Performance of the Population Bioequivalence (PBE) Statistical Test with Impactor Sized Mass Data

Area of Science:

• Pharmaceutical Sciences
• Biostatistics
• Regulatory Science

Background:

• The US Food and Drug Administration (FDA) recommends population bioequivalence (PBE) testing for orally inhaled and nasal drug products.
• Existing PBE methods may have issues with falsely concluding equivalence and low power when between-batch variability is present.

Purpose of the Study:

• To compare the performance of the FDA-recommended PBE approach with modified and alternative bioequivalence tests.
• To investigate methods for improving PBE test reliability, particularly in the presence of between-batch variability.

Main Methods:

• A simulation study was conducted using a metered dose inhaler database.
• Evaluated the standard PBE, an extended PBE modeling within- and between-batch variability, and average bioequivalence (ABE) tests with various log-transformation and batch variability modeling options.

Main Results:

• Separately modeling within- and between-batch variability, along with increased batch sampling, addressed PBE issues of false equivalence and low power.
• Similar modifications were necessary for ABE tests to achieve expected performance.
• These modifications did not resolve asymmetric performance where conclusions depended on the direction of mean differences.

Conclusions:

• Improved bioequivalence testing strategies, particularly for inhaled products, require careful consideration of batch variability.
• Accurate assessment of decision-making error rates is crucial for developing robust regulatory standards for bioequivalence.
• Further research may be needed to address asymmetric performance in bioequivalence testing.