The pilot BE study was conducted to evaluate the pharmacokinetic performance of the test formulation relative to the RLD Based on the results of the pilot, the intra-subject CV for the pharmacokinetic parameter Cmax was observed to be approximately 38%, while for AUC, it was approximately 28%. These values indicate that the formulation, particularly in terms of peak concentration, exhibits high intra-subject variability, which classifies it as a Highly Variable Drug (HVD) under current regulatory guidelines (i.e., CV ≥ 30%).
This degree of variability significantly impacts the design and success probability of a pivotal bioequivalence study. In a conventional 2-treatment, 2-period, 2-sequence crossover study, high variability inflates the width of the confidence intervals around the point estimates, making it difficult to demonstrate bioequivalence within the standard 80.00–125.00% acceptance range. For a drug with 38% variability in Cmax, a 2-way design would require more than 120 subjects to achieve adequate statistical power (≥80%), which is neither cost-effective nor operationally efficient. Furthermore, the risk of study failure remains high even with large sample sizes due to the fixed BE limits.
To overcome this challenge, a replicate design study is recommended. Specifically, a partial or full replicate crossover design (e.g., 3-period, 3-sequence or 4-period, 2-sequence) will be employed. This design allows for within-subject variability of the reference product to be estimated directly, enabling the use of Reference-Scaled Average Bioequivalence (RSABE) methodology. Under the RSABE approach, the conventional 80–125% limits can be widened up to a regulatory cap (e.g., 75.00–133.00%) depending on the variability of the reference product.
Considering the point estimates from the pilot study—95% for Cmax and 98% for AUC—along with acceptable Tmax profiles and similar PK shapes, the test formulation is pharmacokinetically comparable to the reference. Thus, with the application of RSABE and a replicate design, it is expected that the pivotal study can demonstrate bioequivalence with high confidence.
Preliminary sample size estimations based on simulations using the observed CV suggest that a 3-period replicate study would require approximately 60 subjects to achieve ≥90% power, while a 4-period design may provide even greater precision with approximately 48–52 subjects. The final choice of replicate design will consider logistical feasibility, regulatory acceptance, and analytical throughput.
In summary, the pilot data strongly support the adoption of a replicate design BE study, coupled with RSABE analysis, to account for the inherent variability of the reference product. This approach ensures scientific rigor, regulatory compliance, and cost-effective execution, while minimizing the risk of pivotal study failure due to high intra-subject variability.
Resource Person: Moinuddin Syed. Ph.D, PMP®
