Review Special Focus Issue: Clinical Chemistry
For reprint orders, please contact [email protected]
Heterogeneous and homogeneous immunoassays for drug analysis
Immunoassays are very useful techniques to perform screening and semi-quantitative analysis of hundreds of different xenobiotics. Small sample volumes are required and pretreatment is usually unnecessary (e.g., homogeneous immunoassays). Fully automated and high-throughput systems are available, which help physicians to take timely decisions. However, immunoassays do suffer from interference from both endogenous and exogenous factors that limit their application in quantitative analysis. These assays use different labels (e.g., colorimetric, fluorescent, chemiluminescent or electrochemiluminescent) and different methods for generating and measuring signals, but the basic principles are usually similar. This review outlines the practical aspects of immunoassays in bioanalysis and describes their application in clinical chemistry for xenobiotic analysis, namely medicines and drugs of abuse.
Ricardo Jorge Dinis-Oliveira IINFACTS - Institute of Research & Advanced Training in Health Sciences & Technologies, Department of Sciences, Advanced Institute of Health Sciences – North (ISCS-N), CESPU, CRL, Gandra / Department of Legal Medicine & Forensic Sciences, Faculty of Medicine, University of Porto, Porto / REQUIMTE, Laboratory of Toxicology, Department of Biological Sciences, Faculty of Pharmacy, University of Porto, Porto, Portugal Tel.: +351 222 073 850 ricardinis@ sapo.pt
Keywords: drugs • heterogeneous format • homogeneous format • immunoassays • interferences
Background Immunoassays are quick analytical techniques that use different antibodies, labels (e.g., colorimetric, fluorescent, chemiluminescent or electrochemiluminescent) and methods for generating and measuring signals related to the presence or concentration of analytes. The first immunoassay (radioimmunoassay) was described by Berson and Yalow in 1959 for insulin quantification [1] . For this discovery they were awarded the Nobel Prize in Medicine in 1977. Since then, in just half a century, immunoassays for hundreds of other analytes have been developed [2] . In clinical and forensic chemistry/toxicology, immunoassays are widely used as a bioanalytical technique to screen biological samples for the presence of a xenobiotic in a variety of situations [3–5] . Indeed, immunoassays for drugs of abuse and therapeutic drug monitoring are routinely used for screening (qualitative) and semi-quantitative analysis, respectively [4,6] . Table 1 summarizes other common immunoassay applications for drug analysis. Particularly in forensic toxicology,
10.4155/BIO.14.208 © 2014 Future Science Ltd
scientific societies [7] have established that all positive results must be confirmed by more specific and sensitive techniques, preferably GC–MS and LC–MS. In other words, immunoassays usually provide a preliminary result. In fact, it is well known that chemically and non-chemically related xenobiotics can sometimes interfere with analysis, leading to false-positive and false-negative results [8,9] and the risk of sample adulteration must be taken into account [10] . Therefore, one may question if there is still a role for immunoassay in clinical and forensic toxicology. The answer to this is definitely yes. Table 2 details some of the most relevant advantages and limitations of immunoassays for drug analysis. Here, the practical aspects of all immunoassays devoted to xenobiotic analysis, namely medicines and drugs of abuse, are reviewed. Articles written in English, German, French, Spanish and Portuguese that detailed immunoassays applied to drug analysis (e.g., amphetamines, benzodiazepines, opioids, hallucinogens,
Bioanalysis (2014) 6(21), 2877–2896
part of
ISSN 1757-6180
2877
Review
Dinis-Oliveira
Key terms Immunoassay: An analytical technique based on antibody–antigen recognition. Xenobiotic: (Of the Greek, xenos ‘stranger’; bios ‘life’). Foreign chemical (e.g., medicines, pesticides and drugs of abuse) found within an organism that is not naturally produced by or expected to be present within that organism. It is opposed to the endobiotic concept. Point-of-care testing: Analyses are performed outside of the traditional central laboratory in close proximity to the patient or other subject.
cannabinoids, cocaine, barbiturates, antiepileptic, tricyclic antidepressants, immunosuppressants, cardioactive compounds, aminoglycosides and vancomycin) were searched for using the national library of medicine’s PubMed MedLine database and the web of knowledge. Classfication of immunoassays Different approaches can be used to classify/characterize immunoassays. Table 3 highlights the most important details for classification, which are further developed below: • Number of processed samples: point-of-care testing (POCT) or automated immunoassays could be used to analyze one or several samples, respectively; • Analytical platforms (e.g., microplate, microcolumn, flow, membrane and bead) using either closed or open designs. The former design can only make use reagents of the platform manufacturer’s. Moreover, existing methods could not be improved and new assays for different analytes to those on
the equipment’s repertoire could not be validated. Furthermore, the scientific spirit of the analyst is very difficult to be applied, and existing errors could not be corrected. An interesting review on this subject can be found in [11] ; • Existence or absence of competition: –– Competitive design uses only one antibody and is mostly applied for analytes with small molecular weight (e.g., xenobiotics). Predetermined amounts of labeled xenobiotic (i.e., hapten) and antibody are added to the specimen followed by incubation. In the basic design, the analyte molecules present in the specimen compete with labeled xenobiotic molecules for the antibody binding sites molecules [12,13] . If the signal is generated when the labeled xenobiotic binds to an antibody molecule, the signal is inversely proportional to analyte concentration in the specimen (e.g., fluorescence polarization immunoassay [FPIA]). On the other hand, if the signal is generated by an unbound labeled xenobiotic, the assay signal is directly proportional to the xenobiotic concentration (e.g., enzyme multiplied immunoassay technique [EMIT]); –– Immunometric or noncompetitive (‘sandwich’) design uses two antibodies that recognize different parts of the analyte, usually of large molecular weight (i.e., to accommodate two different antibody molecules), such as proteins, peptides or polysaccharides. An excess of captured antibodies directed to the analyte is usually immobilized on a solid support (e.g., microparticle bead, a coated tube or,
