Application of chemical pattern recognition and similarity evaluation in electrochemical and HPLC-DAD fingerprints for quality consistency study of herbal medicines.

Background: Herbal medicines and their preparations play a significant role in healthcare systems, yet concerns remain about their quality consistency. Chemical fingerprinting and multi-component quantitative analysis are the commonly used analytical methods and are widely applied in the quality analysis of herbal medicines. The study uses Gegen Qinlian tablets (GQTs) as a case to propose a comprehensive quality consistency evaluation system.

Results: Initially, the evaluation system is based on three quality components and three mixtures representing RPL, RS, and RC, categorizing all samples into eight levels. Subsequently, a four-wavelength fusion HPLC profiling (FWFP) method was established, yielding a relative standard deviation (RSD) of 0.43 % for mean relative retention times (RRA) and 21.12 % for relative retention area using the normalized fingerprint method (NFM). The systematically quantified fingerprint method (SQFM) was employed, resulting in qualitative similarity (Sm) and quantitative similarity (Pm) ranges of 0.878-0.978 and 74.9%-120.4 %, respectively. Concurrently, the Electrochemical Fingerprint method (ECFM) was applied for joint evaluation with FWFP, producing S and projection quantitative similarity (C) ranges of 0.962-1.000 and 70.6-155.2 %, respectively. Ultimately, the series spectra from FWFP and ECFM were used to comprehensively assess sample quality, with S and C ranges of 0.891-0.979 and 87.5-120.9 %, leading to the classification of the 22 GQT batches into five grades.

Significance: The study first proposes using characteristic parameters of the ECFM combined with S and C to evaluate the similarity of electrochemical fingerprints. It also comprehensively describes and uses SQFM to evaluate the quality of herbal medicines, including methodological validation, qualitative similarity (S), quantitative similarity (P), and reliability assessment. These methods may provide new insights for the similarity evaluation of different types of fingerprints, which can be applied in the quality consistency study of herbal medicines.