ALBERT

Description: The JAK2V617F somatic point mutation has been described in patients with myeloproliferative disorders (MPDs). Despite this progress, it remains unknown how a single JAK2 mutation causes 3 different MPD phenotypes, polycythemia vera (PV), essential thrombocythemia, and primitive myelofibrosis (PMF). Using an in vivo xenotransplantation assay in nonobese diabetic-severe combined immunodeficient (NOD/SCID) mice, we tested whether disease heterogeneity was associated with quantitative or qualitative differences in the hematopoietic stem cell (HSC) compartment. We show that the HSC compartment of PV and PMF patients contains JAK2V617F-positive long-term, multipotent, and self-renewing cells. However, the proportion of JAK2V617F and JAK2 wild-type SCID repopulating cells was dramatically different in these diseases, without major modifications of the self-renewal and proliferation capacities for JAK2V617F SCID repopulating cells. These experiments provide new insights into the pathogenesis of JAK2V617F MPD and demonstrate that a JAK2 inhibitor needs to target the HSC compartment for optimal disease control in classical MPD.

Description: Introduction The acquisition of ABL1 Kinase Domain (KD) mutations represent the most frequent resistance mechanism in CP-CML patients (pts) treated with tyrosine kinase inhibitors (TKI). Currently, the standard assay relies on a poorly sensitive technique, Sanger Sequencing (SS). Thus, the detection of these mutations using SS might be too late to trigger a timely treatment change. In a national phase III academic trial (PETALs, EudraCT 2013-004974-82), we evaluated prospectively the value of a more sensitive technique, Next Generation Sequencing (NGS) to detect KD ABL1 mutations in newly diagnosed CP-CML patients randomized to get nilotinib 600 mg/d for 6 years ± Pegylated-IFN-α2a (Peg-IFN) 45 μg/wk for 2 years in combination. Methods Newly diagnosed CP CML pts ≤65 years were randomized 1:1 to get NIL 300 mg BID alone (M0 to M48, arm A) vs Peg-IFN alone for 30 days (M-1→M0) 30 mg/wk as priming, prior to NIL 300 mg BID + Peg-IFN 30 μg/wk 2 weeks, upgraded to 45 μg/wk thereafter, for up to 2 y (M0 to M24, arm B) followed by NIL alone for 4 more years unless pts enter a treatment-free remission phase. In addition to KD mutational analysis performed by SS as per protocol, patients also had KD mutational analysis performed by NGS at M3, M6, M12 and 6-monthly thereafter until achievement of a stable MMR, regardless of response. NGS assay was performed as previously described (Kizilors et al. Lancet Haematol 2019). Results Two hundred pts were randomized (99 in A, 101 in B), of which 96 patients (51/99 in A, 45/101 in B, p=0.399) underwent a KD mutational analysis performed by NGS as part of this study. The remaining 104 patients are currently being screened and the full dataset will be presented. Among the 96 patients tested, there was no difference in the distribution between the 2 arms with respect to gender, age [median 45 years (18-66)] or risk factors distribution (p=0.862 and 0.328 for Sokal and ELTS respectively in patients tested at 3 months). The median follow-up of this cohort is 45.0 (33.2-58.7) months. By 12 months, 11 patients [8/51 (11.8%) in A, 3/45 (6.6%) in B] had developed a KD mutation. After only 3 months of TKI therapy, 3 patients were found mutated (Y253H 2 pts, T315I 1 pt), of whom 2 pts were only detected using NGS. At M6, a KD mutation was found in 8 pts [A: 7 patients, B: 1 pt, (p= 0.055), of which 6/8 were not detected by SS, due to either low level Variant Allele frequency (VAF, n=5) or low level BCR-ABL transcript levels (n=1). Y253H mutations were found in 4 pts, T315I in 2 pts and E255K in 1 pt. Consecutively to KD mutation identification, 6/8 patients lost their response and were withdrawn from study (1 pt with a Y253H detected at M3 progressed to advanced phase), while 1 pt lost MMR at last follow-up and another pt with a mutation sensitive to nilotinib achieved MMR. KD mutations were detected while pts were in optimal response at M6 [BCR-ABL

