Pharmaceuticals & Life sciences

Pharma 2020: Virtual R&D Which path will you take?

Pharma 2020: Virtual R&D 1 TableTable ofof contentscontents

Innovation essentials 1 Getting to know ourselves 4 Making research more predictive 6 Expediting development 9 Collaborating to bring treatments to the market 15 Imperatives for change 18 Acknowledgements 19 References 20 Contacts 21

Pharma 2020: Virtual R&D – Which path will you take? is the second in a series of papers published by PricewaterhouseCoopers exploring the future of the pharmaceutical industry. Published in June 2007, Pharma 2020: The vision – Which path will you take? highlighted a number of issues that will have a major bearing on the industry over the next 13 years and outlined the changes we believe will best help pharmaceutical companies realise the potential the future holds to enhance the value they provide to shareholders and society alike. This paper explores the opportunities to improve the R&D process. It proposes new technologies will enable the adoption of virtual R&D; and by operating in a more connected world, the industry in collaboration with researchers, governments, healthcare payers and providers, can address the changing needs of society more effectively. Innovation essentials

Pharma is at a pivotal point in Seven major trends reshaping the pharmaceutical marketplace its evolution. As we* indicated The pharmaceutical marketplace is changing dramatically, with huge ramifications in “Pharma 2020: The vision” by for the industry as a whole. We have identified seven major socio-economic trends. PricewaterhouseCoopers, published in June 20071, the social, demographic The burden of chronic disease is clinical advances render previously fatal and economic milieu in which the soaring. The prevalence of chronic diseases chronic. The self-medication industry operates is undergoing huge diseases like diabetes is growing sector is also increasing as more changes (see sidebar, Seven major everywhere. As greater longevity forces prescription products are switched to trends reshaping the pharmaceutical many countries to lift the retirement age, over-the-counter status. The needs marketplace). These challenges have more people will still be working at the of patients are changing accordingly. been compounded by the dearth of point at which these diseases start. The Where treatment is migrating from the good new compounds in its pipeline. social and economic value of treatments doctor to ancillary care or self-care, for chronic diseases will rise accordingly, patients will require more comprehensive but Pharma will have to reduce its information. Where treatment is prices and rely on volume sales of such migrating from the hospital to the products because many countries will primary-care sector, patients will require otherwise be unable to afford them. new services such as home delivery. Healthcare policy-makers and payers The markets of the developing world, are increasingly mandating what where demand for medicines is likely doctors can prescribe. As treatment to grow most rapidly over the next 12 protocols replace individual prescribing years, are highly varied. Developing decisions, Pharma’s target audience countries have very different clinical is also becoming more consolidated and economic characteristics, and more powerful, with profound healthcare systems and attitudes implications for its sales and marketing towards the protection of intellectual model. The industry will have to work property. Any company that wants to much harder for its dollars, collaborate serve these markets successfully will with healthcare payers and providers, therefore have to devise strategies that and improve patient compliance. are tailored to their individual needs. Pay-for-performance is on the rise. Many governments are beginning A growing number of healthcare payers to focus on prevention rather than are measuring the pharmacoeconomic treatment, although they are not yet performance of different medicines. investing very much in pre-emptive Widespread adoption of electronic measures. This change of emphasis medical records will give them the will enable Pharma to enter the realm outcomes data they need to determine of health management. But if it is to best medical practice, eschew do so, it will have to rebuild its image, products that are more expensive since healthcare professionals and or less effective than comparable patients will not trust the industry to therapies and pay for treatments provide such services unless they are based on the outcomes they deliver. sure it has their best interests at heart. So Pharma will have to prove that its The regulators are becoming more medicines really work, provide value for risk-averse. The leading national and money and are better than alternative multinational agencies have become forms of intervention. much more cautious about approving The boundaries between different truly innovative medicines, in the wake * PricewaterhouseCoopers refers to the network forms of healthcare are blurring. The of the problems with Vioxx. of member firms of PricewaterhouseCoopers primary-care sector is expanding as International Limited, each of which is a separate and independent legal entity.

Pharma 2020: Virtual R&D 1 Table of contents

Pharma’s traditional strategy of Figure 1: The decline in R&D productivity placing big bets on a few molecules, promoting them heavily and turning them into blockbusters worked well 50,000 60 for shareholders for many years. s 45,000 al However, its productivity in the lab 50 40,000 is now plummeting, as it switches pprov

35,000 A 40 its attention from diseases that are A 30,000 relatively common and easy to treat to 25,000 30 those that are much more complex or 20,000 20 unusual. In 2007, the US Food and Drug 15,000 Administration (FDA) approved only 19 10,000 10 No. of NME, BL new molecular entities and biologics – a R&D Spending (US$ Millions) 5,000 smaller number than at any time since 0 0 2 1983 (see Figure 1) . 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Moreover, the patents on many of the medicines the industry launched in R&D Spending NME and New Biologics Approved (BLA) the glory days of the 1990s will expire NME: over the next few years, leaving Big New chemical or biological prescription therapeutic: excludes vaccines, Pharma very exposed. US research firm antigens and combination therapies which do not include at least one new constituents Sanford C. Bernstein estimates that generic erosion will knock between 2% and 40% off the revenues of the top Source: FDA CDER, PhRMA and PricewaterhouseCoopers analysis 10 companies between now and 2015 Note: Data on R&D spending for non-PhRMA companies are not included here. (see Figure 2). Worse still, it calculates that only four of the 10 have pipelines containing products sufficiently valuable to offset these losses.3

2 PricewaterhouseCoopers Figure 2: The impact of generic erosion on Big Pharma’s revenues

“Base Revenue Estimates” (US$ Millions) Company 2008 2015 % Change Novartis $40,529 $45,714 13% Schering-Plough $20,595 $20,216 (2%) Wyeth $22,367 $20,537 (8%) GlaxoSmithKline $22,858 $20,294 (11%) sanofi-aventis $43,177 $36,186 (16%) Merck & Co $29,724 $24,428 (18%) Bristol-Myers Squibb $21,603 $16,364 (24%) Eli Lilly $20,275 $15,286 (25%) Pfizer $48,639 $34,075 (30%) AstraZeneca $31,522 $18,878 (40%)

Source: Bernstein estimates and analysis Note: These figures show each company’s “base revenues” from products that are already on the market. They exclude any future pipeline contributions

This “innovation deficit” has enormous We believe that, if the industry is to strategic implications for the industry become more innovative and cut its as a whole. Many pharmaceutical R&D costs, four features will be vital: companies need to decide what they • A comprehensive understanding of want to concentrate on doing and how the human body works at the identify the core competencies they will molecular level require, a process which may involve exiting from some parts of research and • A much better grasp of the development (R&D). But even those that pathophysiology of disease (by regard research and development as a which we mean the functional core element of their business will have changes associated with, or arising to make fundamental alterations in the from, disease or injury) way they work. They may, for example, • Greater use of new technologies to have to focus more heavily on speciality “virtualise” the research process and therapies, since most of the diseases accelerate clinical development; and for which there are currently no effective medications or cures are not amenable • Greater collaboration between the to mass-market treatments, as well as industry, academia, the regulators, reducing the time and costs involved governments and healthcare in researching and developing such providers. medicines to ensure that society can We shall discuss some of the changes afford them. we consider necessary in more detail in the following pages.

