目的 采用星点设计-效应面法（CCD-RSM），以平均粒径、粒径分布（多分散指数，PDI）及Zeta电位为考察指标，确定欧前胡素脂质微球（ILM）的最佳处方，并进行质量评价。方法 以平均粒径、PDI、Zeta电位为评价指标，考察蛋黄卵磷脂用量、泊洛沙姆用量、大豆油占油相总量的百分比对ILM制剂性质的影响。用二项式拟合建立指标与因素之间的数学关系，根据评价指标的最佳数学模型描绘效应面，经效应面法预测最佳处方，并进行预测分析。按照优化出来的处方制备ILM，对其常规的质量评价进行研究。结果 优化后的最佳处方为蛋黄卵磷脂1.5 g，泊洛沙姆188 0.35 g，注射用中链脂肪酸甘油脂（MCT）10 g，优化处方各指标和目标值接近。按照优化的处方制备出来的ILM的平均粒径是（165.00±0.22）nm，PDI为0.058±0.070，Zeta电位为（-30.10±0.13）mV，包封率约为90.09%，载药量为1.0 mg/mL。结论 星点设计-效应面法适用于脂质微球的处方优化，所建立的数学模型预测性良好；优化后的ILM符合药剂学要求。

Objective To optimize prescription and evaluate the quality of imperatorin lipid microspheres (ILM) by central composite design-response surface methodology (CCD-RSM), therefore to use the mean particle size, particle size distribution (polydispersity index, PDI) and Zeta potential as the primary indicator. Methods Mean particle size, particle size distribution and Zeta potential of lipid microspheres were measured to investigate the effect of the concentration of egg yolk lecithin, poloxamer and percentage of soybean oil in oil phase on the properties of ILM preparation. Mathematic relation between indicator and factor was constructed by binominal fitting. Then the response surface method according to the best mathematical model of evaluation indicators was used to make predictive analysis of the best prescription. Optimized prescription was used to prepare ILM and evaluate its quality. Results Optimized prescription was egg yolk lecithin 1.50 g, F68 0.35 g, and MCT 10 g. All items of optimized prescription were similar to target values. According to the optimized prescription, mean particle size of ILM was (165.00 ±0.22) nm, particle size distribution was 0.046 ±0.070, Zeta potential was (-30.30 ±0.13) mV, encapsulation efficiency was about 90.09%, drug-loading rate was 1.0 mg/mL. Conclusion Due to the better predictability of constructed mathematical model, CCD-RSM can be applied to optimize prescription of lipid microspheres and the optimized ILM meet pharmacy requirements.

江西省青年科学家培养对象（20153BCB23019）；"赣鄱英才555"工程领导人才培养计划（赣财教指[2013]296号）；青年科学基金资助（20161ACB21020）