Novel germ line DDX41 mutations define families with a lower age of MDS/AML onset and lymphoid malignancies.

Recently our group and others have identified DDX41 mutations both as germ line and acquired somatic mutations in families with multiple cases of late onset myelodysplastic syndrome (MDS) and/or acute myeloid leukemia (AML), suggesting that DDX41 acts as a tumor suppressor. To determine whether novel DDX41 mutations could be identified in families with additional types of hematologic malignancies, our group screened two cohorts of families with a diverse range of hematologic malignancy subtypes. Among 289 families, we identified nine (3%) with DDX41 mutations. As previously observed, MDS and AML were the most common malignancies, often of the erythroblastic subtype, and 1 family displayed early-onset follicular lymphoma. Five novel mutations were identified, including missense mutations within important functional domains and start-loss and splicing mutations predicted to result in truncated proteins. We also show that most asymptomatic mutation carriers have normal blood counts until malignancy develops. This study expands both the mutation and phenotypic spectra observed in families with germ line DDX41 mutations. With an increasing number of both inherited and acquired mutations in this gene being identified, further study of how DDX41 disruption leads to hematologic malignancies is critical.

State-dependent evolutionary models reveal modes of solid tumour growth.

Spatial properties of tumour growth have profound implications for cancer progression, therapeutic resistance and metastasis. Yet, how spatial position governs tumour cell division remains difficult to evaluate in clinical tumours. Here, we demonstrate that faster division on the tumour periphery leaves characteristic genetic patterns, which become evident when a phylogenetic tree is reconstructed from spatially sampled cells. Namely, rapidly dividing peripheral lineages branch more extensively and acquire more mutations than slower-dividing centre lineages. We develop a Bayesian state-dependent evolutionary phylodynamic model (SDevo) that quantifies these patterns to infer the differential division rates between peripheral and central cells. We demonstrate that this approach accurately infers spatially varying birth rates of simulated tumours across a range of growth conditions and sampling strategies. We then show that SDevo outperforms state-of-the-art, non-cancer multi-state phylodynamic methods that ignore differential sequence evolution. Finally, we apply SDevo to single-time-point, multi-region sequencing data from clinical hepatocellular carcinomas and find evidence of a three- to six-times-higher division rate on the tumour edge. With the increasing availability of high-resolution, multi-region sequencing, we anticipate that SDevo will be useful in interrogating spatial growth restrictions and could be extended to model non-spatial factors that influence tumour progression.