Pharma 2020: Virtual R&D 3 Getting to know ourselves

At present, when pharmaceutical predictive models and generate further Virtual vermin companies start investigating biological knowledge. The American Diabetes Association targets, they may know relatively little Bioinformatics experts aim to create a and US biopharmaceutical company about how those targets are involved complete mathematical model of the Entelos have developed a diabetic in the diseases they want to treat. The molecular and cellular components of the virtual mouse that is being used to information they possess usually comes human body – a “virtual” man – which study cures for Type 1 diabetes. from academic literature and patents, can be used to simulate the physiological Researchers can simulate the effect and is often based on animal studies, effects of interacting with specific targets, of new medicines, including different which may not be relevant to the way in identify which targets have a bearing on dosing levels and dosing regimens, which a disease progresses in humans. the course of a disease and determine on different therapeutic targets, It is generally only in Phase II clinical what sort of intervention is required biological pathways and functions. trials that companies test whether The model is based on years of data modulating a particular target with a (i.e. an agonist, antagonist, inverse from experiments on real animals, particular molecule is efficacious in agonist, opener, blocker etc.). However, but virtual animals could be created treating a disease in man. developing such a model will require a for any species for which there are monumental global effort far exceeding This helps to explain why just 11% of sufficient data. that of any similar work, e.g., the Human the molecules that enter pre-clinical Genome Project. development reach the market,4 and hence why costs per drug are so Numerous organizations are building high. The Tufts Center for the Study of models of different organs and cells, or Drug Development puts the average creating three-dimensional images from bill for producing a new medicine at the resulting data (see sidebar, Virtual 8 $868m.5 Clearly, there are significant vermin). One of the biggest problems variations, depending on the therapeutic with these models is that they are only area concerned (see Figure 3).6 as good as the data on which they are Nevertheless, given average costs and based, and little is currently known average attrition rates in each phase of about many physiological processes. the R&D process, we estimate that the Ultimately, they must also be integrated cost of conducting R&D is $454m per into a validated model in order to product.7 predict the effects of modulating a biological target on the whole system, It is now widely recognised that one and that model must be capable of the elements required to overcome of reflecting common genetic and the problems in the research process phenotypic variations. The computing and make significant advances in the power required to run such a model will treatment of disease is a comprehensive be enormous. understanding of how the human body works at the molecular level, Despite these difficulties, various together with a much better grasp of academic collaborations to create a the pathophysiology of disease. This digital representation of the human knowledge can then be used to build body are already underway. The Step

4 PricewaterhouseCoopers Figure 3: Costs per drug for medicines in selected therapeutic areas

1400 Average Cost of R&D by Therapy Area 1200 1000 800 600

US$ Millions 400 200 0

HIV Blood Cancer ological Sensory diovascular Respiratory Genitourinary Neur Antiparasitic Car Musculoskeletal

Source: Christopher P. Adams & Van V. Brantner, “Spending on New Drug Development”

Consortium is, for example, developing biosimulation is already playing a a methodological and technological growing role in the R&D process. framework for investigating the human Scientists at University College, London, body as a single complex system.9 have, for example, used computer Meanwhile, the Living Human Project modelling to simulate the efficacy of is working on an in silico model of the an HIV treatment in blocking one of human musculoskeletal apparatus,10 the key proteins used by the virus.12 and the Physiome Project aims to Similarly, when Roche was developing create a computational framework for Pegasys, its combination treatment for understanding the integrative function hepatitis C, it used computer modelling 11 of cells, organs and organisms. The to determine the optimum dose for commercial potential of virtual man different subpopulations of patients on might also attract the interest of some whom the therapy had not been tested of the largest technology providers, and in real life.13 We anticipate that this trend grid computing will deliver the resources will continue and that, by 2020, virtual required to support R&D on demand. cells, organs and animals will be widely It is probably unrealistic to think that employed in pharmaceutical research, virtual patients will be available within reducing the need to experiment on the next 12 years. However, predictive living creatures.

Pharma 2020: Virtual R&D 5 Making research more predictive

In silico methods are currently used predicting the efficacy of a molecule to design new molecules, where the in man (especially when the target is a structure of the target is known and new one). So Pharma needs a faster the interactions between the target and and more predictive way of testing virtual molecules can be modelled. But molecules before they go into man. researchers more commonly use in vitro With the advent of virtual patients, it screens to find molecules that “hit” a will be possible to “screen” candidates designated target, and further screens in a digital representation of the to test the physical and toxicological human body which can be adjusted to properties of these molecules. reflect common genetic variations and Thereafter, they test the most promising disease traits, such as a weakened candidates in animals before forwarding cardiovascular system. This will show them for initial testing in man (see whether a molecule interacts with any Figure 4). unwanted targets and produces any This approach has numerous side effects, and in what circumstances drawbacks. It usually produces a it does so. Predictive analysis will then reasonably clear picture of how the enable researchers to assess how molecule being researched interacts the molecule is likely to be absorbed, with the target, and how safe it is. But distributed, metabolised and excreted; in vitro assays and animal models of a what long-term side effects it might disease are often unrepresentative of have; what free plasma concentration is that same disease in human beings, needed to provide the optimal balance and are thus an unreliable means of between efficacy and safety; and what

Figure 4: The current research process

Hit ID/ Testing of Testing of Design of Synthesis of Initial testing Target ID screening of molecule molecule molecule molecule in man molecule in vitro in vivo

