Manual for the determination of narcotics and dangerous drugs in the urine

Sections

Introduction
Discussion
Procedure for Test A
Procedure for Test B
Results
References

Details

Author: K. D. PARKER, C. H. HINE
Pages: 51 to 57
Creation Date: 1967/01/01

Manual for the determination of narcotics and dangerous drugs in the urine *

K. D. PARKER
C. H. HINE

Introduction

Detailed procedures are described for the determinnation in urine of common narcotics and some dangerous drugs.

The methods are based on liquid-liquid extractions at controlled pH's, solvent concentration and finally identification and semiquantitation by thin-layer chromatography (T.L.C.).

Two different treatments may be applied to the urine sample. These are termed Test A and Test B, respectively. Test A identifies morphine derived from heroin, morphine, codeine, nalorphine, oxycodone, methyldihydromorphine, dihydromorphinone, and dihydrocodeinone. Test B identifies methadone, pethidine, cocaine, amphetamine, methamphetamine, cyclazocine and D-proproxyphene.

The procedures were designed for the determination of narcotic usage by measurement of the drug or its metabolite on large numbers of samples at a low unit cost. Sixty ml of urine are required. Samples can be mailed to a central laboratory for purposes of parolee monitoring, law enforcement and forensic toxicology.

Tests A and B were developed as screening tests to determine narcotic usage in conjunction with the Nalline pupil test. Sensitivity, sample size, low unit cost and specificity were major considerations in the development of the tests needed for monitoring of parolees in the State of California.

During a part of 1964, 4,800 samples or about 600 urines a month were examined for the presence of narcotics. This was done as a contracted laboratory service by the Parole and Community Services Division of the State of California Department of Corrections, with the Hine Laboratories, Inc. at a unit cost of $3.50. Test A was used to a much greater extent than Test B.

_____________

This article has been reproduced from Psychopharmacology Bulletin, July 1966, with permission of the National Clearinghouse for Mental Health Information, National Institutes of Mental Health, U.S. Department of Health, Education and Welfare, Public Health Service, Bethesda, Maryland.

Test A was performed on urine collected from all geographical areas of the State as a part of the check on parolees for narcotic usage. Both tests were performed on urine from parolees whose pupil test was positive or equivocal, or where there was other evidence that suggested the use of narcotics.

Other agencies, city and county law enforcement and county probation, have used the laboratory's services on an individual sample basis. In these instances provision was made for tamper-proof sealing of the urine sample mailer. Certain additional confirmatory tests were performed including gas-liquid chromatography, thin-layer chromatography using additional developing solvent systems and other color developing reagents for spot location on the thin-layer plates.

The Hine Laboratories, Inc. developed and evaluated these tests during the period 1963-1965 under contract with the National Institute of Mental Health (PH 4363-586 and 43-4-931). Inmate volunteers were administered narcotics using double-blind technique. [ 3] The confined subjects were given pupil tests and the collected urine specimens were tested chemically. Comparisons were made on a double-blind basis of the nalorphine test and urinary analysis for the detection of narcotics use. Some 900 subjects were used during this three-year interval in controlled studies. The accuracy of the test, the time after administration that it remained positive and the quantity of and form in which the narcotic was excreted [ 2] , [ 6] were evaluated at intervals after administrations of known doses of heroin, morphine, codeine, oxycodone, pethidine, methadone, amphetamine and methamphetamine. Using these procedures a trained technician with one assistant can analyze 30 urines daily employing either test A or B. An additional assistant is required for an additional 30 samples or 10-20 urines when working with Test B

Discussion

Test A-begins with an acid hydrolysis step. The urine sample is acidified with concentrated hydrochloric acid and heated for 30 minutes at 120° C. The sample mixture is cooled and made extremely alkaline by the addition of 16 N potassium hydroxide. The alkaline sample mixture is then extracted with butanol. The narcotic is then extracted from the butanol phase with 1 N sulfuric acid. The aqueous sulfuric acid phase is adjusted to pH 8.6 using potassium hydroxide and sodium bicarbonate and extracted with 10% ethanol in chloroform. The chloroform phase is removed and filtered and evaporated to dryness. The residue is redissolved in acetone methanol solution and applied to a thin-layer chromatography plate for development and fractionation by T.L.C. The various colorless fractions are located on the plate as colored spots after the plate is sprayed with the appropriate reagents.

