SF3B1 mutations in the Driver Clone Increase the Risk of Evolution to Myelofibrosis in Patients with Myeloproliferative Neoplasms (MPN)

Introduction:

Next generation sequencing (NGS) studies identified additional somatic mutations impacting disease evolution and prognosis in MPN. SF3B1, a component of the U2 small nuclear ribonucleoprotein splicing complex, is frequently mutated in myelodysplastic syndromes where it has been proposed to define a new entity (Malcovati et. al. Blood 2020). In MPN, SF3B1 is mutated in approximately 10% of patients with primary myelofibrosis (PMF) and 3-5% with polycytemia vera or essential thrombocytemia (ET). Recent reports suggested that spliceosome mutations may adversely affect myelofibrosis free survival (MFS) in ET (Tefferi et. al. Br J Haematol. 2020). The main objective of this study was to evaluate the impact of the concomitant presence of driver (JAK2, MPLor CALR) and SF3B1 clonal or sub-clonal mutations on MPN phenotype and evolution in a large single center cohort of MPN patients.

Methods:

A total of 1243 consecutive patients were diagnosed with MPN according to WHO criteria between January 2011 and May 2020 in our center, of whom 707 had molecular analysis by NGS targeting a panel of 36 myeloid genes performed at diagnosis and/or during follow-up. Significant variants were retained with a sensitivity of 0.5%. Patients were grouped according to variant allele frequencies (VAF) determined by NGS as “driver” SF3B1mutated patients when driver and SF3B1 mutations VAF were similar (double mutated clone), and as “non-driver” SF3B1 mutants when SF3B1VAF was lower than that of the driver mutation, suggestive of a sub-clone. 4 patients with SF3B1 mutations but no driver mutation were excluded. We then compared the characteristics and outcomes of 3 groups of patients according to their SF3B1 mutational status: wild type (WT), driver and non-driver SF3B1 mutations.

Results:

A total of 39/703 (5.6%) patients had SF3B1 mutations, of whom 11/39 (28.2%) and 28/39 (71.8%) harbored driver and non-driver SF3B1mutations, respectively.

Driver SF3B1 mutations were associated with PMF (OR 6.1, _95%CI [1.1; 33.6], p= 0.039) and MPN unclassified (OR 15.6,_95%CI [1.1; 116.5], p= 0.007) subtypes, presence of immature myeloid cells ≥ 2% (OR 9.3, CI [2.6; 32.8], p= 0.001) and peripheral blasts ≥ 1% (OR 5.0,_95%CI [1.0; 24.7], p= 0.047) at diagnosis (Figure A). Other variables were not significantly different between patients with driver, non-driver and WT SF3B1, including age, driver mutation type, MPN-related symptoms, cytogenetics and high molecular risk mutations (ASXL1, EZH2, SRSF2, IDH1/2or U2AF1). There was no significant difference in the response to therapy: complete hematological response was seen in 5/11 (45.5%), 10/28 (35.7%) and 327/664 (49.3%) of patients with driver, non-driver and WT SF3B1 respectively.

After a median follow-up of 103.7 months IQR [47.2; 175.6], evolution to myelofibrosis occurred in 5/7 (71.4%), 6/20 (30.0%) and 99/564 (17.6%) of patients with driver, non-driver and WT SF3B1 respectively. Interestingly, driver SF3B1 but not non-driver SF3B1 mutational status adversely impacted MFS (OR 7.56,_95%CI[2.95; 19.38], p<0.001)(Figure B). Other variables independently associated with adverse MFS in multivariate COX regression analysis included age at MPN diagnosis (OR 1.02, _95%CI[1.00; 1.04], p=0.003), JAK2V617F allele burden (OR 1.03, _95%CI[1.02; 1.04], p<0.001), MPL (OR 13.94, _95%CI[4.90; 39.70], p<0.001) and CALR (OR 7.06, _95%CI[3.69; 13.51], p<0.001) mutations. SF3B1 mutational status had no impact on overall survival, transformation to MDS/AML or thrombotic/hemorrhagic events free survival.

Conclusion:

This study in a large cohort of MPN patients highlights for the first time to our knowledge the adverse impact on MFS of SF3B1 and MPN-driver co-mutated clones. In contrast, presence of an SF3B1 mutation at sub-clonal level didn’t increase the risk of MF development. In line with our findings, a recent study reported an association between rapid progression to myelofibrosis and SF3B1 mutations in patients with age-related clonal hematopoiesis (Bartels et al. Leukemia 2020). Further studies are warranted to confirm our results on independent cohorts and to investigate the mechanisms of bone marrow fibrosis development in patients with SF3B1 and MPN-drivermutations.