Mixed Computer/Lab Lab work Testing in man

Source: PricewaterhouseCoopers

6 PricewaterhouseCoopers Figure 5: What the research process might look like when virtual man exists Taking the pain out of the process When bioinformatics provider Design of Testing of Synthesis of Initial testing BioWisdom wanted to identify which Target ID treatment treatment treatment in man ion channels might be promising targets for developing pain treatments, Primary it began by creating a vocabulary In silico pharmacology for 349 human ion channels. Then Lab work Wide-ligand profiling it trawled MEDLINE to find every Testing in man (testing for side-effects) document in which these 349 ion ADME properties channels, or any of their 4,000 Pharmaceutical science synonyms, were mentioned in Toxicology conjunction with the central and peripheral sites that are known Source: PricewaterhouseCoopers to be relevant in mediating pain. Further analysis of the literature on the 59 ion channels this process formulation and dosing levels might (e.g. headache and migraine) in different produced showed that 11 ion work best. This suggests that much of sources, it cannot readily be connected. channels were clearly associated with the three key mechanisms of pain the work currently undertaken in the Conversely, where two concepts share (central sensitisation, allodynia and clinical environment can be tackled the same name but are fundamentally hyperalgesia), while another 10 looked much earlier within discovery. Figure different, they will be treated as if worthy of additional exploration. 5 shows what we think the research they are identical. As a result, it is process might look like, once robust very difficult to aggregate data from computer models of the entire human multiple sources and make meaningful body are available. associations between them, without the application of intelligent reasoning. However, as we have already indicated, the “birth” of virtual man is still many Semantic technologies, by contrast, will years off. In the more immediate allow scientists to connect disparate future, two other advances – semantic data sets, query the data using “natural technologies and computer-aided language” and make correlations molecule design – will play a much bigger that would otherwise have been unobservable (see sidebar, Taking the part in enhancing the research process. pain out of the process).14 This will Traditional informatics systems are make it much easier to identify the links constrained both by the structure of the between a particular disease and the data they represent and how they can biological pathways it affects, or the represent those data. Thus, if the same links between a particular molecule and concept is called by different names its impact on the human body.

Pharma 2020: Virtual R&D 7 Table of contents

Meanwhile, computer-aided molecule a medicine which is known to work in now pays its researchers a bonus only design (utilising greater knowledge the clinic has no impact in a “predictive” when a candidate molecule reaches the about biological targets and their in vivo model, it will be clear that the proof-of-concept stage or when they structure) will give researchers a much model is flawed. solve major problems, such as figuring better starting point in the search for out how to make a previously insoluble Similarly, the industry can change the potent molecules and reduce the need compound soluble.15 way in which it rewards research. Most to run high throughput screens to find companies have traditionally promoted This has two advantages: it encourages hits, which is effectively like looking researchers to focus on creating for a needle in a haystack. It will still their best scientists to management compounds which have a real chance of be necessary to test these molecules positions, although scientific expertise is success in the clinic; and it strengthens in in vitro and in vivo assays, until no guarantee of managerial competence. complete models of the anatomical They also reward researchers for the links between the research and and physiological characteristics of the getting candidate molecules to the development functions. But it is not human body in a healthy and diseased stage preceding submission of the enough merely to reward success; state are available. But some parts investigational drug application for it is equally important to promote a of the research process will become testing in man, thereby encouraging “fail early, fail cheaply” mindset, by increasingly virtualised within the next those researchers to push unviable providing incentives for pulling the plug 12 years (see Figure 6). molecules further down the pipeline so on unpromising candidates as fast as that they can meet their targets. possible. Moreover, while Pharma waits for virtual nirvana, it can take several other A better way of stimulating genuine steps to improve the way in which it innovation would be to reward scientists conducts research. It can, for example, for “what they do, not for what the rest pay more attention to validating in of the company does”, as Jean-Pierre vivo disease models and making them Garnier, outgoing chief executive of more predictive by using products with GlaxoSmithKline recently noted. GSK proven clinical efficacy to test them. If has overhauled its incentive scheme and

Figure 6: What the research process might look like in 2020

Design & Further testing Testing of Synthesis of Initial testing Target ID initial testing of treatment treatment treatment in man of treatment in vitro in vivo

Mixed Computer/Lab Lab work In silico Testing in man

Source: PricewaterhouseCoopers

8 PricewaterhouseCoopers Expediting development

Of course, even the most robustly When is a medicine not a medicine? modelled molecules will still have to be tested in real human beings – just as Most medicines come in oral formulations and many patients do not take them the Boeing 777, which was completely properly. However, several emerging sciences will enable Pharma to develop designed on computers, had to be better ways of delivering existing therapies and totally new forms of treatment flown by test pilots before it could be which improve compliance – and hence outcomes. used to carry passengers. However, the With advances in nanotechnology, for example, it will be possible to deliver development process will also change therapeutic agents to specific cells in the human body, and develop nano-scale dramatically. machines for monitoring how medicines are distributed and metabolised. More Some of the new therapies Pharma than 100 nanotech-based medicines and delivery systems are already being develops will not be conventional developed. pharmacological agents capable of Gene therapies also hold great promise. Dozens of human trials are underway being tested in conventional ways (see to assess the efficacy of gene therapies in treating diseases that are currently sidebar, When is a medicine not a treated with pharmacological agents, one such instance being heart disease. In medicine?).16 These new treatments the short term, these efforts could result in the development of single treatments will be more difficult to make and they that remain effective for many months or years. In the long term, gene therapies will have a very different commercial may even provide a cure for hypertension and its related pathophysiology, but profile from traditional medicines taken they have many other applications, too. on a long-term basis, but they will also be much harder to replicate and will Regenerative medicine – the replacement or regeneration of human cells, tissues therefore help the industry to protect its or organs to establish or restore normal function – also has the potential to reverse intellectual property more effectively. the course of disease. Commercial products for treating skin ulcers and knee cartilage damage already exist, but a number of other medical conditions, such The development of clinical biomarkers as heart disease, insulin-dependent diabetes, spinal cord injuries and Parkinson’s and new technology platforms will disease, likewise look as though they could be amenable to cell-based therapy. likewise have a profound impact on

Pharma 2020: Virtual R&D 9 the way in which all new therapies are Pervasive monitoring tested. When biomarkers for diagnosing Pervasive monitoring – the use of miniature devices and wireless networks to and treating patients more accurately monitor patients on a real-time basis outside a clinical setting – has numerous are more widely available, the industry applications in clinical trials and everyday medical practice alike. Most of the will be able to stratify patients with monitors currently on the market are wearable devices with limited applications, different but related conditions and test but several devices that are reliable enough for medical purposes have either new medicines only in patients who recently been launched or are in the pipeline. suffer from a specific disease subtype. That will, in turn, allow it to reduce the Theranos has, for example, developed a handheld device for detecting adverse number and size of the clinical studies drug reactions in real time. The device tests tiny blood samples, using a required to prove efficacy. Using clinical biochip, and transmits the data to a central database. Evidence of too high a biomarkers that are reliable surrogates concentration of medication in the bloodstream automatically triggers an alert. for a longer-term endpoint, like Similarly, Microsoft Research is working on the development of a wearable survival, will also help to cut endpoint system for monitoring a wide range of physiological signals, which has been observation times. piloted on 20 patients in a study of sleep apnoea. The regulators are backing various Technological advances will also facilitate the development of embedded initiatives to create new biomarkers. monitors. In a recent paper on the future of healthcare, for example, British The FDA is, for example, providing Telecommunications (BT) suggests that electronic circuits could be painlessly scientific and strategic support for the “printed” onto the skin. These circuits might comprise an upper layer with a Predictive Safety Testing Consortium, polymer display and a deeper, more permanent player housing components an alliance between the C-Path Institute which are in contact with a patient’s blood capillaries and nerve endings. They and 15 pharmaceutical companies, to might even include a smart membrane that opens on command to let medication validate preclinical biomarkers for drug- through. induced nephrotoxicity, hepatotoxicity and other forms of toxicity.17 It has also established a biomarker qualification

