Prognostic Impact of Splicing Factor Mutations in Patients with Myelofibrosis Undergoing Allogeneic Hematopoietic Stem Cell Transplantation

Background: Recurrent somatic mutations in splicing machinery components, including SRSF2, SF3B1 and U2AF1 have been recently identified and frequently found in a large spectrum of myeloid malignancies. The highest reported frequencies were 85% for myelodysplastic syndromes associated with ring sideroblasts (MDS-RS), 44% for MDS without RS and 55% for chronic myelomonocytic leukemia (CMML), 26% in therapy-related MDS or acute myeloid leukemia, 7% in primary acute myeloid leukemia (AML), and 9% in myeloproliferative neoplasms (MPNs). In primary myelofibrosis (PMF), SRSF2 mutations are associated with advanced age, higher DIPSS scores and poor prognosis. On the other hand, SF3B1 and U2AF1 mutations had no prognostic relevance for either overall or leukemia-free survival.

Aim: The aim of this study was to investigate the clinical characteristics and prognosis of splicing factor mutations in SRSF2, SF3B1 and U2AF1 in patients with primary myelofibrosis or post ET/PV myelofibrosis that underwent allogeneic stem cell transplantation (alloHSCT).

Methods: 158 patients with WHO defined primary myelofibrosis (n=112, 71%) or post-ET/post-PV myelofibrosis (n=46, 29%) underwent allogeneic hematopoietic stem cell transplantation. Bone marrow or peripheral blood samples were obtained before transplantation. U2AF1, SRSF2, and SF3B1 were amplified by PCR and were sequenced and confirmed with Sanger sequencing besides JAK2, CALR and MPL.

Results: Mutations in SRSF2, SF3B1 and U2AF1were detected in 12 (8%), 6 (4%) and 9 (6%) patients, respectively. In total, 27 patients (18%) carried at least one of the three main spliceosome pathway mutations. Baseline characteristics were similarly distributed between SRSF2, SF3B1 and U2AF1 mutated and wildtype patients, respectively (primary vs secondary myelofibrosis, donor match, cytogenetic characteristics), except for higher median age of U2AF1 mutated compared to wildtype patients (67 vs. 58 years, P=0.023). We observed no associations between the presence of SRSF2/SF3B1/U2AF1 mutations and the presence of JAK2, CALR and MPL mutation status, respectively. The majority of patients (n=110, 70%) were transplanted with a matched donor (33 from related and 77 from unrelated donors). 48 patients (30%) received a mismatched unrelated allograft. Univariate analysis disclosed no differences in cumulative incidence of relapse (CIR) between mutated and wildtype patients (CIR: SRSF2, P=0.78; SF3B1, P=0.19; U2AF1, P=0.51; all three splicing genes, P=0.83), although none of the six SF3B1 mutated patients relapsed after alloHSCT. Non-relapse mortality (NRM) was not significantly different between mutated and wildtype patients (NRM: SRSF2, P=0.58; SF3B1, P=0.6; U2AF1, P=0.11; all three splicing genes, P=0.77). Furthermore, patients with mutations in one or any of the three main spliceosome genes had a similar 2-year overall survival after allogeneic HSCT as wildtype patients (2y-OS mutant vs wildtype: SRSF2, 75% vs 68%, P=0.8; SF3B1, 83% vs 69%, P=0.4; U2AF1, 38% vs 70%, P=0.13; all three splicing genes, 68% vs 70%, P=0.78).

Conclusions: In our study we show that (i) splicing gene mutations are recurrent molecular aberrations in myelofibrosis patients, (ii) are not associated with an adverse prognostic outcome in patients with myelofibrosis undergoing allogeneic transplantation, and (iii) the potential negative prognostic effect of SRSF2 mutations can be overcome by alloHSCT. Splicing gene mutations do not constitute a risk factor for alloHSCT in patients with myelofibrosis.