Detection of Testosterone Administrationby Increased Ratio between Serum Concentrations of Testosterone and 17a-Hydroxyprogesterone

Transcription

1 CLN.CHEM.38/9, (1992) Detection of Testosterone Administrationby ncreased Ratio between Serum Concentrations of Testosterone and 17a-Hydroxyprogesterone Kjell Carlstr#{26}m, Elzbieta Palonek,2 Mats Garle,3 Helge Oftebro,5 Johan Stanghelle,6 and ngemar Bj#{26}rkhem2 An increased ratio between urinary testosterone (1) and epitestosterone (epit) has been accepted by the nternational Olympic Committee as a marker for T doping. However, in a few subjects, we and others have observed constantly above-normal urinary T/epiT ratios that are unlikely to be related to exogenous T administration. To find a better test for T doping, we studied several serum and urinary androgens and androgen precursors, estrogens, and luteinizing hormone (LH) in seven healthy volunteers for 35 days after an intramuscular injection of 25 mg of testosterone enanthate. Among urinary analyses, only the T/epiT ratio was a suitable marker of T doping; of the serum assays, 17a-hydroxyprogesterone (17OHP), T/1 7OHP ratio, LH, and T/LH ratio were fair to good markers of doping. The serum T/1 7OHP ratio was the best marker of those tested, with all seven subjects having above-normal values for this in the first 3 days of the observation period. No other marker showed abnormal values in all subjects at any time. Moreover, the 1/1 7OHP ratio was affected by neither diurnal variation nor physical stress. The value of this marker for T doping was further supported by the finding of normal T/1 7OHP ratios in a subject with increased urinary T/epil ratios caused by an abnormally low testicular epit production, probably related to genetic factors. Addtonal Keyphrases: sports medicine - steroid hormones - abused drugs. genetic variation urine vs serum markers About a decade ago, the increasing efficiency of the antidoping laboratories to detect synthetic anabolic steroids led to abuse of the natural androgen testosterone (T) among athletes.7 However, exogenously adininistered T causes marked and characteristic changes in the pattern of steroids excreted in the urine. As shown by Donike et al. (1), not only is practically no epitestosterone (epit) formed from exogenous T, but also its formation is inhibited by exogenous T. Thus, administration of exogenous T leads to an increased ratio between T and epit, such that an above-normal urinary T/epiT ratio is considered a reliable indicator of T doping Departments of Obstetrics and Gynecology,2 Clinical Chemistry, and 3Clinical Pharmacology, and the Clinical Research Center,Karolinska nstitutet, Huddinge University Hospital, S-1186 Huddinge, Sweden. 5Norwegian nstitute forsports Medicine, Ullev#{225}l, Oslo, Norway. 5Sunnaas Hospital, Oslo,Norway. 7Nonstandard abbreviations: T, testosterone; 17OHP, 17a-hydroxyprogesterone; L!, luteinizing hormone (lutropin); epit,epitestosterone; A-, -androstene-3,17-dione; te2,totalestradione- 17p; and tel, totalestrone. Received January9, 1992; accepted March 31, (nternational Olympic Committee Medical Committee Meeting, Los Angeles, CA, February 1982). However, results from several antidoping laboratories, including our own, indicate that genetic, environmental, or dietary factors may affect this ratio and may even cause falsely positive results in some rare cases (2-). On the basis of the statistical distribution of values for the T/epiT ratio in teenage athletes, Dehennin and Scholler (3) calculated that the rate of false-positive results would be -1.5/1. n our opinion, additional criteria to confirm T doping are needed. Circulating concentrations of luteinizing hormone (LH) in men are suppressed by T admimstration; therefore, decreased urinary LH excretion or increased urinary TLH ratios might be useful as such additional criteria (5, 6). n national competitions in Sweden, doubtful cases with low absolute concentrations of epit and with T/epiT ratios between 6 and 1 are never condemned unless a decreased urinary LH excretion is documented. However, because of the difficulty of distinguishing between normal and decreased L concentrations in young men by standard immunoassays and because of the possibility of reporting falsely high urinary concentrations, this is not an optimal complementary test. n addition, strenuous exercise itself may affect L concentrations (7). Another strategy is to measure