Description: The combination of 2GTKI+pegylated IFN-α (Peg-IFN) is an attractive approach for first-line treatment of CP CML, inducing high rates of deep molecular responses in phase II trials. Thus, we evaluated nilotinib (NIL) alone versus NIL+Peg-IFN in newly diagnosed CP-CML patients (pts) in a randomised phase III trial (PETALs, EudraCT 2013-004974-82). Newly diagnosed CP CML pts ≤65 y, without prior history of arterial occlusion were randomized 1:1 to get NIL 300 mg BID alone (M0 to M48, arm A) vs Peg-IFN alone for 30 days (M-1→M0) 30 μg/wk as priming, prior to NIL 300 mg BID + Peg-IFN 30 μg/wk 2 wks, upgraded to 45 μg/wk thereafter, for up to 2 y (M0 to M24, arm B) followed by NIL alone for 4 more years unless pts enter treatment-free remission (TFR). The primary endpoint is the rate of MR4.5 by 1 y. As a secondary endpoint, pts reaching MR4.5 ≥2 y are allowed to stop NIL and enter a TFR phase in both arms. The trigger for treatment resumption is loss of MMR. All molecular assessments are centralised, quantifications are expressed as BCR-ABL/ABL1 (IS) in % with ≥32,000 copies of ABL1 as control. Two hundred pts were randomized (99 in A, 101 in B), 130 M and 35 F in each arm, median age of 46 (18-66) y. Median follow-up is 43.8 (34.3-55.9) Mo. Results are analysed in intention-to-treat. Sokal and EUTOS LTS scores were H in 25% and 2.5%, Int. in 33% and 16.5% and L in 42% and 81% pts respectively equally balanced. Median age is 46 (18-66) y, 18 pts (9%) had ACAs, all pts have a "Major" BCR transcript. CHR was obtained in 9.6% of pts at M0 (in B) and 88% of pts in A and 90.4% of pts in B at M1. CCyR rates at M3 were 63% vs 75% in A and B (p=ns), and BCR-ABL1 ≤1% at M6 were 87% in A vs 93% in B (p=ns). By M12, the rates of MMR were 68.1% vs 70.1% (p=0.44), MR4 were 34% vs 47.5% (p=0.041), MR4.5 were 15.9% vs 21.5% (p=0.049), MR5 11.7% vs 23.71% (p=0.023), in A vs B respectively. By M36 the rates of MMR were 83% vs 86.6% (p=0.31), MR4 were 70.2% vs 71.13% (p=0.50), MR4.5 were 37.2% vs 49.5% (p=0.05), MR5 33% vs 42.3% (p=0.12), in A vs B respectively The overall cumulative incidence of MR4.5 is superior in B (54.6 [43.7-65.5]%) vs A (44 [31.5-54]%) close to significance (unilateral Fisher test, p=0.05, see Figure). Seven patients were mutated by Sanger in A (5 Y253, 1 E255K, 1 T315I) vs 2 in B (2 T315I). One pt (A) progressed toward AP and then myeloid BC with a Y253H mutation, is still alive in CMR on Ponatinib. Twenty nine (29%) pts were withdrawn from study in A (toxicity 9, cancer 3, resistance 14, investigator decision 2, lost for FU 1) vs 26 (26%) pts for B (toxicity 13, resistance 8, investigator decision 5), 1 pt died from cervix cancer (A). Median overall doses of NIL delivered by M36 were 600 mg/d in both arms (p=ns). The median overall dose of Peg-IFN delivered in B by M24 was 37.5 mg/wk. The overall rate of grade 3-4 hematologic toxicities was 22%; with 2% and 7% thrombocytopenia, 4% and 6% neutropenia, and 1% and 1% pancytopenia in A vs B respectively. Major grade 3-4 non-hematologic toxicities consisted in 9% of cardiac disorders in A (2 coronaropathies, 1 myocardial infarction, 2 thoracic pains, 2 atrial fibrillation, 1 bradycardia, 1 palpitations, 1 pericarditis) vs 8% in B (2 coronaropathies, 1 myocardial infarction, 3 atrial fibrillation, 1 palpitations, 1 pericarditis), 4% vascular disorders in A (1 thrombophlebitis + PE, 1 transient ischemic attack, 1 PAOD, 1 carotid stenosis) vs 3% in B (1 thrombophlebitis, 1 PAOD, 1 transient ischemic attack). Three % of gastro-intestinal disorders were observed in A (2 pancreatitis, 1 anal fissure) vs 6% in B (2 pancreatitis, 1 anal fissure, 1 abdominal pain, 2 cholecystectomies); 5% auto-immune disorders in B (1 recurrent pericarditis, 2 hemolytic anemia, 1 ITP, 1 thyroiditis); 5 and 8 pregnancies (2 pts + 3 partner Arm 1, 3 pts + 5 partner Arm B), despite recommended contraceptive methods. Secondary tumours were diagnosed in 4% (1 breast, 1 cervix, 1 thyroid, 1 neuroendocrine) in A vs 2% of pts (1 neuroendocrine and 1 testis) in B. Of note 8% psychiatric episodes were reported in B pts (2 unsuccessful suicide attempts), vs 