Exploring the spectrum of non-malignant complications in germline DDX41-associated hematologic disorders

Abstract

Abstract First discovered by our group in 2015 Cancer Cell. 2015 May 11;27(5):658-701 in an index family with history of MDS/AML, DDX41 is mutated in a significant proportion of hereditary myeloid neoplasia. The presence of germline loss of function DDX41 mutations identifies a group of AML with male predominance, late onset, and favorable prognosis. These patients show concurrent somatic hits, predominantly R525H. Individuals who inherit DDX41 pathogenic variants have an increased lifetime risk to develop MDS and AML. It has been also noted an increased risk of developing other malignancies including the ones affecting the lymphoid system and multiple myeloma. However, given the RNA helicase function of DDX41 and the various pathways in which this function can play a role, one can envision that the effects produced by germline DDX41 variants might be broader and also affect other organ systems. To date, extra-hematologic manifestations of inherited DDX41 alterations have not been systematically studied. Clearly, such an analysis is difficult due to expected long disease anticipation and thereby competing morbidities. Similarly, in analogy to the hematopoietic consequences of inherited mutations, the effects of DDX41 hits may be subtle and without hematologic disease difficult to differentiate. Thus, the current study is limited to carriers with hematologic abnormalities. Here we have reviewed 154 cases with DDX41 mutations obtained through genetic screen of 11,500 cases with hematologic symptoms or overt hematologic disease. Of those cases 61% had single hits considered pathogenic and suspicious VUS. Molecular data, clinical parameters, past medical and family history were reviewed. We interrogated presence of signs of immunodeficiency, autoimmune, metabolic, neurologic and cardiovascular disease (CVD) both in hematologic patients. We found patients carrying single germline DDX41 to have other associated manifestations such as immune (20%), neurologic (25%), rheumatologic (30%), and acute renal injury and chronic kidney disease (15%). Particularly striking was the presence of CVD. Of note is that prior integrative analyses of disease causing gene variants and phenotypes among individuals with CVD identified DDX41 to be part of a set of genes clustering with CVD. Moreover, RNA expression analysis showed that DDX41 levels were decreased in CVD patients compared to controls.Genomics. 2023 Mar;115(2):110584.2 In our cohort, in 95 patients with single germline, 38 cases had documented past medical history of CVD (40%). Median age was 72 [33-91]; 1:2.2 M:F. In 34% of these patients there was also family history of CVD but by comparison family history of blood disease and cancer was present in 42% of the cases. Patients' diagnoses were AML (n=9), MDS (n=8), MDS/MPN (n=9) and history of cytopenias (n=12). CVD was registered as atherosclerotic vascular disease (ASVD; n=22), ASVD with heart failure (HF; n=2), ASVD with conduction abnormality (n=1), conduction abnormality (n=7), HF (n=4) and structural cardiovascular abnormality (n=2). In total ASVD were 66% (25/38) of the total CVD. To validate our findings, we sought a cohort with high numbers of germline DDX41 (NIH All of Us Research Program) to extract data from 1453 participants. 50% (717/1453) of carriers were affected by CVD (Median age: 61 [20-100]; 1:1.42 M:F). We then analyzed the molecular profile of our cohort of patients with germline DDX41 and CVD. The most common DDX41 hits were missense (n=24), c.3G>A (n=7), D140fs (n=3), and splice sites (n=4). The most common somatic events following germline DDX41 were in TET2 (9/38; 24%; Median VAF: 37% [5.1-52] and ASXL1 (7/38; 18%; Median VAF: 22% [6.2-31]). Although the role of DDX41 in CVD is unknown, the most plausible explanation resides in the function of DDX41 in cGAS/STING mediated interferon inflammation linked to adverse outcomes in coronary artery disease. The role of DDX41 in cGAS/STING mediated pathway is part of a body of literature also known in MDS and AML; although still under investigation. In conclusion, although our study has caveats (limited sample size, retrospective nature of study, age of our cohort, comorbidity, evaluation of risk factors) we believe that it is worth noting the frequency of CVD in our patients and given the involvement of both DDX41 and TET2 genes in the inflammatory network, further studies are warranted. We also thank the NIH All of Us Research Program for making available the participant data.