the T/epiT ratio in urine samples collected at various intervals after the collection of the first sample with an increased ratio. f T has been administered, the T/epiT ratio can be expected to change with time. However, this time-consuming method is also not entirely reliable, particularly if the athlete knows why extra urine samples are being collected. Exogenously administered T is to some extent converted into estrogens (8, 9). Thus we considered that urinary estrogens might be a marker for T doping. Because only urine samples have thus far been used in the testing for androgen doping, we also considered the possibility that supplementary analysis of a serum sample could give additional useful information. n the present work we monitored the concentrations of several serum and urinary androgens, androgen precursors, estrogens, and LH in healthy volunteers during 35 days after a single intramuscular injection of testosterone enanthate. Materals and Methods Subjects and Design of the Study Seven sedentary men, ages years (mean 27., SD 2.), volunteered for the pharmacokinetic/metabolic study. They were in general good health, were taking no CLNCAL CHEMSTRY, Vol.38, No.9,

2 Table 1. Serum and Urinary Hormone Concentrations n Healthy Sedentary Men, Ages 2-39 Years Age, y.srs Hormone concn n mean SEM mean Rang. S-T, nmol/l S-i 7OHP,nmol/L S-T/17OHP ratio S-LH, lu/l S-T/LHratio, nmol/lu S-A-, nmol/l S-E2, pmol/l S-tEl, nmol/l U-T/epT ratio U-LH/Cr,klU/mol U-T/LHrato U-tEl/Cr, x io U-tEa/Cr, x S. Serum; U, urine; Cr, creatinine. medications, and had no clinical or laboratory signs of endocrine, hepatic, biliary, renal, or intestinal malfunction. Each subject received an intramuscular injection of 25 mg of testosterone enanthate (Testoviron-Depot#{17}; Schering AG, Berlin, FRG) on day. Samples of venous blood and of urine (untimed samples) were collected on the day before injection (defined as day values) and on days 1, 2,, 7, 9, 11, 17, 25, and 35 after injection. The blood samples were taken at 8 and serum was separated after centrifligation at 8 x g. Serum and urine samples were stored at -2 #{176}C until analyzed. All samples from the same subject were analyzed in the same assay to avoid between-assay variation. Reference values for endogenous hormones and metabolites were obtained from a population of apparently healthy, sedentary men, who matched the abovementioned inclusion criteria. Because the number of subjects differed for different compounds analyzed, these data and the data on age are given in Table 1. To study diurnal variations in the serum T/17a-hydroxyprogesterone (17OHP) ratio, we monitored the ratio in 15 elite ice hockey players and in 2 spectators, matched with the athletes for physical fitness, during a 26-h tournament. Data from these two latter groups were presented in a previous paper (1). The validity of the serum T/17OHP ratio was further investigated in a man with a well-documented constantly increased urinary T/epiT ratio, for whom this increase was not related to T administration (). n three unannounced doping tests, this 23-year-old elite athlete, who denied doping, delivered urine samples with above-normal T/epiT ratios (ci. below). He was further subjected to a suppression test with the antifungal imidazole derivative ketoconazole, which efficiently decreases circulating T to orchidectomy values (11). n subjects with normal testicular secretion of T but abnormally low epit production, this test will result in decreased urinary T/epiT ratios because of the decrease in T secretion. n T abusers, on the other hand, circulating T concentrations would be marginally reduced and un- 178 CLNCALCHEMSTRY, Vol. 38, No. 9, 1992 nary T/epiT ratios would remain above normal or even increase as a consequence of a decreased testicular secretion of epit. n this subject, ketoconazole suppression caused peripheral T concentrations to decrease from 2 to 2 nmol/l and the urinary T/epiT ratio fell to -1. These and other findings, including results from luliberin (LH-releasing hormone), corticotropin, and dexamethasone tests, indicated that the high urinary T/epiT ratio was not caused by exogenoust administration, but rather resulted from an abnormally low testicular production of epit, probably of genetic origin. Serum and urinary samples were collected on two different days, 3 days apart, and analyzed for serum T and 17OHP and for urinary T, epit, LH, and creatinine. The