Asciminib monotherapy in patients with chronic-phase chronic myeloid leukemia with the T315I mutation after ≥1 prior tyrosine kinase inhibitor: 2-year follow-up results

Asciminib targets the BCR::ABL1 myristoyl pocket, maintaining activity against BCR::ABL1T315I, which is resistant to most approved adenosine triphosphate–competitive tyrosine kinase inhibitors. We report updated phase I results (NCT02081378) assessing safety/tolerability and antileukemic activity of asciminib monotherapy 200 mg twice daily in 48 heavily pretreated patients with T315I-mutated chronic-phase chronic myeloid leukemia (CML-CP; data cutoff: January 6, 2021). With 2 years’ median exposure, 56.3% of patients continued receiving asciminib. Overall, 62.2% of evaluable patients achieved BCR::ABL1 ≤1% on the International Scale (IS); 47.6% and 81.3% of ponatinib-pretreated and -naive patients, respectively, achieved BCR::ABL1IS ≤1%. Of 45 evaluable patients, 48.9% achieved a major molecular response (MMR, BCR::ABL1IS ≤0.1%), including 34.6% and 68.4% of ponatinib-pretreated and -naive patients, respectively. MMR was maintained until data cutoff in 19 of 22 patients who achieved it. The most common grade ≥3 adverse events (AEs) included increased lipase level (18.8%) and thrombocytopenia (14.6%). Five (10.4%) patients experienced AEs leading to discontinuation, including 2 who discontinued asciminib and died due to COVID-19; these were the only deaths reported. These results show asciminib’s effectiveness, including in almost 50% of ponatinib pretreated patients, and confirm its risk-benefit profile, supporting its use as a treatment option for T315I-mutated CML-CP.

A previous analysis of the phase I X2101 trial (NCT02081378) in heavily pretreated patients with Ph+ CML-CP/AP, including T315Imutated CML-CP/AP, first demonstrated the safety of asciminib monotherapy QD or twice daily (BID) at 10 to 200 mg [26].After a median follow-up of approximately 14 months, asciminib demonstrated a favorable safety and tolerability profile, with 11% and 6% of patients with CML-CP with or without the T315I mutation, respectively, discontinuing therapy due to AEs [26].The maximum tolerated dose of asciminib was not reached [26].Preclinical observations suggested that a 4-to 13-fold higher asciminib concentration was required for adequate inhibition of BCR::ABL1 T315I compared with non-mutated BCR::ABL1, and in X2101 the majority of patients with T315I-mutated CML-CP who achieved responses had received doses of ≥150 mg BID [12,23,26,27].Use of the highest tested dose of asciminib, 200 mg BID, was justified by overall major molecular response (MMR; BCR::ABL1 IS ≤0.1%) rates predicted by pharmacokinetic/ pharmacodynamic analysis in patients with T315I-mutated CML-CP and may improve the probability of response by maximizing the predicted proportion of patients with asciminib exposure above the preclinical 90% maximal effective concentration [28].Results from this trial supported full approval of asciminib 200 mg BID for patients with T315I-mutated CML-CP in the US with subsequent approvals worldwide [29][30][31].
We report efficacy and safety data in 48 patients with T315Imutated CML-CP receiving asciminib 200 mg BID monotherapy with a median duration of exposure of 2 years in X2101; separate analyses of patients with CML-CP receiving asciminib 150 or 160 mg and patients with CML-AP receiving asciminib 200 mg BID are also included.

Study oversight
The study was designed by the sponsor (Novartis Pharmaceuticals) in collaboration with study investigators.The sponsor collected and analyzed data in conjunction with the authors.All authors contributed to the development and writing of the manuscript and vouch for the accuracy and completeness of the data and the fidelity of the study to the protocol.

Study design
This phase I, first-in-human, dose-finding study was described previously [25,26].The current analysis focused on 48 patients with CML-CP with a confirmed T315I mutation who received a starting dose of asciminib monotherapy 200 mg BID in the dose-escalation or -expansion phases of this study arm.BCR::ABL1 mutational analyses were performed centrally by ICON (Portland, OR, USA) using Sanger sequencing.Patients were ≥18 years old with cytogenetically confirmed Ph+ CML-CP with the T315I mutation who were resistant to or intolerant of ≥1 prior TKI (per 2009 European LeukemiaNet recommendations) [32].Efficacy and safety data for patients with T315I-mutated CML-CP who received asciminib 150 mg BID (n = 5) and 160 mg BID (n = 6) and efficacy data for patients with T315I-mutated CML-AP who received asciminib 200 mg BID (n = 4) are reported.The primary objective was to determine the maximum tolerated dose and/or recommended dose for expansion of asciminib monotherapy.Secondary objectives included assessing the safety and tolerability, pharmacokinetics, and preliminary antileukemic activity of asciminib.

