CAMBRIDGE, Mass.—The Collaborative Trajectory Analysis Project (cTAP) recently announced results from the largest multinational, collaborative study in Duchenne muscular dystrophy (DMD). The study indicates that both real-world data (RWD) and natural history data (NHD) are highly comparable to data from patients treated with placebo in multiple recent clinical trials.
The results, published recently in Neurology, demonstrate that use of RWD or NHD could supplement or potentially even replace a placebo arm in future DMD research.
“In 2017, Collaborative Trajectory Analysis Project (cTAP) initiated a collaborative study to compare natural history data from multiple clinical centers with placebo controls from prior clinical research in Duchenne muscular dystrophy,” says Dr. Susan J. Ward, executive director of cTAP. “Clinical regulators and drug developers had raised well-founded concerns about risks of bias with external controls in DMD—especially for effort-based outcomes. For example, might boys simply try harder on walk test outcomes in clinical trials than in clinical practice?”
“We designed our study to address this empirically, using the best data available. We wanted to use data and evidence to assess concerns that RWD/NHD would be biased relative to clinical trials,” she continues. “We designed an experiment that would clearly show and quantify such bias if present.”
cTAP researchers analyzed disease progression in 383 DMD patients randomized to placebo arms in six clinical trials, and compared it to data observed in a real-world setting for 430 patients from five clinical registries in the United States and Europe. Patient outcomes were assessed based on similar inclusion/exclusion criteria and adjusted for known prognostic factors.
“This study found a striking level of consistency in the six-minute walk distance assessment in DMD patients from six clinical trial placebo arms and patients from five different real world and natural history studies,” added Craig McDonald, professor and chair of the Department of Physical Medicine and Rehabilitation at the University of California, Davis, and a co-author of the study.
“This study is essentially a ‘negative control’ for use of RWD as external controls—i.e., an experiment in which seeing no difference supports the reliability of parallel experiments where we do hope to see a difference, in this case future comparisons of novel treatments vs. RWD,” notes Dr. James Signorovitch, managing principal of Analysis Group. “In drug development, negative controls are standard practice for bench science, but are less common in studies of RWD. Our study shows that RWD researchers can learn from bench scientists in this respect.”
“Results from this study provided evidence that progression of DMD observed in RWD/NHD does not differ from that observed among patients randomized to placebo arms in clinical trials. Further, sensitivity analyses did not identify any differences in outcomes by NHD source or geography,” Signorovitch points out.
The potential for bias in clinical trials versus clinical practice has limited the use of natural history and real-world data to supplement or replace placebo controls. Now, this study demonstrates that potential for bias is low, and provides a foundation for drug developers to consider application of NH/RWD in registration trials.
“Drug developers and regulators always need to be cautious when using real-world or natural history controls in clinical trials. The devil is in the details, and the suitability of external controls must be evaluated for each trial on a case-by-case basis. Our study provides a solid evidence base for this evaluation,” states Ward. “We show that external controls can avoid disqualifying levels of bias and demonstrate the types of data sources and statistical methods that are suitable.”
“While the primary goal of the study was to achieve this basic level of understanding, these results provide a foundational assessment of the ability to confirm the compatibility of NHD and placebo control data, potentially opening the door to consideration of a range of modifications in trial protocols that can better leverage NHD,” she adds.
According to Ward, “Drug developers and regulators will only consider the use of NHD and RWD in clinical research if they have the confidence that these data are applied against highly rigorous standards. This study reflects an empirical method of evaluating and confirming the validity of RWD and NHD that could be applied to other disease states in the years ahead.”
“In rare disease research, a significant percentage of clinical trials do not achieve their primary end point or they produce results that are open to interpretation and debate regarding efficacy benefit and as a result may not be approved by regulators. Often, the only way forward is to repeat the study with more patients or better endpoints,” says Dr. Nathalie Goemans of University Hospitals Leuven, Belgium. “A failed trial, which happens often, costs millions of dollars and can expand a development timeline by several years.”
“The use of RWD/NHD in DMD clinical trials is of particular interest because it could make efficient use of existing patient data, facilitate faster trial enrollment, and enable more patients to access active therapies as opposed to placebo. In DMD, recent drug approvals have relied on dystrophin expression as a biomarker while seeking to reach some level of confidence in clinical benefit, but clinical benefits can take longer to develop and can be more difficult to tease out amid the high variability in DMD disease progression,” she notes. “Being able to use larger sample sizes from external controls can significantly improve confidence in clinical effects without the need for more patients on placebos.”
“This study is the largest collaboration between multiple registries and geographies ever conducted in Duchenne, but without such a comprehensive scope, we would not be able to draw such conclusions so firmly,” concludes Goemans. “The results reinforce the importance of data collection, data sharing, consistency of analysis, and collaboration in clinical research that cTAP enables. If drug developers and regulators can achieve confidence in appropriate ways to use RWD and NHD in drug evaluation—based on evidence, such as our study—it will mean that fewer DMD patients need to be on placebo arms, and that trials can be faster to enroll and reach reliable conclusions.”