2% in A. We observed 9% lipase elevations in A, 6% in B, 2% cholestatic episodes in A, 6% in B; 3% of transaminase elevations in A vs 2% in B. Infections were detected in 3% A vs 7% in B. The combination of NIL + Peg-IFN seems to provide somewhat higher MR4.5 rates by M36 in newly diagnosed CP CML pts without inducing significant higher toxicities than NIL alone. Whether this will translate in higher TFR rates is under evaluation. Final updated results at M36 will be presented Disclosures Nicolini: Sun Pharma Ltd: Consultancy; Novartis: Research Funding, Speakers Bureau; Incyte Biosciences: Honoraria, Research Funding, Speakers Bureau. Etienne:Novartis: Consultancy, Honoraria, Speakers Bureau; BMS: Honoraria, Speakers Bureau; Incyte Biosciences: Honoraria, Speakers Bureau; Pfizer: Honoraria, Speakers Bureau. Huguet:Servier: Honoraria; Amgen: Honoraria; Novartis: Honoraria; Incyte Biosciences: Honoraria; Jazz Pharmaceuticals: Honoraria; Pfizer: Honoraria; BMS: Honoraria. Guerci-Bresler:Novartis: Honoraria, Speakers Bureau; BMS: Honoraria, Speakers Bureau; Pfizer: Honoraria, Speakers Bureau; Incyte Biosciences: Honoraria, Speakers Bureau. Charbonnier:Incyte Biosciences: Honoraria, Speakers Bureau; Novartis: Consultancy; Pfizer: Consultancy. Legros:Novartis: Honoraria; Pfizer: Honoraria, Research Funding; Incyte Biosciences: Honoraria, Research Funding; BMS: Honoraria. Coiteux:Pfizer: Honoraria, Speakers Bureau; BMS: Honoraria, Speakers Bureau; Incyte Biosciences: Honoraria, Speakers Bureau; Novartis: Honoraria, Speakers Bureau. Cony-Makhoul:BMS: Consultancy, Honoraria, Speakers Bureau; Pfizer: Consultancy; Incyte Biosciences: Honoraria, Speakers Bureau; Novartis: Consultancy. Roy:Incyte Biosciences: Consultancy. Rousselot:Pfizer: Research Funding; Incyte: Research Funding. Quittet:Novartis: Honoraria, Speakers Bureau. Ame:Incyte Biosciences: Honoraria, Speakers Bureau. Rea:Pfizer: Honoraria, Membership on an entity's Board of Directors or advisory committees; Novartis: Honoraria, Membership on an entity's Board of Directors or advisory committees; Incyte Biosciences: Honoraria; BMS: Honoraria. Dulucq:Novartis: Honoraria, Speakers Bureau; Incyte Biosciences: Honoraria, Speakers Bureau; BMS: Honoraria, Speakers Bureau. Mahon:Novartis: Consultancy, Honoraria, Speakers Bureau; Pfizer: Honoraria, Speakers Bureau; BMS: Honoraria, Speakers Bureau; Incyte Biosciences: Honoraria, Speakers Bureau. OffLabel Disclosure: Pegylated Interferon alpha 2 a is not licensed in this setting

Description: Introduction Deep molecular response (DMR) are now highly desirable goals in the treatment of CP-CML, especially in the front-line setting, because it can lead to a definitive treatment-free remission (TFR). However, such a goal is difficult to attain and does not concern the majority of patients (pts), but currently the precise number of pts able to access to TFR is unknown. Aims We aim to determine the number or newly diagnosed CP-CML pts reaching DMR, stable DMR, and access to TFR, on Imatinib (IM, Glivec®) first-line. Methods We retrospectively analyzed in an observational study, a cohort of newly diagnosed CP-CML pts treated with IM first-line 400 mg daily alone in our 3 reference centers between 2000→2018. All pts were followed according to the ELN recommendations 2006, 2010 and 2013. Clinical data were extracted from medical files, and responses (hematologic, cytogenetic, molecular) were analysed according to standard methods. Molecular results were standardised according to the ELN/Eutos programs since 2003, and were all expressed as BCR-ABLIS in %. DMR have been defined according to the ELN (NCP. Cross et al., Leukemia 2015). Stability of DMR has been defined as a stable if ≥2 years at least on 4 datapoints. TFR has been proposed to pts presenting the only current recommended criteria: MR4.5 ≥2 years at least on 4 datapoints [(Rea et al., Cancer 2018)], in the 3 centers involved, within clinical trials, pioneered in our country, or now as a clinical routine recommendation. Loss of MMR was the trigger for TKI resumption after IM cessation for TFR. Overall