study was approved by the local ethical committee, and full informed consent was obtained from each subject in the study. Analytical Methods All assays were carried out in duplicate. Serum concentrations oft were determined directly in untreated serum by using a commercial radioimmunoassay kit (Coat-a-Count#{17}; Diagnostic Products Corp., Los Angeles, CA). Serum concentrations of unconjugated estradiol-17/3 (E2) were determined by direct radioiinmunoassay in untreated serum by a modification of a commercial kit (Estradiol Double Antibody; t)iagnostic Products Corp.), performed with the following changes: (a) After addition of the E2 tracer, the samples were incubated overnight at #{176}C; (b) Bound and free radioactivity were separated by adding 1 ml of a cold suspension of 1 g of charcoal (Norit A or equivalent) and 1 mg of Dextran T7 (Pharmacia AB, Uppsala, Sweden) in 1 ml of phosphate-buffered saline, thorough mixing, and incubation for 15 mm at #{176}C, followed by centrifugation at 2 x g for 1 mm at #{176}C and counting the radioactivity in the supernate. This modification was undertaken to reduce the nonspecifically bound radioactivity. n the original precipitation procedure, certain samples had a higher nonspecific binding

3 than did the standard, resulting in falsely high radioactivity and falsely low or even negative E2 values. The modification reduced nonspecific binding from -5% to 1-2% of the activity in the zero calibrator, and no E2 values < were obtained. Serum -androstene-3,17-dione (A-) and 17OHP and serum and urinary total estrone (tel; sum of conjugated + unconjugated estrone; in serum, te consistsof 85% estrone sulfate) were determined by radioimmunoassay after extraction with diethyl ether by methods developed at our departments (12-1), with the antibodies specified in the paper of Stege et al. (1). n the assay of tel, the estrogen conjugates were hydrolyzed by partially purified sulfatase + f3-glucuronidase preparation (from Helix pomatia) before extraction. Urinary concentrations of total estradiol (te2; sum of conjugated + unconjugated E2) were determined as described for tel, except that we dissolved the evaporated ether extract in zero calibrator supplied by the manufacturer, and quantified the hormone with a commercial radioimmunoassay kit (Coat-a-Count). Serum and urinary concentrations of LH were determined by time-resolved fluoroimmunoassay with a commercial kit (Delfia hlh Spec#{17}; Pharmacia Diagnostics AB). For the urine assay, the procedure was slightly modified to include larger sample volumes (2 tl) and an increased ionic strength of the assay buffer. The LH concentrations were expressed as international units (U)L or fl.j/mol of creatinine of the WHO 2nd nternational Standard for pituitary LH for immunoassay 8/ 552. Urinary T and epit were determined by gas chromatography-mass spectrometry after hydrolysis of the conjugates with p-glucuronidase. A slight modification of a previously published method (15) was used, in which deuterium-labeled T and epit were included as internal standards (Garle et al., to be published). Because untimed urinary samples were used in the pharmacokinetic/metabolic study, we related the urinary estrogen and LH concentrations to the urinary creatimne content. Minimum quantifiable concentrations and within- and between-assay coefficientsof variation, respectively, were as follows: serum T:.1 nmol/l, 6%, and 1%; serum E2: 6 pmol/l, 5.%, and.6%; A-:.6 nmol/l, 6%, and 1%; 17OHP:.1 nmol/l, 7%, and 1%; serum and urinary tel:.33 nmol/l, 7%, and 9%; serum and urinary te2:.5 nmol/l, 6%, and 1%; serum L:.6 lu/l, 5%, and 9%; urinary LH:.2 lu/l, 6%, and 9%; urinary T: 2.8 nmol/l, 6%, and 9%; and urinary epit: 3.5 nmol/l, 6%, and 9%. Statistical Methods Statistical analysis was performed by the Wibcoxon signed-rank test. Because most variables were not normally distributed, the values at day and on subsequent days are given as geometric mean and range, except where noted. Results Serum and urinary hormone concentrations and ratios in the seven volunteers receiving an intramuscular injection of testosterone enanthate are shown in Figures 1 and 2; Table 1 lists the corresponding values for the reference population. The number of abnormal values for the individual analytes recorded at the different test occasions are given in Table 2. njection of testosterone enanthate caused significant changes from the day concentrations for each