Molecular monitoring of BCR-ABL transcripts after allogeneic stem cell transplantation for chronic myeloid leukemia

The monitoring of minimal residual disease (MRD) through low sensitivity real-time (RT) polymerase chain reaction (PCR) analysis of BCR-ABL transcripts allows early detection of chronic myeloid leukemia (CML) relapse after allogeneic hematopoietic stem cell transplantation (HSCT). The introduction of more sensitive techniques, such as RT quantitative (Q)-PCR, may lead to an overestimation of the risk of CML relapse. In this study, we reviewed the results of peripheral blood RT Q-PCR in CML patients who underwent allogeneic HSCT from 1983 to 2007. In our laboratory, RT Q-PCR analysis was routinely performed since 2002. Eighty-seven of 189 patients had available RT Q-PCR data; 63 patients had at least 3 RT Q-PCR analyses assessable. Fifty-two of 63 patients (83%) had, at least once, detectable transcript levels, but with an BCR-ABL/ABL ratio <.1% defined as <MR3 (molecular remission <0,1%), whereas 11 (17%) had persistent undetectable BCR-ABL transcripts. Six of 52 patients with <MR3 relapsed, defined as BCR-ABL transcript numbers >.1% confirmed by the finding of Ph+ cells in bone marrow. No patients with persistent undetectable transcripts relapsed (P = .19). Relapse did not correlate with the number of occurrences of <MR3 or with the time to the first <MR3 result. Finally, of 46 patients with detectable transcripts who did not relapse, 35 had undetectable transcripts at last contact. RT Q-PCR analysis had low specificity (19%) and low positive predictive value (12%) in predicting relapse of CML patients after allogeneic HSCT. Our data suggest that detection of low BCR-ABL transcript levels by RT Q-PCR analysis has a poor accuracy in defining the risk of CML relapse and should not be considered as the unique indication to treatment. Fluctuation of BCR-ABL transcripts levels is common as late as ≥10 years posttransplantation, possibly suggesting the long-term persistence of CML stem cells.