Table 1. Most important applications of immunoassays for xenobiotic analysis. Applications (urine, sweat, serum, plasma, blood, exhaled air, hair and oral fluid) Evaluate potential drug abuse or overdose in the emergency room Diagnosis and management of pregnant drug abusers Evaluate newborns for in utero drug exposure Ensure abstinence in drug dependency rehabilitation programs Therapeutic drug monitoring Workplace drug testing Pre-employment drug screening Bioequivalence studies in drug discovery and pharmaceutical industries Driving under the influence of psychoactive drugs (including ethanol) Self-monitoring during therapeutic Doping control (e.g., anabolic androgenic steroids)
2878
Bioanalysis (2014) 6(21)
future science group
Heterogeneous & homogeneous immunoassays for drug analysis
more commonly, the surface of a microplate). The specimen is then added and incubation with the antibody is left to occur for a predetermined time [12,13] . After washing, a liquid reagent containing the secondary antibody conjugated with a molecule for signal production (e.g., an enzyme) is added. The secondary antibody binds to the captured analyte forming a ‘sandwich’. A second washing step to remove unbound secondary antibody and the addition of a substrate for signal development follows this. Xenobiotic concentration is directly proportional to the intensity of the signal;
Review
• Outcomes: –– Qualitative (screening) immunoassays only indicates that xenobiotics present in the sample are above a particular cutoff concentration but do not give information about which xenobiotic is present or respective concentrations. These cutoff values can be the lower limit of quantification of the method or a higher value. Laws can also define these cutoffs. A full calibration line is normally unnecessary and, therefore, less complex since only one calibrator at the cutoff concentration is needed. Negative results reflect concentrations below the cutoff and do not
Table 2. Advantages and limitations of immunoassays for xenobiotic analysis. Advantages
Limitations
Relatively easily automated and integrated into the laboratory workflow
Are intrinsically non-specific (not entirely true since it is dependent on appropriate assay development, particularly with respect to antiserum generation) for xenobiotic analysis, particularly with respect to metabolites
Require less technical sophistication on the part of staff (simplicity)
Compared with other bioanalytical techniques, requires a long assay development time (i.e.,’usually some months to generate the reagent antibody, whether monoclonal or polyclonal), with no guarantee of success
Speedy analysis (e.g., results can be reported in 10–30 min) that allow physicians to take timely decisions
Reduced precision
Very sensitive assays with very small sample volumes (10–50 μl)
Not commercially available for all drugs (finite number) currently monitored in clinical practice
There is generally no need for a Interferences from drug metabolites, other drugs with similar sample preparation or extraction step structures, herbal supplements as well as various endogenous prior to analysis (e.g., samples can be factors assayed directly) High-throughput methods
Cannot raise antibodies to some inorganic or very small organic compounds, such as carbon monoxide, cyanide ion, ethanol, ethylene glycol and lithium
Much less complex than LC and GC combined with mass spectrometry
Many drugs (e.g., cardioactives, psychoactives) have active metabolites, which should be measured together
Antibodies directed to a certain class of drugs are useful for screening purposes
Cannot raise selective antibodies to certain drugs (e.g., amiodarone cross reacts with thyroxine)
Are cost-effective since testing can usually performed in existing automated platforms, and laboratories can rapidly introduce new immunoassays if necessary
Typically the kits are approved for certain biological matrices (usually urine and serum or plasma). To be applied to other samples (e.g., solid tissues), extraction or dilution is necessary followed by validation
New compounds are introduced every day in the market and, as Reagents can be stored in the analyzer and they are often stable for such, the absence of antibodies could lead to false-negative results 1–2 months Auto-flagging to alert poor specimen quality (e.g., hemolysis, high bilirubin and lipemic)
future science group
www.future-science.com
2879
Review
Dinis-Oliveira
Table 3. Classification of immunoassays. Immunoassay
Format
Common assay signals
RIA
Heterogeneous
Radioactive counting
ELISA
Heterogeneous/competition or sandwich
Colorimetry
ELFA†
Heterogeneous/competition or sandwich
Fluorescence
MEIA†
Heterogeneous/sandwich
Fluorescence
EMIT†
Homogeneous/competition
Absorbance at 340 nm
CEDIA
Homogeneous/competition
Colorimetry
KIMS / PETINIA
Homogeneous
Turbidimetry
FPIA
Homogeneous
Fluorescence polarization
LOCI
Homogeneous/competition or sandwich
Chemiluminescence
CMIA
Heterogeneous/competition
Chemiluminescence (acridinium ester label)
ECLIA
Heterogeneous/competition or sandwich
Electrochemiluminescence
LIA
Homogeneous or Heterogeneous/ competition or sandwich
Electrochemical (e.g., voltammetric, potentiometric, capacitive, conductometric and impedimetric)
†
†
These are enzyme Immunoassays. CEDIA: Cloned enzyme donor immunoassay; CMIA:Chemiluminescent microparticle immunoassay; ECLIA: Electro-chemiluminescence immunoassay; ELFA: Enzyme-linked fluorescent assay; ELISA: Enzyme-linked immunosorbent assay; EMIT: Enzyme-multiplied immunoassay technique; FPIA: Fluorescence polarization immunoassay; KIMS: Kinetic interaction of microparticles in solution; LIA: Liposome immunoassay; LOCI: Luminescent oxygen channeling immunoassay; MEIA: Microparticle enzyme immunoassay; PETINIA: Particle enhanced turbidimetric inhibition immunoassay; RIA: Radioimmunoassay.
