The Natural History of RTQ-PCR Levels After the Achievement of Complete Molecular Remission (CMR): Implications for ‘Stopping' Studies

Abstract 605

Variations in RTQ-PCR estimations of BCR-ABL1 transcript numbers between laboratories have resulted in recognised difficulties in interpreting results and have led to a global effort of harmonisation via an international reporting scale (IS). Currently this is achieved in a limited number of laboratories worldwide by exchange of samples and will hopefully be replaced by the production of internationally accredited reference reagents. Differences in the limits of sensitivity of assays in different laboratories pose particular problems in the definition and interpretation of molecular negativity, so-called complete molecular remission (CMR), leading some investigators to suggest distinctions between assays capable of detecting 4, 4.5 and 5 log reductions in tumour load and introducing the terms CMR4, CMR4.5 and CMR5. These definitions take on particular relevance when designing studies of de-escalation and/or stopping tyrosine kinase inhibitor (TKI) therapy. In the French STIM trial, criteria for stopping were relatively stringent in that patients were required to have at least 5 results of RTQ-PCR negativity in their local laboratory sustained over at least 2 years and confirmed on one further occasion in the centralised laboratory. Negative results of BCR–ABL 1 amplification were reported only if the RNA was of good quality and quantity (50 000 copies of normal ABL1). Subsequently several groups are designing similar studies. As our ability to stop treatment must in large part be determined by the level of residual disease at the time of cessation, it is important to have robust definitions of CMR.

We maintain a comprehensive database of all our CML patients on TKI. For chronic phase this database now contains 521 patients (273 [52%] male) of median age 48 yrs (range 13–86). 212 patients received interferon prior to TKI therapy. The median follow up for surviving patients is 76 mths (range15-137). 88 (37 [42%] male) of these patients have achieved RTQ-PCR molecular negativity on more than one occasion and prompted us to identify the proportion that would satisfy entry criteria for a stopping study and hence the natural history of RTQ-PCR results in such patients. Confirmed complete molecular response (cCMR) was defined as two consecutive samples with no detectable transcripts at least 4 weeks apart with an ABL1 control >40,000 copies (median ABL1 control in the CMR samples was 84,000 copies). 64 patients met our criteria for cCMR, the remaining 24 patients had at least two negative results but never consecutively. 56 patients achieved cCMR on their first line TKI (imatinib in all but 2). Times from diagnosis to MMR and cCMR in this cohort were a median of 24 (range 3–77) and 46 mths (range 5–118) respectively. The median time from MMR to cCMR was 26 mths (range 0–89). Excluding 8 patients in whom follow-up since cCMR is less than 24 months the median duration from cCMR is 53 mths (range 24–113), Only one patient has subsequently lost MMR confirming the excellent prognosis of this cohort. However, only 10 patients (21%) have sustained RTQ-PCR negativity over a 2 year period that would deem them eligible for a STIM-equivalent study. If we were to define a less stringent CMR4.5 as a BCR-ABL ratio of 0.0032 in the international scale the number of eligible patients increases to 18/48 (37.5%). If we applied CMR4.5 to the 24 patients without consecutive RTQ-PCR results a further 3 patients would meet the criteria for a stopping study, total 21/88 (24%).

In conclusion the numbers of patients eligible for stopping studies confined to sustained cCMR is relatively few although we cannot exclude the possibility that some patients were not entirely compliant. Although not proven, reducing the stringency of the definition of CMR is likely to lead to higher relapse rates in subsequent stopping studies than in the original STIM trial. This must be considered when interpreting the results of first-line second generation TKI where the rates of achievement of MMR and CMR may be higher than with imatinib. In these studies CMR may not be synonymous with a 50% chance of discontinuing treatment permanently and future studies might more appropriately consider strategies of de-escalation rather than cessation.