Clinical data consortium seeks to establish best practices for supplemental controls

The Clinical Research Data Sharing Alliance (CRDSA) is leading a demonstration project that seeks to establish best practices around the use of clinical trial and real-world data (RWD) in building supplemental control arms, as well as set standards for data provenance and patient selection methodologies when mixing data sources.
 
“While external control arms have been previously submitted to the [US Food and Drug Administration] in support of regulatory assessment, there still remain a lot of questions surrounding the best methodological approach to identifying and working with ‘fit-for-purpose’ RWD in the regulatory space,” said C.K. Wang, MD, chief medical officer at COTA, a real-world data company focused on oncology. COTA is involved in the demonstration project as a part of the CRDSA’s Innovative Trial Design Working Group.
 
The demonstration project will collect and analyze completed clinical trial data and RWD in non-small cell lung cancer (NSCLC) to create supplemental control arms, building on earlier work done in glioblastoma. The idea is to address gaps in current guidance and develop consensus standards that can inform decisions by FDA and other regulators on the use of patient-level clinical trial data and RWD.
 
“When we have looked at what worked well in helping FDA and other regulators draft and move guidance forward, it is demonstration projects. Show. Don’t tell. That’s exactly why we’re doing this,” said Aaron Mann, CEO of the CRDSA, which includes biopharma companies, nonprofit data sharing platforms, academic institutions, and service and technology partners.
 
Results later in 2023
 
CRDSA formally launched the project in mid-2022 and began bringing together data contributors from both RWD and sponsored trials. The group is currently in the process of drafting data sharing and research agreements. Analysis of the data will begin in mid-2023, with initial findings by the end of the year, Mann said.
 
The first step in the project is to focus on data provenance and selection methodology. CRDSA will work with research partners at academic institutions to examine the best way to select individual patients to go into a supplemental control to ensure that there is not selection bias or cherry picking. “The output from that will probably be a range of options from best to workable,” Mann said. “That’s our starting point.”
 
Once some of the data quality work has been done, the project will examine how the use of clinical trial data and RWD together can create a different clinical picture than any one data source alone. For instance, the RWD may have a temporal component that the trial data does not while the trial data could have endpoints not contained within the RWD.
 
“The industry acknowledges that real-world data and clinical trial data are not the same. They are generated and collected for very different purposes, and each has its own strengths and weaknesses. However, in order to use them together we have to somehow address the gap that exists between the two,” Wang said.