spacer
home > epc > spring 2012 > evolution or revolution
PUBLICATIONS
European Pharmaceutical Contractor

Evolution or Revolution



The current economic pressures on the industry are now creating a unique opportunity within R&D to consider the industrialisation of the adaptive trial design principle.

The Industrial Revolution was a major turning point in history that saw a shift from an agricultural-based manual labour economy to a machine-based manufacturing economy, which started in Great Britain, before sweeping throughout western Europe, North America and eventually affecting most of the world. This article draws parallels between the Industrial Revolution and recent developments in the pharmaceutical industry, and examines how advances in technology affecting the design and execution of clinical trials have become the springboard for a greater uptake of adaptive clinical trial designs. The full potential to reduce clinical development costs can be achieved through improved efficiency in decision making, unleashed by industrialising approaches originally developed in the context of adaptive trial designs. Specifically, these approaches can be applied broadly across design and execution of the entire spectrum of the clinical development space to all trials, with and without adaptive design elements.

Birth of the Capitalist Economy

This shift would not have come about without a number of crucial enablers:

  • Technology facilitating the production of mass production tools, as well as facilitating a scalable and increasingly widespread availability of energy, ultimately culminating in gas and electricity networks
  • Technology facilitating a scalable and increasingly widespread network of transport, including the application of the steam and combustion engine to shipping, rail and motor transport, followed by an expanding network of canals, railways and roads
  • An organised approach to the distribution and spreading of knowledge, through learned societies, and the emergence of a newly skilled mobile workforce

Response of an Industry under Stress

Fast-forwarding to the 21st century, there are considerable challenges to the pharmaceutical industry. These include the patent expiration cliff, which is causing the loss of billions of dollars in revenue with blockbusters going off patent, and the tightening of government healthcare budgets and downward pressure on drug prices. At the same time there is an unsustainably high failure rate in costly late stage Phase 3 clinical development projects.

In an attempt to reshape and redefine their businesses, many large pharmaceutical companies have reorganised their R&D operations, focusing on speciality areas combined with an emphasis on efficiency savings and productivity improvements. The call to do more with less has never been stronger. These combined pressures have not come without disruption, lay-offs and a workforce suffering from reorganisational fatigue. Nevertheless, the metamorphosis of the industry may not yet be complete, and the challenge to be more efficient and increase productivity remains a high priority.

The era of the blockbuster is over, and companies are in all likelihood going to need to register more compounds to replace the loss of revenue from single blockbuster drugs. In that context it is encouraging that the FDA cleared 30 new treatments in 2011 – the highest number since 2004 (36) and an important increase compared to 2010 (21).

The economic pressures on companies continue to focus attention not only on identifying potential new candidate medicines, but also on the need for increased productivity and more efficient approaches to clinical development.

In Search of Improved Efficiency and Productivity

The pharmaceutical industry continues to explore multiple options to reduce costs. These include off shoring certain activities to low cost centres and reducing internal overheads by relying more on external providers for services, also allowing for volume discounts achieved through preferred supplier arrangements. However, this has been complemented by a more radical, innovation based revolution in the making: the application of the adaptive principle to the clinical development paradigm, from planning and designing to implementing and executing clinical trials. This revolution has the potential to reduce the rate of late stage clinical development failures, and thereby allow the reallocation of funding to more promising compounds in the portfolio.

Some of the factors that have been proposed as contributing to the high rate of Phase 3 failures include:

  • A lack of success due to inappropriately sized trials
  • Late termination of compounds that were unlikely to succeed
  • Incomplete understanding of the dose response, leading to an incorrect choice of dose(s) entering Phase 3 trials. This issue does not only lead to Phase 3 failures, but can also result in post-marketing withdrawals of unsafe drugs for which the minimal efficacious dose has not been appropriately established, or pricing challenges, if later studies establish comparable efficacy of a lower dose than the one tested in Phase 3

A Paradigm Shift in Clinical Development

The traditional approach to clinical trial design works with the best assumptions available prior to the start of the trial, without the ability to adapt later on. The trial is conducted, and only at the end of the trial is the data analysed. If the initial assumptions are correct and if the correct populations and treatment regimens are studied, the trial may turn out to be a success. But if the initial assumptions are proved to be incorrect, the trial may fail, leading to rework, additional investment and a protracted clinical development process.

