Overcoming clinical BCR-ABL1 compound mutant resistance with combined ponatinib and asciminib therapy

Abstract

BCR-ABL1 compound mutations can lead to resistance to ABL1 inhibitors in chronic myeloid leukemia (CML), which could be targeted by combining the ATP-site inhibitor ponatinib and the allosteric inhibitor asciminib. Here, we report the clinical validation of this approach in a CML patient, providing a basis for combination therapy to overcome such resistance.

Figure 1.. Clinical history and response of a CML patient harboring a compound mutation to combined therapy with ponatinib and asciminib.

(A) Schematic highlighting a relevant timeline of the patient’s diagnosis and response to past TKI therapies. CML-CP: chronic myeloid leukemia in chronic phase. (B) Gene expression levels of a panel of drug transporter genes from different timepoints on past treatments. Levels for K562 and asciminib-resistant K562 cells are included as controls. (C) Schematic highlighting timeline, dosing changes, and response of the patient, who harbored a BCR-ABL1 T315I/E355G compound mutant at baseline, to combination treatment with ponatinib and asciminib. CHR: complete hematologic response. (D) Changes in levels of the patient’s relevant clinical lab values over the course of combination treatment.

Figure 2.. In vitro sensitivity of the BCR-ABL1 T315I/E355G compound mutant to single-agent versus combination treatment with ponatinib and asciminib.

(A) Ba/F3 cells expressing BCR-ABL1 T315I/E355G were treated in vitro with ponatinib and asciminib alone or in combination at the indicated concentrations for 72 h. Viability normalized to untreated cells is shown, with each bar representing the mean ± SEM for four independent replicates. (B) Synergy of the combination in Ba/F3 BCR-ABL1 T315I/E355G cells across a dose matrix of possible concentrations. Synergy was calculated using the zero interaction potency (ZIP) model, where the heatmap reflects regions of synergy (red), additivity (white), and antagonism (blue). The mean synergy value across the full matrix of combined doses is shown. (C) Immunoblot analysis of Ba/F3 cells expressing BCR-ABL1 with a T315I, E355G, or T315I/E355G mutation. Cells were treated in vitro for 4 h with the indicated concentrations of each inhibitor, then probed for CRKL phosphorylation levels by western blot, as a readout of BCR-ABL1 kinase activity. β-tubulin was included as a loading control.

Figure 3.. Evidence of dysregulated, BCR-ABL1 kinase-independent resistance pathway activation prior to combination treatment discontinuation.

(A) Violin-plot distributions of overall fluorescence intensity values for phosphoprotein markers from single-cell imaging analysis of primary patient cells conducted at various timepoints on combination treatment. Mononuclear cells isolated at the indicated timepoints were fixed, permeabilized, and stained for each of the indicated phosphoproteins along with CD34. Analysis shown is confined to those cells expressing CD34. (B) Breakdown of relative expression levels of each phosphoprotein with the CD34+ cells from the patients at the indicated timepoints on combination treatment. Expression level for each marker in each cell analyzed per sample was binned into high (orange), intermediate (green), and low (blue) groups defined by the top quartile, middle 50%, and bottom quartile of fluorescence intensity, respectively. (C) Select fields of view and inset zoom on a representative cell comparing pERK expression from baseline and Day 75 timepoints. Cells of interest within the view are highlighted with dashed yellow boxes, and example zoomed views of one of those cells is show in the inset. Images of the identical field of view for each stain (differential interference contrast (DIC) only, CD34 (green), and pERK (purple)) are shown for both timepoints on treatment.

Figure 4.. NMR-based structural impact of the T315I/E355G compound mutant on ABL1 kinase conformational state.

(A) 2D heteronuclear correlation NMR analysis of the impact of the ABL1 E355G mutation. The relative abundance of both the inactive conformation of ABL1 kinase to which imatinib binds (I₂) and the active conformation of the kinase are indicated in the ¹H-¹³C plots shown for key kinase residues. The ABL1 T389Y mutation serves as a control which induces an approximately equal abundance of both conformational states for visualization; addition of the E355G mutant on top of this background control reveals a dramatic shift toward favoring the active conformation. (B) Mapping chemical shift perturbations (CSP) caused by the E355G mutation onto ABL1 kinase domain. The Glu side chain is shown with stick representation, and methyl groups throughout are shown as spheres, where the coloration (yellow through red) indicate the increasing degree of chemical shift perturbation observed at that position as a result of the mutation. For reference, the activation loop is highlighted in blue. (C) ¹H-¹³C correlations are shown for select residues identified as having large or small to medium degree of CSP as a result of the E355G mutation.