Riding pharma’s R&D rollercoaster

The pharmaceutical industry's R&D productivity has undergone a slump in recent years. Although the number of new medicines reaching the market picked up in 2012, pharma's annual output has flatlined for the past decade. Developing new medicines is becoming an increasingly expensive business. At the same time, the regulatory environment has become more rigorous.

The European Medicines Agency (EMA) recently introduced a new, three-pronged approach to the management of adverse reactions, while the Food and Drug Administration (FDA) is building an active surveillance system, called Sentinel, to oversee the safety of all medicines on the US market.

While the tide may be starting to turn, the situation is still very serious. Only a few fully integrated organisations will remain by 2020. Some companies will be acquired and stripped of their assets. Others will separate their R&D from their revenue-generating activities to reduce risk and unlock shareholder value. There are two aspects to the productivity problem: one scientific, the other managerial. In this article, the focus is on the science side.

Some ugly facts

Between 2002 and 2011, the pharma and biotech sectors spent nearly $1.1trn on R&D. What has this investment produced? Clearly, new medicines originate in many countries, but most of them are eventually launched in the US. FDA approvals are thus a reasonable proxy for the industry's overall output over time.

In the ten years to 2011, the FDA approved 308 new molecular entities (NMEs) and biologics. Given how much the industry has invested in R&D during the same period, that means the annual average cost per approved molecule was in the range of $2.3-4.9bn. In recent years, costs have risen relentlessly. During 2007-12, the average cost per molecule was $4.1bn, nearly 50% more than during 2002-06.

While there was some improvement between 2010 and 2012, it is questionable whether this is a sustainable trend - only time will tell. Industry majors realise they need to get a handle on productivity costs. But while many industry leaders are all trying to do 'more with less', there is no sign of a big surge in productivity. Research from KMR Group shows that the number of NMEs required to achieve one new drug approval is increasing at every stage of development. In 2007-11, it took an average 30.4 NMEs in preclinical development to secure one approval, compared with just 12.4 NMEs in 2003-07.

Frontloading the R&D process

So, what accounts for pharma's recent poor performance in R&D? One of the many arguments put forward is that the industry is now focusing on more complex diseases involving novel targets. That is true, but it is by no means the whole story.

The most important - and arguably hardest - decision a pharma company makes during the R&D process is which target or mechanism to focus on. It usually starts by collating numerous sources of evidence, drawn largely from the public domain, to create a hypothesis about the role of a mechanism in a given disease.

But there is rarely compelling data validating the mechanism's role in the underlying pathophysiology of the disease. And, even if there is, the data may be incorrect. Moreover, very little is known about the feasibility of intervening pharmacologically or demonstrating the desired clinical effect at this stage of research. If a pharma company backs the wrong research project, it could end up eight or nine years later with a failure that costs more than $1bn.

Investing more money early on in understanding the molecular basis of a disease and the role a particular mechanism plays reduces the risk of losing a lot more money further down the line. Yet, on average, pharma companies spend only 7% of their R&D budgets on target/mechanism selection and validation - a fraction of the sum they spend on clinical trials. Here, again, there are signs of a shift beginning, with industry majors committing to focus more on understanding the disease process.

Lighting the way

Genetics and genomics could play a big role in improving future productivity. By the end of 2011, there were 1,068 published genome-wide association studies. Such studies normally compare the DNA of patients suffering from a specific disease with a control group to identify the alleles associated with that disease. They cannot alone determine which genes are causal, but, by covering the entire genome, they can point to new regions for research and validate or rule out mechanisms in human populations without conducting clinical trials.

At present, genomics plays a relatively small role in the lab. It is estimated that the industry spends just $6bn a year - less than 7% of its total R&D investment in 2011 - on such research. But this is an area of study that is advancing very rapidly. Witness the fact that scientists have recently identified approximately four million gene switches in DNA that was once dismissed as junk. So, by 2020, pharma could be investing as much as 20% of its R&D budget in genetics and genomics, for discovering and commercialising new drugs.