By contrast, the adaptive paradigm starts with a scenario analysis of options: simulation-guided clinical trial design formally compares the operating characteristics of a traditional design with trial designs that enable the trial to be adjusted, if the observations within the trial suggest that some of the initial assumptions differ from what is observed during the conduct of the clinical trial. Software packages are now available to apply modelling and simulation techniques to formally establish in silico how different clinical trial designs perform. Adaptive designs enable a trial to react to discrepancies between initial assumptions and emerging observations during the conduct of the trial. For instance, if the data variability on the primary endpoint turns out to be different from the initial assumption, sample size re-estimation techniques can be applied during an interim analysis, increasing the probability of declaring success if the investigational compound does provide benefit, or by facilitating the stopping of a trial early for futility should the compound not provide the minimal acceptable benefit. Regardless of whether an adaptive design is chosen at the end of this exercise or not, the understanding of the operating characteristics of the design taken forward into production is greatly improved by simulating the performance of a particular design under different conditions in silico. None of this would be possible without the considerable advances in computing power over the last decade, the emergence of commercially available simulation software, and the development of a pool of experts trained in modelling and simulation applications.

The adaptive paradigm then calls for the timely collection of data. It took a decade for electronic data capture (EDC) to become a mainstream feature in the conduct of clinical trials. However, this breakthrough in technology also became the first enabler in expanding the opportunities for utilising adaptive approaches. Through the introduction of EDC, interim analyses have become more possible, providing the opportunity for the trial to adapt to emerging data. Interim analyses are an integral component of an adaptive design that requires the trial to be conducted in stages. The approach uses accumulating data collected during the trial to make decisions about the conduct of future stages of the trial. For example, following an interim analysis, a decision may be made to re-estimate the size of the trial, stop the overall trial for reasons of futility or efficacy, drop ineffective treatments from future stages in the trial, or perhaps change the randomisation allocation in favour of more effective treatments/ populations in the trial. These are not ad hoc changes to the trial design but are design changes that are pre-specified and planned in advance.

This concept of real-time learning is widely applicable from Phase 1 to Phase 3, and we have seen regulatory agencies both in the US and Europe publish guidance documents on adaptive designs. These support the application of the adaptive paradigm in ‘learn’ trials and lay the ground rules for applying adaptive design elements to ‘confirm’ trials.

Benefits of Adaptive Designs

Some see and promote the promise of adaptive approaches as their potential to reduce the cost of trials due to reduced sample sizes. This is an oversimplification. The real value of the adaptive approach lies in increasing the information value, allowing for improved decision-making by ensuring at the end of the trial the correct decision can be made as to how to proceed to the next step. The costeffectiveness in the approach is to provide increased information value per dollar spent on the trial and reducing rework. Increasing the information value of studies enables management to be more confident in their decision making, leading to reduced rework, more confidence in selecting treatments to carry forward into future studies, and greater confidence in declaring futility decisions earlier in development. Even simple adaptive approaches, such as sample size re-estimation, can be beneficial cost-effective strategies akin to building in trial insurance through the design to mitigate against either an undersized or oversized trial, leading to increased chances of success for effective compounds.

The uptake of adaptive designs over the past decade has faced many challenges. Nevertheless, the initial debate about adaptive approaches has been replaced by a cautious acceptance of the methodology and an increasing number of adaptive trials have been conducted. However, at this point, the clinical operation of an adaptive design still largely relies on solutions using systems that were not optimised for real-time learning. Imagine by contrast that the operational aspect of adaptive designs can be conducted based on monitoring techniques, data management systems and data monitoring committee structures that are tailored to enable real-time learning combined with systems that support dynamic randomisation, drug supply management infrastructures that minimise drug wastage, while enabling adaptive treatment allocation in multi-centre trials across a large number of doses, and firewall systems that optimally support the validity and integrity of the ongoing trial.

Scalability: Moving Beyond Work-Around Solutions

To push the adaptive principle to achieve scalability, it is necessary to build and apply integrated systems, facilitating real-time learning and near real-time decision making at a large scale. These systems need to include all aspects of clinical execution, reducing manual intervention wherever possible and ultimately reducing the cost of clinical trial conduct. It is key that the integrated system also de-pressurises the process through the introduction of analysis automation, and partial automation to generate the data monitoring committee (DMC) report, thereby removing the burden of producing routine elements of production like tables, listings and graphs through automation, enabling statisticians to focus on the main components of interims and DMC reporting. However, for this to be executed successfully, it must also be combined with standardisation and up-front tools to facilitate a more streamlined and effective data cleaning process.

The integrated system will have to be supported by welldocumented processes and a workforce knowledgeable in the application of these processes. Expertise in this area has grown steadily over the last decade within pharma companies, which have established specialised groups to set standards, and the emergence of some CROs with a focus on the adaptive principle. One objective in streamlining the process has been to move away from a stop-and-go sequential data cleaning process that creates a substantial lag-time between the initial observation and the eventual cleaning of the data, to a close to real-time continuous data cleaning effort involving cross-functional review of the data.

