Regarding the Pfizer Sandwich closure

Pfizer is to close its historic research facility in Sandwich, Kent. It would be difficult to overstate the scale of this closure, and the magnitude of its impact on the pharmaceutical research community in the UK. I've been to the Sandwich labs. They are a sprawl of buildings with the population the size of a small town. They have their own (subsidised) Starbucks.

I'm not really qualified to talk at any great length about the state of the pharmaceutical industry - I will defer to whatever Derek Lowe has to say on the subject - but I've had a thought I'd like to share. Pulling a successful drug from the chemical space has never been easy, with success rates in the just-double-figures or lower. Never the less, the drug business built a reasonably efficient and quite profitable machine to tease out these leads.

The pharma system is built on blockbusters, big money makers that have to support a handful of complete failures or niche drugs, and in the glory days of the 1980s and 1990s it appears that quite enough were arriving for the whole operation to come out in the black. More recent history, however, is littered with high-profile failures. It's not so much that the well of molecular possibility has gone dry - it was never gushing - but that it's gone from recalcitrant to downright surly.

Against this background it seems one approach to success might be to cut down the failures. To get to my idea, it seems that university and government biotech spin-off companies, having performed much of the preliminary study and trimmed down the possibilities under the umbrella of pure science, have a better chance of then bringing a commercially and clinically successful molecule to market than the pharmaceutical business, which traditionally starts from a broader search.

Given the ongoing economic crisis and the longstanding pressure from on universities to balance their books, might we be on the brink of a sea-change in the way medicines are created? I am probably overstating something that's comically obvious, but there you go.