Remission in CML: is DNA useful?

Imatinib is the current standard of care for patients with CML; however, the ability of imatinib to eradicate the CML clone is uncertain. Indeed, according to in vitro studies, leukemic stem cells are seemingly resistant to imatinib-induced apoptosis.¹  Thus, imatinib treatment would not appear to be curative. Most patients in long-term remission after allogeneic stem cell transplantation would be considered “cured,” in spite of very occasionally being found positive for low-level BCR-ABL mRNA detected by RT-PCR.² 

In this issue of Blood, Sobrinho-Simões and colleagues describe a patient-specific quantitative polymerase chain reaction (PCR) strategy using genomic DNA designed to test minimal residual disease in chronic myeloid leukemia (CML) patients in complete molecular response (CMR) after either allogeneic stem cell transplantation (SCT) or imatinib.³  This shrewd strategy needs the patient-specific genomic BCR-ABL fusion sequence to be characterized. Indeed, in CML the genomic breakpoints in the BCR and ABL gene are dispersed over intervals of 3.0 kb and 150 kb, respectively. The variability on the genomic DNA located in the introns is very important. Using this technique, each leukemic fusion sequence is therefore unique, unlike the mRNA transcript in which the variable intronic sequences containing the breakpoints are spliced out leading to 2 types of transcripts. Generally, to evaluate and monitor the response after treatment, the “BCR-ABL load ”—that is, the amount of BCR-ABL mRNA in the peripheral blood—is measured as a ratio of BCR-ABL to a control gene. To perform this analysis, the quantitative reverse transcriptase–PCR technique is routinely used and major molecular response is defined as a 3-log reduction in BCR-ABL mRNA. The CMR with no detectable residual disease corresponds to 4- to 5-log reduction of BCR-ABL due to the limit of the reverse transcriptase–PCR sensitivity compared with diagnosis.⁴ 

BCR-ABL DNA is unique and may be detected by PCR after sequencing and so provides a patient-specific marker of minimum residual disease. Although this technique cannot be used as a routine test (appropriate primers and probes need to be designed), the usefulness of this approach is nicely illustrated by Sobrinho-Simões and colleagues. Hence, it was applied to trace the persistence of the original leukemic clone in 12 patients who had received allogeneic SCT (9 in long-term molecular remission) and 5 imatinib-treated patients in confirmed CMR. The results show that: (1) in patients on imatinib, the leukemic clone is detectable after achievement of CMR (5/5) although it may disappear with continued therapy (4/5) and (2) in patients in long-term remission after SCT, the original leukemic clone almost invariably seems to disappear although Bcr-Abl–transcript positivity may occasionally occur.

Regarding imatinib treatment, even in patients who are in CMR (ie, with undetectable Bcr-Abl transcripts) there is likely to be a residual population of CML cells which could be detected by specific PCR on DNA. It has been reported previously that Bcr-Abl transcript numbers continue to decrease for some years after initiating treatment with imatinib, and that increasing numbers of patients achieve transcript-undetectable status.⁵  This report goes further and demonstrates that continuing imatinib therapy after achieving CMR leads to further reduction in residual disease. It has also been reported that discontinuation may be possible for patient CMR maintained strictly for more than 2 years.⁶ 

In a pilot study, 50% of patients previously treated by interferon were in sustained CMR after discontinuation of imatinib. In a multicenter study entitled “Stop Imatinib” (STIM), these results were confirmed: CMR persisted after imatinib discontinuation in approximately 40% of patients only treated with imatinib.⁷  In these 2 studies, sustained CMR was defined as BCR-ABL/ABL levels below a detection threshold corresponding to at least a 5-log reduction and undetectable signal using real-time quantitative PCR for at least 2 years. By specific PCR on DNA it would be possible to define more strict criteria and increase the proportion of patients in sustained CMR after discontinuation. Concerning allogeneic SCT, this work confirms that the rare late relapse corresponds to the same disease. In only 1 of 9 patients in long-term remission after SCT was there evidence for the survival of the original leukemic clone, which is consistent with the low risk of relapse observed many years after SCT.

Finally, even with a single molecular abnormality such as BCR-ABL, CML is a heterogeneous disease and we still do not know whether the variability of the DNA breakpoint may influence the disease, although not demonstrated using RNA. The complexity of patient-specific DNA PCR may limit the use of this method for the routine monitoring of CML, but taking into account both progress and rapidity of DNA sequencing, this kind of investigation would be able to provide useful information about residual disease in selected patients.