20
14 BEST PRACTICES IN BIOLOGICAL SAMPLE PREPARATION FOR LC-MS BIOANALYSIS Guowen Liu and Anne-Franc ¸oise Aubry 14.1 WHY DO SAMPLE PREPARATION? Liquid chromatography coupled with mass spectrometry (LC-MS) or tandem mass spectrometry (LC-MS/MS) has become the benchmark for bioanalysis for small molecules in past decades (Watt et al., 2000; Jemal and Xia, 2006; Aubry, 2011). LC-MS/MS is also now becoming an impor- tant tool for quantitative analysis of larger molecules (pep- tides, oligonucleotides, and proteins) in biological matrices (Ewles and Goodwin, 2011; Li et al., 2011). As powerful and sophisticated a tool as an LC-MS system can be, its perfor- mance may be compromised by the nature of the samples analyzed. There are certain physical and chemical require- ments for samples that are being analyzed by LC-MS. Most biological specimens cannot be directly introduced to the LC-MS system without some pretreatment. The principles of sample pretreatment are the same regardless of the par- ticular MS detection technology (single quadrupole, triple quadrupole, other tandem MS technology, or high-resolution MS) being used for analysis. This chapter reviews the most commonly used sample preparation techniques in the phar- maceutical and biopharmaceutical industry for LC-MS or LC-MS/MS analyses (for simplicity, we will use LC-MS in the rest of the chapter). The discussion deals primarily with small molecules but we also pointed out where some of the techniques can be applied to large molecules. As a hybrid instrument, LC-MS needs to follow the requirements of both of its components. The LC part is designed to deal with liquid samples. The flow path from the solvent bottle, through HPLC pump, injection valve, LC col- umn to the LC column exit is a long and delicate system. Sam- ples injected into the LC system are required to be in liquid form, and essentially free of particles. Therefore, samples not already in liquid form, such as tissue samples (solid), blood samples (suspension), have to be processed or converted into liquid form before being loaded into the LC system. The MS is designed to generate ions into the gas phase and detect them under a high vacuum condition. Electrospray ionization (ESI) and atmospheric pressure chemical ionization (APCI) are the two most common ion sources used in an LC-MS sys- tem. To ensure good ionization efficiency, nonvolatile acid, base, or salts should not be introduced into the ESI or APCI ion source. In addition, unless a special column/mobile phase combination is used, proteins cannot be injected in the LC as they tend to precipitate at the head of the column and cause a rapid deterioration of column performance. Further- more, matrix effect, interfering effects (ion enhancement or ion suppression) from matrix components to the analyte of interest during ionization process, is a well-described phe- nomenon in ESI/APCI MS (Fu et al., 1998; King et al., 2000; Dams et al., 2003; Wu et al., 2008b; Liu et al., 2009). Matrix effect is unpredictable and can lead to inaccurate mea- surements if not addressed properly (Wang et al., 2007; Liu et al., 2010a). Using a stable isotope labeled internal standard (SIL-IS) can compensate matrix effects in most cases. How- ever, when matrix effect is dramatically different from sample to sample (Liu et al., 2010a) or the SIL-IS does not exactly coelute with the analyte chromatographically (Jemal et al., Handbook of LC-MS Bioanalysis: Best Practices, Experimental Protocols, and Regulations, First Edition. Edited by Wenkui Li, Jie Zhang, and Francis L.S. Tse. C 2013 John Wiley & Sons, Inc. Published 2013 by John Wiley & Sons, Inc. 165

Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

Embed Size (px)

Citation preview

Page 1: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

14BEST PRACTICES IN BIOLOGICAL SAMPLEPREPARATION FOR LC-MS BIOANALYSIS

Guowen Liu and Anne-Francoise Aubry

14.1 WHY DO SAMPLE PREPARATION?

Liquid chromatography coupled with mass spectrometry(LC-MS) or tandem mass spectrometry (LC-MS/MS) hasbecome the benchmark for bioanalysis for small moleculesin past decades (Watt et al., 2000; Jemal and Xia, 2006;Aubry, 2011). LC-MS/MS is also now becoming an impor-tant tool for quantitative analysis of larger molecules (pep-tides, oligonucleotides, and proteins) in biological matrices(Ewles and Goodwin, 2011; Li et al., 2011). As powerful andsophisticated a tool as an LC-MS system can be, its perfor-mance may be compromised by the nature of the samplesanalyzed. There are certain physical and chemical require-ments for samples that are being analyzed by LC-MS. Mostbiological specimens cannot be directly introduced to theLC-MS system without some pretreatment. The principlesof sample pretreatment are the same regardless of the par-ticular MS detection technology (single quadrupole, triplequadrupole, other tandem MS technology, or high-resolutionMS) being used for analysis. This chapter reviews the mostcommonly used sample preparation techniques in the phar-maceutical and biopharmaceutical industry for LC-MS orLC-MS/MS analyses (for simplicity, we will use LC-MS inthe rest of the chapter). The discussion deals primarily withsmall molecules but we also pointed out where some of thetechniques can be applied to large molecules.

As a hybrid instrument, LC-MS needs to follow therequirements of both of its components. The LC part isdesigned to deal with liquid samples. The flow path from the

solvent bottle, through HPLC pump, injection valve, LC col-umn to the LC column exit is a long and delicate system. Sam-ples injected into the LC system are required to be in liquidform, and essentially free of particles. Therefore, samples notalready in liquid form, such as tissue samples (solid), bloodsamples (suspension), have to be processed or converted intoliquid form before being loaded into the LC system. TheMS is designed to generate ions into the gas phase and detectthem under a high vacuum condition. Electrospray ionization(ESI) and atmospheric pressure chemical ionization (APCI)are the two most common ion sources used in an LC-MS sys-tem. To ensure good ionization efficiency, nonvolatile acid,base, or salts should not be introduced into the ESI or APCIion source. In addition, unless a special column/mobile phasecombination is used, proteins cannot be injected in the LCas they tend to precipitate at the head of the column andcause a rapid deterioration of column performance. Further-more, matrix effect, interfering effects (ion enhancement orion suppression) from matrix components to the analyte ofinterest during ionization process, is a well-described phe-nomenon in ESI/APCI MS (Fu et al., 1998; King et al.,2000; Dams et al., 2003; Wu et al., 2008b; Liu et al., 2009).Matrix effect is unpredictable and can lead to inaccurate mea-surements if not addressed properly (Wang et al., 2007; Liuet al., 2010a). Using a stable isotope labeled internal standard(SIL-IS) can compensate matrix effects in most cases. How-ever, when matrix effect is dramatically different from sampleto sample (Liu et al., 2010a) or the SIL-IS does not exactlycoelute with the analyte chromatographically (Jemal et al.,

Handbook of LC-MS Bioanalysis: Best Practices, Experimental Protocols, and Regulations, First Edition. Edited by Wenkui Li, Jie Zhang, and Francis L.S. Tse.C© 2013 John Wiley & Sons, Inc. Published 2013 by John Wiley & Sons, Inc.

165

Page 2: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

166 BEST PRACTICES IN BIOLOGICAL SAMPLE PREPARATION FOR LC-MS BIOANALYSIS

2003; Wang et al., 2007), even a SIL-IS may not be able tocompensate for all of the matrix effects. Many matrix com-ponents are hydrophilic and can be eliminated by simpleextraction procedures for assays of moderately polar tohydrophobic compounds. Phospholipids have been singledout for their propensity to cause matrix effects for pharma-ceuticals (Liu et al., 2009; Xia and Jemal, 2009) and muchinformation is available regarding the removal of phospho-lipids from plasma (Liu et al., 2009; Pucci et al., 2009; Xiaand Jemal, 2009; Jiang et al., 2012). Sensitivity enhance-ment is another reason for sample preparation. To detecttrace amounts of analyte in a complex system, such as abiological matrix, the analyte of interest often needs to beenriched to reach the desired detection limit. In these cases,the extraction procedure is used to preconcentrate the ana-lyte prior to LC-MS analysis. To summarize, the goals ofsample preparation are to change the sample form (e.g., pro-cess a solid or heterogeneous specimen to a clear solution),simplify the sample composition (i.e., reduce matrix back-ground), or enrich the analyte of interest (i.e., preconcentratethe sample). Sample preparation is also one of the most, ifnot the most, critical and time consuming part of a bioanalyt-ical assay. A successful analytical strategy should not onlyachieve the goals of sample preparation but also preserve theintegrity of the analyte (i.e., avoid degradation of the ana-lyte of interest). Shown in Figure 14.1 is a schema of themost commonly used strategies by bioanalytical scientistsfrom simply converting the specimen to a sample physically

amenable to LC-MS analysis to using an analyte-specificextraction.

Many up-to-date, high quality and comprehensive reviewson the principles and advancements of sample prepara-tion techniques for LC-MS bioanalysis have been published(Chang et al., 2007; Novakova and Vlckova, 2009; Ashri andAbdel-Rehim, 2011; Kole et al., 2011). Technique-specificor application-oriented reviews are also available: Moreno-Bondi et al. (2009) reviewed the sample preparation proce-dures for LC-MS analyses of antibiotics in environmental andfood samples; Samanidou et al. ( 2011) reviewed novel strate-gies for sample preparation in forensic toxicology; Rudewicz(2011) reviewed the applications of turbulent flow (TFC)in drug metabolism and pharmacokinetics; Vuckovic et al.(2010) updated the status of solid-phase microextraction inbioanalysis; Chen et al. (2009) summarized the applicationof online SPE for liquid chromatography.

In this chapter, we are describing the commonly usedsample preparation techniques in the pharmaceutical andbiopharmaceutical industries, namely, protein precipitation(PPT), liquid–liquid extraction (LLE), and solid phaseextraction (SPE), focusing on the need for balancing recov-ery and selectivity, and on the operational details of eachtechnique. A scientifically sound bioanalytical method canstill fail due to incorrect execution of the sample preparationprocedures. Many details deemed trivial are, in fact, criticalto the assay’s successful execution: nonspecific absorptionon plasticware (Ji et al., 2010; Li et al., 2010), ineffective

Dilution(all inclusive)

PPT(excludingproteins)

Nonliquidsamples (tissue,organ,dried blood

Liquidsamples(plasma,urine, tissue homogenate,

LLE(similarpolarity

compoundsextracted)

Injectionsolution

LC-MS

spot, etc.) etc.)

SPE(analyte alike

compoundextracted)

Immuno-affinity extraction(only the analyteextracted)

FIGURE 14.1 General overview of sample preparation of biological samples. PPT, protein pre-cipitation; LLE, liquid–liquid extraction; SPE, solid phase extraction.

Page 3: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

WHAT SHOULD YOU KNOW ABOUT THE SAMPLE? 167

AH2

% F

orm

ati

on

AH-

A2-

100

120

60

80

100

20

40

-20

00 2 4 6 8 10 12

pH

FIGURE 14.2 Illustration of different species formation at different pHs.

sample mixing prior to pipetting (Fu et al., 2011), cross-contamination during sample handling and liquid transfers,carryover from fixed-tip liquid handler, etc. In our opinion,sample preparation is all about execution of the details, espe-cially for those new to this field. Many good bioanalyticalassays fail in cross-laboratory transfers due to lack of suffi-cient sample preparation details in the written method. There-fore, our focus in this chapter is explaining the specificity ofeach sample preparation techniques and the advantages anddisadvantages and the applications of each technique. Recentvariations of these sample preparation techniques will also bediscussed briefly. Before we discuss each specific technique,we thought it important to review the fundamental chemicalprinciples of sample extraction.

14.2 WHAT SHOULD YOU KNOW ABOUTTHE SAMPLE?

14.2.1 Analyte of Interest

Before one starts working on a sample preparation method,it is always a good practice to know the relevant chemi-cal and physical properties of the compound(s) of interest.Knowledge of the properties of organic compounds (log P,log D, pKa, solubility, protein binding, stabilities, etc.) iscritical to understand their compatibility with the varioussample preparation techniques and the selection of exper-imental conditions. Therefore, it is strongly recommendedto spend some time reviewing these properties and all theoptions in planning experiments on paper before heading tothe laboratory.

