Michael Coladonato and Lucas Tatem, MCRA08.12.22
A prospective, randomized controlled trial (RCT) is the gold standard and the U.S. Food and Drug Administration’s (FDA) preferred study design because it provides the highest level of scientific evidence.
Prospective RCTs generate robust and reliable evidence through bias minimization (randomizing subjects to investigational or control groups), variability control (through carefully constructed eligibility criteria), and stringent data collection requirements (in accordance with a pre-defined protocol). A study’s randomization also ensures that populations are well matched by balancing subject characteristics (both observed and unobserved).1 In premarket studies, the RCT design is used to evaluate new interventions or treatments seeking Premarket Approval (PMA) and/or New Drug Application (NDA) consent from the FDA. PMAs compare the investigational product to an active control cohort, while NDAs traditionally evaluate the investigational product against a placebo. While the RCT study design is considered the benchmark, this approach can be impractical or excessively challenging to conduct due to the lack of viable treatment alternatives or unethical randomization of trial subjects.
When a prospective RCT is not possible or considered overly burdensome to clinicians and/or patients, alternative trial designs can be utilized. One alternative solution involves substituting an active concurrent control with a viable historical control. Using a historical control requires a robust justification and early involvement with the FDA via pre-submissions or Type C meetings. There is regulatory precedence to support applying historical controls to partially or fully replace a concurrent control2 but organizations should be aware of the pros and cons of such a technique before making the switch.
Historical control refers to the practice of using data from past studies and administrative databases to estimate potential response(s) to a placebo or standard care treatment among trial patients. When considering historical evidence in clinical trial design and data analysis planning, many factors must be considered. Ultimately, the historical control data characteristics should closely align with the characteristics of the study’s investigational arm. Historical data may be obtained from various sources as long as the data is relevant and reliable. Data reliability and relevancy are necessary to generate valid scientific evidence for regulatory decisions.
Real-world data (RWD) is one of the primary resources for constructing historical control datasets, and include patient registries, medical charts, and systematic reviews of published clinical data. Regardless of the source, it is imperative to accurately and reliably capture RWD at clinically relevant time intervals in order to meet the FDA’s threshold for RWE (real-world evidence) and serve as a benchmark for providing a reasonable assurance of safety and efficacy compared to the investigational product.
Since RWD can be obtained from different sources with varied structures and quality, many factors must be taken into account when assessing the viability of leveraging data as a historical control. Several important considerations include propensity score matching, performance outcomes, available sample size, and available follow-up time intervals. Propensity score matching attempts to reduce treatment assignment bias and mimic randomization by creating an investigational treatment subject sample and comparing it (with all observed and unobserved covariates) to patients receiving a placebo or standard treatment. Performance outcomes for the historical control should align with the proposed investigational arm and should provide data on both safety and effectiveness of the target population. The historical control’s sample size must be sufficient to power the statistical analysis and should be representative of the target population. Lastly, follow-up time and visit intervals should be relevant and aligned with the investigational arm. When historical information is leveraged appropriately, the evidence generated can be used to support regulatory decision-making and provide a least burdensome regulatory path to market.
It is very important to develop a thorough analytical plan before accessing, retrieving, and analyzing RWD for a historical control so that source and selection bias—as well as post-hoc analyses—are minimized. Further, the study protocol and statistical analysis plan (SAP) should address the same elements a traditional clinical trial protocol and SAP would cover.2 Additionally, sponsors should consider the availability of patient-level data, including source statistics, when selecting RWD; information ideally should be provided in the required standardized file formats and data structures in accordance with statutory standardized variables and/or definitions, when applicable. The FDA recently published several relevant resources specifically related to the use of RWD in investigational new drug applications, including the draft guidance documents “Considerations for the Use of Real-World Data and Real-World Evidence To Support Regulatory Decision-Making for Drug and Biological Products: Draft Guidance for Industry,” and “Data Standards for Drug and Biological Product Submissions Containing Real-World Data: Draft Guidance for Industry.”
Using a historical control arm instead of enrolling an active comparator can help reduce or eliminate the need for control patient enrollment and shorten clinical study periods, which ultimately can help cut the costs of generating the required evidence to support product approvals. Additional time and cost-reducing factors include less overhead for investigational sites and easier enrollment of the single-arm investigational study due to clear treatment assignment.