†
exclude the presence of the xenobiotic. Samples close to the cutoff may give negative or positive results depending on the precision of the assay and influence of matrix effects; –– In semi-quantitative immunoassays, the xenobiotic concentration can be obtained by comparing the produced signal with calibrators. There is almost always a non-linear relationship (since the number of antibody binding sites is finite) between the measured response in the assay and the analyte concentration [14] . Since there is no universal mathematical function that fits the results, bias could be introduced. Moreover, greater errors are registered at the extremes (e.g., becoming asymptotic) than in the central linear region of the calibration curve. Polynomial or logit functions ares usually applied. Computer software usually helps to aid the decision of ‘best fit’ according to the obtained standard points (e.g., spline fit) or on a mathematical model that reflects the physical principles underlying immunoassays (e.g., four-parameter logistic fit) [14] . The latter is preferred since it is less subjected to the bias of any inaccurate standard points [14] .
2880
Bioanalysis (2014) 6(21)
• Existence/no existance of a separation step: –– Heterogeneous format – antibody-bound label must be separated from the free label before measuring the signal. Consequently, there is less opportunity for interference in comparison with many homogenous immunoassays [2] . The separation step can be simplified by using 96-well microplates or paramagnetic particles; –– Homogeneous assays – after incubation, a separation step is not required to distinguish the signal produced by antibody-bound and free label. Nevertheless, extraction of the analytes with solvents may be useful before to the immunoassay to increase selectivity. Although easier to perform than heterogeneous immunoassays, the matrix effects may influence the results [2] ; • Signal produced – non-isotopic (e.g., enzyme immunoassay [EIA]) and isotopic (radioimmunoassay [RIA]) immunoassays could be considered. Characteristics of both types are discussed in the following sections.
future science group
Heterogeneous & homogeneous immunoassays for drug analysis
Point-of-care-testing (POCT) Aiming to provide an immediate result, POCT or bedside testing is any analysis performed outside of the traditional central laboratory and in close proximity to the patient or other subject. For xenobiotic analysis, urine, blood, oral fluid or sweat are usually applied to screening in roadside and workplace testing [4,15] . Supplementary Table 1 lists some of the advantages and limitations of POCT. In the majority of the POCT devices, a competitive lateral flow immunoassays format (also known as immunochromatographic assay) is used. In addition, most of the devices use a negative-indicating reaction, meaning that in the absence of a visible test line, the reaction is positive. Briefly, the strips (normally made of nitrocellulose), often enclosed by a plastic test cassette, have a reservoir (well) with labeled antibodies antixenobiotic (e.g., chromogenic, fluorescent, paramagnetic monodisperse latex particle, etc.) into which the sample is pipetted (Figure 1) . By manipulating the concentration of antibodies, POCT can be manufactured to provide the required cut-off values. At an appropriate distance from the origin, the xenobiotic(s) immobilized (in a line format) on the strip will capture any antibody not bound to a xenobiotic. Therefore, a visible line of labeled antibody is produced, the signal being maximal if sample is negative. To assess that the device is working properly (e.g., enough sample volume, quality of
Key terms Heterogeneous format: Antibody-bound label must be separated from the free label before measuring the signal. Interference: It is the alteration of the correct value (accuracy) of the analyte, usually expressed as concentration or activity, due to the presence of a substance in the sample. Homegeneous format: A separation step is not required to distinguish the signal produced by antibody-bound and free label.
reagents), a control line normally exists at an appropriate distance from the immobilized xenobiotic. The excess of free labeled antibody further migrates and will be captured by the anti-immunoglobulin antibody to form a second visible line, which must be always present both in positive and negative analysis. In positive analysis, the first line is not produced since the xenobiotic in the sample will bind all the antibody. If the device uses labeled xenobiotic, and the immobilized antibodies being bound onto the membrane, the positive result is indicated by the appearance of a line. With either type device, faint lines should be read as negative [16] . Modern radioimmunoassay (RIA) presents a scheme of RIA for negative and positive analysis. Typically, 125I is used
Supplementary Figure 1
Control line line Control
Test line
A A
Review
visibleline line 11visible B B
2 visible lines 2 visible lines Xenobiotic Xenobiotic
Chromogeniclabeled labelled Chromogenic antibody antibody anti-xenobiotic anti-xenobiotic
Anti-IgG antibody
Flow Flowdirection direction
Figure 1. Lateral flow immunoassay. (A) The excess of sample xenobiotic binds all antibody preventing line formation. (B) Labeled antibody binds to immobilized xenobiotic giving a visible line.
