Quality Approaches Key to Transforming Research

Empty drug pipelines and ever-rising clinical research costs are prompting extensive soul-searching for ways to make clinical trials more productive and better able to support drug discovery and development. Everyone agrees on the need for more relevant studies, more efficient research methods, broader public support for clinical trials, a stronger pool of investigators, and more rational regulation and oversight of research activities. There's a healthy debate over how to get there.

Jill Wechsler

Food and Drug Administration officials promote a shift towards a quality-based approach to protocol design and study management, along with revisions in specific policies to support risk assessment. Leaders of the US biomedical research community contemplate more "transformative strategies" that integrate clinical research into the nation's healthcare system so it can tap electronic health records (EHRs). That would make it easier to identify patients suitable for clinical trials, record study data accurately and quickly, and document how new therapies can improve outcomes and reduce costs.

Evaluating risks

Adoption of quality-by-design (QbD) and quality risk management methods for clinical trial management is the current mantra at FDA. This strategy was discussed by agency officials and research executives at a workshop in August 2011 organized by the Clinical Trials Transformation Initiative (CTTI), a partnership between FDA and the Duke University Medical Center that has gained strong support from industry and the biomedical research community; future CTTI workshops will further explore the QbD approach to clinical research.

Anne Meeker-O'Connell, Associate Director of the Office of Scientific Investigations (OSI) in the Office of Compliance, Center for Drug Evaluation and Research (CDER), outlined the shortcomings of FDA's current oversight model for clinical trials, noting that lax internal controls by study sponsors often lead to significant data problems and systemic errors in applications. Inefficient practices, she noted, consume valuable resources and undermine research quality. Ken Getz of the Tufts Center for the Study of Drug Development emphasized the need to adopt a quality approach, starting with protocol design to reduce study complexity and thus curb the need for multiple amendments as the study progresses.

The next step is for sponsors to develop an integrated quality management plan (IQMP) as part of protocol development. Under an OSI pilot, sponsors identify important risks and quality objectives to guide implementation of the clinical program. Metrics could include the number of subjects that fail to meet inclusion/exclusion criteria or a percentage of subjects with inadequate informed consent. Failure mode analysis can identify for sponsors what could go wrong and why, and how severely such problems would affect patient safety, data integrity, and protocol compliance. IQMPs may be particularly valuable for ensuring quality data from multi-regional clinical trials and for managing updates to research protocols in response to new information, problems with drug supplies, and unexpected events.

A recent FDA proposal for reducing excessive and costly clinical trial monitoring illustrates the tenor of the risk-based quality approach. Sponsors feel they have to verify all data at every clinical site to avoid regulatory problems, Meeker-O'Connell acknowledged, even though FDA permits a variety of monitoring approaches based on risks to human subject protection and data integrity of a trial. To promote change, FDA published draft guidance in August 2011 that encourages sponsors to design monitoring plans based on key risk factors: the complexity of study design and study population, investigator experience, data capture systems, and the relative safety of the investigational product. Sponsors should perform a risk assessment to identify what could go wrong with critical data processes, what would be the impact, and whether such problems could be detected readily, Meeker-O'Connell explained at the October 2011 FDA Inspections Summit sponsored by FDA News.

FDA also seeks to improve the quality and value of adverse event reporting from clinical trials, with an eye to reducing reports on less important, low-risk events in order to focus on more consequential safety issues. FDA issued revised regulations in September 2010 for safety reporting on experimental drugs, which clarifies responsibilities of investigators and sponsors for reporting serious, unexpected adverse events. CTTI is further assessing thresholds for reporting serious AEs and ways to relate safety reports to underlying disease to help sponsors implement the new rules.

Internally, OSI is testing a risk-based model for planning its own clinical site inspections, initially for good clinical practice (GCP) inspections related to applications for new drugs and biologics; the plan is to extend this approach to oversight of institutional review boards (IRBs) and bioequivalence studies. This strategy also may be extended to OSI's collaboration with the European Medicines Agency on parallel/joint GCP inspections. The regulators are sharing information on practices and schedules in order to improve methods, gain confidence in each other's programs, and conduct more joint and parallel inspections to reduce redundant oversight.

Broader change

A quality approach to clinical trial management also should be part of a more comprehensive transformation of the US clinical trials enterprise, according to members of the Institute of Medicine Forum on Drug Discovery, Development and Translation. The forum has been weighing options for such change for several years and wound up its deliberations at a November 2011 workshop on "Envisioning a Future Clinical Trials Enterprise for 2020"—a final report is expected this spring.

The main conclusion of these experts is that clinical research needs to be linked more closely to the nation's healthcare system in order to attract sufficient financial support to facilitate the development of new, important medicines. Despite years of discussion, the envisioned "learning healthcare system has not really come together," commented Robert Califf of Duke University. Instead we have a process that "is slow, expensive, and often asks the wrong questions." Protocols are complex and burdensome, he said; over 90% of trials don't meet desired timelines, and many studies posted on clinicaltrials.gov are small and inconsistently utilize randomization, blinding, and data monitoring committees.

CDER Director Janet Woodcock noted that the demand for personalized therapies targeted to small patient populations illustrates the importance of integrating biomedical research into medical practice. Professional societies, payers, healthcare systems, and pharma companies all are important for establishing disease-based networks and supporting participation in and payment for clinical research; FDA, in turn, needs to expand concepts for assuring quality data from trials.

Amgen Senior Vice President Paul Eisenberg called for even more extensive reform of FDA's "onerous IND process," pointing out that the European Union's simpler notification system doesn't require a review of clinical trial materials and manufacturing processes at the research stage. US studies at multiple research institutions, he noted, are complicated by diverse contracting systems; variable informed consent and IRB procedures; and difficulties in managing privacy requirements. Central IRBs would help, as would harmonization of clinical trial applications, nomenclature, and electronic systems for reporting and assessing adverse events. Bryan Luce, Senior Vice President at United BioSource, urged "real regulatory reform" that permits earlier, conditional approval of new therapies based on evolving evidence, acceptance of new methods and "trust across regulators, manufacturers, and payers."

Linking up

Health system executives highlighted how health system data systems can provide an infrastructure to support clinical research. Richard Platt of the Harvard Pilgrim Health Care Institute and Harvard Medical School described the use of EHRs to identify and enroll children in a study on reducing obesity and to support a massive randomized study on how hospitals can identify strategies for reducing antibiotic resistant infections.

Another example is the Veterans Administration healthcare system, which has incorporated clinical studies into its healthcare program for years, according to Louis Fiore, head of the VA Cooperative Studies Program. Similarly, leading health plans around the country have joined the HMO Research Network, which now has some 20 sites at integrated healthcare systems that represent 10 million covered lives available for collaborative studies on vaccine safety, mental health, diabetes management, cancer, and other conditions. The aim is to conduct pragmatic, low-cost trials using EHRs to recruit, randomize, intervene and collect follow-up data, explained Karen Margolis of the HealthPartners Research Foundation. Centralized IRBs and template agreements on data sharing streamline operations, leading to strong patient retention, efficient adverse event surveillance, and ready dissemination of results.

The discussion ended optimistically as Douglas Cropper, CEO of Genesis Health System, a community-based health group in Iowa and Illinois with little involvement in clinical research, expressed a high interest in using EHRs to support trials and outcomes studies. He encouraged researchers to get health system CEOs and clinicians more involved by appealing to their business and clinical interests in improving best practices and better managing chronic disease, and by embracing quality improvement methods.

Jill Wechsler is the Washington Editor of Applied Clinical Trials, (301) 656-4634, jwechsler@advanstar.com