Validation of thrombotic risk factors in 1381 patients with essential thrombocythaemia: A multicentre retrospective real-life study.

Thrombosis and haemorrhage are frequent in patients with essential thrombocythaemia (ET). The 2016 revised International Prognostic Score for Thrombosis in Essential Thrombocythaemia-thrombosis (r-IPSET-t) score stratifies patients into very-low- (VLR), low- (LR), intermediate- (IR) and high-risk (HR) groups. We validated the r-IPSET-t in the biggest population of patients with ET (n = 1381) to date and found it to be a better fit than the earlier IPSET-t score. With an average follow-up of 87.7 months, there were 0.578 thrombotic events/person-year and 0.286 bleeding events/person-year after diagnosis. The 10-year thrombosis-free survival was 88% and 99% for the r-IPSET-t LR and VLR groups (p < 0.001). Cytoreduction was a thrombotic risk factor in younger patients (aged <60 years, hazard ratio 9.49, p = 0.026; aged ≥60 years, hazard ratio 1.04, p = 0.93). In multivariable Cox regression analysis, anti-aggregation after diagnosis was protective for thrombosis (hazard ratio 0.31, p = 0.005) but a risk factor for major bleeding (hazard ratio 10.56, p = 0.021). Of the IPSET-t HR and LR groups, 132/780 and 249/301 were re-classified as LR and VLR respectively (p < 0.001). The European LeukemiaNET (ELN) does not recommend aspirin for VLR patients but in this real-life analysis 83.1% of VLR patients received it. Our results validate the r-IPSET-t score as more predictive for thrombosis than the ELN-recommended IPSET-t and raise concerns about unnecessary cytoreductive and anti-aggregative therapy.