Although total duration of tyrosine kinase inhibitor (TKI) therapy and of molecular response at 4 log reduction or deeper (MR4) correlates with treatment‐free remission (TFR) success after TKI discontinuation, the optimal cut‐off values of the duration remain unresolved. Thus, 131 patients were enrolled into the Canadian TKI discontinuation study. The molecular relapse‐free survival (mRFS) was defined from imatinib discontinuation till molecular recurrence, that is, major molecular response (MMR) loss and/or MR4 loss. We evaluated mRFS at 12 months after imatinib discontinuation, analyzed it according to the imatinib treatment duration and MR4 duration, and calculated
Pvalue, positive (PPV) and negative predictive value (NPV) in the yearly cut‐off period of time. The shortest cut‐off was sought that met the joint criteria of a Pvalue ≤ 0·05, PPV ≥ 60% and NPV ≥ 60%. We propose six years as the shortest imatinib duration cut‐off with a Pvalue 0·01, PPV 68% and NPV 62%: The patients treated with imatinib duration ≥ 6 years showed a superior mRFS rate (61·8%) compared to those with less treatment (36·0%). Also, 4·5 years MR4 duration as the shortest cut‐off with a Pvalue 0·003, PPV 63% and NPV 61%: those with MR4 duration ≥ 4·5 years showed a higher mRFS rate (64·2%) than those with a shorter MR4 duration (41·9%).