The incidence of SF3B1 mutations in patients with RARS is high. Recently, it has been shown that SF3B1 and DNMT3A mutations overlap more often than expected, although it is not clear how this could affect the disease. We studied SF3B1 and DNMT3A in 123 RARS patients: 101 out of 123 samples (82%) had somatic mutations in SF3B1, and 13 of them (13%) showed a co-mutation (SF3B1mutDNMT3Amut). All co-mutated patients had a normal karyotype, and 12 of them (92%) were lower-risk patients (IPSS and IPSS-R). Despite their favorable profile, SF3B1mutDNMT3Amut patients showed a higher RBC transfusion dependency (92% versus 48%, p = .007), a shorter overall survival (OS) (median, 30 versus 97 months, p = .034), and a higher risk of progression to acute myeloid leukemia (AML) at 5 years (25% versus 2%, p = .023) than SF3B1mutDNMT3Awt patients. In conclusion, DNMT3A mutations are present in a significant proportion of SF3B1mut patients with a negative clinical impact.
Keywords: DNMT3A mutations; SF3B1 mutations; methyltransferase domain; ring sideroblasts.