Extending analysis methods for population pharmacokinetic to account for individual differences and study design

2022 Fiscal Year Final Research Report

• Project/Area Number: 17K00057
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Statistical science
• Research Institution: Teikyo University
• Principal Investigator: Suzuki Asuka (根本明日香) 帝京大学, 公私立大学の部局等, 講師 (20722482)
• Project Period (FY): 2017-04-01 – 2023-03-31
• Keywords: 非線形混合効果モデル / 母集団薬物動態解析 / 階層モデル母数のベイズ推定 / モデリングでの共変量組み入れ基準 / 尤度比検定 / デビアンス統計量 / 数値実験

Outline of Final Research Achievements

The objective of population pharmacokinetic analysis is a covariate selection problem to find an association between inter-individual differences and patient attributes. Although it is desirable to obtain data over the entire range of the time-concentration curve, there are practical limitations on sampling design to obtain blood drug concentration data, which has been a challenge in practices of clinical pharmacology. As the research outcomes, as an extension of the objective function for estimation by the maximum likelihood method, we proposed a further approximation, pseudo-OFV, based on a linear approximation of the objective function. This method is robust in the vulnerability where a detection power varies depending on the sampling design. In addition, we showed that sample size calculation assuming Bayesian modeling with prior information is useful for limited sampling problems because it has the advantage of reducing the burden on subjects.

Free Research Field

臨床薬理学、生物統計学

Academic Significance and Societal Importance of the Research Achievements

母集団薬物動態解析のために、適切な計画に基づく臨床薬理試験の実施が求められているが、多数の時点で繰り返し採血を行い、また、生活上不都合のある時間帯に採血を行う研究計画は被験者に過大な負担をかけるものであり、実際には実施が難しいために本来の研究の目的を達成することができないという問題があった。限定されたサンプリングにもとづくデータの情報不足への対処法として、疑似OFVを目的関数とする新しい方法を考案し、また、ベイズ流モデルパラメータ推定を行う状況での研究計画法として症例数計算の事例を示し、問題の解決に向けて一定の進歩を達成した。