CML: how low can you go?

The use of imatinib mesylate as a specific inhibitory treatment of the BCR-ABL tyrosine kinase has revolutionized the treatment of CML. Although results of treatment in patients in accelerated phase or blast crisis remain suboptimal, the rate of progression to advanced phase from chronic phase has dramatically been reduced such that approximately 90% of people treated with imatinib at the time of diagnosis remain alive with follow-up that now approaches 8+ years. The results in chronic phase are not perfect, however, and intent-to-treat analyses of large cohorts of chronic-phase patients suggest that approximately 60%-65% of patients remain on imatinib therapy 5 years after initiation of treatment.¹ 

Monitoring of response is a critical feature of the management of chronic myelogenous leukemia (CML) and international consensus guidelines have been developed suggesting therapeutic end points that should be achieved after specific durations of therapy.²  It is generally accepted that at a minimum, patients should attain a complete cytogenetic response (CCyR) because there is almost no disease progression in patients in CCyR after 2-3 years of treatment. In this issue of Blood, Hughes and colleagues report on a subgroup of patients from the IRIS trial and emphasize the clinical value of achieving a greater and more rapid reduction in tumor burden termed “major molecular response” (MMR) as quantified by quantitative polymerase chain reaction (Q-PCR).³  MMR is defined as a 3-log reduction in BCR-ABL transcript number compared with a pretreatment baseline. Importantly, the “starting point” is not the value for an individual patient (which is usually not available), but rather represents the average results of initial transcript numbers as assessed in 3 research laboratories of approximately 40 patients entered on the IRIS trial and treated with imatinib.

Approximately 40% of the subgroup of patients entered on the IRIS trial who had serial molecular analyses achieved this end point by 12 months and all remained without disease progression with an event-free survival of approximately 90% with long-term follow-up (see figure).³  Certainly, this is reassuring for this group of patients and therefore clinically helpful, but, only if one can measure MMR reliably. Q-PCR assays can be fickle and as documented by Branford et al, values can vary widely among different molecular laboratories.⁴  These same authors demonstrated that with exchange of clinical material, results could become relatively standardized on what is termed the “International Scale” (IS; in which 100% is the baseline value, with < 0.1% as an MMR and < 1% roughly corresponding to CCyR). However, they also emphasize the need for continued recalibration to maintain consistency of results.

And, the molecular assays are not perfect, even in the hands of experts. Hughes et al mention 6 patients apparently in MMR by molecular analyses who had the Philadelphia chromosome detected by cytogenetic assays done simultaneously. Indeed, assessment of MMR is not available in most molecular laboratories in the United States because they have not calculated the correction factor needed to correlate their results with the IS. It remains unclear how often the absence of this information adversely affects clinical management of individual patients.⁵ 

Much of this uncertainty could be alleviated by the development of internal BCR-ABL “standards” which could be used to create a reference curve to generate the Q-PCR results. Efforts to create such a standard reagent from K562 cells are in progress.⁶  A minority of patients become completely Q-PCR “negative” and a recent study, with short follow-up, has suggested that perhaps a third of patients who are repeatedly PCR negative using sensitive assays may not have disease recurrence after imatinib is discontinued.⁷  Thus, further standardization of the assay at the lowest end of detection would also be important.

So, how should these sensitive molecular assays be used clinically? I have noted that some clinicians now rely on serial Q-PCR exclusively and fail to confirm CCyR by marrow cytogenetics. This approach can be used only if one has some certainty about the numerical Q-PCR level which corresponds to a CCyR; I have asked this question of clinicians at multiple meetings and virtually no one is familiar with that number in their laboratory. Therefore, it is recommended that a CCyR be confirmed by bone marrow cytogenetics before relying exclusively on molecular monitoring. In my practice, I follow patients serially with peripheral blood FISH (fluorescence in situ hybridization) analyses, doing a marrow when the FISH becomes negative. Thereafter, patients are followed every 3-4 months with peripheral blood Q-PCR.

Values frequently fluctuate and it is critical to recognize that modifications of therapy should not be made on the basis of a change in a single Q-PCR assessment. Rather, the value should be repeated and if persistently elevated, bone marrow evaluation and assessment for BCR-ABL kinase mutations should be considered. In addition, it is critical to be certain that the patient continues to be compliant with the medication.⁸  Some patients skip doses or occasionally stop taking the imatinib due to persistent low-grade side effects; it is likely that variations in adherence may increase in the future due to higher copays for this expensive medication in the United States. Indeed, given the extremely low relapse rate with continued therapy, many clinicians feel that a major reason for follow-up every 4-6 months is to reinforce the need for continued compliance.

Lastly, recently completed trials comparing imatinib with nilotinib or dasatinib have demonstrated improved results, at least in the short-term, with the second-generation tyrosine kinase inhibitors.^9,10  As these newer agents are used more routinely as first-line therapy, long-term follow-up of molecular outcomes analogous to the study by Hughes et al will be needed because the kinetics of response are more rapid with the newer agents. Thus, we can expect an amplification of the number of articles documenting (and debating) the clinical merits of molecular quantification in the future.