variable on at least one test occasion. However, only the T/17OHP ratio showed abnormal values for all seven subjects at the same test occasion. Data for the T/17OHP ratio in elite athletes and in matched spectators (9) at different times of day during a tournament are shown in Figure 3. The T/17OHP ratio was not affected by diurnal variation or physical stress. n the subject with an increased urinary T/epiT ratio of probable genetic origin, the ratio on the first day of sampling was 5.8 and 3 days later was 6.3, both values clearly greater than the reference values in Table 1. Corresponding results for urinary LHlcreatinine were.39 and 1.56 kllj/mol, for serum LH 6.7 and 7.5 lu/l, for serum T 22.7 and 21.9 nmol/l, and for serum 17OHP 5.6 and 6.1 nmol/l. The serum T/17OHP ratios were.5 and 3.59, respectively. Except for slightly increased serum LH values and one increased value for the urinary LHlcreatinine ratio, all the values for this subject were in the normal range. Discussion Most of the circulating estrogens in men are formed by peripheral aromatization of androgens, notably A- and T (cf. ref. 16). Because exogenous adminigtration of T leads to increased concentrations of E2 (8, 9), estrogen assays could theoretically give confirmatory evidence in suspected cases of T doping. n the present single-dose study, increased estrogen concentrations were detected in only a few of the subjects during the first days of observation. This probably reflects the rather modest increase in substrate steroids, i.e., A- and T. However, long-term treatment with T preparations may give more significant increases in serum and urinary estrogens (9). 17OHP is one of the key precursors in testicular T synthesis, and - 75% of circulating 17OHP is of testicular origin (16). Endogenous concentrations of T and 17OHP are closely correlated in healthy men (1, 11, 17), and administration of synthetic androgenic/anabolic steroids reportedly suppresses 17OHP concentrations by -75% (8). 17OHP is easily determined accurately by conventional radioimmunoaasay, and low concentrations of 17OHP or increased T/17OHP ratios may serve as additional markers for T doping. We confirmed this possibility in the present study, where the serum T/17OHP ratio was the most sensitive marker investigated. Lack of diurnal variation and insensitivity towards physical stress further increase the value of the T/17OHP ratio for this use. During the last 5 years, we have investigated several cases with above-normal urinary T/epiT ratios in which T doping seemed highly unlikely. All of these subjects had normal urinary LH concentrations. n one subject, suppression and stimulation tests, including inhibition CLNCALCHEMSTRY,Vol.38, No. 9,

4 amos/. $-T/LH 1 -T.stostSfOns $. amos/ Hyd,osy-proq.sseoas amos/. S-AfldOstflsdons S-T/ TOHP STotsS Sit, on. amos/. :1f\ #{19} U/. S-Lut.u,in9 hormons.1 pmo/. $.1 1\\,,_,,,,jt\\*%\\#{19} = #{19}_ 1 21 i 1 o a o Fig. 1. Serum concentrations oftestosterone (T) and 17a-hydroxyprogesterone (17OHP) and the T/17OHP ratio; of lutelnizinghormone (LH) andthe T/LH ratio; and of-androstene-3,i 7-dione(A-),total estrone (tel), and estradlol-i7p (E2)in sevenhealthymenafterintramuscular njection of 25 mg of testosteroneenanthate Geometric mean (thldc line) and range (thin fines). Values significantlydifferent from day values: *** = P<.O1, ** = P<O.1, * = P<O.5 of testicular steroidogenesis by ketoconazole, showed that the above-normal T/epiT ratio was due to an abnormally low testicular production of epit, probably attributable to genetic factors (Oftebro et a!., unpublished observations). The finding of perfectly normal serum T/17OHP ratios (.5 and 3.59) in this subject further illustrates the value of this index as a suitable marker for T administration. As could have been expected, the ratio between serum T and LH was almost as sensitive a marker fort doping as was the urinary T/epiT ratio. Measurement of the peptide hormone L might give additional valuable 1782 CLNCALCHEMSTRY,Vol.38, No. 9, 1992

