目的 构建一种适用于中药复方透皮促渗剂筛选的评价体系，并将其用于二仙汤复合促渗剂的优选。方法 以薄荷脑、氮酮、冰片为促渗剂，设计L₉(3³)正交试验。测定10个指标成分（黄柏碱、芒果苷、阿魏酸、小檗碱、仙茅苷、朝藿定A1、朝藿定A、朝藿定B、朝藿定C、淫羊藿苷）的48 h单位面积累积渗透量（Q₄₈）、稳态渗透速率（steady-state flux，J_ss）与增渗比，利用主成分分析（principal component analysis，PCA）对多维渗透参数进行降维处理，同时引入指纹图谱相似度（S）作为质量约束条件。将上述主成分得分与指纹图谱相似度通过满意度函数进行加权整合，计算综合满意度（D），并以D作为响应值进行极差和方差分析，筛选最佳促渗剂配方。结果 筛选得到最优复合促渗剂配方为2%薄荷脑、1%氮酮。验证实验表明，该组成在具备卓越促渗效果的同时，指纹图谱相似度S保持在0.9以上，兼顾“高效促渗”和“质量保真”。结论 构建的“PCA降维-指纹图谱约束-满意度函数整合”系统，能够有效解决多指标优化中信息冗余与决策矛盾的问题，为中药复方复合促渗剂筛选提供可借鉴的研究范式。

Objective This study aimed to establish an evaluation system for screening transdermal penetration enhancers suitable for traditional Chinese medicine (TCM) compound formulations and to apply it to the optimization of a composite penetration enhancer for Erxian Decoction (EXD, 二仙汤). Methods Menthol, azone, and borneol were selected as penetration enhancers, and an L₉(3³) orthogonal design was employed. The 48 h cumulative permeation per unit area (Q₄₈), steady-state flux (J_ss), and enhancement ratio (ER) of ten marker components (phellodendrine, mangiferin, ferulic acid, berberine, curculigine, epimedin A1, epimedin A, epimedin B, epimedin C, icariin) were measured. Principal component analysis (PCA) was performed separately for Q₄₈ and J_ss datasets to reduce dimensionality, while fingerprint similarity (S) was incorporated as a quality constraint. The PCA scores and S were weighted and integrated via the satisfaction function to calculate the comprehensive satisfaction (D). With D as the response value, range analysis and analysis of variance were performed to screen the optimal formulation of penetration enhancers. Results The optimal compound penetration enhancer formulation was identified as 2% menthol combined with 1% azone. Validation studies showed that this formulation exhibited excellent penetration-enhancing effect, while the fingerprint similarity S remained above 0.9, achieving both “high-efficiency penetration enhancement” and “quality fidelity”. Conclusion The constructed “PCA dimension reduction-fingerprint constraint-satisfaction function integration” system can effectively solve the problems of information redundancy and decision contradiction in multi-index optimization, providing a referable research paradigm for screening composite penetration enhancers of TCM compound prescriptions.

R283.6

国家自然科学基金资助项目(8206140176)