Clinical Trials Face New Enforcement Models

The Food and Drug Administration is moving forward on several fronts to bolster assurances of the integrity and quality of data compiled in clinical trials, as well as the safety of study participants. The rising volume of overseas clinical trials, together with FDA's limited resources for onsite inspections, requires new approaches to clinical research oversight, particularly for studies utilizing more innovative, adaptive protocol designs.

Jill Wechsler

FDA traditionally inspects about four or five clinical sites listed on a new drug application (NDA), usually those conducting pivotal studies with high enrollment. These site visits most often occur after the trial is completed and too late to correct basic problems in the data or in how the study was conducted, explains Leslie Ball, director of the Division of Scientific Investigations (DSI) in the Office of Compliance of the Center for Drug Evaluation and Research (CDER). Even though DSI is conducting more bioresearch monitoring (BIMO) inspections and issuing more warning letters to investigators and sponsors, the globalization of clinical trials and several high-profile cases involving data fraud have raised questions about whether FDA's post-hoc inspection program can ensure quality clinical programs. Congressional Committees, the HHS inspector general and the Government Accountability Office continue to examine FDA oversight of clinical investigators, Institutional Review Board operations, and monitoring of foreign clinical trials, among other issues.

Calculating risk

One strategy for improving FDA's ability to assess sponsor and investigator adherence to good clinical practices (GCPs) is to utilize risk management to schedule BIMO inspections more efficiently and effectively. DSI is establishing an Inspection Risk Model that can better calculate how inspectional findings from a few sites may translate across the entire application, Ball explained at a meeting in March on Proactive GCP Compliance sponsored by ExL Pharma. Such approaches are becoming more common among sponsors, which increasingly use quality risk management systems internally to determine which sites to monitor to best prevent and detect deviations while trials are underway.

CDER's Site Selection Model aims to assess risk attributes based on the type of application (new molecular entity or supplement), the importance and complexity of a study (pivotal trial, adaptive design), and site location. Other key factors include a site's enrollment volume, efficacy data results, past protocol violations, number of serious adverse events, subject discontinuations, and important financial conflicts of interest. Now in a pilot phase, FDA expects to issue guidance explaining its algorithm for site selection as the program evolves.

Robert O'Neill, director of CDER's Office of Biostatistics, cited the adoption of risk-based auditing of clinical trials and site selection at the April Statistics Forum co-sponsored by FDA and the Drug Information Association (DIA). In discussing important innovations in drug development and regulation, O'Neill referred to this approach as "a big-ticket item."

Innovative challenges

At that meeting, O'Neill outlined a range of initiatives for improving drug discovery and the review process, including more pre-submission planning between FDA and sponsors and more guidance on designing innovative trials, trial enrichment, and handling missing data. Increased use of adaptive studies also requires new approaches in FDA oversight to address the greater prospect of producing false conclusions about treatment effect, Ball explained at DIA's March meeting on Adaptive Design for Clinical Trials. FDA considers it particularly important to ensure the integrity of adaptive trials, which may involve separate trial management and data assessment operations. And because it is important to detect problems before this kind of study is completed, the usual post-trial inspections may not be appropriate.

Consequently, FDA is proposing a pilot program that will provide prospective review of plans for a limited number of Phase III adaptive trials, Ball noted. The reviews will be done by CDER review divisions and statisticians, along with DSI, and will be combined with real-time inspections to evaluate the specific risks posed by the trial. The inspection team, which will include DSI and the Office of Biostatistics, will audit the operations and performance of independent "trial integrity" committees, data monitoring committees, sponsors, study statistical centers, among others.

The focus will be more on evaluating the process, and not the data, to provide feedback to sponsors in real time, Ball explained. The aim is not to issue enforcement reports, but to help ensure that the research is conducted in a way to avoid bias—and to give FDA officials a higher comfort level about the integrity of the study.

Sharing resources

The need to better evaluate increasingly complex and far-flung clinical trials is encouraging more FDA collaboration with its European counterpart in order to use resources more efficiently. FDA and the European Medicines Agency (EMA) have launched a pilot program to share information on drug applications and GCP inspections, with an eye to relying more on each other's inspection findings and to avoiding duplicative efforts.

The EMA coordinates GCP inspections for those innovative drugs and biologics seeking market approval under its centralized application procedure, explained Gabriele Schwarz, head of GCP inspection services at the German Federal Institute for Drugs and Medical Devices (BfArM), at the ExL Pharma meeting; Schwarz also is a member of the EMA GCP Inspectors Working Group, which sets standards, conducts training, and coordinates communication of inspection findings. Routine GCP inspections are conducted by the competent authorities of Member States upon the request of EMA staff. The inspectors issue findings, which are compiled into an integrated inspection report. The EMA also may request "triggered" inspections when concerns arise about unusual recruitment patterns, inconsistent results, or other issues, Schwarz pointed out. GCP inspections have been rising, prompting the EMA, as with FDA, to better target inspection resources.

The EMA-FDA GCP initiative, at its basic level, aims to share GCP inspection planning information and the results of inspection outcomes. While continuing to conduct their own inspections, both agencies may request that the other expand the scope of an inspection to cover additional areas of common interest. The two agencies are exchanging more information on inspection procedures and best practices, as well as GCP-related policy developments.

The next step involves a range of joint inspection activities. These include observational inspections, where one party conducts the inspection and the other observes, primarily to share information and to enhance training. For sequential inspections, the agency that first receives an application initiates the inspection, while providing planning and outcomes information to the other, with an eye to avoiding duplication. Parallel inspections involve applications of common interest; EMA and FDA discuss inspection questions and priorities, and each agency inspects different sites and shares findings under what Schwarz describes as a "divide-and-conquer approach" that makes the best use of resources.

Joint inspections are conducted by a team of inspectors from both sides of the Atlantic and are valuable for gaining a better understanding of each other's policies and practices. Although the visits are planned and conducted jointly, each authority writes up its own report. The two or three joint inspections so far reveal fairly similar ethical and scientific standards, Schwarz observed, but "relevant differences" in inspection procedures and regulatory framework. Summaries of inspections often differ as to major and minor findings, and FDA and EMA inspectors have very distinct "closeout" procedures following inspections.

These activities and joint training programs aim to establish more similar GCP standards and inspector practices. More extensive harmonization, though, will have to recognize differences in European and United States inspection goals and practices. Sponsors still receive different inspection reports from FDA and EMA and have to respond separately to each one, but these may become more similar in the future.

Ensuring quality

Even with more collaboration and cooperation, regulators at FDA and EMA will never be able to conduct enough inspections to ensure the proper conduct of all clinical trials. FDA officials thus are urging sponsors to establish a Quality System/Quality Risk Management approach to managing and monitoring clinical trials. This involves building quality into clinical research operations so that problems can be detected and corrected as close to real time as possible, Ball explained at the ExL Pharma conference.

A Quality Risk Management approach should focus on key parameters of risk to trial integrity and data quality, as well as to subject safety and protection, Ball pointed out. "Building in" quality and integrity to clinical trials involves establishing a quality-based rationale for protocol design and statistical analysis in the clinical plan, plus close attention to proper trial governance and operations and to procedures for handling data. Combined with FDA's risk model for targeting BIMO inspections, this quality approach will allow FDA and sponsors to set priorities for using resources more appropriately.

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