DRUML), which uses omics data to produce ordered lists of >400 drugs based on their anti-proliferative efficacy in cancer cells. To reduce noise and increase predictive robustness, instead of individual features, DRUML uses internally normalized distance metrics of drug response as features for ML model generation.