Mutations of the Spliceosome Complex Genes Occur In Adult Patients but Are Very Rare In Children with Myeloid Neoplasia

Abstract 2797

Somatic point mutations occur frequently in adult patients with myelodysplastic syndromes (MDS) and are associated with distinct clinical features such as overall survival, genomic aberrations and clonal transformation. Recently a number of new genes had been identified which are involved in methylation and chromatin regulation (TET2, DNMT3A, ASXL1). Although mutations of these genes are present at high frequency in adult cohorts, they are very rare events in children with myeloid disease. Based on the results from a large-scale whole exome sequencing study of adult patients with myeloid neoplasia, a new group of genes involved in the formation of spliceosome complex was discovered to be recurrently somatically mutated in approximately 40% of patients with MDS and 55% of chronic myelomonocytic leukemia (CMML) cases (Yoshida, et al., unpublished data). Spliceosomes are macromolecular RNA-protein complexes that remove noncoding introns from precursor mRNA. We hypothesized that the disruption of the spliceosome complex might play a driving role in the leukemogenesis in pediatric MDS. Using targeted re-sequencing we investigated the 3 exclusive hotspots of 2 spliceosome genes that were found to be mutated in adult MDS: U2AF1 (S34), U2AF1 (Q157) and SFRS2 (P95).

We analyzed a total of 339 WHO-defined pediatric cases: juvenile myelomonocytic leukemia (JMML), n=112; secondary CMML, n=12; Down syndrome-associated MDS, n=6; de novo acute myeloid leukemia (AML), n=12, MDS-related AML, n=50; refractory anemia with excess blasts (RAEB), n=52; RAEB in transformation (RAEB-T), n=11; and refractory cytopenia (RC), n=84.

We found two heterozygous missense mutations: SFRS2 c.284C>T/P95L in a patient with JMML, and U2AF1 c.101C>A/S34Y in a patient with refractory cytopenia. The first patient presented with non-syndromic JMML with somatic mutation of PTPN11 c.226G>C/E76Q, normal cytogenetics, increased HbF(80%), splenomegaly, monocytosis, and was 4.5 years old at diagnosis. The patient underwent successful SCT. The analysis of buccal-swab DNA confirmed the somatic status of the SFRS2 mutation. The second patient was diagnosed at age of 17 years with systemic mastocytosis positive for c-kit D816V mutation, and with associated clonal hematological non-mast cell lineage disease: refractory cytopenia with normal cytogenetics. SCT was also performed with success on this patient. To ensure that the low rate of mutations found in our pediatric cohort was not due to technical issues, we also sequenced major mutation sites of U2AF1 and SFRS2 genes in 19 adult patients with MDS: refractory anemia (RA), n=8; RC with multilineage dysplasia (RCMD), n=3; RAEBI/II, n=5; CMML, n=3. We found 4 heterozygous changes: SFRS2 missense mutation c.284C>T/P95L in two patients with RA and RAEBII; SFRS2 deletion c.284_307/P95RfsX120 in one RA patient, and U2AF1 c.470A>C/Q157P in one RA patient. Interestingly, 3 out of 4 patients with a mutated spliceosome gene harbored a mutation of the ASXL1 gene, but none was mutated for DNMT3A.

In summary, re-sequencing of 3 spliceosome gene hotspots revealed the presence of heterozygous mutations in 2/339 children and 4/19 adults with myeloid disease. The analysis of another gene-hotspot implicated in the MDS whole exome study, SF3B1 K700 is ongoing, but all pediatric cases analyzed to date were negative. The drastically reduced frequency of spliceosome mutations in pediatric compared to adult myeloid malignancies suggests a different pathogenetic mechanism in childhood disease, and fits well with previous reports that somatic mutations of non-Ras-pathway genes, such as DNMT3A, are less prevalent in pediatric cohorts. The functional impact of spliceosome mutations on leukemogenesis warrants further study.