10 PricewaterhouseCoopers programme in which the European A number of healthcare providers are it is not until the end of Phase II that Medicines Evaluation Agency (EMEA) already piloting remote monitoring scientists have a reasonable grasp of and Japanese Pharmaceutical and programmes. The new European the safety and efficacy of the molecules Medical Devices Agency are actively Centre for Connected Health, based in they are testing (see Figure 7). Even participating.18 Meanwhile, the Northern Ireland, is testing technologies then, that understanding may be European Commission and European that enable people with chronic, long- fatally flawed; in one recent review of Federation of Pharmaceutical Industries term conditions to live independently 73 clinical candidates which failed in and Associations have launched the at home. The UK government is also Phase III, for example, 31% were pulled Innovative Medicines Initiative (IMI), conducting a large tele-care trial, with because they were unsafe and 50% which aims, among other things, to the installation of remote monitoring because they were ineffective.22 create a framework for developing new devices in the homes of 7,000 patients,20 However, armed with a much better biomarkers and disease-specific centres and new technologies will facilitate understanding of the pathophysiology for validating them.19 the development of increasingly of disease and how the body behaves sophisticated, embedded systems Semantic technologies will also play a at a molecular level, and much better (see sidebar, Pervasive monitoring).21 major role in improving the development systems for monitoring patients, Indeed, with nano-scale devices that process. They will enable the industry to pharmaceutical companies will be able can measure absorption rates, it will link clinical trial data with epidemiological to refine their trial designs to reduce even be able to identify variations in and early research data, identify any the number of studies they perform efficacy as a result of non-compliance. significant patterns and use that and the number of patients on whom information to modify the course of its We believe that these scientific new medicines are tested. They will studies without compromising their and technological advances will start by administering a treatment to a statistical validity. Similarly, pervasive ultimately render the current model single patient who has been screened monitoring will allow Pharma to track of development, with its four distinct to ensure that he or she meets the patients on a real-time basis wherever phases of clinical testing, obsolete. inclusion/exclusion criteria, which are they are. At present, as we have already noted, likely to include specific genotypic and

Figure 7: The current development process

8 years

1.5 years 2 years 3 years 1.5 years

Submission Phase I Phase II Phase III Phase IIIb/IV MAA/NDA

Submission of CTA/IND CIM CIS Launch

CIM Confidence in Mechanism CIS Confidence in Safety IND Investigational New Drug CTA Clinical Trial Application MAA Marketing Authorisation Application

Source: PricewaterhouseCoopers

Pharma 2020: Virtual R&D 11 Table of contents

phenotypic characteristics as well as variant of diabetes will be used to shape Model trial the relevant disease subtype. the development of medicines for other Entelos has developed a virtual variants of diabetes. Once there is evidence that the research lab for simulating clinical treatment does not cause any One last change to the way in which trials of new treatments for a range immediate adverse events, it will be trials are designed will help to ensure of diseases, such as asthma, sequentially administered to other that Pharma directs its efforts more rheumatoid arthritis and diabetes. patients – from as few as 20 to as many productively. The industry has Scientists can model the impact of a therapy using multiple variables, for as 100 – all of whom have also been traditionally focused on establishing example in genotype, phenotype and screened to ensure that they have the whether new molecules are safe pathophysiology. right medical profile. The data they and efficacious, not on whether they generate will be compared to data provide value for money. In future, The virtual lab has already proved from the modelling that preceded the it will have to address the payer’s its worth. When Johnson & Johnson study and subjected to techniques like perspective. We believe that, by wanted to design a Phase I trial of Bayesian analysis to adapt the course 2020, pharmaceutical companies will a diabetes treatment with a novel of the study, but the study itself will be collaborate with healthcare payers mechanism of action, Entelos conducted in a single, continuous phase in different jurisdictions to develop simulated the effects of using various (see sidebar, Model trial).23 criteria for assessing the value of new dosing levels. As a result of this work, treatments – i.e. measurable increases Johnson & Johnson redesigned The development process will also in efficacy and ease of compliance or the trial, with a 40% saving in time become much more iterative, with decreases in healthcare costs – and and a 66% saving in the number data on a molecule for one disease they will integrate the criteria into their of patients on whom the treatment subtype getting fed back into the trial protocols. needed to be tested. The real-life trial development of new molecules for other subsequently confirmed the accuracy disease subtypes in the same cluster We predict by 2020 the clinical of the predictions the simulation had of related diseases (see Figure 8). So, environment will marry the needs of produced. for example, information that is derived patients, payors and providers, and from developing a medicine for one regulators by working much more

12 PricewaterhouseCoopers closely on a common agenda agreed sector has long been a major barrier Numerous countries are currently between them. They will share a to “interoperability”. However, the developing national health information common infrastructure and access to FDA and EMEA are actively promoting networks, and several European Union outcomes data and results. the creation of common formats for member states have already made collecting and reporting biological data. considerable progress. One of the These are not the only elements of the Various standards-setting organizations, main planks of the UK Connecting for trial design and development process including the Clinical Data Interchange Health programme, for example, is the that will change. The current system Standards Consortium and Society for creation of a new IT system that links of conducting trials at multiple sites Clinical Data Management, are also data about patients from different parts is very inefficient. By 2020, we think working towards this end. of the National Health Service, so that that it will be replaced by a system healthcare practitioners throughout the based on clinical supercentres – one They have already made considerable country can access the information or two per country, perhaps – to recruit progress in simplifying the standards safely, securely and easily, whenever patients, manage trials and collate trial that are used to exchange clinical data and wherever it is needed, and data. The supercentres will be owned between pharmaceutical companies, transmit prescriptions electronically.25 and run independently of the industry, contract research organizations, trial Similarly, France has embarked on an possibly by a new generation of site investigators and the regulators, ambitious programme to develop a management organizations, and they although many challenges remain. There national system of patient smart cards will act as centres of excellence in the is still, for example, no consensus on and electronic health records, which delivery of new medicines to patients. how different data and applications it aims to have working by the end should be integrated, or a set of Two technological advances will be of this year;26 Austria is developing a common business processes for necessary to facilitate this transition decentralised system that includes performing many clinical activities.24 – electronic data interchange and electronic health records, electronic Nevertheless, these problems will be electronic medical records – but both prescriptions, electronic referrals and resolved within the next 12 years. are already on the horizon. The diversity electronic medication histories;27 and and complexity of the information Use of electronic medical records Portugal is piloting an electronic identity that is generated by the life sciences will also be widespread by that time. card with a chip that will ultimately be

