Getting new drugs to market: how individuals could do this without leaving their desks

BMJ 2006; 333 doi: (Published 21 December 2006) Cite this as: BMJ 2006;333:1315
  1. Joe Collier, professor of medicines policy and consultant in clinical pharmacology
  1. 1Department of Basic Medical Sciences, St George's Hospital and Medical School, London SW17 0RE
  1. jcollier{at}

    At one time, all of the products that came out of a pharmaceutical company were researched, devised, developed, and manufactured within its four walls; the company offered the complete package. Now, although vast amounts of money, intellectual property, and management capacity remain within the imposing headquarters of the multinationals, much of the real work is done elsewhere—by start-up companies, contract houses, university departments, and public bodies.

    One way or another, the work of big drug companies has changed, and inevitably this change, which will continue and probably accelerate, will lead to the collapse of the monolithic system and its replacement by smaller, leaner, and more nimble companies. As this process continues, virtual companies, run by individuals working from their desks in cyberspace, could well develop.

    Developing drugs

    The development of medicines for human use is a tricky business and one that is continually evolving as new technologies, such as genetic engineering, emerge. At its simplest, the process involves finding “new” molecules, purifying them, checking that they do what they are designed to do, showing that they work in clinical practice without causing unacceptable harm, and then scaling up their manufacture to make the product of acceptable quality in commercial amounts. Along the line the product needs to pass legal hurdles to allow it to be sold and marketed to ensure returns. All of these steps, some running simultaneously, take time (currently 10-12 years), need money (estimated as being between $500m and $800m1), and require careful coordination.

    These steps consist of discrete packages of work relating to, for example, molecule synthesis, target validation, animal efficacy, animal toxicology, animal and human pharmocokinetics and pharmacodynamics, teratogenicity, carcinogenicity, effects in patients (clinical trials through phases I-IV), drug side effects, and drug interactions. Added to these are applications for patents and regulatory approval, followed by product manufacture, marketing, promotion, distribution, sales, and supply.

    Each of these discrete packages can be (and is being) undertaken, or managed by, small and highly skilled independent companies. In some instances, as with start-up companies, small groups develop ideas and products and then sell, or license, them on to the larger companies (the antiflu drug zanamivir was discovered by a biotech company in Australia, then licensed to GlaxoWellcome ultimately to be marketed as Relenza). In other instances larger companies will contract out areas of work, for instance to public relations companies who will be in charge of marketing their products, or to development companies who will undertake clinical trials. So, for instance, the website for advertising agency McCann Erickson ( tells that it is their philosophy to “single out your story from the crowded advertising page, endowe it with the gift of vitality, create from it the buying impulse” and the contract research house Quintiles ( states that for clinical research work ‘we have global resources to quickly assemble not just a project team, but the best project team for local studies or multi-country mega-studies.” In some instances the drug companies expressly set up arrangements designed to court those with new ideas so that the two can develop new products to mutual advantage, for example the GlaxoSmithKline Centre of Excellence for External Drug Discovery (

    This notion of contracting in or contracting out work is part of the pattern of how drug companies are now working. If the contracting business can work for drug companies, why not for an individual entrepreneur acting as a virtual company? Moreover, if the key features that make the big drug companies “tick” are centred primarily on their financial and intellectual resources, then perhaps an entrepreneur with a bias towards an interest in health would be a welcome change. After all, it is the drug industry's obsession with secrecy (a position that will serve to guard its intellectual resources) and its excessive pricing (a strategy designed to protect its wealth), that make drug companies particularly unpopular. It is also these assets (money and information) that give the industry its power and influence.

    We have to assume that the changes in drug development have come about for some reason, rather than being the product of chance. We do know that the large (multinational) drug companies are bulky, cumbersome, and costly to run and that they have, at least in the recent past, failed to produce the new products needed to guarantee returns, and have appeared (or have chosen to be) sluggish as often they have left the initiative and risk for new advances, particularly of biological agents, to small highly specialised biotech companies. Also, companies have a poor image and a high level of distrust. The collapse of the old system, at least as it relates to producing new drugs for the developing world, is echoed in the comment by Daniel Vasella, chief executive officer of Novartis, that “We have no model which would (meet) the need for new drugs in a sustainable way” and if you want drug companies to systematically invest in this area, you will “need a different system.”2 Even the large drug companies seem to recognise the need for change; now, with a myriad of independent contract houses, the wherewithal for change is in place.

    What would an individual need to do?

    It would probably cost around $150m-$240m to bring a new drug to market if the project is well managed,1 so the head of the virtual company will need to have access to such funds, sustained over 10-12 years. Stepwise development with support from seed investors, then venture investors, and then institutional investors would probably be the wisest approach and would share the risk, but investors, like drug companies themselves, will not spend their money unless they feel secure about returns.

    Assuming the money was there, entrepreneurs would need to surround themselves with a successful team (the members of the virtual company). The head of one large multinational is purported to have claimed that with 20 or so of his senior (and handpicked) colleagues he could run a company with ease. In addition to the chief executive, the members of the virtual company would probably need to include a talented senior scientist, researcher, financier, accountant, administrator, clinician, chemist, lawyer, marketeer, information technologist, and manager. Moreover a core of the team would need to know its way around the industry, academia, and the regulatory business. Clearly the team would need to understand drug development and the processes involved in “contracting” work. Somehow the new virtual company would probably need to develop a “persona” with an agreed modus operandi, qualities that the big drug companies see as particularly important. The company could be profit led (any drug would do, provided it raised money), disease led (focusing on drugs that target a particular disease, tuberculosis perhaps), or drug led (developing biotech products, or drugs from products now used as folk or herbal medicines, for example).

    What might be the advantages or disadvantages of the virtual company?

    The drug industry is made up of a few vast companies which reside in definable and tangible bases (“homes”), which, by and large, have identity, stability (staying power capable of long term investments), and familiarity (including their foibles); they tend to produce perhaps dozens of drugs, have a vast science base, have an unsurpassed worldwide sales network, and have some shared understanding with both doctors and society at large. Over the years the roles and responsibilities of drug companies have been defined in law and practice, as have their financial arrangements (public company status), which bring with them the openness (albeit limited) that the business world demands.

    Virtual companies would be very different. Gone would be any dividend arising out of stability and familiarity, any advantage from having a dedicated science force working to a recognised process, any insights arising from any shared understanding, and we could find ourselves dealing with a myriad new companies producing only a few products. In the new order, society would have to revise the ways it might influence the companies. Instead of lobbying shareholders and the financial sector, we would, for instance, have to lobby the venture capitalist, and this would require new insights and understanding. But in return, a more dynamic and imaginative drug producing process could result, the overpowering and stifling influence of drug companies on governments and policy makers would be lifted, and the price of medicines should fall.

    Summary points

    • The power and money of the old system of developing drugs are stifling, and companies' process for innovation is grinding to a halt

    • A virtual drug company, possibly run by a single entrepreneur working with a group of like minded people, could contract in or out all that needs to be done to produce and sell new medicines

    • Virtual drug companies producing just one or two drugs will bring dynamism and economies but will need careful oversight if the public interests are to be fully met


    • Competing interests: None declared.


    View Abstract