Routine Radioimmunoassay of Plasma Testosterone, and Results for ...

3 downloads 0 Views 760KB Size Report
Assay of plasma testosterone is an index of andro- genic function that is useful in evaluating feminine hirsutism, virilizingtumors, polycystic ovaries, and.
CUN.CHEM.

21/2.

206-210 (1975)

Routine Radioimmunoassay of Plasma Testosterone, and Results for Various Endocrine Disorders Pui-Yuen Wong, Donald E. Wood, and Thomas Johnson

We describe a modification of published methods for radloimmunoassay of plasma testosterone. This simpler method involves no chromatographiC steps, and the necessary reagents, including tritiated testosterone and testosterone anti-serum, are commercially available. A digital computer is used for the calculation. Without difficulty, a technician can complete 100 assays in three

working days. Mean testosterone concentrations (±2 SD) in the plasmas of 21 normal men and 26 women a normal menstrual cycle were 684 ± 300 and 45 zk 2O ng/dl, respectively. Within- and between-assay precision (coefficient of variation) were 5.2% (n = 29) and 6.7% (n = 26), respectively. We have assayed more than 1000 samples during the past year. We give with

data on the concentrations of testosterone in plasma of patients with various endocrine

disorders.

Addftlonal Keyphrases: diagnostic ers

#{149} normal

values

aid

#{149} endocrine

disord-

#{149} steroids

Assay of plasma testosterone is an index of androgenic function that is useful in evaluating feminine hirsutism, virilizingtumors, polycystic ovaries, and the adrenogenital syndromes, and in the male in evaluating impotence, primary and secondary hypogonadism, and delayed and precocious puberty. It is also useful in monitoring therapy in several of these conditions. Techniques proposed for assay of testosterone in plasma include gas chromatography (1), the doubleisotope derivative method (2, 3), and competitive protein-binding analysis (4-6). Some of these methods are accurate, sensitive, and specific, but all are tedious. The recent successful production of an antiserum for testosterone provides a more reliable and facile technique that retains accuracy and sensitivity (8,9). In this communication we wish to report our experience in routine radioimmunoassay of plasma testosterone with use of [1,2,6,7(n)-3H]-testosterone for recovery determination as well as for radioimmunoassays. Toronto General Hospital, 101 College St., Canada M5G 1L7. Received July 8, 1974; accepted Nov. 6, 1974.

206

Toronto,

CLINICAL CHEMISTRY, Vol. 21, No. 2, 1975

Ontario,

Materials and Methods Materials Testosterone antiserum: This was obtained from Endocrine Sciences, Tarzana, Calif. 91356. Testosterone standard: This was obtained from E. Merck, EM Laboratories Inc., Elmsford, N.Y. 10523, and [1,2,6,7(n)-3H]-testosterone (which we call here “31I-labeled testosterone”) from Amersham/Searle, Arlington Hei’ghts, Ill. 60005 (cat. No. TRK-402, 90

Ci/mmol). Organic solvents: All solvents, including water, were redistilled in glass. Borate buffer: 1 g of gelatin was dissolved in 100 ml of hot water. Boric acid, 4.0 g, and 1.0 g of sodium azide were dissolved in 500 ml of water and the pH was adjusted to 8.0 with sodium hydroxide. The gelatin solution was added and the mixture diluted to 1000 ml. Saturated

aqueous

ammonium

sulfate.

Dextran-coated charcoal: 50 mg of dextran and 500 mg of Norit A charcoal (Schwartz/Mann, Orangeburg, N. Y. 10962) were added to 100 ml of the borate buffer. Testosterone antiserum solution: The stock antiserum was diluted with borate buffer containing 3Hlabeled testosterone (about 5000 cpm/ml) to obtain 50-60% bound at 0 pg testosterone concentration. Liquid scintillation fluid: “Unogel” (Schwartz! Mann). Pipettor: Model 25000, Micromedic Systems, Inc., Philadelphia, Pa. 19105. Liquid scintillation counter: Unilux II with Teletype print-out (Searle Analytic Inc., Des Plaines, Ill. 60018). Digital computer: PDP 11/40 (Digital Equipment Corp., Maynard, Mass. 01754). Methods Extraction of plasma testosterone: The procedure is summarized in Figure 1. Generally, 1.0 ml of plasma (or in the case of men, 0.1 ml of plasma) was pipetted into test tubes containing 3000 cpm of 3H-labeled testosterone (about 10 pg). The contents were

Plasma

taming 0, 10, 25, 50, 75, 100, and 150 pg of testosterone and 200 cpm of 3H-labeled testosterone were added. The contents of these tubes were evaporated. The tubes containing the dried extracts were stored at -20 #{176}C until assayed. Radioimmunoassay: Figure 2 summarizes the experimental procedure. Testosterone antiserum solution 0.5 ml was added to the test tubes containing the dry residue, and tubes were incubated at 37 #{176}C for an hour followed by an additional hour at 4 #{176}C. Cold dextran-coated charcoal, 0.5 ml, was added to each tube. They were vortex-mixed and then centrifuged at 3000 rpm for 10 mm at 4 #{176}C. The supernatant fluid was poured into counting vials containing 10 ml of Unogel. The activity in all the vials was measured.

