Associations of Mutations in Spliceosome Genes With Ring Sideroblasts Verified in Myelodysplastic Syndromes

Following new proposals for the classification of myeloid malignancies, researchers have confirmed the association of mutations in the splicing pathway gene SF3B1 with low bone marrow blast counts and increased ring sideroblasts in myelodysplastic syndromes (MDS) and demonstrated significant associations between mutations in other splicing genes (SRSF2 and ZRSR2) with high bone marrow blast counts and low ring sideroblasts in MDS, according to a letter published in Leukemia.

The investigators weighed in on 2 new proposals for the classification of myeloid malignancies, the 5th edition of the WHO Classification (WHO 2022) and the International Consensus Classification (ICC).

Their aim was to address differences between WHO 2022 and ICC regarding MDS-SF3B1 and to evaluate the meaningfulness of the WHO term “MDS with low blasts and ring sideroblasts” (MDS-LB-RS) as an alternative for SF3B1 wild-type cases. They estimated the incidence of the MDS-LB-RS, presence of other splicing gene mutations, and clinical outcome.

They analyzed samples from an initial cohort of 704 patients with de novo MDS (58% men and 42% women; median age, 73 years; range 23-93; 58% with <5% bone marrow blasts). All samples underwent whole genome sequencing.

They also analyzed samples from a validation cohort of 1804 patients with de novo MDS (64% men and 36% women; median age, 76; range, 24-96; 56% with <5% bone marrow blasts). All samples underwent targeted panel sequencing during routine diagnostic testing.

Of the initial cohort, 660 of the 704 (94%) MDS cases had data available on the presence of ring sideroblasts. Of these cases, the researchers found 40% had ring sideroblasts ≥15%. They found 45% had low blasts (bone marrow blasts <5%) and also did not fulfill the criteria for the WHO 2022 entities MDS with low blasts and isolated 5q deletion (MDS-5q; n=98) or MDS with biallelic TP53 inactivation (MDS-biTP53; n=41). Among these patients, they observed ring sideroblasts <15% in 38% and ring sideroblasts ≥15% in 62%.

The team found splicing mutations in 97% of patients with low blast counts and ring sideroblasts ≥15%; these included SF3B1 (87%), SRSF2 (9%), U2AF1 (3%), ZRSR2 (3%). They found only 28% of patients with low blasts and ring sideroblasts <15% harbored mutations in at least 1 of the 4 analyzed splicing genes, SF3B1 in 4%, SRSF2 in 8%, U2AF1 in 8%, and ZRSR2 in 10%.

The researchers investigated the associations of the splicing gene mutations (SF3B1, SRSF2, U2AF1, and ZRSR2) with the presence of ring sideroblasts (<15% vs ≥15%) and bone marrow blasts (<5% vs ≥5%).

In the initial cohort, they found SF3B1 mutations were significantly associated with bone marrow blasts <5% (P <.001) and mutations in other splicing genes were significantly associated with bone marrow blasts ≥5% (SRSF2, P <.001 and U2AF1; P =.006; except ZRSR2, P =.083) In the validation cohort, the investigators observed similar associations (SF3B1, P <.001; SRSF2, P <.001; U2AF1, P =.003; ZRSR2, P =.037), with ZRSR2 mutations reaching statistical significance.

In the initial cohort, they found SF3B1 mutations, but not those in SRSF2, U2AF1, and ZRSR2, were significantly associated with ring sideroblasts ≥15% (P <.001). They observed a trend for the association of ring sideroblasts <15% with SRSF2 mutations (P =.056) and a significant association in the case of ZRSR2 mutations (P =.016). In the validation cohort, the team confirmed these findings (SF3B1, P <.001; SRSF2, P <.001; ZRSR2, P <.001), with SRSF2 mutations reaching statistical significance. In both cohorts they did not observe an association between U2AF1 mutations and ring sideroblasts.

“Together, these data clearly indicate that other splicing factor mutations cannot be used as a substitute for SF3B1 mutations and are not useful for the classification of MDS-LB-RS in the absence of SF3B1 mutations,” the researchers concluded in their report. “[We] suggest the alternative term “MDS with low blasts and ring sideroblasts” (MDS-LB-RS) as a second best classification only for cases when SF3B1 mutation analysis is not available as in this setting RS ≥15% represent a good but not perfect surrogate for SF3B1 mutations”

Reference

Huber S, Haferlach T, Meggendorfer M, et al. Mutations in spliceosome genes in myelodysplastic neoplasms and their association to ring sideroblasts. Leukemia. Published online December 3, 2022. doi:10.1038/s41375-022-01783-y

This article originally appeared on Hematology Advisor