future science group
www.future-science.com
2881
Review
Dinis-Oliveira as the radiolabel and requires a γ-counter to measure radioactivity in the tubes. Iodination procedures can either be direct (i.e., the xenobiotic is labeled by replacing an atom of hydrogen with 125I), or indirect (i.e., the xenobiotic is labeled with a pre-iodinated molecule). RIA reagents are available for many xenobiotics, cost is reasonable and has very low limits of detection. Typically, tubes are precoated with antibody anti-xenobiotic. Radiolabeled xenobiotic is then added, which binds to antibody anti-xenobiotic. The xenobiotic present in the sample will compete with the radioactive xenobiotic bound to the antibody. Simple decanting easily separates the bound and free fractions, and allows quantification of the bound fraction. The radioactivity signal is relatively free of matrix interference and less susceptible to influence by urine adulteration, when compared with other immunoassays. The major limitations relate to the use of radioactive materials (e.g., need for radioactive materials licensure and waste disposal, unsafe radioactive handling and storage, cost, and the relatively short shelf-life of the reagents, which is usually less than 60 days). Therefore, in recent years, the trend has been away from RIA and toward non-isotopic immunoassays. ELISA Different ELISA formats exist, including direct, indirect, sandwich and competitive. Unfortunately, these terms may be misleading as the same method may be described by more than one term [16] . Figures 2, Supplementary Figures 2 & 3 present the scheme of the most common ELISA formats for negative and positive analysis. The following topics are highlighted: • It is the most widely used immunoassay; • They are more laborious than homogeneous EIAs but are already semi-automated (i.e., pipetting and washing); • These assays are typically performed in 96-well microplates and thus are not compatible with highthroughput clinical chemistry analyzers; • Since it is heterogeneous, less matrix interferences are registered, especially if a double wash is performed; • Costs are somewhat higher than homogeneous EIA reagents; • It can be used with whole blood, typically diluted with water or buffer, and with proper validation, even for tissue homogenates; • Although less than RIA, ELISA has a good sensitivity and dynamic range, especially if a fluorescent signal is used (limit of detection of ~1–5 pg/well
2882
Bioanalysis (2014) 6(21)
compared with ~5–200 pg/well in clorimetric format) [17] ; • Steps are often long and time precision for addition of substrate and stop reagents is required to obtain accurate results. Xenobiotic-labeled competitive ELISA
In this immunoassay, the wells of a microplate contain a fixed amount of adsorbed antibody (Figure 2) . The sample is added to the wells of the microplate followed by a buffered solution containing a fixed amount of xenobiotic labeled with an enzyme (i.e., horseradish peroxidase or alkaline phosphatase). The xenobiotic labeled with horseradish peroxidase or phosphatase and xenobiotic present in the sample compete for binding to the solid-phase antibody (incubation period). Then the wells of the microplate are washed with a suitable buffer (the separation step) to remove all traces of unbound xenobiotic labeled with an enzyme. A substrate reagent (chromagen) is added to the wells and left to develop color; dilute H2O2 with tetramethylbenzidine and p-nitrophenyl phosphate is the substrate of choice for horseradish peroxidase and alkaline phosphatase, respectively. If an enzymelinked fluorescence assay is used, alkaline phosphatase converts the substrate 4-methylumbelliferyl phosphate to 4-methylumbelliferone, which emits light and can be measured by a fluorometer. The rate of oxidation of tetramethylbenzidine to produce a blue color can be monitored at 370 or 630–650 nm. Alternatively, the color development is stopped by adding 1 M dilute sulfuric acid after 20–30 min of incubation. The developed yellow color is stable for at least 1 h and can be measured at 450 nm using a standard microplate reader. The higher the concentration of analyte in the sample, the less xenobiotic labeled with the enzyme will be bound on the wells of the microplate (inversely proportional) and, therefore, the calibration curve has a negative slope. Consequently, the signal is maximum in the absence of the xenobiotic in the sample. Antibody-labeled competitive ELISA
In this ELISA format, the wells of the microplate contain a fixed amount of xenobiotic, the antibody being labeled with the enzyme (Supplementary Figure 2) . During the assay, competition occurs between the xenobiotic in the sample and the immobilized xenobiotic for binding to the enzyme-labeled antibody. A washing step removes supernatant fraction and lets the antibody labeled with enzyme, attached to xenobiotic that coats the wells of the microplate. The same reporter system, as described above, could be used
future science group
Heterogeneous & homogeneous immunoassays for drug analysis
Review
Figure 2 ELISA A A
TMB
+
TMB
TMB
Sulfuric acid 1M
Wash
B B
TMB
Blue dye Microplate well coated by immobilized antibody anti-xenobiotic
Xenobiotic
Xenobiotic labeled with enzyme (horseradish peroxidase or phosphatase)
Yellow dye
Absorbance (450 nm)
+
Sulfuric acid 1M
Wash
..
.. .
Linear signal
* response
Concentration
Figure 2. Xenobiotic-labeled competitive enzyme-linked immunosorbent assay (ELISA) with xenobiotic labeled with the enzyme horseradish peroxidase. Results for (A) negative and (B) positive analysis. TMB: 3,3’, 5,5;-tetramethylbenzidine (chromogenic compound), H2O2 with tetramethylbenzidine.
for xenobiotic-labeled competitive ELISA. Since the amount of labeled-antibody that is bound is inversely related to the analyte concentration in the sample, the calibration curve has a negative slope. Sandwich (non-competitive) ELISA
This immunoassay uses two or more antibodies (Supplementary Figure 3) . Since it requires that the analyte has more than one binding site, it is mostly applied for those of high molecular weight, such as proteins and polysaccharides, and not xenobiotics. An excess of a primary (capture) antibody is adsorbed to the well and incubated with the sample. After washing (to remove the unbound analyte), a second enzymelabeled (detection) antibody is added and allowed to equilibrate. A second wash is included to remove the label excess and the same reporter system, as described above, could be used for xenobiotic-labeled competitive ELISA. The signal produced is proportional to the amount of analyte in the sample (calibration curve with a positive slope). This immunoassay potentially is much more sensitive than competitive ELISAs. Microparticle enzyme immunoassay (MEIA) Figure 3 presents a scheme of MEIA for competitive for-
mat for negative and positive analysis. Beads are coated with capture antibodies against the xenobiotic. A xenobiotic labeled with an enzyme (e.g., alkaline phospha-
future science group
tase) is incubated with the beads and competes with the xenobiotic present in sample. A washing step is included to remove the unbound xenobiotic labeled with enzyme. A suitable substrate (e.g., 4-methylumbelliferone phosphate) is added and the fluorescence of the product (e.g., 4-methylumbelliferone) is measured in a fluorometer. The signal is proportional to analyte concentration in the sample. Sandwich format (with antibody anti-xenobiotic labeled with an enzyme) is also available. It is the most common assay method for determining blood tacrolimus concentration, however, it provides overestimated concentrations in samples with lower hematocrit values (
For reprint orders, please contact [email protected]
Heterogeneous and homogeneous immunoassays for drug analysis
Immunoassays are very useful techniques to perform screening and semi-quantitative analysis of hundreds of different xenobiotics. Small sample volumes are required and pretreatment is usually unnecessary (e.g., homogeneous immunoassays). Fully automated and high-throughput systems are available, which help physicians to take timely decisions. However, immunoassays do suffer from interference from both endogenous and exogenous factors that limit their application in quantitative analysis. These assays use different labels (e.g., colorimetric, fluorescent, chemiluminescent or electrochemiluminescent) and different methods for generating and measuring signals, but the basic principles are usually similar. This review outlines the practical aspects of immunoassays in bioanalysis and describes their application in clinical chemistry for xenobiotic analysis, namely medicines and drugs of abuse.
