Multi‐omics studies promise the improved characterization of biological processes across molecular layers. However, methods for the unsupervised integration of the resulting heterogeneous data sets are lacking. We present Multi‐Omics Factor Analysis (MOFA), a computational method for discovering the principal sources of variation in multi‐omics data sets. MOFA infers a set of (hidden) factors that capture biological and technical sources of variability. It disentangles axes of heterogeneity that are shared across multiple modalities and those specific to individual data modalities. The learnt factors enable a variety of downstream analyses, including identification of sample subgroups, data imputation and the detection of outlier samples. We applied MOFA to a cohort of 200 patient samples of chronic lymphocytic leukaemia, profiled for somatic mutations, RNA expression, DNA methylation and ex vivo drug responses. MOFA identified major dimensions of disease heterogeneity, including immunoglobulin heavy‐chain variable region status, trisomy of chromosome 12 and previously underappreciated drivers, such as response to oxidative stress. In a second application, we used MOFA to analyse single‐cell multi‐omics data, identifying coordinated transcriptional and epigenetic changes along cell differentiation.

Tumour heterogeneity encompasses both the malignant cells and their microenvironment. While heterogeneity between individual patients is known to affect the efficacy of cancer therapy, most personalized treatment approaches do not account for intratumour heterogeneity. We addressed this issue by studying the heterogeneity of nodal B-cell lymphomas by single-cell RNA-sequencing and transcriptome-informed flow cytometry. We identified transcriptionally distinct malignant subpopulations and compared their drug-response and genomic profiles. Malignant subpopulations from the same patient responded strikingly differently to anti-cancer drugs ex vivo, which recapitulated subpopulation-specific drug sensitivity during in vivo treatment. Infiltrating T cells represented the majority of non-malignant cells, whose gene-expression signatures were similar across all donors, whereas the frequencies of T-cell subsets varied significantly between the donors. Our data provide insights into the heterogeneity of nodal B-cell lymphomas and highlight the relevance of intratumour heterogeneity for personalized cancer therapy.

As new generations of targeted therapies emerge and tumor genome sequencing discovers increasingly comprehensive mutation repertoires, the functional relationships of mutations to tumor phenotypes remain largely unknown. Here, we measured ex vivo sensitivity of 246 blood cancers to 63 drugs alongside genome, transcriptome, and DNA methylome analysis to understand determinants of drug response. We assembled a primary blood cancer cell encyclopedia data set that revealed disease-specific sensitivities for each cancer. Within chronic lymphocytic leukemia (CLL), responses to 62% of drugs were associated with 2 or more mutations, and linked the B cell receptor (BCR) pathway to trisomy 12, an important driver of CLL. Based on drug responses, the disease could be organized into phenotypic subgroups characterized by exploitable dependencies on BCR, mTOR, or MEK signaling and associated with mutations, gene expression, and DNA methylation. Fourteen percent of CLLs were driven by mTOR signaling in a non–BCR-dependent manner. Multivariate modeling revealed immunoglobulin heavy chain variable gene (IGHV) mutation status and trisomy 12 as the most important modulators of response to kinase inhibitors in CLL. Ex vivo drug responses were associated with outcome. This study overcomes the perception that most mutations do not influence drug response of cancer, and points to an updated approach to understanding tumor biology, with implications for biomarker discovery and cancer care.

BCL-2 inhibition has been shown to be effective in acute myeloid leukemia (AML) in combination with hypomethylating agents or low-dose cytarabine. However, resistance and relapse represent major clinical challenges. Therefore, there is an unmet need to overcome resistance to current venetoclax-based strategies. We performed high-throughput drug screening to identify effective combination partners for venetoclax in AML. Overall, 64 antileukemic drugs were screened in 31 primary high-risk AML samples with or without venetoclax. Gilteritinib exhibited the highest synergy with venetoclax in FLT3 wild-type AML. The combination of gilteritinib and venetoclax increased apoptosis, reduced viability, and was active in venetoclax-azacitidine–resistant cell lines and primary patient samples. Proteomics revealed increased FLT3 wild-type signaling in specimens with low in vitro response to the currently used venetoclax-azacitidine combination. Mechanistically, venetoclax with gilteritinib decreased phosphorylation of ERK and GSK3B via combined AXL and FLT3 inhibition with subsequent suppression of the antiapoptotic protein MCL-1. MCL-1 downregulation was associated with increased MCL-1 phosphorylation of serine 159, decreased phosphorylation of threonine 161, and proteasomal degradation. Gilteritinib and venetoclax were active in an FLT3 wild-type AML patient-derived xenograft model with TP53 mutation and reduced leukemic burden in 4 patients with FLT3 wild-type AML receiving venetoclax-gilteritinib off label after developing refractory disease under venetoclax-azacitidine. In summary, our results suggest that combined inhibition of FLT3/AXL potentiates venetoclax response in FLT3 wild-type AML by inducing MCL-1 degradation. Therefore, the venetoclax-gilteritinib combination merits testing as a potentially active regimen in patients with high-risk FLT3 wild-type AML.

Mosunetuzumab, a CD20×CD3 T-cell engaging bispecific antibody, redirects T cells to eliminate malignant B cells. We present updated efficacy and safety data of a pivotal phase 1/2 study after a median follow-up of 37.4 months in 90 patients with relapsed/refractory (R/R) follicular lymphoma (FL) and ≥2 prior lines of therapy treated with fixed-duration mosunetuzumab. Investigator-assessed complete response (CR) rate and objective response rate were 60.0% (95% confidence interval [CI], 49.1-70.2) and 77.8% (95% CI, 67.8-85.9), respectively. Among 70 responders, median duration of response was 35.9 months (95% CI, 20.7 to not estimable [NE]). Among 54 patients who achieved CR, 49 remained in CR at the end of treatment; median duration of CR was not reached (NR; 95% CI, 33.0 to NE); Kaplan-Meier-estimated 30-month remission rate was 72.4% (95% CI, 59.2-85.6). Estimated 36-month overall survival (OS) rate was 82.4% (95% CI, 73.8-91.0); median OS was NR (95% CI, NE to NE). Median progression-free survival was 24.0 months (95% CI, 12.0 to NE). Median time to CD19+ B-cell recovery was 18.4 months (95% CI, 12.8-25.0) after 8 cycles of mosunetuzumab treatment. No new cytokine release syndrome events or fatal, serious, or grade ≥3 adverse events were reported. With extended follow-up, mosunetuzumab demonstrated high response rates, durable remissions, and manageable safety with no long-term concerns. This supports outpatient mosunetuzumab administration as an off-the-shelf, fixed-duration, safe, and effective treatment for patients with R/R FL, including those with high-risk disease. This trial was registered at www.clinicaltrials.gov as #NCT02500407.

Splenic marginal zone B-cell lymphoma (SMZL) is a heterogeneous clinico-biological entity. The clinical course is variable, multiple genes are mutated with no unifying mechanism, and essential regulatory pathways and surrounding microenvironments are diverse. We sought to clarify the heterogeneity of SMZL by resolving different subgroups and their underlying genomic abnormalities, pathway signatures, and microenvironment compositions to uncover biomarkers and therapeutic vulnerabilities. We studied 303 SMZL spleen samples collected through the IELSG46 multicenter international study (NCT02945319) by using a multiplatform approach. We carried out genetic and phenotypic analyses, defined self-organized signatures, validated the findings in independent primary tumor metadata and determined correlations with outcome data. We identified 2 prominent genetic clusters in SMZL, termed NNK (58% of cases, harboring NF-κB, NOTCH, and KLF2 modules) and DMT (32% of cases, with DNA-damage response, MAPK, and TLR modules). Genetic aberrations in multiple genes as well as cytogenetic and immunogenetic features distinguished NNK- from DMT-SMZLs. These genetic clusters not only have distinct underpinning biology, as judged by differences in gene-expression signatures, but also different outcomes, with inferior survival in NNK-SMZLs. Digital cytometry and in situ profiling segregated 2 basic types of SMZL immune microenvironments termed immune-suppressive SMZL (50% of cases, associated with inflammatory cells and immune checkpoint activation) and immune-silent SMZL (50% of cases, associated with an immune-excluded phenotype) with distinct mutational and clinical connotations. In summary, we propose a nosology of SMZL that can implement its classification and also aid in the development of rationally targeted treatments.