目的 基于中药复方化学成分群整体指纹信息，考察桃红四物汤（TSD）汤剂制备过程中化学成分的变化，并分析其影响质量的波动成分和主要贡献成分，为其质量评控提供参考。方法 按照经典名方物质基准制备工艺，制备TSD一次煎煮（A组）、二次煎煮（B组）、冷冻干燥（C组）3个制备过程样品，建立18批药材3组样品的HPLC指纹图谱，采用中药指纹图谱相似度评价系统进行数据处理，SPSS 20.0进行系统聚类分析（HCA）和主成分分析（PCA），SIMCA 14.1软件进行偏最小二乘法-判别分析（PLS-DA），分析3组样品中化学指纹信息的动态变化。结果 选取17个色谱峰作为指纹图谱共有峰，3个组别各18批样品相似度分别均大于0.900；共指认出没食子酸、5-羟甲基糠醛、原儿茶酸、苦杏仁苷、咖啡酸、羟基红花黄色素A、芍药苷、阿魏酸、洋川芎内酯A、藁本内酯共10个成分；HCA、PLS-DA将54批样品分为2类，A组样品一类，B、C组样品一类；PLS-DA分析筛选出6个差异性色谱峰；根据PCA综合模型计算，综合评分结果为B组>C组>A组，TSD煎煮2次较为合理，冷冻干燥能较好保留煎液中的成分，色谱峰16（藁本内酯）、1（没食子酸）、13、12（阿魏酸）、7（苦杏仁苷）、15（洋川芎内酯A）是影响样品质量的标志性差异物质。结论 通过HPLC指纹图谱和化学模式识别够较好地辨识制备过程中影响TSD质量的标志性成分，为经典名方制备工艺评价和质量标志物辨识思路提供了参考。

Objective Based on the overall fingerprint information of the chemical constituents of traditional Chinese medicine compound, the changes in the chemical constituents of Taohong Siwu Decoction (TSD, 桃红四物汤) in the preparation process were investigated, and the fluctuation and main contributing constituents affecting the quality were analyzed to provide references for its quality evaluation and control. Methods According to the standard preparation process of the classic prescriptions, three preparation process samples were prepared:one-time decoction (group A), secondary decoction (group B), and freeze-drying (group C) were prepared, and the HPLC fingerprint of 18 batches of medicinal materials were established in three groups and processed by the fingerprint similarity evaluation system of traditional Chinese medicine, the SPSS 20.0 was used for systematic clustering analysis (HCA) and principal component analysis (PCA), and the SIMCA 14.1 software was used for partial least squares-discrimination analysis (PLS-DA), to analyze the dynamic changes of chemical fingerprint information in three groups of samples. Results Seventeen chromatographic peaks were selected as the common peaks of the fingerprint. The similarities of 18 batches of samples in each of the three groups were all greater than 0.900; Ten ingredients including gallic acid, 5-hydroxymethyl furfural, protocatechuic acid, amygdalin, caffeic acid, hydroxysafflor yellow A, paeoniflorin, ferulic acid, senkyunolide A, ligustilide were identified. HCA and PLS-DA divided 54 batches of samples into two types, one for group A, and one for group B and C; PLS-DA analysis screened out six different chromatographic peaks. According to the PCA comprehensive model calculation, the comprehensive scoring result was group B > group C > group A. It was reasonable to decoct TSD twice and freeze-drying can retain the components in the decoction. Chromatographic peaks 16 (ligustilide), 1 (gallic acid), 13, 12 (ferulic acid), 7 (amygdalin), and 15 (senkyunolide A) were the marked differential substances affecting the quality of the samples. Conclusion The HPLC fingerprint and chemical pattern recognition can better identify the iconic components that affect the quality of TSD during the preparation process. This paper provides a reference for the preparation process evaluation and identification of quality indicators of the classic prescriptions.

R286.02

四川省科技计划项目（2021YFS0256）；成都中医药大学杏林学者学科人才科研提升计划（QNXZ2018027）