Ricardo Jorge Dinis-Oliveira IINFACTS - Institute of Research & Advanced Training in Health Sciences & Technologies, Department of Sciences, Advanced Institute of Health Sciences – North (ISCS-N), CESPU, CRL, Gandra / Department of Legal Medicine & Forensic Sciences, Faculty of Medicine, University of Porto, Porto / REQUIMTE, Laboratory of Toxicology, Department of Biological Sciences, Faculty of Pharmacy, University of Porto, Porto, Portugal Tel.: +351 222 073 850 ricardinis@ sapo.pt
Keywords: drugs • heterogeneous format • homogeneous format • immunoassays • interferences
Background Immunoassays are quick analytical techniques that use different antibodies, labels (e.g., colorimetric, fluorescent, chemiluminescent or electrochemiluminescent) and methods for generating and measuring signals related to the presence or concentration of analytes. The first immunoassay (radioimmunoassay) was described by Berson and Yalow in 1959 for insulin quantification [1] . For this discovery they were awarded the Nobel Prize in Medicine in 1977. Since then, in just half a century, immunoassays for hundreds of other analytes have been developed [2] . In clinical and forensic chemistry/toxicology, immunoassays are widely used as a bioanalytical technique to screen biological samples for the presence of a xenobiotic in a variety of situations [3–5] . Indeed, immunoassays for drugs of abuse and therapeutic drug monitoring are routinely used for screening (qualitative) and semi-quantitative analysis, respectively [4,6] . Table 1 summarizes other common immunoassay applications for drug analysis. Particularly in forensic toxicology,
10.4155/BIO.14.208 © 2014 Future Science Ltd
scientific societies [7] have established that all positive results must be confirmed by more specific and sensitive techniques, preferably GC–MS and LC–MS. In other words, immunoassays usually provide a preliminary result. In fact, it is well known that chemically and non-chemically related xenobiotics can sometimes interfere with analysis, leading to false-positive and false-negative results [8,9] and the risk of sample adulteration must be taken into account [10] . Therefore, one may question if there is still a role for immunoassay in clinical and forensic toxicology. The answer to this is definitely yes. Table 2 details some of the most relevant advantages and limitations of immunoassays for drug analysis. Here, the practical aspects of all immunoassays devoted to xenobiotic analysis, namely medicines and drugs of abuse, are reviewed. Articles written in English, German, French, Spanish and Portuguese that detailed immunoassays applied to drug analysis (e.g., amphetamines, benzodiazepines, opioids, hallucinogens,
Bioanalysis (2014) 6(21), 2877–2896
part of
ISSN 1757-6180
2877
Review
Dinis-Oliveira
Key terms Immunoassay: An analytical technique based on antibody–antigen recognition. Xenobiotic: (Of the Greek, xenos ‘stranger’; bios ‘life’). Foreign chemical (e.g., medicines, pesticides and drugs of abuse) found within an organism that is not naturally produced by or expected to be present within that organism. It is opposed to the endobiotic concept. Point-of-care testing: Analyses are performed outside of the traditional central laboratory in close proximity to the patient or other subject.
cannabinoids, cocaine, barbiturates, antiepileptic, tricyclic antidepressants, immunosuppressants, cardioactive compounds, aminoglycosides and vancomycin) were searched for using the national library of medicine’s PubMed MedLine database and the web of knowledge. Classfication of immunoassays Different approaches can be used to classify/characterize immunoassays. Table 3 highlights the most important details for classification, which are further developed below: • Number of processed samples: point-of-care testing (POCT) or automated immunoassays could be used to analyze one or several samples, respectively; • Analytical platforms (e.g., microplate, microcolumn, flow, membrane and bead) using either closed or open designs. The former design can only make use reagents of the platform manufacturer’s. Moreover, existing methods could not be improved and new assays for different analytes to those on
the equipment’s repertoire could not be validated. Furthermore, the scientific spirit of the analyst is very difficult to be applied, and existing errors could not be corrected. An interesting review on this subject can be found in [11] ; • Existence or absence of competition: –– Competitive design uses only one antibody and is mostly applied for analytes with small molecular weight (e.g., xenobiotics). Predetermined amounts of labeled xenobiotic (i.e., hapten) and antibody are added to the specimen followed by incubation. In the basic design, the analyte molecules present in the specimen compete with labeled xenobiotic molecules for the antibody binding sites molecules [12,13] . If the signal is generated when the labeled xenobiotic binds to an antibody molecule, the signal is inversely proportional to analyte concentration in the specimen (e.g., fluorescence polarization immunoassay [FPIA]). On the other hand, if the signal is generated by an unbound labeled xenobiotic, the assay signal is directly proportional to the xenobiotic concentration (e.g., enzyme multiplied immunoassay technique [EMIT]); –– Immunometric or noncompetitive (‘sandwich’) design uses two antibodies that recognize different parts of the analyte, usually of large molecular weight (i.e., to accommodate two different antibody molecules), such as proteins, peptides or polysaccharides. An excess of captured antibodies directed to the analyte is usually immobilized on a solid support (e.g., microparticle bead, a coated tube or,
