Fairly early on in our exploratory project we had identified one particularly interesting compound that demonstrated very good in vitro potency and oral in vivo efficacy comparable to DuP-721, and was found to have similar PK properties in the rat. This was the indanone oxazolidinone ( + )-U-82965 (5, PNU-82965),24,25 one of two cyclic ketones we had targeted in order to explore the consequence of restricting the rotational conformers of the DuP-721 methyl ketone, via constraint in a five- or six-membered ring. Not long after U-82965 was in hand, we became aware of sketchy information that DuPont had dropped their oxazolidinone program, putatively on the basis of toxicological findings of their clinical candidates.20'25 This surprising news obviously represented a very critical juncture for our project, for should we have been unable to rapidly identify an active oxazolidinone that could be differentiated from the DuPont lead (i.e., with a significantly improved toleration profile), our continued work in this area would likely be short lived.
As will be seen, it would turn out to be most fortuitous that at that time we had identified U-82965. It was no less auspicious that we also happened to establish a working relationship with Richard Piper, an Upjohn pathologist, who offered excellent leadership and assistance in developing a strategy for potentially moving the project past this significant hurdle. Furthermore, we also benefited considerably from the expert opinion and guidance on this matter and one concerning the potential for oxazolidinone inhibition of monoamine oxidases, which was provided by our resident infectious diseases clinical expert, Donald Batts.
Piper graciously volunteered the efforts of his laboratory, staffed by his associates John Palmer and Thomas Platte; this was also enabled under the Upjohn 10% free-time policy. Piper's laboratory proceeded to design and carry out a protocol for an exploratory 1 month duration comparative toxicology study in the rat, with a side-by-side evaluation of ( + )-U-82965 and ( + )-DuP-721. John Greenfield conducted the PK determinations that provided decisive data supporting this comparative study, by establishing that both compounds had similar exposure levels upon oral administration in the rat.
The drug-sparing 30-day toxicology protocol designed by Piper required that my laboratory need only prepare 8-12 g of each test compound. The use of such limited drug quantities was feasible, as the protocol would involve the dosing of only three rats per sex. As the rationale of the study was to enable the expeditious identification of oxazolidinones having at least a 10-fold therapeutic index in the rat, the test compounds were dosed orally, twice daily (b.i.d.), at a dosage level 10-fold the ED50 (the effective dose (mgkg_ 1) that protected 50% of the mice from death after an injection of a lethal dose of S. aureus). As the comparative study progressed, the toxicologists reported several distinct toxicological findings that were readily apparent by clinical observation in the group of rats dosed orally with ( + )-DuP-721 at 100mgkg_ 1 day_ 1. Those findings included alopecia, severe anorexia, ataxia, and the death of one of the six animals; another two animals observed in a moribund state were euthanized prior to the end of the study. On histopathological examination there was evidence of bone marrow toxicity and terminal circulatory failure.26 In contrast, we were elated to find that all six animals treated with U-82965 at the same dose fared very well; the compound was tolerated very well over the course of the 30 day trial.26
Following our filing of the Upjohn patent application on the promising U-82965 compound, we learned of a subsequently published DuPont US patent27 (which had been filed earlier than our application) that claimed this indanone compound. Hence, while this was disappointing, we nevertheless had gained through the study of U-82965
invaluable insight and knowledge concerning the vastly different toleration in a rodent species of two closely related ketones - one cyclic, the other an acetophenone - which clearly established that an STR existed for this series.
The outcome of this study had a major impact on our project - one that would definitively play a significant role in our eventual successful discovery of eperezolid and linezolid. The importance of Piper's role in that success cannot be over-emphasized. First, a principal initial objective of the study was met in substantiating the rumored toxicological problems with the lead oxazolidinone. Second, the protocol had identified an equipotent analog of DuP-721 of close structural similarity with a clean toxicological profile. Third, Piper's study design had clearly demonstrated the utility of the drug-sparing protocol, thereby allowing us to plan on using this for expedited subsequent lead oxazolidinone toxicology evaluations. Finally, it shaped what would become our enabling strategy, allowing us to proceed. That strategy pivoted on the need to establish an understanding of STRs as a means of acquiring confidence in the selection of our advanced leads for further progression to drug candidate status. The necessity of conducting multiple, early, multiday toleration studies presented obvious additional hurdles for our team, obstacles that were considerably above those normally encountered in a typical SAR-driven program at that time. The strategy we instituted may arguably be one of the earliest of the few projects in the industry that succeeded in delivering a first-in-class drug to market, from a research program heavily reliant on early toxicological evaluation as a means to vet numerous, promising lead compounds.
Another series of compounds we focused on following the successful outcome with U-82965 would add substantially to our understanding of the STRs, and led to the identification of structural features that would eventually be incorporated into our drug candidates. We had chosen to examine various fused-ring heterocyclic oxazolidinone derivatives22'23 for the explicit purpose of replacing the indanone ketone of U-82965, while retaining a five-membered benzo-fused ring. This interest included a series of active indazoles22 and 5'-indolines, among others. With the principle of locating a carbonyl with an orientation proximate to that of the ketone in U-82965 or DuP-721, we prepared a series of amides attached to a 3-(5'-indolinyl)-5-acetamidomethyl-2-oxazolidinone core. This series of 5'-indolinyl amides23 had superior activity to the isomeric 6'-indolinyl analogs.
Two of these active 5'-indoline amides, U-97456 (6) and U-85910 (7), were tolerated extremely well in the 30 day rat toxicology protocol.28 Those results thereby established for the first time that a nitrogen atom substituted at the oxazolidinone phenyl para position could lead to compounds with improved toleration profiles. The thiophene amide 7 was of interest in that it was determined by James Kilburn and Suzanne Glickman at the Centers for Disease Control and Prevention (CDC) to have potent in vitro activity against Mycobacterium tuberculosis.29 While the Gram-positive antibacterial potency of U-97456 was slightly below our targeted profile desired for a clinical candidate, the excellent safety profile of this compound and 7 laid the foundation for the eventual synthesis of many other oxazolidinone series similarly substituted with nitrogen-containing heterocycles at the ^ara-phenyl position. This includes our first drug candidate, the piperazinyl fluorophenyl eperezolid, and the morpholinyl fluorophenyl analog linezolid.
Was this article helpful?