14.2.1.1 pKa Many organic compounds contain acidic(H+ donor) or basic (H+ acceptor) functional groups.Knowing the pKa of a compound can inform the charge state

of the compound at a given pH. By definition, at pH = pKa,50% of the molecules are in the basic (protonated) form and50% in the acidic (deprotonated) form; at pH � pKa, almost100% of the compounds will be unionized for acids and ion-ized for bases; at pH � pKa, almost 100% of the compoundswill be ionized for acids and unionized for bases. One can,therefore, map out the charge state of the molecule at variouspHs according the formula log [AH]/[A−] = pKa – pH. Forexample, assuming an acid (AH2) with two pKa (pKa1 = 3,pKa2 = 7), the proportion of different species (AH2, AH−,and A2−) can be mapped out at different pHs as illustratedin Figure 14.2. pH is a critical parameter in sample extrac-tion for ionized or ionizable compounds. In LLE, the bestrecovery will be at a pH at which the analyte is not charged;however, in ion-exchange SPE, the analyte must be ionizedto interact with the stationary phase. Compounds containingboth an acidic and a basic group form zwitterions at pHs inbetween the two pKa. Zwitterions are particularly difficult toextract. These basic rules regarding pH are often overlookedduring the initial stage of method development.

14.2.1.2 log P and log D The partition coefficient, P, is aconstant to characterize the hydrophilicity–lipophilicity bal-ance of a compound. It is the concentration ratio of the neutralform of a compound in two immiscible solvents (normallywater and octanol) at equilibrium. To measure the P value ofan ionizable compound, the predominant form of the com-pound in aqueous phase should be unionized by adjustingthe pH. The “log P” value of a compound is often referred toas its lipophilicity, which is the logarithm of the ratio of theconcentrations of the unionized compound in the solvents(see below for a formula for log P):

log P = log10[Co/Caq],

Page 4: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

168 BEST PRACTICES IN BIOLOGICAL SAMPLE PREPARATION FOR LC-MS BIOANALYSIS

where Co is the concentration of the neutral compound in thewater-immiscible solvent and Caq is the concentration in theaqueous phase.

The distribution coefficient, D, is another parameter tomeasure the lipophilicity of a compound and is defined asthe ratio of the sum of the concentrations of all forms of thecompound between an aqueous phase and a water-immiscibleorganic solvent. This parameter is both solvent and pH depen-dant for ionizable compounds. The “log D” value is oftengiven for a specific compound, which is derived from thefollowing formula:

log D = log10 Co/[Caq ion + Caq neu],

where Co is the concentration of the compound in the water-immiscible organic solvent, Caq ion is the concentration of theionized compound in water, and Caq neu is the concentrationof the neutral compound in the aqueous phase.

For unionizable compounds, log P = log D at any givenpH. For ionizable compounds, a relationship can be estab-lished as the following between log P and log D at a givenpH with the known pKa, providing that charged moleculesdo not enter the organic phase:

Acids : log D = log P − log [1 + 10(pH−pKa)]Bases : log D = log P − log [1 + 10(pKa−pH)]

A more detailed description of log D and log Pcan be found in this website: http://en.wikipedia.org/wiki/Partition_coefficient (accessed Mar 19, 2013). Unlessspecified, the log D and log P values of a given compoundare measured using octanol as the partitioning solvent bydefault.

The recovery of a compound at given conditions (pH,organic solvent, volume ratio between aqueous and organic)can be predicted for LLE methods and to an extent forreversed-phase SPE methods, based on their pKa, log D,and log P. The knowledge of the properties of the analyte,such as chemical stability, protein binding, solubility, and soon, will also help eliminate some obviously bad choices forsample preparation. Typically, more than one sample prepa-ration techniques can be used for an analyte of interest and thechoice is, by in large, a matter of personal preference. How-ever, for analytes with unique physicochemical properties,some techniques or experimental parameters are obviouslynot good choices: for example, choosing LLE for compoundwith log D < 0; modifying the pH to a range where the com-pound is unstable; PPT for analytes that will likely coprecip-itate with the proteins. In summary, understanding the ana-lyte well is the first step of developing a sample preparationmethod. In addition, the expected concentration of the ana-lyte in the matrix should be a factor to consider in choosingthe extraction technique.

14.2.2 Matrix

One challenge in bioanalysis is dealing with a variety of bio-logical matrices: plasma, serum, whole blood, urine, CSF,tissues, etc. The composition and complexity of the differentmatrices is significantly different. Knowing the nature of thesample is one of the key factors in deciding on a techniquefor sample preparation. Factors such as the normal pH of thematrix, nature and concentration of proteins and lipids, saltcontent, may all potentially impact the extraction. Under-standing the interactions between the analyte and the matrixit resides in also helps to decide a suitable sample prepa-ration technique. For example, when working with wholeblood samples, the distribution of the compound betweenplasma and red blood cells (RBC) (Brockman et al., 2007),the enzymatic stability of the analyte in blood (Li et al., 2011),protein binding, etc., should be considered. A sample prepa-ration method that is capable of releasing the analyte fromthe RBC is needed when the analyte distribution into RBCsis significant (Brockman et al., 2007). For tissue samples,checking that the analyte is released from the solid tissue is amust (Chng et al., 2010); for urine and CSF samples, nonspe-cific binding are often a concern in the absence of proteinsand the procedure must be able to release the analyte fromsurfaces (Gu et al., 2010; Ji et al., 2010). Finally, matrixeffects in the LC-MS analysis, already discussed in Section14.1, is an important aspect of the matrix to evaluate and onethat may drive the selection of the extraction method. Onething worth mentioning is that extracted matrix backgroundfor a specific technique/method is not compound dependantbut sample preparation method dependant (Liu et al., 2009);this will be discussed more in Section 14.3.3.

14.3 KNOW YOUR TOOLS

Liquid handling is a predominant part of sample preparation.The steps of pipetting biological samples, adding extractionsolvent, transferring intermediate solutions, and so on canbe done either manually using pipettes or using automation(e.g., robotic liquid handlers). Automation is the trend and thepreferred solution for liquid handling for sample preparationbecause of the better precision achieved, as well as ergonomicconcerns with repeated motion. Many sophisticated liquidhandling workstations are available, such as the FreedomEVO R© from Tecan (http://www.tecan.com, accessed Mar19, 2013), MICROLAB R© STAR Liquid Handling Work-stations from Hamilton (http://www.hamiltoncompany.com,accessed Mar 19, 2013), and JANUS R© Automated Work-station from PerkinElmer (http://www.perkinelmer.com,accessed Mar 19, 2013). A good review by Vogeser andKirchhoff (2011) on the progress of automation for LC-MSbioanalysis has been recently published. A chapter on the useof automation can also be found in this book.

Page 5: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

KNOW YOUR TOOLS 169

14.3.1 Dilution

Dilution is the simplest sample preparation technique(Casetta et al., 2000; Dams et al., 2003; Rashed et al., 2005),and is also referred to as “dilute-and-shoot” (McCauley-Myers et al., 2000; Xue et al., 2007). It is an inclusivemethod, which yields 100% recovery of all analytes but with-out selectivity. All the components of a sample are loaded intothe LC-MS system in this approach. The sample is simplydiluted using a solution amenable for LC-MS (Wood et al.,2004; Raffaelli et al., 2006; Bishop et al., 2007; Gray et al.,2011). The applications of dilution have been mainly lim-ited to situations in which matrix effect is not a concern andsensitivity is not an issue. The advantages of this approachinclude minimum sample handling, good sample integrity,and low cost. Its disadvantages include dilution of the sig-nal and high potential for matrix effect. For these reasons,dilution is often limited to relatively simple liquid matricescontaining no or few macromolecules, such as urine, tears,and CSF, and to applications that do not require high sen-sitivity. In reality, with the availability of more and moresensitive instrumentation and stable isotope labeled internalstandards (ISs), sample preparation using dilution becomes aviable solution in bioanalysis for more applications and onethat should be considered first because of its simplicity. Forthe most part, simple dilution is unsuitable for more complexbiological samples that contain proteins or lipids.

14.3.2 Protein Precipitation

Biological samples, such as plasma and serum, whichcontain abundant soluble proteins, require more complexpretreatment. The process of removing the majority of pro-teins from the biological samples by precipitating out theproteins is known in bioanalysis as PPT. It is a simple,quick, and convenient sample preparation technique, favoredfor high-throughput, non-GLP applications. Although manyapproaches (Polson et al., 2003) can be used to denatureproteins: such as water miscible organic solvent, acid, salts,and metal ions, PPT using an organic solvent (methanol,acetonitrile, or ethanol) seems to be dominant in bioanalyt-ical practice (Dams et al., 2003; Flaherty et al., 2005; Xueet al., 2006b; Ma et al., 2008; Chng et al., 2010). Polsonet al. (2003) reported a comprehensive study on the effec-tiveness of protein removal using acids (trichloroacetic acid,TCA), metal ions (zinc sulfate), salts (ammonium sulfate),and organic solvent (acetonitrile, ethanol, methanol) as theprecipitants. Based on the results from this study, more than90% of the total proteins can be removed from rat, dog,mouse, and human plasma with a precipitant-to-plasma ratioof 0.5 (10% TCA, v/v), 2 (10% zinc sulphate, w/v), 3 (satu-rated ammonium sulfate solution at room temperature), 1.5(acetonitrile), 3 (ethanol), and 2.5 (methanol). A typical pro-tocol of PPT with an organic solvent is shown later with

practical details and tips provided at each step. Since 96-wellplates are the most popular sample preparation format of bio-analysis in the pharmaceutical industry, all examples givenin this chapter will be assuming this format is used unlessotherwise stated.

14.3.2.1 A Typical Example of PPT Protocol In theexample of a typical PPT method for plasma samples usingacetonitrile as the organic solvent, the overall process can becompleted in 1–3 h for one batch, depending on whether adry-down and reconstitution process is included:

� Transfer 50 μl of standards (STDs), quality controls(QCs), and incurred samples into a 96-well plate.� This step can be performed either manually or using

a robotic liquid handler. The volume transferred foreach sample does not have to be accurate but has tobe precise. The 96-well plate used must have a coverwith perfect seal to prevent any solvent leaking andavoid well-to-well cross-contamination.

� Add 50 μl of IS working solution into each sample well.� The IS is added before the organic solvent. The IS may

also be prepared in the organic solvent and then thisstep can be combined with the next step. When doingso, one must be wary of less than perfect trackingby the IS, especially if the analyte is highly proteinbound. This step can easily be performed using roboticliquid handler. The volume delivered during this stepneeds to be precise but not necessarily accurate.

� Add 600 μl acetonitrile to each sample, seal the plateand vortex for 1 min.� Water-miscible organic solvent was added to the

aqueous solution. The organic solvent will dena-ture the soluble proteins and cause the proteinsto precipitate out of solution. To remove a major-ity of the proteins (e.g., >90%) in the sample, aminimum ratio of organic to aqueous of 1.5 for ace-tonitrile is recommended. This step can also be per-formed using automation. The plate has to be sealedseamlessly to avoid cross-well contamination. Eventhough automation of the plate sealing and vortexingsteps is possible, it is not prevailing in the bioana-lytical community. This step does not require highvolume precision or accuracy.

� Centrifuge the sample plate at 3000 rpm for 5 min.� Manual intervention is usually required: for example,

the plate is manually transferred to a centrifuge forcentrifugation and returned to the robotic workstationafter that.

� Transfer 500 μl of the supernatant after centrifugationinto a collection plate� Depending on the sensitivity requirement, different

amounts of supernatant may be transferred out. The

Page 6: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

170 BEST PRACTICES IN BIOLOGICAL SAMPLE PREPARATION FOR LC-MS BIOANALYSIS

supernatant contains high organic content; althoughthis transfer step can be easily automated, cautionmust take to avoid liquid dripping during the transferto avoid cross-contamination. Since the IS will com-pensate for any variation in the pipetting, this stepdoes not require high precision or accuracy.

� Dry the sample plate under nitrogen flow and reconsti-tute each sample with 200 μl reconstitution solution.� There are two purposes in this step. First, the sample

can be concentrated as necessary by reconstituting ina smaller volume (2.5 times in this example); sec-ond, the injection solution strength can be adjustedto match the starting LC conditions. This step maybe omitted if no sample concentration is needed andhigh organic content solution can be injected directlyinto the LC system (e.g., on a HILIC column) directly.This step usually takes most of the sample prepara-tion time because of the time required for evaporatingwater.

� Vortex the plate for 2 min and centrifuge the plate for2 min and load the final sample plate to the LC-MSsystem for analysis.� This step is to make sure that all the dry residue in

each well has been redissolved in the reconstitutionsolution and any solution on the well wall or the topis brought down to the bottom of each well. For somesamples, not all the solid residue will be completelyredissolved in the reconstitution solution. However,this may not be critical as long as the analyte of interestand its IS are fully dissolved in the final solution andsolid particles are not in suspension.