The size and intensity of the narcotic spot provides an estimate of the amount present. The R f value is the major parameter used for identification in Test A and Test B. The R f value for a defined T.L.C. system is the ratio of the distance the narcotic has moved to the distance the developing solvent has moved with respect to the point of sample application. Test A identifies those narcotic substances which are not destroyed by the acid hydrolysis treatment and are extracted by the Test A extraction procedure. These include the following: morphine derived from heroin, morphine, codeine, nalorphine, oxycodone, methyldihydromorphinone, dihydromorphinone and dihydrocodeinone. The T.L.C. solvent system used for Test A contains benzene, dioxane and ammonium hydroxide.

The urine sample is not hydrolyzed for Test B. The urine sample is adjusted to pH 10 with potassium carbonate. Sodium chloride is also added and the mixture is extracted twice with chloroform. The chloroform phase each time is removed and filtered. The pooled chloroform extracts are washed with a weak solution of ammonium hydroxide. The washed chloroform is then extracted twice with 1 N sulfuric acid. The pooled sulfuric acid extracts are then adjusted to pH 10 with concentrated potassium hydroxide and potassium carbonate. Sodium chloride is also added and the mixture is extracted twice with chloroform. The filtered and pooled chloroform is then carefully evaporated after the addition of one drop of a solution of 0.5% sulfuric acid in methanol.

The residue is redissolved in acetone methanol solution and applied to a T.L.C. plate for development. Test B solvent system contains methanol and ammonium hydoxide. Test B detects methadone, pethidine, cocaine, amphetamine, methamphetamine, cyclazocine and D-propoxyphene. Many other organic bases would be extracted by this procedure and appear on the T.L.C. plate.

It would be ideal to have a single test or a single screening procedure which would sensitively and simply determine in the blood and urine all the narcotics, synthetic narcotics, sympathomimetic amines, phenothiazine tranquilizers and sedative hypnotics. Until that test is found, tests narrower in scope must be developed and used which take maximum advantage of the similarities and the differences in the properties among various compounds.

The acid hydrolysis step employed in Test A procedure is necessary for a test of sufficient sensitivity to detect morphine at the 0.5 µg/ml level with a 20 ml urine sample. The water soluble morphine and codeine glucuronides excreted in the urine to the extent of 10-40% of the total narcotic present, are hydrolyzed to release the total narcotic present for organic solvent extraction.

The detectability and the apparent increase in the sensitivity of the determinations which follow are increased about five times due to the increase in extractable narcotic. Acid hydrolysis chemically alters the more labile compounds which suggests the need for a test without this harsh treatment. Both Test A and Test B are based on liquid-liquid extractions at controlled pH's, solvent concentration with identification and semi-quantitation by T.L.C. The solvent system for Test A was selected because it provides a satisfactory separation of morphine, nalorphine, codeine, oxycodone and the components extracted by Test A present in normal urine. The solvent system for Test B was similarly selected for its separation of methadone, pethidine, cocaine, amphetamine, methamphetamine, cyclazocine and D-propoxyphene.

Test B sample preparation followed by either Test A or B T.L.C. resulted in a less specific and less sensitive test for morphine and codeine. Test A sample preparation with Test A T.L.C. provides maximum detectability and specificity for morphine and codeine. The acid hydrolysis step provides for the recovery of the total morphine and codeine and simultaneously increases the specificity of the test due to the destruction of many possibly interfering substances such as the phenothiazines, methadone, pethidine, the amphetamines and D-propoxyphene. Further, Test A T.L.C. simultaneously separates morphine, codeine and nalorphine while Test B T.L.C. fails to separate morphine and codeine. Other solvent systems have been described by Cochin and Daly [ 1] , Mule [ 5] and Mannering, Dixon, Carrol and Cope [ 4] .

