论文部分内容阅读
目的应用基于模型化的荟萃分析的方法,建立紫杉醇种属间相关性的定量评价方法,为新药研发的种属间外推与剂量确定提供参考。方法以紫杉醇药动学为关键词,检索Pub-Med、中国知网(CNKI)、万方等数据库建库相关文献。按照纳入排除标准对检索文献进行筛选,并按种属进行分类,摘录每篇文献的血药浓度数据,应用非线性混合效应模型法(NONMEM)分别对人、大鼠、小鼠进行建模。采用正态化预测分布误差(normalized prediction distribution errors,NPDE)法对建立的模型进行验证,并依据相关生长规律法对药物种属间相关系数进行计算。结果通过非线性混合效应模型法法模型化,人、大鼠、小鼠的药动学行为均符合二室模型,与文献检索结果一致。正态化预测分布误差对最终模型结果进行可视化检验,模型结果准确可靠。依据相关生长规律法(allometric scaling)对3个种属的清除率CL和总表观分布容积Vtotal的相关系数进行计算,结果分别为r2=0.997 4和r2=0.937 2,种属间相关系数的线性关系良好。结论以紫杉醇为例,成功地建立了基于模型的Meta分析方法,能够定量的评价和预测种属间相关性。
OBJECTIVE: To establish a quantitative evaluation method for the correlation between taxoids in taxa based on the model-based meta-analysis and to provide reference for the extra-genomic extrapolation and dose determination of new drug research and development. Methods Taking paclitaxel pharmacokinetics as the key word, we searched PubMed, CNKI and Wanfang databases. According to inclusion criteria, the search documents were screened and classified according to their genus. The blood concentration data of each reference were extracted. Nonlinear mixed-effects model (NONMEM) was used to model human, rat and mouse respectively. The established model was validated by normalized prediction distribution errors (NPDE), and the correlation coefficient of drug species was calculated according to the relevant growth law. The results were modeled by the non-linear mixed-effects model method. The pharmacokinetic behaviors of human, rat and mouse were in accordance with the two-compartment model and were consistent with the results of literature search. The error of normalized prediction distribution can verify the result of the final model visually, and the result of the model is accurate and reliable. Correlation coefficients of clearance CL and total apparent volume Vtotal of three species were calculated by allometric scaling, and the results were r2 = 0.997 4 and r2 = 0.937 2 respectively. The correlation coefficient Linear relationship is good. Conclusion Taking paclitaxel as an example, we successfully established a model-based meta-analysis method that can quantitatively evaluate and predict inter-species correlation.