14.3.2.2 Advantages and Limitations There are manyadvantages of using PPT for sample preparation: (1) it isa sample preparation method suitable for almost all types ofsmall molecules, no matter their polarity; (2) operationally,a standard procedure can be applied to all assays; (3) it is arapid process with good potential for automation or semiau-tomation (Ma et al., 2008; Tweed et al., 2010); and (4) therecovery of the analyte is literally 100% and almost all smallmolecules can be retained, which is, therefore, particularlyadvantageous for metabolite profiling (Wilson, 2011). Theless attractive part of this technique lies in the fact that theprocessed samples after PPT still contain all small endoge-nous molecules (no selectivity) that may interfere with theMS ionization process, decrease the performance of the LCcolumn, and so on. PPT has been associated with high matrixeffect (Dams et al., 2003). These caveats, if not addressedproperly, may significantly impact the data quality (Wanget al., 2007). PPT is very popular in drug discovery, wheresample turnaround time is critical to keep up with the fastpace of discovery activities. In drug development, it wasextremely popular when ESI-LC-MS was first introduced as

the best tool for bioanalysis, until the issue of matrix effectbecame a major concern. It is regaining its popularity, espe-cially when a stable isotope labeled IS is available, becausewith new instruments becoming more and more sensitive,only a small amount of biological material is injected practi-cally eliminating concerns with matrix interference.

14.3.2.3 Other Forms of PPT To simplify the procedureand improve sample throughput further, membrane-basedPPT filter plates (e.g., Unifilter R© from Whatman, Strata R©

ImpactTM from Phenomenex, Isolute R© PPT + from Biotage,SiroccoTM from Waters) are available to remove proteins inplasma or serum. In recent years, more advanced PPT plateswith packed materials specifically retaining phospholipidsare also available to remove both proteins and the abundantphospholipids in a plasma/serum sample (e.g., CaptivaTM

NDLipids from Agilent, OstroTM from Waters, PhreeTM fromPhenomenex, and HybridSPE R© from Sigma). By using theseplates, an extracted sample free of protein and phospholipidsare readily available for the subsequent LC-MS analysis. Theoverall PPT process can be done in a 96-well plate formatwithout the needs for centrifugation or supernatant trans-fer steps, which leads to short sample preparation time andhigher solvent recovery. It is a very helpful approach for lowsample volume and is amenable to full automation. Eachof the commercially available PPT plates comes with stan-dardized recommended procedures. A detailed comparisonof different PPT plates/tubes can be found in a recent reviewarticle by Kole et al. (2011).

14.3.3 Liquid–Liquid Extraction

LLE is another popular sample cleanup technique in bioanal-ysis. It is based on partition between immiscible solvents andconsists in extracting compounds from the aqueous sampleinto a water-immiscible solvent. The distribution of the ana-lyte between the two phases is related to its log P or log Dvalues. When LLE approach is used, the biological sampleusually is mixed with an aqueous buffer (sample extractionbuffer to bring the pH of the matrix to a pH range in which ananalyte of interest can be readily transferred from the aque-ous phase to the organic phase). The buffer may also containthe IS. The buffered sample is then mixed with a speci-fied volume of a water-immiscible organic solvent. Com-monly used organic solvents include methyl tert-butyl ether(MTBE) (Pin et al., 2012), ethyl acetate (Kosovec et al., 2008;Cai et al., 2012), hexane (Cai et al., 2012), 1-chlorobutane(Saar et al., 2010), or a combination of two solvents and areselected based on the polarity and solubility of the analyte.By changing the organic solvent or adjusting pH and bufferionic strength, the target analyte(s) can be favorably extractedfrom the aqueous phase to the organic phase, leaving mostof matrix components, such as phospholipids, proteins, inor-ganic salts in the aqueous phase. The resulting extracted

Page 7: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

KNOW YOUR TOOLS 171

sample is much cleaner/simpler than the original biologicsample. Compared with PPT and even with SPE, LLE is amore selective sample cleanup technique. Below is a typicalprotocol of LLE in 96-well format. The overall process cannormally be completed in less than 2 h with automation.

14.3.3.1 A Typical Example of LLE Protocol

� Transfer 50 μl of plasma samples (STDs, QCs, andincurred samples) to each well in a 96-well plate.� Same as PPT.

� Add 50 μl of IS working solution to each sample.� Operationally, this step is the same as for PPT. How-

ever, it is strongly recommended to prepare the ISworking solution in a solvent that contains the mini-mum amount of water-miscible organic solvent, as thepresence of the water-miscible solvents will greatlyimpact the cleanness of the LLE extracts. This steprequires high pipetting precision but not necessarilyhigh accuracy.

� After 10 min, add 100 μl sample extraction buffer(0.5 M ammonium acetate with 2% formic acid) to eachsample.� A 10-min equilibrium time is enough to allow the IS to

equilibrate with the matrix and proteins in particular tobetter track the analyte extraction recovery. A strongextraction buffer is to make the analyte in a neutralform so that it can be readily extracted by the organicsolvent. This step can be easily automated. It (andthe following steps) does not require high precisionor accuracy in pipetting as the IS will compensate forany variability.

� Add 600 μl of MTBE solvent to each sample.� Water-immiscible organic solvent is added. This step

can be easily automated, although some solvents aredifficult to pipette and tend to drip. Adjustments inthe setting of the liquid handler are needed based onthe solvent type.

� Shake the whole plate for 15 min.� Automation for the shaking step is possible but dif-

ficult. However, it has been reported (Wang et al.,2006) that this step can be replaced by aspirating anddispensing multiple times, which can then be readilyautomated. A good seal of the plate before shaking iscritical to avoid cross-well contamination. It was alsoreported (Aubry, 2011) that a sweet spot may existfor shaking time with regard to analyte recovery andmatrix effect.

� Centrifuge the plate for 5 min at 4000 rpm.� This step can be automated but automation for this

step is not popular.� Transfer 500 μl of the top organic layer of each sample

after centrifugation to a collection plate.

� Since manual transfer of the organic layer may causeliquid dripping, it is recommended to perform thisstep using a liquid handler, which can be easily imple-mented.

� Dry down the plate under nitrogen flow.� Operationally this step is similar to PPT but usually

evaporation takes less time because the organic sol-vent is more volatile than water.

� Reconstitute each sample with 200 μl of reconstitutionbuffer.� Operationally this step is similar to PPT. This step

is almost always needed in this case, as the extrac-tion solvent is rarely compatible with RP-LC mobilephases. However, it has been reported (Song andNaidong, 2006; Xue et al., 2006b) that the organicphase can be directly injected into the LC systemwhen HILIC column is used.

Since LLE is a moderately selective sample preparationmethod, the sample preparation process can be adjustedto balance recovery and selectivity (matrix backgroundremoval) depending on the assay needs. When matrix inter-ference is a concern, recovery may be sacrificed for selectiv-ity. Normally, when a SIL-IS is used, the extraction processcan be perfectly tracked by the IS. Therefore, near-completerecovery or even consistent recovery from sample to sampleis not critical. It is, however, desirable to achieve consistentrecovery, which is a good indication of the ruggedness of theprocedure. Otherwise, sample preparation will always haveto be focused on both the absolute recovery (the higher thebetter) and the consistency of the recovery from sample tosample. It is strongly recommended to use a SIL-IS or aclosely related chemical analog whenever possible.

As for its applications, LLE is ideal for nonpolar to moder-ately polar analytes, which have favorable distribution in thewater-immiscible organic solvent and, therefore, can be read-ily extracted from the aqueous phase. In reality, many drugcandidates are nonpolar or moderately polar compounds, anda suitable LLE extraction method can often be developed.However, when a bioanalytical assay is targeting more thanone analyte, for example, a drug and its metabolites, a finetuning of the LLE conditions may be required to achieveacceptable recoveries for both analytes (Patel et al., 2008).

14.3.3.2 Advantages and Limitations One advantage forLLE is scalability, for example, a large sample volume can beused to increase the assay sensitivity if needed. Because theextracted sample is relatively clean, the signal gain from thesample volume increase will most likely surpass the signalloss from additional matrix suppression. Another practicaladvantage for LLE is that the extracted matrix backgroundis predictable once the extraction conditions are determined(e.g., ethyl acetate at pH 7) for a given biological matrix

Page 8: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

172 BEST PRACTICES IN BIOLOGICAL SAMPLE PREPARATION FOR LC-MS BIOANALYSIS

100H

0E

90H10

E

80H20

E

70H30

E

60H40

E

50H50

E

40H60

E

30H70

E

20H80

E

10H90

E

0H10

0E PPT

LLE Extraction solvent combination

Rel

ativ

e P

C r

emai

nin

g (

%)

0

5

10

15

20

0

2

4

6

8

10

12(a)

(b)

0

50

100

150

200

0

20pH10

pH7

pH3

40

60

80

100

120

FIGURE 14.3 Phospholipids profiles for two human plasma lots after LLE at different conditions:(a) normal human plasma and (b) human plasma with high fat content. Numbers under the shadow areaare multiplied by a factor of 10 for relative phospholipids response. For solvent combination, 100H0Erepresents hexane/ethyl acetate, 100/0, v/v; PPT represents supernatant of protein precipitation (PPT)plasma sample. Please note that PPT sample was prepared without pH adjusted (reprinted withpermission from Liu et al., 2009; copyright C© 2009 American Chemical Society).

(e.g., serum and plasma). Once generated, matrix backgroundinformation after LLE obtained with one compound can beapplied to guide method development for another. Extractedphospholipids from plasma/serum are a major concern formatrix suppression in ESI-MS (Wu et al., 2008b; Lahaieet al., 2010). Establishing a knowledge database for theextracted phospholipids under different extraction conditionshas proved very useful in our laboratory. As shown in Figure14.3 (Liu et al., 2009), the amount of extracted phospho-lipids differs significantly under different extraction condi-tions. Different combinations of hexane and ethyl acetatewill also have a dramatic effect on the recovery of a spe-cific analyte, as shown in Figure 14.4. A combination ofthe recovery and selectivity (matrix background) can guidethe optimization of sample extraction. Taking advantage ofthe established data set of matrix background (phospholipids)removal, the method development process for a new com-pound using similar LLE can be greatly shortened (Liu et al.,2009). Other advantages for LLE include low cost, straight-forward and repeatable process, ease of method transfer, andrelative short method development time.

LLE is often seen in regulated bioanalytical assays fordrug candidates in late stage of development. At that stage,there are usually a large number of samples to be analyzed,

and automation of the sample preparation is usually desired toincrease the overall sample throughput. One common mis-conception is that LLE is relatively difficult to automate.However, with the recent developments in robotic equipmentand extraction plates (Dotsikas et al., 2006; Hussain et al.,2009; Tweed et al., 2010), most of the LLE steps can be auto-mated as indicated in the protocol, which allows for very goodthroughput even though full automation is still difficult. LLEis not suitable for hydrophilic compounds, and developmentcan be challenging when trying to extract multiple analyteswith different lipophilicity.

14.3.3.3 Other Forms of LLE Other than the conven-tional LLE, salting-out assisted LLE (SALLE) and solidsupported LLE (SLE) are the two major variations that havebeen used by bioanalytical scientists. SALLE is based onthe “salt-induced phase separation” phenomenon to enhancethe LLE efficiency. It extracts the analyte of interest usingwater-miscible organic solvent (such as acetonitrile) and thenuses high concentration of salts (magnesium sulfate) (Zhanget al., 2009), potassium carbonate (Rustum, 1989), sodiumchloride (Yoshida et al., 2004), or ammonium acetate (Wuet al., 2008a) to induce phase separation of the water misci-ble organic phase. By using SALLE, the extraction solvent

Page 9: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

KNOW YOUR TOOLS 173

0

0.5

1.0

pH 10

pH 7

pH 3

1.5R

elat

ive

pea

k ar

ea (

A/IS

)

LLE Extraction solvent combination

2.0

A-100

H

90H10

E

80H20

E

70H30

E

60H40

E

50H50

E

40H60

E

30H70

E

20H80

E

10H90

E

0H10

0E

FIGURE 14.4 Recovery profiles for ketoconazole under different pH and extraction solvent com-binations. The relative responses are normalized to IS response (example for solvent combination,100H0E: hexane/ethyl acetate, 100/0, v/v) (reprinted with permission from Liu et al., 2009; copyrightC© 2009 American Chemical Society).

for LLE is no longer limited to water-immiscible organicsolvent. Compared to conventional LLE, the advantagesfor SALLE include broader application (suitable for ana-lytes from low to high lipophilicity), better analyte recov-ery and compatibility with RP and HILIC chromatography.Conversely, extracts typically contain more endogenouscompounds and are associated with higher matrix effect. Sup-ported liquid extraction (SLE) is a high throughput techniqueperformed in a 96-well format that is analogous to tradi-tional LLE. It uses an inert solid support with high surfacearea to improve the interface between the aqueous samplesand the water-immiscible organic solvent. Briefly, biologi-cal samples are mixed with aqueous buffer and then loadedonto the solid support. Subsequently, water-immiscible sol-vent is added, passing through the solid support to extractthe analyte of interest. The extraction mechanism involvedis mainly a partition of the analyte(s) between the organicsolvent and the absorbed aqueous phase on the solid sup-port. The advantages of SLE include no emulsion formation,easy automation, and high extraction efficiency (Jiang et al.,2009, 2012).