Table 1. Most important applications of immunoassays for xenobiotic analysis. Applications (urine, sweat, serum, plasma, blood, exhaled air, hair and oral fluid) Evaluate potential drug abuse or overdose in the emergency room Diagnosis and management of pregnant drug abusers Evaluate newborns for in utero drug exposure Ensure abstinence in drug dependency rehabilitation programs Therapeutic drug monitoring Workplace drug testing Pre-employment drug screening Bioequivalence studies in drug discovery and pharmaceutical industries Driving under the influence of psychoactive drugs (including ethanol) Self-monitoring during therapeutic Doping control (e.g., anabolic androgenic steroids)
2878
Bioanalysis (2014) 6(21)
future science group
Heterogeneous & homogeneous immunoassays for drug analysis
more commonly, the surface of a microplate). The specimen is then added and incubation with the antibody is left to occur for a predetermined time [12,13] . After washing, a liquid reagent containing the secondary antibody conjugated with a molecule for signal production (e.g., an enzyme) is added. The secondary antibody binds to the captured analyte forming a ‘sandwich’. A second washing step to remove unbound secondary antibody and the addition of a substrate for signal development follows this. Xenobiotic concentration is directly proportional to the intensity of the signal;
Review
• Outcomes: –– Qualitative (screening) immunoassays only indicates that xenobiotics present in the sample are above a particular cutoff concentration but do not give information about which xenobiotic is present or respective concentrations. These cutoff values can be the lower limit of quantification of the method or a higher value. Laws can also define these cutoffs. A full calibration line is normally unnecessary and, therefore, less complex since only one calibrator at the cutoff concentration is needed. Negative results reflect concentrations below the cutoff and do not
Table 2. Advantages and limitations of immunoassays for xenobiotic analysis. Advantages
Limitations
Relatively easily automated and integrated into the laboratory workflow
Are intrinsically non-specific (not entirely true since it is dependent on appropriate assay development, particularly with respect to antiserum generation) for xenobiotic analysis, particularly with respect to metabolites
Require less technical sophistication on the part of staff (simplicity)
Compared with other bioanalytical techniques, requires a long assay development time (i.e.,’usually some months to generate the reagent antibody, whether monoclonal or polyclonal), with no guarantee of success
Speedy analysis (e.g., results can be reported in 10–30 min) that allow physicians to take timely decisions
Reduced precision
Very sensitive assays with very small sample volumes (10–50 μl)
Not commercially available for all drugs (finite number) currently monitored in clinical practice
There is generally no need for a Interferences from drug metabolites, other drugs with similar sample preparation or extraction step structures, herbal supplements as well as various endogenous prior to analysis (e.g., samples can be factors assayed directly) High-throughput methods
Cannot raise antibodies to some inorganic or very small organic compounds, such as carbon monoxide, cyanide ion, ethanol, ethylene glycol and lithium
Much less complex than LC and GC combined with mass spectrometry
Many drugs (e.g., cardioactives, psychoactives) have active metabolites, which should be measured together
Antibodies directed to a certain class of drugs are useful for screening purposes
Cannot raise selective antibodies to certain drugs (e.g., amiodarone cross reacts with thyroxine)
Are cost-effective since testing can usually performed in existing automated platforms, and laboratories can rapidly introduce new immunoassays if necessary
Typically the kits are approved for certain biological matrices (usually urine and serum or plasma). To be applied to other samples (e.g., solid tissues), extraction or dilution is necessary followed by validation
New compounds are introduced every day in the market and, as Reagents can be stored in the analyzer and they are often stable for such, the absence of antibodies could lead to false-negative results 1–2 months Auto-flagging to alert poor specimen quality (e.g., hemolysis, high bilirubin and lipemic)
future science group
www.future-science.com
2879
Review
Dinis-Oliveira
Table 3. Classification of immunoassays. Immunoassay
Format
Common assay signals
RIA
Heterogeneous
Radioactive counting
ELISA
Heterogeneous/competition or sandwich
Colorimetry
ELFA†
Heterogeneous/competition or sandwich
Fluorescence
MEIA†
Heterogeneous/sandwich
Fluorescence
EMIT†
Homogeneous/competition
Absorbance at 340 nm
CEDIA
Homogeneous/competition
Colorimetry
KIMS / PETINIA
Homogeneous
Turbidimetry
FPIA
Homogeneous
Fluorescence polarization
LOCI
Homogeneous/competition or sandwich
Chemiluminescence
CMIA
Heterogeneous/competition
Chemiluminescence (acridinium ester label)
ECLIA
Heterogeneous/competition or sandwich
Electrochemiluminescence
LIA
Homogeneous or Heterogeneous/ competition or sandwich
Electrochemical (e.g., voltammetric, potentiometric, capacitive, conductometric and impedimetric)
†
†
These are enzyme Immunoassays. CEDIA: Cloned enzyme donor immunoassay; CMIA:Chemiluminescent microparticle immunoassay; ECLIA: Electro-chemiluminescence immunoassay; ELFA: Enzyme-linked fluorescent assay; ELISA: Enzyme-linked immunosorbent assay; EMIT: Enzyme-multiplied immunoassay technique; FPIA: Fluorescence polarization immunoassay; KIMS: Kinetic interaction of microparticles in solution; LIA: Liposome immunoassay; LOCI: Luminescent oxygen channeling immunoassay; MEIA: Microparticle enzyme immunoassay; PETINIA: Particle enhanced turbidimetric inhibition immunoassay; RIA: Radioimmunoassay.
