PB1930 CORRELATION BETWEEN LEVELS OF EARLY MOLECULAR RESPONSE AND BCR‐ABL TRANSCRIPT TYPE IN STABLE DMR ATTAINMENT IN CML PATIENTS TREATED WITH IMATINIB

Background: Stable deep molecular response (sDMR), is considered a prerequisite for treatment discontinuation in chronic myeloid leukemia (CML) patients treated with tyrosine kinase inhibitors (TKIs). Depth and kinetics of molecular response (MR) during the first year of TKI treatment correlate with the probability of sDMR. However, no data concerning the impact of transcript types across different levels of early MR on the achievement of a subsequent sDMR have been reported. Aims: We sought to evaluate the probability of sDMR according to BCR‐ABL transcript type (b2a2 vs b3a2) and depth of MR at 3 and 6 months of frontline therapy with imatinib standard dose in a real‐life population of chronic phase (CP) CML patients. Methods: We retrospectively analyzed our cohort of CP‐CML patients treated with imatinib 400 mg daily as front‐line therapy. Patients who switched to other treatments for intolerance to imatinib were excluded. MR was defined as BCL‐ABR IS ratio >1% ‐ ≤10% (MR1), >0.1% ‐ ≤1% (MR2), >0.01% ‐ ≤0.1% (MR3) and ≤0.01% (MR4 or better). Deep molecular response (DMR) was defined as MR4 or better; patients with DMR lasting ≥2 years and at least a Q‐PCR test every 6 months were defined as sDMR while patients with any sample >0.01% BCR‐ABL IS after the achievement of DMR were defined as unstable DMR. Frequencies were compared by Fisher's exact test. Univariate and multivariate regression analysis were performed using the competing risk model of Fine and Gray. Death and switch to other TKIs for primary or secondary cytogenetic or molecular resistance to imatinib were considered competing events. Results: Our analysis included 185 patients (male: 109; median age at diagnosis: 60 years ‐ range 20‐85; Sokal score: low 74, intermediate 85, high 25, unknown 1). BCR‐ABL transcript types were b2a2 (n = 74, 40%), b3a2 (n = 89, 48%), or b2a2/b3a2 (n = 22, 12%). At 3 months patients with BCR‐ABL>10% IS were 51 (28%), and depth of response in early responders was as follows: MR1 (n = 67, 36%), MR2 (n = 56, 30%), or MR3 (n = 11, 6%). At 6 months patients with no MR, MR1, MR2, MR3, and MR4 were 36 (19%), 44 (24%), 53 (29%), 39 (21%) and 13 (7%), respectively. With a median duration of imatinib therapy of 50 months (range 3‐165), 86 (46%) patients never achieved DMR, 33 (18%) had an unstable DMR and 66 (36%) reached a sDMR. Overall, patients with MR2 or better at 3 months had a cumulative incidence of sDMR of 70.0% (95% CI: 54‐80.5%) at 5 years, significantly superior to those with MR1 or no MR (p = 0.006) Similarly, cumulative incidence of sDMR at 5 years was significantly different across level of MR, ranging from 13.4% (95%CI 0‐29.4%) for MR1 to 78.6% (95%CI 56.1‐89.6%) for MR3 or better (p < 0.001). Patients with b3a2 or b3a2/b2a2 transcripts had a higher probability of sDMR compared to those with b2a2 transcript (60.3% vs 44.6%, p = 0.018). Considering the impact of transcript type across different depth levels of MR, we found that among patients with MR1 at 3 months, those with b3a2 had a significantly higher probability of sDMR (39.5% vs 16.7%, p = 0.029), Summary/Conclusion: This real‐life, retrospective study suggests that patients with very good MR after 3 or 6 months of standard dose imatinib, defined as BCL‐ABR IS ratio ≤1% and ≤0.1% respectively, have a high probability of achieving a sDMR, a prerequisite for therapy discontinuation. Among patients with BCR‐ABL ≤10% but >1% at 3 months, b3a2 transcript identifies patients with a higher possibility of subsequent sDMR.