The development process, like the research process, needs to undergo major changes to reduce the time and costs associated with bringing new medicines to market. As Dr Scott Gottlieb, the FDA’s Deputy Commissioner for Medical and Scientific Affairs, recently noted, the highly empirical, statistical method that currently predominates is inflexible; it restricts innovation and results in “overly large” trials that yield information “about how large populations with the same or similar conditions are likely to respond to a treatment.
But doctors don’t treat populations, they treat individual patients.” Of course doctors still lack many of the diagnostic tools and medicines they need to treat patients individually because “stratified medicine”125 depends on the ability to identify the patients who are most likely to respond to a particular therapy –and without a sufficient understanding of the multifactorial causes of disease it is impossible to devise a means of distinguishing between patients with different disease subtypes–.
However, this is where clinical biomarkers have already begun to revolutionise clinical development and medical practice alike. As the authors of an excellent article on the subject explain, developing biomarkers to stratify patients with related but distinct conditions will enable Pharma to make different treatments for different patient subpopulations, test them only in patients who suffer from those conditions, and thus reduce both the number and size of the trials required to prove efficacy.
It will also help to cut endpoint observation times when a clinical biomarker is an accepted surrogate for a longer-term endpoint such as survival. In all, the authors estimate, better use of safety and efficacy biomarkers could halve development costs.
Moreover, targeted treatments have a very different economical model from that of conventional medicines. Clearly, the potential number of patients any one such treatment can serve is smaller than the number for whom a mass-market therapy can be prescribed.
But targeted treatments, by definition, offer superior clinical results for the patient subpopulations whose distinct conditions they address, so they can generally command premium prices and are more rapidly adopted. The biomarkers themselves also provide additional opportunities for creating value, and using biomarkers to monitor patients’ progress can improve long-term compliance.
Combining biomarkers and medicines will thus help Pharma to make safer, more effective therapies more economically. In-silico testing will likewise improve its ability to predict the safety and efficacy of new medicines in different patient populations.
US life sciences company Entelos is one of several firms leading the way in the virtual domain; Entelos has created mathematical models of various diseases, including CVD, asthma, obesity, diabetes and rheumatoid arthritis, which it is using to acquire a better understanding of disease, identify targets and test potential medications.
Lastly, “pervasive healthcare” –the use of remote devices to monitor patients on a real-time basis wherever they are– will allow the industry to test new medicines outside a clinical setting. Pervasive computing is still in its infancy, and the infrastructure required to support it has yet to be fully developed. But, by 2020, robust portable monitoring devices and the wireless networks across which the data they collect can be sent will both be in place (see below, Anytime, anywhere healthcare).
Together with EMRs, “smart cards” containing details of patients’ individual health records (much as store cards track their shopping habits) and semantic technologies to link different kinds of data, pervasive healthcare will create a day-to-day environment that equates with the controlled environment in which clinical trials are conducted today.
All these changes will facilitate the refinement of the development process. A company will begin by defining the minimum amount and kind of information it needs to secure approval for “in-life testing” of a new medicine.
It will then perform a series of small, highly targeted clinical studies, using simulation, modelling and other technologies, to ensure that it understands the efficacy and safety of the product concerned, before submitting the data to the relevant regulatory agency –thereby rendering the traditional four-phase approach to clinical development redundant–.
If the regulator is satisfied with the evidence, it will issue a “live licence” permitting the company to market the medicine on a very restricted basis. The company will thereafter conduct in-life testing of that medicine in a small population of patients (many of whom will be referred via specialist centres or patient advocacy groups). With each substantive increase in evidence of the medicine’s safety and effectiveness, the regulator will extend the licence to cover a larger number of patients, a different patient population or multiple indications (see Figure 13).
This process has several advantages. It will reduce clinical development costs still further and allow pharmaceutical companies to recoup some of their costs more quickly, thereby enabling them to charge lower prices for new therapies.
It will facilitate testing for polypharmacy in wider populations. And it will align the bench and the bedside more closely. Indeed, it might ultimately culminate in the complete integration of clinical trials with clinical practice, as is already starting to happen in the treatment of cancer.
So, for example, a patient who suffered from diabetes and lived in Paris would be automatically given the opportunity to enrol in clinical trials in the area at the same time as receiving treatment. In effect, clinical trial participation would become part of normal care.
Clearly, some of the reforms we have outlined depend on the willingness of the regulators, as well as the political and legislative changes required to alter any regulatory regime. However, the European Medicines Agency (EMEA) and FDA have already shown that they are ready to grant conditional marketing approvals for some therapies, subject to certain obligations, including the completion of in-life testing.
The EMEA authorised the use of conditional approvals for orphan drugs and therapies for life-threatening conditions in April 2006 under Regulation EC 507/2006, and the FDA is piloting the concept under the Prescription Drug User Fee Act (PDUFA) III.
