Primary asciminib resistance in a chronic myeloid leukemia patient with the atypical BCR::ABL1 e13a3 transcript: a case study

Treatment with tyrosine kinase inhibitors (TKI) has significantly improved the survival of patients with chronic-phase (CP) chronic myeloid leukemia (CML), with the majority now attaining a normal life expectancy. Nevertheless, approximately 20-30% of patients still experience therapy failure under conventional TKI, often caused by the emergence of resistance mutations at the inhibitor’s binding site.^1,2 The newly approved BCR::ABL1 inhibitor asciminib provides new therapeutic perspectives for CML patients, who have experienced treatment failure due to side effects or insensitivity towards conventional TKI. Asciminib functions as an allosteric BCR::ABL1 inhibitor by binding to the myristate pocket of the ABL1 kinase domain - a mechanism that differs from established ATP-competitive TKI.³ Asciminib‘s unique target region makes it especially suitable for patients with resistance mutations within the ATP-binding site, including the T315I mutation. As well as being used in heavily pretreated patients, asciminib has recently also received Food and Drug Administration approval for treating newly diagnosed CP-CML patients.⁴

This novel approach is particularly relevant due to the critical regulatory role of the ABL1 kinase. The ABL1 gene encodes for a tyrosine kinase, which plays a pivotal role in cell growth regulation, differentiation, and preserving genomic integrity. In healthy individuals this critical protein is equipped with a built-in regulatory mechanism: the myristoylated ABL1 N-terminus binds to its myristate pocket, which leads to the proximity of the SH2 and the SH3 domains. The newly formed SH2-SH3 clamp then attaches to the kinase domain, where it suppresses ABL1 activity by allosterically blocking the active center of the kinase domain.^5,6 In individuals with CML, the fusion of BCR to ABL1 cuts off the myristoylated N-terminus, which results in a loss of the autoinhibitory cascade and an overly active kinase. In contrast to the standard ATP-competitive TKI, asciminib does not bind the kinase domain’s ATP site, but rather its myristate pocket.⁷ In this manner, it can trigger the formation of the SH2-SH3-complex and effectively restore kinase autoinhibition in BCR::ABL1-positive cells. Hence, the integrity of the SH2 and SH3 domains is strictly required for the asciminib-mediated suppression of kinase activity. The SH3 domain is encoded by ABL1 exons 2 and 3, which are located N-terminally of the kinase domain within the ABL1 protein. In 98% of BCR::ABL1-positive patients, the defining mRNA fusion occurs at BCR exon 13 or 14 and ABL exon 2, resulting in the e13a2 and e14a2 isoforms.^8,9 Much rarer transcripts are e1a2, followed by e19a2, accounting for 1.3% and 0.7% of newly diagnosed patients at our laboratory between 2021 and 2024. In these transcripts the SH3, SH2 and kinase domains remain functionally intact. However, BCR::ABL1 gene fusion can also occur with ABL1 exon 3, forming the e13a3 and e14a3 variants. Patients with these isoforms lack the entire ABL1 exon 2 and therefore also parts of the SH3 domain’s coding sequence (Figure 1A). Fusions lacking ABL1 exon 2 accounted for 0.7% of newly diagnosed CML patients in our laboratory between 2021 and 2024.

Recent in vitro and protein modeling studies demonstrate that the loss of exon 2 functionally impairs the correct folding of the SH3 domain and consequently ABL1 autoinhibition. This truncation of the SH3 domain has been demonstrated to confer a high degree of resistance towards the TKI asciminib.^10,11 In view of asciminib‘s relatively recent approval and brief history of therapeutic use, however, there is so far only very limited data to verify the described resistance mechanism in clinical use. We here report clinical evidence of the naturally conferred resistance against the novel BCR::ABL1 inhibitor asciminib in a CML patient with e13a3 transcript type who was treated with asciminib before the original publication of the resistance mechanism.

In October 2019, a 67-year-old male’s peripheral blood was referred to our laboratory for evaluation of CML. The patient gave informed consent for the use of laboratory results and clinical data for research purposes according to the Declaration of Helsinki. The study was approved by the laboratory’s institutional review board. The patient presented with a white blood cell count of 14.3x10⁹/L, a hemoglobin level of 15.2 g/dL and a platelet count of 213x10⁹/L. A BCR::ABL1 -specific multiplex polymerase chain reaction identified a rare e13a3 fusion transcript with a BCR::ABL1/ABL1 ratio of 204.66% upon quantitation with the in-house-developed BCR::ABL1 p210 quantitative assay.¹² It is important to highlight that the assay is optimized for the quantitation of the major breakpoints e13a2 and e14a2 (Figure 1B). In the typical e13a2/e14a2 fusions, both the BCR::ABL1 fusion and the ABL1 reference amplicons are amplified within exon 2, thus yielding BCR::ABL1/ABL1 ratios below 100%. In contrast, in a3 fusions lacking exon 2, this quantitation method can yield ratios above 100%, as only ABL1 transcripts are amplified in the reference gene replicates.

Other MPN-driver mutations in the genes JAK2, CALR and MPL were excluded by sequencing of the respective mutation hot spots. Morphological examinations of the bone marrow smears showed a cell-rich bone marrow with hyperplasia of the granulopoiesis and evidence of all stages of maturation. Erythropoiesis was reduced and the megakaryocytes were regular. Myeloid blasts could not be detected, confirming the diagnosis of CML. Fluorescence in situ hybridization on the bone marrow sample confirmed the BCR::ABL1 fusion in 71 of 100 interphase nuclei and no additional cytogenetic alterations were found. According to the EUTOS long-term survival score, the patient was at low risk. For all follow-up investigations of the e13a3 transcript, the BCR::ABL1 p210 quantitative assay was applied.

