目的 探讨NK细胞及Treg细胞在慢性粒细胞白血病（chronic myeloid leukemia， CML）应用酪氨酸激酶抑制剂（tyrosine kinase inhibitor， TKI）停药患者中的变化及二者对TKI停药的预测价值。 方法 随机选取确诊为CML的患者200例，分为停药组（n = 100）和未停药组（n = 100），停药组患者进一步分为复发组（n = 41）和未复发组（n = 59），对患者的临床资料及随访资料进行回顾性分析。Logistic回归分析各变量对停药后结局的影响，探讨患者停药后复发的独立影响因素。采用受试者工作特征（receiver operating characteristic， ROC）曲线分析NK细胞及Treg对TKI停药后结局的预测价值。 结果 停药组中应用干扰素治疗的患者比例显著高于未停药组（P < 0.05），而前者NK细胞（P < 0.05）及Treg细胞（P < 0.01）的数量显著高于后者。与复发组患者相比，未复发组患者中应用干扰素的比例显著高于前者（P < 0.05），TKI治疗时间及深度分子学反应（deep molecular response， DMR）持续时间均显著大于复发组患者（P < 0.05）。未复发组的NK细胞及Treg细胞的数量显著高于复发组（P < 0.01）。Logistic回归分析发现，是否应用干扰素（OR= 1.25， 95%CI： 1.11 ~ 2.03， P < 0.001）、DMR持续时间（OR = 1.16， 95%CI： 1.08 ~ 1.92， P < 0.05）、NK细胞（OR = 1.64， 95%CI： 1.14 ~ 2.28， P < 0.01）、Treg细胞（OR = 1.83， 95%CI： 1.15 ~ 2.42， P < 0.01）均是停药后患者是否复发的关键影响因素。ROC曲线分析结果表明，NK细胞联合Treg细胞预测TKI停药后是否复发的AUC = 0.892（95%CI： 0.857 ~ 0.927， P < 0.001）。 结论 NK细胞和Treg细胞在TKI停药未复发的患者中显著增加，二者联合对于CML患者停药是否复发具有良好的预测作用。

Objective To investigate the dynamics of NK cells and Treg cells， as well as their potential prognostic significance in relation to TKI discontinuation among patients diagnosed with chronic myeloid leukemia （CML）. Methods In this study， a total of 200 patients diagnosed with CML were randomly selected and divided into two groups： the discontinuation group （n = 100） and the non-discontinuation group （n = 100）. Within the discontinuation group， patients were further categorized into a recurrence subgroup （n = 41） and a non-recurrence subgroup （n = 59）. Clinical data and follow-up information of these patients were retrospectively analyzed. Logistic regression analysis was performed to investigate the impact of various variables on patient outcomes following drug discontinuation， as well as to explore independent factors influencing recurrence in these individuals. Receiver operating characteristic （ROC） curve analysis was employed to assess the predictive value of NK cells and Treg cells for TKI discontinuation outcomes. A significance level of P < 0.05 was considered statistically significant. Results The proportion of patients treated with interferon in the discontinuation group was significantly higher than that in the non-discontinuation group （P < 0.05）. Moreover， the former group exhibited a significantly higher number of NK cells （P < 0.05） and Treg cells （P < 0.01） compared to the latter group. Compared to the recurrence group， there was a significant increase in the proportion of patients using interferon in the non-recurrence group （P < 0.05）， along with longer durations of TKI treatment and deep molecular response （DMR） duration （P < 0.05）. The number of NK cells and Treg cells in the non-recurrence group was significantly higher than that in the recurrence group （P < 0.01）. Logistic regression analysis found that the use of interferon （OR = 1.25， 95% CI： 1.11 ~ 2.03， P < 0.001）， duration of DMR （OR = 1.16， 95% CI： 1.08 ~ 1.92， P < 0.05）， NK cells （OR = 1.64， 95% CI： 1.14 ~ 2.28， P < 0.01）， and Treg cells （OR = 1.83， 95% CI： 1.15 ~ 2.42， P < 0.01） were all influencing factors for the recurrence of patients after drug discontinuation. The results of ROC curve analysis showed that the AUC of NK cells combined with Treg cells for predicting the recurrence of TKI after discontinuation was 0.892 （95% CI： 0.857 ~ 0.927， P < 0.001）. Conclusion The frequencies of NK cells and Treg cells were significantly elevated in patients who remained recurrence-free following TKI discontinuation， highlighting the potential predictive value of combined NK cell and Treg cell analysis for drug cessation in CML patients.