14.3.4 Solid Phase Extraction

SPE has been used for several decades for extracting andconcentrating trace amounts of analytes from biological andenvironmental samples (Lee and Esnaud, 1988; Peng et al.,2000; Mornar et al., 2012). The SPE process involves load-ing biological samples to an SPE cartridge/plate/column thatis packed with solid sorbents. The analyte(s) of interest isretained by interacting with the packing material through dif-ferent interaction mechanisms; the interfering matrix compo-nents are either directly passing through the sorbents during

the loading step; being washed away during the washing stepor retained on the stationary phase after elution, analytes ofinterest are retained on the stationary phase, then eluted fromthe sorbents using a suitable elution solvent and collected forLC-MS analysis. SPE is technically a form of low-pressurechromatography. Like for LC columns, different modes ofSPE and a great variety of commercial phases are availabletoday. When developing an SPE-based bioanalytical method,parameters needed to be determined include the type andamount of sorbent, the sample volume that can be appliedwithout loss of recovery, the optimum loading, washing andelution conditions (time, volume, composition). SPE can becarried out either off-line (Peng et al., 2000) or online withthe chromatographic system (Chen et al., 2009). Since eachcartridge or each well in a multiple-well plate is essentiallyan individual chromatographic column, the consistency ofthe SPE cartridge or each well in a plate, as well as batch-to-batch variability in the SPE cartridge or plate, will haveimpact on the performance of an SPE method.

Based on the packing materials used, conventional SPEcan be divided into three major categories. The first typeis reversed-phase (RP)-SPE, which uses a nonpolar sta-tionary phase such as the alkyl- or aryl-bonded silica;the second is normal-phase SPE using a polar stationaryphase such as silica with polar functional groups (Si-CN,Si-NH2, Si-Diol, and pure silica); the third type is ion-exchange using ionic functional groups (strong or weakorganic acids and bases bonded to the supporting base).Table 14.1 lists the major types of SPE, their mechanismsof retention and elution, and their applications. An SPEcartridge or plate usually comes with standard proceduresfor operation from the supplier. These conditions can be avery good starting point for method development. However,

Page 10: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

174 BEST PRACTICES IN BIOLOGICAL SAMPLE PREPARATION FOR LC-MS BIOANALYSIS

TABLE 14.1 Commonly Used SPE Types and Their Mechanism of Retention and Applications

RP-SPE NP-SPE Ion-exchange

Solid PhaseFunctionalGroups

Alkyl- or aryl-bonded silica(Si-C4, Si-C8, Si-C18,Si-Ph, etc.)

Si-CN, Si-NH2, Si-Diol andpure Silica, etc.

Quaternary amine bonded silica (StrongAnion Exchange), sulfonic acid bondedsilica (Strong Cation Exchange),carboxylic acid bonded silica (WeakCation Exchange), Neutral amine (WeakAnion Exchange)

RetentionMechanism

Nonpolar–nonpolarinteractions, van der waalsand dispersion forces

Polar–polar interactions;Hydrogen bonding;dipole–dipole interactions;dipole-induced dipoleinteractions

Electrostatic attraction of the oppositelycharged functional groups between theanalyte and the sorbent

Loading Samples prepared in and SPEplate conditioned withaqueous buffer with pHadjustment that analyte willnot be charged

Samples prepared in and SPEplate conditioned withnonpolar solvent (e.g.,hexane, dichloromethane,etc.)

An aqueous or organic, low salt solutionwith a pH at which the analytes and thefunctional groups on the stationary phasewill be oppositely charged. e.g., low pHfor basic compounds and strong cationexchange

Elution Polar organic solvent such asmethanol, acetonitrile, etc.with pH adjustment to theopposite of load conditions

Polar organic solvent such asmethanol, acetonitrile,acetone, isopropanol

Disrupt the electrostatic interactions by pH,ionic strength and solvent modificationsto neutralize groups of opposite chargeoruse a more selective counter-ion tocompete for ion-exchange interactionsites.

Applications Nonpolar to moderate polarcompounds: e.g., organiccompound with alkyl,aromatic, alicyclic groups

Polar compounds: such assmall organic compoundswith hydroxyl groups,carbonyls, hetero atoms

Cation-exchange: basic compounds such asprimary, secondary, tertiary andquaternary amines, etc.Anion-exchange:acidic acids such as carboxylic acids,sulphonic acids, etc.

since SPE is like any other chromatography, the optimumconditions for a specific bioanalytical method are compoundspecific. Fine tuning of experimental conditions for eachcompound is usually necessary to achieve the best results.SPE plates/cartridges are available from all the major suppli-ers, such as Oasis R© series from Waters (www.waters.com,accessed Mar 19, 2013), HyperSep series from Ther-moFisher Scientific (www.thermofisher.com, accessed Mar19, 2013), Bond Elut series from Agilent (www.agilent.com,accessed Mar 19, 2013), EVOLUTE R© series from Bio-tage (www.biotage.com, accessed Mar 19, 2013), Strata R©

series Phenomenex (www.phenomenex.com, accessed Mar19, 2013), Discovery R© and EmporeTM series from Sigma-Aldrich (www.sigma-aldrich.com, accessed Mar 19, 2013).It is also worth mentioning that mixed mode (a combina-tion of multiple retention mechanisms) SPE products are alsoavailable from almost all major vendors: Oasis R© MCX/MAXfrom Waters, ISOLUTE R© HCX/HAX from Biotage, etc. Themost commonly used mixed-mode SPE is a combination ofreversed-phase and ion-exchange. It utilizes dual retentionmechanisms of hydrophobic and electrostatic interactions toretain basic, acidic, neutral, and zwitterionic compounds.Due to its broad applicability to a wide range of compounds,

mixed-mode SPE is widely used (Shou et al., 2001; Mulleret al., 2002; Jenkins et al., 2004; Ge et al., 2005; Xu et al.,2005; Yue et al., 2005; Xue et al., 2006a), however, providesless sample cleanup than other more selective SPE phases.

SPE products are available in different formats, for exam-ple, single cartridge, multiple-well plate (96-well, 384-well)or online SPE column. In the earlier days, SPE was donemanually in single cartridge mode (Khan et al., 1999; vander Heeft et al., 2000; Cavaliere et al., 2003). However, withthe demand for high throughput in pharmaceutical indus-try, 96-well format with automation is rapidly becoming thedominant platform in bioanalysis (Kaye et al., 1996; Shouet al., 2001; Shou et al., 2002; Mallet et al., 2003; Xu et al.,2007; Helle et al., 2011). No matter which format is chosen,the same principles apply. Below is a general protocol basedon the 96-well format.

14.3.4.1 A Typical Sample Preparation Protocol UsingWaters HLB 96-Well SPE Plate

� Pipette 50 μl of each of the blank, STDs, and QC sam-ples to individual wells in a 96-well plate.� Same as PPT and LLE.

Page 11: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

KNOW YOUR TOOLS 175

� Add 50 μl of IS working solution to each sample.� The IS working solution should be prepared in aque-

ous solution. After mixing the IS working solutionwith the biological samples, some time (e.g., 10 min)should be given to allow the IS to interact withthe biological matrix to closely mimic the analyte’sbinding.

� Add 400 μl of 10 mM ammonium acetate in water toeach sample.� This step is to buffer each sample to a given pH while

maintaining a suitable solvent strength for the analyteto bind on the packing material. Protein binding con-cerns should be addressed during this step. Modifierscan be added to the buffer to disrupt the drug–proteinbinding so that the drug molecules are available forbinding to the packing material during the loadingstep.

� Condition the 96-well SPE plate with 450 μl ofmethanol, followed by 450 μl of 10 mM ammoniumacetate in water.� This step is to clean and precondition the SPE plate

for sample loading. Vacuum is usually involved forefficiency and consistency.

� Transfer the sample mixtures to the preconditioned SPEplate.� This step is to load the samples onto the 96-well

plate. Vacuum is normally needed during this step.The majority of the proteins, other macromolecules,and salts will pass through the plate to the waste. It iscritical to make sure all the wells are free of previoussolution before loading the samples.

� Wash SPE plate with 450 μl of water.� This step is to further remove residual proteins, salts,

and highly polar compounds. Sometimes, the wash-ing solution may also contain a small percentage oforganic solvent to remove some less polar compounds.The desired balance of recovery and selectivity is nor-mally decided during this step. A stronger solventwash results in a cleaner extract but lower recovery.Vacuum is needed for this step.

� Elute SPE plate twice with 250 μl of methanol.� This step is to release the bound analyte of interest

from the SPE packing material with strong organicsolvent. Vacuum is needed for this step. It is alsocritical to make sure all the wells are free of previ-ous solution before the elution solvent is added. Alsothe distance between the 96-well SPE plate and thecollection plate has to be carefully adjusted to avoidcross-contamination. A double elution is not neces-sary in all cases but will results in better recovery.

� Evaporate the sample under nitrogen flow to completedryness and reconstitute the sample with 200 μl ofreconstitution solution.

� This step is to change the solvent system for the sam-ples to make it suitable for LC injection. Sample con-centration can also be achieved during this step byreconstituting the sample in a smaller volume. Thisstep can be omitted if the concentration and solventare compatible with the LC-MS assay.

All liquid transfer steps in the above procedures can beroutinely done using robotic liquid handler. The whole pro-cess can be easily automated using commercially availableworkstations. However, well-to-well and plate-to-plate dif-ferences always exist in 96-well SPE plates. It is not uncom-mon to observe solutions passing through one well at adifferent speed than another. Therefore, it is critical to makesure all the wells in the 96-well plate are free of the previoussolution before applying the next solution. Unfortunately,this process normally requires visual checks by the operator,which makes walk-away automation less practical.

14.3.4.2 Advantages and Limitations The main advan-tages for SPE method include high-to-moderate selectivity,versatility, and suitability for full automation. Due to theavailability of different types of packing material, SPE issuitable for a wide variety of compounds ranging from polarto nonpolar, acidic to basic, low to high molecular weight,and for various sample matrices. When optimized, it canachieve both high recovery and good selectivity of an analytewith very good reproducibility. Limitations of this techniqueinclude high cost, a dependence on the quality of the supplies(e.g., lot-to-lot variability of SPE plate/cartridge) and longmethod development time. SPE also involves many steps andrequires high operational skills and good knowledge of sep-aration science to fully actualize the advantages of SPE. Agood understanding of the chemistry of the analyte and thepacking material and of the retention mechanism is essentialto design a good SPE method. An unoptimized SPE methodis no better than a PPT method.

14.3.4.3 Other Forms of SPE The SPE format workswell for other separation mechanism including immunoaffin-ity SPE (Delaunay-Bertoncini and Hennion, 2004), molecu-larly imprinted polymer (MIP) SPE (Lasakova and Jandera,2009), and restricted access material (RAM) SPE (Souverainet al., 2004). Immunoaffinity and MIP SPE are discussed insome details in Sections 14.3.5.1 and 14.3.5.2. The RAMterm was first introduced in 1991 by Desilet et al. (1991).Instead of the solid support for regular sorbents, RAM sor-bents are a class of support materials with pores only acces-sible to small molecules. Their outer surface is coated withhydrophilic functional groups, and the pore inner surface canbe coated with different functional groups, such as hydropho-bic (Amini and Crescenzi, 2003) or ion exchange (Chiapet al., 2002) groups. RAM SPE takes the advantages of bothregular SPE and size exclusion chromatography. Biological

Page 12: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

176 BEST PRACTICES IN BIOLOGICAL SAMPLE PREPARATION FOR LC-MS BIOANALYSIS

proteins can be easily removed from the matrix and the smallorganic analytes can then access the pores, and interact withthe functional groups. RAM SPE is most often used in onlineSPE mode as it allows for direct injection of untreated plasma.A comprehensive review is available for online RAM SPEsample preparation (Souverain et al., 2004).

