Malcolm Young at e-Therapeutics plc and Alan Whitmore at Lumemed Ltd warn that a gap has opened between current science and commercial drug discovery science which the latter had better close
Science costs money. In mainstream science, and in universities and research institutes, the decision to award money for new science is usually dependent on two factors. One is the extent to which the research leader and team have delivered important new results in the past. The other, generally more important factor is whether the current research proposal is framed creatively in the most up-to-date science, and would therefore contribute substantially to knowledge. Both factors are assessed by groups of competitive peers, whose current expertise and critical faculties are very keen.
Scientists often suspect that more weight is placed on the proposing researcher’s track record of turning funding into highly successful outputs, and that science that really pushes the boundaries by challenging established dogma is often felt by peers to be more risky and less likely to get funding. Nevertheless, despite its many shortcomings, this system does at least require scientific proposals to be assessed by up-to-date expertise in the field, able to understand and evaluate the quality of methodology and strategy, as well as current context and present relevance.
The coming-together of money and science is similar but different when it comes to commercial drug discovery research. The similarities are that historical performance is an important factor, and that nature is never fooled by poor science, wishful thinking or hubris. But the differences can be striking to the experienced mainstream science research leader entering the commercial sphere.
ASSESSMENT
One glaring disparity between the commercial and the academic worlds lies in the framework for assessment of fresh science in drug discovery. The gatekeepers to the research dollar in commercial drug discovery are typically either investment fund managers for relatively small biotech enterprises, or internal research managers for more mature ones. In the case of investment fund managers, their funding decisions are made almost invariably without current and detailed participation in, or exposure to top-flight science or cutting-edge medicine, and without an upto- date experience base that can evaluate the current competitiveness, specific scientific plausibility and medical practicality of the proposals. No one can doubt the ability of fund managers to evaluate the business proposal, but the business proposal in biotech is typically framed by the science.
For internal research within larger enterprises, funding decisions are most often made by application of exactly the same critical faculties and assumptions that led to some historical successes, but also to an unmistakable recent history of clinical failure. Neither assessment framework is primarily motivated by the desire to discover the best new treatments for patients, but to support the most profitable programmes with the lowest risk and cost. For these reasons, in both cases the great danger is that assessment and funding decisions in commercial drug discovery research strongly favour the conventional over the radical, the older over newer science, and the more familiar over the most deserving. Assessment in the commercial sphere hence suffers a variety of science lags that frequently puts it out of step with current understanding in mainstream science – a science lag to which, ironically, a very high cost may be attached, mainly to investors and shareholders in the companies making the decisions. In a global industry, at present characterised by unsustainably low productivity in drug discovery, this seems to us a problem worth addressing with alacrity.
WATER UNDER THE BRIDGE
Another dissimilarity derives from the sheer length of time required for drug discovery and development. Despite heroic efforts in many quarters to reduce the development period where possible, by the time a candidate drug reaches Phase III, the original data that motivated it – perhaps five or 10 years before – can be revealed by constantly evolving external science as partial or mistaken, so that the mechanism of action is questionable or improvable. For example, those working on ERK/MAPK inhibitors are likely to know of the advances in science since the identification of these kinases as targets in cancer, which have now shown a variety of previously unsuspected network feedback loops whose function will tend towards therapeutic leakage, and which has revealed a variety of much better motivated targets (1). If it is assumed that, not much more time can realistically be shaved from the discovery and development process, then perhaps the role of current science here will be limited to more informed decisions as to whether to progress particular candidates, and choosing companion molecules rationally that may mitigate therapeutic leakage through the now better understood pathways.
SCIENCE LAG
Both the science, and the assessment of the science, can suffer a lag in commercial drug research, which can put it out of step with the best up-to-the-minute global knowledge. The extent of this lag may be related to the separation between the advances and the commercial researcher’s background. As stated above, news of one’s favourite kinase may reach commercial researchers quite quickly, since these advances are very close to the scientific background of the researchers concerned. But several of the really fundamental problems facing the industry, such as a limited ability to address complex diseases, which tend to be the only diseases in which a premium-priced drug might be prescribed in preference to a generic, and the attrition of drug candidates in late-stage clinical trials, are clearly addressed in current science that may lie very far from the scientific comfort zone of private sector research funding decision-makers.
For example, the problems of disease complexity and late-stage attrition, which together frame the low productivity of the industry, are picked out in flashing fluorescent lights in current science, albeit in a science of which many decisionmakers appear not to even have heard – namely complex systems science. Complex systems science is populated by mathematical analysis, computation and very large data resources, and so can seem very inhospitable to the molecular biologist or synthetic chemist. But its lessons for disease complexity and latestage attrition are straightforward. All drug molecules make multiple interventions on complex networks in and between cells, through their direct and indirect effects on multiple proteins. There are synthetic effects present when more than one intervention is made, some of which are positively or negatively synergetic. This means – unless a rare scientific miracle happens – that the net effect of a molecule on cells cannot be accurately predicted from the potency of its affinity at a single protein target. The presence of negative synthetic effects can yield the nasty surprises in clinical efficacy seen so often for molecules that have high potency at a ‘validated target’. The presence of positive synthetic effects often yields the unpleasant surprises in clinical safety and tolerance that beset the industry.