It will also draw on a growing number of population-based studies with well-characterised phenotypes and, by 2020, online genetic testing companies will be another key source of information. Together with better biomarker screening technologies and cheaper genomic technologies, this will help pharma decipher the messages encoded in our genes.

It won't be easy and some of the insights the industry is unearthing will require years of multidisciplinary research before they can be used to create new medicines. And, since most companies are organised by disease area or indication, as well as being geographically scattered, they will need to remove the barriers to information sharing.

But despite all the challenges, pharma will be in a much better position to dissect the molecular basis of many conditions by 2020. It can then start developing targeted medicines to treat them, much as it is now doing with cancer. To get there, companies will need to become more selective about the therapeutic areas they cover. A lot of companies try to investigate numerous diseases and spread themselves very thin. It is better to focus on a few therapeutic areas, prune your portfolio accordingly and bolster your expertise by hiring or collaborating with the best people in your chosen fields of research.

Many pharma executives now also recognise the merits of open innovation. The industry majors are actively linking up with universities. Some are also joining precompetitive discovery federations, where public and private institutions pool resources to overcome bottlenecks. A number of historical rivals have also entered into co-development pacts.

This pattern will continue and, by 2020, most pre-competitive challenges will be tackled collectively. But collaboration is only part of the equation. The other part is specialisation: focusing on a select range of diseases, rather than trying - and failing - to cover all the bases.

Cutting to the chase

It is equally important to devise a clear path to clinical proof of concept for all compounds entering development and test this in humans as soon as possible, using the best tools for selecting subjects and end points. Biomarkers have a significant contribution to make here by narrowing down the subset of patients on whom a molecule should be tested and exposing likely failures more rapidly, so they should be treated as an integral part of the route to market, rather than being bolted on in late-stage development.

Some companies might also want to consider novel forms of testing, such as n-of-1 trials (where a single subject receives two treatments in an alternating fashion) and in-life trials. Most organisations still focus on performing traditional randomised controlled trials, believing that is what the regulators favour. But the EMA has explicitly stated that it is prepared to consider evidence from pre-planned sequential n-of-1 trials. And the FDA recently approved Xarelto for the prevention of strokes in patients with atrial fibrillation on the basis of a large in-life trial. Both forms of testing can provide insights that traditional trials cannot yield.

Turning to new treatment types

Scientists are also making significant progress with new forms of medical intervention. Novel drug delivery technologies could reduce non-compliance, while new vaccines and regenerative medicine could provide a way of preventing or curing certain chronic conditions.

The realm of man/machine interfaces is proving equally exciting; for example, scientists at Massachusetts Institute of Technology are developing a 'smart tattoo' with a nanoparticle ink that can track glucose levels in patients with diabetes. And researchers at China's Chongqing University have built a prototype temperature-controlled drug-release system using titanium nanotubes covered in a layer of hydrogel. Both have obvious uses in pharma.

With regenerative medicine, it may eventually be possible to do away with some biomechanical aids altogether. Several tissue-repair products, such as ChondroSelect, are available on the market. But that is just the first step. The second is tissue replacement, using 3D bioprinters to print living tissue with 'ink' derived from human cells.

Some have already made headway in this field. Replacing damaged neurological tissue and entire organs is, of course, the end goal - and, despite the enormous challenges, it is less of a far-fetched fantasy.

Return of the blockbuster

So, conventional pharmacological agents are not the industry's only hope. Options are getting steadily wider and, by 2020, there will be far more diversification. Many of these new therapies will require much more complex development, manufacturing and distribution processes than those used to produce conventional medicines. Yet they will also generate enormous clinical and commercial value.

A prophylactic vaccine for a common chronic condition or stem cell therapy that cures a neurodegenerative disorder won't earn revenues from repeat prescriptions. But it will command a higher price because it provides a permanent solution. Such products will be tomorrow's blockbusters.

Keeping an open mind

Whatever diseases and forms of medical intervention a company decides to focus on, and whatever methods it chooses to discover and develop new treatments, one thing is vital: keeping an open mind until clinical proof of concept. It is always painful to see a beautiful hypothesis slain by an ugly fact. It is even more painful when that hypothesis has consumed a lot of money.

References available on request.