Figure 8: What the development process might look like in 2020

1.5 years

Confidence in Mechanism from Research work 1 year 0.5 year Epidemiological Data First into Man Automated Limited Disease Knowledge (Adaptive Design) 20-100 pts submission/approvals Clinical Use Knowledge / Data from clinical usage or similar products Proof of value requirements CIE Launch CIS

Clinical Data / Knowledge incorporated into studies on future indications/populations Source: PricewaterhouseCoopers

Pharma 2020: Virtual R&D 13 used to store various kinds of personal Managing the trial process in 2020 information, including medical records.28 When John Doe’s doctor asks whether he would be willing to participate in a With common data standards, electronic study for a new treatment for AD3, the subtype of Alzheimer’s disease with which medical records and electronic data he was recently diagnosed, he jumps at the chance. So his doctor forwards interchange, it will be possible to John’s medical details to the national clinical supercentre. Soon afterwards, John manage clinical trials long-distance and receives an email from the supercentre confirming that his clinical profile fits the almost completely electronically, using screening criteria and asking him to call one of the nursing staff there to discuss a small number of supercentres which the process in more depth. have been globally accredited by the Three weeks later, John attends the clinical supercentre, where he is given a slow- regulators and have direct links with release implant containing the new medicine and closely supervised on an inpatient specialist medical units. They will source basis for five days. All the evidence suggests that John is responding well, so he is patients from local healthcare providers then injected with a saline solution containing thousands of micron-sized robots, and satellite centres, use remote which will track how he responds in the longer term. monitoring devices to track how those patients respond to new medicines The trial investigator explains how the data these robots capture will be and analyse the data before forwarding transmitted to a central database at the supercentre and electronically filtered, them (in blinded form) to the sponsoring using intelligent algorithms and pre-programmed safety parameters, to detect companies (see sidebar, Managing the any abnormalities. If John experiences an adverse reaction, the system will trial process in 2020).29 immediately notify the supercentre, which will then contact him to discuss the implications. John’s doctor will also be informed, if he requires medical care. For This approach has numerous his part, John will only have to make one additional visit to the supercentre for a advantages. It will accelerate trial final check. recruitment and ensure that trials are managed more efficiently and When John goes home, he is reassured by the knowledge that he is being consistently. It will also make all trial monitored around the clock. Most of the time he forgets about the steady stream data more transparent, enabling them of data his nano-monitors are transmitting, but he knows that the supercentre to be scrutinised more comprehensively is analysing and collating them with clinical data from other patients, using (and more impartially, since it is often continuous feedback loops to refine its research. difficult for scientists who have spent several years developing a new medicine to retain their objectivity). Lastly, it may encourage more patients and healthcare providers to participate in clinical studies.

14 PricewaterhouseCoopers Collaborating to bring treatments to the market

The regulatory approval process will commercial failure on its hands. restricted basis. With each incremental change equally substantially. At present, Moreover, securing the marketing increase in evidence of safety, efficacy a regulator reviews the evidence on licence and getting a product approved and value, the regulator will extend the a candidate molecule at the end of for reimbursement may take several licence to cover more patients, different development, when the sponsoring years – years in which the clock on the indications or different formulations company submits the supporting dossier. patent is ticking away. (see Figure 9). If it is satisfied with the data, it issues a By 2020, we believe that decisions marketing licence permitting the company about reimbursement will fall within With the advent of the “live licence” to market the medicine throughout the the remit of the regulatory body which approach will need to modify our area over which it has jurisdiction. conducts the quality, safety and efficacy existing practices to realise its full Thereafter, the new medicine is review and that this cumbersome, all- potential. An analogy today is where increasingly likely to be subjected to a or-nothing approach will be replaced organizations can obtain conditional health technology assessment – which by a cumulative process, based on approvals in smaller patient populations is usually undertaken by a completely the gradual accumulation of data. The utilising the Orphan Drug Regulations. different agency – to determine sponsoring company will collaborate This approach is similar to the general whether or not it should be eligible for with the regulator to establish the “live licence” approach envisioned reimbursement. A growing number evidence it is required to provide, in 2020, but is not as seamless or of healthcare payers now conduct and when it is required to do so. It pharmacoeconomic evaluations of new will submit the data electronically at connected with the other stakeholders treatments and refuse to reimburse predetermined milestones, in line with in the healthcare environment. A move products for which they do not believe this timetable. toward the live-licence approach will there is sufficient evidence of economic require the use of novel clinical trial Once there is sufficient evidence to as well as clinical value. designs and changes in regulatory show that a medicine genuinely works Any pharmaceutical company which and is cost-effective in the initial trial strategies whilst also leveraging is launching a new treatment must population, the regulator will issue a electronic health records and pervasive thus overcome two external hurdles “live licence” allowing the sponsoring monitoring in a way that has never been to ensure that it does not have a company to market the treatment on a done before.

Figure 9: What the regulatory process might look like in 2020

1.5 years Clinical use US Clinical use US 1 year 0.5 year Clinical use FR Clinical use FR First into Man Automated Up to 1.5 years Clinical use DE (Adaptive Design) submission Clinical use DE 20-100 pts /approvals Clinical use UK Up to 1 year Up to 0.5 year Clinical use UK Automated Clinical use ES Data New Indication Clinical use ES from /Population Studies submission/ approval of CIE Limited Clinical (Adaptive Design) Clinical use CA Clinical use CA 20-100 pts new indication/ CIS Launch Use population Data Clinical use IT Clinical use IT from Clinical Clinical use SW Use Clinical use SW CIE Broader Enhanced CIS Launch Clinical use FI Clinical use FI Clinical use IS Clinical use IS Clinical use IN

