P198 | BONE MARROW FIBROSIS IN POLYCYTHEMIA VERA AND ESSENTIAL THROMBOCYTHEMIA: CLINICAL, MOLECULAR AND PROGNOSTIC IMPLICATIONS

Background Mild BM fibrosis (MF-1) can be identified in up to 20% of patients (pts) with polycythemia vera (PV) and essential thrombocytemia (ET), and may play a role –albeit still poorly defined–in pathophysiology, clinical presentation ad long-term outcome. Aim&Methods To assess the prevalence of MF-1 and investigate its clinical, molecular and prognostic implications in PV and ET. All diagnosis were retrospectively confirmed, and ET cases in which a diagnosis of prefibrotic primary myelofibrosis (PMF) could not be excluded were omitted. Results The study included 989 patients with PV (n=507) and ET (n=482). Median age was 59 (range 19-92) years (PV/ET: 60/59), and 44 (61%) were male (PV/ET: 305 [60%]/187 [39%]). Targeted NGS analysis was available for 211 (21%) pts (PV/ET: 116 [23%]/95 [20%]). MF-1 was reported in 194 (19%) pts, including 126 (25%) and 68 (14%) PV and ET pts, respectively. In PV, MF-1 was associated with higher leukocytes (10.3 vs 9.5 x10⁹/L, P=.0250) and LDH (285 vs 257 U/L, P=.0016). The latter was confirmed also in ET (260 vs 238 U/L, P=.0450). While MF-1 was associated with higher JAK2^VF mutant burden in PV (41% vs 34%, P=.0074), ET pts with MF-1 had higher CALR mutant burden (51% vs 38%, P=.0087), with no differences in the prevalence and allele burden of the other driver mutations. Among the 211 NGS-annotated pts, PV pts with MF-1 were more likely to harbor mutations in SRSF2 (9% vs 0%, P=.0052) and NF-E2 (44% vs 3%, P=.0002), with no significant correlation in the ET cohort. The rate of arterial thrombosis (AT) was greater in PV pts without MF-1 (7% vs 2%, P=.0492), who showed accordingly a trend for higher 10-year cumulative incidence (CuI) of AT (7% vs 4%, Grey test P=.0596). We did not report any significant association with bleeding. Median follow-up was 7 (95% CI 6- 8) years. After competing risk analysis, the 10-year CuI of MF progression was significantly higher in MF-1 pts in both PV (24% vs 7%, Grey test P<.0001; Fig. 1A) and ET (20% vs 2%, Grey test P<.0001; Fig. 1B). In addition, MF-1 was associated with shorted overall survival (OS) in ET (14 years vs NR; HR 7.2, 95% CI 4.1-10.3, P<.0001; Fig. 1D), unlike in PV (Fig. 1C). Conclusions Increased BM fibrosis is more prevalent in PV than ET, and is associated with distinct clinical and molecular features, including higher JAK2^VF and CALR mutant burden, respectively. More importantly, MF-1 predicted increased risk of MF progression in both cohorts and shorter OS in PV.