Molecular Long-Term Surveillance of CML Patients on Imatinib Therapy. Follow-Up of German Patients Treated within the IRIS Trial.

High rates of complete cytogenetic response (CCR), the availability of sensitive methods to detect residual disease, and direct therapeutic consequences are leading motives to integrate regular molecular monitoring into the standards for the management of patients (pts) with chronic myeloid leukemia (CML). We sought to determine long-term dynamics of BCR-ABL mRNA expression levels in 132 CML pts (75 m, 57 f, median age 51, range 20–71 yrs) recruited into the IRIS study in 17 German centers. Pts were randomized to receive imatinib (n=69) or interferon α+Ara-C (IFN, n=63). Due to intolerance or lack of response 41 pts crossed over from IFN to imatinib. Response to therapy was sequentially monitored by conventional cytogenetics from bone marrow metaphases (n=806). BCR-ABL transcripts were determined in 1414 peripheral blood samples by quantitative real time RT-PCR (RQ-PCR) using the LightCycler technology. In case of low level (<10 transcripts/2μl cDNA) or neg RQ-PCR, nested PCR was performed. Total ABL transcripts were quantified as internal controls. A single series of BCR-ABL plasmid dilutions served as standard for both BCR-ABL and ABL transcripts. In pts on 1st-line imatinib therapy median ratios BCR-ABL/ABL gradually decreased: 4.8% at mo 3, 0.88% at mo 6, 0.22% at mo 12, 0.17% at mo 18, 0.058% at mo 24, 0.066% at mo 30, and 0.023% at mo 36. After crossover to imatinib results were not significantly different: 15.5% at mo 3, 1.6% at mo 6, 0.28% at mo 12, 0.068% at mo 18, 0.045% at mo 24, and 0.041% at mo 30. After a median follow-up of 40 mo (1–47) 31/69 pts (45%) on 1st-line imatinib were still RQ-PCR pos, 20 pts (29%) were RQ-PCR neg and nested PCR pos, and in 4 pts (5.8%) BCR-ABL became undetectable by RQ- and nested PCR. After a median time of 25 mo (3–43) on 2nd-line imatinib therapy 19/41 pts (46%) were RQ-PCR pos, 9 pts (22%) were RQ-PCR neg and nested PCR pos, and in 5 pts (12%) BCR-ABL was undetectable by RQ- and nested PCR. Considering adequate RNA quality BCR-ABL became repeatedly undetectable in 4 pts after 18–33 mo of 1st-line imatinib therapy and in 5 pts 9–33 mo after crossover from IFN to imatinib. In one patient, BCR-ABL remained undetectable after a treatment free interval of 4 weeks. After achieving CCR, 5 pts (7.2%) on 1st-line and 2 pts (4.9%) on 2nd-line imatinib therapy experienced cytogenetic relapse after a median time of 10 mo (4–21). In none of these pts mutations of the tyrosine kinase domain of BCR-ABL were detected. BCR-ABL/ABL ratios after 12 mo of imatinib therapy were significantly lower in pts in continuous CCR vs pts with subsequent relapse (0.18 vs 0.60%, respectively, p=0.04). None of the relapsing patients had achieved a ratio BCR-ABL/ABL <0.12% after 12 mo, which represents a 3-log reduction from baseline. During total follow-up ratios BCR-ABL/ABL <0.12% have been achieved in 51 pts (74%) on 1st-line and in 21 pts (51%) on 2nd-line imatinib therapy. We conclude that (i) treatment with imatinib in newly diagnosed CML pts is associated with a rapid and steady decrease of BCR-ABL transcript levels, (ii) a short trial of IFN does not jeopardize molecular response to subsequent imatinib therapy, (iii) an increasing minority of pts achieve complete molecular remission, and (iv) ratios of BCR-ABL/ABL <0.12% after 12 mo of therapy predict for long-term response.