Operationally, other than the commonly used car-tridge/plate formats of SPE, there are also many other typesof SPE formats for specialized applications, such as onlineSPE, dispersive SPE (dSPE, also referred as QuEChERS,which stands for Quick, Easy, Cheap, Effective, Rugged,and Safe), disposable pipette extraction (DPX), microex-traction SPE techniques (microextraction by packed sorbent(MEPS), solid phase microextraction (SPME), and stir barsorptive extraction (SBSE)). Online SPE is very similar totwo-dimensional chromatography, which couples an SPEcartridge to an LC column with a diverting valve in between.Two different sets of pumps are needed to deliver solvents tothe SPE cartridge and the LC column. A detailed descriptionof the hardware configuration can be found in a recent review(Chang et al., 2007). Online SPE has gained popularity forits operational advantages, in particular the minimum samplepretreatment needed before prior to LC-MS analysis. Oncethe system is set up and a method is developed, the produc-tion stage can be done with limited supervision. The progress,advantages, and limitations of online SPE for LC-MS bio-analysis have been extensively covered in several reviews(Xu et al., 2007; Chen et al., 2009; Novakova and Vlckova,2009; Kole et al., 2011). In QuEChERS, the analyte of inter-est is first extracted from the biological matrix using organicsolvent, and then sorbents are added to remove matrix com-ponents from the organic phase. It has been widely used formeasuring pesticides and drug residues in food and environ-mental samples (Anastassiades et al., 2003; Lehotay et al.,2005; Posyniak et al., 2005). In recent years, applicationswere also found in the analysis of drug/metabolites in bio-logical samples such as tissue (Fagerquist et al., 2005), milk(Whelan et al., 2010), and whole blood (Plossl et al., 2006).DPX is a variant of traditional SPE. The sorbent is looselypacked into a standard pipette tip in between two frits. Sam-ples are loaded by aspirating through the tip from the bot-tom. The major advantage of DPX is its simplicity. DPX tipsare commercially available with a variety of sorbents. Onetrend for sample preparation is miniaturization. New devel-opments for SPE include several forms of microextractiontechniques, such as SPME, MEPS, and SBSE. A commonfeature for these techniques is that a small amount of sor-bents are used for the extraction, which allows for elutingthe adsorbed analytes in a small volume, and injecting theentire volume of extract directly in the LC-MS system foranalysis. The detailed description of these techniques, theiradvantages, and limitations are covered in several compre-hensive reviews (Novakova and Vlckova, 2009; Ashri andAbdel-Rehim, 2011; Kole et al., 2011); a focused review

(Vuckovic et al., 2010) for SPME is also available. No mat-ter which operating mode is used, the principles are the same:the analyte of interest is selectively/semiselectively retainedon the solid sorbent, the interfering compounds are washedaway and then the analyte of interest is released and ana-lyzed. Although there is increased interest in adopting thesemethods in bioanalysis, the traditional SPE in 96-well plateformat is still the mainstream in bioanalysis.

14.3.5 Analyte Specific Extraction Techniques

The sample preparation techniques discussed earlier arelisted in order of increasing selectivity, from the simpledilution to regular SPE methods but none of them is trulycompound specific. Even though they can meet bioanalyt-ical needs for the most part, some special cases do existthat require a more selective technique to achieve super highsensitivity or remove stubborn interferences. Matrix interfer-ences that are most likely to interfere with the LC-MS analy-sis would have similar physicochemical properties to those ofthe analyte of interest. They are very likely to be coextractedwith the analyte using the above sample extraction methods.Orthogonal or hybrid sample preparation techniques suchas LLE/SPE, PPT/SPE, and PPT/LLE have sometimes beenused to mitigate this issue. Compound specific extraction,such as immunoaffinity solid phase extraction (IA-SPE), SPEwith MIP packing material (MIP-SPE), is a more effectiveapproach to deal with this problem. In this section, we dis-cuss them in detail in regard to their advantages, applicationsand limitations.

14.3.5.1 Immunoaffinity Extraction An immunosorbentis a selective extracting material that is prepared by immobi-lizing an analyte specific antibody (either polyclonal or mon-oclonal) on a solid support. When exposing the immunosor-bent to a sample containing the analyte (antigen) of interest,a complex can form between the analyte and the immobilizedantibody by selective and reversible antigen–antibody inter-action. Extracting an analyte of interest using immunosor-bents is called immunoaffinity extraction. It has been usedfor decades (Farjam et al., 1988; Haasnoot et al., 1989;Medina-Casanellas et al., 2012) to extract and enrich an ana-lyte of interest specifically from a complex matrix. UsingSPE plate/cartridge packed with immunosorbents for samplepreparation is termed as IA-SPE, which can be done eitheronline (Farjam et al., 1988) or off-line (Aranda-Rodriguezet al., 2003). Applications of IA-SPE in pharmaceuticaland biomedical research have been reviewed (Delaunay-Bertoncini and Hennion, 2004).

Procedures for Immunoextraction SPE. The immunoex-traction process can be divided into four steps: condi-tioning, percolation of the sample, washing, and elution.Detailed description of each step can be found in the review

Page 13: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

KNOW YOUR TOOLS 177

(Delaunay-Bertoncini and Hennion, 2004) of Delaunay-Bertonicini. Briefly, the immunoextraction procedure is asfollows:

Conditioning: the conditioning step is to establish an envi-ronment that favors of the interaction between the ana-lyte and the antibody. The immunosorbents are normallystored in a PBS buffer containing a small percentage ofazide to preserve the antibody activity. This buffer hasto be replaced by a solution in which the specific inter-action between the analyte of interest and the antibodycan occur.

Loading: the biological sample after pretreatment is loadedor incubated with the immunosorbents. During this step,the capacity of the immunosorbents should be taken intoaccount to avoid overloading the immunosorbents. Thecapacity of an immunosorbent is defined as the maxi-mal amount of analyte that can be bound onto the sor-bent based on the total number of accessible antibodiesimmobilized on the solid support.

Washing: This step serves to remove the interfering com-pounds that were absorbed on the sorbents throughnonspecific interactions, such as hydrophobic and ionicinteractions. Similarly to regular SPE procedures, thewashing solution should be optimized to effectivelyremove the interferences but without disrupting theantigen–antibody interaction.

Elution: Any solution which can effectively disrupt theinteraction between the analyte and the antibody can beused to elute the analyte and the IS, such as displaceragents, chaotropic agents, pH variations, or organic sol-vent; organic solvents are the most commonly usedelution reagent.

The main advantage for immunoaffinity extraction is itsselectivity. The immobilized antibodies on the support mate-rial are raised by immunization against a specific antigen (theanalyte). Therefore, the interactions between the antibodyand the antigen (analyte) are very specific. Even though anti-bodies can be raised against small molecules, it is much easierto raise antibodies against proteins. Immunoaffinity extrac-tion is a powerful and efficient tool in bioanalysis. The off-line (Berna et al., 2007) or online (Dufield and Radabaugh,2012) coupling of immunoextraction with a chromatographicseparation and mass spectrometric detection allows for thedetection and quantification of drugs with a high sensitiv-ity, selectivity, and reproducibility. However, compared withother sample preparation methods, such as PPT, LLE, andregular SPE, tremendous efforts have to be engaged for thepreparation of immunosorbents. Producing an antibody spe-cific to the analyte of interest and then immobilizing it on asolid support is neither a trivial nor a cheap process. There-fore, immunoextraction is often a solution of last resort for

bioanalytical scientists. With the increasing interest in quan-tifying protein drug candidates in biological samples usingLC-MS (Li et al., 2011), immunoaffinity purification of thetarget protein from the biological matrix (Dubois et al., 2007)or the surrogate peptides after digestion (Neubert et al., 2010)may be a necessary approach to achieve a sensitivity compa-rable to that of ligand binding assays, which remain the goldstandard for protein drug measurements. When technologyevolves to allow easy antibody generation, simple and cheapimmunosorbent production for a specific analyte, or in thecase when a specific analyte is widely and routinely mea-sured, for example, biomarkers in clinical diagnostics, usingstandardized immunoaffinity extraction procedures (kit) forbiological sample cleanup can be easily justified and alsowill be cost-effective.

14.3.5.2 MIP-SPE

Molecularly Imprinted Polymers. An MIP is a highlycrosslinked polymer that is polymerized in the presence of atemplate molecule. The template molecule is removed afterthe polymerization, leaving complementary cavities that canthen be used for molecular recognition. The cavities in theMIP specifically bind the template molecules or sometimesother molecules with very similar shape and functionality.This is very similar to the interaction between an antigenand its antibody in terms of specificity. Therefore, MIP isalso referred to as a “synthetic antibody” by analogy withprotein antibodies. The molecular recognition property canthen be used for analyte-specific extraction in complex bio-logical matrices. The use of MIPs as packing materials forSPE sample extraction dates back to 1994 (Sellergren, 1994).Since then, numerous applications have been reported usingMIP-SPE for sample preparation in various analytical assays,including environmental sample analysis (Guan et al., 2012),food contaminants analysis (Baggiani et al., 2007; Piletskaet al., 2012), and drug bioanalysis (Mullett and Lai, 1999;Yang et al., 2006; Mirmahdieh et al., 2011). Several reviewsare available on the progress of using MIP-SPE for samplepreparation in bioanalysis (Lasakova and Jandera, 2009; TseSum Bui and Haupt, 2010).

MIPs offer many advantages over protein antibodies.Proteins are difficult and expensive to purify, easily dena-tured (pH, heat, proteolysis), difficult to immobilize, andimpractical to reuse. Antibodies to small molecules aredifficult to produce as small molecules are not immunogenicby themselves. In contrast, MIPs are easy to synthesize,resistant to harsh environment (elevated temperature andpressure, extreme pHs, organic solvents), have high sampleload capacity, and can be used over a long period oftime. Similarly to immunoaffinity SPE, MIP-SPE can veryselectively extract the target analyte from complex samples.Therefore, MIP-SPE is a practical way to concentrate theanalyte of interest from a large-volume sample without

Page 14: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

178 BEST PRACTICES IN BIOLOGICAL SAMPLE PREPARATION FOR LC-MS BIOANALYSIS

retaining significant amounts of matrix interferences. Thisis especially helpful for an assay that needs ultrahighsensitivity to measure trace amount of compounds in a verycomplex system. Limitations do exist for this technology,such as bleeding of the template molecule or heterogene-ity of binding sites (nonspecific interactions). However,approaches were reported to address these concerns, makingMIP-SPE a powerful analytical tool. Like any other SPEmethods, MIP-SPE can be done either online or off-line.

Technically, MIPs can be prepared based on any typeof template molecule. Although many reported researcheson MISPE were carried using homemade MIP materials,MIP-SPE cartridges for selective compounds or a class ofcompounds are commercially available. Some vendors alsoprepare customized MIPS.

In summary, both IA-SPE and MIP-SPE use packingmaterial specifically interacting with the analyte of interest.They offer superior selectivity, as well as the best ability toenrich the analyte of interest over other sample preparationmethods. These advantages do come with a cost. In general,the initial cost for both IA-SPE and MIP-SPE is high andtremendous efforts have to be engaged to get the method upand running. Therefore, they usually tend to be a last resort inmost bioanalytical laboratories. However, there is increasedinterest in adopting these technologies in bioanalytical appli-cations in recent years. On one hand, with the combinationof these compound-specific sample extraction techniques andavailability of high sensitive instruments, many applicationspreviously out of reach of LC-MS, such as trace analysis ofbiomarkers, become possible. On the other hand, althoughthe initial cost for customized MIP or immunosorbents maybe high, the overall cost may be still comparable with othertechniques when a large number of samples are to be assayedusing these techniques. The initial efforts and cost for cus-tomized MIPs or immunosorbents can be paid off by theefficiency of sample analysis and high quality of data.

14.3.6 Sample Preparation Techniquesfor Special Samples

Plasma, serum, and urine samples are the most common sam-ples in the daily work of bioanalytical scientists. However,LC-MS bioanalysis is not limited to these ordinary matrices.LC-MS bioanalytical laboratories can be asked to measuredrug/metabolites/biomarkers concentration in all kinds ofsamples, such as tissues, saliva, tears, bile, and most recentlydried blood spots (DBS) samples. New classes of analytessuch as proteins, peptides, oligonucleotides are now com-monly analyzed by LC-MS. In this section, we briefly dis-cuss sample preparation for DBS, tissue samples and proteintherapeutics in plasma/serum.

