- Email: [email protected]
Chemical analysis of marker substances
Chemical Analysis of Marker Substances Roger P. Maickel Department of Pharmacology and Toxicology, Purdue University, West Lafayette, Indiana
ABSTRACT: The development of chemical assay procedures for potential marker substances will depend on, and be determined by, three basic factors: (1) the nature of the substance used, (2) the characteristics of the biological samples obtained, and (3) the varieties of information required. Once these have been delineated, the analytical methodology can be specifically characterized in terms of specificity and sensitivity; accuracy, precision, and reliability; and ease, cost, and safety. The specific details of a method must be based on the three factors mentioned above. Thus, a procedure having an ultimate sensitivity limit of more than a micromole per milliliter will be useless if the maximum expected marker concentration in the sample is less than a micromole per milliliter. Similarly, a procedure utilizing gas-liquid chromatography mass spectrometry costing many thousands of dollars is overkill if only a simple, inexpensive thin-layer chromatography procedure is needed to merely confirm the presence of the marker in a sample of urine or blood. Finally, the potential utility of the marker substance will be constrained by nonscientific factors such as cost, time, and convenience. These factors may depend on the ultimate applications of marker substance use, but they must also be addressed in any assay development program. The overall characteristics of a chemical assay procedure for marker substances will be discussed in terms of a model based on similar procedures for the chemical assay of foreign organic compounds in biological materials.
Analytical chemistry has been developing rapidly over the past few decades. Dramatic advances have been m a d e in the development of techniques and technologies for spectrophotofluorometry, gas-liquid chromatography (GLC), mass spectrometry (MS) and gas chromatography to mass spectrometry (GC/MS), high-performance liquid chromatography (HPLC), and other procedures. Nevertheless, most of the dramatic advances have been in instrumentation and engineering and not in basic analytical methodology. For example, the entire analytical technology involving HPLC is based on detector systems using absorption spectrometry, fluorometry, or polarography--techniques that are 50 years old. W h e n one considers the problems inherent in the accurate and precise determination of quantities of exogenous chemicals in biological materials or biological samples, it becomes apparent that the limiting factors are far more complex than the instrumentation involved. Some readily identified problems
Address reprint requests to: Roger P. Maickel, Ph.D., Department of Pharmacology and Toxicology, School of Pharmacy and Pharmacal Sciences, Purdue University, West Lafayette, Indiana 47907. Controlled Clinical Trials 5:556-567 (1984) © Elsevier Science Publishing Co., Inc. 1984 52 Vanderbilt Ave., New York, New York 10017
556 0197-2456/84/$03.00
557
Chemical Analysis Of Marker Substances i
I SAMPLE COLLECTION
1
/ QUANTITATIVE MEASUREMENT i
QUALITATIVE CONFIRMATION(?)
I
INTERPRETATION &
APPLICATION FIGURE i ,
are presented in Figure 1. Although this entire approach may seem oversimplistic, its purpose is to emphasize the problems that must be recognized. The first step, sample collection, is not as simple as might be seen. An initial question is whether the sample can be obtained by noninvasive methods (feces, respired air, urine, sputum) or by an invasive procedure (blood). Two subfacets must be considered: (1) the sample size required for assay and the practicality of obtaining such a sample and (2) the pharmacokinetic factors that may restrict the eventual interpretation of the data obtained. The second step, qualitative confirmation and quantitative measurement, may go hand-in-hand by virtue of the nature of the assay procedure or may demand separate or even independent technology. The former situation is often characteristic of procedures such as GLC (combining a specific detector and a specific retention time under given column/instrument conditions), GC/MS, or RIA. The latter may be exemplified by combinations of thin-layer chromatography (TLC) (for confirmation of identity) and absorption spectrometry (for quantitation). The final step, interpretation and application, requires the integration of the numerical data obtained with known factors of the marker substance and the overall system involved, that is, dosage, distribution and metabolism,
558
Roger P. Maickel and pharmacokinetic data. The net result should be an estimation (with an inherent degree of confidence) of patient compliance. Although chemical analysis of marker substances requires that both general and specific aspects of the various portions of Figure 1 be addressed, the identification and measurement aspects of potential markers will be emphasized.
SAMPLE COLLECTION Sample collection is best considered as a conglomerate of a number of subitems. Some of these subitems are equally relevant to physical and immunoassay tests and can be listed as (1) sample quantitation and (2) sample stability and storage. Prior to applying technology for quantitative measurement of the marker, the sample must also be quantitatively defined. The simplest method is to utilize a defined aliquot of the total sample collected; a less satisfactory but acceptable alternative is to utilize a stoichiometric transfer of an entire, premeasured sample. If the sample is liquid or gaseous, appropriate care must be taken to ensure that leakage or evaporative losses do not occur in this step. Samples must be stored under optimal conditions to maintain stability between the time of collection and the time of analysis. The ideal situation would be to have virtually no time elapse between sample collection and analysis; in the absence of this, storage situations must be characterized and sample stability maintained. Depending on the nature of the marker substance, this may require low temperature, control of pH, inhibition of enzymes, removal of oxygen and/or metallic ions, protein precipitation, or other specific treatment. Not only should degradation of the marker substance be prevented, but production of any interfering substances (from endogenous biological materials) should be retarded or prevented.
QUALITATIVE CONFIRMATION For any procedure used to determine a marker substance, the analytical process will have to confirm that the substance being assayed is, in fact, what it is. Although this may seem facetious, it is necessary to confirm the identity of the substance being quantitatively determined. This confirmation may often be an inherent part of the quantitative analytical methodology (Rr, retention time, absorption maxima/minima, activation/fluorescence wavelengths) or may be an independent step. Nevertheless, the validity of any marker assay will depend on the ability of the overall procedure to measure only the marker substance itself.
QUANTITATIVE MEASUREMENT The measurement system used to determine the amount of a marker substance in a biological sample can be a single instrument, procedure, or tech-
Chemical Analysis of Marker Substances
559
nology, or a combination of several instruments, procedures, or technologies. One can assume that an electronic instrument will be employed to convert a physical measurement into numerical data. The constraints are simple: the process must be accurate and precise and have a suitable measuring sensitivity.
INTERPRETATION AND APPLICATION The final step in the overall procedure is a combination of interpreting the data obtained from the analysis and applying it to the situation. In the case of any proposed marker system, the question is whether the data support the conclusion that there was satisfactory patient adherence to the therapeutic regimen. Optimally, the data should answer this question with a qualitative yes or no and indicate the qualitative deviation from perfect compliance. Although such an answer is, in part, dependent upon nonanalytical factors such as pharmacokinetics, it is crucially dependent upon the analytical data obtained and the inherent limits such data place upon any subsequent interpretations and conclusions.
