Key Points• For CML patients on TKI therapy, 70% of double mutations in the BCR-ABL1 kinase domain detected by direct sequencing are compound mutations.• Sequential, branching, and parallel routes to compound mutations were observed, suggesting complex patterns of emergence.

Cyclin-dependent protein kinases (CDKs) are central to the appropriate regulation of cell proliferation, apoptosis, and gene expression. Abnormalities in CDK activity and regulation are common features of cancer, making CDK family members attractive targets for the development of anticancer drugs. Here, we report the identification of a pyrazolo[1,5-a]pyrimidine derived compound, 4k (BS-194), as a selective and potent CDK inhibitor, which inhibits CDK2, CDK1, CDK5, CDK7, and CDK9 (IC₅₀= 3, 30, 30, 250, and 90 nmol/L, respectively). Cell-based studies showed inhibition of the phosphorylation of CDK substrates, Rb and the RNA polymerase II C-terminal domain, down-regulation of cyclins A, E, and D1, and cell cycle block in the S and G₂/M phases. Consistent with these findings, 4k demonstrated potent antiproliferative activity in 60 cancer cell lines tested (mean GI₅₀= 280 nmol/L). Pharmacokinetic studies showed that 4k is orally bioavailable, with an elimination half-life of 178 min following oral dosing in mice. When administered at a concentration of 25 mg/kg orally, 4k inhibited human tumor xenografts and suppressed CDK substrate phosphorylation. These findings identify 4k as a novel, potent CDK selective inhibitor with potential for oral delivery in cancer patients.

BACKGROUND Tyrosine kinase inhibitors (TKIs) are the cornerstone of successful clinical management of patients with chronic myeloid leukemia (CML). Quantitative monitoring of the percentage of the fusion transcript BCR-ABL1 (breakpoint cluster region–c-abl oncogene 1, non-receptor tyrosine kinase) BCR-ABL1IS (%BCR-ABL1IS) by reverse transcription–quantitative PCR (RT-qPCR) is the gold standard strategy for evaluating patient response to TKIs and classification into prognostic subgroups. However, this approach can be challenging to perform in a reproducible manner. Reverse-transcription digital PCR (RT-dPCR) is an adaptation of this method that could provide the robust and standardized workflow needed for truly standardized patient stratification. METHODS BCR-ABL1 and ABL1 transcript copy numbers were quantified in a total of 102 samples; 70 CML patients undergoing TKI therapy and 32 non-CML individuals. 3 commercially available digital PCR platforms (QS3D, QX200 and Raindrop) were compared with the platform routinely used in the clinic for RT-qPCR using the EAC (Europe Against Cancer) assay. RESULTS Measurements on all instruments correlated well when the %BCR-ABL1IS was ≥0.1%. In patients with residual disease below this level, greater variations were measured both within and between instruments limiting comparable performance to a 4 log dynamic range. CONCLUSIONS RT-dPCR was able to quantify low-level BCR-ABL1 transcript copies but was unable to improve sensitivity below the level of detection achieved by RT-qPCR. However, RT-dPCR was able to perform these sensitive measurements without use of a calibration curve. Adaptions to the protocol to increase the amount of RNA measured are likely to be necessary to improve the analytical sensitivity of BCR-ABL testing on a dPCR platform.