DDX41 is thought to be a tumor suppressor gene involved in pre-mRNA splicing, innate immunity and rRNA processing. Myeloid neoplasms with germline DDX41 mutations have been included as a new diagnostic category in the 2016 WHO classification. However, there are limited studies describing the mutation profile of myeloid neoplasms and acute leukemias associated with DDX41 mutation. We analyzed the prevalence and characteristics of DDX41 mutations in an unselected cohort of 1764 patients with myeloid neoplasms and acute leukemias, including 720 subjects with AML, 91 with MDS, 41 with MPN, 16 with MDS/MPN, 760 with ALL, and 42 with MPAL. Next-generation sequencing was performed on 86 genes closely related to hematologic neoplasms. The fingernail specimens or blood samples in remission were taken as control samples to verify the mutation from possible germline sources.
We identified 21 different DDX41 mutations in 16 unrelated patients (6 MDS/AML, 1 CMML, 9 ALL) that were classified as causal (n=17) and uncertain significance (n=4) variants. The acquisition of a somatic DDX41 mutation was also considered as a very strong criterion for causality, the uncertain significance variants were excluded. Nine causal variants have not been reported. 53% of variants were located on the DEAD domain and 24% on the Helicase C domain, the rest were located upstream of the DEAD domain. Ten variants were germline that the majority (80%) were located upstream of the Helicase C domain, 7 variants were somatic and were scattered.
In 6 patients with MDS/AML and DDX41 mutations, the median age was 49 years (range, 28-78y) and 57% were male. None of the patients had del 5/5q. Five (83%) patients had personal history of cytopenia prior to MDS/AML diagnosis, while only one patient had a family history of anemia and one patient's aunt died of leukemia. Four (67%) patients harbor DDX41 germline/somatic biallelic mutation, two with typical biallelic mutation (N-terminal germline nonsense and C-terminal somatic missense), the other two with atypical biallelic mutation (N-terminal germline missense and C-terminal somatic missense). The average age of patients with DDX41 atypical biallelic mutation (48y) seems lower than that with typical biallelic mutation (74y). The rest two patients harbor single germline mutations and one of them concomitant with SF3B1 mutation, which is a component of spliceosome complex also involving in mRNA splicing. DDX41 mutations were identified in 7 patients with B-ALL and one with T-ALL. The median age was 9 years (range, 4-2 y) and 56% were male. None of the patients had a family history of hematological malignancy and del 5/5q. Unlike in myeloid neoplasms, no DDX41 biallelic mutations were identified that 5 patients had single somatic mutation (3 missenses, 1 nonsense) and 4 had single germline mutation (all are missenses).
Among MDS/AML patients with DDX41 biallelic mutation, only one received treatment who relapsed after HSCT and received second HSCT, the time of overall survival (OS) was 74 months, the other 3 quite after diagnosed. In patients with MDS/AML and DDX41 single germline mutation, one received 4 courses of treatment with decitabine and half-dose CAG regimen, then transformed to AML and abandoned, the other one received 10 courses of chemotherapy and showed continuous no remission. The time of OS was 17 and 31 months, respectively. Among ALL patients with DDX41 single somatic mutation, 80% (4/5) received HSCT, 80% (4/5) were in complete remission (CR), one died of post-transplant infection, the median OS was 25 months. Among ALL patients with DDX41 single germline mutation, all the three patients received HSCT and were in CR, the median OS was 37 months.
The genotype-phenotype correlations regarding germline DDX41 mutations should be clarified more specifically, the most prevalent loss of function mutations, predisposes to myeloid disease at the same age as sporadic disease, whereas point mutations in the DEAD domain (this study) or helicase C domain (previous report) were speculated to cause earlier onset disease. Moreover, this study reported for the first time that DDX41 mutations have also been found in ALL, which expanded its phenotypic spectrum. The characteristics of DDX41 mutation in ALL are different from myeloid neoplasm, the age of onset is young, and no germline/somatic biallelic mutation have been observed, suggesting that it might be involved in different pathogenesis mechanisms.
Disclosures
No relevant conflicts of interest to declare.
Biallelic mutation of
FBXL7
suggests a novel form of Hennekam syndrome
