The end of drug discovery?
Half a century ago, in the drug industry's golden era, we were bestowed with countless pills to lower blood pressure, control blood sugar and get rid of infections. But today it costs about $1bn to bring a new medicine to market, a process that can take 15 years.
The industry faces multiple crises as budgets are squeezed and tough scientific challenges loom.
Academics and patients' groups are concerned we will not have the drugs necessary to treat future ills.
Last year, the World Health Organization's director general, Margaret Chan, warned that the world was heading for a "post-antibiotic era".
She raised fears that many common infections might no longer have a cure and could once again "kill unabated".
Major problems, Dr Chan suggested, included growing resistance to antibiotics.
And as the population ages, many of us will battle with neurological conditions such as Alzheimer's and Parkinson's, where new drugs are also lacking.
But the old model of drug discovery is no longer working well.
This so-called blockbuster method - which commits large sums of money to finding a drug that promises to treat a huge proportion of the population, and generate swathes of cash to cover other experimental losses - can no longer sustain the industry.
The model is inefficient. Nine times out of 10, molecules which show promise at the early stage are destined for failure at later trials.
Pharmaceutical companies are also under pressure as patents for some of their most lucrative drugs expire.
Manufacturers face an uphill struggle with science too. Some argue that in its most successful years, the big pharmaceutical firms picked off the "low-hanging fruit" - the compounds that were easiest to discover and turn into drugs.
Chas Bountra, professor of translational medicine at Oxford University, says: "We don't understand enough about human disease or enough about the mode of action of existing drugs. Take paracetamol for example. We all take it but we don't know how it works, we don't know what the site of action of paracetamol is.
"So if we don't know that, how can we design better, more superior molecules?"
But finding new targets - proteins in the body that could be modified by a drug to help relieve disease - is like a lottery, he says.
"There are more than 20,000 proteins in the human body, and any of those proteins could be a target for drug discovery. Our ability, even in 2012, to be able to say one protein is going to work in this patient group or subset of patients is appalling."
Alongside many other people working on drug development, he points to another difficulty that threatens the process - duplication.
"We don't publish our failures, or if we do publish them, we publish too late. As a consequence other academics, other companies, carry on working on the same target, carry on wasting resources and careers and exposing patients to molecules that have a chance of failure," Prof Bountra says.
But Prof Patrick Vallance, president of pharmaceuticals research and development at GlaxoSmithKline, thinks the industry is becoming more open: "One of the things we have done is to be very public about when things go into the clinic.
"And we publish our results - we make our protocols available to reviewers when papers are submitted. Those sorts of things help to start better visibility of what is in the pipeline."
GlaxoSmithKline has put 13,500 structures which have the potential to kill malaria into the public domain.
Prof Vallance says the firm has done this because it is a "tough problem to crack".
He adds: "Why not let everyone have a look at those compounds and see if they can think of something smarter than we are doing, and see if they can group them in a certain way or spot a pattern we have missed?"
But, he argues, we have to be realistic about this. "There will always be a degree of duplication, because that comes with the competition."
Prof Bountra's group at Oxford has also made a decision to publish all their successes and failures.
Prof Paul Workman at the Institute of Cancer Research points to another pressing issue: "With the problems of the financial crisis there has developed something of a vacuum that many of us describe as 'the valley of death'.
"This is the valley between basic research and innovation on the one hand, and patient benefit and commercial success on the other, with this chasm in between into which there is a lack of funding and a lot of failure."
One possible solution is for charitable foundations such as the Wellcome Trust to offer funding to fill this gap.
Prof Workman, on the other hand, works on a model where many processes of drug discovery and development come under one roof.
"We go all the way from discerning a new biological concept to developing the chemicals that block a certain pathway or gene function, and then all the way into clinical trials," he says.
He points out that they don't do the whole thing alone, often partnering with smaller and larger companies in later stages.
Prof Workman is optimistic that the science is moving forward, and bringing new chances of success with it: "The science is taking us in the opposite direction to blockbuster drugs, to personalised medicine. You would identify the patient who would benefit from what drug according to a gene test.
"A relatively small number of patients will benefit but they will benefit extremely well."
What all these newer models of drug development have in common, it seems, is collaboration - perhaps presenting a cultural change for an industry that many people believe has had competition at the centre of its working.
But, says Prof Vallance, while there is need for more collaboration in the early stages of drug development for some of the more complex diseases, that will turn into competition at a later point.
"There will be fierce competition to get the best drug first, to make sure you do the right trial and show you've got the best medicine."
You can hear more on Radio 4: The end of drug discovery? , Tuesday 22 May at 20:00 BST. The programme is repeated on Sunday 27 May at 17:00 BST.