+

3H-Testosterone

1 h

at 37’C

lh

at

4#{176}C

(NH4)2S04 Precipitation

extraction

Calculation very

determinatio Fig. 1. Flow chart one from plasma

for the procedure for extracting testoster-

Using a digital PD? 11/40 computer off-line in the batch processing mode, we constructed the standard curve by plotting percent antibody bound vs. concentration of unlabeled testosterone. Percent antibody bound testosterone for standards is = [C/(A and for samples = [C/(A

where C

=

+ S)] X 100 + E)]

x 100

cpm of supernate

after addition

of dextran

coated charcoal, A

=

S E T

= =

=

cpm in 0.5 ml of 3H-labeled-testosterone antiserum solution, cpm added to standard tubes, cpm in the aliquot used for assay, and total cpm added to plasma samples before extraction.

The testosterone content of samples was read off the standard curve in picograms per sample = (pg read off standard curve water blank) X (T/E) -

Results Fig. 2. Flow chart of the radioimmunoassay procedure

vortex-mixed for 30 s and incubated at 37 #{176}C for 1 h and for an additional hour at 4 #{176}C. Nine volumes of 53% saturated ammonium sulfate (4 #{176}C) was added to each tube and the contents were mixed. The tubes were centrifuged at 3000 rpm for 10 mm. The supernatant fluid was discarded. The precipitate was dissolved in 2 ml of water (room temperature) and extracted once with 3 ml of hexane-ethyl acetate (9:1 by vol). The organic (upper) phase was separated by centrifugation and 0.3 ml aliquots of it were transferred to three 10 X 75 mm glass disposable tubes and a liquid-scintillation counting vial. The contents were evaporated in a vacuum oven at 45 #{176}C. Ten milliliters of Unogel was added to the counting vials and the activity measured. To another set of test tubes, standards con-

Analytical

Variables

Recovery of 3H-labeled testosterone added to plasma sample. The mean percent recovery ±SD of 3H-labeled testosterone added to 143 samples from women and 125 samples from men was 68.7% ± 10.2% and 55.0% ± 10.1%, respectively. Antibody titration curve. We made serial dilutions ranging from 10 000- to 70 000-fold of the antiserum in borate buffer containing, per milliliter, 5000 cpm of 3H-labeled testosterone. Figure 3 shows the titration curve. Water blank. Distilled water was used to investigate blank values obtained in this system. The mean ±SD, expressed as pg equivalents of testosterone per tube, was 4 ± 2 (n = 20). Linearity. The linearity of the procedure was checked by assaying increasing volumes of the extracts of plasma sample. The results are shown in

Figure 4. CLINICAL CHEMISTRY,

Vol. 21, No. 2, 1975

207

were sent to Inter Sciences Institute (Los Angeles, Calif. 90025) for determination of testosterone concentrations, with use of A1203 chromatography as a purification step. The testosterone concentrations of these samples were also determined by assay after the described (NH4)2S04 precipitation in our laboratory. The results are summarized in Table 1.

BOUND

80

70

60

Clinical 50

Normal subjects. Mean testosterone concentrations (±2 SD) for 26 women (whose menstrual cycles were normal) and 21 normal men, as assayed by the present method, were 44.5 ± 19.8 and 684 ± 300 ng/ dl, respectively. Pregnancy. The meafl testosterone concentration (±2 SD) fOr nine pregnant women (gestation >16 weeks), as measured by the present method, was 122 ± 94 ng/dl. Hirsutism and adrenogenital disorders. For a group of 13 patients with hirsutism and adrenogenital disorders, the mean testosterone concentration ± 2 SD was 103 ± 30 ng/dl (range, 78.8-127 ng!dl). Hypogonadism. For a group of 21 patients with either primary or secondary hypogonadism, the mean (2 SD) was 245 ± 190 ng/dl (range, 80.8-390 ng/dl). Treatment of hypogonadism. After i.m. administration of 2.0 ml of “Depo-Testosterone” (100 mg/mi, Upjohn) to each of a group of hypogonadal patients, plasma testosterone was determined at 0, 7, 14, 21, and 28 days. Generally, the plasma testosterone concentration was highest at day 7, and it had dropped to the original value by day 28. A typical pattern is’ shown in Figure 5.

