CCR5 antagonists prevent entry of HIV into host cells by binding to the CCR5 co-receptor and stopping the gp120-CD4 complex from making the crucial interaction with CCR5 that leads to fusion of the viral and host cell membranes (see Wikipedia article for more information). Pfizer’s Maraviroc (Selzentry, or Celsentri outside the U.S.) was the first such compound approved for clinical use in 2007. The discovery of CCR5 antagonists with the requisite drug-like properties has been very challenging and whilst Maraviroc has excellent anti-viral activity it has only moderate bioavailability in humans (23%) and is dosed twice daily in the clinic. In this interesting paper from the Pfizer group they detail their efforts to identify follow-up compounds with improved oral absorption (ideally to allow once daily dosing) and activity against maraviroc-resistant virus.Earlier work had shown that building in the desired properties whilst minimizing hERG inhibition was the key challenge to improving the oral absorption of CCR5 antagonists like (1) (MW=513, logD=1.9). The team opted to retain the central tropane group which, in earlier analogues, had demonstrated excellent antiviral activity in the absence of issues such as CYP inhibition, and planned to optimise the C- and N-substituents to achieve the required follow-up profile. In compound (1) the difluoro moiety was found to be poorly tolerated by hERG. Furthermore, truncating the amide was also found to reduce the affinity for hERG and provided lower molecular weight compounds with good activity. Caco-2 permeability could be improved by replacing the 1,2,4-triazole with other heterocycles (this being attributed to changes in dipole moment and the consequent extent of hydration of the heterocycle) and compound (3) (MW=416) was identified as an attractive lead for further optimisation.
Whilst the benzimidazole in (3) introduced a degree of hERG inhibition it was felt that introduction of polarity on the right-hand side, to disrupt putative lipophilic interactions (from modelling work) with the hERG channel, provided the opportunity to identify novel antagonists with the appropriate properties and with sufficient structural variation (relative to (1)) to lead to differences in the viral resistance profiles. The imidazopiperidine group was chosen as an alternative to the benzimidazole in (3) and, based on earlier SAR, analogues of this type with logD in the range 1.5-2.5 were targeted. The paper then goes on to detail the emergent SAR of this series and the process that led to the identification of 3 lead compounds (all lying within the targeted logD range). Interestingly, the meta-fluoro substituent (see (41f) below, logD=2.0) was found to be crucial to reducing hERG affinity and whilst a number of substituents (R below) were tolerated on the imidazopiperidine, the nature of this group and it’s spatial disposition were also found to be key to controlling any hERG liability (with some substituents demonstrating orthogonal SAR in the primary gp160 assay relative to the hERG assay) as well as to obtaining good caco-2 permeability.
All 3 lead compounds demonstrated in vivo pharmacokinetics in preclinical species which offered the potential for complete absorption and acceptable clearance rates in man (i.e. the possibility of once daily dosing in the clinic). The lead compounds were also shown to have antiviral activity against HIV isolates known to be resistant to (1). A superior preclinical safety profile led to 41f (PF-232798) being progressed to clinical studies. Phase I data was favourable and a Phase II trial was completed in 2009. Data from this latter trial is promised in due course.