Printer Friendly

The growth of adaptive designs: adaptive trials have the potential to change the drug development process.

THE START OF THE 21ST CENTURY saw the pharmaceutical industry facing one of its toughest challenges to date. Decreasing success in bringing new compounds to launch, coupled with spiraling development costs, threatened to hinder the development and availability of new treatments and destabilize the industry. The race was on to replenish portfolios with successful compounds ahead of the looming patent expirations for various blockbuster compounds (which have historically driven company profits). The recent global economic crisis further accentuated this issue. The dire predictions of yesterday have now become today's reality, and the industry is very much aware of the crisis at hand. Reorganizations, lay-offs, downsizing, off-shoring and budget cuts are evident across the entire industry, as companies aim to reduce costs and increase efficiencies in their approach to drug development.

The crisis has far-reaching implications; the pharmaceutical industry is an industry that society cannot afford--or allow--to disappear. Perhaps the structure of the industry will be transformed, but ultimately society is the customer and the customer is dependent on an industry that must continue to deliver safe and efficacious medicines. Equally important, society, as the voice of payers, will demand that drug development be both efficient and cost effective, making the need for more efficient approaches to drug development paramount to our industry's survival.

[ILLUSTRATION OMITTED]

Even while predictions of difficult times were being voiced by management, other activities were quietly occurring across the industry. Initially limited to small pockets within companies, groups of individuals were starting to advocate a new approach to clinical trial design. This approach, called adaptive designs, is a flexible approach to design where trials are conducted in stages and interim data is used to plan the next stage of the trial. These are not ad hoc changes to the design, but are adaptive fry design--using an approach that emphasizes upfront planning.

The crisis of the industry was also felt at the FDA, leading to the 2004 Critical Path Initiative, which called for a review of the drug development process, from bench to launch, with the goal of identifying improvements that would lead to a cheaper, faster, improved process.

What was initially a debate about the merits of the methodology has been replaced with understanding gained through experience, and leading to a cautious encouragement and acceptance of adaptive trail designs. The stage has been set for a greater application of adaptive trial designs within clinical development strategies and greater utilization across the industry.

The Problem with Traditional Trial Designs

Rob Hemmings of the MHRA stated at the EFPIA/EMA workshop on adaptive designs in 2007 that "the cost of drug development is not in the cost of success but in the cost of failure." The high attrition rates and consequently wasted study costs, currently seen in Phase III trials, where failure rates approach 50%, lead to opportunity cost and reduced resources available to fund other assets in development.

Some of the reasons speculated for this high attrition rate have been 1) inappropriately sized trials, 2) progressing compounds that were unlikely to succeed, and 3) failing to understand the dose response adequately, leading to the wrong dose(s) or/and population being selected for Phase III. This is echoed further by drugs being withdrawn from the market for safety due to doses being too high.

Benefits of Adaptive Trial Designs

Adaptive designs can be useful in these situations. While adaptive trial designs can range from the simple to the complex, dramatic improvements to risk-management can be achieved through routinely applying even simple adaptive approaches. Many large-scale clinical outcome trials in Phase II and III provide an opportunity for interim analyses, either for sample size re-estimation (SSR) (1) or an early futility stop. In any large trial it is conceivable that the assumed data variability and/or treatment effect, derived from the literature or historical data, differs from the data actually observed in the trial itself. SSR allows for adjusting the sample size in light of the observed data. Furthermore, pre-specified futility analyses can be applied to either drop a particular treatment arm or to stop the study altogether if the probability of ultimately reaching treatment success falls below a pre-specified level. SSR therefore addresses the planning dilemma of uncertainty about size by allowing those initial planning assumptions to be revisited during the trial, and adjusting the sample size accordingly, so we ultimately achieve an "appropriately" sized trial.

This is not just important but critical in Phase III, where an underpowered trial may fail to achieve statistical significance, and require a further study. For a potentially effective drug the additional cost of another study can be a severe setback for the sponsor. The alternative, an unnecessarily large trial, is more expensive and takes longer to complete. Both situations are clearly not ideal for the sponsor and, more importantly, are not ideal for the patient population. Both lead to delays in bringing an effective medicine to market as quickly as possible. Preplanned SSR thus helps to mitigate against the possibilities of either an under- or over-sized trial.