14.3.6.1 Dried Blood Spots DBS, a sample collectiontechnique that has been used for decades in clinical diag-

nostics, has recently emerged as a choice of microsamplingtechnique in the pharmaceutical industry (Li and Tse, 2010).As a new sample format, sample processing of DBS sam-ple has its own challenges and is attracting a lot of inter-est from bioanalytical scientists. Several sample preparationapproaches have been reported (Deglon et al., 2009; Liu et al.,2010b; Abu-Rabie and Spooner, 2011). A two-step strategywas reported (Liu et al., 2010b) to best accommodate DBSsamples with the approaches used for regular liquid samples.The DBS samples were first soaked with aqueous buffer tomake the analyte in the blood samples more accessible tosubsequent extraction. The reconstituted liquid samples canbe processed in a similar way as other liquid samples usingPPT, LLE, and SPE. A key factor affecting DBS samplepreparation is how to efficiently and completely recover theanalyte from the DBS card into the liquid phase. The term“elution efficiency” has been proposed to represent the ana-lyte recovery from the solid DBS sample. Since this processusually is not tracked by an IS, high and consistent recoveryof the analyte from the DBS sample to the liquid solutionis desired. Different direct elution methods have also beenreported (Deglon et al., 2009; Thomas et al., 2010; Abu-Rabieand Spooner, 2011) to elute the analyte of interest from theDBS card directly into the detection system.

14.3.6.2 Tissues and Organs Pharmaceutical drug can-didates are often measured in tissue samples to better under-stand their distribution or for pharmacodynamic evaluation.Compared with liquid samples, sample preparation for tissuesamples is more challenging. The solid tissue samples needto be converted into liquid-like samples often by homog-enization (Liang et al., 2011). Organic solvent precipita-tion of the tissue homogenate (Gurav et al., 2012) can bedone to generate a supernatant suitable for LC-MS analysis.The supernatant can also be further processed using LLE orSPE for a cleaner extract. Diluting tissue homogenate withplasma, and then processing the diluted tissue homogenateas plasma sample and quantifying it with plasma STDs wasalso reported (Jiang et al., 2011). A chapter devoted to tissuesample analysis using LC-MS can be found in this book.

14.3.6.3 Proteins Using LC-MS to quantitatively mon-itor biologics (proteins) has drawn more and more attentionin recent years. Many assays have been published that useLC-MS to quantify proteins by monitoring one or severalsurrogate peptides after enzymatic digestion (Yang et al.,2007; Heudi et al., 2008; Ewles and Goodwin, 2011; Liet al., 2011; Mesmin et al., 2011; Wu et al., 2011). Unlikesmall molecules, protein drugs usually have to be digestedinto small peptides for sensitive quantitative analysis usingLC-MS. Measuring proteins in plasma or serum usingLC-MS usually involves extraction of protein drugs beforedigestion, such as immunoaffinity purification of targetprotein, or extraction of surrogate peptides after digestion.

Page 15: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

CONCLUSION AND PERSPECTIVES 179

The analysis of crude digested plasma or serum sampleswas also reported directly (Ouyang et al., 2012). No matterwhich sample preparation method is used, digestion is acritical step for protein bioanalysis using LC-MS. Manyinnovative approaches have been reported to quickly,effectively, and consistently perform the digestion step, suchas microwave-assisted digestion (Vaezzadeh et al., 2010),organic solvent digestion (Strader et al., 2006), and pelletdigestion (Ouyang et al., 2012). A digested whole plasmaor serum sample is a very complex sample containinghundreds of peptides. Even though it had been demonstratedthat it is feasible to quantify proteins by direct LC-MSanalysis of the crude digested plasma samples, it is stronglyrecommended to simplify the sample composition usingavailable techniques, such as immunoaffinity purification,MIP-SPE, or other SPE techniques.

14.4 KNOW YOUR NEEDS

As a bioanalytical scientist, after examining the basic infor-mation available regarding the samples and the analyte andchecking all the tools in our toolbox, the next thing is todecide what approach to use on a given project. In mostcases, scientists will turn to the technique they are mostfamiliar with or is most readily available. Sample prepara-tion is a matter of personal preference but it is good practiceto brainstorm all the options and eliminate some obviouslybad choices before heading to the laboratory for experi-ments. Before starting a method, two questions should beasked: “what can we do?” and “what do we need to do?” Agood bioanalytical strategy should derive from a comprehen-sive evaluation of the properties of both the analyte and thematrix, available options for sample preparation, the projectneeds and the cost. As illustrated in Figure 14.5, the final bio-analytical strategy should be the right balance between thetechnical capabilities and the cost (time, money and otherresources) based on the project needs. For example, an assaythat does not require high sensitivity may only need a sim-ple “dilution” approach; a one-time-use method for a smallstudy can settle with PPT; an assay which is expected to beused for many studies with a large number of samples shouldbe very rugged and the initial cost of the assay may notbe a concern as long as the final cost/sample is acceptable;assays for bioequivalence studies should be of the highestprecision. These approaches are frequently discussed as “fit-for-purpose” or “risk-based.” The best bioanalytical strategyis the most economical solution that fits the project needs.

14.5 CONCLUSION AND PERSPECTIVES

With the advancements in technology and understanding ofmatrix effect, sample preparation for small molecule drug

Sample(analyte +matrix)

Availabletools

Technicalcapabilities

(sensitivity, assayruggedness, etc.)

Cost (time, money,other resources)

Projectneeds (intended

use of the data, requiredassay sensitivity, expected

sample number, etc.)

FIGURE 14.5 Illustration of bioanalytical strategy selection.

candidates in ordinary matrices (e.g., plasma, urine, serum,and blood) has become straightforward in most cases. Withthe availability of increasingly more sensitive mass spec-trometers, it is now possible to adopt a simple sample dilu-tion strategy, in many applications, in which only a very smallamount of matrix (e.g., less than 1 μl of plasma per injection)will need to be injected into the LC-MS system to achievethe required sensitivity. Also if a SIL-IS is available, matrixeffects can usually be well compensated for in a bioanalyti-cal method, which makes dilution or simple PPT even moreattractive. Ideally, a sample preparation method is preferredwith minimum change of the original samples to preservetheir information as much as possible.

With the expanding scope of LC-MS bioanalysis, in par-ticular in the area of biomarkers and biologics, the demandfor novel and better sample preparation approaches is contin-uing. For routine sample analysis, greater emphasis has beenplaced on truly walk-away automation. The sample prepa-ration process could be fully automated eventually. Anothertrend for sample preparation is the handling of small sam-ple volumes. Microsampling technology is becoming moreand more popular now that the instrument detection limitsallow bioanalysis of the equivalent of 1-μl or 2-μl samplesper analysis. The currently popular sample handling and pro-cessing techniques in routine bioanalysis seem to be fallingbehind the pace of more advanced detection techniques. Intraditional sample preparation, the majority of the samplecollected is not used for the final analysis but wasted. Weexpect that better usage of precious biological samples willbe the trend and sample preparation techniques, which caneffectively deal with 1-μl or 2-μl samples, will become pop-ular in bioanalysis.

Finally, as bioanalysis-related technologies evolve, cus-tomized bioanalysis kits for a specific analyte, using

Page 16: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

180 BEST PRACTICES IN BIOLOGICAL SAMPLE PREPARATION FOR LC-MS BIOANALYSIS

technologies such as IA-SPE, MIP-SPE, with “standardizedsample preparation procedures,” can become a practical andcost-effective solution for the analysis of common drugs.

REFERENCES

Abu-Rabie P, Spooner N. Dried matrix spot direct analysis: eval-uating the robustness of a direct elution technique for use inquantitative bioanalysis. Bioanalysis 2011;3:2769–2781.

Amini N, Crescenzi C. Feasibility of an on-line restrictedaccess material/liquid chromatography/tandem mass spectrom-etry method in the rapid and sensitive determination oforganophosphorus triesters in human blood plasma. J Chro-matogr B Analyt Technol Biomed Life Sci 2003;795:245–256.

Anastassiades M, Lehotay SJ, Stajnbaher D, Schenck FJ. Fastand easy multiresidue method employing acetonitrile extrac-tion/partitioning and “dispersive solid-phase extraction” for thedetermination of pesticide residues in produce. J AOAC Int2003;86:412–431.

Aranda-Rodriguez R, Kubwabo C, Benoit FM. Extraction of 15microcystins and nodularin using immunoaffinity columns. Tox-icon 2003;42:587–599.

Ashri NY, Abdel-Rehim M. Sample treatment based on extractiontechniques in biological matrices. Bioanalysis 2011;3:2003–2018.

Aubry AF. LC-MS/MS bioanalytical challenge: ultra-high sensitiv-ity assays. Bioanalysis 2011;3:1819–1825.

Baggiani C, Anfossi L, Giovannoli C. Solid phase extraction of foodcontaminants using molecular imprinted polymers. Anal ChimActa 2007;591:29–39.

Berna MJ, Zhen Y, Watson DE, Hale JE, Ackermann BL. Strategicuse of immunoprecipitation and LC/MS/MS for trace-level pro-tein quantification: myosin light chain 1, a biomarker of cardiacnecrosis. Anal Chem 2007;79:4199–4205.

Bishop MJ, Crow BS, Kovalcik KD, George J, Bralley JA. Quan-tification of urinary zwitterionic organic acids using weak-anionexchange chromatography with tandem MS detection. J Chro-matogr B Analyt Technol Biomed Life Sci 2007;848:303–310.

Brockman AH, Hatsis P, Paton M, Wu JT. Impact of differentialrecovery in bioanalysis: the example of bortezomib in wholeblood. Anal Chem 2007;79:1599–1603.

Cai X, Zhong B, Su B, Xu S, Guo B. Development and validationof a rapid LC-MS/MS method for the determination of JCC76,a novel antitumor agent for breast cancer, in rat plasma and itsapplication to a pharmacokinetics study. Biomed Chromatogr2012;26:1118–1124.

Casetta B, Romanello M, Moro L. A rapid and simple method forquantitation of urinary hydroxylysyl glycosides, indicators ofcollagen turnover, using liquid chromatography/tandem massspectrometry. Rapid Commun Mass Spectrom 2000;14:2238–2241.

Cavaliere C, Curini R, Di Corcia A, Nazzari M, Samperi R. Asimple and sensitive liquid chromatography-mass spectrometryconfirmatory method for analyzing sulfonamide antibacterialsin milk and egg. J Agric Food Chem 2003;51:558–566.

Chang M, Ji Q, Zhang J, El-Shourbagy T. Historical review ofsample preparation for chromatographic bioanalysis: pros andcons. Drug Development Research 2007;68:107–133.

Chen L, Wang H, Zeng Q, et al. On-line coupling of solid-phaseextraction to liquid chromatography–a review. J Chromatogr Sci2009;47:614–623.

Chiap P, Rbeida O, Christiaens B, et al. Use of a novel cation-exchange restricted-access material for automated sample clean-up prior to the determination of basic drugs in plasma by liquidchromatography. J Chromatogr A 2002;975:145–155.

Chng HT, New LS, Neo AH, Goh CW, Browne ER, Chan EC. Asensitive LC/MS/MS bioanalysis assay of orally administeredlipoic acid in rat blood and brain tissue. J Pharm Biomed Anal2010;51:754–757.

Dams R, Huestis MA, Lambert WE, Murphy CM. Matrix effectin bio-analysis of illicit drugs with LC-MS/MS: influence ofionization type, sample preparation, and biofluid. J Am SocMass Spectrom 2003;14:1290–1294.

Deglon J, Thomas A, Cataldo A, Mangin P, Staub C. On-line desorp-tion of dried blood spot: a novel approach for the direct LC/MSanalysis of micro-whole blood samples. J Pharm Biomed Anal2009;49:1034–1039.

Delaunay-Bertoncini N, Hennion MC. Immunoaffinity solid-phaseextraction for pharmaceutical and biomedical trace-analysis-coupling with HPLC and CE-perspectives. J Pharm BiomedAnal 2004;34:717–736.

Desilets CP, Rounds MA, Regnier FE. Semipermeable-surfacereversed-phase media for high-performance liquid chromatog-raphy. J Chromatogr 1991;544:25–39.

Dotsikas Y, Kousoulos C, Tsatsou G, Loukas YL. Development andvalidation of a rapid 96-well format based liquid-liquid extrac-tion and liquid chromatography-tandem mass spectrometry anal-ysis method for ondansetron in human plasma. J Chromatogr BAnalyt Technol Biomed Life Sci 2006;836:79–82.

Dubois M, Becher F, Herbet A, Ezan E. Immuno-mass spectrometryassay of EPI-HNE4, a recombinant protein inhibitor of humanelastase. Rapid Commun Mass Spectrom 2007;21:352–358.

