Direct measurement of in vivo BCR::ABL1 kinase inhibition reveals stronger potency of asciminib as compared with imatinib and nilotinib

To the Editor,

Asciminib is the first approved BCR::ABL1 inhibitor that Specifically Targets the ABL Myristoyl Pocket (STAMP) [1, 2]. In ASC4FIRST, with longer follow-up, asciminib continued to demonstrate a favourable benefit-risk profile over investigator-selected (IS)-tyrosine kinase inhibitors (TKIs) and imatinib [3]. Its specificity minimizes off-target toxicity [1–5]. The degree of BCR::ABL1 inhibition achieved in patients receiving asciminib compared to 2^nd generation TKI (2G-TKI), has not been assessed.

CRKL is a direct substrate of BCR::ABL1, and a lower degree of its phosphorylation correlates with BCR::ABL1 kinase inhibition. In vivo kinase inhibition (IVKI) is defined as the percent reduction in BCR::ABL1 kinase inhibition from day 0 (baseline) to day 7 of frontline therapy, quantified by phospho‑CRKL (Supplementary Methods and Supplementary Fig 1) [6, 7]. Previously we have studied IVKI on patients treated frontline with imatinib (n = 173) or nilotinib (n = 73) [6]. We showed that IVKI with 2G‑TKI (nilotinib) is higher than with imatinib, and low imatinib IVKI ( < 11%) is associated with inferior molecular outcomes and a higher risk of treatment failure and transformation to blast phase, especially in patients who present with significant splenomegaly [6].

In this study, we assessed IVKI in patients treated with frontline asciminib and compared its potency to imatinib and nilotinib (Supplementary Methods). We also examined its relationship with molecular response and other clinically relevant variables. Blood samples at diagnosis (baseline, pre-TKI therapy) and after the first 7 days of treatment were available for IVKI assessment from adult chronic phase CML (CP-CML) patients in 70 of 101 CP-CML patients treated frontline with asciminib enroled in the ASCEND study [8]. Patient demographics are summarized in (Supplementary Table 1).

Pre-treatment median kinase level based on phospho‑CRKL was similar across imatinib [6], nilotinib [6], and asciminib cohorts (56% vs 57% vs 59%, respectively; Fig. 1A). After 7 days of asciminib treatment, the level of phospho‑CRKL was significantly lower than in baseline samples (22% vs 59%, p < 0.0001; Fig. 1B). Moreover, IVKI was significantly higher with asciminib (median 83%) than with imatinib (23%; p < 0.0001) or nilotinib (56%; p < 0.0001; Fig. 1C). Notably, no patients receiving asciminib had low IVKI (defined as below 11% from our previous study [6]), compared with 4% for nilotinib and 23% for imatinib. Unlike our previous findings with imatinib [6], there was no statistically significant IVKI differences observed between the transcript types (e14a2 (84%), e13a2 (82%) and both (e13a2 and e14a2) transcript types (85%); p = 0.73; Supplementary Fig 2A), ELTS (Low: 83%, Intermediate: 84%, High: 85%, p = 0.87; Supplementary Fig 2B), and BCR::ABL1 halving time [9] (p = 0.79; Supplementary Table 2).

Fig. 1. The level of in vivo p-CRKL kinase inhibition for asciminib, nilotinib and imatinib.

A Dotplot shows the dynamic of in vivo phospho-CRKL kinase level (%p-CRKL) for each patient at baseline (day 0) for each tyrosine kinase inhibitor (TKI). The horizontal bar indicates median. B Pairwise lineplot shows in vivo phospho-CRKL kinase level (%p-CRKL) at baseline and at day 7 since asciminib treatment. C Dotplot shows the in vivo kinase inhibition (IVKI) for each patient after the first 7 days of TKI treatment. The horizontal bar indicates median. The horizontal dotted line indicates the IVKI threshold cutoff based on imatinib as previously reported [6].

ASXL1 variants identified at diagnosis recently emerged as a predictor of treatment failure [10, 11], even in the context of highly potent TKIs [12, 13]. We then investigated the relationship between IVKI and CP-CML patients with previously described ASXL1 variants [10, 12] at the time of diagnosis. In asciminib-treated patients, IVKI was similar in patients with ASXL1 variants versus wild-type (77% vs 84%, p = 0.11; Fig. 2A). No correlation was observed for nilotinib (45% vs 57%, p = 0.35; Fig. 2B), nor imatinib (32% vs 22%, p = 0.19; Fig. 2C and Supplementary Fig 3A–C). Furthermore, based on multivariable analysis, asciminib (Odds Ratio (OR):57.2; 95% CI: 52.1–62.4; p < 0.001) and nilotinib (OR:29.6; 95% CI: 24.4–34.7; p < 0.001) were associated with significantly higher IVKI, whereas ASXL1 variants had no significant effect (OR:0.7; 95% CI: −6.6–8.1; p = 0.8; Fig. 2D). To further determine whether the degree of kinase inhibition had an impact on failure-free survival (FFS) as defined by European LeukemiaNet 2020 [14], multivariable Cox proportional hazards modelling was performed, incorporating IVKI, TKI, and ASXL1 variant status. ASXL1 variants (p = 0.002) remained a statistically significant independent predictor of inferior FFS as shown previously [12], but not IVKI (p = 0.34) nor TKI (nilotinib vs imatinib: p = 0.09; asciminib vs imatinib: p = 0.11) were predictors. These findings suggest that the negative impact of ASXL1 on FFS was not due to inadequate kinase inhibition, but rather due to BCR::ABL1 independent pathways. For instance, Miyashita et al showed that the mutant ASXL1 (Y591* truncating mutation) reduces the sensitivity of 3 CML cell lines to multiple TKIs (imatinib, nilotinib, dasatinib, ponatinib and asciminib) through upregulation of ALOX5 expression, which activates downstream AKT signalling and promotes cell survival [15]. Furthermore, they showed that ALOX5 overexpression induces TKI resistance, whereas ALOX5 knockout or pharmacologic inhibition of its downstream signalling using the leukotriene B4 receptor (BLTR) antagonist LY293111 restores TKI sensitivity by suppressing AKT phosphorylation [15].