CHARACTERISTICS OF ACCEPTABLE PROCEDURES The characteristics of an acceptable assay procedure may be summarized in terms of specificity, sensitivity, speed, simplicity, reliability, economy, and safety. A clear understanding of each characteristic will enable anyone to design a method appropriate for a given marker or, conversely, to define the optimal characteristics of a substance to be used as a marker. Each characteristic is independently important; a failure to satisfy the demands of any characteristic can seriously compromise the overall procedure.
Specificity To determine the concentration of a marker in a biological sample, t h e marker must be differentiated from other substances in the sample. Specificity must be independent of variations in the composition of the sample and must be capable of determining the marker accurately, even in the presence of very high concentrations of other substances. A number of situations may confer specificity upon an analytical method: 1. The measurement technology may be specific, as in spectrophotofluorometry, which requires a precise pairing of activation and emission wavelengths. 2. A chemical reaction may be performed prior to the final measurement step, with the reaction process conferring specificity, as in the o-phthalaldehyde reaction with serotonin.
Roger P. Maickel
560
3. One or more physical separations may be performed, thereby separating substances by virtue of differing physiochemical properties, as in the numerous variants of chromatography. Most assay procedures utilize some combination of these situations. The crucial point is that the final measurement process must determine only the substance of interest or must correct for any interfering substances.
Sensitivity An absolute definition of sensitivity is impossible. A reasonable working definition is that the analytical procedure should be capable of accurately measuring the marker substance in a biological sample at a concentration at least one order of magnitude lower than the expected concentrations. In addition, the numerical response of the measuring system at this limit of sensitivity should be a value at least double that of the "blank" value obtained from an identical biological sample containing none of the compound of interest. Presumably, a marker substance would be present in the body at a concentration of 10-9 to 10-3 M; thus, the ultimate sensitivity level should be in the lower end of this range.
Speed Time is a critical factor in any analytical procedure. For a marker substance, one may project a need for an assay that could be completed in the 30-90minute range. An assay requiring 2-3 days to complete would be cumbersome and might not be cost effective. A relevant point is whether samples would be assayed singly or in groups.
Simplicity The less complicated a method is, the fewer opportunities for error, the shorter the time, and the lower the cost. A procedure should be foolproof and, if possible, idiotproof. The ideal method is one that can be performed by someone with minimal training and should be of minimal complexity. It must be emphasized, however, that specificity and reliability should not be compromised.
Reliability This term covers two aspects of analytical methodology. Reproducibility from sample to sample, from day to day, and from laboratory to laboratory should be sufficient to ensure an error of less than + 5%. Replicate analyses of the same sample, on the same day, and in the same laboratory should also yield values with an error of less than + 5%. A high degree of reliability is particularly important, since a marker may be used to standardize comparison of different dosage forms or therapeutic regimens.
Chemical Analysis of Marker Substances
561
Economy Cost factors in the health care system cannot be ignored. A single analysis that is exhorbitant in cost will severely restrict the usefulness of the marker substance. Cost factors to be stringently examined include labor, instrumentation, and materials. The usefulness of a marker substance may ultimately depend on the fiscal attractiveness of the assay procedure demanded. Safety This aspect of analytical methodology is often ignored, sometimes with disastrous consequences. One need only cite the hazards of using perchloric acid as a protein precipitant if a subsequent evaporation to dryness is performed without removal of the perchloric acid as insoluble potassium perchlorate. SEPARATION A N D MEASUREMENT--BASIC CONCEPTS Virtually every analytical procedure can be divided into two principles or procedures: separation and measurement. Depending on the actual procedure involved, these may be relatively physically independent (as in the case of a paper chromatographic separation followed by elution and a spectrophotometric measurement), or they may be a single procedure (as in GLC with a specific detector). For the purposes of this presentation, the two concepts are considered independently.
Separation Procedures It is uncommon to find that an exogenous substance, such as a marker, can be determined in a biological sample directly, that is, without the application of some sort of procedure involving a physical separation. For most exogenous compounds, such procedures serve two purposes. First, they separate the desired compound from nonspecific gross interferences of a physical nature, such as macromolecules (i.e., proteins), that might otherwise create turbidity, plug a capillary tube, or adhere to a surface. Secondly, they separate the desired compound from more specific interfering substances that, because of chemical similarities, may react in derivatization procedures or yield positive responses in measurement devices or systems. A list of common separation procedures and some details of their characteristics are presented in Table 1. Precipitation may be used to separate proteins and other macromolecules from a biological sample such as blood plasma or serum, thereby permitting direct assay of a marker in the filtrate/supernatant liquid. A number of procedures are available using substances such as trichloracetic, metaphosphoric, or perchloric acids or neutral salts such as zinc hydroxide or barium sulfate. Alternatively, one may selectively precipitate out the compound of interest (by the use of a reineckate salt) and then redissolve it after filtration to remove soluble interfering materials.
562
Roger P. Maickel Table 1
Comparative Characteristics of Separation Procedures Processing time/ sample (manOperator training Procedure hours) Cost/sample Initial cost requirements Precipitation 0.5 $2.50 $250 minimal Liquid-liquid extraction 0.5 $2.50 $500 minimal Column chromatography 1-3 $5.00 $500 minimal Paper chromatography 6 $5.00 $500 minimal Thin-layer chromatography 1-2 $5.00 $500-2000 minimal Gas chromatography (GC)° 0.5 $5.00 $4000-9000 B.S. Liquid chromatography (HPLC)a 0.5-1 $10.00 $1000-4000 B.S. °GC and HPLC used in separation (nonquantitative) mode only.
Liquid-liquid extraction may accomplish the separation of a desired substance from interfering materials by applying the principles of partition between two immiscible liquid phases. This system, which forms the basis for all chromatographic separation procedures, depends on the Nernst distribution law, which says that the sum of the fractions of a solute in two immiscible solvent phases equals the total solute present. Mathematically, this can be expressed as:
K = f/f2
and
fl +f2 = 1.
