色谱系统显著变动对定量指纹图谱评价中成药质量的影响

目的考察在色谱系统显著变动条件下用系统指纹定量法(SQFM)对中成药进行评价产生的误差,并进行t检验。方法在理论上阐述SQFM 3个指标Sm、Pm和α间的关联性,在洗脱程序EP1条件下用HPLC测定12批朱砂安神丸(ZSASPs)在228、286和326 nm指纹图谱,分别以均值法整合三波长下定量指纹图谱的鉴定结果;11个月后在显著变动的洗脱程序EP2条件下同法试验和评价指纹图谱。对2次获得指纹图谱群的评价结果进行误差对比和t检验,并讨论质量鉴定等级和宏定量相似度变化。结果以小檗碱为参照物峰,EP1确定53(228 nm)、50(286 nm)和52(326 nm)个共有指纹峰,EP2确定50(228 nm)、40(286 nm)和35(326 nm)个共有指纹峰。经SQFM评价后的三波长指纹图谱结果按照均值法整合后对S4、S6^S8、S10S8、S10^S12等7批样品鉴定结果一致,对余下5批样品鉴定结果波动仅1级,排除S7和S8,EP2下各样品的定性相似度平均降低0.03,定量相似度误差基本在±5%之内,对2次获得鉴定结果进行T检验证明无显著性差异。对EP1和EP2三波长HPLC-FPs鉴定数据取平均,最终鉴定S1S12等7批样品鉴定结果一致,对余下5批样品鉴定结果波动仅1级,排除S7和S8,EP2下各样品的定性相似度平均降低0.03,定量相似度误差基本在±5%之内,对2次获得鉴定结果进行T检验证明无显著性差异。对EP1和EP2三波长HPLC-FPs鉴定数据取平均,最终鉴定S1^S4和S9质量极好(1级),S5、S6、S10和S12质量很好(2级),S11质量好(3级),S7和S8质量劣(8级)。结论在有效期内SQFM对同一总体样品以定量指纹图谱进行评价时,色谱条件的显著变化不影响鉴定结果,SQFM对中成药进行质量评判时具有包容性和耐用性。

Effect of the larger change in chromatographic elution program on evaluation of traditonal Chinese medicine by quantifi ed fi ngerprints

objective To determine how the deviation in the chromatographic system affect the evaluation of Chinese patent medicine with systematically quantifi ed fi ngerprint method（SQFM） and to perform T test. Methods Three indicators of SQFM（Sm, Pm and α） were theoretically elaborated to display the close ties in them. The fi ngerprints of 12 batches of Zhusha Anshen pills（ZSASPs） under elution program 1（EP1） by HPLC at 228, 286 and 326 nm were established and these 3 wavelength fingerprint identification results were integrated by the average method. After 11 months, the same procedures were repeated in the same samples under the signifi cantly changed elution program 2（EP2） to obtain the deviations examined by t test. The changes in quality grade and the macro quantitative similarity were both discussed. Results While berberine（BBR） was selected as the reference peak, 53（228 nm）, 50（286 nm） and 52（326 nm） common peaks were identifi ed by EP1, while 50（228 nm）, 40（286 nm） and 35（326 nm） common peaks by EP2. After evaluating the the 3 wavelength fi ngerprints with SQFM through averaging method, S4, S6- S8 and S10- S12 were identifi ed to be consistent under 2 EP conditions, and the remaining 5 batches of samples fl uctuated by only 1 grade. Excluding S7 and S8, the qualitative similarity of each sample under condition EP2 decreased 0.03 averagely, and the quantitative similarity errors were all within ±5%, and t test was performed to prove no signifi cant difference in the 2 identifi cations. The 3 wavelength HPLC-FPs identifi cation data under EP1 and EP2 identifi ed S1, S2, S3, S4 and S9 as of the best quality（grade 1）, and S5, S6, S10 and S12 of better quality（grade 2）, S11 of good quality（grade 3）, S7 and S8 of worse quality（grade 8）. Conclusion When the same samples are evaluated by quantitative fi ngerprints in SQFM, signifi cant changes in gradient elution conditions do not affect the identifi cation, indicating SQFM for Chinese Patent medicine is r