Fig. 2. The level of in vivo p-CRKL kinase inhibition and its association with mutant ASXL1.

A Dotplot shows the IVKI for each patient after the first 7 days of asciminib treatment between patients with and without ASXL1 variants. The horizontal bar indicates median. B Dotplot shows the IVKI for each patient after the first 7 days of nilotinib treatment between patients with and without ASXL1 variants. The horizontal bar indicates median. The horizontal dotted line indicates IVKI threshold cutoff based on imatinib. C Dotplot shows the IVKI for each patient after the first 7 days of imatinib treatment between patients with and without ASXL1 variants. The horizontal bar indicates median. The horizontal dotted line indicates IVKI threshold cutoff based on imatinib. D Forest plot of multivariable linear regression shows the association of ASXL1 mutation and TKI with IVKI i.e. (IVKI ~ ASXL1 + TKI). Squares represent effect estimates (odd ratio) with 95% confidence intervals. ASXL1 wild-type and imatinib served as reference categories.

We next determined the impact of IVKI, as stratified by quartiles, on molecular response rates in patients treated with asciminib. There were no statistically significant differences across IVKI quartiles, when analysed for MMR achievement by 24 months (MMR: 84%, 88%, 76%, and 88%; p > 0.9; Supplementary Fig 4A), nor MR4.5 by 3 years (61%, 66%, 55%, and 59%; p = 0.8; Supplementary Fig 4B). This suggests that inadequate BCR::ABL1 inhibition is unlikely to be the primary driver of the lack of MMR in these patients.

Our in vivo data indicate that asciminib achieves robust BCR::ABL1 kinase inhibition, relative to imatinib and nilotinib, and may partially explain the clinical observations of faster and deeper molecular responses on asciminib. The absence of a low‑IVKI subgroup [6] among asciminib-treated patients may explain why IVKI appears to have limited prognostic relevance in this setting, in contrast to imatinib, where low IVKI was observed.

In summary, this study provides the first clinical evidence that frontline asciminib delivers uniform, very high levels of ABL kinase inhibition with no patient falling within the low‑IVKI subgroup. The potency and tolerability of asciminib make it a suitable frontline option, especially for patients who would otherwise exhibit low IVKI on imatinib. Additionally, based on our previous observation that levels of IVKI only had a negative impact on response when below 11% [6], we speculate that the frontline dose of asciminib could potentially be reduced without impacting molecular response or treatment failure rate. However, without results from a prospective clinical trial testing this hypothesis, universal dose reductions of frontline asciminib cannot be recommended. ASXL1 variant‑associated risk of treatment failure and resistance appears independent of the level of kinase inhibition achieved, underscoring the need for genomic profiling at diagnosis and for therapeutic strategies that address other drivers of treatment failure. Future work should integrate early IVKI readouts with molecular predictors to refine risk‑adapted frontline management.

Author contributions

CHK designed and planned data analysis, analysed data, interpreted data and wrote the paper. VAS performed research, analysed data, interpreted data and critically reviewed the paper. PD performed research, interpreted data and critically reviewed the paper. NS and SB performed the genomic studies, provided clinical information and critically reviewed the paper. DW critically reviewed the paper. DY and TPH designed and coordinated the correlative studies, contributed key concepts and methodology, interpreted data, and assisted in writing the paper.

Funding

TPH received support from the National Health and Medical Research Council of Australia (APP1135949) and has the financial support of Cancer Council SA’s Beat Cancer Project on behalf of its donors and the State Government of South Australia through the Department of Health. The TIDEL-II study was sponsored by the Australasian Leukaemia and Lymphoma Group. NS received scholarship funding from the Royal Adelaide Hospital Research Foundation Dawes Scholarship. SB received support from the National Health and Medical Research Council of Australia (APP1117718, APP1138935, APP1027531), from the Ray and Shirl Norman Research Trust, and Tour de Cure.

Competing interests

NS received research funding from Novartis and honoraria from Takeda. SB received research funding from Novartis, Terns Pharmaceuticals and Cepheid and honoraria from GSK and Novartis. DY received research funding from Novartis and BMS; and received honoraria from Novartis, Takeda, Pfizer and Amgen. DLW received honoraria from Amgen. TPH received research funding from Novartis and BMS and honoraria from Novartis, BMS, and Takeda. The remaining authors have no conflict of interest to disclose.

Supplementary information

The online version contains supplementary material available at 10.1038/s41375-026-03002-4.