Abstract: The clinical trial of OBCA (OCALIVA) in the treatment of primary biliary cirrhosis (PBC) shows its efficacy. As it is difficult to conduct sufficient clinical studies in moderate and severe hepatic impairment population, the applicant and the FDA theorem have different opinions based on the same model, such as population PK, exposure-response and physiologically-based PK (PBPK). The applicant considers that the increase in the exposure of drug in liver tissue is limited, and there is no need for dose adjustment, that is, 5 mg, once a day. FDA believes that the influencing factors of the PBPK model have not been fully taken into account and the validation of the PBPK is not robust with a wide variability, and there is also a risk of high blood drug exposure in patients. It is recommended to significantly reduce the dose, that is, 5 mg, once a week, no more than 10 mg, per week at least 3 days interval, and accordingly written into the medication instructions. After approval many patients with hepatic impairment did not take medicine according to the instructions,therefore overdosed, resulting in death. The results fully prove that the original considerations and decisions of FDA have been verified, and the experience and lessons of this example once again suggest that modeling and simulation need bold assumptions and careful verification.

Key words: model-Informed drug development, specific patients, exposure-response relationship, pharmacometrics