早产儿药物不良反应率高，CYP1A2酶活性变化规律尚未知，通过传统药动学研究方法获得安全用药资料伦理和操作上不可行。国际上正推行使用生理药动学（PBPK）模型预测早产儿用药剂量和器官暴露，但基础数据多来自动物、体外或基于年龄估算，用于准确建模的早产儿重要器官血流、体积及酶活性数据奇缺，模型容易产生偏倚。本研究拟在前期工作基础上，利用临床进行咖啡因（CYP1A2体内探针底物）治疗的早产儿剩余血样，通过咖啡因和其代谢物浓度，利用非线性混合效应模型法阐明0-3月龄（校正胎龄40周）CYP1A2酶活性变化规律以及基因多态性对酶活性的影响，并应用无创的二维多普勒和三维超声技术明确0-3月龄早产儿肝、肾、心、脑血流速率和体积变化规律，然后将获得的数据替代PBPK软件中稀缺或基于年龄估算的数据，考察CYP1A2底物PBPK建模的准确性。本研究对于提高早产儿PBPK临床给药方案设计和风险评估具有重要意义。

While the adverse drug reactions are commonly seen in premature infants, the developmental changes of one important drug metabolizing enzyme, cytochrome P450 1A2 (CYP1A2), in premature infants still remains unclear. Furthermore, data collected from premature infants by using conventional pharmacokinetic methods cannot be ethically and practically applied in this population. Physiologically based pharmacokinetics (PBPK) modeling is a mathematical modeling technique which can be used for predicting the drug dosage and organs exposure in premature infants. However, this model is usually built based on the studies from the animal and in vitro models, or the extrapolated data gathered from adults or older children. The developing activity of drug metabolizing enzyme, blood flow rate and volume of the vital organs in premature infants are actually very important data for PBPK simulation. The current problem is that these data are extremely rare. For these reasons, the result from PBPK simulation is not so reliable. In this study, spare blood sample from premature infants who were treated with caffeine, the only CYP1A2 probe substrate in human, will be collected and the concentration of caffeine and its metabolites paraxanthine (1,7-dimethylxanthine) in the blood will be measured with ultra-performance liquid chromatography massspectrometry (UPLC-MS) method. Then the data will be analyzed by using non-linear mixed-effects model (NONMEM) to illuminate the rules of developmental changes of CYP1A2 enzyme activity in premature infants from 0 to 3 months (adjusted gestational age is about 40 weeks) . The study will also assess the influence of gene polymorphisms on enzyme activity. Furthermore, the developmental changes of blood flow and volume in main organs including lung, kidney, heart, and brain will be detected by using a noninvasive Doppler imaging and three - dimensional ultrasound technology . These data, collected from the premature infants, will finally replace the estimate data or data still lacked in PBPK. After that, the accuracy of PBPK simulation will be evaluated through importing the concentration data of CYP1A2 substrates. The current study will increase the accuracy of PBPK simulation for preterm infants in clinical trial and will enable more accurate dose selection.