BMS-962212, a parenteral Factor XIa inhibitor, was scaled-up for toxicity studies. Two steps of supercritical fluid chromatography (SFC) were developed for the chiral resolution of the penultimate and achiral purification of final active pharmaceutical ingredient (API), BMS-962212. A robust SFC process using Chiralcel OD-H with methanol-acetonitrile as modifier in CO2 was established to achieve a stable and uninterrupted operation with reduced mobile phase viscosity and system pressure drop. More than 230 g of the racemic penultimate was chirally resolved to reach >99% chiral purity, ready for final tert-butyl ester deprotection to provide the API. There were a significant number of impurities in BMS-962212 generated from the final step that needed to be removed. In contrast to conventional SFC conditions, an SFC method exploiting water and ammonia as additives in both the mobile phase and sample solution was developed to accomplish purification and desalting (i.e. removing TFA) of the zwitterionic API in one step. Water as an additive eliminated salt precipitation and improved the resolution while ammonia contributed to the desalting, details of which will be discussed in this article. A throughput of 2 g/h was achieved, and >80 g of the crude API was purified. The same strategy was applied to another Factor XIa API (compound A) and its penultimate.

An efficient and “endotoxin-free” purification of a cyclic dinucleotide (CDN) STING agonist was achieved to produce multigram quantities of pure BMT-390025, an active pharmaceutical ingredient (API), for toxicological studies. A two-step sub/supercritical fluid chromatography (SFC) procedure was developed for the achiral purification and desalting of the polar ionic CDN. A robust SFC process employing methanol-acetonitrile-water with ammonium acetate as co-solvent in CO2 on BEH 2-ethylpyridine was established and scaled up as the first step to achieve a successful purification. The desalting/salt-switching (i.e. removing acetate and acetamide) was conducted using methanol-water with ammonium hydroxide as co-solvent on the same column in the second step to convert the final API to the ammonium salt. Water with additive was essential to eliminating salt precipitation and improving the peak shape and resolution. Due to the extreme hydrophilicity of BMT-390025, 65% of co-solvent was needed to adequately elute the target in both steps. More than 40 g of crude API was purified and desalted producing >20 g of pure BMT-390025 as the ammonium salt which was obtained with a chemical purity of >98.5% and met the endotoxin requirement of <0.1 EU/mg. In addition, >80 g of its penultimate prior to the deprotection of the silyl group was purified at a high throughput of 6.3 g/h (0.42 g/day/g SP).

Inhibition of the bromodomain and extra-terminal (BET) family of adaptor proteins is an attractive strategy for targeting transcriptional regulation of key oncogenes, such as c-MYC. Starting with the screening hit 1, a combination of structure-activity relationship and protein structure-guided drug design led to the discovery of a differently oriented carbazole 9 with favorable binding to the tryptophan, proline, and phenylalanine (WPF) shelf conserved in the BET family. Identification of an additional lipophilic pocket and functional group optimization to optimize pharmacokinetic (PK) properties culminated in the discovery of 18 (BMS-986158) with excellent potency in binding and functional assays. On the basis of its favorable PK profile and robust in vivo activity in a panel of hematologic and solid tumor models, BMS-986158 was selected as a candidate for clinical evaluation.

A regioisomeric mixture of the nucleoside derivative, Intermediate 1, required resolution by preparative supercritical fluid chromatography (SFC) in order to obtain the desired regioisomer as a key intermediate in a STING agonist program. Various chiral columns and solvents including methanol, acetonitrile, isopropanol, and the mixture of acetonitrile and isopropanol as organic modifiers in carbon dioxide at different temperatures were screened to obtain the best regioisomeric resolution. A key issue associated with interconversion between the regioisomers via silyl migration during purification was investigated in methanol, acetonitrile, and the mixture of acetonitrile and isopropanol, and the optimal organic modifier in CO2 was established to mitigate the interconversion to an acceptable level (<5%). Taking into account peak resolution, throughput, interconversion and operation robustness, an efficient SFC method for large-scale purification was successfully developed and scaled up onto a 5 cm I. D. Chiralcel OJ-H column using 25% acetonitrile: isopropanol [1:1 (v/v)] with 0.1% ammonium hydroxide as the modifier in CO2 at a total flow rate of 270 mL/min and a temperature of 30°C. In addition, continual evaporation (i.e. every hour) of the desired isomer fraction stream post-separation ensured minimal further interconversion. A total of 258 grams were separated at a high throughput of 8.6 g/h. Regioisomeric purity of the desired isomer of Intermediate 1 was ≥98.2% and the recovery was ≥90.2%. A similar purification strategy was applied to the regioisomeric resolution of Intermediate 2, an analog of Intermediate 1. In total, 1028 grams of Intermediate 2 were processed at a high throughput of 12.5 g/h on a Viridis BEH 2-EP column. The regioisomeric purity of the desired isomer was ≥96.8% and the recovery was ≥90.7%.