From our experience with Test A and B in controlled experiments with narcotics [ 2] , [ 3] , [ 6] and from data on excretion reported in the literature [ 7] , certain statements can be made relative to the biotransformation products of narcotics and their determination in the urine. After administration, codeine, norcodeine and morphine are usually identified in the urine by Test A. As most of a dose of heroin is changed to and excreted as morphine, and because the chemical procedures used in Test A cause deacetylation of both heroin and monacetylmorphine, the finding of morphine in the urine indicates that heroin or morphine or both were probably used. However, morphine only has been found in the urine after 48 to 72 hours of abstinence in about 7% of the 60 subjects tested receiving oral codeine every six hours for five days. Presumably, the codeine was excreted more quickly than the biotransformed morphine. The finding of morphine and codeine in the urine suggests that codeine alone or any combination with morphine or heroin or both was used. Normorphine has not been detected in the urine of subjects even after receiving 30 mg of morphine every six hours for five days. Norpethidine metabolite is almost always found in the urine of subjects administered pethidine when Test B is applied. The lowest concentration of morphine codeine and oxycodone detectable by Test A is 0.5 µg/ml. Methadone and pethidine are detected by Test B in concentrations as low as 2 µg/ml while amphetamine and methamphetamine are detected at concentrations of 3 µg/ml.

Certains aspects of storage of urine specimens for analysis by Test A and B have been evaluated. Urine specimens submitted for narcotic urinalysis services may be in transit via the mail for one or two days and then might spend one or two days (Saturday-Monday) in the Laboratory before the analysis is begun. Four days without refrigeration does not appreciably affect the suitability of the specimen for analysis either by Test A or B. Since sediment is often grossly present, the specimens should always be agitated to insure homogeneity before removing a portion for testing. Urine containing morphine and codeine, which has either been excreted or added, has been kept without refrigeration for 30 days without a significant change in the levels. Urine analyzed by Test B shows an increase in background with the age of urine which partially interferes with the location of the spots after T.L.C. Amphetamine, methamphetamine, pethidine and methadone are more labile compounds than morphine and codeine and more susceptible to decomposition and chemical change during storage or during testing. Specimens kept at room temperature for 4 to 8 days have been suitable for analysis by Test B. Specimens collected during the research experiments were routinely refrigerated from the time of collection. Under these circumstances, results could be duplicated after storage for a month.

A cost study for Test A and Test B services has been made which might be useful to those initiating a urinalysis program. Test A and Test B require about equal expenditures of time, labor and materials. Under part time supervision (10%) of an analytical toxicologist, a full time chemist and a laboratory technician can perform 30 tests a day or 600 tests a month. By the addition of one more laboratory technician or chemist, 1200 tests a month can be performed.

The unit cost in the U.S.A. without a profit factor for each test done en masse would be about $3.50 with the following breakdown: mailer and sample container $0.17, expendable supplies of $0.16, labor $1.72, overhead $1.29, glass breakage and equipment depreciation $0.16.

Procedure for Test A

  1. Measure 20 ml of urine into a 50 ml glass-stoppered centrifuge tube and 2.0 ml of concentrated hydrochloric acid and autoclave for 30 minutes at 20 p.s.i. or about 120° C.

  2. Cool and add 10 ml of 16 N potassium hydroxide and 20 ml of butanol and shake for 5 minutes and centrifuge.

  3. Transfer the upper butanol phase using a pipet control on a 20 ml pipet to another 50 ml tube containing 22 ml of 1 N sulfuric acid. Shake for 5 minutes and centrifuge.

  4. Aspirate off and discard the upper butanol phase. Add 1 drop of 1% phenolphthalein in methanol indicator to the aqueous acid layer and adjust the pH to 7.0 by the cautious dropwise addition of 16 N potassium hydroxide (about 1.5 ml) until the phenolpthalein turns pink. Add about 1 g of sodium bicarbonate, replace the stopper and invert the tube to dissolve the sodium bicarbonate. The phenolphthalein indicator turns colorless.

  5. Add 21 ml of chloroform containing 10% by volume denatured alcohol to the centrifuge tube, shake for 5 minutes and again centrifuge.

  6. Aspirate off and discard the upper aqueous phase. Filter the chloroform layer into a 50 ml beaker through a funnel containing filter paper and about 8 g of anhydrous sodium sulfate. Evaporate (at a slow boil) to near dryness in the vacuum oven maintained at 10 p.s.i., and 90° C.

  7. Transfer the residue in the beaker to a 3 ml micro tube using 3 small volumes (about 0.5 ml) of 1:1 acetone-methanol.

  8. Evaporate the acetone-methanol (less than 2 ml in the 3 ml micro tube to dryness in the vacuum oven as described previously.