40

30

1/110 ANTISERUM

DILUTION

Fig. 3. Antibody titration curve Endocrine

Sciences’

testosterone

antiserum

T3-1 25 was used

0 500

w

z

0

Studies

‘100

300

200

100

100 i

200 EXTRACT

300

ASSAYED

Fig. 4. Relationship between the aliquots of extract used in the radioimmunoassay and their testosterone content

Table 1. Comparison of Plasma Testosterone Values for 13 Subjects, as Determined by Radioimmunoassay, with Use of A1203 Chromatography and (NH4)1S04Precipitation as Purification Steps AIzO, chromatography

(NH42S04 precipitation ng/dl

Males

606 Within-assay precision was examined by measuring the testosterone concentrations of a pool of plasma from women. The coefficient of variatioh for 29 samples was 5.2% and the thean value (±2 SD) was 47.3 ± 2.5 ng/dl. Between-assay precision was determined by measuring the testosterone in the same pool plasma on 26 separate days. The coeITicient of variation was 6.7% and the mean value (±2 SD) was 45.8 ± 3.1 ng/dl. Precision.

Comparison in

subjects

of plasma as

without

testosterone

determined chromatography.

by

concentration radioimmunoassay

Thirteen

with

and

208

CLINICALCHEMISTRY, Vol. 21, No. 2, 1975

samples

589 420

442 20.0

14.6 35.8 120

33.0 123 52.0 932 727 182 465 402

52.6

979 714 218 555 473 F’emales

34.0 23.0

40.7 30.4

ng%

Table 2. Plasma Testosterone Values for a 29-Year-Old Man with Hypogonadism upon

TES1OSTERJE

Choriogonadotropin

Stimulation

1501

Test ng /d I

Diagnosis: Hypogonadism Baseline:

60.0

3 days after HCG: 0800h 2000h

125

123

1001

Table 3. The Laboratory Results for a 26-Year-Old Man with Panhypopituitarism HGH: j TSH: 1. 13 14 , Plasma cortisol: 0800h 1000 h FSH: Testosterone: 614 ng/dl --#{176}

500

-

-

0

7

hi

21

20

DAYS

Fig. 5. A typical plasma testosterone concentrations of a hypogonad patient after i.m. administration of 200 mg of testosterone cypionate

Pituitary tumors. A group of five male patients with pituitary tumors gave values of 12.7, 14.6, 20.5, 25.0, and 829 ng/dl. Choriogonadotropin stimulation test. For a male patient with secondary hypogonadism, the testosterone values before and after a choriogonadotropin stimulation test are shown in Table 2. Panhypopituitarism. Results are summarized in Table 3 for a male patient with panhypopituitarism. Discussion A globulin fraction rich in bound testosterone was separated from serum by ammonium sulfate precipitation before extraction. This preliminary precipitation step enables the radioimmunoassay to be done without a chromatographic purification step. Recovery. 3H-labeled testosterone is used to determine recovery as well as for radioimmunoassay. Because of the specific activity of [1,2,6,7(n)-3H]testosterone is greater than that of [1,2(n)-3H]testosterone, lessthan 10 pg of labeled steroid is required for recovery determination, an amount can be ignored in the final calculations. In addition, less antiserum is required when this labeled compound is used. As Figure 3 shows, maximum sensitivity was obtained at about a 30 000-fold dilution. Specificity of the method was examined by analysis of blanks, determination of antibody specificity, and by comparison of results obtained by the present method with those obtained by using established methods (3, 5, 8,, 9). Water was used to investigate

blank values obtained in this system. The mean (±SD) was 4 ± 2 pg equivalents of testosterone (n = 20). Cross-reactivity. Data on the cross reaction of the antiserum with other steroids was obtained from the supplier. Generally, dihydrotestosterone, delta-idehydrotestosterone and delta-i-testosterone crossreact significantly with the antibody because cross reactions with these compounds would be expected with antibodies to testosterone conjugated at the no. 3 carbon. A separate experiment has shown that 300 pg of 17-methyl-testosterone and 1-dehydrotestosterone were processed through the percent method as though they were equivalent to 118 and 199 pg of testosterone, respectively. Linearity. Figure 4 shows the linearity of the procedure. From this, it may be assumed that the material assayed in plasma samples and that used in the standard curve are identical in their reactivity with the antibody. Interlaboratory comparison. Comparison of results for 13 samples where testosterone assay was determined by using a chromatographic purification (Inter Science Laboratories) and the present method, are summarized in Table 1. An excellent agreement was found between these two methods. Normal range. The plasma testosterone concentrations (±2 SD) of normal subjects were established by assaying samples from 26 women with normal menstrual cycles and 21 normal men. As shown in Table 4, the normal ranges are in good agreement with several published methods, such as double-isotope dilution (4), competitive protein-binding analysis (5), radioimmunoassay with (8) and without (9) chromatography. Plasma testosterone concentrations increase during pregnancy. Clinically, these patients showed no signs of hirsutism or virilization, probably because of the increased concentration of testosterone-binding CLINICAL CHEMISTRY,

Vol. 21, No. 2, 1975

209

Table 4. Plasma Testosterone

Values (ng/ dl) for Normal Persons Female

Author

Rivarola and Migeon (3) Mayes and Nugent (5) Furuyama et al. (8) Ismail et al. (9) Present authors

Method

No.