Study assessments
Coding and grading of AEs, assessment of molecular response, and BCR::ABL1 mutational analysis were described previously [25].

Statistical analyses
Assessment of molecular response rates by time point was defined previously [25].The analyses herein are based on data collected by the January 6, 2021, cutoff, when all patients had completed their week 60 follow-up visit or discontinued earlier and represent cumulative rates unless otherwise specified.
In all 48 patients, the Kaplan-Meier estimated rate of event-free survival at week 96 was 87% (95% CI, 77-97%) (Supplementary Fig. S2A).The median time to event was not reached.Two patients with additional mutations at baseline did not achieve MMR (Supplementary Table S7); one remained on therapy and achieved BCR::ABL1 IS ≤1% by data cutoff; the other discontinued the study due to physician decision.Six patients acquired additional mutations during the study (M244V, M351T, F359I, A337T, and F359V).The patient who acquired F359V did so after they had achieved MMR and lost it at week 96 (no mutations detectable at that time); at the next assessment, F359V was detected.BCR::ABL1 IS levels continued to increase until the patient discontinued treatment to receive a transplant.The 5 remaining patients did not achieve MMR; 4 discontinued treatment (3 due to lack of efficacy, 1 due to disease progression), and 1 (with M244V) remained on study drug and achieved BCR::ABL1 IS ≤1% by data cutoff.
In patients who received asciminib 150 mg BID (n = 5) or 160 mg BID (n = 6), BCR::ABL1 IS ≤1% and MMR rates did not increase after week 24 (Supplementary Tables S8 and S9).Among evaluable patients, 1 patient with 150 mg BID and 2 with 160 mg BID were in MMR by week 96.
Of 4 patients in AP (Supplementary Table S10), 2 remained on treatment by data cutoff for ≥72 and ≥168 weeks.These 2 patients had BCR::ABL1 IS >1% to ≤10% at baseline.One achieved MMR by week 48, the other achieved first MR 4.5 by week 24, and both maintained responses at cutoff.Two patients had BCR::ABL1 IS >10% at baseline and discontinued treatment, one after treatment for ≥24 weeks without achieving BCR::ABL1 IS ≤1% and the other after treatment for 23 days without postbaseline assessment.
Four (8.3%) patients had AOEs (Supplementary Table S14); 2 (4.2%) were grade 1 and 2 (4.3%) were grade 3. When adjusted for patient-year exposure, the incidence of all-grade and grade 3 AOEs was 4.3% and 2.2%, respectively (Supplementary Table S15).One patient experienced cerebrovascular accident (grade 3) 13 days after hospitalization due to COVID-19 and died 23 days posthospitalization due to pneumonia with COVID-19 as a contributing factor.Another patient experienced 3 grade 3 AOEs (peripheral arterial occlusive disease of the right and left side and coronary artery disease), all of which occurred separately over >1300 days of treatment.Percutaneous transluminal angioplasty, electrocardiography, and heart catheterization were performed, and a stent was implanted.Peripheral arterial occlusive disease of both sides later improved to grade 2. Grade 1 AOEs included left carotid artery disease (n = 1) and cerebrovascular accident (n = 1; reported 2 days after the patient discontinued asciminib due to disease progression).Of the 4 patients with AOEs, all had prior exposure to ≥3 TKIs (imatinib and ponatinib [n = 3]).A lower exposureadjusted incidence of AOEs was observed in ponatinib-naive (2.4%) than in ponatinib-pretreated (5.9%) patients.All patients with AOEs had ≥1 past and/or active CV risk factor at baseline, including arterial hypertension (n = 4), coronary artery disease (n = 1), and cerebrovascular accident (n = 1).No AOEs led to asciminib dose adjustment, interruption, or discontinuation.
Cardiac failure occurred in 1 patient (Supplementary Table S16) who had exposure to 4 prior antileukemic therapies and multiple baseline CV risk factors.On study day 22, pre-existing degenerative aortic valve disease worsened to grade 2, and aortic valve disease (aortic valve vitium) and tricuspid valve incompetence (both grade 1) developed.The patient continued receiving asciminib until hospitalization due to COVID-19 (day 888), a cause of death.
The achievement of MMR is an important treatment milestone that predicts improved OS, and progression-free survival [1,33,36,37], and is associated with improved durations of CCyR [38].Almost half of patients (22 [48.9%]), including most patients who achieved BCR::ABL1 IS ≤1% (16 of 23), achieved MMR.MMR was achieved by patients with all observed baseline BCR::ABL1 IS levels and in patients who discontinued prior ponatinib for both intolerance and resistance.Responses were sustained, with most patients (19 of 22) who achieved MMR maintaining MMR by data cutoff.In PACE, 56% and 58% of patients with T315I-mutated CML-CP receiving ponatinib 45 mg QD achieved MMR by 12 and 57 months, respectively; 60% of patients who achieved a response by 57 months were estimated to maintain MMR at 5 years [13,35].In OPTIC, 34.4%, 24.7%, and 23.1% of patients with CML-CP with or without the T315I mutation at ponatinib starting doses of 45, 30, and 15 mg, respectively, had MMR at cutoff (32 months' median follow-up) [16].
Deep molecular response (MR 4 and MR 4.5 ) is a strong predictor of OS [39] and can minimize risk of loss of CCyR or MMR [40].MR 4  is required for treatment-free remission eligibility [1,41], and MR 4.5  has been associated with better event-free survival and failure-free survival than CCyR [39,42].In this analysis, >50% of patients who achieved MMR improved to a deeper response level, with 13 (28.9%)and 11 (24.4%)patients achieving MR 4 and MR 4.5 , respectively.
After 2 years of median exposure, safety results were consistent with the known safety profile of asciminib; no new or worsening safety signals arose in this heavily pretreated patient population.Few patients (5 [10.4%]) experienced AEs leading to discontinuation.Most AEs occurred early (within the first 6 months of treatment); this pattern was also observed in pretreated patients with CML-CP without the T315I mutation from this trial and the phase III ASCEMBL trial [25,43].New AEs occurring in ≥10% of patients predominantly after 6 months of treatment included increased lipase level, upper respiratory tract infection, cough, edema, pyrexia, dizziness, increased amylase level, vomiting, anemia, and lower respiratory tract infection; most were grade 1/2.
