How to Stay Competitive in 2015: Three Pillars of New Device Development

Manufacturers that implement adaptive trial design, health economics, and risk-based monitoring will have a competitive advantage.

Vicki Anastasi and James Eaton

Regulatory and reimbursement reform is moving rapidly on a global level. Regulators and payers expect real-world evidence that demonstrates efficacy, safety, and an improvement over standard-of-care relative to the cost of the device.

To stay competitive, medical device manufacturers must take a three-pillared approach to product development:

These pillars are not mutually exclusive, but they tackle different issues encountered when bringing a device to market.

Adaptive Clinical Trials

Collection of health economic data and risk-based monitoring work best in the framework of an intelligently designed trial. Adaptive trial designs, which use preplanned adaptations to a trial to correct initial incorrect assumptions, enhance the efficiency of device development by reducing the need to repeat trials that narrowly miss a clinical endpoint and eliminating unnecessary patient recruitment for overpowered designs. For small patient populations, this can reduce recruitment time by many months. Adaptive designs that incorporate economic models can ensure that a device launches with the right clinical and safety data, as well as a strong economic value proposition.

In the worst-case scenario, adaptive designs can also prevent unnecessary expenses by terminating trials that show no benefit early on over a predicate. Though an adaptive design will not be possible for all medical device studies, the upfront feasibility analysis that is required focuses the trial and improves sponsor decision-making even if an adaptive design is not used.

Although adaptive trial designs are not as well known in the medical device space, they are employed in 20% of trials in the pharmaceutical space and their prevalence is expected to increase. Adaptive design makes an interim data analysis possible and so can shorten a trial’s timeline, reduce operating and development costs, limit unnecessary patient exposure, and, importantly, generate higher quality data that enable sponsors to ask better questions and make better decisions.

Three types of adaptive design are particularly applicable to medical device trials. Sample size reestimation allows sponsors to use interim data to right-size their trial by adding or removing patients in order to achieve statistical significance. This is the most popular type of adaptive design for medical device trials, making up 80% of adaptive design submissions to FDA’s Center for Devices and Radiological Health from 2007 to 2012.¹ Adaptive group sequential designs, the second type, allow for early stopping for futility, efficacy, harm, or safety. Finally, seamless adaptive designs combine the pilot and pivotal studies into one trial and allow the same patients to be used for the entire duration of the trial.

Regulators are generally open to the use of adaptive designs in medical device trials when appropriate, as evidenced by the release of regulatory framework on the topic. The greater hurdle is obtaining organization-wide buy-in for the additional upfront planning time and infrastructure to design and execute these types of trials. Interim analyses require fast and efficient electronic data capture and cleaning. An independent data monitoring committee that controls data flow will minimize operational bias associated with interim analyses and maintain trial integrity.

Health Economics

Device reimbursement is increasingly contingent upon the presentation of solid health economic (HE) data that demonstrate a device’s value both in terms of patient benefit and in comparison to competitors.

Collecting HE data before reimbursement hurdles appear is not a common practice in the device industry. However, understanding how a device offers value and capturing relevant economic endpoints early can be vital to maximizing a device’s health economic story. Data collection strategies differ when justifying a technology that may not improve efficacy, but does reduce resource use and costs through demonstrated safety or patient usability improvements.

To avoid the unnecessary expense of a separate study, the best strategy is to collect HE data during regulatory trials or prospective registries. Measuring product attributes against existing comparators during preclinical development can help evolve the technology and clarify what additional clinical and economic data are needed to support the value story.

Early evaluation of HE data cultivates a better understanding of a device’s value proposition and thus a more informed discussion about the likely price that the market will bear. Naming a competitive price is particularly important in countries outside the United States that have adopted health technology appraisal with cost-effectiveness. Although only the United Kingdom specifies a cost per quality adjusted life year threshold, representing the amount the National Health Service is willing to spend on a new intervention, many other countries have implicit thresholds beyond which a product is highly unlikely to be reimbursed.

Health economic arguments must be adaptable to different reimbursement agencies, as there is no singular solution appropriate for every jurisdiction. Building long-term relationships with HE advisors can cultivate greater understanding and appreciation of the value of the product on a global level. Onboarding advisors early in the process will ensure that manufacturers receive the right HE advice early on in the development cycle.

Risk-based Monitoring

As clinical trials become more complex, clinical investigators are burdened with more complicated procedures, frequent protocol amendments, and lengthier case report forms. At the same time, clinical research associates (CRAs) are unable to properly train and communicate with investigators because they spend the majority of their time at trial sites manually verifying all source data. These issues can lead to delayed recruitment and protocol non-compliance, both of which lengthen trial duration and slow time to market.

