Stepwise optimization approach for improving LC-MS/MS analysis of zwitterionic antiepileptic drugs with implementation of experimental design Nađa Kostić, a Yannis Dotsikas, b Anđelija Malenović, a * Biljana Jan  cić Stojanović, a Tijana Rakić, a Darko Ivanović a and Mirjana Medenica c In this article, a step-by-step optimization procedure for improving analyte response with implementation of experimental design is described. Zwitterionic antiepileptics, namely vigabatrin, pregabalin and gabapentin, were chosen as model com- pounds to undergo chloroformate-mediated derivatization followed by liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) analysis. Application of a planned stepwise optimization procedure allowed responses of analytes, expressed as areas and signal-to-noise ratios, to be improved, enabling achievement of lower limit of detection values. Results from the current study demonstrate that optimization of parameters such as scan time, geometry of ion source, sheath and auxiliary gas pressure, capillary temperature, collision pressure and mobile phase composition can have a positive impact on sensitivity of LC-MS/MS methods. Optimization of LC and MS parameters led to a total increment of 53.9%, 83.3% and 95.7% in areas of derivatized vigabatrin, pregabalin and gabapentin, respectively, while for signal-to-noise values, an improvement of 140.0%, 93.6% and 124.0% was achieved, compared to autotune settings. After defining the final optimal conditions, a time-segmented method was validated for the determination of mentioned drugs in plasma. The method proved to be accurate and precise with excellent linearity for the tested concentration range (40.0 ng ml À1 –10.0 Â 10 3 ng ml À1 ). Copyright © 2013 John Wiley & Sons, Ltd. Additional supporting material can be found in the online version of this article. Keywords: optimization; antiepileptics; ion source parameters; experimental design; response surface methodology Introduction Liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) constitutes the method of choice for determination and quantification of many structurally different compounds. Its inherent sensitivity can vary per case, depending on analyte properties, such as chargeability and relative hydrophobicity, [1] along with many other parameters. However, the need for more sophisticated protocols, requiring increased sensitivity in minimal volumes of biological media, is constantly growing. [2,3] To this purpose, there are two ways to follow: (1) purchase of new MS detectors with improved characteristics or (2) maximization of abilities of existing instrumentation. Experience and understanding of signal generation via MS are prerequisite for maximizing potential of the detector, as reflected in the ability to evaluate the effect of a series of LC-MS/MS pa- rameters on analyte signal. In general, most researchers conduct one basic optimization of MS parameters, mainly via (auto)tuning procedure, using the optimization software supplied by the equipment manufacturer, followed by selection of LC parame- ters, such as (1) appropriate stationary phase, (2) composition of mobile phase including amount and type of organic modifier, with or without adjustment of pH value, (3) flow rate and (4) col- umn temperature. Usually, this procedure is adequate, but when the need for sensitivity intensifies, additional optimization of MS parameters could be considered towards signal increment achievement. Optimization of experimental conditions in such complex sys- tem can be tedious or ineffective if not planned accordingly. In most cases, further optimization of MS parameters is attempted by performing one-factor-at-a-time (OFAT) experiments. [4–10] The drawback of this procedure is that the possible interactions among examined factors are not taken into account. [11] There- fore, application of optimal conditions, estimated via OFAT, is very probable to result in failure of method performance. Incorporation of chemometrics-based techniques offers valu- able means in many fields of science, enabling determination of * Correspondence to: Anđelija Malenović, Vojvode Stepe 450, 11000 Belgrade, Serbia. E-mail: andja@pharmacy.bg.ac.rs a University of Belgrade, Faculty of Pharmacy, Department of Drug Analysis, Vojvode Stepe 450, Belgrade, Serbia b University of Athens, School of Pharmacy, Department of Pharmaceutical Chemistry, Panepistimioupoli Zografou, Athens, Greece c University of Belgrade, Faculty of Pharmacy, Department of Physical Chemistry and Instrumental Methods, Vojvode Stepe 450, Belgrade, Serbia J. Mass Spectrom. 2013, 48, 875–884 Copyright © 2013 John Wiley & Sons, Ltd. Research article Received: 1 April 2013 Revised: 26 April 2013 Accepted: 9 May 2013 Published online in Wiley Online Library (wileyonlinelibrary.com) DOI 10.1002/jms.3236 875