EU regulators offer new guidance on complex clinical trials
Regulatory News | 03 June 2022 |
European regulators published a question and answer document that they hope will give more clarity on how complex clinical trials (CCT) should be conducted. The document aims to answer some basic questions, such as what to consider in the planning and conduct of CCTs, how to justify Bayesian approaches to regulators, and how to use biomarkers.
On 2 June, the European Commission (EC), European Medicines Agency (EMA) and the Heads of Medicines Agencies (HMA) published a joint Q&A intended to provide guidance on CCTs. They note that the structure of the results from a CCT in a marketing authorization (MAA) may differ significantly from its submission(s) for clinical trial authorization (CTA). For example, the regulators write, “A MAA submission that is based on an active and a control arm which had been submitted as different CTAs yet were part of a shared scientific framework,” noting that a CCT may refer to a trial within a single EU clinical trial number or to multiple related trials with different EUCT numbers under the same master protocol.
To address how CCTs may operate differently than standard clinical trials, the regulators ask seven broad questions, such as what needs to be considered in the planning and conduct of CCTs, what to consider in the study design, and what to consider in terms of the safety, rights and well-being of study participants.
A key portion of the document delves into the use of Bayesian methodologies. They are often used in situations when there is insufficient uniform data to draw a conclusion but using statistical tools can help overcome those limitations.
“In trials exploring pharmacokinetics and pharmacodynamics, Bayesian approaches have often been used to analyze data and to generate results, including empirical Bayesian approaches where prior distributions are estimated from the newly collected data, such as for population pharmacokinetic analyses,” the regulators noted. “Bayesian approaches have also been used in the context of dose ranging, dose escalation and selection of a dose for further trials. Results from Bayesian work have subsequently been assessed in regulatory submissions and enabled inference on pharmacokinetic and pharmacodynamic parameters as well as dose selection.”
While Bayesian methodologies offer a lot of potential, the regulators note that there are few examples where the tools have been successfully used for a submission. They also note that the methodologies have evolved significantly in recent years and sponsors should talk to regulators early on about how to best use them.
The EU regulators go on to note that if a sponsor decides to use Bayesian methodologies, it’s important that they provide certain documentation, such as the reasons for using the Bayesian methods, and provide relevant details of the Bayesian approach and data on prior predictive simulations.
Another key area the regulators touch on is the use of biomarkers in CCTs.
Generally, the document states that sponsors need to provide sufficient information in their trial documentation to allow them to assess the appropriateness of the biomarker assays being used. They also note key details about the in vitro diagnostic (IVD) being used to assess the biomarkers need to be provided and the parameters should be included in the Instructions for Use (IFU) for CE marked devices.
“In addition, where multiple biomarkers are used in complex clinical trials, the consequences of their interplay on the selected patient population(s) needs to be elaborated in the protocol,” the regulators added. “Depending on the trial design, biomarker assays and related analyses (as function of their purpose and associated claims) may be described either in the common (master protocol) part or in the sub-protocols, per IMP(s) or target patient populations.”
The document also addresses questions regarding what to do in CCTs with interventions that target multiple biomarkers, and how to present rationales and designs for when planning performance studies. It also goes into what to do when the trials include the use of combination diagnostics (CDx). In certain situations, however, the sponsor may need to look at their local regulations to conduct such clinical trials.
“Sponsors of clinical trials investigating medicinal products and, in parallel, IVDs are encouraged to consult national guidance documents and contact the respective National Competent Authorities for performance studies prior to clinical trial submission,” the regulators said. “A timely consultation is also encouraged when a particular biomarker assay is used for different purposes/intended uses and/or has different regulatory status in sub-protocols (e.g. CE marked when used for one intended purpose in one sub-protocol, while not being CE-marked for intended purpose in another sub-protocol).”
“In clinical trials, where the assay is not a device for self-testing or a device for near-patient testing, the applicant is recommended to perform the assay (e.g. immunohistochemistry assay) at a central laboratory to generate consistent and robust data,” they added. “The laboratory should be adequately accredited and the IVDs used in EU laboratories have to comply with the IVDR.”