survival (OS), progression-free survival (PFS), failure-free survival (FFS, defined as progression to advanced phases death, loss of CHR, CCyR, or MMR, discontinuation of IM for toxicity, primary cytogenetic resistance) were analysed since IM initiation in intention-to-treat. Results Four hundred and eighteen pts have been included in this study, with a median age of 60.7 (48-70) years at diagnosis, with 57% males and 43% females. Sokal score (n=401) was low in 32%, intermediate in 51% and high in 17%. ACA were present at diagnosis in 5.5% of the pts (NA in 1.44%). Major BCR transcripts were found in 98% of pts, and atypical transcripts in 1.9%. CHR was reached in a median of 1 (0.85 to 1.64) month of IM,

Description: Key Points Nilotinib induced deeper molecular responses than continued imatinib in patients with minimal residual disease on long-term imatinib. These deeper responses may enable more patients to benefit from treatment-free remission trials.

Description: Abstract 2298 We previously reported that imatinib treatment could be safely discontinued in chronic myeloid leukemia (CML) patients who had achieved a sustained complete molecular remission (CMR) lasting for at least two years in succession. However, in the multicenter French STIM study, 60 % of patients loose CMR mainly within the first 6 months after discontinuation and the probability to remain in CMR 2 years after imatinib discontinuation is around 40 %. The molecular relapse originates in persisting leukemic cells that are undetectable by current RQ-PCR techniques. The sensitivity of these routinely used techniques is below a detection threshold corresponding to a 5-log reduction in the leukemic burden. We thus hypothesized that improving the sensitivity of leukemic cell detection would be of valuable help to better select those patients who can possibly be cured after imatinib treatment. In the present work, we have increased the sensitivity of the current RQ-PCR, based on repeated PCR and on an increased number of normal ABL copies that were analyzed, in order to augment the probability to amplify BCR-ABL. For each sample, 5 microg of RNA were used to synthesize 5 cDNA and for each cDNA, 2 PCR for BCR-ABL and one for ABL were carried out, i.e 10 BCR-ABL PCR points were performed per sample. For each microg of RNA the number of ABL copies was over 20 000, so the number of ABL copies analyzed was over 100 000. The detection threshold of BCR-ABL was calculated and considered at 40 cycles of PCR (theshold Ct). The control plasmid (pME-2) (Kindly given by Martin Müller and Andreas Hochhaus from European LeukemiaNet) used as standard curve dilution was also included to test the sensitivity and for instance 4 copies were detected in 19/20 cases, 2 copies 7/10 cases,1 copy in 2/10, 0.4 copy in 1/10. Thirty one samples from healthy donors or non CML patients (BCR-ABL negative) served as controls. Among the 310 (10 × 21) PCR for BCR-ABL we found only one positive well. 65 patients enrolled in the STIM study were analyzed at the time of imatinib discontinuation, using this new RQPCR technique. In the STIM study relapse was defined as the positivity of Bcr-Abl transcripts using classical QRT-PCR confirmed by a second analysis point indicating the increase in relation to the first analysis point performed on 2 successive assessments. Among 650 PCR for BCR-ABL, 46 wells were found positive. 9 patients were found at least 3 /10 positive PCR and among them 6 relapsed. 22 patients were found only at least 1 /10 positive and among them 15 relapsed. In 43 patients, BCR-ABL was never detected using the more sensitive RQ-PCR technique and 22 of them relapsed. To conclude, the use of a new RQPCR technique with a sensitivity of detection of BCR-ABL close to 6-log does not allow the prediction of molecular relapse following imatinib discontinuation in patients in CMR. Our results are in agreement with what was previously reported using PCR on genomic DNA. The persistence of leukemic cells in CMR patients does not automatically lead to CML relapse. Disclosures: Belanger: Novartis Pharma: Employment.