†
exclude the presence of the xenobiotic. Samples close to the cutoff may give negative or positive results depending on the precision of the assay and influence of matrix effects; –– In semi-quantitative immunoassays, the xenobiotic concentration can be obtained by comparing the produced signal with calibrators. There is almost always a non-linear relationship (since the number of antibody binding sites is finite) between the measured response in the assay and the analyte concentration [14] . Since there is no universal mathematical function that fits the results, bias could be introduced. Moreover, greater errors are registered at the extremes (e.g., becoming asymptotic) than in the central linear region of the calibration curve. Polynomial or logit functions ares usually applied. Computer software usually helps to aid the decision of ‘best fit’ according to the obtained standard points (e.g., spline fit) or on a mathematical model that reflects the physical principles underlying immunoassays (e.g., four-parameter logistic fit) [14] . The latter is preferred since it is less subjected to the bias of any inaccurate standard points [14] .
2880
Bioanalysis (2014) 6(21)
• Existence/no existance of a separation step: –– Heterogeneous format – antibody-bound label must be separated from the free label before measuring the signal. Consequently, there is less opportunity for interference in comparison with many homogenous immunoassays [2] . The separation step can be simplified by using 96-well microplates or paramagnetic particles; –– Homogeneous assays – after incubation, a separation step is not required to distinguish the signal produced by antibody-bound and free label. Nevertheless, extraction of the analytes with solvents may be useful before to the immunoassay to increase selectivity. Although easier to perform than heterogeneous immunoassays, the matrix effects may influence the results [2] ; • Signal produced – non-isotopic (e.g., enzyme immunoassay [EIA]) and isotopic (radioimmunoassay [RIA]) immunoassays could be considered. Characteristics of both types are discussed in the following sections.
future science group
Heterogeneous & homogeneous immunoassays for drug analysis
Point-of-care-testing (POCT) Aiming to provide an immediate result, POCT or bedside testing is any analysis performed outside of the traditional central laboratory and in close proximity to the patient or other subject. For xenobiotic analysis, urine, blood, oral fluid or sweat are usually applied to screening in roadside and workplace testing [4,15] . Supplementary Table 1 lists some of the advantages and limitations of POCT. In the majority of the POCT devices, a competitive lateral flow immunoassays format (also known as immunochromatographic assay) is used. In addition, most of the devices use a negative-indicating reaction, meaning that in the absence of a visible test line, the reaction is positive. Briefly, the strips (normally made of nitrocellulose), often enclosed by a plastic test cassette, have a reservoir (well) with labeled antibodies antixenobiotic (e.g., chromogenic, fluorescent, paramagnetic monodisperse latex particle, etc.) into which the sample is pipetted (Figure 1) . By manipulating the concentration of antibodies, POCT can be manufactured to provide the required cut-off values. At an appropriate distance from the origin, the xenobiotic(s) immobilized (in a line format) on the strip will capture any antibody not bound to a xenobiotic. Therefore, a visible line of labeled antibody is produced, the signal being maximal if sample is negative. To assess that the device is working properly (e.g., enough sample volume, quality of
Key terms Heterogeneous format: Antibody-bound label must be separated from the free label before measuring the signal. Interference: It is the alteration of the correct value (accuracy) of the analyte, usually expressed as concentration or activity, due to the presence of a substance in the sample. Homegeneous format: A separation step is not required to distinguish the signal produced by antibody-bound and free label.
reagents), a control line normally exists at an appropriate distance from the immobilized xenobiotic. The excess of free labeled antibody further migrates and will be captured by the anti-immunoglobulin antibody to form a second visible line, which must be always present both in positive and negative analysis. In positive analysis, the first line is not produced since the xenobiotic in the sample will bind all the antibody. If the device uses labeled xenobiotic, and the immobilized antibodies being bound onto the membrane, the positive result is indicated by the appearance of a line. With either type device, faint lines should be read as negative [16] . Modern radioimmunoassay (RIA) presents a scheme of RIA for negative and positive analysis. Typically, 125I is used
Supplementary Figure 1
Control line line Control
Test line
A A
Review
visibleline line 11visible B B
2 visible lines 2 visible lines Xenobiotic Xenobiotic
Chromogeniclabeled labelled Chromogenic antibody antibody anti-xenobiotic anti-xenobiotic
Anti-IgG antibody
Flow Flowdirection direction
Figure 1. Lateral flow immunoassay. (A) The excess of sample xenobiotic binds all antibody preventing line formation. (B) Labeled antibody binds to immobilized xenobiotic giving a visible line.
