目的 建立HPLC指纹图谱与多成分含量测定方法，结合化学计量学分析，评价暖宫七味丸（Nuangong Qiwei Pills，NQP）的质量。方法 采用RP-HPLC法，色谱柱为Grace Alltima HP C₁₈柱（250 mm×4.6 mm，5 μm）；流动相为甲醇-水，梯度洗脱；体积流量为1.0 mL/min；检测波长为220 nm。对4家生产企业14批NQP进行指纹图谱及4种指标性化学成分含量测定，利用化学模式识别法通过SPSS 20.0对含量测定数据进行聚类分析和通过SIMCA 14.1软件对NQP 4种定量成分进行主成分分析（principal component analysis，PCA）并将共有峰按峰面积进行差异性分析。结果 建立了NQP指纹图谱，共匹配出21个共有峰，分别归属沉香、丁香、肉豆蔻、豆蔻4味药材，指纹图谱相似度>0.98；对照品指认沉香四醇（6号峰）、丁香酚（8号峰）、肉豆蔻木脂素（14号峰）、去氢二异丁香酚（20号峰）4种化学成分，质量分数分别为0.120~0.416、1.574~5.018、0.103~0.205、0.093~0.139 mg/g；当平方欧式距离为15时，14批样品通过聚类分析分成3类，PCA结果与聚类分析结果一致；通过对共有峰峰面积为变量进行正交偏最小二乘-判别分析建模分析，R_X²为0.696，Q²为0.597，均大于0.5，建立的模型稳定可靠，对照品指认的4种化学成分均为重要差异性标志化合物。结论 指纹图谱及含量测定方法准确可靠，可用于NQP质量控制及综合评价。

Objective To establish HPLC fingerprint and multi-component content determination method, combined with chemometrics, to evaluate the quality of Nuangong Qiwei Pills (NQP, 暖宫七味丸). Methods RP-HPLC was performed on a Grace Alltima HP C₁₈ (250 mm×4.6 mm, 5 μm) column with methanol-water gradient elution as mobile phase. The volume flow rate was 1.0 mL/L, and the detection wavelength was 220 nm. A total of 14 batches of sample from four pharmaceutical enterprise were studied by fingerprint analysis and determination of four index chemical components. Cluster analysis of chromatographic data was performed with SPSS 20.0 by chemical pattern recognition method and principal component analysis was performed with SIMCA 14.1 of NQP. Results The fingerprint of NQP was established, and 21 common peaks were calibrated, which belonged to four traditional Chinese medicinal materials:Chenxiang (Aquilariae Lignum Resinatum), Dingxiang (Caryophylli Flos), Roudoukou (Myristicae Semen), Doukou (Amomi Fructus Rotundus), fingerprint similarity were all above 0.98, four chemical components of agarotetrol (peak 4), eugenol (peak 8), myrislignan (peak 14), dehydrodiisoeugenol (peak 20) were identified by reference substances. The contents were 0.120-0.416, 1.574-5.018, 0.103-0.205, 0.093-0.139 mg/g. When the squared euclidean distance is 15, 14 batches of samples are divided into three class according to the pharmaceutical company, and the results of cluster analysis and principal component analysis were consistent. Through orthogonal partial least squares-discriminant analysis modeling analysis with common peak-peak area as variable, R_X² was 0.696 and Q² was 0.597, both of which were greater than 0.5, the established model was stable and reliable, and the four chemical components identified by the reference were all important difference marker compounds. Conclusion The fingerprint and content determination method were accurate and reliable, which can be used for quality control and comprehensive evaluation of NQP.

R283.6

内蒙古自治区科技计划项目（2020GG0199）；内蒙古自治区自然科学基金项目（2024FX13）；内蒙古自治区市场监督管理局服务产业高质量发展“蒙中药标准勘误项目”（［2021］09197）