A pure β-​D-​Glucopyranosiduronic acid metabolite (≥98.0 % purity and a single impurity ≤0.50 %) was requested for biological studies. Due to its unusual instability, the purification of the glucuronide metabolite was extremely challenging. Initially, the crude sample (89 % HPLC area purity) was purified on a Waters SunFire C8 OBD column with 40 mM ammonium acetate buffer and acetonitrile as the mobile phase under a gradient program. The purified glucuronide metabolite solid was obtained by evaporation and lyophilization. However, this procedure yielded the target compound with 97.6 % HPLC area purity and did not meet the requirements. Through the investigation, lyophilization was identified as the key step for the purity of the metabolite, and further lyophilization resulted in an increased amount of the degraded impurities. To better understand the compound, stability studies of the purified metabolite were conducted under sample media, organic solvent, acid, base, and light exposure. The compound was observed to be extremely unstable in water, acid, base and methanol, and sensitive to light, but relatively stable in ammonium acetate buffer (pH 5.0). Taking into account compound stability and the initial purification method, the improved purification procedure was successfully developed and the purified glucuronide metabolite was obtained with 99.2 % HPLC area purity and 0.39 % of the largest single impurity.

Abstract Background: The bromodomains and extra-terminal domain (BET) proteins are a family of 4 adapter proteins, BRD2, BRD3, BRD4, and BRDT, that bind to specific acetylated lysine residues on the histone tails of chromatin and recruit additional proteins to regulate gene transcription. The c-MYC oncogene, which is amplified and deregulated in 40% to 70% of all cancers, is directly regulated by BET proteins. Preclinical studies provide a strong rationale for pursuing transcriptional regulation via BET inhibition in cancer treatment (Lenhart, et al. Mol Cancer Ther. 2015;14:2167-2174; Filippakopoulos, et al. Nature. 2010;468:1067-1073). Here, we present results of crystal structure-guided structure-activity relationship (SAR) studies that resulted in the identification of BMS-986158, a highly potent BET inhibitor. Methods: Using fluorescence resonance energy transfer (FRET), we screened a library of compounds and identified a carbazole series of BET inhibitors. Alkylation of the carbazole nitrogen resulted in a 10-fold boost in potency against BET. We then created a differently oriented carbazole series and, subsequently, a carboline series of compounds to improve potency and pharmaceutical properties. A thermal shift assay was used to evaluate selectivity for binding to the BET family of bromodomains. Results: Crystal structure and subsequent SAR studies demonstrated that the isoxazole moiety formed critical interactions with the BET bromodomains. Lead compounds demonstrated potent binding to BRD4 and reduction in c-MYC expression and proliferation in cell lines such as KMS-11. Accessing a second lipophilic pocket in the BRD4 binding site increased potency significantly. Modification of the lead series from a carbazole carboxamide to a carboline resulted in significant improvement in pharmaceutical properties and led to the identification of BMS-986158, which demonstrated in vitro and in vivo potency against a variety of tumor types. In c-MYC-driven cancer cell lines, BMS-986158 caused dose-dependent downregulation of c-MYC expression and induced cancer cell death. BMS-986158 demonstrated > 70% tumor growth inhibition at tolerated doses in patient-derived xenograft models (lung, colorectal, and triple-negative breast cancers). Antitumor activity in mice and pharmacokinetic properties in animal studies support oral dosing in humans. Conclusions: Structure-based drug design led to the discovery of BMS-986158, a highly potent BET inhibitor. With promising antitumor activity in preclinical studies, BMS-986158 is currently being evaluated in a phase 1/2a clinical trial in patients with advanced cancers. Citation Format: Ashvinikumar V. Gavai, Derek Norris, David Tortolani, Daniel O'Malley, Yufen Zhao, Claude Quesnelle, Patrice Gill, Wayne Vaccaro, Tram Huynh, Vijay Ahuja, Dharmpal Dodd, Christopher Mussari, Lalgudi Harikrishnan, Muthoni Kamau, John S. Tokarski, Steven Sheriff, Richard Rampulla, Dauh-Rurng Wu, Jianqing Li, Huiping Zhang, Peng Li, Dawn Sun, Henry Yip, Yingru Zhang, Arvind Mathur, Haiying Zhang, Christine Huang, Zheng Yang, Asoka Ranasinghe, Celia D'Arienzo, Ching Su, Gerry Everlof, Lisa Zhang, Nirmala Raghavan, John T. Hunt, Michael Poss, Gregory D. Vite, Richard A. Westhouse, Susan Wee. Discovery of clinical candidate BMS-986158, an oral BET inhibitor, for the treatment of cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 5789.