Frontiers in Oncology (Jan 2024)

Risk stratification for early mortality in newly diagnosed acute promyelocytic leukemia: a multicenter, non-selected, retrospective cohort study

  • Suhyeon Kim,
  • Jiye Jung,
  • Seo-Yeon Ahn,
  • Mihee Kim,
  • So Yeon Jeon,
  • Chang-Hoon Lee,
  • Dae Sik Kim,
  • Se Ryeon Lee,
  • Hwa Jung Sung,
  • Chul Won Choi,
  • Byung-Soo Kim,
  • Hyeoung-Joon Kim,
  • Jae-Yong Kwak,
  • Yong Park,
  • Jae-Sook Ahn,
  • Ho-Young Yhim

DOI
https://doi.org/10.3389/fonc.2024.1307315
Journal volume & issue
Vol. 14

Abstract

Read online

IntroductionDespite the current effective treatments for acute promyelocytic leukemia (APL), early mortality (EM), defined as death within 30 days of presentation, is a major hurdle to long-term survival.MethodsWe performed a multicenter retrospective study to evaluate the incidence and clinical characteristics of EM in patients with newly diagnosed APL and to develop a risk stratification model to predict EM.ResultsWe identified 313 eligible patients diagnosed between 2000 and 2021 from five academic hospitals. The median age was 50 years (range 19-94), and 250 (79.9%) patients were <65 years. Most patients (n=274, 87.5%) received their first dose of all-trans retinoic acid (ATRA) within 24 hours of presentation. EM occurred in 41 patients, with a cumulative incidence of 13.1%. The most common cause of EM was intracranial hemorrhage (n=22, 53.6%), and most EMs (31/41, 75.6%) occurred within the first seven days of APL presentation. In a multivariable analysis, we identified three independent factors predicting EM: age ≥65 years (HR, 2.56), white blood cell count ≥8.0 x 109/L (HR, 3.30), and ATRA administration >24 hours of presentation (HR, 2.95). Based on these factors, patients were stratified into three categories with a significantly increasing risk of EM: 4.1% for low risk (54.3%; no risk factors; HR 1), 18.5% for intermediate risk (34.5%; 1 factor; HR 4.81), and 40.5% for high risk (11.2%; 2-3 factors; HR 13.16).DiscussionThe risk of EM is still not negligible in this era of ATRA-based therapies. Our risk model serves as a clinically useful tool to identify high-risk patients for EM who may be candidates for novel treatments and aggressive supportive strategies.

Keywords