Assessment of Molecular Response in CML Patients with Atypical BCR-ABL Tanscripts. Recommendations By the EUTOS Collaboration

Approximately 2‐3% of CML patients (pts) have variant BCR‐ABL fusions transcripts that cannot be monitored by RQ‐PCR using standard methodologies. Most rare variants are accounted for by seven variant fusions (e1a2/3, b2/3a3, e6a2, e8a2, e19a2). Within the European Treatment and Outcome Study (EUTOS) we sought to establish and validate robust RQ-PCR methods for each of these abnormalities. RNA was extracted from leukocytes obtained from 20 mL peripheral blood. Samples were received either locally or by mail and spent between 1 to 3 days in transit. Multiplex reverse transcription polymerase chain reaction (RT-PCR) experiments were performed to identify BCR-ABL1 transcripts (Cross et al., Leukemia 1994). In all cases, gel electrophoresis showed an unusual band. PCR products were then sequenced to characterize BCR-ABL1 rearrangements. Monitoring of residual disease was carried out by RQ-PCR with specific forward and/or reverse primers and probes using a serial dilution of individual BCR-ABL1 or GUSB plasmid DNA calibrators. GUSB transcripts were quantified as internal control and results were expressed as the BCR-ABL1/GUSB ratio with no conversion according to the international scale (IS) as it is recommended for only typical major (M) BCR-ABL1 and cannot be applied to other transcripts in CML. 43 pts from 8 prospective studies and pts outside of studies were investigated. A total of 435 samples (1-72 per patient, median 8) were analyzed. Pts expressed 6 different atypical BCR-ABL1 transcripts (e19a2, n=18; e1a2, n=10; b2a3, n=5; b3a3, n=5; b2a3&b3a3, n=2, e8a2, n=3). At the start of the observation period, ratios BCR-ABL1/GUSB ranged between 0.01% and 179.1% (median 13.75%). Follow-up results were determined in comparison to the starting level and expressed as relative log reduction. 9 pts were in deep molecular remission considering a minimum GUSB transcript level of 20,000/reaction. During the observation period, six pts relapsed with a significant increase of BCR-ABL1/GUSB ratios. We conclude that very few pts are diagnosed with unusual BCR-ABL1 transcripts, but their characterization is important for proper assessment of treatment response in affected patients, and to avoid false negative assessment of residual disease status. Based on the individual PCR strategies, results cannot be expressed with precision on the International Scale and thus the common molecular milestones and triggers for treatment discontinuation are difficult to apply. We suggest that residual disease in these cases should be assessed primarily by considering trends in BCR-ABL/reference gene ratios over time.