Each partition that occurs may be considered to have as a basis the characteristic coefficient (K), which is determined by the solute and solvents involved. Repetitive partitioning is known as countercurrent distribution. The general procedures for developing a liquid-liquid extraction process were described some 35 years ago by Brodie et al. [1]. The organic solvent selected should be the one that is least polar, yet will extract the greatest fraction of the marker compound from the (aqueous) biological sample while removing the least amount of interfering substances. The relative polarity of solvents can be estimated from properties such as the dielectric constant, as reported by Craig and Craig [2]. Column chromatography procedures have the disadvantages of limited resolution and lengthy development time. Column packings include a variety of alternatives: alumina, cellulose, ion-exchange resins, silica gel, and others. Development utilizes frontal displacement or elution analysis. The applications of procedures of column chromatography to marker assays will probably be limited by the nature of the marker substance, the concentration in a sample, and the need for rapid processing. The general background of column
Chemical Analysis of Marker Substances
563
chromatography may be found in a number of references [3]; an excellent overview of application of the technique to exogenous organic compounds can be found in Hirtz [4]. Paper chromatographycan also be used to separate compounds, utilizing the principles of repetitive countercurrent distributions between the "bound" water in the paper sheets and appropriate organic solvent systems. The water phase may be replaced by nonpolar compounds or may be adjusted in pH. Many variations on them exist [3]. One serious constraint of paper chromatographic separations is that they are often slow, time-consuming processes and may require additional technology after development to prepare the marker substance for quantitation. Thin-layer chromatographyconsists of a layer of sorbent (separatory material) such as alumina, cellulose, or silica gel on a solid support such as a sheet of glass, aluminum, or a polymeric material. The separation process is accomplished as an appropriate solvent (mobile) phase ascends the sorbent phase by capillarity. The basic principles and practice of TLC are thoroughly described in several sources [5,6]. As a separation technique, TLC is rapid, inexpensive, and widely applicable. Quantitation may be difficult, but a number of procedures are available to facilitate stoichiometric application of samples and quantitative measurements. Technology is also available to enhance separation and speed up development [7]. Gas-liquid chromatography is a relatively recent addition to the armamentarium of the analytical scientist. In this technique, the components of a mixture are separated while being carried (in a stream of inert gas) through a column packed with a stationary phase consisting of some solid material (such as fire brick) generally coated with a high-boiling inert liquid. There are many descriptive reviews of GLC theory, applications, and practice [8-11]. In general, the technique is rapid and has few limitations. The most likely problem with a marker substance might be the relative degree of volatility. One potential advantage of GLC procedures for the analysis of marker substances lies in the diversity of available detector systems. These units, appended at the outflow end of the column, permit separation and measurement to be combined into a single instrumental system. From the early thermal conductivity detectors has come a steady stream of developments: flame ionization, electron capture, and specific detectors for nitrogen, halogens, sulfur, or phosphorous. The ultimate in GLC detection systems is the combination of a gas chromatograph and a mass spectrometer (GC/MS), often coupled with a computer. Unfortunately, the degree of sophistication of such systems results in an extremely high cost of acquisition and maintenance and demands a high level of operator competence. High-performance liquid chromatographyutilizes small bore columns containing appropriate packing materials, with or without coating, and high pressure flow of a mobile phase. Various combinations of stationary and mobile phases can be used to achieve the desired polarity/partition characteristics; the principles of such techniques appear in a number of reviews [!2-14]. A number o f detector systems are available, including ultraviolet absorption, fluorescence, and electrochemical.
564
Roger P. Maickel The technology of HPLC permits relatively rapid processing of samples; the combination of separation and measurement (as in GC) may well be both time and cost effective.
Measurement Procedures Once the marker substance has been "separated" from other substances in a biological sample, a number of technologies can be applied to quantitatively determine the level of the marker substance per unit of sample. As with the prior discussion of separation procedures, only a brief set of comments are made about each of the measurement procedures listed in Table 2. Ultraviolet (UV) absorption spectrometry as a measurement technique requires that the marker substance, or a derivative thereof, have the property of absorbing light in the wavelength range of 200-400 nm. The basic principles and procedures have been reviewed many times; some particularly relevant references may be cited [15-17]. This measurement technique is rapid and simple; nevertheless, limitations on sensitivity and specificity are severe and often demand considerable prework. Systems employing UV detectors are commonly associated with column chromatography of HPLC technologies to permit combined separafion-measurement procedures. Visible (VIS) absorption spectrometry is basically similar to UV technology, with the exception that the range of wavelengths for absorption is 400-750 nm; the basic principles are similar [1,16,17]. In the case of marker substances, the presence of a chromophoric group would make application of VIS methodology inexpensive and simple, especially if the chromophore is of sufficient sensitivity to permit direct measurement, thereby bypassing or simplifying the separation step(s). Infrared absorption spectrometry utilizes the ability of molecules to absorb light at wavelengths >800 n m because of vibrational or rotational energies. This technology has significant potential, although specificity and sensitivity may be critical limitations [18,19]. Spectrophotofluorometry has good criteria for specificity and sensitivity, since it involves both specific absorption and emission wavelengths and has a high quantum yield. Indeed, fluorescence detection and measurement procedures can be utilized with a variety of separation techniques, including TLC or HPLC. The basic principles and a wide variety of practical applications can be seen in a number of reports [20-23]. Many compounds that do not possess fluorescent characteristics per se may be readily converted to a fluorescent compound by chemical reactions or derivatization. Radioisotope derivatization is a technology in which a radiolabeled reagent is chemically "attached" to the substance (such as a marker) to be determined. These procedures can often be carefully tailored to fit a specific compound, although, in doing so, the premeasurement work-up time and cost may be significantly increased. The procedures and principles involved are basically those of organic chemistry [24]. The application of similar technology utilizing radioisotopes in combination with immunoassays is discussed elsewhere in this issue.
moderate to excellent considerable
1.0
2.0-4.0
Radioisotope derivatization
1.0-2.0
excellent
moderate
fair-good
moderate to good considerable
considerable
moderate to fair-good considerable
fair
good
excellent
good
excellent
excellent
10-9 g/ml
10-~° g/ml
good
good
10-1'~/~,1 good
10-12g/ml
10:~ g/ml
10-~ g/ml
10~ g/ml
10-6 g/ml
$17,000
0.25
B.S./M.S.
1.0
$ 7,000
$10,000
1.0 B.S./M.S.
B.S./M.S.
$75,000
2.0
Ph.D.
$20,000
$10,000
0.50
B.S./M.S.
2.0
$ 4,000
0.25
B,S.
B.S./M.S.
$ 7,000
0.25
B.S.
°GC and HPLC are included since use of appropriate ctetector systems permits separation and measurement.
