Validation of the New European LeukemiaNet (ELN) Recommendations for Bcr-Abl Kinase Domain Mutation Analysis In Chronic Myeloid Leukemia: An Analysis of the GIMEMA CML Working Party Studies

BCR-ABL kinase domain (KD) mutation analysis may be an useful tool for physicians and is being performed in a growing number of laboratories. Recommendations aimed to rationalize the use of mutation testing in chronic myeloid leukemia (CML) have recently (Blood 2011) been compiled by a panel of experts appointed by European LeukemiaNet (ELN) – including specific recommendations as to when mutation analysis should be performed. They came from the expert opinion of the panel members whenever published data were insufficient or contradictory. In order to provide further data to validate or refine these recommendations, we have analyzed the GIMEMA CML WP database recording the results of mutation analyses performed in CML pts (n=1301) receiving imatinib and/or 2nd generation TKIs between January 2004 and July 2011. At dagnosis, mutation analysis was recognized to be useful in the few pts who present in accelerated phase or blast crisis (BC), while it was not recommended in chronic phase (CP) pts. Interrogating our database, we could retrieve 58 mutation analyses in newly diagnosed pts in CP and 12 in newly diagnosed pts in BC. Imatinib-resistant mutations were detected in 0 and 2 pts, respectively. In pts receiving 1st-line imatinib, mutation analysis was recommended both in case of failure and in case of suboptimal response. We have analyzed 399 chronic phase (CP) CML pts receiving first-line imatinib because they were found to meet one of the criteria for failure or suboptimal response. Overall, 45/166 (27.1%) failures were found to be positive for one or more BCR-ABL KD mutations. In particular, mutations were detected in 3/16 (18.8%) pts with less than CHR at 3 months, 1/9 (11.1%) pts with no CyR at 6 months, 4/24 (16.7%) pts with less than PCyR at 12 months, 6/36 (16.7%) pts with less than CCyR at 18 months, 15/49 (30.6%) pts who lost CCyR and 16/32 (50%) pts who lost CHR. More interestingly, only 11/233 (4.7%) suboptimal responders we analyzed were positive for mutations. Among ‘cytogenetic' suboptimal responders, mutations were detected in 1/15 (6.7%) pts with no CyR at 3 months, 1/20 (5.0%) pts with less than PCyR at 6 months, 5/51 (8.2%) pts with less than CCyR at 12 months. Among ‘molecular' suboptimal responders, mutations were detected in 0/52 pts with less than MMR (but having achieved CCyR) at 18 months and in 4/95 (4.2%) pts who lost MMR (but not CCyR). Which rise in Bcr-Abl transcript level should trigger a mutation analysis was the most difficult issue to provide recommendations upon, given the lack of convincing and reproducible data in the literature. It was finally agreed to recommend mutation analysis only in case of MMR loss. In 159 of the CP pts we have analyzed, mutation analysis was specifically requested because of a transcript increase at a single RQ-PCR assessment: 29 pts had less than 1-log increase and 41 pts had a 1-log increase or more – but with no loss of MMR. None of these pts was found to have mutations. Another 36 pts had less than 1-log increase and 53 had a 1-log increase or more, leading to loss of MMR. Mutations were detected in 1 (2.8%) and 3 (5.7%) pts, respectively. In pts receiving dasatinib or nilotinib as 2nd-line agents, mutation analysis was recommended at baseline and then in case of failure according to the provisional definitions proposed by Baccarani et al (J Clin Oncol 2009). Nineteen among the pts we analyzed met these criteria; overall, mutations were detected in 11 (57.8%), including 5/7 pts with no CyR at 3 months, 6/9 pts with minimal CyR at 6 months, 1/4 pts with less than PCyR at 12 months. In addition, newly acquired mutations were detected in 93/131 (71%) pts who lost a previously achieved HR or CyR. We also tested 19 pts who met the provisional definitions for suboptimal response to dasatinib or nilotinib 2nd-line. Mutations were detected in 4/19 pts (21%), including 2/5 pts with minor CyR at 3 months, 1/7 pts with PCyR at 6 months, 1/7 pts with less than MMR at 12 months. Our data indicate that: a) pts harbouring mutations can more frequently be found among cytogenetic suboptimal responders than among molecular suboptimal responders; b) any Bcr-Abl transcript increase that is not associated with MMR loss shouldn't indeed trigger a mutation analysis; c) although definitions of response to dasatinib or nilotinib 2nd-line are still provisional and might soon be refined, not only failures but also suboptimal responses are frequently associated with mutations.