Clinical use CHI

Clinical use MX

Source: PricewaterhouseCoopers Clinical use BZ

Pharma 2020: Virtual R&D 15 Moreover, the regulator will decide The industry has often argued that the Second, as the delineation between whether or not to licence a medicine regulatory process is an impediment pharmaceuticals, medical devices, gene using specific risk/benefit analyses to innovation. However, the leading therapies and other treatments gradually rather than data on average outcomes. agencies have clearly signalled that diminishes, the regulation of such It will ask sponsoring companies to they are willing to consider new ways products could be brought under the disclose the gaps in their knowledge of developing and regulating medicines same roof (as has already happened in about the risks associated with any (see sidebar, Agencies of change).30,31 the UK). At one time, distinct boundaries medicines they submit for approval, In November 2007, for example, the existed between the medical device and and it will make reimbursement of new FDA and Duke University Medical biotechnology industries, and different therapies contingent on performance. Center launched a partnership to procedures are still used to regulate modernise clinical trials. This initiative them. But the development of drug- The burden of proof will thus become will, among other things, explore the eluting stents, implanted wafers for the more precise, but the benefits of controlled release of chemotherapy opportunities for streamlining clinical this precision will outweigh the agents and other such drug-device trials, minimising the administrative disadvantages. This process will combinations highlights the increasing load in multi-site trials and switching to enable a company to reduce its time artificiality of this distinction. electronic data management systems to market and earn revenue sooner that enable researchers to monitor data By 2020, then, there may well be a single thereby recouping its costs more in real time and help them spot safety regulatory regime covering all healthcare quickly. Peak sales will most likely be problems more rapidly.32 products. Indeed, there may even be a lower, but through step wise revenue single global system, administered by growth as each new part of the licence Two further trends could reinforce national or federal agencies responsible is approved the revenue stream will the need for closer links between for ensuring that new treatments meet the be boosted (see Figure 10). It will Pharma and its regulators. First, needs of patients within their respective also enable the regulators to manage a number of national and regional domains. This would help to reduce their resources more effectively, since agencies have begun to share safety the costs of regulatory compliance and they will be able to forecast their and efficacy data under mutual accelerate times to market even more, workload much more accurately. recognition agreements – data which but it would also mean that any product Lastly, and equally importantly, greater could be very useful in reducing the time which failed to pass muster with one collaboration between the industry and and costs associated with developing agency would be very unlikely to get its regulators, and greater transparency, new products. If the industry is to get licensed elsewhere, unless it treated a will help to restore public confidence in access to this information, it will have disease caused by a genetic variation in a Pharma’s integrity. to be equally open. very restricted ethnic population.

16 PricewaterhouseCoopers Figure 10: How the R&D cost/revenue curve might look by 2020 Agencies of change The US and European Union (EU) have both recognised the need for major changes in pharmaceutical research

nue ve Future R&D and development. In March 2004,

Re cost/revenue curve - with live licensing the FDA launched its Critical Path Step wise revenue Initiative, which aims to bridge the gap Faster to market - increases on automated reduced spend in approvals between basic scientific research and initial R&D the development process. In March $0 Traditional R&D 2006, it published its Critical Path cost /revenue Opportunities List, which identifies curve 76 projects for researching how new Time scientific discoveries – in fields such as genomics and proteomics, imaging and bioinformatics – can be used Source: PricewaterhouseCoopers to predict the safety and efficacy of candidate molecules more accurately, streamline clinical trials, improve For all these reasons, it is clear that pharmaceutical manufacturing and Pharma must work much more closely deliver treatments that address urgent public health needs. with the regulators than it has done in the past. Some companies have already Meanwhile, EMEA has produced a recognised this; it is no accident that the “Road Map to 2010”, which lays the most successful are those that are also foundations for major changes in the the most willing to share information, way in which medicines are regulated listen and adapt. By 2020, we think that and thus how R&D is performed. every company will have to operate The ultimate objective of the Road in the same fashion and that working Map is to ensure that EMEA and its with the regulators will be built into the partners in the EU medicines system remuneration packages of development adequately prepare the ground for scientists. future scientific advances.

Pharma 2020: Virtual R&D 17 ImperativesTable of contents for change

If Pharma is to remain at the forefront of medicines that only deliver marginal well as the skills of those who support medical research and continue helping enhancements in safety or efficacy. them. patients to live longer, healthier lives, it Indeed, the regulators may even refuse Those that regard R&D as an integral must become much more innovative, as to approve such products. part of their activities may also need to well as reducing the time and money it The new technologies we have review the way in which they manage spends developing new therapies. We identified can play a major role in their R&D and remunerate their scientific believe that incremental improvements helping Pharma move forward. They staff. We now know that one-size are no longer enough; the industry will will enhance its ability to produce medicines don’t fit all patients, and need to make a seismic shift to facilitate treatments which deliver measurable the same is true of the R&D process further progress in the treatment of improvements in safety, efficacy and itself. The limitations of the approach disease. ease of compliance – treatments on which the industry has relied for It will have to learn much more about which have value in the eyes of many years have become increasingly how the human body functions healthcare payers as well as those of clear and, in future, each company at the molecular level and the the companies making them. They will will have to chart its own course – or, pathophysiological changes disease also deliver substantial savings; indeed, rather, different courses for each of the causes. Only then will it be able to we estimate that they could collectively projects it undertakes. develop a better understanding of how halve development times and attrition The challenges Pharma faces are to modify or reverse these changes. rates, thereby reducing costs per drug enormous, but we are confident that it This is a huge undertaking – and one dramatically. can succeed. When Charles Babbage that Pharma cannot complete alone. It However, technology is not the answer to first proposed building his difference will require the support of academia, all Pharma’s problems. Many companies engine nearly two hundred years ago, governments, technology vendors, as well as regulators and vendors that nobody could have envisaged the healthcare providers and the regulators. support the industry will have to make connected world that exists today. Patients must play their part, too; significant strategic, organizational and Connectivity – technological, intellectual without access to medical data and behavioural changes. Pharma will, for and social – will ultimately enable us volunteers for clinical studies, the example, have to decide whether they to make sense of ourselves and the industry will be unable either to make want to produce mass-market medicines diseases from which we suffer. theoretical advances or to translate or speciality therapies; where they want Equally we as a society must those advances into practice. to be located geographically to have acknowledge that we cannot afford to Pharma will also need to virtualise much access to the best skills or cost base suffocate the investments made into of the research it currently performs in wherever they may be; and whether they R&D by the pharmaceutical industry; the lab, transform the way in which it want to outsource some or even most of a concern that should be high on the designs and manages clinical studies, their research and development or keep socio-political agenda. We have to face and pay greater heed to the views of it in-house. The choices they make will the issue that if Pharma is no longer healthcare payers. By 2020, aggressive have a profound bearing on the business financially capable of this, where will the marketing will not be enough to salvage models and mix of skills they require as new medicine come from?

18 PricewaterhouseCoopers Acknowledgements

We would like to thank the many people at PricewaterhouseCoopers who helped us to develop this report and contributed to the content. We would also like to express our appreciation for the input we have received from clients, and our particular gratitude to the following external experts who so generously donated their time and effort to the project.

David Jefferys, CEO, Eisai Europe Limited

Dr Brian Albert Gennery MB FRCP FFPM Pharmaceutical Medical Consultant Consultant CMO to PharmaKodex Previously Director of the Clinical Research Centre, University of Surrey

Dr Steven Manos Research Fellow Centre for Computational Science, Department of Chemistry, University College London

Dr Gordon Baxter Chief Executive Officer Biowisdom Ltd

Clive Page Professor Pharmacology, King’s College London Director, Sackler Institute of Pulmonary Pharmacology, King’s College London Co-founder of Verona Pharma

Dr John Horton MA, MB, BChir, MRCGP, FFPM. Adjunct Professor, Murdoch University Perth Western Australia, Ex- GlaxoSmithKline

The views expressed herein are personnal and do not reflect the views of the organizations represented by the individuals concerned.