Liquid Chromatography (HPLC)°
Mass spectrometry (GC/MS) 1.0-4.0 Gas chromatography (GC)" 1.0-2.0
considerable
2.0
fair-good
moderate
1.0
fair-good
moderate
1.0
$6O
$40
$80
$70
$40
$50
$30
$30
Estimated cost Limit(s) of Operator Time/assay Specificity sensitivity Reliability expertise man-hours To set-up lab Per assay
UV absorption spectrometry VIS absorption spectrometry IR absorption spectrometry Spectrophotofluorometry
Prework Man-hours Complexity
Comparative Characteristics of Overall Analytical Procedures
Measurement process
Table 2
O1 rjn
566
Roger P. Maickel
Mass spectrometry (MS) is the most recent development in terms of measurement technology. When an MS system is appended to the output of a GC system and combined with an appropriate computerized data processing package, the resultant overall system has extremely high ratings for sensitivity and specificity. Unfortunately, the initial and operating costs of such a system are extremely high, and the expertise required for effective operation is considerable. The potential applications of MS technology to the analysis of marker substances may be seen in a number of reports [25-27]. SUMMARY A N D OVERVIEW
The application of chemical methods of analysis to determine potential marker substances is a complex situation. Several questions must be considered to determine the choice of an appropriate procedure: What is the chemical nature of the marker substance? What type(s) of biological materials will constitute the samples to be analyzed? Will qualitative or quantitative analyses be required? What type of specificity will be needed? What level of sensitivity will be demanded? What constraints of time can be defined? What are reasonable limits for cost per sample or per analysis? The answers to these questions should enable one to develop an appropriate chemical analytical procedure for any potential marker substance.
REFERENCES 1. Brodie BB, Udenfriend S, Baer JE: The estimation of basic organic compounds in biological material. I. General principles. J Biol Chem 168:229-240, 1947 2. Craig LC, Craig D: Extraction and distribution. In: Technique of Organic Chemistry, 2nd ed, Weissburger A, Ed. New York: Interscience, 1956, vol 3, p 172 3. Heftmann E (Ed): Chromatography, A Laboratory Handbook of Chromatographic • and Electrophoretic Methods, 3rd ed. New York: Van Nostrand Reinhold, 1976 4. Hirtz JL: Analytical Metabolic Chemistry of Drugs. New York: Marcel Dekker, 1971 5. Stahl E: Thin-layer Chromatography--A Laboratory Handbook, 2nd ed. New York: Springer-Verlag, 1969 6. Butler TJ: Thin-layer chromatography. In: Guidelines for Analytical Toxicology Programs, Thoma JJ, Bondo PB, Sunshine I, Eds. Cleveland: CRC Press, 1977, vol I, pp 217-251 7. Jupille TH, Perry JA: Programmed multiple development. Spot behavior during solvent advance: A mathematical treatment. J Chromatogr 99:231-239, 1974 8. Vanden Heuvel WJA, Zacchei AG: Gas-liquid chromatography. In: Drug Fate and Metabolism, Garrett ER, Hirtz JL, Eds. New York: Marcel Dekker, 1978, vol 2, pp 49-142 9. Schupp OE: Gas Chromatography. New York: Interscience, 1968 10. Anders MW: Gas chromatography. In: Handbook of Experimental Pharmacology. Brodie BB, Gillette JR, Eds. New York: Springer-Verlag, 1971, vol 28, part 2, pp 63-87
Chemical Analysis of Marker Substances
567
11. Gudzinowicz BJ: Gas Chromatographic Analysis of Drugs and Pesticides. New York: Marcel Dekker, 1967 12. Perry SG, Amos R, Brewer PI: Practical Liquid Chromatography. New York: Plenum, 1972 13. Kirkland JJ (Ed): Modern Practice of Liquid Chromatography. New York: Wiley, 1971 14. Done JN, Knox JH, Loheac J: Applications of High Speed Liquid Chromatography. New York: Wiley, 1974 15. Josephson ES, Udenfriend S, Brodie BB: The estimation of basic organic compounds in biological material. 6. Estimation by ultraviolet spectrophotometry. J Biol Chem 168:341-349, 1947 16. Maickel RP, Bosin TR: Absorption spectrometry. In: Handbook of Experimental Pharmacology, Brodie BB, Gillette JR, Eds. New York: Springer-Verlag, 1971, vol 28, part 2, pp 9-18 17. Cohen SL, Bondo PB: Ultraviolet-visible spectrophotometry. In: Guidelines for Analytical Toxicology Programs, Thoma JJ, Bondo PB, Sunshine I, Eds. Cleveland: CRC Press, 1977, vol II, pp 117-152 18. Kendall DN (Ed): Applied Infrared Spectroscopy. New York: Reinhold, 1966 19. Fales HM: Isolation and identification procedures--spectral methods. In: Fundamentals of Drug Disposition and Drug Metabolism, La Du BN, Mandel HG, Way EL, Eds. Baltimore: Williams & Wilkins, 1971, pp 437-457 20. Ackerman HS, Udenfriend S: Fluorometry. In: Handbook of Experimental Pharmacology, Brodie BB, Gillette JR, Eds. New York: Springer-Verlag, 1971, vol 28, part 2, pp 21-40 21. Udenfriend S: Fluorescence Assay in Biology and Medicine. New York: Academic, 1969 22. Terhaar DA, Porro TJ: Fluorescence spectrophotometry. In: Guidelines for Analytical Toxicology Programs, Thoma JJ, Bondo PB, Sunshine I, Eds. Cleveland: CRC Press, 1977, vol II, pp 153-170 23. Schulman SG, Naik DV: Fluorescence spectroscopy. In: Drug Fate and Metabolism, Garrett ER, Hirtz JL, Eds. New York: Marcel Dekker, 1978, vol 2, pp 195-256 24. Kuntzman RH, Cox RH, Maickel RP: Radioactive techniques: Radioactive isotope derivatives of non-labeled drugs. In: Handbook of Experimental Pharmacology, Brodie BB, Gillette, JR, Eds. New York: Springer-Verlag, 1971, vol 28, part 2, pp 58-62 25. Guarino AM, Fales HM: Gas chromatography-mass spectrometry. In: Handbook of Experimental Pharmacology, Brodie BB, Gillette JR, Eds. New York: Springer -Verlag, 1971, vol 28, part 2, pp 178-194 26. Jenden DE, Cho, AK: Applications of integrated gas chromatography/mass spectrometry in pharmacology and toxicology. Annu Rev Pharmacol Toxicol 13:371-393, 1973 27. Foltz RL: Mass spectrometry. In: Guidelines for Analytical Toxicology Programs, Thoma JJ, Bondo PB, Sunshine I, Eds. Cleveland: CRC Press, 1977, vol II, pp 89-116
ABSTRACT: The development of chemical assay procedures for potential marker substances will depend on, and be determined by, three basic factors: (1) the nature of the substance used, (2) the characteristics of the biological samples obtained, and (3) the varieties of information required. Once these have been delineated, the analytical methodology can be specifically characterized in terms of specificity and sensitivity; accuracy, precision, and reliability; and ease, cost, and safety. The specific details of a method must be based on the three factors mentioned above. Thus, a procedure having an ultimate sensitivity limit of more than a micromole per milliliter will be useless if the maximum expected marker concentration in the sample is less than a micromole per milliliter. Similarly, a procedure utilizing gas-liquid chromatography mass spectrometry costing many thousands of dollars is overkill if only a simple, inexpensive thin-layer chromatography procedure is needed to merely confirm the presence of the marker in a sample of urine or blood. Finally, the potential utility of the marker substance will be constrained by nonscientific factors such as cost, time, and convenience. These factors may depend on the ultimate applications of marker substance use, but they must also be addressed in any assay development program. The overall characteristics of a chemical assay procedure for marker substances will be discussed in terms of a model based on similar procedures for the chemical assay of foreign organic compounds in biological materials.