Double.isotope dilution CPBA RIA with chromatography RIA without chromatography RIA without chromatography

globulin. Separate experiments showed that the human placental lactogen concentrations of these specimens were also increased. Estimation of plasma testosterone concentrations appears to be an important adjunct when investigation virilism and hirsutism, particularly when adrenal function tests were normal and an ovarian source of excess androgen is suspected. It has also been used in studying the pathophysiology of hypogonadism, in psycho-endocrine studies, and in evaluating the relationship of adrenocortical hormone secretion to surgical trauma. In these studies, we found that plasma testosterone concentrations agreed not only with the clinical conditions, but also with values in the literature (10-14). In development of the routine radioimmunoassay for plasma testosterone in a clinical laboratory, certain problems must be considered, such as the time required to report a result. With the present method 100 samples, in triplicate, can easily be processed in three working days. Turn-around time is improved by counting at night and the calculations are completed by using a digital computer. In our protocol, automatic and precalibrated pipettors replace most of the hand pipetting operations. The reagents, tntiated testosterone and testosterone antiserum are all commercially available.

Male

Mean ± 2 SD

9 18 12 6 26

47 ± 40 ± 33 ± 46 ± 45 ±

30 28 14

No.

Mean ± 2 SD

13 16 13

42

9

20

21

551 ± 680 ± 590 ± 596 ± 684 ±

302 360 298

404 300

References 1. Brownie, A. C., Vander Molen, H. J., and Nishizawa, E., Determination of plasma testosterone in human peripheral blood using gas liquid chromatography with electron capture detection. J. Clin. Endocrinol. 24, 1091 (1964). 2. Riondel, A., Tail, J. F., Gut, M., et aL, Estimation of testosterone in human peripheral blood using S35-thiosemicarbazide. J. Clin. Endocrinol. 23, 620 (1963). 3. Rivarola, M. A., and Migeon, C. J., Determination of testosterone and androst-4-ene-3,17-dione concentration in human plasma. Steroids 7, 103 (1966). 4. Fritz, G. R., and Knobil, E., The measurement of testosterone in plasma by competitive protein-binding analysis. Fed. Proc. 25, 757 (1967). 5. Mayes, D., and Nugent, C. A., Determination terone by the use of competitive protein-binding. nol. 28, 1169 (1968).

of plasma

testos-

J. Clin. Endocri-

6. Demetrion, J. A., and Austin, F. G., Quantitation tosterone by improved competitive protein-binding Clin. Chem. 16, 111 (1970).

of plasma testechnique.

7. Midgley, A. R., Jr., Niswender, G. D., Goy, V. L., and Reichart, L. E., Jr., Use of antibodies for characterization of gonadotropins and steroids. Recent Prog. Horm. Res. 27, 235 (1971). 8. Furuyama, S., Mayes, D. M., and Nugent, C. A., A radioimmunoassay for plasma testosterone. Steroid 16, 415 (1970). 9. Ismail, A. A., Niswender, F. D., and Midgley, A. R., Jr., Radioimmunoassay without chromatography. J. Clin. Endocrinol. 34, 177 (1972). 10. Hancock, K. W., Levell, M. J., and Coulson, A., The value of plasma testosterone estimation in the diagnosis of virilizing ovarian neoplasia. 11. Ettigen, G., Goldfield, R. B., and Burrill, K. D., Plasma testosterone stimulation suppression dynamics in hirsute woman. Correlation with long-term therapy. Amer. J. Med. 54, 195 (1973). 12. Kent, tosterone

J. R., Scaramuzzi, R. J., and Lawrence, W., Plasma tesestradial, and gonadotrophins in hepatic insufficiency. Gastroenterology 64, 111 (1973).

We thank the staff in the Division of Endocrinology for supplying the specimens, and Dr. T. D. Cradduck for writing the computer program. This work was originally presented at the 17th Annual meeting of the Canadian Society of Clinical Chemists in Banff, Canada, in August 1973.

210

CUNICALCHEMISTRY,Vol. 21, No. 2,

1975

13. Sacher, E. J., Halpern, F., and Rosenfeld, R., Plasma and urinary testosterone levels in depressed men. Arch. Gen. Psychol. 28, 15 (1973). 14. Carstensen, H., Amen, B., and Amen, I., The post-operative

decrease relation

of plasma to plasma

testosterone FSH

in man, after

and LH. J. Steroid

major

Biochem.

surgery

4, 45 (1973).

in