Purpose: In the first years of imatinib treatment, BCR-ABL remained detectable in all but a small minority of patients with chronic myeloid leukemia. We determined whether BCR-ABL continues to decline with longer imatinib exposure and the incidence and consequence of undetectable BCR-ABL.
Experimental design: BCR-ABL levels were measured in a subset of 53 imatinib-treated IRIS trial patients for up to 7 years (29 first-line, 24 second-line). Levels were deemed undetectable using strict PCR sensitivity criteria.
Results: By 18 months, the majority achieved a 3-log reduction [major molecular response (MMR)]. BCR-ABL continued to decline but at a slower rate (median time to 4-log reduction and undetectable BCR-ABL of 45 and 66 months for first-line). The probability of undetectable BCR-ABL increased considerably from 36 to 81 months of first-line imatinib {7% [95% confidence interval (95% CI), 0-17%] versus 52% (95% CI, 32-72%)}. Undetectable BCR-ABL was achieved in 18 of 53 patients and none of these 18 lost MMR after a median follow-up of 33 months. Conversely, MMR was lost in 6 of 22 (27%) patients with sustained detectable BCR-ABL and was associated with BCR-ABL mutations in 3 of 6. Loss of MMR was recently defined as suboptimal imatinib response. There was no difference in the probability of achieving molecular responses between first- and second-line patients but first-line had a significantly higher probability of maintaining MMR [P = 0.03; 96% (95% CI, 88-100%) versus 71% (95% CI, 48-93%)].
Conclusions: With prolonged therapy, BCR-ABL continued to decline in most patients and undetectable BCR-ABL was no longer a rare event. Loss of MMR was only observed in patients with sustained detectable BCR-ABL.