In the treatment of CML, TKIs have been associated with CV toxicity [18,29,[44][45][46][47][48][49][50], and CV risk profile is a clinical concern in selecting, starting, and monitoring TKI treatment [46].Results from preclinical models suggest that asciminib may have less potential for cardiotoxicity than ponatinib [51,52].In the current analysis, few patients (4 [8.3%]) experienced AOEs.The role of asciminib in these events remains uncertain due to multiple confounding factors, including treatment with multiple prior TKIs and baseline CV risk factors; data from ongoing studies of asciminib in treatment-naive patients will further clarify the CV risk profile of asciminib.In a cohort of 115 patients with CML-CP without the T315I mutation from this trial treated with asciminib doses ranging from 10 to 200 mg QD or BID, with 4.2 years' median exposure, 10 (8.7%) patients experienced AOEs; none led to treatment discontinuation [25].In the ASCEND clinical trial of asciminib 40 mg BID starting dose in 101 newly diagnosed patients with median follow-up of 23 months, 1 patient had 2 vascular events [53].In ASCEMBL, AOEs occurred in 3.2% of patients who received asciminib 40 mg BID with a median followup of 1.2 years [24].In an updated analysis of ASCEMBL with a median follow-up of 2.3 years, the frequency of AOEs rose to 5.1%, but the exposure-adjusted AOE rate (3.0 per 100 patient-years) decreased from that of the primary analysis (3.3) [43,54].Patients in these analyses who experienced AOEs while receiving asciminib received several prior TKIs and had multiple baseline CV risk factors [25,43,55].Of 270 patients with CML-CP who received ponatinib 45 mg QD in PACE, 84 (31%) experienced ≥1 AOEs (including CV [16%], cerebrovascular [13%], and peripheral vascular [14%] events) after a median exposure of 32.1 months (range, 0.1-73.0)[13].OPTIC used ponatinib dose-reduction strategies to assess treatment efficacy while limiting the occurrence of AOEs; after 32 months median follow-up, AOEs were reported in 9.6%, 5.3%, and 3.2% of patients receiving ponatinib 45, 30, and 15 mg, respectively [16].Ponatinib response- based dose-reduction strategies reduced the risk of AOEs by approximately 60% [21].In this analysis of X2101, only 1 patient experienced cardiac failure, which occurred contemporaneously with COVID-19 pneumonia from which the patient later died.The role of asciminib remains uncertain due to pretreatment with multiple prior TKIs, baseline CV risk factors, and lack of pretreatment assessment of ejection fraction.While the reported CV events do not constitute a safety signal, risk factors and comorbidities should be closely monitored and managed during asciminib therapy in accordance with clinical practice recommendations.Pancreatic toxicity poses a safety concern for patients with CML receiving TKI therapy and warrants clinical awareness as enzyme elevations may occur in the absence of evidence of pancreatitis, although the risk factors and mechanism of TKI-associated pancreatic toxicity are unknown [27,56,57].In this analysis with asciminib, pancreatic toxicity events (31.3%) were mainly asymptomatic enzyme elevations.Only 1 patient experienced pancreatitis, which was grade 2 and did not require treatment changes.In a previous analysis of this trial, which included patients with CML-CP/AP with and without the T315I mutation receiving asciminib monotherapy 10 to 200 mg QD or BID (14 months' median followup), clinical pancreatitis occurred in 3% of patients and resolved in all patients within 10 days of discontinuing asciminib [26].In a cohort of 115 patients with CML-CP without the T315I mutation from this trial, pancreatic enzyme elevations and clinical pancreatitis were reported in 46 (40.0%) and 8 (7.0%) patients, respectively [25].In contrast, no cases of pancreatitis were reported in ASCEMBL [24,43].The asciminib prescribing information recommends monitoring patients monthly or as clinically indicated for pancreatic toxicity, with dose modifications as needed [29].
In the current study, across all observed molecular endpoints, more patients achieved milestones receiving asciminib without prior ponatinib treatment; overall rates were 1.7-, 2-, 2.2-, and 2.4fold higher in ponatinib-naive than -pretreated patients for BCR::ABL1 IS ≤1%, MMR, MR 4 , and MR 4.5 , respectively.A similar trend was observed in a real-world analysis of patients with CML treated with asciminib, including 4% with the T315I mutation; in ponatinib-pretreated and -naive patients, respectively, without the indicated response at baseline, 27.3% and 53.3% achieved CCyR, 20% and 52.2% achieved MMR, and 10.5% and 19.4% achieved MR 4.5 after a median follow-up of 30 months [58].In the current study, rates of BCR::ABL1 IS ≤1% and MMR were 2.3-and 3.7-fold higher, respectively, in ponatinib-pretreated patients who discontinued ponatinib due to intolerance versus resistance.Responses were sustained regardless of prior treatment with ponatinib.The safety profile of asciminib was independent of prior ponatinib treatment, with no differences observed between populations in rates of grade ≥3 AEs or AEs requiring additional therapy, reinforcing the safety of asciminib across lines of treatment.AOEs were more frequent in ponatinib-pretreated than -naive patients.The tolerability of asciminib and higher efficacy in ponatinib-naive patients may provide rationale for using asciminib before ponatinib in patients with T315I-mutated CML-CP, although asciminib also retains activity after ponatinib.
In conclusion, asciminib exhibited clinical efficacy, a sustained safety profile, and tolerability in patients with T315I-mutated CML-CP who have limited treatment options.Sustained responses were observed in both ponatinib-naïve and -pretreated patients, and >50% of patients remained on therapy at 2 years' median exposure.This analysis reinforces the utility of asciminib as a treatment option for patients with T315I-mutated CML-CP, regardless of ponatinib pretreatment status.Mainly mild dermatologic events, including rash, rash maculopapular, dermatitis acneiform, eczema, and urticaria.One event of allergic rhinitis was also observed.