5 amos/u U-T/LH 2 2 U-1/.p-T a o U/ma,o 1. U-LH/C,.5 U-1E,/Cr a 1..$..2 i a S S 2 8 Fig.2. Ratios between urinary concentrationsof testosteroneand epitestosterone (TepiT), between LH and creatinine(lh/cr),between testosteroneand LH (T/LH), between total estrone andcreatinine(tel/cr x 1#{176}), and between total estradloland creatinine(te2fcrx 1#{176}) n sevenhealthymen afterntramuscularinjectionof 25 mgof testosteronenanthate Symbols and significance as in Fig. 1 Table 2. Number of the Seven Test Subjects Having Abnormala Serum and Urinary Hormone Vaiues after injection of after njection of 25 mg of testosterone enanthats S-T 5 S-i 7OHP S-T/i7OHP 7 S-LH S-T/LH S-A- S-E2 S-tEl U-T/eprr U-LH/Cr U-T/LH U-tEl/Cr U-tEZ/Cr the normalrange,exceptfor S-17OHP,S-LH,and U-LH/Cr. Abbreviationsas in Table 1. information in suspected cases of T doping. However, given the lack of a reference method for measuring and identifying L, we believe that LH measurements may be less effective in meeting legal challenges. For future antidoping work, it is important that the analytical procedures in the antidoping laboratories be as accurate as possible and that all possible risks for falsely positive results be eliminated. We therefore CLNCAL CHEMSTRY, Vol.38, No. 9,

6 #{19} a #{176}#{176} 13 1#{176}18#{176} Fg.3. Ratios (arithmetic mean ± SEM)between serum testosterone and 17a-hydroxyprogesterone (T/17OHP) in 15 elte ce hockey players and in 2 spectatorsmatchedfor physicalfitnessduringa 26-h tournament suggest supplementary immunological analysis for 17OHP, T, and L in serum in all suspectedcasesof T doping. Although most commercial assay kits are sufficient to assay T in serum from men, this is not necessarily true of all kits for 17OHP, some of which are intended for screening for adrenocortical 21-hydroxylass deficiency and thus have a rather low sensitivity. A highly sensitive immunoassay of 17OHP, preferably including an extraction step to improve specificity, should be used. n cases where the results of these assays may have legal consequences, one can quantify T (18)-and most probably also 17OHP-in serum by isotope dilution mass spectrometry. a Players This work was supported by grants from the nternational Amateur Athletic Foundation and from the Norncolon Confederation of Spain. References exogenous 1. Donike ML, testosterone. Barwald n: KR, Sport: Klosterman Leistung K, und et al. Gesundheit. The detection Heck of H, ed. KOn: Deutsche Artzte-Verlag, 1983: Falk, Palonek E, Bjorkhem. Effect of ethanol on the ratio between testosterone and epitestosterone in urine. Cliii Chem 1988;3: Dehennin L, SchollerR. Depistage de la prise de testosterone comme anaboliant chez lea adolescentspar ladetermination de rapport des excretions urunairesdetestosterone etd epitestosterone. PatholBiol(Paris)199;38:92-2. Spectators. Oftebro H. Evaluating an abnormal steroid inprofile. Lancet 1992;359: Brooks RV, Jeremiah G, Webb WA, etal Detection ofanabolic steroid administration to athletes. J Steroid Biochem 19 79;11: T 6. Kicman AT, Brooks RV, CollyerSC, etal.criteria to indicate testosterone administration. Br J Sports Med 1992: Aakvaag A, Sand T, Opstad PK, Fonnum F. Hormonal changes in serum in young men during prolonged physical strain. Eur J Appl Physiol 1978;39: Al#{233}n M, Rahkila P, Reinilfi M, et al. Androgenic-anabolic steroid effects onserumthyroid, pituitary and steroidhormonesin athletes. Am J Sports Med 1987;15: Al#{233}n M, ReinilAM, Vthko H. Response of serum hormones to androgen administration inpowerathletes. Med SciSports 1985; 17: Tegelman H, CarlstrOm K, Pousette A. Hormone levels in male ice hockey players during a 26-hour cup tournament, hit J Androl 1988;11: Trachtenberg J, Halpern N, Pont A. Ketoconazole, a noveland rapid treatment foradvancedprostate cancer. J Urol 1983;13: Brody 5, CarlstrOm K, Lagrelius A, et al.serum levelsof -androstene-3,17-dione in menstruating and postmenopausal women. Acts Obstet Gynecol Scand 1983;62: CarlstrOm K, Skoldefors H. Determination oftotalestrone in peripheral serum from non-pregnant humans. J SteroidBiochem 1977;8: Stege B., Eriksson A, Henriksson P, et al.orchidectomy or estrogen treatment in prostatic cancer effects on serum levels of adrenal androgens and related steroids. hit J Androl 1987;1: Donike M, Zimmerman J, Barwald KR, et al. Routinebestimmung von Anabolica inham. DtschZ Sportmed 198;35: CrillyRG, FrancisRM, Nordun BEC. Steroid hormones, ageing and bone. Clin EndocrinolMetab 1981;1: Vermeulen A, Verdonck L Radioimmunoassay of17p-hydroxy- 5o-androstan-3-one, -androstene-3,17-dione, dehydroepiandrosterone, l7aiydroxyprogesterone and progesterone and its application to human male plasma. J Steroid Biochem 1976;7: BjOrkhem, Lantto, Svensson L Serum testosterone determinationby masefragmentography. Clin Chim Acta 1975;6: CLNCALCHEMSTRY,Vol.38, No. 9, 1992