In September 2023, a study led by Professor Qian Jiang from Peking University Institute of Hematology was published in the international academic journal ——Leukemia(IF=11.4). The title of the study is “External Validation of Predictive Scoring Systems for Molecular Responses in Chronic Myeloid Leukemia Receiving Imatinib Therapy“. This study not only solidifies the role of predictive scoring models in the personalized management of CML patients but also opens avenues for future research aimed at refining these models and exploring their applicability across different therapeutic agents and disease phases.
This comprehensive study embarks on a mission to craft and rigorously externally validate predictive scoring models specifically designed for patients in the chronic phase of chronic myeloid leukemia (CML), who are embarking on their journey with initial imatinib therapy. The essence of this research is to neatly stratify subjects into distinct risk cohorts—low, intermediate, and high—with marked disparities in the cumulative incidences of major molecular response (MMR) and MMR4, paving the path toward a more tailored therapeutic approach.
Utilizing a substantial dataset comprising 2184 subjects diagnosed with chronic phase CML and initiating treatment with imatinib, this study meticulously validates the predictive scoring models developed through the refined lens of the Fine-Gray model. Subjects were stratified into risk cohorts based on their MMR and MR4 scores, drawing from a comprehensive analysis of co-variates detailed in Supplementary Table 1, ensuring a robust methodological framework.
The study’s results are illuminating, revealing that a significant majority (64%) of subjects achieved an MMR, and a substantial proportion (43%) maintained a stable MR4. Over a seven-year horizon, cumulative incidences of MMR and MR4 were observed at 70% and 55%, respectively. Risk cohorts were delineated based on predictive scoring, leading to significant disparities in outcomes, with a lower platelet concentration notably associated with reduced MMR incidences, as further detailed in Supplementary Table 2.
These findings underscore the efficacy of the predictive scoring models in identifying patients at an elevated risk of suboptimal molecular responses to initial imatinib therapy. This insight is pivotal, enabling clinicians to tailor treatment strategies more precisely, thereby enhancing patient outcomes. The association of lower platelet concentration and co-morbidities with diminished MMR and MR4 incidences further enriches the clinical utility of these models.
(Leukemia . 2023 Sep;37(9):1922-1924. )
The model’s predictive performance was meticulously evaluated through a series of ROC curves, calibration plots, and decision curve analyses (DCAs), demonstrating commendable predictive accuracy for both MMR and MR4 outcomes. The study further validates the models’ independence and accuracy in predicting cumulative incidences of MMR and MR4, with concordance analyses revealing significant insights into risk stratification concordance between MMR and MR4 cohorts.
An in-depth exploration of the impact of various co-morbidities and baseline characteristics on the predictive accuracy of the scoring models reveals nuanced insights into the multifactorial nature of response to imatinib therapy. This section delves into the biological and clinical rationale behind the observed associations, offering a richer understanding of the underlying mechanisms and potential pathways for intervention.
This study not only solidifies the role of predictive scoring models in the personalized management of CML patients but also opens avenues for future research aimed at refining these models and exploring their applicability across different therapeutic agents and disease phases. The implications for clinical practice are profound, offering a tangible framework for risk-adapted therapy that promises to elevate patient care standards.