Dufield DR, Radabaugh MR. Online immunoaffinity LC/MS/MS.A general method to increase sensitivity and specificity: how doyou do it and what do you need? Methods 2012;56:236–245.

Ewles M, Goodwin L. Bioanalytical approaches to analyzing pep-tides and proteins by LC–MS/MS. Bioanalysis 2011;3:1379–1397.

Fagerquist CK, Lightfield AR, Lehotay SJ. Confirmatory and quan-titative analysis of beta-lactam antibiotics in bovine kidney tissueby dispersive solid-phase extraction and liquid chromatography-tandem mass spectrometry. Anal Chem 2005;77:1473–1482.

Farjam A, de Jong GJ, Frei RW, et al. Immunoaffinity pre-columnfor selective on-line sample pre-treatment in high-performanceliquid chromatography determination of 19-nortestosterone.J Chromatogr 1988;452:419–433.

Flaherty JM, Connolly PD, Decker ER, et al. Quantitative determi-nation of perfluorooctanoic acid in serum and plasma by liquidchromatography tandem mass spectrometry. J Chromatogr BAnalyt Technol Biomed Life Sci 2005;819:329–338.

Page 17: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

REFERENCES 181

Fu I, Woolf EJ, Matuszewski BK. Effect of the sample matrix onthe determination of indinavir in human urine by HPLC withturbo ion spray tandem mass spectrometric detection. J PharmBiomed Anal 1998;18:347–357.

Fu Y, Li W, Smith HT, Tse FL. An investigation of incurred humanurine sample reanalysis failure. Bioanalysis 2011;3:967–972.

Ge L, Yong JW, Tan SN, Yang XH, Ong ES. Analysis of posi-tional isomers of hydroxylated aromatic cytokinins by micellarelectrokinetic chromatography. Electrophoresis 2005;26:1768–1777.

Gray N, Musenga A, Cowan DA, Plumb R, Smith NW. A sim-ple high pH liquid chromatography-tandem mass spectrometrymethod for basic compounds: application to ephedrines in dop-ing control analysis. J Chromatogr A 2011;1218:2098–2105.

Gu H, Deng Y, Wang J, Aubry AF, Arnold ME. Development andvalidation of sensitive and selective LC-MS/MS methods for thedetermination of BMS-708163, a gamma-secretase inhibitor, inplasma and cerebrospinal fluid using deprotonated or formateadduct ions as precursor ions. J Chromatogr B Analyt TechnolBiomed Life Sci 2010;878:2319–2326.

Guan W, Han C, Wang X, et al. Molecularly imprinted polymersurfaces as solid-phase extraction sorbents for the extraction of2-nitrophenol and isomers from environmental water. J Sep Sci2012;35:490–497.

Gurav SD, Jeniffer S, Punde R, et al. A strategy for extending theapplicability of a validated plasma calibration curve to quanti-tative measurements in multiple tissue homogenate samples: acase study from a rat tissue distribution study of JI-101, a triplekinase inhibitor. Biomed Chromatogr 2012;26:419–424.

Haasnoot W, Schilt R, Hamers AR, et al. Determination of beta-19-nortestosterone and its metabolite alpha-19-nortestosteronein biological samples at the sub parts per billion level by high-performance liquid chromatography with on-line immunoaffin-ity sample pretreatment. J Chromatogr 1989;489:157–171.

Helle N, Baden M, Petersen K. Automated solid phase extraction.Methods Mol Biol 2011;747:93–129.

Heudi O, Barteau S, Zimmer D, et al. Towards absolute quantifica-tion of therapeutic monoclonal antibody in serum by LC-MS/MSusing isotope-labeled antibody standard and protein cleavageisotope dilution mass spectrometry. Anal Chem 2008;80:4200–4207.

Hussain S, Patel H, Tan A. Automated liquid-liquid extractionmethod for high-throughput analysis of rosuvastatin in humanEDTA K2 plasma by LC-MS/MS. Bioanalysis 2009;1:529–535.

Jemal M, Schuster A, Whigan DB. Liquid chromatography/tandemmass spectrometry methods for quantitation of mevalonic acidin human plasma and urine: method validation, demonstrationof using a surrogate analyte, and demonstration of unacceptablematrix effect in spite of use of a stable isotope analog internalstandard. Rapid Commun Mass Spectrom 2003;17:1723–1734.

Jemal M, Xia YQ. LC-MS Development strategies for quantitativebioanalysis. Curr Drug Metab 2006;7:491–502.

Jenkins KM, Young MS, Mallet CR, Elian AA. Mixed-mode solid-phase extraction procedures for the determination of MDMAand metabolites in urine using LC-MS, LC-UV, or GC-NPD.J Anal Toxicol 2004;28:50–58.

Ji AJ, Jiang Z, Livson Y, Davis JA, Chu JX, Weng N. Challengesin urine bioanalytical assays: overcoming nonspecific binding.Bioanalysis 2010;2:1573–1586.

Jiang H, Cao H, Zhang Y, Fast DM. Systematic evaluation of sup-ported liquid extraction in reducing matrix effect and improv-ing extraction efficiency in LC-MS/MS based bioanalysis for10 model pharmaceutical compounds. J Chromatogr B AnalytTechnol Biomed Life Sci 2012;891–892:71–80

Jiang H, Randlett C, Junga H, Jiang X, Ji QC. Using supported liq-uid extraction together with cellobiohydrolase chiral stationaryphases-based liquid chromatography with tandem mass spec-trometry for enantioselective determination of acebutolol andits active metabolite diacetolol in spiked human plasma. J Chro-matogr B Analyt Technol Biomed Life Sci 2009;877:173–180.

Jiang H, Zeng J, Zheng N, et al. A convenient strategy for quantita-tive determination of drug concentrations in tissue homogenatesusing a liquid chromatography/tandem mass spectrometry assayfor plasma samples. Anal Chem 2011;83:6237–6244.

Kaye B, Herron WJ, Macrae PV, et al. Rapid, solid phase extractiontechnique for the high-throughput assay of darifenacin in humanplasma. Anal Chem 1996;68:1658–1660.

Khan JK, Bu HZ, Samarendra ZZ, Maiti N, Micetich RG. A rapidand reliable solid-phase extraction–LC/MS/MS assay for thedetermination of two novel human leukocyte elastase inhibitors,SYN-1390 and SYN-1396, in rat plasma. J Pharm Biomed Anal1999;20:697–703.

King R, Bonfiglio R, Fernandez-Metzler C, Miller-Stein C, Olah T.Mechanistic investigation of ionization suppression in electro-spray ionization. J Am Soc Mass Spectrom 2000;11:942–950.

Kole PL, Venkatesh G, Kotecha J, Sheshala R. Recent advances insample preparation techniques for effective bioanalytical meth-ods. Biomed Chromatogr 2011;25:199–217.

Kosovec JE, Egorin MJ, Gjurich S, Beumer JH. Quantitation of5-fluorouracil (5-FU) in human plasma by liquid chromatogra-phy/electrospray ionization tandem mass spectrometry. RapidCommun Mass Spectrom 2008;22:224–230.

Lahaie M, Mess JN, Furtado M, Garofolo F. Elimination ofLC-MS/MS matrix effect due to phospholipids using spe-cific solid-phase extraction elution conditions. Bioanalysis2010;2(6):1011–1021.

Lasakova M, Jandera P. Molecularly imprinted polymers and theirapplication in solid phase extraction. J Sep Sci 2009;32:799–812.

Lee CR, Esnaud H. Determination of melatonin by GC-MS: prob-lems with solid phase extraction (SPE) columns. Biomed Envi-ron Mass Spectrom 1988;15:677–679.

Lehotay SJ, de Kok A, Hiemstra M, Van Bodegraven P. Validationof a fast and easy method for the determination of residuesfrom 229 pesticides in fruits and vegetables using gas and liquidchromatography and mass spectrometric detection. J AOAC Int2005;88:595–614.

Li F, Fast D, Michael S. Absolute quantitation of protein therapeu-tics in biological matrices by enzymatic digestion and LC-MS.Bioanalysis 2011;3:2459–2480.

Li W, Luo S, Smith HT, Tse FL. Quantitative determination ofBAF312, a S1P-R modulator, in human urine by LC-MS/MS:

Page 18: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

182 BEST PRACTICES IN BIOLOGICAL SAMPLE PREPARATION FOR LC-MS BIOANALYSIS

prevention and recovery of lost analyte due to container surfaceadsorption. J Chromatogr B Analyt Technol Biomed Life Sci2010;878:583–589.

Li W, Tse FL. Dried blood spot sampling in combination with LC-MS/MS for quantitative analysis of small molecules. BiomedChromatogr 2010;24:49–65.

Li W, Zhang J, Tse FL. Strategies in quantitative LC-MS/MS anal-ysis of unstable small molecules in biological matrices. BiomedChromatogr 2011;25:258–277.

Liang X, Ubhayakar S, Liederer BM, et al. Evaluation of homoge-nization techniques for the preparation of mouse tissue samplesto support drug discovery. Bioanalysis 2011;3:1923–1933.

Liu G, Ji QC, Arnold ME. Identifying, evaluating, and control-ling bioanalytical risks resulting from nonuniform matrix ionsuppression/enhancement and nonlinear liquid chromatography-mass spectrometry assay response. Anal Chem 2010a;82:9671–9677.

Liu G, Patrone L, Snapp HM, et al. Evaluating and defining samplepreparation procedures for DBS LC-MS/MS assays. Bioanalysis2010b;2:1405–1414.

Liu G, Snapp HM, Ji QC, Arnold ME. Strategy of acceleratedmethod development for high-throughput bioanalytical assaysusing ultra high-performance liquid chromatography coupledwith mass spectrometry. Anal Chem 2009;81:9225–9232.

Ma J, Shi J, Le H, et al. A fully automated plasma protein precip-itation sample preparation method for LC-MS/MS bioanalysis.J Chromatogr B Analyt Technol Biomed Life Sci 2008;862:219–226.

Mallet CR, Lu Z, Fisk R, Mazzeo JR, Neue UD. Performanceof an ultra-low elution-volume 96-well plate: drug discoveryand development applications. Rapid Commun Mass Spectrom2003;17:163–170.

McCauley-Myers DL, Eichhold TH, Bailey RE, et al. Rapid bio-analytical determination of dextromethorphan in canine plasmaby dilute-and-shoot preparation combined with one minute persample LC-MS/MS analysis to optimize formulations for drugdelivery. J Pharm Biomed Anal 2000;23:825–835.

Medina-Casanellas S, Benavente F, Barbosa J, Sanz-Nebot V.Preparation and evaluation of an immunoaffinity sorbent forthe analysis of opioid peptides by on-line immunoaffinity solid-phase extraction capillary electrophoresis-mass spectrometry.Anal Chim Acta 2012;717:134–142.

Mesmin C, Fenaille F, Ezan E, Becher F. MS-based approachesfor studying the pharmacokinetics of protein drugs. Bioanalysis2011;3:477–480.

Mirmahdieh S, Mardihallaj A, Hashemian Z, Razavizadeh J, Ghazi-askar H, Khayamian T. Analysis of testosterone in human urineusing molecularly imprinted solid-phase extraction and coronadischarge ion mobility spectrometry. J Sep Sci 2011;34:107–112.

Moreno-Bondi MC, Marazuela MD, Herranz S, Rodriguez E. Anoverview of sample preparation procedures for LC-MS multi-class antibiotic determination in environmental and food sam-ples. Anal Bioanal Chem 2009;395:921–946.

Mornar A, Sertic M, Turk N, Nigovic B, Korsic M. Simultaneousanalysis of mitotane and its main metabolites in human blood and

urine samples by SPE-HPLC technique. Biomed Chromatogr2012.

Muller C, Schafer P, Stortzel M, Vogt S, Weinmann W. Ion suppres-sion effects in liquid chromatography-electrospray-ionisationtransport-region collision induced dissociation mass spectrom-etry with different serum extraction methods for systematic tox-icological analysis with mass spectra libraries. J Chromatogr BAnalyt Technol Biomed Life Sci 2002;773:47–52.

Mullett WM, Lai EP. Rapid determination of theophylline in serumby selective extraction using a heated molecularly imprintedpolymer micro-column with differential pulsed elution. J PharmBiomed Anal 1999;21:835–843.

Neubert H, Gale J, Muirhead D. Online high-flow peptideimmunoaffinity enrichment and nanoflow LC-MS/MS: assaydevelopment for total salivary pepsin/pepsinogen. Clin Chem2010;56:1413–1423.

