DEFINITIONS AND STANDARDISATION OF ‘COMPLETE’ MOLECULAR RESPONSE IN CHRONIC MYELOID LEUKEMIA

Sborník

The international, collaborative effort to standardize BCR-ABL qRT-PCR testing for CML has focused largely on determining whether a patient has or has not achieved MMR (≤0.1% BCR-ABL on the International Scale; equivalent to ≥3-log reduction in BCR-ABL transcript levels from the IRIS standard baseline). Many patients on imatinib achieve MMR, but only a minority progress to what has been termed complete molecular response (CMR), defined by the European LeukemiaNet (ELN) as undetectable BCR-ABL transcripts by qRT-PCR and/or nested PCR in 2 consecutive samples with a sensitivity >104. This definition, however, is difficult to implement in a standardised fashion across multiple centres and does not adequately take into account variations in assay sensitivity within and between centres. Improved definition of molecular milestones is a pressing issue since second-generation TKIs result in deeper molecular responses compared to imatinib. Furthermore, there is considerable interest in conducting studies (e.g. EURO-SKI) to assess the possibility of stopping TKI therapy once significant and sustained molecular responses are achieved. There is a general consensus that it is not possible to have a single workable definition of CMR, but rather the level of response needs to be defined by an upper boundary. So, just as MMR corresponds to ≤0.1% BCR-ABLIS, the terms CMR4, CMR4.5 and CMR5 have started to be used to indicate levels of disease ≤0.01% BCR-ABLIS (4-log reduction from IRIS baseline), ≤0.0032% BCR-ABLIS (4.5-log reduction) and ≤0.001% BCR-ABLIS (5-log reduction), respectively. However there are two immediate problems. The first is semantic: the fact that cut offs are defined by an upper boundary means that disease may still be detectable at a lower level, which does not fit well with the term ‘complete’. We therefore propose that the terms for low level disease are modified to ‘molecular response’, i.e. MR4, MR4.5 etc. The second problem concerns laboratory standardisation: how is MR4, MR4.5 etc. actually determined in the testing laboratory and how comparable are results across different laboratories? ENEST1st is a phase IIIb, open-label study of nilotinib in adult patients with newly diagnosed CML (ClinicalTrials.gov NCT01061177). The primary study aim is to establish the rate of MR4 at 18 months and to work with EUTOS (European Treatment and Outcome Study) laboratories to improve the sensitivity and standardisation of qRT-PCR for low level BCR-ABL detection. Preliminary analysis of data from the 12 ENEST1st molecular monitoring laboratories shows substantial variation in scoring of low level disease results caused principally by technical differences but exacerbated by differences in laboratory definitions. Since conversion factors are of questionable value when disease is undetectable, we suggest the following working criteria should be used to define molecular response: MR4 = either (i) detectable disease ≤0.01% BCR-ABLIS or (ii) undetectable disease in cDNA with ≥10,000 ABL or ≥24,000 GUSB transcripts MR4.5 = either (i) detectable disease ≤0.0032% BCR-ABLIS or (ii) undetectable disease in cDNA with ≥32,000 ABL or ≥77,000 GUSB transcripts . Details of these definitions, their implementation in the testing laboratory and what further work needs to be performed will be discussed.

Haematologica, 2012; 97(s1):  308