  9. The sample is now ready for application to a thinlayer plate.

  10. Spot sample residues, including controls and pure alkaloids on a thin-layer plate along a sample application line located 2.5 cm from and parallel to the bottom of the plate. This permits 11 points of application located on 2 cm centers. Dissolve the residues in 20 µl of 1-1 acetone-methanol and spot the dissolved sample using a 10 µl microsyringe. Repeat the spotting twice adding solvent each time to insure that all of the dissolved residue is transferred. Apply 10 µg each of morphine and nalorphine on one spot and 10 µg each of codeine and oxycodone on another spot toward the center of the application line. Apply the authentic reference compounds as 1 µg/µl solutions in 95% ethyl alcohol to every plate using a 10 µl microsyringe.

  11. Activate the spotted plate in an oven for 10 minutes at 110°C and allow to cool to room temperature before development.

  12. Place the plate into a developing tank containing 200 ml of the following solvent system: 10 ml of No. 3-A denatured alcohol, 11 ml of concentrated ammonium hydroxide, 80 ml of 1,4 dioxane, and 100 ml of benzene. Equilibrate at least 10 minutes before the plate is inserted. Place two plates at a time into each tank and re-use the solvent in the tank once. Allow about 30 minutes for development (front movement about 14 cm).

  13. Remove the plates from the tank and place them in an oven at about 110°C for 10 minutes.

  14. Remove them from the oven and allow the plates to cool to room temperature.

  15. Examine the batch of plates under ultraviolet light and then spray them with potassium iodoplatinate reagent within one hour after removal from the oven for maximum contrast against the background. Evaluate the batch of plates comparing the R f values and color intensities of the authentic reference compounds, the controls and the samples. Spray with Phenol Reagent for a 5 fold increase in sensitivity for morphine. Non-phenolic compounds such as codeine do not produce the intense blue color. Read and interpret the plates as the spray is applied and again after 5 minutes and 20 minutes.

Procedure for Test B

  1. Measure 20 ml of urine into a 50 ml glass-stoppered centrifuhe tube. Add 1 g of potassium carbonate to adjust the pH to 10. Add 4 g of sodium chloride. Add the salts using a powder funnel and measuring spoons. Shake to dissolve the salts.

  2. Add 20 ml of chloroform and shake for 5 minutes and centrifuge.

  3. Aspirate off the lower chloroform layer and filter into another tube.

  4. Add 20 ml of chloroform for a second extraction. Shake for 5 minutes and centrifuge.

  5. Aspirate off the lower chloroform layer and filter into the second tube.

  6. Wash the filtered pooled chloroform with 10 ml of pH 9 aqueous ammonium hydroxide solution as follows: shake for 5 minutes and centrifuge and aspirate off and discard the upper wash phase.

  7. Add 10 ml of 1 N sulfuric acid to the tube, shake for 5 minutes and centrifuge. Aspirate off the upper acid phase and tranfer it to a third tube.

  8. Repeat the extraction with another 10 ml portion of 1 N sulfuric acid and pool with the first extraction in the third tube. Discard the lower chloroform phase.

  9. To the acid phase in the third tube add 16 N potassium hydroxide dropwise (about 1.3 ml) to adjust the pH to about 7. Add 1 g of potassium carbonate to adjust the pH to 10. Add 4 g of sodium chlorides and shake to dissolve the salts.

  10. Add 20 ml chloroform and shake for 5 minutes and centrifuge. Aspirate the lower chloroform phase and filter into a 50 ml beaker.

  11. Repeat the extraction with a second 20 ml of chloroform and aspirate off and discard the upper aqueous phase. Decant and filter the chloroform phase into the beaker. Add 3 ml of chloroform wash to the tube and filter into the beaker.

  12. Add one drop of 0.5 % sulfuric acid in methanol to the pooled chloroform in the beaker.

  13. Evaporate carefully to near dryness in the vacuum oven at a temperature of 90°C and a vacuum of 10 p.s.i. Remove the beaker and allow the final few drops of solvent to air dry.

  14. Transfer the residue to a 3 ml microcentrifuge tube using small portion (0.5 ml) of 1:1 acetone-methanol. Again evaporate at a slow boil to near dryness in the vacuum oven maintained at 10 p.s.i, and 60° C.

