How Cured Are CML Patients in Complete Molecular Remission (CMR) after Stem Cell Transplantation or Imatinib?.

It is unclear if CMR, i.e., absence of BCR-ABL mRNA, is synonymous with, or even required for, cure of chronic myeloid leukemia (CML). This is particularly relevant for management of the minority of patients who achieve CMR with imatinib (IM). Although most patients in long-term remission (LTR) post-stem cell transplantation (SCT) are considered “functionally cured”, BCR-ABL mRNA is occasionally detected in their peripheral blood (PB), at a level similar to that detectable in the PB of healthy individuals. Most CML patients in CMR on IM relapse shortly after treatment discontinuation. We sought to elucidate the quality of the molecular response in these two groups of CML patients by using a genomic DNA (gDNA) real-time quantitative PCR (RQ-PCR) with patient-specific primers/probe combinations for detection of BCR-ABL genomic fusions (gBCR-ABL). gBCR-ABL - a molecular signature of each CML case - was sequenced from pre-SCT or pre-/early-IM therapy using inverse PCR (I-PCR) or long-range genomic PCR (LR-PCR). The I-PCR involved digestion of gDNA with RsaI, circularization of the fragments, and amplification with 2 sets of inverse primers located in the 5′ end of the RsaI-fragments of the major breakpoint region of BCR for cloning and sequencing of the BCR-ABL band. The LR-PCR is a multiplex reaction with forward primers on BCR exons 13 or 14, and 20 reverse primers, spanning ∼150kb of the ABL breakpoint region. The patient-specific products are then directly sequenced. Knowledge of the sequence allowed us to design patient-specific primers/probe combinations that were then used to test gDNA from PB follow-up (FU) samples using novel single-step or nested RQ-PCR assays. When tested in serial dilutions of the sample from which the breakpoint sequence was obtained, both methods generated standard curves of similarly good quality; the nested approach did not improve the sensitivity of the assay, with both methods being capable of detecting one single target DNA molecule per reaction. The specifity of the assay was demonstrated using at least 2 different BCR-ABL-positive gDNAs and a no-gDNA negative controls, whereas the sensitivity was maximized by testing a minimum of 7.2μg gDNA in multiple reactions. From 6 patients in LTR post-SCT (median time post-SCT 186 months; range: 87 to 333) we tested 9 FU samples collected between 87 and 321 months post-SCT (median: 168). From 3 IM-treated patients we tested 5 samples in CMR collected between 36 and 75 months (median: 63) after the start of IM. Six of the 9 post-SCT samples had been classified as low-level positive for BCR-ABL transcripts (BCR-ABL/ABL 0.001 to 0.012%): only 1 was positive for gBCR-ABL (BCR-ABL/ABL from this sample - 0.003%). Of the 3 patients in CMR on IM, 1 had 1 sample negative for gBCR-ABL; 1 had 1 sample positive and 1 sample negative 37 months later; 1 had 2 positive samples separated by 36 months. The FU samples positive for gBCR-ABL were positive at a very low level, with only 1 or 2 positive reactions out of a minimum of 24 replicates, with Ct values very close to the threshold of detection of the standard curves. In conclusion, the results so far suggest that, in post-SCT patients in LTR, the original BCR-ABL positive clone is rarely detected, and in most instances may not be the cause of low-level positivity for BCR-ABL mRNA. The leukemic clone may be more frequently present in IM-treated patients in CMR, which suggests the need for continuing IM even after achievement of CMR.