On 2 June, the European Commission (EC), European Medicines Agency (EMA) and the Heads of Medicines Agencies (HMA) published a joint Q&A intended to provide guidance on CCTs. They note that the structure of the results from a CCT in a marketing authorization (MAA) may differ significantly from its submission(s) for clinical trial authorization (CTA). For example, the regulators write, “A MAA submission that is based on an active and a control arm which had been submitted as different CTAs yet were part of a shared scientific framework,” noting that a CCT may refer to a trial within a single EU clinical trial number or to multiple related trials with different EUCT numbers under the same master protocol.
To address how CCTs may operate differently than standard clinical trials, the regulators ask seven broad questions, such as what needs to be considered in the planning and conduct of CCTs, what to consider in the study design, and what to consider in terms of the safety, rights and well-being of study participants.
A key portion of the document delves into the use of Bayesian methodologies. They are often used in situations when there is insufficient uniform data to draw a conclusion but using statistical tools can help overcome those limitations.
“In trials exploring pharmacokinetics and pharmacodynamics, Bayesian approaches have often been used to analyze data and to generate results, including empirical Bayesian approaches where prior distributions are estimated from the newly collected data, such as for population pharmacokinetic analyses,” the regulators noted. “Bayesian approaches have also been used in the context of dose ranging, dose escalation and selection of a dose for further trials. Results from Bayesian work have subsequently been assessed in regulatory submissions and enabled inference on pharmacokinetic and pharmacodynamic parameters as well as dose selection.”
While Bayesian methodologies offer a lot of potential, the regulators note that there are few examples where the tools have been successfully used for a submission. They also note that the methodologies have evolved significantly in recent years and sponsors should talk to regulators early on about how to best use them.
The EU regulators go on to note that if a sponsor decides to use Bayesian methodologies, it’s important that they provide certain documentation, such as the reasons for using the Bayesian methods, and provide relevant details of the Bayesian approach and data on prior predictive simulations.
Another key area the regulators touch on is the use of biomarkers in CCTs.
Generally, the document states that sponsors need to provide sufficient information in their trial documentation to allow them to assess the appropriateness of the biomarker assays being used. They also note key details about the in vitro diagnostic (IVD) being used to assess the biomarkers need to be provided and the parameters should be included in the Instructions for Use (IFU) for CE marked devices.
“In addition, where multiple biomarkers are used in complex clinical trials, the consequences of their interplay on the selected patient population(s) needs to be elaborated in the protocol,” the regulators added. “Depending on the trial design, biomarker assays and related analyses (as function of their purpose and associated claims) may be described either in the common (master protocol) part or in the sub-protocols, per IMP(s) or target patient populations.”
The document also addresses questions regarding what to do in CCTs with interventions that target multiple biomarkers, and how to present rationales and designs for when planning performance studies. It also goes into what to do when the trials include the use of combination diagnostics (CDx). In certain situations, however, the sponsor may need to look at their local regulations to conduct such clinical trials.
“Sponsors of clinical trials investigating medicinal products and, in parallel, IVDs are encouraged to consult national guidance documents and contact the respective National Competent Authorities for performance studies prior to clinical trial submission,” the regulators said. “A timely consultation is also encouraged when a particular biomarker assay is used for different purposes/intended uses and/or has different regulatory status in sub-protocols (e.g. CE marked when used for one intended purpose in one sub-protocol, while not being CE-marked for intended purpose in another sub-protocol).”
“In clinical trials, where the assay is not a device for self-testing or a device for near-patient testing, the applicant is recommended to perform the assay (e.g. immunohistochemistry assay) at a central laboratory to generate consistent and robust data,” they added. “The laboratory should be adequately accredited and the IVDs used in EU laboratories have to comply with the IVDR.”
© 2025 Regulatory Affairs Professionals Society.