Description: Abstract 3246 Poster Board III-183 Introduction BCR-ABL confers to several cell lines an apoptosis-resistant phenotype that is critically dependent on its kinase activity. It has been shown that BCR-ABL can block the mitochondrial step of apoptosis and that tyrosine kinase inhibition induces apoptosis in BCR-ABL expressing cells. Several pro- and anti-apoptotic proteins of the Bcl-2 family modulate the mitochondrial apoptotic signal. Amongst them, anti-apoptotic Bcl-xL and Mcl-1 are induced by BCR-ABL. Several studies underlined the probable role of Bim expression during tyrosine kinase inhibition-induced apoptosis and we previously confirmed that cells which were depleted in Bim were unable to undergo apoptosis when treated by two tyrosine kinase inhibitors (TKI) : imatinib (IMA) and nilotinib (NIL). Altogether, these results strongly support the prominent role of the balance between pro- and anti-apoptotic proteins of the Bcl-2 family in the apoptotic response of CML cells after TKI treatment. It has been shown that Bim actually is an indirect activator of apoptosis through its anti-Bcl-xL, -Mcl-1 and -Bcl-2 effects. Recently, ABT-737, a small molecule which binds Bcl-2 and Bcl-xL but not Mcl-1, was shown to induce apoptosis in several tumour cell types. In this study, we verify if ABT-737 could cooperate efficiently with the Bim stabilizing TKI. Methods and Results K562 cells were incubated with increasing concentrations of either imatinib or nilotinib, alone or in combination with ABT-737, in a constant ratio. Both TKI and ABT-737 induced a dose dependant apoptosis and the combination of IMA or NIL with ABT-737 resulted in a synergistic cooperation to induce apoptosis. The combination index (CI) calculated using the Calcusyn software was 0.05 for ABT-737 and either IMA or NIL at ED90. Similar experiments were performed on CD34 expressing cells from 12 CML patients, and the strong synergism of ABT-737 with both IMA and NIL was confirmed with a mean CI of 0.25. To elucidate the mechanisms underlying this cooperation, we analyzed the Bcl-2 family proteins expression and confirmed the increase of Bim and the decrease of Bcl-xL due to TKI treatment while ABT-737 was without effect on the expression of these proteins. Pull-down experiments on K562 cells using anti-Bcl-2 and anti-Bcl-xL antibodies followed by analysis of Bim showed that ABT-737 decreases the interaction between BIM and Bcl-2 or Bcl-xL in cell free system as in intact cells. Moreover, Bim was co-immunoprecipitated more efficiently by anti-MCL-1 antibodies when it was displaced from Bcl-2 and Bcl-xL by ABT-737. This brings some mechanistic support to the complementary effects of TKI (stabilizing Bim) and ABT-737. The expression of several anti-apoptotic proteins was also analyzed. Surprisingly, the combination of TKI with ABT-737 induced a dramatic decrease in XIAP content while each inhibitor alone was without effect on this protein. The decrease in XIAP was accompanied by an increase in caspase 3 cleavage. Cell sorting experiments showed that the decrease in XIAP preceded the drop in mitochondrial membrane potential and the activation of caspase 3 and it was only partly inhibited by a pan-caspase inhibitor and not at all by proteasome or cathepsin B inhibitors. Conclusion ABT-737 cooperates with TKI to induce apoptosis of CML cells. This cooperation can be explained in part by the increase in Bim activity induced by TKI associated to the inactivation of Bcl-2, -xL by ABT-737. However, a side effect at the level of the caspase inhibitor XIAP also seems to participate, resulting in a strong synergism. Thus, the association between TKI and Bcl-2 inhibitors could be a facilitating strategy to induce apoptosis in some BCR-ABL expressing cells resistant to TKI alone. Disclosures Mahon: Amgen: Honoraria; Novartis Pharma: Consultancy, Honoraria, Research Funding; Alexion: Consultancy, Honoraria.

