Avoiding Unnecessary Delays, Failures

The deployment of adaptive design approaches has been seen at the strategic / operational level of many pharmaceutical and biotech companies as a means to improve risk management, increase the speed and efficiency of their drug development programs, and more efficiently manage their budget and resources. Most importantly these approaches are seen as a mechanism to help identify failures early, while for potentially successful drugs they provide a strategy for bringing them to market earlier, leading to an increased return on investment (ROI).

Furthermore, the adoption of an adaptive design strategy across the product development process brings a number of important benefits; most importantly increased probability of success at various stages of clinical development. We are all too familiar with the worrying industry statistic that more than half of Phase II and phase III studies fail due to lack of efficacy; see Schuhmacher et al. [1]. Adaptive design trials offer the potential to change this industry statistic and dramatically increase the ability of pharmaceutical companies to successfully bring more effective treatments to the market. Companies that adopt a comprehensive adaptive design strategy across their product pipeline will make better development decisions and ultimately bring effective products to the market more quickly.

Adaptive design enhances success for dose-ranging studies

When the dose–response relationship is not well defined then designing dose-ranging studies can be challenging. Suboptimal trial design may lead to repeat trials or, worse, selecting the wrong dose for a Phase III trial.

In some cases, the risk introduced by suboptimal dose-finding trial designs can be extreme. Mullard  [1] reported one development program in which dose selection unexpectedly required three trials over three years, and tens of millions of dollars in unplanned costs. In this case, pairwise comparisons were employed in sequential trials. It appeared that the appropriate dose level was overestimated in the planning stage. The first trial tested three doses ranging from 80 to 160 mg against placebo. All active doses were effective, but no efficacy difference was detected. A second three-dose trial, ranging from 40 to 120 mg against placebo, also found no difference. It took a third trial examining three doses from 2.5 to 40 mg against a placebo to detect the increasing part of the dose response curve.

The traditional pairwise comparison approach to dose finding lends itself, in economic terms, to assessing just a few drug doses in relatively large treatment groups. Due to the limited scope of a dose–response relationship that can be explored with just two or three subgroups, it is not uncommon to generate poor dose-finding data.

In fact, Sacks et al. [2] reported  that failure to determine the appropriate dose of a drug for clinical use was the leading cause for FDA non-approval, affecting nearly 16% of all unsuccessful first time new drug applications, and significantly contributed to delays in new drug approval. The report suggested that using adaptive trial designs could help to reduce number of patients allocated to non-informative, unsafe and clinically ineffective doses; thereby optimizing patients’ exposure by determining and allocating patients to clinically better doses, maximizing information gained and minimizing study expenses.

ICON’s ADDPLAN®  is a software for designing, simulating, and analysing adaptive trials. In 2013, ICON incorporated Multiple Comparison Procedures (MCP-Mod) innovative dose finding approach in a simple user interface, allowing a standardized design and reporting approach to adaptive study designs. MCP-Mod is particularly advantageous when used at an interim analysis because resolving uncertainty regarding the dose-response model early can support optimized patient allocation for the remainder of the study, reducing the risk of poor dose selection. MCP-Mod used in the interim analysis for futility testing controls the probability of continued investment into non-working drugs at a predefined level, while asserting a high power for effective drugs. ICON’s ADDPLAN® software is the first to incorporate adaptive MCP-Mod, enabling the most powerful strategies for dose ranging study design to be used together. ADDPLAN® utilizes the power of model-based design and analysis with MCP-Mod to optimize dose-finding studies and support team discussions with its extensive simulation capabilities in all phases of the trial to:

  • Determine optimized decision rules prior to the study
  • Enable interim decision making based on conditional predictive simulations
  • Support sensitivity analysis under model uncertainty after completion of the study

ADDPLAN®  neo is the latest version of ADDPLAN® that was released in Sep 2018. This version combines ADDPLAN BASE, MC and DF modules on one platform, thereby allowing the ease of use for designing, simulating and analysing trials.

Furthermore, ADDPLAN®  neo allows researchers to link their R code with ADDPLAN® and benefit from simulating and executing for MCP-Mod and nonlinear regression procedures.

Requirements for using Adaptive Trial Design

It is important to ensure that adaptive trial design utilizes appropriate adaptive software to maintain the validity of the trials; that is to utilize correct statistical inferences. ADDPLAN®  neo is an adaptive statistical software that has been used since 2002 by industry and academia. ICON  has conducted over  250 adaptive trials using ADDPLAN® software. The other important element in conducting adaptive trials is to maintain the integrity of the trial without introducing any operational bias. Figure 1 describes the hierarchical structure of what is required to maintain the integrity of the adaptive trials.

At ICON, harmonizing technology, processes and people are key components of a successful adaptive trial. ICON approach resulted in:

  • Facilitation of data driven monitoring and real time data cleaning, which in turn enables timely database lock for interim reporting. Critical features for the effective execution of adaptive trials; specifically designed to meet regulatory agency requirements to minimize operational bias and maintain trial’s integrity. That is ensuring that information is only made available to the appropriate trial participants
  • A key starting point in the deployment of effective firewalls is the development of, and adherence to, a defined set of standard operating procedures (SOPs). These SOPs govern workflow processes as to who sees what; when they see it; and they ensure that the appropriate data is delivered to the appropriate set of trial participants
  • FlexAdvantage, is created to limit the number of connection points in order to provide a flexible and seamless adaptive environment. The utilization of an integrated technology platform is aimed at efficiency in execution, and at minimizing handovers in data transfer.

References

1 Schuhmacher, Alexander, et al. “Changing R&D models in research-based pharmaceutical companies” J Transl Med. Vol. 14 (2016): 105.

2 Mullard, Asher. “Regulators and Industry Tackle Dose Finding Issues.” Nature Reviews Drug Discovery Vol. 14 (June 2015): 371-372

3 Sacks, Leonard V., et al. “Scientific and Regulatory Reasons for Delay and Denial of FDA Approval of Initial Applications for New Drugs.” Journal of the American Medical Association Vol. 311 No. 4 (Jan. 22-29, 2014): 378-384

4 https://www.iconplc.com/innovation/flex-advantage/