Figure 1.Fusion of BCR to ABL1 exon 3 leads to a shortened SH3 domain and altered amplification via p210 LightCycler assay. (A) Schematic of BCR::ABL1 domains and exons with BCR (red), SH3 (blue), SH2 (green) and the kinase domain (purple) in the a3- and a2 transcript types. (B) Schematic of p210 LightCycler assay design with arrows indicating the primer binding sites. The ABL1 reference forward primer is capable of binding in ABL1 exon 2 of the reference gene, but also the BCR::ABL1 fusion gene in the case of a2 fusion transcripts. In a3 fusion transcripts double binding is not possible and BCR::ABL1/ABL1 ratio calculation is altered, leading to ratios exceeding 100%.

The patient received first-line treatment with the second -generation TKI dasatinib at a dose of 100 mg/day. This caused a decrease in BCR::ABL1/ABL1 ratio by three orders of magnitude to 0.14% within 3 months (Figure 2), indicating optimal treatment response. After experiencing a pleural effusion, therapy was switched to 600 mg/day of nilotinib. Six months of nilotinib treatment led to an even further decrease of the BCR::ABL1 ratio by more than four orders of magnitude from the baseline value, indicating an individual molecular response (IMR) of four.¹³ This response level remained constant over time, yet treatment had to be discontinued after 40 months due to the increased risk for cardiovascular adverse events under nilotinib. This 7-week treatment-free period was accompanied by a rapid increase in BCR::ABL1 ratio. Consequently, imatinib 400 mg/day was started, which had to be discontinued due to the onset of diarrhea and edema. Despite the necessity to pause and adjust the patient’s treatment several times, satisfactory response rates were achieved under each of the conventional TKI.

In January 2024 the therapeutic regimen was changed to asciminib in response to the patient‘s persistent side effects. Within the first 3 months under asciminib, BCR::ABL1/ ABL1 ratio increased by two orders of magnitude, from 0.26% to 3.98% and subsequently reached 10.49%, indicating a loss of IMR. As a consequence of the patient´s non-responsiveness towards asciminib, in April 2024 the BCR::ABL1 kinase domain was analyzed for resistance mutations by next -generation sequencing of amino acids 184 to 510.¹⁴ No kinase domain mutation was detected, indicating an alternative underlying resistance mechanism.

Figure 2.Patient with a13a3 fusion is resistant towards asciminib. Course of BCR::ABL1/ ABL1 ratio under ATP-competitive tyrosine kinase inhibotors (TKI) and asciminib over time measured via the BCR::ABL1 p210 Light-Cycler assay. Each symbol indicates one measurement of BCR::ABL1/ABL1 ratio at different treatment stages: grey triangles indicate no treatment, blue circles indicate treatment with ATP-competitive inhibitor and red squares indicate treatment with asciminib. Bars above the graph indicate in blue period of treatment with ATP-competitive TKI, in red treatment period with asciminib, and no bar indicates a treatment-free period.

After 3 months of asciminib treatment, dasatinib was rein-troduced at a reduced dose. Under 50 mg/day of dasatinib, treatment response could be successfully restored, leading to an IMR of four.

This case, along with prior in vitro evidence and a recent clinical case report, demonstrates that BCR::ABL1 fusions lacking ABL1 exon 2 lead to SH3 domain disruption, which significantly compromises the efficacy of asciminib.^10,11 Although e13a3 and e14a3 are the predominant fusions lacking ABL1 exon 2, further transcripts with the a3 break point, such as e1a3, e6a3 or e19a3 are possible.

Asciminib is gradually being used in an increasing number of patients. Thus, resistance mechanisms are only now unraveling. Besides point mutations surrounding the myristoyl-binding pocket, mutations occurring in the SH3 domain, such as Y115N, have been found to also mediate asciminib resistance.^3,15 This further underscores the essential role of the SH3 domain in ABL1 autoinhibition, as demonstrated in this case study.

Superior efficacy and a favorable safety profile led to Food and Drug Administration approval of asciminib also as a first-line treatment.^3,4 However, as shown in this case, patients with the e13a3 or e14a3 BCR::ABL1 fusion should be considered at risk of primary resistance to asciminib. Even more critically, patients with newly diagnosed CML typically exhibit high cancer cell burdens and in some of these cases ineffective treatment with asciminib could favor a fast progression to blast phase.

Despite experiencing side effects, which resulted in several treatment interruptions, the here investigated patient responded well to ATP-competitive TKI. Further, following asciminib failure, the reintroduction of dasatinib was able to restore the patient’s therapeutic response. These findings strongly suggest the usage of ATP-competitive TKI in patients with e13a3 or e14a3 transcripts and highlight the need for identifying CML patients‘ BCR::ABL1 isoforms at diagnosis.

Although the majority of CML patients carry the e14a2 or 13a2 BCR::ABL1 isoforms, the described asciminib resistance is highly relevant for patients with the a3 transcript type. Our findings are consistent with previous reports of asciminib resistance mediated by the BCR::ABL1 e13a3 and e14a3 fusion transcripts. The case described here contributes much-needed clinical data that confirms its relevance in a real-world scenario.

Footnotes

• Received July 23, 2025
• Accepted October 7, 2025

Correspondence

F. DICKER - frank.dicker@mll.com

Disclosures

TH discloses part ownership of the Munich Leukemia Laboratory (MLL). BKJ, SW, MM and FD are employed by the MLL Munich Leukemia Laboratory. CK discloses part ownership of the Outpatient Therapy Center for Hematology and Oncology Offenburg.

Contributions

Acknowledgments

The authors thank all MLL Munich Leukemia Laboratory co-workers for their dedicated work and all physicians who provided samples, cared for patients, and collected data.