  15. Remove a thin-layer plate from the desiccator just before it is to be spotted. Spot the sample residues, procedure controls and reference compounds on a thin-layer plate on the sample application line located 2.5 cm and parallel to the bottom edge of the plate. The 11 points of application are located on 2 cm centers. Dissolve the residues in 20 µl of 1:1 acetone-methanol and spot the dissolved sample from the micro centrifuge tubes using a 10 µl microsyringe. Repeat the spotting twice adding solvent each time to insure that all of the dissolved residue is transferred. Apply 30µg of methadone, 60 µg each of amphetamine and methamphetamine on one spot and 20 µg each of pethidine and cocaine to another spot toward the center of the application line. Apply 10 µg each of morphine and nalorphine on another spot and 10 µg each of codeine and oxycodone on another spot. Apply to every plate the authentic reference compounds methadone, amphetamine, methamphetamine, pethidine and cocaine as 1 µg/µl solutions in 95% ethyl alcohol. Apply them separately until the order of separation is established and consistent. Apply the morphine, nalorphine, codeine, oxycodone to at least one plate in the batch.

  16. Place the spotted plated into the developing tank containing 3 ml of conc. ammonium hydroxide in 200 ml of methanol which has equilibrated for 10 minutes. Allow the development to proceed until there is about 14 cm of front movement in about 30 minutes.

  17. Remove the developed plate and allow it to air dry for about one hour. Examine the plate under ultraviolet light for absorbing or fluorescent spots and circle the spots on the uncoated side of the plate using a china marking pencil. Spray with Dragendorff's Reagent and make notes of spots, colors and intensities. Then spray with potassium iodoplatinate reagent and repeat the observations. Examine the batch of plates making appropriate comparisons. Spray the plates within two hours after development and read and interpret them as the sprays are applied and again 10 minutes later. Use the Phenothiazine reagent spray on a duplicate plate to identify or exclude the phenothiozine tranquilizers and their sulfoxide biotransformation products when indicated.

Results

Table 1 lists the T.L.C. R f values obtained for certain drug substances and metabolites with Test A and Test B solvent systems.

Figures 1-7 illustrate typical Test A and Test B results. These figures, in color, are available on request. Address correspondence to the attention of the Senior Author, the Hine Laboratories, Inc., 1099 Folsom Street, San Francisco, California, 94103.

TABLE 1

T.L.C. Rf values

 

R f Values (X100)

Compound

Test A1

Test B2

Atropine
5 8
Norcodeine
19 19
Morphine
16 40
Dihydromorphinone (Dilaudid)
24 30
Nalorphine (Nalline)
36 71
Dihydrohydroxymorphinone (Numorphan)
55 71
Methyldihydromorphinone (Metopon)
40 30
Normorphine
5 20
Codeine
53 39
Dihydrocodeinone (Dicodid)
62 32
Oxycodone (Percodan)
88 57
D-propoxyphene (Darvon)
98 72
Pethidine (Demerol)
97 56
Methadone (Dolophine)
97 59
Methamphetamine (Methedrine)
60 44
Amphetamine (Benzedrine)
57 54
Heroin
75 63
Monacetylmorphine
63 63
Norpethidine (Nordemerol)
37 42
Levallorphan (Lorfan)
90 81
Cocaine
93 80
Quinine
37 68
Cyclazocine (Win 20,740)
83 67
Norcyclazocine
12 20
1

Solvent System for Test A - Mixture of 10 ml of No. 3A denatured alcohol, 11 ml of concentrated ammonium hydroxide, 80 ml of 1,4-dioxane and 100 ml of benzene.

2

Solvent System for Test B - Mixture of 200 ml of methanol and 3 ml o concentrated ammonium hydroxide.

References

Full size image: 0 kB

Figure 1

T.L.C. separations of compounds applied directly to rhe plat. The Test A developing solvent system consisted of 10 ml of denatured alcohol, 11 ml of ammonium hydroxid, 80 ml of 1,4-dioxane, and 100ml of benzene. The plate was sprayed with potassium iodoplatine. The following compounds were applied to the origins 1 to 10: 1norcodeine, 2 morphine and nalorphine, 3 codeine and oxycodone, 4 dihydromorphinone, 5 dihydrohydroxymorphinone, 6 atropine, 7 normorphine, 8 dihydrocodeinone, 9 D-propoxyphene, 10 pethidine.

