C016 | GENOMIC PROFILING FOR CLINICAL DECISION MAKING IN POST-POLYCYTHEMIA VERA AND POST-ESSENTIAL THROMBOCYTHEMIA MYELOFIBROSIS.

Background: Myelofibrosis (MF) encompasses primary (PMF) and secondary MF (SMF). Most PMF patients have additional non-driver myeloid neoplasms-associated cancer genes variants (CGVs), some correlated with overall survival (OS). Aims: to assess the distribution and correlations of CGVs in SMF patients of the MYSEC (MYelofibrosis SECondary) project. Methods: A myeloid-panel next generation sequencing was performed. Characteristics of study cohort and associations were described by standard statistics. Kaplan-Meier curves with log-rank tests for comparison were used for exploratory time to event analyses. A semiparametric Cox model was used for regression. Results: Out of 644 SMF cases, 429 (66.6%) reported CGVs: 234 (36.3%) one, 120 (18.6%) two, 75 (11.6%) three or more. The most prevalent CGVs were in (Figure): ASXL1 (n=191, 29.7%), TET2 (n=128, 19.9%), DNMT3A (n=43, 6.7%), EZH2 (n=33, 5.1%), SF3B1 (n=30, 4.7%), U2AF1 (n=28, 4.3%), TP53 (n=27, 4.2%), IDH1/2 (n=25, 3.9%), CBL (n=22, 3.4%), ZRSR2 (n=22, 3.4%), RUNX1 (n=21, 3.3%), NRAS (n=16, 2.5%), CSF3R (n=14, 2.2%), SRSF2 (n=14, 2.2%), and NFE2 (n=13, 2%). Phenotype-genotype correlations of CGVs will be presented. The median follow-up was 4.2 years (range, 0.6-28.1). Median OS was 9.2 years (95%CI: 8.3-11.2). Among the most prevalent CGVs, patients carrying the following ones had a significantly reduced median OS compared to those who did not: ASXL1 (7.4 vs 10.8 years, p=0.003), EZH2 (6.0 vs 9.5 years, p=0.02), U2AF1 (2.5 vs 9.7 years, p<0.001), TP53 (2.5 vs 9.2 years, p=0.002) and SRSF2 (2.0 vs 9.5 years, p<0.001). In semi-parametric Cox models including these individual CGVs and the MYSEC-Prognostic Model (MYSEC-PM), four CGVs maintained their prognostic value: ASXL1 (HR 1.53, 95%CI: 1.16-2.00; p=0.02); U2AF1 (HR 2.94, 95%CI: 1.84-4.69; p<0.0001); TP53 (HR 2.24, 95%CI: 1.31-3.81; p=0.003) and SRSF2 (HR 4.30, 95%CI: 2.32-7.94; p<0.0001). With respect to ASXL1, CGVs in U2AF1, TP53 and SRSF2 genes collectively conferred a significantly worse impact on outcome (median OS 8.2 vs 3.9 years, p<0.01), also considering the MYSEC-PM strata in a Cox model (data will be shown). Conclusion: Our study shows that genomic profiling in SMF refines current risk-stratification allowing for enhanced clinical decision making.