Zhang Zizhong, Jiang Weiwei, Ding Sen, Shen Xuliang
Department of Haematology, Heping Hospital Affiliated to Changzhi Medical College, Shanxi Clinical Medical Research Center for Hematologic Diseases (Myeloproliferative Neoplasms), Changzhi, China.
Jining Institute of Education, Jining, China.
Medicine (Baltimore). 2025 Jul 18;104(29):e43120. doi: 10.1097/MD.0000000000043120.
This study aims to comprehensively assess the effects of imatinib, nilotinib, and flumatinib in treating chronic myeloid leukaemia and to explore the main factors affecting its efficacy. Ninety-nine chronic myeloid leukaemia patients initially diagnosed and treated with one of these 3 tyrosine kinase inhibitors at a tertiary hospital in Shanxi Province between June 2018 and June 2023 were selected and divided into an imatinib group (n = 32), nilotinib group (n = 30), and flumatinib group (n = 37). Hematological response rates, cytogentic response rates, molecular response rates, and adverse reactions were compared among the 3 groups to statistically analyze efficacy and safety. Univariate analysis and logistic regression were used to explore the related factors influencing the curative effect. A nomogram prediction model of influencing factors of efficacy was constructed in R software and validated according to receiver operating characteristic and calibration curves, with a clinical decision curve and clinical impact curve further drawn to confirm its clinical practicability. The complete cytogenetic response at 3 months differed significantly, with rates of 53.13%, 76.67%, and 78.38% for the imatinib, nilotinib, and flumatinib groups, respectively (P < .05). Major molecular response (MMR) rates at 3 months were 25.00%, 53.33%, and 51.35%, reaching 78.13%, 90.00%, and 83.78% at 12 months, respectively. Deep molecular response (DMR) rates at 12 months were 50.00%, 76.67%, and 75.68% in each respective group (P < .05). Multivariate logistic regression indicated early molecular response, white blood cell count, red cell distribution width and platelet count as independent influencing factors of MMR. Age, drug type, early early molecular response, and red cell distribution width were identified as independent influencing factors of DMR (P < .05). The areas under the receiver operating characteristic curves for MMR and DMR nomogram models were 0.912(95% confidence interval: 0.833, 0.990)and 0.874 (95% confidence interval: 0.801, 0.946), respectively, indicating satisfactory model calibration. Nilotinib and flumatinib demonstrate superior efficacy over imatinib, with effectiveness influenced by various factors including sociodemographic characteristics, clinical heterogeneity, and drug side effects. The proposed clinical prediction model may provide valuable insights for decision-making and demonstrates generalizability and practical application value.
本研究旨在全面评估伊马替尼、尼洛替尼和氟马替尼治疗慢性髓性白血病的效果，并探讨影响其疗效的主要因素。选取2018年6月至2023年6月在山西省某三级医院初诊并接受这3种酪氨酸激酶抑制剂之一治疗的99例慢性髓性白血病患者，分为伊马替尼组（n = 32）、尼洛替尼组（n = 30）和氟马替尼组（n = 37）。比较3组的血液学缓解率、细胞遗传学缓解率、分子学缓解率及不良反应，对疗效和安全性进行统计学分析。采用单因素分析和逻辑回归探索影响疗效的相关因素。在R软件中构建疗效影响因素的列线图预测模型，并根据受试者工作特征曲线和校准曲线进行验证，进一步绘制临床决策曲线和临床影响曲线以确认其临床实用性。3个月时的完全细胞遗传学缓解率差异有统计学意义，伊马替尼组、尼洛替尼组和氟马替尼组分别为53.13%、76.67%和78.38%（P < 0.05）。3个月时的主要分子学缓解（MMR）率分别为25.00%、53.33%和51.35%，12个月时分别达到78.13%、90.00%和83.78%。各治疗组12个月时的深度分子学缓解（DMR）率分别为50.00%、76.67%和75.68%（P < 0.05）。多因素逻辑回归显示早期分子学缓解、白细胞计数、红细胞分布宽度和血小板计数是MMR的独立影响因素。年龄、药物类型、早期分子学缓解和红细胞分布宽度被确定为DMR的独立影响因素（P < 0.05）。MMR和DMR列线图模型的受试者工作特征曲线下面积分别为0.912（95%置信区间：0.833，0.990）和0.874（95%置信区间：0.801，0.946），表明模型校准良好。尼洛替尼和氟马替尼的疗效优于伊马替尼，疗效受社会人口学特征、临床异质性和药物副作用等多种因素影响。所提出的临床预测模型可为决策提供有价值的见解，并具有普遍性和实际应用价值。