Description: The Bcr-Abl inhibitor imatinib is now the first line therapy for all newly diagnosed CML patients in chronic phase. Nevertheless resistance to the drug emerges as CML progresses to an acute deadly phase. Additional cellular targets should thus be identified to develop alternative therapeutic strategies. The transcription factor NF-kB is a pro-survival factor, found abnormally active in numerous hematologic malignancies. In the present study we show that the constitutive and abnormal activation of NF-kB in Bcr-Abl transformed BaF3 cells and in the LAMA84 CML line could be downregulated after inhibition of Bcr-Abl. Pharmacological blockade of NF-kB by the IKK2 inhibitor AS602868 (Serono International S.A.) prevented proliferation of BaF3/Bcr-Abl, LAMA84 and primary CML cells. Importantly, AS602868 led to apoptosis of an imatinib resistant variant of LAMA84 and of BaF3 clones expressing mutated form of Bcr-Abl derived from imatinib resistant patients. Moreover, NF-kB inhibition affected proliferation and hematopoietic colony formation of primary imatinib resistant CML cells. Finally, the IKK2 inhibitor prolonged survival of mice intravenously injected with the imatinib resistant clone LAMA84-r. Our data strongly suggest that NF-kB mediates important survival functions in CML cells for bcr-abl and that targeting NF-kB with the IKK2 inhibitor AS602868 may represent a new promising therapeutic strategy for CML treatment.

Description: Abstract 1684 The early assessment of molecular responses to tyrosine kinase inhibitors (TKI) used as a front line therapy for chronic phase chronic myeloid leukemia (CML) might represent an attractive surrogate marker for survival and help to discriminate poor prognosis patients (pts). So, we studied in a retrospective bicentric institutional analysis a series of 142 de novo CP CML pts diagnosed between 2000 and 2011, receiving imatinib (IM) 400 mg daily (n=85 pts), second generation TKI (TKI2) (n=57 pts) with 37 dasatinib (DAS) pts 100 mg daily and 20 nilotinib (NIL) pts 600–800 mg daily as front-line therapy, enrolled or not in clinical trials. All pts were assessed for their cytogenetic and molecular responses (local RQ-PCR for BCR-ABL, expressed as BCR-ABL/ABL ratios (IS) in %). Failure to TKI was defined as progression to accelerated phase or blast crisis, death, loss of complete hematologic response, loss of complete cytogenetic response, confirmed loss of major molecular response (MMR), discontinuation of TKI because unacceptable toxicity, primary cytogenetic resistance. The definition of progression included the same variables except the two last. Pts were considered in MMR with a BCR-ABL/ABL ratio ≤0.1% (IS). There were 87 males (61%) and 55 females with a median age of 55 years (21–83) at diagnosis. Six pts (4%) had an additional clonal abnormality, 4 a masked Philadelphia (Ph) chromosome, and 4 a variant Ph. Three pts had an atypical BCR-ABL transcript, all in IM400 group. Sokal scores were low for 37 (26%), intermediate for 61 (43%) and high for 43 (31%) (1 NA), and Euro scores were low for 49 (35%), intermediate for 78 (55%), and low for 14 (10%) (1 NA). These scores were similarly balanced in the IM400, NIL and DAS groups. The median time between diagnosis and TKI was 1 (0.9–1.28) month, not significantly different between TKI groups. The median follow-ups were 56, 37, and 19 months for IM400, NIL and DAS groups respectively. Fifteen percent, 6%, and 14% of pts achieved MMR at 3 months (M3) (p=ns); 31%, 37%, and 40% at M6 (p=ns); 55%, 70% and 71% at M12 (p=ns); 67%, 90% and 87% at M18 (p=ns) in IM400, NIL and DAS groups respectively, taken into account the limited numbers of pts at latest time-points in the 2 TKI2 arms. Twenty-three pts achieved 4-log molecular response MR4 in IM400, 10% in NIL and 11% in DAS at latest follow-ups. Four pts died, 3 in blast crisis in IM400 group and 1 for unrelated reason to disease or treatment in NIL group. A multivariate analysis adjusted on progression-free survival (PFS) performed, did not identify any significant parameter including age, gender, Sokal and Hasford scores, ACA, type of transcript, interval between diagnosis and TKI start. The progression-free survival (PFS) was 94 (83–100)% for pts achieving M3 MMR vs 86 (74–100)% if MMR occurred at later time points vs 86 (73–100)% for pts that never achieved MMR (p=0.02). There was no difference for PFS between the 3 TKI (log-rank p=0.9). None of the M3 MMR switched to another treatment, whereas 10 % of other pts did so for resistance (n= 10, 1 and 2 for IM400, NIL and DAS respectively), 9% for intolerance (n=10, 1 and 4), and 3% for other reasons (n=4, 0 and 0) with no difference between the 3 drugs (p=0.74). The failure-free survival (FFS) was 95% for M3 MMR pts vs 65% for pts with MMR occurring at later time points, versus 0% if MMR has not been achieved (p