future science group
www.future-science.com
2881
Review
Dinis-Oliveira as the radiolabel and requires a γ-counter to measure radioactivity in the tubes. Iodination procedures can either be direct (i.e., the xenobiotic is labeled by replacing an atom of hydrogen with 125I), or indirect (i.e., the xenobiotic is labeled with a pre-iodinated molecule). RIA reagents are available for many xenobiotics, cost is reasonable and has very low limits of detection. Typically, tubes are precoated with antibody anti-xenobiotic. Radiolabeled xenobiotic is then added, which binds to antibody anti-xenobiotic. The xenobiotic present in the sample will compete with the radioactive xenobiotic bound to the antibody. Simple decanting easily separates the bound and free fractions, and allows quantification of the bound fraction. The radioactivity signal is relatively free of matrix interference and less susceptible to influence by urine adulteration, when compared with other immunoassays. The major limitations relate to the use of radioactive materials (e.g., need for radioactive materials licensure and waste disposal, unsafe radioactive handling and storage, cost, and the relatively short shelf-life of the reagents, which is usually less than 60 days). Therefore, in recent years, the trend has been away from RIA and toward non-isotopic immunoassays. ELISA Different ELISA formats exist, including direct, indirect, sandwich and competitive. Unfortunately, these terms may be misleading as the same method may be described by more than one term [16] . Figures 2, Supplementary Figures 2 & 3 present the scheme of the most common ELISA formats for negative and positive analysis. The following topics are highlighted: • It is the most widely used immunoassay; • They are more laborious than homogeneous EIAs but are already semi-automated (i.e., pipetting and washing); • These assays are typically performed in 96-well microplates and thus are not compatible with highthroughput clinical chemistry analyzers; • Since it is heterogeneous, less matrix interferences are registered, especially if a double wash is performed; • Costs are somewhat higher than homogeneous EIA reagents; • It can be used with whole blood, typically diluted with water or buffer, and with proper validation, even for tissue homogenates; • Although less than RIA, ELISA has a good sensitivity and dynamic range, especially if a fluorescent signal is used (limit of detection of ~1–5 pg/well
2882
Bioanalysis (2014) 6(21)
compared with ~5–200 pg/well in clorimetric format) [17] ; • Steps are often long and time precision for addition of substrate and stop reagents is required to obtain accurate results. Xenobiotic-labeled competitive ELISA
In this immunoassay, the wells of a microplate contain a fixed amount of adsorbed antibody (Figure 2) . The sample is added to the wells of the microplate followed by a buffered solution containing a fixed amount of xenobiotic labeled with an enzyme (i.e., horseradish peroxidase or alkaline phosphatase). The xenobiotic labeled with horseradish peroxidase or phosphatase and xenobiotic present in the sample compete for binding to the solid-phase antibody (incubation period). Then the wells of the microplate are washed with a suitable buffer (the separation step) to remove all traces of unbound xenobiotic labeled with an enzyme. A substrate reagent (chromagen) is added to the wells and left to develop color; dilute H2O2 with tetramethylbenzidine and p-nitrophenyl phosphate is the substrate of choice for horseradish peroxidase and alkaline phosphatase, respectively. If an enzymelinked fluorescence assay is used, alkaline phosphatase converts the substrate 4-methylumbelliferyl phosphate to 4-methylumbelliferone, which emits light and can be measured by a fluorometer. The rate of oxidation of tetramethylbenzidine to produce a blue color can be monitored at 370 or 630–650 nm. Alternatively, the color development is stopped by adding 1 M dilute sulfuric acid after 20–30 min of incubation. The developed yellow color is stable for at least 1 h and can be measured at 450 nm using a standard microplate reader. The higher the concentration of analyte in the sample, the less xenobiotic labeled with the enzyme will be bound on the wells of the microplate (inversely proportional) and, therefore, the calibration curve has a negative slope. Consequently, the signal is maximum in the absence of the xenobiotic in the sample. Antibody-labeled competitive ELISA
In this ELISA format, the wells of the microplate contain a fixed amount of xenobiotic, the antibody being labeled with the enzyme (Supplementary Figure 2) . During the assay, competition occurs between the xenobiotic in the sample and the immobilized xenobiotic for binding to the enzyme-labeled antibody. A washing step removes supernatant fraction and lets the antibody labeled with enzyme, attached to xenobiotic that coats the wells of the microplate. The same reporter system, as described above, could be used
future science group
Heterogeneous & homogeneous immunoassays for drug analysis
Review
Figure 2 ELISA A A
TMB
+
TMB
TMB
Sulfuric acid 1M
Wash
B B
TMB
Blue dye Microplate well coated by immobilized antibody anti-xenobiotic
Xenobiotic
Xenobiotic labeled with enzyme (horseradish peroxidase or phosphatase)
Yellow dye
Absorbance (450 nm)
+
Sulfuric acid 1M
Wash
..
.. .
Linear signal
* response
Concentration
Figure 2. Xenobiotic-labeled competitive enzyme-linked immunosorbent assay (ELISA) with xenobiotic labeled with the enzyme horseradish peroxidase. Results for (A) negative and (B) positive analysis. TMB: 3,3’, 5,5;-tetramethylbenzidine (chromogenic compound), H2O2 with tetramethylbenzidine.
for xenobiotic-labeled competitive ELISA. Since the amount of labeled-antibody that is bound is inversely related to the analyte concentration in the sample, the calibration curve has a negative slope. Sandwich (non-competitive) ELISA
This immunoassay uses two or more antibodies (Supplementary Figure 3) . Since it requires that the analyte has more than one binding site, it is mostly applied for those of high molecular weight, such as proteins and polysaccharides, and not xenobiotics. An excess of a primary (capture) antibody is adsorbed to the well and incubated with the sample. After washing (to remove the unbound analyte), a second enzymelabeled (detection) antibody is added and allowed to equilibrate. A second wash is included to remove the label excess and the same reporter system, as described above, could be used for xenobiotic-labeled competitive ELISA. The signal produced is proportional to the amount of analyte in the sample (calibration curve with a positive slope). This immunoassay potentially is much more sensitive than competitive ELISAs. Microparticle enzyme immunoassay (MEIA) Figure 3 presents a scheme of MEIA for competitive for-
mat for negative and positive analysis. Beads are coated with capture antibodies against the xenobiotic. A xenobiotic labeled with an enzyme (e.g., alkaline phospha-
future science group
tase) is incubated with the beads and competes with the xenobiotic present in sample. A washing step is included to remove the unbound xenobiotic labeled with enzyme. A suitable substrate (e.g., 4-methylumbelliferone phosphate) is added and the fluorescence of the product (e.g., 4-methylumbelliferone) is measured in a fluorometer. The signal is proportional to analyte concentration in the sample. Sandwich format (with antibody anti-xenobiotic labeled with an enzyme) is also available. It is the most common assay method for determining blood tacrolimus concentration, however, it provides overestimated concentrations in samples with lower hematocrit values (