Analytical chemistry has been developing rapidly over the past few decades. Dramatic advances have been m a d e in the development of techniques and technologies for spectrophotofluorometry, gas-liquid chromatography (GLC), mass spectrometry (MS) and gas chromatography to mass spectrometry (GC/MS), high-performance liquid chromatography (HPLC), and other procedures. Nevertheless, most of the dramatic advances have been in instrumentation and engineering and not in basic analytical methodology. For example, the entire analytical technology involving HPLC is based on detector systems using absorption spectrometry, fluorometry, or polarography--techniques that are 50 years old. W h e n one considers the problems inherent in the accurate and precise determination of quantities of exogenous chemicals in biological materials or biological samples, it becomes apparent that the limiting factors are far more complex than the instrumentation involved. Some readily identified problems
Address reprint requests to: Roger P. Maickel, Ph.D., Department of Pharmacology and Toxicology, School of Pharmacy and Pharmacal Sciences, Purdue University, West Lafayette, Indiana 47907. Controlled Clinical Trials 5:556-567 (1984) © Elsevier Science Publishing Co., Inc. 1984 52 Vanderbilt Ave., New York, New York 10017
556 0197-2456/84/$03.00
557
Chemical Analysis Of Marker Substances i
I SAMPLE COLLECTION
1
/ QUANTITATIVE MEASUREMENT i
QUALITATIVE CONFIRMATION(?)
I
INTERPRETATION &
APPLICATION FIGURE i ,
are presented in Figure 1. Although this entire approach may seem oversimplistic, its purpose is to emphasize the problems that must be recognized. The first step, sample collection, is not as simple as might be seen. An initial question is whether the sample can be obtained by noninvasive methods (feces, respired air, urine, sputum) or by an invasive procedure (blood). Two subfacets must be considered: (1) the sample size required for assay and the practicality of obtaining such a sample and (2) the pharmacokinetic factors that may restrict the eventual interpretation of the data obtained. The second step, qualitative confirmation and quantitative measurement, may go hand-in-hand by virtue of the nature of the assay procedure or may demand separate or even independent technology. The former situation is often characteristic of procedures such as GLC (combining a specific detector and a specific retention time under given column/instrument conditions), GC/MS, or RIA. The latter may be exemplified by combinations of thin-layer chromatography (TLC) (for confirmation of identity) and absorption spectrometry (for quantitation). The final step, interpretation and application, requires the integration of the numerical data obtained with known factors of the marker substance and the overall system involved, that is, dosage, distribution and metabolism,
558
Roger P. Maickel and pharmacokinetic data. The net result should be an estimation (with an inherent degree of confidence) of patient compliance. Although chemical analysis of marker substances requires that both general and specific aspects of the various portions of Figure 1 be addressed, the identification and measurement aspects of potential markers will be emphasized.
SAMPLE COLLECTION Sample collection is best considered as a conglomerate of a number of subitems. Some of these subitems are equally relevant to physical and immunoassay tests and can be listed as (1) sample quantitation and (2) sample stability and storage. Prior to applying technology for quantitative measurement of the marker, the sample must also be quantitatively defined. The simplest method is to utilize a defined aliquot of the total sample collected; a less satisfactory but acceptable alternative is to utilize a stoichiometric transfer of an entire, premeasured sample. If the sample is liquid or gaseous, appropriate care must be taken to ensure that leakage or evaporative losses do not occur in this step. Samples must be stored under optimal conditions to maintain stability between the time of collection and the time of analysis. The ideal situation would be to have virtually no time elapse between sample collection and analysis; in the absence of this, storage situations must be characterized and sample stability maintained. Depending on the nature of the marker substance, this may require low temperature, control of pH, inhibition of enzymes, removal of oxygen and/or metallic ions, protein precipitation, or other specific treatment. Not only should degradation of the marker substance be prevented, but production of any interfering substances (from endogenous biological materials) should be retarded or prevented.
QUALITATIVE CONFIRMATION For any procedure used to determine a marker substance, the analytical process will have to confirm that the substance being assayed is, in fact, what it is. Although this may seem facetious, it is necessary to confirm the identity of the substance being quantitatively determined. This confirmation may often be an inherent part of the quantitative analytical methodology (Rr, retention time, absorption maxima/minima, activation/fluorescence wavelengths) or may be an independent step. Nevertheless, the validity of any marker assay will depend on the ability of the overall procedure to measure only the marker substance itself.
QUANTITATIVE MEASUREMENT The measurement system used to determine the amount of a marker substance in a biological sample can be a single instrument, procedure, or tech-
Chemical Analysis of Marker Substances
559
nology, or a combination of several instruments, procedures, or technologies. One can assume that an electronic instrument will be employed to convert a physical measurement into numerical data. The constraints are simple: the process must be accurate and precise and have a suitable measuring sensitivity.
INTERPRETATION AND APPLICATION The final step in the overall procedure is a combination of interpreting the data obtained from the analysis and applying it to the situation. In the case of any proposed marker system, the question is whether the data support the conclusion that there was satisfactory patient adherence to the therapeutic regimen. Optimally, the data should answer this question with a qualitative yes or no and indicate the qualitative deviation from perfect compliance. Although such an answer is, in part, dependent upon nonanalytical factors such as pharmacokinetics, it is crucially dependent upon the analytical data obtained and the inherent limits such data place upon any subsequent interpretations and conclusions.
