The importance of matrix effect investigations in human biological matrices for accurate and sensitive quantification of polyphenols with LC-ESI/MS/MS

摘要

The application of LC-ESI/MS/MS analysis has expanded in bioanalytical and pharmaceutical research. Residual co-eluting matrix components, e.g. endogenous phospholipids, salts and mobile phase modifiers affect ESI-ionization of the target analytes. This can contribute to matrix effects (ME), resulting either ion suppression or ion enhancement. Aside from diverse matrices such as plasma and serum a high individual variability of different sources of e.g. plasma from the same species needs to be considered. Current FDA and EMA guidance documentation require that ME have to be evaluated in different lots of matrix as a part of quantitative LC/MS/MS method development and validation. There exist few main strategies to overcome ME like an adequate sample preparation, adjustments in the chromatography or calibration. We compared more than 30 different sample preparation techniques like protein precipitation, liquid-liquid extraction and solid-phase extraction in human plasma for quantification selected polyphenols at trace levels and evaluated the sample preparation recovery, quantitative ME and process efficiency. Besides the post-column infusion method for qualitative determination of ME, we used the post-extraction spike method which assesses quantitative ME by comparing the signal response of an analyte spiked postextracted into blank matrix to the response present in neat mobile phase. Moreover, we focused on studying and quantifying ME with the optimized sample clean-up in six different lots of plasma and in human pooled plasma, which is commonly used for calibration. Target analytes-free blank matrix for calibration might not always be available. We also investigated the influence of the total analyt concentration in human pooled and individual plasma on ME. Furthermore, we present a comparison of the accuracy and thus the performance of the quantification with a structural and a stable isotope labelled (SIL-) internal standard (IS) carried out using the example of an analyt with the use of spiked individual plasma samples. It is generally believed that SIL-IS yield better assay performance, but their use is rather expensive and for many compounds SIL-IS are not commercially available. By the selection of an appropriate sample clean-up and the associated resulting compensated ME the accuracy of our quantification with structural IS in different lots of plasma can compete with SIL-IS. Our investigations emphasize the importance of evaluating ME during development and validation of analytical methods in complex biological fluids such as plasma or serum. ME might also impact the sensitivity of an analytical method. If utilized for pharmacokinetic or clinical studies it is essential to examine the relative ME due to the individual variability of the matrix samples, which can compromise accuracy and yield to uncorrected quantification results.