In June 2004, AstraZeneca trotted out its new anti-coagulant Exanta with great fanfare. The first oral blood-thinner introduced in nearly 60 years, it was the result of more than 20 years' R&D. Nineteen months later, the company pulled the drug from 12 European countries and halted its efforts to introduce it elsewhere. New clinical safety data had shown that Exanta could cause severe, sudden-onset liver injury. In 2005, the drug earned the company revenues of precisely US$575,000 on its multimillion-dollar investment.

The story is all too familiar. Promising compounds move into costly clinical trails, or beyond, only to be scuppered by safety problems that evaded detection in preclinical testing. With the Exanta experience fresh in its rear-view mirror, AstraZeneca is one of three drug companies — the others are GlaxoSmithKline and Roche — who in October joined with five agencies of the UK government to launch Stem Cells for Safer Medicines (SC4SM), a novel public–private consortium. Structured as an independent company, its mission is to use human embryonic stem (ES) cells to catch drug-safety problems in petri dishes, not in patients. It's also hoped that the technology developed will eventually replace some of the millions of animals currently sacrificed in preclinical testing.

Disease-specific stem cells are already available for studying single-gene maladies such as cystic fibrosis and fragile X syndrome1. And the possibility of deriving stem cells from people with genetic variations relevant to a particular disease or drug came closer in November, with the publication of papers showing that cells taken from adult humans could be genetically engineered into a pluripotent state2. But even more difficult than obtaining cells capable of differentiating into specific cell types is finding robust and scalable techniques to make them do so reliably (Table 1).

Going for liver and heart

“We need to be able to predict clinical safety in our discovery research,” says Jan Törnell, vice-president of translational science at AstraZeneca and a stem-cell expert based at the company's laboratory in Mölndal, Sweden. “The big thing for us is that when we get into patients, we want no surprises.” Indeed, unpleasant surprises at the clinical stage are the biggest contributor to the cost of bringing a new drug to market, estimated to be as high as $1.7 billion.

AstraZeneca is hoping that the SC4SM project will make unpleasant surprises a lot less likely. Starting with a modest budget of just over £1 million (about $2 million) in its first, pilot year, with three-quarters coming from the UK government, the consortium's first goal is to spur ways of creating functioning liver cell lines from stem cells. The reason is simple: as the body's garbage processor, the liver is particularly susceptible to the toxic side effects of the drugs that it breaks down. Access to human liver cells in vitro would thus be a boon for drug-company scientists, who currently have to rely on a combination of animal toxicity studies and scarce, variable, and often diseased human liver samples. And while human ES cells can be coaxed to make neurons with a fair degree of reliability, similar reliability is proving much, much harder to achieve for other cell types.

Table 1. Selected companies working on stem cells for drug screening. Some of these companies are also working on therapies and a variety of other cell lines. Also, scientific reagent companies such as Invitrogen and Millipore sell stem cells to researchers.

“If we could crack the liver cell differentiation problem, it would really move things forward for everybody” Philip Wright, chief executive of Stem Cells for Safer Medicines

“If we could crack the liver cell differentiation problem, it would really move things forward for everybody,” says Philip Wright, chief executive of SC4M, who is also director of science and technology at the Association of the British Pharmaceutical Industry based in London.

The consortium may also focus in future on the cell type that's next on the industry's wish list: cardiomyocytes. New heart medications are major targets for most firms, and drug discovery programmes could be hugely helped if compounds could be screened against human heart cells. For example, some drugs interact with a particular ion channel in heart cells in a way that can cause rare, fatal reactions. And in the post-Vioxx era, the effects of all drugs on the heart will be closely scrutinized by regulators.

Pharmaceutical companies already run batteries of tests aimed at detecting cardiovascular side effects, but that arsenal lacks a critical tool. “The one they can't use because they can't get a reliable source is human adult cardiodmyocytes,” says Alan Colman, executive director of the Singapore Stem Cell Consortium.

Meeting of minds

Considering that AstraZeneca spent nearly $4 billion on R&D in 2006, the companies' contributions of £100,000 each in the project's first year doesn't seem much. But experts say that, as important as the money — which the companies are expected to add to during the five-year life of the project — is the fact that SC4SM's launch signifies a meeting of minds among big pharma. The firms have made a decision to pool resources and expertise to hasten a largely outsourced quest for tools that won't generate revenue in themselves but which should ultimately prove invaluable for both safety testing and drug discovery. Each company is also providing four to seven high-level scientists for consortium work. These include toxicologists, stem-cell scientists and experts with experience of high-throughput screens.

It makes sense for drug companies to collaborate to avoid duplicating research, says Colin Pouton, a professor of pharmaceutical biology at Monash University in Melbourne, Australia, who works on the use of differentiated stem cells in drug discovery3. “It's very much pre-competitive research. The pharma industry doesn't particularly have a desire to make money out of it; they just want to use it,” he says

The consortium's launch is a sign that large drug companies are clamoring for products that are still far from perfected, says Tim Allsopp, chief scientific officer at Stem Cell Sciences, an Edinburgh-based firm that provides customized stem cells for basic research and drug discovery. The consortium's financial commitment, he says, means that companies like his can “actually get some bandwidth in terms of maturing this technology”.

Swedish stem-cell company Cellartis, together with an unidentified partner from the UK, responded quickly to SC4SM's first call for proposals, which closed this December. The consortium's commitment indicates that “the pharmaceutical industry really are the future customers that we hope them to be,” says Petter Björquist, a cell biologist and former AstraZeneca researcher who now directs Cellartis's programme to derive liver cells from human ES cells.

Björquist adds that the savvy of the big drug companies is a boon for scientists trying to derive practical applications of stem-cell technologies. “They are the end users, so they see even more clearly than we do how the cell material should look. They know if they are going to run 96-well plates or 384-well plates. They know if the media we are using are compatible with large-scale screening,” he says.

Wright agrees that that expertise is key to the consortium's value. “What a lot of the [smaller] companies haven't got is access to a large number of safety scientists across the industry,” he says. “We can help them produce the standards that can say: we haven't just got a hepatocyte-like cell, we have hepatocytes.”

Indeed, SC4SM has already started to map out the parameters that it sees as essential for ensuring that differentiated cells are physiologically relevant, such as the range of markers that bona fide liver cells generated from human ES cells should express.

And the smaller companies are recognizing that the consortium's launch brings new opportunities. Within two months of announcing its first call for proposals in October, the consortium had received 12 expressions of interest that resulted in six applications for funding.

The consortium approach, although unusual, is not without precedent. In the early 2000s, the Consortium for Metabonomic Toxicology (COMET) — five major drug companies working with researchers at Imperial College, London — used NMR spectroscopy to study reactions to 147 toxins and treatments in rat and mouse urine and blood4. With these findings, the consortium developed a database for the companies to use in-house to predict kidney or liver toxicity in candidate drugs.

In the United States, the Critical Path project, launched in part by the US Food and Drug Administration, has initiated a data-sharing project with for-profit companies to identify the most accurate toxicity tests. Given the timelines necessary to bring a drug to market, the contribution of such projects is difficult to assess. But some drug companies at least are voting with their feet: a second COMET project is now being launched to probe drug safety and toxicity mechanisms has strong industry backing.

But the leap into human stem cells is technologically daunting, and SC4SM's five-year timeline may be ambitious. Nevertheless, the role of stem cells in drug screening is coming to be seen as inevitable, not just for drug safety but also for small-molecule drug discovery. “That's not to say that, in ten year's time, stem cells will have totally replaced the use of animals in testing,” says Allsopp. “But one would hope that it would be a platform technology in routine use by all the major pharma companies.”