CHARACTERISTICS OF ACCEPTABLE PROCEDURES The characteristics of an acceptable assay procedure may be summarized in terms of specificity, sensitivity, speed, simplicity, reliability, economy, and safety. A clear understanding of each characteristic will enable anyone to design a method appropriate for a given marker or, conversely, to define the optimal characteristics of a substance to be used as a marker. Each characteristic is independently important; a failure to satisfy the demands of any characteristic can seriously compromise the overall procedure.
Specificity To determine the concentration of a marker in a biological sample, t h e marker must be differentiated from other substances in the sample. Specificity must be independent of variations in the composition of the sample and must be capable of determining the marker accurately, even in the presence of very high concentrations of other substances. A number of situations may confer specificity upon an analytical method: 1. The measurement technology may be specific, as in spectrophotofluorometry, which requires a precise pairing of activation and emission wavelengths. 2. A chemical reaction may be performed prior to the final measurement step, with the reaction process conferring specificity, as in the o-phthalaldehyde reaction with serotonin.
Roger P. Maickel
560
3. One or more physical separations may be performed, thereby separating substances by virtue of differing physiochemical properties, as in the numerous variants of chromatography. Most assay procedures utilize some combination of these situations. The crucial point is that the final measurement process must determine only the substance of interest or must correct for any interfering substances.
Sensitivity An absolute definition of sensitivity is impossible. A reasonable working definition is that the analytical procedure should be capable of accurately measuring the marker substance in a biological sample at a concentration at least one order of magnitude lower than the expected concentrations. In addition, the numerical response of the measuring system at this limit of sensitivity should be a value at least double that of the "blank" value obtained from an identical biological sample containing none of the compound of interest. Presumably, a marker substance would be present in the body at a concentration of 10-9 to 10-3 M; thus, the ultimate sensitivity level should be in the lower end of this range.
Speed Time is a critical factor in any analytical procedure. For a marker substance, one may project a need for an assay that could be completed in the 30-90minute range. An assay requiring 2-3 days to complete would be cumbersome and might not be cost effective. A relevant point is whether samples would be assayed singly or in groups.
Simplicity The less complicated a method is, the fewer opportunities for error, the shorter the time, and the lower the cost. A procedure should be foolproof and, if possible, idiotproof. The ideal method is one that can be performed by someone with minimal training and should be of minimal complexity. It must be emphasized, however, that specificity and reliability should not be compromised.
Reliability This term covers two aspects of analytical methodology. Reproducibility from sample to sample, from day to day, and from laboratory to laboratory should be sufficient to ensure an error of less than + 5%. Replicate analyses of the same sample, on the same day, and in the same laboratory should also yield values with an error of less than + 5%. A high degree of reliability is particularly important, since a marker may be used to standardize comparison of different dosage forms or therapeutic regimens.
Chemical Analysis of Marker Substances
561
Economy Cost factors in the health care system cannot be ignored. A single analysis that is exhorbitant in cost will severely restrict the usefulness of the marker substance. Cost factors to be stringently examined include labor, instrumentation, and materials. The usefulness of a marker substance may ultimately depend on the fiscal attractiveness of the assay procedure demanded. Safety This aspect of analytical methodology is often ignored, sometimes with disastrous consequences. One need only cite the hazards of using perchloric acid as a protein precipitant if a subsequent evaporation to dryness is performed without removal of the perchloric acid as insoluble potassium perchlorate. SEPARATION A N D MEASUREMENT--BASIC CONCEPTS Virtually every analytical procedure can be divided into two principles or procedures: separation and measurement. Depending on the actual procedure involved, these may be relatively physically independent (as in the case of a paper chromatographic separation followed by elution and a spectrophotometric measurement), or they may be a single procedure (as in GLC with a specific detector). For the purposes of this presentation, the two concepts are considered independently.
Separation Procedures It is uncommon to find that an exogenous substance, such as a marker, can be determined in a biological sample directly, that is, without the application of some sort of procedure involving a physical separation. For most exogenous compounds, such procedures serve two purposes. First, they separate the desired compound from nonspecific gross interferences of a physical nature, such as macromolecules (i.e., proteins), that might otherwise create turbidity, plug a capillary tube, or adhere to a surface. Secondly, they separate the desired compound from more specific interfering substances that, because of chemical similarities, may react in derivatization procedures or yield positive responses in measurement devices or systems. A list of common separation procedures and some details of their characteristics are presented in Table 1. Precipitation may be used to separate proteins and other macromolecules from a biological sample such as blood plasma or serum, thereby permitting direct assay of a marker in the filtrate/supernatant liquid. A number of procedures are available using substances such as trichloracetic, metaphosphoric, or perchloric acids or neutral salts such as zinc hydroxide or barium sulfate. Alternatively, one may selectively precipitate out the compound of interest (by the use of a reineckate salt) and then redissolve it after filtration to remove soluble interfering materials.
562
Roger P. Maickel Table 1
Comparative Characteristics of Separation Procedures Processing time/ sample (manOperator training Procedure hours) Cost/sample Initial cost requirements Precipitation 0.5 $2.50 $250 minimal Liquid-liquid extraction 0.5 $2.50 $500 minimal Column chromatography 1-3 $5.00 $500 minimal Paper chromatography 6 $5.00 $500 minimal Thin-layer chromatography 1-2 $5.00 $500-2000 minimal Gas chromatography (GC)° 0.5 $5.00 $4000-9000 B.S. Liquid chromatography (HPLC)a 0.5-1 $10.00 $1000-4000 B.S. °GC and HPLC used in separation (nonquantitative) mode only.
Liquid-liquid extraction may accomplish the separation of a desired substance from interfering materials by applying the principles of partition between two immiscible liquid phases. This system, which forms the basis for all chromatographic separation procedures, depends on the Nernst distribution law, which says that the sum of the fractions of a solute in two immiscible solvent phases equals the total solute present. Mathematically, this can be expressed as:
K = f/f2
and
fl +f2 = 1.
