Long Term Study of the Impact of Quantitative Molecular Monitoring of Bcr-Abl Transcripts on the Risk of Relapse of CML After Allogeneic HSCT.

Monitoring of minimal residual disease through RT-PCR analysis of bcr-abl transcripts allows early detection of CML relapse after allogeneic HSCT. However, the introduction of more sensitive techniques, such as quantitative PCR, may result in decreased specificity, leading to false positive results.

In this study we reviewed the results of molecular analysis of bcr-abl transcripts in all patients with p210+ CML who underwent allogeneic HSCT from 1983 through 2007. Q-PCR analysis was started in 2002. Out of 189 patients, 87 patients had available Q-PCR data; of these, 63 patients with at least three separate Q-PCR data were included in the study. Median time to the 1^st Q-PCR analysis was 2196 days (35-7823). Median age was 36 years (13-56), 62%/38% received transplant from a related/unrelated donor, 62% with BM. 32% were in accelerated phase (AP). Patients with at least one positive Q-PCR value (measured as a ratio of bcr-abl to abl of > 0.001) were classified as Major Molecular Remission (MMR) patients. Each event was defined as one or more consecutive positive results.

60/63 patients are alive after a median follow up of 3693 days (898-9405). 6 have relapsed 2142 (1419-3746) days after transplant. 52 (83%) patients had at least one positive result (28 with a value of >0.01, 24 with a value of <0.01), whereas 11 (17%) had persistent undetectable transcripts. In MMR patients, 94 events occurred. 29 patients had only one event, while 6 had >3 events. In 10 patients, the event occurred within 1 year after transplant, whereas in 28 it occurred after >5 years. 6/52 MMR patients relapsed, as compared to 0/11 with persistent undetectable transcripts (p=0.19). Relapse did not correlate with the Q-PCR value, the number of events or the time to the event. Finally, of 46 MMR patients who did not relapse, 35 had undetectable transcripts at last follow up. Positive Q-PCR had low specificity (19%) and positive predictive value (12%) in predicting relapse after allogeneic HSC transplantation.

Our data confirm that the detection of low levels of bcr-abl transcripts (as based on Q-PCR) has a poor accuracy in predicting relapse, and it should not be considered as the sole indication to start treatment. It appears that fluctuation of bcr-abl transcript levels is common as late as > 10 yrs after transplant, possibly suggesting the long term persistence of CML stem cells.

Rosti:Novartis: Consultancy, Honoraria; BMS: Consultancy, Honoraria. Martinelli:Novartis: Consultancy, Honoraria; BMS: Consultancy, Honoraria; Pfizer: Consultancy.