Relative RATIO Of BCR-ABL1 Transcript Clearance Improves The Prognostic VALUE Of Single PCR Assessment At The 3 and 6 Month Time Points In Patients With Chronic Myeloid Leukemia In Chronic PHASE Treated In The Dasision Study

A cut-off value of 10% BCR-ABL1^IS has been retrospectively validated by several groups as predictive of favorable long-term outcomes. The NCCN guidelines establish that transcripts >10% represent failure to TKI therapy that warrants a change in therapy. It remains controversial whether the predictive value of combining PCR values at 3 and 6 mos improves that of the use of a single PCR value at 3 mos. We aimed to improve the accuracy of early time point PCR prognostication by using the fractional clearance of BCR-ABL1 transcripts between baseline and 3 mos and between 3 and 6 mos in pts with CML in chronic phase (CML-CP) treated in the DASISION study.

Pts with CML-CP were randomized to receive dasatinib (DAS) 100 mg QD (n=259) or imatinib (IM) 400 mg QD (n=260). Methods have been previously reported (NEJM 2010 362 2260). PFS rates were obtained from Kaplan-Meier estimates. Qualifying events for PFS included: increasing WBCs, loss of complete hematologic response or major cytogenetic response, transformation to AP/BP, or death. Outcomes of pts reaching similar PCR milestones were similar regardless of therapy and they were merged for global outcome analyses.

We first evaluated the prognostic value of combining PCR values at 3 and 6 mos in the DASISION study. Overall, 473 pts (234 IM, 239 DAS) remained progression-free at 3 mos and 464 (233 IM, 231 DAS) at 6 mos with PCR assessments available. We used as a cut-off transcript levels determined by ROC analysis that optimally predicted for PFS at 3 months (lower or higher than 10.04%) and 6 months (lower or higher than 2.34%).

(3-year PFS for pts receiving IM or DAS was 87.0% vs 87.9%, respectively. A total of 103 (45.6%) pts on IM and 150 (65.8%) on DAS had low transcripts both at 3 and 6 mos (LO/LO) and their 3-year PFS was 98.0% vs 92.5%, respectively (94.7% for both cohorts combined). Seventy-seven (34.1%) IM treated pts and 33 (14.5%) DAS treated pts had high transcript levels (HI/HI) at 3 and 6 mos and their 3-year PFS was 72.0% (p<0.0001) for IM, 75.9% (p<0.127) for DAS (73.2%, p<0.0001 for combined dataset). Seventy-five pts (40 (17.7%) IM, 35 (15.3%) DAS) pts were classified as LO/HI and had PFS of 89.7% (IM), 88.1% (DAS) and 88.9% (merged dataset). Only 16 (6 IM, 10 DAS) pts were HI/LO but none experienced progression (similar to LO/LO pts). Thus, combining PCR values at 3 and 6 mos added little prognostic discrimination to the single 3-mo PCR assessment.

We then examined the prognostic value of combining PCR values at 0 and 3 mos (i.e. rate of transcript elimination during the first 3 mos of therapy, Table 1, Figure 1). The relative ratio at 3 mos (RR3) was defined as: (PCR at 0 mos-PCR at 3 mos/PCR at 0 mos), while RR6 was: (PCR at 3 mos-PCR at 6 mos/PCR at 3 mos). ROC analysis established cut-off values of 64.5% at 3 mos and 77.5% at 6 mos. Of note, at 3 mos, BCR-ABL1<10.04% and RR3<64.5% identified 14 pts with 3-year PFS of only 79% (vs 90% for those with RR3 >64.5%) in spite of having achieved an optimal response by standard criteria (i.e. PCR<10%). Similarly, at 6 mos, BCR-ABL1<2.34% and RR6<77.5% identified a group of pts with PFS of 87% (vs 97% for those with RR6>77.5%).

Figure 1

View largeDownload slide

KM plots according to RR and PCR at 3 months

Figure 1

View largeDownload slide

KM plots according to RR and PCR at 3 months

Close modal

The combined use of cut-off PCR values at 3 and 6 mos do not significantly improve the prognostic value of a single 3 mo PCR assessment. However, the combination of these cut-offs with the RR of BCR-ABL1 transcript decrease between 0 and 3 mos and between 3 and 6 mos identifies pts with unfavorable outcomes among those with optimal PCR values at 3 and 6 mos, and pts with favorable outcomes among those with suboptimal PCR values at 3 and 6 mos. Validation of these data using an independent dataset of pts treated with imatinib at our institution is underway.

Quintás-Cardama:Novartis, BMS, Ariad: Consultancy. Cortes:Ariad: Consultancy, Honoraria, Research Funding; Bristol-Myers Squibb: Honoraria, Research Funding; Novartis: Honoraria, Research Funding; Pfizer: Consultancy, Honoraria, Research Funding; Teva: Consultancy, Honoraria, Research Funding.