Adopting a risk-based approach results in a more efficient deployment of resources that allows for a more meaningful CRA interaction at each trial site. Traditional clinical trial onsite monitoring practices are frequency-based and rigid, with most employing 100% source data verification (SDV) at all times regardless of the risk profile of the trial or site. In pharma trials, employing 100% SDV consumes an estimated 25–30% of the total cost of a Phase III trial² and detects random errors that are not likely to significantly impact the outcome of a trial.³ Though these numbers may differ in magnitude for device trials, it is generally accepted throughout both industries that 100% SDV is extremely time-consuming, can provide a false sense of data integrity, and may not be as effective as intended.

Regulators and pharmaceutical industry working groups support centralized monitoring strategies that reduce SDV, including risk-based monitoring.⁴ According to FDA, a risk-based approach does not result in any less vigilance in site monitoring and is “dynamic, more readily facilitating continual improvement in trial conduct and oversight.”^4c

Taking a statistical approach to site monitoring can improve the effectiveness and efficiency of the process by helping a CRA prioritize and guide site visits. A novel approach called Verification by Statistical Sampling (VSS)⁵ uses a risk-based, adaptive monitoring algorithm to reduce the number of data points to be verified at each visit when appropriate. Each trial has a unique risk profile that is determined by factors such as study phase, objectives, protocol design, complexity, size, and endpoint. The risk profile dictates an acceptable quality limit (AQL) that then defines the number of acceptable errors in site data. Sites whose data are near or below the AQL require fewer monitoring visits, while sites whose data are above the AQL require more visits and greater scrutiny, in some cases 100% SDV.

VSS and a risk-based approach to site monitoring can reduce the cost of premarket, postmarket, and IVD studies while effectively managing trial risk and without compromising data quality. For example, in a hypothetical premarket cardiovascular device study involving 194 subjects across 30 sites in a single country, with a duration of 44 months from start of activity to final study report, deploying VSS can reduce monitoring visits per site by 63%, from eight to three, and can cut monitoring costs by 46%.⁶

There are significant technology requirements for implementing an offsite, centralized risk-based monitoring approach.⁷ Sponsors will require the ability to adjust risk profile and monitoring approaches based on incoming data. Clinical investigators and study personnel will benefit from a platform that includes training, performance support, and study management tools.

An Integrated Approach

Increased regulatory burden and the demand for a compelling health economic story are signs of a changing tide in the medical device industry.

Adaptive design, health economics, and risk-based monitoring can provide the opportunity for a manufacturer to make informed commercial and development decisions without compromising trial integrity. When integrated, these approaches can ensure that technology not only launches with the right clinical and safety data, but also a strong economic value proposition.

Manufacturers that are able to implement these approaches will not only have a competitive advantage, but also the confidence that their device truly benefits patients.

Vicki Anastasi is Vice President and Global Head of Medical Device and Diagnostics Research at ICON plc.

James Eaton is Director of Health Economics at ICON plc.

References

1.Gao, Y. C., DIA/FDA workshop on adaptive designs. 2012.

2.(a) Eisenstein, E. L.; Lemons, P. W., 2nd; Tardiff, B. E.; Schulman, K. A.; Jolly, M. K.; Califf, R. M., Reducing the costs of phase III cardiovascular clinical trials. Am. Heart. J. 2005, 149, 482-488; (b) Funning, S.; Grahnen, A.; Eriksson, K.; Kettis-Linblad, A., Quality assurance within the scope of Good Clinical Practice (GCP): what is the cost of GCP-related activities? A survey with the Swedish Association of the Pharmaceutical Industry (LIF)'s members. Qual. Assur. J. 2009, 12, 3-7.

3.Sheetz, N.; Wilson, B.; Benedict, J.; Huffman, E.; Lawton, A.; Travers, M.; Nadolny, P.; Young, S.; Given, K.; Florin, L., Evaluating source data verification as a quality control measure in clinical trials. Therapeutic Innovation & Regulatory Science 2014, 48, 671-680.

4.(a) Risk-adapted approaches to the management of clinical trials of investigational medicinal products. MDC/DH/MHRA Joint Project: 2011; (b) "Position paper: Risk-based monitoring methodology;" TransCelerate Biopharma, Inc.: 2013; (c) Guidance for industry: Oversight of clinical ivestigations - A risk-based approach to monitoring. U.S. Food and Drug Administration (FDA): 2013; (d) Reflection paper on risk based quality management in clinical trials. European Medicines Agency (EMA): 2011.

5.(a) Grieve, A. P., Source data verification by statistical sampling: issues in implementation. Drug. Inf. J. 2012, 46, 368-377; (b) Grieve, A. P.; Fardipour, P.; Zippel, E. "AptivInSite and verification by statistical sampling (VSS): A novel approach to risk-based monitoring;" Aptiv Solutions White Paper: 2013.

6."An innovative adaptive monitoring methodology for medical device & diagnostic trials;" Aptiv Solutions White Paper: 2013.

7.Barnes, S.; Katta, N.; Sanford, N.; Staigers, T.; Verish, T., Technology considerations to enable to the risk-based monitoring methodology. Therapeutic Innovation & Regulatory Science 2014, 48, 536-545.

[Image courtesy of STUART MILES/FREEDIGITALPHOTOS.NET]