Pfizer Sandwich Laboratories: where drug metabolism first met drug discovery

Sandwich may just have been the centre of a revolution. I like to think it was and hope that it will be remembered favourably by those that witnessed its work, either from within or amongst its many collaborators (maybe even its competitors). Twenty five years ago drug metabolism departments were aligned with safety evaluation. The priority was the conduct of radiolabelled studies in animals and humans examining the excreta and later the circulation for metabolites of development candidate drugs. The work was aimed at demonstrating that the species used for safety evaluation were the correct choice, even though the choice was extremely limited. Sandwich almost uniquely reported drug metabolism into Drug Discovery. The Toxicology Department was about 300 mile away in the picturesque French town of Amboise. With this reporting relationship it was natural that the Department would begin to use its skills to support Discovery. At the time an uncharted area but the initial thrust was to incorporate bioanalysis and plasma concentration data into pharmacological screening programmes. From some of these programmes major drugs were discovered and developed, such as fluconazole and amlodipine, which succeeded clinically, in part because of excellent and superior pharmacokinetics to those agents in the same or similar pharmacological class. Fluconazole provided 100% absorption without food effects, a 30-hour half life and once a day administration in the area of fungal infections compared to existing agents, used at the time, which were mainly topical or lipophilic agents with poor dissolution. Amlodipine is described in the accompanying preface by Simon Campbell (Campbell, 2012).

Already Sandwich reached a dilemma. Extrapolation of the pharmacokinetics to man from animal data of a renally cleared antifungal (fluconazole) or a 48 hour half-life (in dogs) anti-hypertensive (amlodipine) was straightforward, but there was always the spectre of species differences hanging over other projects with less clear endpoints. How could we add a human dimension to our discovery work? In vivo techniques gradually were therefore, augmented with in vitro experiments. These initially were hampered by the shortage of humanised reagents. For some time we had access to a single liver and for a period our human predictions were based on this. Fluconazole was a marvellous drug and remains the drug of choice for candidemia, except for infections with one of the resistant species such as Candida krusei and some strains of Candida glabrata. It however did not have robust activity in severe invasive aspergillosis. The laboratories at Sandwich therefore aimed for a new broad spectrum antifungal, but trying to retain as many as possible of fluconazole’s attributes. The in vitro pathogen testing indicated that voriconazole had all the antifungal properties we wished for, but testing in infected animal models was hampered by the drug’s rapid metabolic clearance in rodents. The then existing paradigm would have rejected the drug, but our drug metabolism studies using human and rodent liver(s) indicated much greater metabolic stability in man. These results were confidently put to senior management and voriconazole entered drug development. Only later, as more human liver samples became available, did we discover our single liver seemed incapable of catalysing any metabolism of any compound except p-nitroanisole, our positive control. Voriconazole was more ‘stable’ in humans (luckily) due to probably a lower rate of catalysis heavily influenced by complex time- and concentration-dependent kinetics. Voriconazole has proved since its launch an extremely valuable drug in the treatment of serious fungal infections.

It seems hard to imagine drug discovery and development occurring without the knowledge we now have on CYPs: the various isoforms present in animals and man, their substrate structure relationships, and their almost pivotal role in many drug -drug interactions. Sandwich played its own role in this process, publishing possibly the first analysis of the substrate structure relationships, producing and publishing active site models, beginning to analyse and comment in the literature on the structures which were most likely to lead to potent inhibition of the enzymes etc. Generally there was a growing understanding that protein binding displacement was not the cause of drug-drug interactions but the major factor was inhibition of the enzymes of drug clearance. In a large organisation, as Pfizer was becoming, it is very difficult to reach all parts of the organisation or check all the information going out. We were approached by the Clinical Leader of the Darifenacin project concerning the clinical investigators concern with the drug being highly protein bound and the worry on protein binding interactions. We produced some elegant scientific briefing notes as to why this wasn’t a worry but despite their distribution the worry continued. Further, even more detailed and referenced briefing notes were produced for distribution and incorporated into the Clinicians Brochure but to no avail. By chance the whole Brochure was examined (rather than carefully proof reading the Drug Metabolism Section) and point one of Precautions was the need to avoid con-meds with high plasma protein binding due to darifenacin being highly bound itself. The statement inserted in an early draft (and probably copied from a very dated template) was never edited out despite all our efforts to allay the problem. That period, however, was a time when Clinical Pharmacology and Drug Metabolism moved to being a seamless partnership. Phenotyping, genotyping, clinical drug-drug interaction studies were planned in unison between the departments as the human enzymology emerged. PK/PD provided a rich mutual ground with scientists spending periods working in either location to ensure best practices, vocabulary and beliefs were common.

In 2004, Drug Metabolism had a name change to Pharmacokinetics, Dynamics and Metabolism (PDM) partly to reflect some of these changes, partly to reflect we had left Mother Discovery and like a fledgling infant became our own global line. Pfizer was on a trajectory of getting bigger by acquisition and not organic growth. However it was unable to maintain the large number of research campuses in its remit. The number of research sites became a continuing contention for the last decade and Fresnes, Ann Arbor, Amboise, Nagoya, St Louis became harbingers of the eventual demise of Sandwich.

For a period the programmes at Sandwich addressed drug targets that meant the compounds were moderately to highly permeable. Combination of human in vitro (now with a more robust liver bank) and animal pharmacokinetics allowed confident prediction. Drugs such as eletriptan (for migraine) and sildenafil (Viagra) were optimised for human PK based on these studies. Within some targets the relationships between in vitro enzymology and in vivo data began to become strained. Compounds, which were metabolically stable, showed high clearance in vivo. The “Rise of the Transporters” was underway and it correlated with targets that required lipophilicity together with high hydrogen bonding potential. These compounds were low to moderate permeability and allowed the influence of transporters to become dominant in the flu into and out of cells.

The papers in this special edition reflect Sandwich as it closed. Still a vibrant, innovative centre of Drug Metabolism. Hungry to discover drugs, hungry to discover more about Drug Metabolism and eager to share its learning with fellow scientists. The breadth of contribution ranges through transporters, MIST, inhaled drug delivery to the abstract (Why do…?). Sandwich has had many collaborators outside in the Academic Community and I think most (if not all) regarded it as a true partnership. Working on the science, teaching the science, bringing on the next generation was a centrepiece to Sandwich’s culture. Sandwich Drug Metabolism (PDM) no longer exists, but its work here will live on.

There is a programme in which retired professional football (soccer) players from a particular team reminisce on their careers and the great highlights. The programme is called “Time of our lives”. For those who worked in Sandwich Drug Metabolism I like to think it was.