Increased confidence in adaptive designs has grown over the last decade, illustrated by an ever increasing catalogue of publications on the topic. However, there is still a substantial gap between the potential of the approach and its current scale of application. For the adaptive principle to become truly scalable, the following enablers are proposed as requirements:

  • Technologies to evaluate trial design options – these technologies facilitate simulation-guided clinical trial design, supported by statistical and medical expertise to conduct scenario analyses of options, including recruitment modelling under different assumptions
  • Real-time decision making, enabled by replacing departmental optimisation with a focus on optimisation of the whole process and minimisation of handovers. This includes the creation of the following principles: integrated technology solutions that include EDC, randomisation, drug supply management and analysis automation connecting design and execution. Also essential is the development of tailored operating procedures that are specifically focused on enabling real-time learning, governing behaviours across the process from project management, data/site monitoring, data cleaning, analysis and DMC report generation, through to the conversion of interim decisions into seamless actions executed without disputing the conduct of the trial
  • Industry- and agencyaccepted solutions for controlling interim analysis communication flow
  • Organised approach to the distribution and spreading of knowledge, leading to emergence of a newly skilled workforce

Conclusion

The systems and processes described in this article for real-time learning are not meant as a means of tweaking or patch-working existing systems and processes. Rather, adaptive designs are defining a fundamental shift in the approach to clinical development through a change to both design and execution. Quantifiable evaluation of the risks and benefits to design options through simulation would ultimately become routine. The integrated execution environment required to support flexible approaches like adaptive designs will open the door to an expansion in the repertoire of designs available to industry. The availability of near real-time data will lead to a greater emphasis on analytics to transform data into more sophisticated study metrics and performance indicators, improving the efficiency and cost effectiveness of trial execution. The use of standards and aspects of report automation will spread beyond DMC reporting to all study reporting. Ultimately adaptive designs will not only reduce costs through improved decision making within the clinical development process: the systems and processes required as ‘must haves’ for adaptive execution will ultimately ensure the running of all clinical trials are more efficient and cost effective.

Read full article from PDF >>

Rate this article You must be a member of the site to make a vote.  
Average rating:
0
     

There are no comments in regards to this article.

spacer
Judith Quinlan is Senior Vice President of Trial Design Implementation at Aptiv Solutions, where she oversees the design and implementation of adaptive clinical trials. Judith worked at GSK for 10 years in both the UK and US, taking on roles of increasing responsibility. Judith is an active member of the Adaptive Design Working Group where she co-chaired the case study workstream that collected industry experience on the use of adaptive designs, the drug supply workstream, and currently co-chairs the monthly KOL lecture series on adaptive designs. Email: judith.quinlan@aptivsolutions.com
spacer
Judith Quinlan
spacer
spacer
Print this page
Send to a friend
Privacy statement
News and Press Releases

Huge CPhI & P-MEC China a direct result of surging manufacturing

Shanghai, 22 July, 2019:CPhI and P-MEC China – organized by China Chamber of Commerce for Import & Export of Medicines & Health Products (CCCMHPIE), Informa Markets, and Shanghai Sinoexpo Informa Markets – saw record numbers of attendees as the pharma market in China continues to grow. The three-day Shanghai event featured 3,200+ exhibiting companies, spanning the entire pharma supply chain, and provides a strong indication of the market’s overall growth patterns and future development.
More info >>

White Papers

New Temperature Profiles from a Global Study

Laminar Medica

Recently there has been an influx of temperature mapping studies to collect sufficient ambient data to develop temperature profiles.  Solid temperature profiles on specific shipping routes are required to ensure streamlined packaging and robust lane qualification. The most reliable method to develop temperature profiles is to capture the actual reading from shipments, as well as using the historical and meteorological data to increase the confidence level.
More info >>

 
Industry Events

CPhI & P-MEC India

26-28 November 2019, India Expo Centre, Greater Noida, Delhi NCR, India

As the pharma industry looks increasingly towards India for high quality, low cost pharma solutions, CPhI & P-MEC India is the ideal event for companies wanting to pick up on the latest trends and innovations the market has to offer. At CPhI & P-MEC India, you will meet the movers and shakers from India's pharma machinery, technology and ingredients industries, giving you a competitive advantage that will help grow your business.
More info >>

 

 

©2000-2011 Samedan Ltd.
Add to favourites

Print this page

Send to a friend
Privacy statement