Full size image: 0 kB

Figure 2

Separations of compounds applied directly to the plate using Test A solvent system as in Fig. 20. 1 methadone, 2 methamphetamine, 3 amphetamine, 4 monoacetylmorphine and heroin, 5 norpethidine, 6 levallorphan, 7 quinine and cocaine, 8 N-allyl noroxymorphone (not located with spray reagents used), 9 cyclazocine, and 10 norcyclazocine.

Full size image: 0 kB

Figure 3

T.L.C. separations of compounds applied directly to the plate. The Test B developing solvent system consisted of 3 ml of ammonium hydroxide and 200 ml methanol. The plate was first sprayed with Dragendorff`s Reagent and then with potassium iodoplatinate. The following compounds were applied to the origins 1 to 10: 1norcodeine, 2 morphine and nalorphine, 3 codeine and oxycodone, 4 dihydromorphinone, 5 dihydrohydroxymorphinone, 6 atropine, 7 normorphine, 8 dihydrocodeinone, 9 D-propoxyphene, 10 pethidine.

Full size image: 0 kB

Figure 4

Separations of compounds applied directly to the plate using Test A solvent system as in Fig. 20. 1 methadone, 2 methamphetamine, 3 amphetamine, 4 monoacetylmorphine and heroin, 5 norpethidine, 6 levallorphan, 7 quinine and cocaine, 8 N-allyl noroxymorphone (not located with spray reagents used), 9 cyclazocine, and 10 norcyclazocine.

Full size image: 0 kB

Figure 5

Test A applied to normal and fortified urines and urine samples collected from patients with a history of narcotic use. Urines were extracted by Test A procedure and the Test A T.L.C. solvent system was used. The plate was sprayed with potassium iodoplatinate. A descriiption of applications to origins 1 to 9 follows:

  1. normal urine extract

  2. sample urine extract positive for morphine (1 µg/ml) and codeine (1 µg/ml)

  3. sample urine extract positive for morphine (0.5µg/ml) and codeine (0.5 µg/ml)

  4. fortified normal urine extract morphine (0.5 µg/ml), codeine (0.5. µg/ml), and oxycodone (1 µg/ml)

  5. authentic compounds applied directly to the plate, 10 µg of morphine and 10 µg nalorphine

  6. authentic compounds applied directly to the plate, 10 µg of codeine and 10 µg of oxycodone

  7. sample urine extract positive for nalorphine (1 µg/ml)

  8. sample urine extract positive for oxycodone (0.5 µg/ml)

  9. normal urine extract.

Full size image: 0 kB

Figure 6

Full size image: 0 kB

Figure 7

1

Solvent System for Test A - Mixture of 10 ml of No. 3A denatured alcohol, 11 ml of concentrated ammonium hydroxide, 80 ml of 1,4-dioxane and 100 ml of benzene.

2

Solvent System for Test B - Mixture of 200 ml of methanol and 3 ml o concentrated ammonium hydroxide.

001

Cochin, J., and Daly, W. J., Rapid Identification of Analgesic Drugs in Urine with Thin-Layer Chromatography, Experientia 18 , 294-295 (1962).

002

Elliot, Henry W., Nomof, Norman, Parker, Kenneth D., Urinary Excretion of Narcotics During Controlled Experiments Evaluating the Pupillary Tests, Clin. Pharmacol. & Therap. (in preparation) (1965).

003

Elliot, Henry W., Nomof, Norman, Parker, Kenneth, Dewey, Marjorie L., and Way, E. Leong, Comparison of the Nalorphine Test and Urinary Analysis in the Detection of Narcotic Use, Clin. Pharmacol. & Therap. 5, 405-413 (1964).

004

Mannering, Gilbert J., Dixon, Arthur C., Carroll, Micholas

V. and Cope, Ogle B., Paper Chromatography Applied to theDetection of Opium Alkaloids in Urine and Tissues, J. Lab . Clin. Med. 44 , 292-300 (1954).

005

Mule, S. J., Determination of Narcotic Analgesics in Human Biological Materials, Anal. Chem. 36 , 1907-1914 (1964).

006

Parker, Kenneth D., and Hine, Charles H., Urine Screening Techniques Employed in the Detection of Users of Narcotics and Their Correlation with the Nalline Test, J. Forensic Sci . (In press) (1965).

007

Way, E. Leong, and Adler, T.K., The Biological Disposition of Morphine and Its Surrogates, World Health Organization, Geneva (1962).