There are different perspectives on what constitutes an inadequate response to treatment in CML, and it largely depends on patient-specific goals of therapy. Prior to the availability of TKIs, our ability to improve survival with medical therapies was modest, as was the case with interferon. But with the availability of TKIs came the expectation of a normal life expectancy for most patients with CML, and there is now a desire to raise the bar even higher, particularly for younger patients.

A level of response of 0.01% or deeper that is maintained consistently over a period of 2 or more years, in conjunction with 3 or more years of TKI therapy duration, are needed before patients with chronic phase CML can try discontinuing therapy. In terms of the minimum level of molecular response that predicts for superior survival compared with lesser levels of response, numerous studies have shown that the equivalent of a complete cytogenetic response is associated with a high level of disease-specific survival. In my practice, I do not put much weight on the achievement of a major molecular response because it does not provide any added clinical benefit. I think that anything above a 1% level after 12 months of therapy would likely be considered inadequate, but it is important to view molecular response as a continuous variable (ie, if a patient was at 0.9% or 1.1%, for example, they would have an indistinguishable likelihood of doing well).

If a patient has no appreciable molecular response, they are at increased risk for progressive disease. Just for perspective, however, we estimate that approximately 70% to 75% of these patients will continue to do well on TKI therapy. This is admittedly far less than the 95% of those who achieve a 1% level, but the majority still do not succumb to their disease. Further, although a minority of patients with chronic phase CML develop loss of response, there is a substantial probability of detecting a resistance-conferring mutation in this clinical context. Loss of molecular response would generally necessitate at least a 1-log increase that is not explained by nonadherence or a coadministered medication that may diminish TKI exposure.

When a change in therapy is warranted, some patients who were diagnosed at the beginning of this century are extremely reluctant to consider switching to an alternate therapy because they remember the emotional impact of the diagnosis of leukemia and how imatinib put all of that at ease. In these situations, I tell my patients that, from the United States health care system/payer perspective, they can always go back to imatinib if an alternate therapy does not improve their quality of life or fails to provide a meaningful improvement in their response.