How to Report and Interpret Bioequivalence Data in Solid Organ Transplant Recipients
In a proper bioequivalence study, the 90% confidence intervals (CI) of the generic:original ratio for area under the curve and Cmax has to be within specific acceptance boundaries (for tacrolimus: area under the curve, 0.9-1.11; Cmax, 0.8-1.25) for bioequivalence to be established.2 Why do the authors in the present underpowered parallel study statistically “manipulate” the data to present 90% CI and refer to the bioequivalence acceptance limits instead of comparing the pharmacokinetic variables using standard statistical methods for these kinds of analyses? What does “statistically similar” mean in this context? We worry that this statement may mislead the reader and create an impression that the authors have performed a proper bioequivalence study. They even state, “Nevertheless, appropriate statistical testing, as applied here, is essential to evaluate the pharmacokinetic similarity of different tacrolimus products.” It is quite clear for any reader with knowledge regarding bioequivalence that this statement is speculative, and we wonder if the authors really mean they have complied with this.
It would also be of interest to get an explanation of the clearance values (CL/F) presented in Table 3. We are not quite sure how to interpret the unit of [mg/h] and why this parameter has been dose-normalized.
We also strongly regret to be misleadingly cited. With regards to our previous publication in Transplantation3 they state; “Consistent with our findings, an appropriately-designed bioequivalence study of the tacrolimus generic formulation Tacni, undertaken in elderly de novo transplant patients, also showed statistically similar tacrolimus pharmacokinetics to the originator drug, but the upper values for 90% CIs in that trial again exceeded Food and Drug Administration and EMA criteria.” Our study showed that Tacni was not bioequivalent with Prograf in renal transplant recipients older than 65 years. The 90% CI of the Cmax ratio in our study ranged from 1.35 to 1.65. To clarify, it was not only the “upper value” that exceeded the bioequivalence acceptance criteria, the entire interval was outside the limits. If this is to be “statistically similar,” we are more than a bit worried about the authors' interpretation of their own data. The transplant society strongly needs more studies and interpretations of current generic immunosuppressive medications.