Historical control data may be leveraged as part of various regulatory activities to support the benefit-risk profile of devices at various points in their life cycle. This includes but is not limited to expanded indications for use (e.g., through powered subgroup analyses of registry data), post-market surveillance studies, post-approval device surveillance, and supplemental data. With each use of historical data, there are nuanced factors that must be considered to ensure valid scientific evidence is generated from RWD and the data substantiates the justification of this alternate path.
The use of historical data, including RWD to generate RWE in support of a regulatory decision, can be a daunting task. But expertise from organizations like MCRA can help ease the difficulties associated with such an approach to clinical trials. Leveraging RWD, applying appropriate risk management strategies, and mitigating sources of bias effectively to generate valid scientific evidence can provide the least burdensome path to market. And that’s certainly a path worth taking.
References
Michael Coladonato is a senior associate, regulatory affairs, at MCRA LLC with four years of experience in medical device regulation. At MCRA, Michael provides guidance on regulatory strategies and submissions across various therapies including spine, orthopedics, wound care, and cardiovascular, as well as drugs, and biologics. Michael has a wide range of regulatory submission experience including but not limited to IDEs, INDs, PMAs, 510(k)s, HUDs, De Novos, Breakthrough Device Designations, and Clinical Evaluation Reports. Michael leads and supports numerous projects from regulatory strategy assessments through full submission writing and communications with FDA. Michael earned his bachelor’s degree in biomedical engineering from The Catholic University of America.
Lucas Tatem, M.Eng., is a regulatory affairs associate at MCRA, where he is primarily focused on supporting and advising clients in the completion of various regulatory submissions specifically for spine and orthopedic implants, as well as wound care products. Lucas has supported projects including 510(k)s, pre-submissions, Premarket Approval (PMAs), Investigational Device Exemptions, Investigational New Drug applications and European Clinical Evaluation Report submissions. He attended Clemson University and earned a master’s degree in biomedical engineering with a concentration in biomaterials.
Prospective RCTs generate robust and reliable evidence through bias minimization (randomizing subjects to investigational or control groups), variability control (through carefully constructed eligibility criteria), and stringent data collection requirements (in accordance with a pre-defined protocol). A study’s randomization also ensures that populations are well matched by balancing subject characteristics (both observed and unobserved).1 In premarket studies, the RCT design is used to evaluate new interventions or treatments seeking Premarket Approval (PMA) and/or New Drug Application (NDA) consent from the FDA. PMAs compare the investigational product to an active control cohort, while NDAs traditionally evaluate the investigational product against a placebo. While the RCT study design is considered the benchmark, this approach can be impractical or excessively challenging to conduct due to the lack of viable treatment alternatives or unethical randomization of trial subjects.
When a prospective RCT is not possible or considered overly burdensome to clinicians and/or patients, alternative trial designs can be utilized. One alternative solution involves substituting an active concurrent control with a viable historical control. Using a historical control requires a robust justification and early involvement with the FDA via pre-submissions or Type C meetings. There is regulatory precedence to support applying historical controls to partially or fully replace a concurrent control2 but organizations should be aware of the pros and cons of such a technique before making the switch.
Historical control refers to the practice of using data from past studies and administrative databases to estimate potential response(s) to a placebo or standard care treatment among trial patients. When considering historical evidence in clinical trial design and data analysis planning, many factors must be considered. Ultimately, the historical control data characteristics should closely align with the characteristics of the study’s investigational arm. Historical data may be obtained from various sources as long as the data is relevant and reliable. Data reliability and relevancy are necessary to generate valid scientific evidence for regulatory decisions.
Real-world data (RWD) is one of the primary resources for constructing historical control datasets, and include patient registries, medical charts, and systematic reviews of published clinical data. Regardless of the source, it is imperative to accurately and reliably capture RWD at clinically relevant time intervals in order to meet the FDA’s threshold for RWE (real-world evidence) and serve as a benchmark for providing a reasonable assurance of safety and efficacy compared to the investigational product.
