CXCR7 antagonist ACT-1004-1239 – discovered and developed at Idorsia Pharmaceuticals Ltd. – is a “beautiful beast” of a molecule from a process chemist’s point of view (Scheme 1). The molecule immediately catches one’s eye with two unique heterocyclic amides, which are beautifully arranged in a trans-3S,4S-configuration on a substituted piperidine ring. However, the presence of exactly these two stereocenters, coupled with the intermediacy of the highly polar and fluorinated heterocyclic building blocks, make ACT-1004-1239 also a beast of a molecule to produce on larger scale.  

Initially, a fit-for-purpose racemic route, consisting of 18 chemical steps, was developed in less than 12 months with one single R&D lab. This initial work included the development of robust procedures for the two heterocyclic building blocks and a racemic, highly crystalline piperidine intermediate. After cis-to-trans epimerization of the C3-position, separation of the resulting trans-enantiomers by liquid chromatography on a chiral stationary phase and a final synthetic sequence, the API was isolated as a white solid in high purity. Over 500 g of API were produced in-house for preclinical activities with this racemic route, which was also replicated to produce 5.5 kg of GMP-material at an external manufacturing partner for Phase 1 clinical studies.

After this initial work on the racemic route, the focus was shifted towards the development of a stereoselective 2^ndgeneration route for ACT-1004-1239, as the large-scale chromatographic separation of enantiomers was not a viable option for a resupply campaign that targeted >30 kg of API. The cornerstone of this 2^nd generation route was the highly stereoselective reduction of a chiral piperidyl enamine derived from inexpensive (S)-1-methylbenzylamine. 240 g of API were produced in-house with the novel stereoselective route, which was then also used to manufacture over 30 kg of GMP-material at an external partner for Phase 2 clinical studies.

In addition to the development of two different routes to the API – including scalable routes to all individual building blocks – a thiadiazole-based follow-up compound was also prepared on kg-scale, using the same stereoselective approach as for the frontrunner ACT-1004-1239. Finally, while working on the highly polar pyrimidine building block for ACT-1004-1239, a new methodology to prepare trifluoromethylated imidazo-fused N-heterocycles was serendipitously found, using inexpensive and atom-efficient TFAA as CF₃-source.

Scheme 1: The Process Chemistry Tale of ACT-1004-1239.