The BIM You're Born With: A Germline Biomarker for Tyrosine Kinase Inhibitor Resistance

Clinically successful targeted therapy of kinase-driven malignancies including chronic myeloid leukemia (CML) (driven by BCR-ABL1) and select cases of non-small-cell lung cancer (NSCLC)(driven by activating mutations of the epidermal growth factor receptor [EGFR]) is a testament to the imaginative, creative power of biomedical research. Nonetheless, room for continued progress in the field remains. For example, in the IRIS trial of imatinibtreated patients with newly diagnosed CML, primary hematologic resistance occurred in 2 percent of patients, and ~8 to13 percent of patients failed to achieve either a major or a complete cytogenetic response.^1,2  After a median of five years of treatment, secondary resistance (hematologic or cytogenetic relapse, or progression to accelerated/blast crisis CML) developed in 15 to 20 percent of patients but was uncommon after four years of treatment.¹  Secondary resistance is primarily due to the acquisition of point mutations within the tyrosine kinase domain of BCR-ABL1 and much less frequently due to BCR-ABL1 overexpression. BCR-ABL1-independent mechanisms of resistance include 1)clonal evolution; 2) suboptimal plasma levels of imatinib as a consequence of more rapid drug metabolism mediated by cytochrome P450 enzymes; and 3) reduction of intracellular bioavailability of imatinib due to decreased influx (mediated by the human organic cation transporter [hOCT1]), and/or increased efflux mediated by the same adenosine triphosphate-binding cassette (ABC) membrane transporter ABCB1 that is responsible for multidrug resistance. The basis for primary resistance is less clear; however, it is observed more frequently in patients from East Asia (e.g., Singapore, Malaysia, and Japan) than in patients from North America and Europe, suggesting that a genetic predisposition may account for suboptimal TKI responses.

Prior studies have shown that the BCL2 family member BIM (BCL2L11) is needed for TKIs to induce apoptosis in malignancies driven by kinases, and the BH3 domain of BIM is required for this function. In an elegant study, investigators from the National University and Genome Institute in Singapore identified a BIM isoform lacking the pro-apoptotic BH3 domain. The BH3 domain-deficient isoform is the result of a relatively common germline deletion polymorphism. This 2,903 base-pair genomic deletion in intron 2 alters the splicing pattern of the gene such that transcripts lacking exon-4 (that encodes the BH3 domain) predominate. The deletion polymorphism was found in 12.3 percent of healthy East Asian individuals, but it was absent in African and European populations.

The deletion-containing CML cell line KCL22 (derived from a Japanese patient) exhibited resistance to imatinib-induced apoptosis, with the resistance being reversed either by transfection of the cells with the exon-4-containing, BH3-domain transcript or by treatment with the BH3-mimetic drug ABT-737. When the BIM deletion polymorphism was introduced into imatinib-sensitive K562 CML cells, susceptibility to imatinib-mediated apoptosis decreased, and this effect was reversed either by treatment with ABT-737 or by forced expression of BIMEL, a BH3-containing BIM isoform. In support of the findings in the CML-based studies, the NSCLC cell line HCC2279 that contains the BIM deletion polymorphism was also found to be resistant to EGFR inhibitors, and treatment with ABT-737 sensitized these cells to apoptosis induced by the TKI gefitinib.

The authors retrospectively analyzed East Asian patients with CML and found a statistically significant increase in prevalence of the deletion polymorphism in resistant subjects³  compared with optimal responders. In addition, the polymorphism was more prevalent in patients with resistance without BCR-ABL1 mutations compared with either patients resistant to TKIs with BCR-ABL1 mutations or patients sensitive to TKIs. In patients from Singapore and Japan with NSCLC and activating EGFR mutations, the BIM deletion polymorphism was associated with a shorter progression-free survival (6.6 months vs. 11.9 months), and the BIM deletion polymorphism and the TKI-resistant exon 20 mutation of EGFR were the only independent prognostic factors identified in a multivariate analysis.