Since RWD can be obtained from different sources with varied structures and quality, many factors must be taken into account when assessing the viability of leveraging data as a historical control. Several important considerations include propensity score matching, performance outcomes, available sample size, and available follow-up time intervals. Propensity score matching attempts to reduce treatment assignment bias and mimic randomization by creating an investigational treatment subject sample and comparing it (with all observed and unobserved covariates) to patients receiving a placebo or standard treatment. Performance outcomes for the historical control should align with the proposed investigational arm and should provide data on both safety and effectiveness of the target population. The historical control’s sample size must be sufficient to power the statistical analysis and should be representative of the target population. Lastly, follow-up time and visit intervals should be relevant and aligned with the investigational arm. When historical information is leveraged appropriately, the evidence generated can be used to support regulatory decision-making and provide a least burdensome regulatory path to market.
It is very important to develop a thorough analytical plan before accessing, retrieving, and analyzing RWD for a historical control so that source and selection bias—as well as post-hoc analyses—are minimized. Further, the study protocol and statistical analysis plan (SAP) should address the same elements a traditional clinical trial protocol and SAP would cover.2 Additionally, sponsors should consider the availability of patient-level data, including source statistics, when selecting RWD; information ideally should be provided in the required standardized file formats and data structures in accordance with statutory standardized variables and/or definitions, when applicable. The FDA recently published several relevant resources specifically related to the use of RWD in investigational new drug applications, including the draft guidance documents “Considerations for the Use of Real-World Data and Real-World Evidence To Support Regulatory Decision-Making for Drug and Biological Products: Draft Guidance for Industry,” and “Data Standards for Drug and Biological Product Submissions Containing Real-World Data: Draft Guidance for Industry.”
Using a historical control arm instead of enrolling an active comparator can help reduce or eliminate the need for control patient enrollment and shorten clinical study periods, which ultimately can help cut the costs of generating the required evidence to support product approvals. Additional time and cost-reducing factors include less overhead for investigational sites and easier enrollment of the single-arm investigational study due to clear treatment assignment.
Historical control data may be leveraged as part of various regulatory activities to support the benefit-risk profile of devices at various points in their life cycle. This includes but is not limited to expanded indications for use (e.g., through powered subgroup analyses of registry data), post-market surveillance studies, post-approval device surveillance, and supplemental data. With each use of historical data, there are nuanced factors that must be considered to ensure valid scientific evidence is generated from RWD and the data substantiates the justification of this alternate path.
The use of historical data, including RWD to generate RWE in support of a regulatory decision, can be a daunting task. But expertise from organizations like MCRA can help ease the difficulties associated with such an approach to clinical trials. Leveraging RWD, applying appropriate risk management strategies, and mitigating sources of bias effectively to generate valid scientific evidence can provide the least burdensome path to market. And that’s certainly a path worth taking.
References
- Hariton, E., & Locascio, J. J. (2018). Randomised controlled trials - the gold standard for effectiveness research: Study design: randomised controlled trials. BJOG : an international journal of obstetrics and gynaecology, 125(13), 1716. https://doi.org/10.1111/1471-0528.15199
- Ghadessi, M., Tang, R., Zhou, J., Liu, R., Wang, C., Toyoizumi, K., ... & Beckman, R. A. (2020). A roadmap to using historical controls in clinical trials–by Drug Information Association Adaptive Design Scientific Working Group (DIA-ADSWG). Orphanet Journal of Rare Diseases, 15(1), 1-19.
Michael Coladonato is a senior associate, regulatory affairs, at MCRA LLC with four years of experience in medical device regulation. At MCRA, Michael provides guidance on regulatory strategies and submissions across various therapies including spine, orthopedics, wound care, and cardiovascular, as well as drugs, and biologics. Michael has a wide range of regulatory submission experience including but not limited to IDEs, INDs, PMAs, 510(k)s, HUDs, De Novos, Breakthrough Device Designations, and Clinical Evaluation Reports. Michael leads and supports numerous projects from regulatory strategy assessments through full submission writing and communications with FDA. Michael earned his bachelor’s degree in biomedical engineering from The Catholic University of America.
Lucas Tatem, M.Eng., is a regulatory affairs associate at MCRA, where he is primarily focused on supporting and advising clients in the completion of various regulatory submissions specifically for spine and orthopedic implants, as well as wound care products. Lucas has supported projects including 510(k)s, pre-submissions, Premarket Approval (PMAs), Investigational Device Exemptions, Investigational New Drug applications and European Clinical Evaluation Report submissions. He attended Clemson University and earned a master’s degree in biomedical engineering with a concentration in biomaterials.