Each partition that occurs may be considered to have as a basis the characteristic coefficient (K), which is determined by the solute and solvents involved. Repetitive partitioning is known as countercurrent distribution. The general procedures for developing a liquid-liquid extraction process were described some 35 years ago by Brodie et al. [1]. The organic solvent selected should be the one that is least polar, yet will extract the greatest fraction of the marker compound from the (aqueous) biological sample while removing the least amount of interfering substances. The relative polarity of solvents can be estimated from properties such as the dielectric constant, as reported by Craig and Craig [2]. Column chromatography procedures have the disadvantages of limited resolution and lengthy development time. Column packings include a variety of alternatives: alumina, cellulose, ion-exchange resins, silica gel, and others. Development utilizes frontal displacement or elution analysis. The applications of procedures of column chromatography to marker assays will probably be limited by the nature of the marker substance, the concentration in a sample, and the need for rapid processing. The general background of column
Chemical Analysis of Marker Substances
563
chromatography may be found in a number of references [3]; an excellent overview of application of the technique to exogenous organic compounds can be found in Hirtz [4]. Paper chromatographycan also be used to separate compounds, utilizing the principles of repetitive countercurrent distributions between the "bound" water in the paper sheets and appropriate organic solvent systems. The water phase may be replaced by nonpolar compounds or may be adjusted in pH. Many variations on them exist [3]. One serious constraint of paper chromatographic separations is that they are often slow, time-consuming processes and may require additional technology after development to prepare the marker substance for quantitation. Thin-layer chromatographyconsists of a layer of sorbent (separatory material) such as alumina, cellulose, or silica gel on a solid support such as a sheet of glass, aluminum, or a polymeric material. The separation process is accomplished as an appropriate solvent (mobile) phase ascends the sorbent phase by capillarity. The basic principles and practice of TLC are thoroughly described in several sources [5,6]. As a separation technique, TLC is rapid, inexpensive, and widely applicable. Quantitation may be difficult, but a number of procedures are available to facilitate stoichiometric application of samples and quantitative measurements. Technology is also available to enhance separation and speed up development [7]. Gas-liquid chromatography is a relatively recent addition to the armamentarium of the analytical scientist. In this technique, the components of a mixture are separated while being carried (in a stream of inert gas) through a column packed with a stationary phase consisting of some solid material (such as fire brick) generally coated with a high-boiling inert liquid. There are many descriptive reviews of GLC theory, applications, and practice [8-11]. In general, the technique is rapid and has few limitations. The most likely problem with a marker substance might be the relative degree of volatility. One potential advantage of GLC procedures for the analysis of marker substances lies in the diversity of available detector systems. These units, appended at the outflow end of the column, permit separation and measurement to be combined into a single instrumental system. From the early thermal conductivity detectors has come a steady stream of developments: flame ionization, electron capture, and specific detectors for nitrogen, halogens, sulfur, or phosphorous. The ultimate in GLC detection systems is the combination of a gas chromatograph and a mass spectrometer (GC/MS), often coupled with a computer. Unfortunately, the degree of sophistication of such systems results in an extremely high cost of acquisition and maintenance and demands a high level of operator competence. High-performance liquid chromatographyutilizes small bore columns containing appropriate packing materials, with or without coating, and high pressure flow of a mobile phase. Various combinations of stationary and mobile phases can be used to achieve the desired polarity/partition characteristics; the principles of such techniques appear in a number of reviews [!2-14]. A number o f detector systems are available, including ultraviolet absorption, fluorescence, and electrochemical.
564
Roger P. Maickel The technology of HPLC permits relatively rapid processing of samples; the combination of separation and measurement (as in GC) may well be both time and cost effective.
Measurement Procedures Once the marker substance has been "separated" from other substances in a biological sample, a number of technologies can be applied to quantitatively determine the level of the marker substance per unit of sample. As with the prior discussion of separation procedures, only a brief set of comments are made about each of the measurement procedures listed in Table 2. Ultraviolet (UV) absorption spectrometry as a measurement technique requires that the marker substance, or a derivative thereof, have the property of absorbing light in the wavelength range of 200-400 nm. The basic principles and procedures have been reviewed many times; some particularly relevant references may be cited [15-17]. This measurement technique is rapid and simple; nevertheless, limitations on sensitivity and specificity are severe and often demand considerable prework. Systems employing UV detectors are commonly associated with column chromatography of HPLC technologies to permit combined separafion-measurement procedures. Visible (VIS) absorption spectrometry is basically similar to UV technology, with the exception that the range of wavelengths for absorption is 400-750 nm; the basic principles are similar [1,16,17]. In the case of marker substances, the presence of a chromophoric group would make application of VIS methodology inexpensive and simple, especially if the chromophore is of sufficient sensitivity to permit direct measurement, thereby bypassing or simplifying the separation step(s). Infrared absorption spectrometry utilizes the ability of molecules to absorb light at wavelengths >800 n m because of vibrational or rotational energies. This technology has significant potential, although specificity and sensitivity may be critical limitations [18,19]. Spectrophotofluorometry has good criteria for specificity and sensitivity, since it involves both specific absorption and emission wavelengths and has a high quantum yield. Indeed, fluorescence detection and measurement procedures can be utilized with a variety of separation techniques, including TLC or HPLC. The basic principles and a wide variety of practical applications can be seen in a number of reports [20-23]. Many compounds that do not possess fluorescent characteristics per se may be readily converted to a fluorescent compound by chemical reactions or derivatization. Radioisotope derivatization is a technology in which a radiolabeled reagent is chemically "attached" to the substance (such as a marker) to be determined. These procedures can often be carefully tailored to fit a specific compound, although, in doing so, the premeasurement work-up time and cost may be significantly increased. The procedures and principles involved are basically those of organic chemistry [24]. The application of similar technology utilizing radioisotopes in combination with immunoassays is discussed elsewhere in this issue.
moderate to excellent considerable
1.0
2.0-4.0
Radioisotope derivatization
1.0-2.0
excellent
moderate
fair-good
moderate to good considerable
considerable
moderate to fair-good considerable
fair
good
excellent
good
excellent
excellent
10-9 g/ml
10-~° g/ml
good
good
10-1'~/~,1 good
10-12g/ml
10:~ g/ml
10-~ g/ml
10~ g/ml
10-6 g/ml
$17,000
0.25
B.S./M.S.
1.0
$ 7,000
$10,000
1.0 B.S./M.S.
B.S./M.S.
$75,000
2.0
Ph.D.
$20,000
$10,000
0.50
B.S./M.S.
2.0
$ 4,000
0.25
B,S.
B.S./M.S.
$ 7,000
0.25
B.S.
°GC and HPLC are included since use of appropriate ctetector systems permits separation and measurement.
