Prediction of thrombosis in polycythemia vera: Development and validation of a multiple factor-based prognostic score system
Wenjing Gu, Yuhui Zhang, Ting Sun, Mankai Ju, Xiao Fan Liu, Feng Xue, Yunfei Chen, Wei Liu, Huiyuan Li, Wentian Wang, Ying Chi, Renchi Yang, Rongfeng Fu, Jie Bai, Lei Zhang
Abstract
Background Thrombosis is an important cause of death in patients with polycythemia vera (PV). The conventional stratification of thrombosis may ignore some potential risk factors. Objectives This study aimed to develop and validate a multiple factor-based prediction model of thrombosis for the 2016 World Health Organization-dened PV. Methods Clinical and next-generation sequencing data from 2 cohorts of patients with PV were analyzed. Multivariable Cox regression analyses were conducted for the identification of thrombotic risk factors and model development. Results The study involved 372 patients in the training cohort and another 195 patients in the external validation cohort. Multivariable analyses indicated that age ≥60 years (hazard ratio [HR] 2.56, 95% CI 1.51-4.35, P < .001), cardiovascular risk factors (HR 4.22, 95% CI 2.00-8.92, P < .001), at least 1 high-risk mutation for thrombosis (mutations in DNMT3A , ASXL1 , or BCOR/BCORL1 ) (HR 4.35, 95% CI 2.62-7.21, P < .001), and previous thrombosis (HR 5.93, 95% CI 3.29-10.68, P < .001) were independent risk factors of thrombosis. After assigning coefficient-weighted scores to each risk factor mentioned above, a multiple factor-based prognostic score system of thrombosis (MFPS-PV) was developed, classifying patients into low-risk, intermediate-risk, and high-risk groups. Patients in the 3 groups had notably different thrombosis-free survival rates ( P < .001). The MFPS-PV outperformed the conventional model in discrimination power (C-statistic: 0.87 [95% CI 0.83-0.91] vs 0.80 [95% CI 0.74-0.86]). The MFPS-PV was well calibrated and remained consistent during external validation. Conclusion The MFPS-PV, integrating genetic and clinical characteristics for the first time, shows excellent accuracy and utility for thrombosis prediction in WHO-defined PV.