Discovery of Potent and Selective Next-Generation Pan-BCR-ABL Inhibitors for the Treatment of CML and Ph+ ALL

Introduction

While 6 tyrosine kinase inhibitors (TKIs) are available for patients with chronic myeloid leukemia (CML) that is newly diagnosed (imatinib, dasatinib, nilotinib, bosutinib) or refractory to prior treatments (asciminib, ponatinib), only ponatinib has pan-BCR-ABL activity, i.e., maintains potent activity against all single resistance mutations, including the difficult to inhibit T315I and E255V mutants. BCR-ABL mutations are the major known mechanism of resistance not only in CML, but also in Philadelphia chromosome-positive acute lymphoblastic leukemia (Ph+ ALL), where use of ponatinib in newly diagnosed patients is yielding especially encouraging results, presumably due to its ability to suppress emergence of resistance mutations. However, toxicities observed with ponatinib often lead to dose reduction or termination that can limit the benefit of its pan-BCR-ABL profile. The suboptimal safety profile of ponatinib is consistent with its relative lack of kinome selectivity, and its potent activity against KDR, FLT3, and FGFRs (unique amongst the 6 approved BCR-ABL TKIs) could potentially contribute to its unique toxicities. We therefore set out to identify a selective, potent pan-BCR-ABL inhibitor.

Methods

Compound potency was determined by assessing effects on viability of Ba/F3 cell lines expressing BCR-ABL variants (wild type [WT] and 11 key mutant forms) and off-targets of interest. Preliminary assessment of broader kinome selectivity was performed by calculating the percentage of 88 kinases inhibited by >80% by 1 µM compound (S-score (80)). For efficacy studies, mice implanted with engineered Ba/F3 cell lines were dosed orally once daily.

Results

Ponatinib potently inhibited BCR-ABL WT, T315I, and E255V with IC50s of 2, 4, and 7 nM respectively, and all 9 other BCR-ABL mutants tested were inhibited with IC50s <4 nM. However, ponatinib also potently inhibited the activity of KDR, FLT3, and FGFR1, with IC50s of 5, 2, and 17 nM, respectively, and had a high kinome S-score (80) of 0.30.

Cmpd A is a representative member of a series of compounds we have identified that have a pan-BCR-ABL inhibitory profile and substantially improved selectivity over ponatinib. Cmpd A, potently inhibited BCR-ABL WT, T315I, and E255V, with IC50s of 6, 21, and 55 nM respectively, and 8/9 other BCR-ABL mutants tested were inhibited with IC50s ≤13 nM (Y253H IC50 75 nM). But in contrast to ponatinib, Cmpd A had a high degree of selectivity over FLT3 and FGFR1 (IC50s 1898 and 3323 nM), good selectivity over KDR (IC50 420 nM), and a low kinome S-score (80) of 0.08.

We next compared the efficacy and tolerability of ponatinib and Cmpd A in an efficacy model driven by E255V, a key clinical resistance mutant that both compounds inhibited less potently than T315I. At 10 mg/kg, ponatinib inhibited tumor growth by 64% compared to vehicle-treated mice and was well tolerated. At 25 mg/kg, while ponatinib induced strong tumor regression (by 82% compared to baseline), it was poorly tolerated (eg, keratinized skin in 4/6 mice). In contrast 120 mg/kg Cmpd A induced strong tumor regression, by 55%, and was well tolerated (eg, 0/6 mice had keratinized skin).

Conclusions

We have identified a series of next-generation pan-BCR-ABL inhibitors with substantially improved selectivity over ponatinib. In vitro and in vivo characterization of compounds with further improved selectivity profiles, and the potential to be developed for use in patients with refractory CML and newly diagnosed Ph+ ALL, will be presented.