Liquid Chromatography (HPLC)°
Mass spectrometry (GC/MS) 1.0-4.0 Gas chromatography (GC)" 1.0-2.0
considerable
2.0
fair-good
moderate
1.0
fair-good
moderate
1.0
$6O
$40
$80
$70
$40
$50
$30
$30
Estimated cost Limit(s) of Operator Time/assay Specificity sensitivity Reliability expertise man-hours To set-up lab Per assay
UV absorption spectrometry VIS absorption spectrometry IR absorption spectrometry Spectrophotofluorometry
Prework Man-hours Complexity
Comparative Characteristics of Overall Analytical Procedures
Measurement process
Table 2
O1 rjn
566
Roger P. Maickel
Mass spectrometry (MS) is the most recent development in terms of measurement technology. When an MS system is appended to the output of a GC system and combined with an appropriate computerized data processing package, the resultant overall system has extremely high ratings for sensitivity and specificity. Unfortunately, the initial and operating costs of such a system are extremely high, and the expertise required for effective operation is considerable. The potential applications of MS technology to the analysis of marker substances may be seen in a number of reports [25-27]. SUMMARY A N D OVERVIEW
The application of chemical methods of analysis to determine potential marker substances is a complex situation. Several questions must be considered to determine the choice of an appropriate procedure: What is the chemical nature of the marker substance? What type(s) of biological materials will constitute the samples to be analyzed? Will qualitative or quantitative analyses be required? What type of specificity will be needed? What level of sensitivity will be demanded? What constraints of time can be defined? What are reasonable limits for cost per sample or per analysis? The answers to these questions should enable one to develop an appropriate chemical analytical procedure for any potential marker substance.
REFERENCES 1. Brodie BB, Udenfriend S, Baer JE: The estimation of basic organic compounds in biological material. I. General principles. J Biol Chem 168:229-240, 1947 2. Craig LC, Craig D: Extraction and distribution. In: Technique of Organic Chemistry, 2nd ed, Weissburger A, Ed. New York: Interscience, 1956, vol 3, p 172 3. Heftmann E (Ed): Chromatography, A Laboratory Handbook of Chromatographic • and Electrophoretic Methods, 3rd ed. New York: Van Nostrand Reinhold, 1976 4. Hirtz JL: Analytical Metabolic Chemistry of Drugs. New York: Marcel Dekker, 1971 5. Stahl E: Thin-layer Chromatography--A Laboratory Handbook, 2nd ed. New York: Springer-Verlag, 1969 6. Butler TJ: Thin-layer chromatography. In: Guidelines for Analytical Toxicology Programs, Thoma JJ, Bondo PB, Sunshine I, Eds. Cleveland: CRC Press, 1977, vol I, pp 217-251 7. Jupille TH, Perry JA: Programmed multiple development. Spot behavior during solvent advance: A mathematical treatment. J Chromatogr 99:231-239, 1974 8. Vanden Heuvel WJA, Zacchei AG: Gas-liquid chromatography. In: Drug Fate and Metabolism, Garrett ER, Hirtz JL, Eds. New York: Marcel Dekker, 1978, vol 2, pp 49-142 9. Schupp OE: Gas Chromatography. New York: Interscience, 1968 10. Anders MW: Gas chromatography. In: Handbook of Experimental Pharmacology. Brodie BB, Gillette JR, Eds. New York: Springer-Verlag, 1971, vol 28, part 2, pp 63-87
Chemical Analysis of Marker Substances
567
11. Gudzinowicz BJ: Gas Chromatographic Analysis of Drugs and Pesticides. New York: Marcel Dekker, 1967 12. Perry SG, Amos R, Brewer PI: Practical Liquid Chromatography. New York: Plenum, 1972 13. Kirkland JJ (Ed): Modern Practice of Liquid Chromatography. New York: Wiley, 1971 14. Done JN, Knox JH, Loheac J: Applications of High Speed Liquid Chromatography. New York: Wiley, 1974 15. Josephson ES, Udenfriend S, Brodie BB: The estimation of basic organic compounds in biological material. 6. Estimation by ultraviolet spectrophotometry. J Biol Chem 168:341-349, 1947 16. Maickel RP, Bosin TR: Absorption spectrometry. In: Handbook of Experimental Pharmacology, Brodie BB, Gillette JR, Eds. New York: Springer-Verlag, 1971, vol 28, part 2, pp 9-18 17. Cohen SL, Bondo PB: Ultraviolet-visible spectrophotometry. In: Guidelines for Analytical Toxicology Programs, Thoma JJ, Bondo PB, Sunshine I, Eds. Cleveland: CRC Press, 1977, vol II, pp 117-152 18. Kendall DN (Ed): Applied Infrared Spectroscopy. New York: Reinhold, 1966 19. Fales HM: Isolation and identification procedures--spectral methods. In: Fundamentals of Drug Disposition and Drug Metabolism, La Du BN, Mandel HG, Way EL, Eds. Baltimore: Williams & Wilkins, 1971, pp 437-457 20. Ackerman HS, Udenfriend S: Fluorometry. In: Handbook of Experimental Pharmacology, Brodie BB, Gillette JR, Eds. New York: Springer-Verlag, 1971, vol 28, part 2, pp 21-40 21. Udenfriend S: Fluorescence Assay in Biology and Medicine. New York: Academic, 1969 22. Terhaar DA, Porro TJ: Fluorescence spectrophotometry. In: Guidelines for Analytical Toxicology Programs, Thoma JJ, Bondo PB, Sunshine I, Eds. Cleveland: CRC Press, 1977, vol II, pp 153-170 23. Schulman SG, Naik DV: Fluorescence spectroscopy. In: Drug Fate and Metabolism, Garrett ER, Hirtz JL, Eds. New York: Marcel Dekker, 1978, vol 2, pp 195-256 24. Kuntzman RH, Cox RH, Maickel RP: Radioactive techniques: Radioactive isotope derivatives of non-labeled drugs. In: Handbook of Experimental Pharmacology, Brodie BB, Gillette, JR, Eds. New York: Springer-Verlag, 1971, vol 28, part 2, pp 58-62 25. Guarino AM, Fales HM: Gas chromatography-mass spectrometry. In: Handbook of Experimental Pharmacology, Brodie BB, Gillette JR, Eds. New York: Springer -Verlag, 1971, vol 28, part 2, pp 178-194 26. Jenden DE, Cho, AK: Applications of integrated gas chromatography/mass spectrometry in pharmacology and toxicology. Annu Rev Pharmacol Toxicol 13:371-393, 1973 27. Foltz RL: Mass spectrometry. In: Guidelines for Analytical Toxicology Programs, Thoma JJ, Bondo PB, Sunshine I, Eds. Cleveland: CRC Press, 1977, vol II, pp 89-116