As it is recommended that most assessments for treatment-free remission (TFR) in patients with chronic myeloid leukemia be conducted as prospective trials, we conducted a systematic review and meta-analysis to investigate which study-level factors affected the TFR rate. The MEDLINE, Embase, and Cochrane databases were systematically searched from inception to July 2018. A random effect model was used to estimate the overall mean TFR rate, subgroup differences, and regression coefficients with continuous variables. Overall, 12 tyrosine kinase inhibitor (TKI) stopping studies comprising 1699 chronic myeloid leukemia patients were included in this analysis. The overall mean TFR rate at 24 months after entering TFR phase was 55% [95% confidence interval (CI) 0.51-0.58]. Trials with molecular criteria of MR4.5 or better for stopping TKI reported higher TFR rates than those of MR4.0 (57.2% vs. 50.5%). Trials with eligible criteria for at least 24 months of deep molecular response (DMR) duration demonstrated higher TFR rates than those for 18 or 12 months (60.2% vs. 49.9%). Our results suggest that TKI stopping trials with sufficient duration of DMR and molecular criteria of MR4.5 or better may account for approximately 60% of the TFR rate at 24 months after stopping TKI.

中文翻译：

---

研究水平因素对慢性粒细胞白血病患者无治疗缓解率的影响：系统评价和荟萃分析。

由于建议对大多数慢性粒细胞白血病患者的无治疗缓解（TFR）进行评估，作为一项前瞻性试验，我们进行了系统的综述和荟萃分析，以调查哪些研究水平因素影响了TFR率。从开始到2018年7月，系统地搜索MEDLINE，Embase和Cochrane数据库。使用随机效应模型来估计总体平均TFR率，亚组差异和带有连续变量的回归系数。总体而言，该分析包括12项酪氨酸激酶抑制剂（TKI）终止研究，其中包括1699例慢性髓性白血病患者。进入TFR阶段后24个月的总体平均TFR率为55％[95％置信区间（CI）0.51-0.58]。以MR4的分子标准进行试验。5个或更高的停止TKI报告的TFR率比MR4.0高（57.2％对50.5％）。具有合格标准的至少24个月深分子反应（DMR）持续时间的试验显示，TFR率高于18个月或12个月的TFR率（60.2％比49.9％）。我们的结果表明，在停止TKI后24个月内，DMR持续时间足够长且分子标准为MR4.5或更高的TKI停止试验可能占TFR率的约60％。