The basis for more successfully predicting these effects has been known for at least 10 years in mainstream science (2). The importance of these effects for drug discovery productivity has been pointed out repeatedly, and tools for de-risking these effects have been derived (3-5). Yet, the search for the protein version of the Ehrlich magic bullet continues in face of science that shows unequivocally that such a search will generally fail. It would be hard to overstate how absolutely flabbergasting this perseveration is when viewed from the perspective of the newly commercial mainstream scientist, and when we think of what is at stake. Such a state of affairs could not prevail long in mainstream science, despite its many shortcomings.
BIOTECH SCIENCE INJECTION
Science lag has previously been mitigated by small innovative biotech, based on novel science believed by the founders and investors to be of significance for drug discovery. In previous cycles, this has been a very important source of innovation in the industry. However, at present, innovative new technology companies are stifled by two factors. One is the relative disinterest from conventionally thinking pharma business development, which has several other rather pressing things to think about. Perhaps a novel ‘target’ will be as innovative a technology as current interest can stretch to. Another factor is the same lack of interest from investors that stifles investment in new biotech formed by the large number of researchers having to leave the major pharma companies. Early stage biotechs appear very risky to investors, given the recent history of clinical failures in companies of every size, which small companies are less well-equipped to survive. Investor risk perception is, perversely, higher for the most innovative companies, despite the fact that the likely performance features of more conventional approaches can be quite well predicted from the unhappy performance features of the drug discovery parts of the industry in recent years.
These considerations could reasonably prompt a pessimistic view of the likely role of top-flight science in drug discovery going forward, and also of the likely fate of patients hoping that these activities will generate important new drugs for their poorly treated diseases. The capacity for drug discovery in the major companies is reducing sharply as research staff are laid off in considerable numbers; investor appetite for biotechnology companies is low and reducing with every clinical failure, so that discovery capacity across biotech is also reducing; the prospects for reducing science lag by importing the real innovations that can tilt discovery and development productivity upward once again are unclear.
Perhaps something will nonetheless come through the fire. The solutions are not obvious, but they could involve, on the one hand, much more government support for innovative biotech, perhaps in return for favourable terms for the use of the drugs in the public health system of the country providing the support. Funding and infrastructure that can permit greater interaction between the best academic institutions and the drug discoverers could provide access to upto- date expertise in a wide range of fields and economies of scale, increasing the probability of success and ultimately yielding both innovative new medicines and income for both partners. On the other hand, perhaps extending forms of patent protection, and reducing the level of damages from litigation to something that reflects harm caused, rather than perceived wealth of the company, would benefit us. Changes to the regulatory systems that approve new drugs to make them more flexible, less time consuming and fairer would result in cheaper, better drugs for patients, while still enhancing the ability to build a viable commercial enterprise around drug discovery (6).
One way or another, we need to rebuild a bridge between commercial and current academic science relevant to drug discovery, and to create a new environment in which innovation is protected, at least to some degree, from the vagaries of market forces. If we do not then we will fail to deliver new medicines that can benefit our children and grandchildren, and in a century’s time they will all still be living with the spectres of neurodegeneration, age-related decline, cancer and pandemic infectious disease.
References
- Zehoraia E, Yaoa Z, Plotnikova A and Seger R, The subcellular localization of MEK and ERK – A novel nuclear translocation signal (NTS) paves a way to the nucleus, Molecular and Cellular Endocrinology 314: pp213-220, 2010
- Young MP, Hilgetag C and Scannell JW, On imputing function to structure from the effects of lesions, Phil Trans R Soc Lond B 355: pp147-161, 2000
- Hopkins AL, Network pharmacology: the next paradigm in drug discovery, Nature Chemical Biology 4: pp682-690, 2008
- Young MP, Prediction versus attrition in drug discovery, Drug Discovery World 3: pp9-12, 2008
- Zimmer S and Young MP, From low productivity to efficient networkbased drug discovery, Innovations in Pharmaceutical Technology 18: pp38-41, 2009
- Bartfai T and Lees GV, Drug Discovery from Bedside to Wall Street, Elsevier AP, 2006 7. Cockell SJ, Weile J, Lord P, Wipat C, Andriychenko D, Pocock M, Wilkinson D, Young M and Wipat A, An integrated dataset for in silico drug discovery, Journal of Integrative Bioinformatics 7(3): p116, 2010