b
Reasons for physician decision included loss of response (n = 3), loss or lack of response with transfer to stem cell transplant (n = 3), lack of efficacy (n = 2), to receive stem cell transplant (n = 2), and transfer to investigator-initiated trial (n = 1).c Patients with disease progression to accelerated phase.d Patient travel to trial site restricted due to COVID-19.Moved to a managed access program.e Patient required prohibited medication (hydroxyurea).Continued treatment under a managed access program.

Fig. 3
Fig.3Cumulative MMR by time point in patients not in MMR at baseline.a Cumulative MMR in all evaluable patients (A) overall, (B) by BCR::ABL1 IS at baseline, and (C) by prior treatment with ponatinib.IS International Scale, MMR major molecular response.a Treatment discontinuations for any reason were treated as competing events.

Fig. 4
Fig.4Adverse reactions (≥10% at first occurrence) over time.a,b ADR adverse drug reaction, AE adverse event, LRTI lower respiratory tract infection, URTI upper respiratory tract infection.a Includes reported AEs and adverse drug reactions.b Proportions are calculated based on the number of patients at risk of an event (ie, patients ongoing treatment and event-free at the start of the interval).A patient with multiple occurrences of an event in the same time interval is counted only once in that time interval.The safety topics correspond to either single preferred terms or groups of preferred terms according to the adverse drug reaction definitions.

Table 2 .
Patient demographics and baseline clinical characteristics.Cumulative BCR::ABL1 IS ≤1% by time point in patients with baseline BCR::ABL1 IS > 1% a .Cumulative BCR::ABL1 IS ≤1% in all evaluable patients (A) overall, (B) by BCR::ABL1 IS at baseline, and (C) by prior treatment with ponatinib.IS International Scale.a Treatment discontinuations for any reason were treated as competing events.

Table 3 .
AEs reported irrespective of study treatment relationship by preferred term reported in ≥10% of patients.
a Musculoskeletal pain includes preferred terms for musculoskeletal pain and pain in extremity.

Table 4 .
Adverse events of special interest.represent counts of patients.A patient with multiple severity grades for an adverse event is only counted under the maximum grade.MedDRA version 23.1, CTCAE version 4.03, Case Retrieval Strategy version released February 25, 2021.
a Numbers (n) b