[Comparative study of the prognostic value of survival prognostic scoring systems in patients with essential...

Objective: To validate and compare the applicability of the international prognostic score for ET (IPSET), the mutation-enhanced international prognostic systems for ET (MIPSS-ET), triple A (AAA), and triple A plus (AAA+) survival prognostic scoring systems in Chinese patients with essential thrombocythemia (ET) . Methods: In this external validation and comparative study of clinical prediction models, 423 patients with ET who visited the Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College between January 2015 and December 2022 were retrospectively analyzed. Risk stratification was performed using the IPSET, MIPSS-ET, AAA, and AAA+ survival prognostic scoring systems. Harrell's concordance index (C-index), Brier score, net reclassification improvement (NRI), and integrated discrimination improvement (IDI) were used to comprehensively evaluate and compare the predictive discrimination, calibration, and risk restratification efficacy of each model for overall survival (OS) . Results: The median age at diagnosis of the enrolled patients was 50 years [interquartile range (IQR) : 34-62], and 60.8% (257 cases) were female. Overall, 22.0% (93 cases) of the patients had a history of thrombosis prior to diagnosis. The mutation frequencies of the driver genes JAK2V617F, CALR, and MPL were 56.7% (240 cases), 17.0% (72 cases), and 5.4% (23 cases), respectively, while 21.3% (90 cases) of the patients had no driver gene mutations. After a median follow-up of 63.1 (95% CI: 60.9-68.5) months, 16 patients (3.8% ) died during the follow-up period, with an estimated 5-year OS rate of (97.4±0.8) %. IPSET, MIPSS-ET, AAA, and AAA+ scoring systems could effectively discriminate the OS of patients in different risk groups (all P<0.001) ; the C-index for OS prediction was 0.756 for IPSET, 0.775 for MIPSS-ET, 0.762 for AAA, and 0.823 for AAA+, and the 5-year Brier scores were 0.023 5, 0.022 7, 0.023 2, and 0.022 6, respectively. Regarding risk restratification, the NRI of the AAA+ scoring system compared with the IPSET, MIPSS-ET, and AAA systems were 0.486, 0.352, and 0.268, respectively. The AAA+ system could accurately reclassify some patients who were categorized as low- or intermediate-risk by other systems into higher-risk groups, and the actual OS of these up-stratified patients was significantly inferior to that of the non-up-stratified group. Conclusions: The IPSET, MIPSS-ET, AAA, and AAA+ survival prognostic scoring systems are all effective for assessing survival risk in Chinese patients with ET. Among them, the AAA+ system exhibited optimal discrimination, calibration, and more accurate risk restratification efficacy in OS prediction, making it more suitable for individualized survival prognostic assessment in the Chinese ET population.