P442: ddx41 germline variants: population prevalence, significance and leukemic evolution

Background: Mutations in DDX41 have been linked to the development of myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML). Recent studies have investigated the frequency, spectrum and prognostic significance of DDX41 germline variants in MDS/AML cases. However, much less is known about their prevalence and significance in the general population and several key questions remain unanswered, including the AML/MDS risk in carriers, association with other cancers, the path to leukaemia evolution and the prospects for early detection of those en route to leukaemia. Aims: To determine the frequency of different DDX41 germline variants in the population, estimate their associated risks for MDS/AML or other cancers, investigate links to clonal hematopoiesis and leukemic evolution, and identify early markers of leukemic progression. Methods: We mined DDX41 germline variants from whole exome sequencing data of 454792 United Kingdom Biobank (UKB) participants with detailed linked phenotypic and blood test data available. We also used whole genome sequencing (WGS) data of 153 AML cases to compare somatic mutation rates in DDX41-mutant vs sporadic AML. Results: We identify 3553 non-synonymous variants in 3538 UKB participants (carrier rate of ~1 in 129). Many of the 452 unique variants identified were novel and variant prevalence varied significantly with ancestry. Of the 33 DDX41 germline pathogenic variant (GPV) carriers that developed MDS/AML, 7 had a start-lost, 22 had a truncating (including 1 splice site variant) and 4 had a missense variant, giving these mutation types odds ratios for MDS/AML development of ~13, 15 and 7 respectively, compared to non-carriers. Using logistic regression analysis we found significant associations of DDX41-GPV with MDS, AML and “family history of leukaemia”, but not lymphoma, myeloproliferative neoplasms (MPNs) or other cancers. Baseline blood count parameters did not differ between DDX41 variant carriers and controls, but among DDX41-GPV carriers, participants who later developed MDS/AML had a higher mean red cell volume (MCV) compared to participants who did not. The presence of somatic DDX41 mutations in blood DNA was associated with a higher risk of developing MDS/AML. However, we did not identify any instances of DDX41-R525H (the most common DDX41 somatic variant in MDS/AML), suggesting that leukemic progression may be faster with this variant and we present data supportive of this premise. Next, we investigated the prevalence of clonal haematopoiesis (CH) in DDX41-GPV carriers versus non-carriers in the general population and found no difference. In fact, CH was significantly more common prior to sporadic than DDX41-mutant AML/MDS, revealing distinct evolutionary paths. Furthermore, we analysed somatic mutation rates using the WGS data of 153 AML samples and did not find increased mutagenesis in DDX41-mutant vs sporadic cases. Summary/Conclusion: We map the landscape of DDX41 germline variants in a cohort of 454792 volunteers and show that carriers are at increased risk of MDS/AML, but not other cancers (including MPN and lymphoma). We estimate variant-specific relative and absolute risks of MDS/AML and show that DDX41-mutant differs to sporadic MDS/AML evolution, but is not linked to increased mutagenesis. We also report that DDX41-GPV carriers have normal baseline blood counts, but those en route to MDS/AML often had higher MCV or harboured somatic DDX41 mutations. Keywords: Familial, Myeloid malignancies, DDX41