Perhaps the most powerful application of adaptive designs lies in the area of dose-finding. (2)Model-based designs employing more than just two or three active doses are far superior to traditional parallel group designs in identifying the correct dose(s) to take forward into confirmatory regulatory trials. Importantly, through adaptive approaches, an improved understanding of the dose response can be achieved by allowing more doses to be studied without necessarily increasing the overall cost of the trial, when compared to a traditional design. Within a dose-ranging trial this may translate to dropping doses from further stages in the trial when they are shown to be ineffective. Or alternatively, changing the randomization allocation so fewer subjects are allocated to doses that are less effective, while increasing allocation to the more effective doses being studied. This approach leads to a more efficient allocation of resources within the study--a clear demonstration of the adaptive principle, which strives to achieve maximum information value per resource unit invested. Such practices also benefit the patients in the trial, as they have a higher chance of being allocated to effective treatments than they would with a traditionally designed trial.

One of the more promising opportunities for adaptive designs, beyond individual studies, lies in their potential to streamline development strategies. Decision points can be created within trials that allow objectives traditionally studied in two trials to be combined into one. For example, in early development there may be the opportunity to combine Proof of Mechanism (PoM) with Proof of Concept (PoC), or combine PoC with dose ranging. In the later case, the interim decision point declares POC and the trial continues to explore the dose range in further detail. More complex in application is the seamless II/III design, which is considered a confirmatory study by agencies. For this type of approach, early interactions with agencies are highly recommended. Although more complex, this approach can add significant value to development approaches for orphan indications, rare diseases and pediatrics where study populations are limited.

Adaptive approaches are not limited to large-scale trials. There is also significant opportunity for application in very early stage clinical trials. For example, first in human and multiple ascending dose trials can be designed using a continuous reassessment method (CRM). This is particularly relevant when there is a strong interest in correctly and efficiently establishing the maximal tolerated dose. Traditional Phase I designs may be easier to conduct, but tend to underestimate the MTD. In particular, CRM is often seen as a method for FTIH studies in oncology and as a replacement for the traditional 3+3 design used in this area. Furthermore, in Phase I designs it is possible to include efficacy biomarkers. Although efficacy biomarkers with early readout may not have high positive predictive power, they may be used as a necessary condition in a decision rule, which might allow an early indication of futility for parts of the dose-range.

Opportunity Spotting

For maximum impact, a systematic review of ongoing and future clinical development programs should be completed to seek out opportunities to improve decision-making efficiency by applying adaptive approaches. This should cover the entire spectrum from first-in-human to confirmatory registration studies. Once the opportunities are identified, competing designs should be compared and contrasted by running simulations across different scenarios to establish the operating characteristics of the different designs, under a variety of assumptions. (3) This will help identify the design that will provide the best information value per resource unit invested. The goal is to achieve conclusive answers to key research questions sooner, avoiding rework wherever possible. Simulation-guided clinical trial design is part of the adaptive planning process and should be the norm, whether the design ultimately chosen has adaptations built into it or not. The main value of clinical trial simulation is that it forces a more intense and earlier discussion between the relevant experts, and requires precise definitions of the key research questions, within the clinical trial, and across the entire clinical development plan. Exploring adaptive approaches requires additional upfront thinking time and discussion. However, the investment in time, combined with the value of simulations, ultimately leads to better design choices. Simulations replace subjective assessments and personal design preferences by providing quantification and greater insight into design performance. This ultimately makes the selection of the final design easier.

When looking for opportunities for adaptation it is important to understand at what time-key endpoints will read out. The earlier the readout, the easier its integration into an adaptive decision rule. Even if the primary endpoint reads out late relative to the overall duration over which the trial will be open to recruitment, there still is ample opportunity to use earlier observations and their predictive value to build decision rules. It is important to take the predicted recruitment speed into account as a design variable during evaluation and planning.