Pharma 2020: Virtual R&D 19 Table of contents

References

1. PricewaterhouseCoopers, “Pharma 2020: 12. UCL press release, “Virtual human in 22. Maria A. Gordian, Navjot Singh et al., “Why The vision – Which path will you take?” HIV drug simulation” (January 29, 2008), drugs fall short in late-stage trials”, The June 2007. Available at http://www.pwc. accessed April 16, 2008, http://www.ucl. McKinsey Quarterly (November 2006), com/pharma2020 ac.uk/media/library/HIVcomputing accessed May 3, 2007, http://www. mckinseyquarterly.com/article_abstract_ 2. “FDA Approved Only 19 New Medications 13. Patrick McGee, “Modeling Success with In Silico Tools, Drug Research and visitor.aspx?ar=1879 In 2007, Analyst Reports”, Medical News Development (April 6, 2005), accessed 23. Jeff Trimmer, Chris McKenna et al., “Use Today (January 11, 2008), accessed April 16, 2008, http://www.dddmag.com/ of systems biology in clinical development: February 21, 2008, http://www. modeling-success-with-iin-silico-i.aspx design and prediction of a type 2 diabetes medicalnewstoday.com/articles/93690.php 14. “Using OmniViz to Identify and Prioritize clinical”, PAREXEL’s Pharmaceutical R&D 3. Bernstein Research, “Global Ion Channel Targets for Pain Drug Sourcebook 2004/2005 ( 2004), accessed Pharmaceuticals: Extending Models To Development” accessed April 16, 2008, May 24, 2008, http://www.entelos.com/ 2015 – A Long Term View Of The Generic http://www.biowisdom.com/system/ pubArchive/PAREXEL_FINAL.pdf ‘Cliff’” (March 20, 2008). files/OmniViz_Case_Study_Ion_Channel_ 24. CDISC, “The Benefits of a Life Sciences 4. CMR International survey of 29 Target_Prioritisation.pdf Industry Architecture” (Spring 2007), pharmaceutical companies in1998, 15. Jean-Pierre Garnier, “Rebuilding the R&D accessed April 17, 2008, http://www. reported in PAREXEL’s Pharmaceutical Engine in Big Pharma”, Harvard Business cdisc.org/publications/Benefits_ lifeSciencesIndustryArchitecture.pdf R&D Statistical Sourcebook (2001). Review (May 2008), DOI: 10.1225/R0805D. 25. For further information on NHS 5. J. DiMasi, R. Hansen & H. Grabowski, “The 16. Editorial, “Nanomedicine: A matter of rhetoric?” Nature Materials (April 2006), Connecting for Health, see http://www. price of innovation: new estimates of drug Vol. 5: 243, accessed May 20, 2008, http:// connectingforhealth.nhs.uk/ development costs”, Journal of Health www.nature.com/nmat/journal/v5/n4/full/ 26. “France’s Electronic Health Record – the Economics 22, (2003): 151–185. nmat1625.html; T.A. Khan, F.W. Sellke Dossier Médical Personnel”, eHealth 6. Christopher P. Adams & Van V. Brantner, & R.J. Laham, “Gene therapy progress Europe, accessed May 19, 2008, http:// “Spending on New Drug Development”, and prospects: therapeutic angiogenesis www.ehealtheurope.net/Features/item. Working Paper, Federal Trade Commission for limb and myocardial ischemia”, Gene cfm?docId=194 – Bureau of Economics (March 14, 2008), Therapy (February 2003) Vol. 10: 285–291; 27. “The Austrian e-Health strategy”, e-Health accessed May 21, 2008, http://papers.ssrn. and Chris Mason & Peter Dunhill, “A Europe, accessed May 19, 2008, http:// com/sol3/papers.cfm?abstract_id=869765 brief definition of regenerative medicine”, www.ehealtheurope.net/Features/item. Regenerative Medicine (January 2008) 7. We have used data from a variety of cfm?docId=215 Vol. 3, No. 1: 1-5, accessed May 20, sources to calculate the average costs 28. “New electronic identity card enables 2008, http://www.futuremedicine.com/doi/ of each stage of the R&D process. We patients to look after their records”, full/10.2217/17460751.3.1.1 eHealth Europe, accessed May 19, 2008, estimate that basic research and discovery 17. US Food and Drug Administration, “FDA http://www.ehealtheurope.net/Features/ accounts for about $165m, preclinical and the Critical Path Institute Announce item.cfm?docId=237 development for $87m, Phase I for $130m, Predictive Safety Testing Consortium” 29. Research Rewired, Merging care and Phase II for $190m, Phase III for $268m (March 16, 2006), accessed May 17, 2008, research information to improve knowledge and approval for $26m. The proportion http://www.fda.gov/bbs/topics/news/2006/ discovery”, PricewaterhouseCoopers, of these costs that goes on molecules NEW01337.html 2008 Available at: http://www.pwc.com/hri which fail to reach the market is the sum 18. Janet Woodcock M.D., “The Role of 30. For further information on the FDA’s of the cost of each phase from preclinical Biotechnology and Bioinformatics Critical Path Initiative, see “Innovation/ development to approval, multiplied by the in FDA’s Critical Path Initiative” Stagnation: Challenge and Opportunity attrition rate in each phase. (September 25, 2007), accessed on the Critical Path to New Medical 8. John Gartner, “Virtual Vermin Saves Lab May 17, 2008, http://66.102.9.104/ Products” (March 2004), accessed April Rats”, Wired (20 May 2005), accessed May search?q=cache:QoHkcn- 20, 2008, http://www.fda.gov/oc/initiatives/ dsIgJ:www.ieeeusa.org/volunteers/ 17, 2008, http://www.wired.com/medtech/ criticalpath/whitepaper.pdf; and “Critical committees/MTPC/documents/ health/news/2005/05/67541 Path Opportunities List” (March 2006), JWCPBioeconomicsNIST092507.ppt+EME accessed April 20, 2008, http://www.fda. 9. STEP Consortium, “Seeding the A+%2B+biomarker+qualification&hl=en&ct gov/oc/initiatives/criticalpath/reports/ EuroPhysiome: A Roadmap to the Virtual =clnk&cd=9&gl=uk opp_list.pdf. For further information on Physiological Human (2007), accessed 19. Innovative Medicines Initiative, “Research the IMI’s Strategic Research Agenda, see April 16, 2008, http://www.biomedtown. Agenda” (February 15, 2008), accessed “Research Agenda” (February 15, 2008), org/biomed_town/STEP/Reception/step_ April 19, 2008, http://www.imi-europe. accessed April 20, 2008, http://imi.europa. presentations/RoadMap/plfng_view org/DocStorage/PublicSiteAdmin/ eu/docs/imi-gb-006v2-15022008-research- 10. Marco Viceconti, Fulvia Taddei et al., Publications/IMI-GB-research-agenda- agenda_en.pdf. “Towards The Virtual Physiological 006v2-15022008.pdf 31. For further information on EMEA’s Human: The Living Human Project” (2006), 20. “Belfast gets European Connected Health programme, see “The European Medicines accessed April 16, 2008, http://kmi.open. centre” (January 29, 2008), accessed Agency Road Map to 2010: Preparing the May 19, 2008, http://www.ehealtheurope. ac.uk/projects/irs/resources/CMMBE2006/ Ground for the Future” (March 4, 2005), net/news/3418/belfast_gets_european_ accessed April 20, 2008, http://www. pap_CMBBE2006_viceconti.pdf; and connected_health_centre emea.europa.eu/pdfs/general/direct/ http://www.physiome.org/ 21. Nicholas Davies & Stuart Henderson, directory/3416303enF.pdf 11. Ellen L. Berg, “In the Game”, Drug “Drugs, devices, and the promise of 32. US Food and Drug Administration, Research and Development (February, pervasive computing”, Current Drug “FDA and Duke Launch Public-Private 2008), accessed April 16, 2008, http:// Discovery (October 2003): 25-28; and Partnership to Modernize Clinical Trials” www.dddmag.com/Article-Systems- British Telecommunications, “Pharma (November 26, 2007), accessed April 20, Biology-Integrates-into-Research-and- Futurology Joined-up Healthcare: 2016 2008, http://www.fda.gov/oc/initiatives/ Development-Processes.aspx and beyond” (2007). criticalpath/partnership.html 20 PricewaterhouseCoopers Territory contacts

