A novel power prior approach for borrowing historical control data in clinical trials

A dynamic power prior approach to non-inferiority trials for normal means

Non-inferiority trials compare new experimental therapies to standard ones (active control). In these experiments, historical information on the control treatment is often available. This makes Bayesian methodology appealing since it allows a natural way to exploit information from past studies. In the present paper, we suggest the use of previous data for constructing the prior distribution of the control effect parameter. Specifically, we consider a dynamic power prior that possibly allows to discount the level of borrowing in the presence of heterogeneity between past and current control data. The discount parameter of the prior is based on the Hellinger distance between the posterior distributions of the control parameter based, respectively, on historical and current data. We develop the methodology for comparing normal means and we handle the unknown variance assumption using MCMC. We also provide a simulation study to analyze the proposed test in terms of frequentist size and power, as it is usually requested by regulatory agencies. Finally, we investigate comparisons with some existing methods and we illustrate an application to a real case study.

Enhancing generalizability and efficiency in clinical trials through dynamic information borrowing for both experimental and control arms: A simulation study

AIM

Utilizing external information in clinical trials enhances validity by including a wider population and expedites the implementation of adaptive designs, ultimately improving research efficiency. However, current research focused on scenarios in which only the control group benefited from the utilization of external information, while trials involving external information in both experimental and control arms were more complex and might pose challenges when applied in real-world settings.

METHODS

To address these concerns, our study pioneered the application of test-then-pool, normalized power prior, calibrated power prior, and elastic prior to a two-arm information borrowing framework and systematically compared their operating characteristics through a series of simulation studies under most and least desirable scenarios.

RESULTS

In the most desirable scenarios of information borrowing, all methods managed to control the mean of type I error rates within 5%, among which the normalized power prior, calibrated power prior and elastic prior approaches increased the mean of power from 85.94% to 95%. In the least desirable scenarios, the mean type I error rates for normalized power prior, calibrated power prior and elastic prior approaches exceeded 20%, while the mean power decreased to around 80%.

CONCLUSIONS

Our findings reveal that the normalized power prior, calibrated power prior and elastic prior approaches are suitable for situations with minimal heterogeneity between historical and current data, whereas the test-then-pool approach emerges as a more prudent choice when facing substantial discrepancies between historical and current information for trials consider information borrow in both arms.