Abstract Background SPE of biological matrices isolates analytes of interest from interfering matrix components that can affect LC-MS/MS quantitation. Optimal SPE conditions are dependent upon the sample matrix and analyte properties and typically require multiple experiments to determine optimal conditions for SPE extraction. Here, a 96-well method development plate was utilized to streamline method development for a panel of 47 clinical research drug analytes: 14 antipsychotics, 22 antidepressants and 11 anticonvulsants. Methods An SPE plate with four polymeric sorbents: reversed phase (X), strong and weak cation-exchange and weak anion-exchange mixed-mode phases (X-C, X-CW and X-AW) was used. The 47 drugs analytes were extracted on all four sorbents using three sets of conditions: acidic pretreatment/wash (aq) with basic elution (AB), neutral pretreatment/wash (aq) with neutral elution (NN), and basic pretreatment/wash (aq) with acidic elution (BA). A 50/50 methanol/water wash was also included. Elution used neutral or pH adjusted ethyl acetate/isopropanol or methanol/acetonitrile. A reagent blank, matrix blank, extracted serum samples spiked post-elution (n=2), and spiked serum (n=4) were extracted on each sorbent using each set of conditions. Optimal conditions were determined by calculating absolute percent recovery for each analyte. Results Recovery for antipsychotics was 85-97% using X-CW-AB, except Chlorpromazine (59%). It had better recovery (80%) using X-AB. Recovery for the antidepressants was 75-112% using X-C-AB, except Bupropion (35%) and Selegiline (10%). Bupropion had better recovery (83% and 95%) using X or X-CW and BA. Selegiline had better recovery using X, X-CW and X-AW with BA (83-92%). Anticonvulsants comprise zwitterionic, sulfonamides, neutral, acidic, and basic compounds with widely varying pKa and polarity. Targeting a single method for these analytes is difficult. The best recovery for anticonvulsants used X-CW-AB with several exceptions. Lacosamide, Felbamate, and Levetiracetam are neutral and among the most polar analytes and had 7-31% recovery under all SPE conditions tested. Zwitterionic Gabapentin and Pregabalin had 50-53% recovery from X-CW and X-C sorbents. The stronger protic elution (ammoniated methanol/acetonitrile vs ethyl acetate/isopropanol ) improved solubility for complete elution. Oxcarbazepine had 37% recovery from X-CW-AB, but recovery was 71% under BA conditions. Neutral, polar Zonisamide and Topiramate (Sulfonamide compounds) recovery was 25-26% using X-CW-AB. However, recovery was better (47% and 83%) under NN conditions, suggesting the analytes didn’t undergo complete elution from the X-CW sorbent. A stronger elution solvent would improve recovery from X-CW. Conclusions Most antipsychotics were recovered with X-CW-AB. Most antidepressants could be extracted using X-C-AB. Some analytes had higher recovery using hydrophobic interaction (X) SPE. A single mixed-mode SPE two-step elution protocol using a stronger elution solvent may be feasible. Anticonvulsants had mixed results. Some would require a stronger elution solvent. Some could be included in the other two panels to maximize recovery, but some had low recovery under all conditions tested. The method development plate screened 12 SPE methods for 47 analytes in a single experiment, greatly reducing the time and labor required to identify the best sorbent and starting conditions for SPE extraction optimization for drug analysis. This workflow could be implemented for other SPE method development projects.
Read full abstract