Argentina India Russia Diego Niebuhr Thomas Mathew Christian Ziegler [54] 4850 4705 [91] 22 6669 1234 [7] 495 232 5461

Australia Ireland Alina Lavrentieva [7] 495 967 6250 John Cannings John M Kelly [61] 2 826 66410 [353] 1 792 6307 Singapore Belgium Enda McDonagh Abhijit Ghosh [353] 1 792 8728 [65] 6236 3888 Thierry Vanwelkenhuyzen [32] 2 710 7422 Israel South Africa Brazil (SOACAT) Assaf Shemer Denis von Hoesslin [972] 3 795 4671 [27] 117 974 285 Luis Madasi [55] 11 3674 1520 Italy Spain Canada Massimo Dal Lago Rafael Rodríguez Alonso [39] 045 8002561 [34] 91 568 4287 Gordon Jans [1] 905 897 4527 Japan Sweden Czech Republic Kenichiro Abe Liselott Stenudd [81] 80 3158 5929 [46] 8 555 33 405 Radmila Fortova [420] 2 5115 2521 Luxembourg Switzerland Denmark Laurent Probst Clive Bellingham [352] 0 494 848 2522 [41] 58 792 2822 Erik Todbjerg [45] 3 945 9433 Mexico Peter Kartscher [41] 58 792 5630 Torben TOJ Jensen Ruben Guerra [45] 3 945 9243 [52] 55 5263 6051 Turkey Finland Netherlands Ediz Gunsel Janne Rajalahti Arwin van der Linden [90] 212 326 6060 [358] 3 3138 8016 [31] 20 5684712 United Arab Emirates Johan Kronberg Poland [358] 9 2280 1253 Sally Jeffery Mariusz Ignatowicz [971] 50 252 4907 [48] 22 523 4795 France United Kingdom Jacques Denizeau Portugal Andrew Packman [33] 1 56 57 10 55 José Fonseca [44] 1895 522104 [351] 217 599 601 Germany Volker Booten [49] 30 2636 5217 ForTable further of contents information, please contact:

Global Europe Simon Friend Jo Pisani Partner, Global Pharmaceutical and Life Sciences Industry Leader Partner, European Pharmaceutical and Life Sciences PricewaterhouseCoopers (UK) Advisory Services [email protected] PricewaterhouseCoopers (UK) [44] 20 7213 4875 [email protected] [44] 20 7804 3744

Steve Arlington Sandy Johnston Partner, Global Pharmaceutical and Life Sciences Advisory Partner, European Pharmaceutical and Life Sciences Services Leader Advisory Services PricewaterhouseCoopers (UK) PricewaterhouseCoopers (UK) [email protected] [email protected] [44] 20 7804 3997 [44] 20 7213 1952

Michael Swanick Kate Moss Partner, Global Pharmaceutical and Life Sciences Tax Leader Director, European Pharmaceutical and Life Sciences PricewaterhouseCoopers (US) Advisory Services [email protected] PricewaterhouseCoopers (UK) [1] 267 330 6060 [email protected] [44] 20 7804 2268

United States Asia Pacific Anthony Farino Sujay Shetty Partner, US Pharmaceutical and Life Sciences Advisory Associate Director, Pharmaceutical and Life Sciences Services Leader Advisory Services, India PricewaterhouseCoopers (US) PricewaterhouseCoopers (India) [email protected] [email protected] [1] 312 298 2631 [91] 22 6669 1305

Mark Simon Beatrijs Van Liedekerke Partner, US Pharmaceutical and Life Sciences Industry Leader Associate Director, Pharmaceutical and Life Sciences PricewaterhouseCoopers (US) Advisory Services, China [email protected] PricewaterhouseCoopers (China) [1] 973 236 5410 [email protected] [86] 10 6533 7223 Michael Mentesana Partner, US Pharmaceutical and Life Sciences R&D Advisory Marketing Services Leader PricewaterhouseCoopers (US) Attila Karacsony [email protected] Director, Global Pharmaceutical and Life Sciences [1] 646 471 2268 PricewaterhouseCoopers (US) [email protected] [1] 973 236 5640

Melanie York Senior Manager, European Pharmaceutical and Life Sciences PricewaterhouseCoopers (UK) [email protected] [44] 20 7804 1991 pwc.com/pharma

This publication has been prepared for general guidance on matters of interest only, and does not constitute professional advice. You should not act upon the information contained in this publication without obtaining specific professional advice. No representation or warranty (express or implied) is given as to the accuracy or completeness of the information contained in this publication, and, to the extent permitted by law, PricewaterhouseCoopers does not accept or assume any liability, responsibility or duty of care for any consequences of you or anyone else acting, or refraining to act, in reliance on the information contained in this publication or for any decision based on it.

© 2008 PricewaterhouseCoopers. All rights reserved. ‘PricewaterhouseCoopers’ refers to the network of member firms of PricewaterhouseCoopers International Limited, each of which is a separate and independent legal entity.