Implementation

Examining data as the trial progresses and basing decisions on these interim examinations comes with the risk of violating the integrity and validity of the trial. It is very important to plan the design and the implementation of the data capture, analysis and decision-making infrastructure. An independent data monitoring committee may well review both safety and efficacy data, it is very important that information that might unblind investigators or patients be restricted to the data monitoring committee and not released to any party involved in the conduct of the clinical trial. Implementation also presents challenges when teams are forced to work with systems that were originally designed for inflexible traditional designs. Nevertheless adaptive trials can be managed through workarounds to these systems, though this can prove to be labor intensive.

Past advances, such as EDC and the evolution of new statistical software for designing adaptive trials, imply that the future holds a similar promise in terms of new technology to support adaptive trials. Technology that provides firewalls and integrated solutions for EDC, dynamic randomization and drug supply management will make adaptive trial execution easier and more scalable across industry.

Clinical drug development involves decision-making on a portfolio, program and clinical trial level. In an environment where the cost for clinical research is spiraling out of control, and the number of newly approved drugs is not keeping pace with the ever-increasing investment, a key question is how to invest scarce resources and accelerate the development of promising novel therapies while eliminating inefficacious or unsafe programs. Aiming for improved efficiencies in clinical drug development is useful from many perspectives: there are patients within the trial who might not wish to be allocated to unsafe or inefficacious treatment options, future patients who are awaiting better treatments, drug developers (and ultimately society) who all benefit from investing scarce research funds in the best possible way.

A systematic review of clinical development programs typically yields a number of opportunities for adaptive designs in trials ranging from first-in-human to confirmatory Phase III and Phase IV studies. Upfront planning, simulation-guided clinical trial design and careful implementation are the keys to making adaptive designs a standard feature in clinical research. European and U.S. regulatory agencies have both issued guidelines on adaptive designs, illustrating the increasing interest in this important approach: EMA Reflection Paper in 2007 and draft FDA guidance in 2010. (4), (5)

Adaptive trial designs have application across multiple disease areas and across all phases of development. Several illustrations for how these approaches can benefit drug development and the industry at large have been provided here. Experience with the methodology and execution of these trials continues to mature, and is now converging with industry's need for greater efficiency. The stage is set for adaptive designs to play an increasingly important role in shaping the transition of industry and the process of drug development.

References

(1.) Chuang-Stein C, Anderson K, Gallo P, Collins S. Sample size reestimation: a review and recommendations. Drug Information Journal 2006; 40:475-484.

(2.) Bornkamp B, Bretz F, Dmitrienko A, Enas G, Gaydos B, Hsu CH, Konig F, Krams M, Liu Q, Neuenschwande B, Parke T, Pinheiro J. Roy A, Sax R, and Shen F. Innovative approaches for designing and analyzing adaptive dose-ranging trials. Journal of Biopharmaceutical Statistics 2007; 17:965-995.

(3.) Dragalin V. Designing, monitoring, and modifying an adaptive trial. Am Pharm Outsourcing 2008; 9 (5): 12-16

(4.) Committee for Medicinal Products for Human Use (CHMP). Reflection paper on methodological issues in confirmatory clinical trials planned with an adaptive design. CHMP/EWP/2459/02 [online]. Available from URL: http://www.emea.europa.eu/pdfs/human/ewp/245902enadopted.pdf [Accessed 2011 March 4]

(5.) Food and Drug Administration. Draft Guidance for Industry - Adaptive Design Clinical Trials for Drugs and Biologics. Available on http://www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/ucm201790.pdf [Accessed 2011 March 4]

By Judith A. Quinlan

Aptiv Solutions

Judith A. Quinlan is senior vice president of trial design implementation at Aptiv Solutions. She can be reached at judith.quinlan@clinresearch.com.
COPYRIGHT 2011 Rodman Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:THE GROWTH OF ADAPTIVE DESIGNS
Author:Quinlan, Judith A.
Publication:Contract Pharma
Date:May 1, 2011
Words:2317
Previous Article:Oral dosage formulation development: achieving faster formulation of solid oral dosage forms for FIH supplies.
Next Article:1,000 Talents: assessing the impact of China's talent schemes on innovation in global life sciences.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters