Clustering of lymphoid neoplasms by cell of origin, somatic mutation and drug usage profiles: a multi-trait...

Lymphoid neoplasms (LNs) are heterogeneous malignancies arising from lymphoid cells, displaying diverse clinical and molecular features. Although LNs are collectively frequent, individual subtypes are rare, posing challenges for genetic association studies. Indeed, genome-wide association studies (GWAS) explained only a fraction of the heritability. Shared genetic susceptibility and overlapping risk factors suggest a partially common etiology across subtypes. We employed a multi-trait GWAS strategy to improve discovery power by leveraging pleiotropy among LN subtypes. We defined LN phenoclusters based on cell of origin, somatic mutation profiles, and approved therapeutic agents. Using data from three large cohorts-the UK Biobank, Million Veteran Program, and FinnGen-we analyzed 31,937 LN cases and 1.2 million controls across 8 individual subtypes and 7 phenoclusters. We replicated the novel associations in two independent cohorts (All of Us and the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial) with 2892 LN cases and 165,791 controls. We identified 76 genome-wide significant loci for individual subtypes or subtype clusters, including 20 novel associations. We identified the subtypes contributing to each locus, putative candidate causal variants, and genes underlying the associations, and found enrichment of specific cell types, biological processes, and drugs associated with LN risk genes. Overall, this study identified new LN genetic risk loci and candidate genes, providing insights that may inform novel therapeutic approaches.