Novakova L, Vlckova H. A review of current trends and advancesin modern bio-analytical methods: chromatography and samplepreparation. Anal Chim Acta 2009;656:8–35.

Ouyang Z, Furlong MT, Wu S, et al. Pellet digestion: a simple andefficient sample preparation technique for LC-MS/MS quan-tification of large therapeutic proteins in plasma. Bioanalysis2012;4:17–28.

Patel BN, Sharma N, Sanyal M, Shrivastav PS. Simultaneous deter-mination of simvastatin and simvastatin acid in human plasmaby LC-MS/MS without polarity switch: application to a bioe-quivalence study. J Sep Sci 2008;31:301–313.

Peng SX, King SL, Bornes DM, Foltz DJ, Baker TR, Natchus MG.Automated 96-well SPE and LC-MS-MS for determination ofprotease inhibitors in plasma and cartilage tissues. Anal Chem2000;72:1913–1917.

Piletska EV, Burns R, Terry LA, Piletsky SA. Applicationof a molecularly imprinted polymer for the extraction ofkukoamine a from potato peels. J Agric Food Chem 2012;60:95–99.

Pin H, Hong-Min L, Ming Y, Qin L. A validated LC-MS/MSmethod for the determination of vinflunine in plasma and itsapplication to pharmacokinetic studies. Biomed Chromatogr2012; 26:797–801.

Plossl F, Giera M, Bracher F. Multiresidue analytical method usingdispersive solid-phase extraction and gas chromatography/iontrap mass spectrometry to determine pharmaceuticals in wholeblood. J Chromatogr A 2006;1135:19–26.

Polson C, Sarkar P, Incledon B, Raguvaran V, Grant R. Optimiza-tion of protein precipitation based upon effectiveness of proteinremoval and ionization effect in liquid chromatography-tandemmass spectrometry. J Chromatogr B Analyt Technol BiomedLife Sci 2003;785:263–275.

Posyniak A, Zmudzki J, Mitrowska K. Dispersive solid-phaseextraction for the determination of sulfonamides in chicken mus-cle by liquid chromatography. J Chromatogr A 2005;1087:259–264.

Pucci V, Di Palma S, Alfieri A, Bonelli F, Monteagudo E. Anovel strategy for reducing phospholipids-based matrix effect inLC-ESI-MS bioanalysis by means of HybridSPE. J PharmBiomed Anal 2009;50:867–871.

Page 19: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

REFERENCES 183

Raffaelli A, Saba A, Vignali E, Marcocci C, Salvadori P.Direct determination of the ratio of tetrahydrocortisol + allo-tetrahydrocortisol to tetrahydrocortisone in urine by LC-MS-MS. J Chromatogr B Analyt Technol Biomed Life Sci2006;830:278–285.

Rashed MS, Saadallah AA, Rahbeeni Z, et al. Determination ofurinary S-sulphocysteine, xanthine and hypoxanthine by liq-uid chromatography-electrospray tandem mass spectrometry.Biomed Chromatogr 2005;19:223–230.

Rudewicz PJ. Turbulent flow bioanalysis in drug metabolism andpharmacokinetics. Bioanalysis 2011;3:1663–1671.

Rustum AM. Determination of cadralazine in human whole bloodusing reversed-phase high-performance liquid chromatogra-phy: utilizing a salting-out extraction procedure. J Chromatogr1989;489:345–352.

Saar E, Gerostamoulos D, Drummer OH, Beyer J. Identification andquantification of 30 antipsychotics in blood using LC-MS/MS.J Mass Spectrom 2010;45:915–925.

Samanidou V, Kovatsi L, Fragou D, Rentifis K. Novel strate-gies for sample preparation in forensic toxicology. Bioanalysis2011;3:2019–2046.

Sellergren B. Direct Drug Determination by Selective SampleEnrichment on an Imprinted Polymer. Anal Chem 1994;66:1578–1582.

Shou WZ, Jiang X, Beato BD, Naidong W. A highly automated 96-well solid phase extraction and liquid chromatography/tandemmass spectrometry method for the determination of fentanylin human plasma. Rapid Commun Mass Spectrom 2001;15:466–476.

Shou WZ, Pelzer M, Addison T, Jiang X, Naidong W. An auto-matic 96-well solid phase extraction and liquid chromatography-tandem mass spectrometry method for the analysis of morphine,morphine-3-glucuronide and morphine-6-glucuronide in humanplasma. J Pharm Biomed Anal 2002;27:143–152.

Song Q, Naidong W. Analysis of omeprazole and 5-OH omeprazolein human plasma using hydrophilic interaction chromatographywith tandem mass spectrometry (HILIC-MS/MS)–eliminatingevaporation and reconstitution steps in 96-well liquid/liquidextraction. J Chromatogr B Analyt Technol Biomed Life Sci2006;830:135–142.

Souverain S, Rudaz S, Veuthey JL. Restricted access materials andlarge particle supports for on-line sample preparation: an attrac-tive approach for biological fluids analysis. J Chromatogr BAnalyt Technol Biomed Life Sci 2004;801:141–156.

Strader MB, Tabb DL, Hervey WJ, Pan C, Hurst GB. Efficient andspecific trypsin digestion of microgram to nanogram quanti-ties of proteins in organic-aqueous solvent systems. Anal Chem2006;78:125–134.

Thomas A, Deglon J, Steimer T, Mangin P, Daali Y, Staub C.On-line desorption of dried blood spots coupled to hydrophilicinteraction/reversed-phase LC/MS/MS system for the simulta-neous analysis of drugs and their polar metabolites. J Sep Sci2010;33:873–879.

Tse Sum Bui B, Haupt K. Molecularly imprinted polymers: syn-thetic receptors in bioanalysis. Anal Bioanal Chem 2010;398:2481–2492.

Tweed JA, Gu Z, Xu H, et al. Automated sample preparation for reg-ulated bioanalysis: an integrated multiple assay extraction plat-form using robotic liquid handling. Bioanalysis 2010;2:1023–1040.

Vaezzadeh AR, Deshusses JM, Waridel P, et al. Accelerated diges-tion for high-throughput proteomics analysis of whole bacterialproteomes. J Microbiol Methods 2010;80:56–62.

van der Heeft E, Dijkman E, Baumann RA, Hogendoorn EA. Com-parison of various liquid chromatographic methods involvingUV and atmospheric pressure chemical ionization mass spec-trometric detection for the efficient trace analysis of phenylureaherbicides in various types of water samples. J Chromatogr A2000;879:39–50.

Vogeser M, Kirchhoff F. Progress in automation of LC-MS in lab-oratory medicine. Clin Biochem 2011;44:4–13.

Vuckovic D, Zhang X, Cudjoe E, Pawliszyn J. Solid-phase microex-traction in bioanalysis: New devices and directions. J Chro-matogr A 2010;1217:4041–4060.

Wang PG, Zhang J, Gage EM, et al. A high-throughput liquidchromatography/tandem mass spectrometry method for simul-taneous quantification of a hydrophobic drug candidate andits hydrophilic metabolite in human urine with a fully auto-mated liquid/liquid extraction. Rapid Commun Mass Spectrom2006;20:3456–3464.

Wang S, Cyronak M, Yang E. Does a stable isotopically labeledinternal standard always correct analyte response? A matrixeffect study on a LC/MS/MS method for the determination ofcarvedilol enantiomers in human plasma. J Pharm Biomed Anal2007;43:701–707.

Watt AP, Morrison D, Locker KL, Evans DC. Higher throughputbioanalysis by automation of a protein precipitation assay usinga 96-well format with detection by LC-MS/MS. Anal Chem2000;72:979–984.

Whelan M, Kinsella B, Furey A, et al. Determination of anthelminticdrug residues in milk using ultra high performance liquidchromatography-tandem mass spectrometry with rapid polarityswitching. J Chromatogr A 2010;1217:4612–4622.

Wilson ID. High-performance liquid chromatography-mass spec-trometry (HPLC-MS)-based drug metabolite profiling. MethodsMol Biol 2011;708:173–190.

Wood M, Laloup M, Samyn N, et al. Simultaneous analysis ofgamma-hydroxybutyric acid and its precursors in urine using liq-uid chromatography-tandem mass spectrometry. J ChromatogrA 2004;1056:83–90.

Wu H, Zhang J, Norem K, El-Shourbagy TA. Simultaneous deter-mination of a hydrophobic drug candidate and its metabolite inhuman plasma with salting-out assisted liquid/liquid extractionusing a mass spectrometry friendly salt. J Pharm Biomed Anal2008a;48:1243–1248.

Wu ST, Ouyang Z, Olah TV, Jemal M. A strategy for liq-uid chromatography/tandem mass spectrometry based quan-titation of pegylated protein drugs in plasma using plasmaprotein precipitation with water-miscible organic solventsand subsequent trypsin digestion to generate surrogate pep-tides for detection. Rapid Commun Mass Spectrom 2011;25:281–290.

Page 20: Handbook of LC-MS Bioanalysis (Best Practices, Experimental Protocols, and Regulations) || Best Practices in Biological Sample Preparation for LC-MS Bioanalysis

184 BEST PRACTICES IN BIOLOGICAL SAMPLE PREPARATION FOR LC-MS BIOANALYSIS

Wu ST, Schoener D, Jemal M. Plasma phospholipids implicatedin the matrix effect observed in liquid chromatography/tandemmass spectrometry bioanalysis: evaluation of the use of colloidalsilica in combination with divalent or trivalent cations for theselective removal of phospholipids from plasma. Rapid CommunMass Spectrom 2008b;22:2873–2881.

Xia YQ, Jemal M. Phospholipids in liquid chromatography/massspectrometry bioanalysis: comparison of three tandem massspectrometric techniques for monitoring plasma phospholipids,the effect of mobile phase composition on phospholipids elutionand the association of phospholipids with matrix effects. RapidCommun Mass Spectrom 2009;23:2125–2138.

Xu RN, Fan L, Rieser MJ, El-Shourbagy TA. Recent advancesin high-throughput quantitative bioanalysis by LC-MS/MS.J Pharm Biomed Anal 2007;44:342–355.

Xu Y, Du L, Rose MJ, Fu I, Woolf EJ, Musson DG. Con-cerns in the development of an assay for determination of ahighly conjugated adsorption-prone compound in human urine.J Chromatogr B Analyt Technol Biomed Life Sci 2005;818:241–248.

Xue YJ, Akinsanya JB, Liu J, Unger SE. A simplified proteinprecipitation/mixed-mode cation-exchange solid-phase extrac-tion, followed by high-speed liquid chromatography/mass spec-trometry, for the determination of a basic drug in human plasma.Rapid Commun Mass Spectrom 2006a;20:2660–2668.

Xue YJ, Liu J, Pursley J, Unger S. A 96-well single-pot proteinprecipitation, liquid chromatography/tandem mass spectrometry(LC/MS/MS) method for the determination of muraglitazar, a

novel diabetes drug, in human plasma. J Chromatogr B AnalytTechnol Biomed Life Sci 2006b;831:213–222.

Xue YJ, Yan JH, Arnold M, Grasela D, Unger S. Quantitative deter-mination of BMS-378806 in human plasma and urine by high-performance liquid chromatography/tandem mass spectrometry.J Sep Sci 2007;30:1267–1275.

Yang J, Hu Y, Cai JB, Zhu XL, Su QD. A new molecularly imprintedpolymer for selective extraction of cotinine from urine samplesby solid-phase extraction. Anal Bioanal Chem 2006;384:761–768.

Yang Z, Hayes M, Fang X, Daley MP, Ettenberg S, Tse FL. LC-MS/MS approach for quantification of therapeutic proteins inplasma using a protein internal standard and 2D-solid-phaseextraction cleanup. Anal Chem 2007;79:9294–9301.

Yoshida M, Akane A, Nishikawa M, Watabiki T, Tsuchihashi H.Extraction of thiamylal in serum using hydrophilic acetonitrilewith subzero-temperature and salting-out methods. Anal Chem2004;76:4672–4675.

Yue H, Borenstein MR, Jansen SA, Raffa RB. Liquidchromatography-mass spectrometric analysis of buprenorphineand its N-dealkylated metabolite norbuprenorphine in rat braintissue and plasma. J Pharmacol Toxicol Methods 2005;52:314–322.

Zhang J, Wu H, Kim E, El-Shourbagy TA. Salting-out assistedliquid/liquid extraction with acetonitrile: a new high through-put sample preparation technique for good laboratory prac-tice bioanalysis using liquid chromatography-mass spectrom-etry. Biomed Chromatogr 2009;23:419–425.