By 2020, we believe that all medicines which receive approval will be approved on a real-time basis, with live licences contingent on the performance of extensive in-life testing, including trials in specific patient subpopulations, and a predetermined schedule for reviewing each set of results.
If in-life testing confirms that a medicine is safe and effective, the company making it will be granted an extended licence or special permit – such as the paediatric use marketing authorisations (PUMAs) which are already used in Europe – so it will have an incentive to conduct further studies.
In other words, every medicine on the market will have a prearranged, fully automated pathway throughout its lifecycle, and its development will be a continuous process rather than ending when it is approved (see Figure 14).
But, as the legislation governing new medicines and the way in which they are licensed becomes more complex, the regulators will insist on greater collaboration and expect to be consulted on a regular basis from a much earlier point in development.
The FDA has already signalled its determination to become more involved in the development process with its Critical Path Initiative, which aims to create a new generation of predictive tools for improving safety and efficacy. Similarly, one of the goals of the EMEA’s Road Map to 2010 is to facilitate the formation of “an adequate product development toolkit, able to address the bottlenecks during the development of innovative medicines”.
The European Commission and European Federation of Pharmaceutical Industries and Associations (EFPIA) have now set up the Innovative Medicines Initiative, a pan-European collaboration to produce new drug development tools.
The criteria for approval will also become more challenging and more specific. The regulators are increasingly looking for evidence that new medicines are not just safe and effective, but better than any comparable existing therapies.
The EMEA often requires “comparator studies” where an alternative pharmacological treatment is available, and the FDA recentl hinted that it would only approve prescription painkillers which filled an unmet medical need for patients who have no other “relatively safer” alternatives.
Several agencies are simultaneously beginning to develop more sensitive detection systems and a more sophisticated attitude to risk management.
The EMEA has, for example, launched a new European Risk Management Strategy which requires that all pharmaceutical companies provide detailed information not only on what they know, but also on what they do not know, about the risks associated with any medicines or manufacturing processes they submit for approval.
They may then be required to develop risk minimisation plans. And one US expert has called for a new risk assessment framework that takes individual variations into account (see sidebar, Getting personal). These changes in the burden of proof will be accompanied by demands for greater transparency.
Pharmaceutical companies will, for example, be required to disclose all data from all clinical studies and inlife testing, regardless of whether they are favourable –with punitive treatment of any firm that breaches this rule–.
They will have to submit all product information electronically; supply data on all adverse events to a website managed by an independent intermediary, to which all prescribing doctors will be given access; and cope with additional scrutiny, in the form of third-party auditing of all their functions from R&D to sales and marketing (see Figure 15).
Figure 15.- New auditing bodies and processes will be required.
The EMEA already maintains a database of all clinical trials conducted within the EU, which interested parties can easily access. EudraCT, as the system is known, is rapidly becoming much more comprehensive in what it covers and may ultimately provide the basis for a global platform that ensures the transparency of all trial data.
Similarly, the MHRA has expanded Sentinel, the paperless system it launched in 2002, to cover licence submissions and product safety reporting, and the other EU regulators are likely to follow in its footsteps.
The FDA is also investigating ways in which to create an entirely paperless submission process and build an electronic exchange for sharing clinical research information. And Dr Mark McClellan, a former FDA commissioner, recently called for the creation of a database linking US public and private healthcare payer claims systems to improve monitoring for adverse drug events.
He argued that such a database would make it possible to target in-life studies more accurately, collect information about safety signals more effectively and better assess usage patterns. In future, many agencies will share such safety and efficacy data to create a broader picture of how different medicines perform. Indeed, by 2020, such data may even be managed in one global database to which every regulator has access.
Several national and regional regulators have already begun to collaborate. In September 2004, for example, the FDA and EMEA launched the Joint Scientific Advice programme, a forum for working together to provide companies with input during the development process and thus avoid the unnecessary replication of trials or use of diverse testing methodologies. But with the globalisation of R&D and the markets, as well as the transfer of a growing amount of pharmaceutical manufacturing to the developing world, the regulators of the E7 countries will become increasingly important, too.
The member countries of the Association of Southeast Asian Nations (ASEAN) have already revolutionised the regulatory systems of the Pacific Rim with a framework designed specifically for their patient universe. They are now building a common set of technical application forms for pharmaceutical registration, Good Manufacturing Practice (GMP) inspection and labelling.
The logical conclusion to such cooperation is, of course, the development of a single global agency charged with regulating pharmaceuticals everywhere –although this is unlikely, if for no other reason than national pride–.
Nevertheless, by 2020, there may well be one global regulatory system administered by national or federal agencies responsible for ensuring that new treatments meet the needs of the patient populations within their respective domains. The initial investment required to create the supporting technological infrastructure might be substantial, but such a system would help to reduce the spiralling costs of regulatory compliance.