Oral Testosterone Administration Detected by Testosterone Glucuronidation Measured in Blood Spots Dried on Filter Paper

Clinical Chemistry 46:4 515 522 (2000) Drug Monitoring and Toxicology Oral Testosterone Administration Detected by Testosterone Glucuronidation Measured in Blood Spots Dried on Filter Paper Shi-Hua Peng, 1 Jordi Segura, 1,2* Magí Farré, 1,3 and Xavier de la Torre 1,3 Background: Blood sampling is not a common practice for sports drug testing. Our aim was to investigate whether dried blood spots on filter paper could be an alternative to plasma samples for monitoring steroid profiles in dope testing. Methods: We collected dried blood spots and plasma from six healthy Caucasian subjects after an oral 120-mg dose of testosterone undecanoate (TU). Nonconjugated testosterone, testosterone glucuronide (TG), androsterone glucuronide (AG), and etiocholanolone glucuronide (EtG) were measured by gas chromatography mass spectrometry in both matrices. 17 -Hydroxyprogesterone (17 OHP) and luteinizing hormone (LH) also were measured in the plasma samples. For comparison, similar measurements were done on samples obtained from the same subjects given 25 mg of testosterone propionate (TP) plus 110 mg of testosterone enanthate (TE) intramuscularly after a wash-out period. Results: After oral TU intake, TG, AG, and EtG increased sharply, whereas nonconjugated testosterone did not change significantly. Results on dried blood spots correlated well with those on plasma. The TG/ testosterone ratio in blood or plasma was verified to be a sensitive and specific marker (significantly increased for up to 8 h after intake; P <0.05) for oral TU intake but not for intramuscular administration of TP plus TE. Little suppression of plasma LH and 17 OHP was observed after a single oral dose of TU. One subject did not show a significant increase of blood TG after oral TU intake. Conclusions: The measurement of glucuronide conjugates in blood and plasma samples is relevant for sports drug testing when analyzing the steroid profile. Dried blood spots collected on filter paper are a suitable alternative to plasma for detecting testosterone abuse. 2000 American Association for Clinical Chemistry In recent years, there has been an increase of doping with endogenous-like substances, such as testosterone and some of its metabolic precursors. Discriminating the exogenous intake of these substances from their endogenous origin constitutes a challenge in today s sports drug testing. Pharmaceutical testosterone preparations are commercially available for oral, percutaneous, and intramuscular administration. The International Olympic Committee criterion for suspecting exogenous testosterone intake, a urinary concentration ratio of testosterone glucuronide to epitestosterone glucuronide (TG/EG) 4 6 (1), needs elaborate follow-up for definitive decisions. There are a few subjects who have physiologically increased urinary TG/EG ratios, whereas other individuals do not show a TG/EG ratio 6 even after exogenous intake of testosterone (2, 3). In addition, oral administration of testosterone produces a high but brief increase in the urinary TG/EG ratio (4), which probably would be difficult to detect (as judged by the criterion TG/EG 6) as an indication of exogenous testosterone intake when urine is collected after a sporting event (which usually includes a 1-h delay plus an additional waiting period for the final collection of urine). Apart from measurement of the 13 C content of urinary testosterone metabolites by isotope-ratio mass spectrometry (5), a theoretically better approach might be to analyze one drop of blood collected immediately after the sporting event. The sample may be 1 Institut Municipal d Investigació Mèdica, E-08003 Barcelona, Spain. 2 Universitat Pompeu Fabra, Barcelona, Spain. 3 Universitat Autónoma de Barcelona, Barcelona, Spain. *Address correspondence to this author at: Drug Research Unit, Institut Municipal d Investigació Mèdica (IMIM), Doctor Aiguader 80, E-08003 Barcelona, Spain. Fax 34-93-2213237; e-mail jsegura@imim.es. Received October 22, 1999; accepted January 18, 2000. 4 Nonstandard abbreviations: TG, testosterone glucuronide; EG, epitestosterone glucuronide; TU, testosterone undecanoate; AG, androsterone glucuronide; EtG, etiocholanolone glucuronide; 17 OHP, 17 -hydroxyprogesterone; T-D 3, testosterone-[16,16,17]-d 3 ; Et-D 4, etiocholanolone-[2,2,4,4]-d 4 ; TP, testosterone propionate; TE, testosterone enanthate; GC, gas chromatography; MS, mass spectrometry; TMS, trimethylsilyl; LH, luteinizing hormone; and TG/T, testosterone glucuronide/testosterone. 515

516 Peng et al.: Dried Blood Spots and Testosterone Abuse obtained by fingerprick or earprick, which currently are used for lactic acid and other measurements. The relevant question is whether suitable information from adequate markers can be obtained from such a small amount of sample. In the present study, one 120-mg dose of testosterone undecanoate (TU) was administrated orally to six healthy Caucasian volunteers. Dried blood spots and plasma samples were collected and studied for biological indicators of exogenous testosterone intake. Materials and Methods references, steroids, and chemicals Testosterone, testosterone glucuronide (TG), androsterone, androsterone glucuronide (AG), etiocholanolone, etiocholanolone glucuronide (EtG), 17 -hydroxyprogesterone (17 OHP), and Escherichia coli -glucuronidase were purchased from Sigma. Testosterone-[16,16,17]-d 3 (T-D 3 ) and etiocholanolone-[2,2,4,4]-d 4 (Et-D 4 ) were kindly provided by the late Prof. M. Donike (Deutsche Sporthochschule, Cologne, Germany). N-Methyl-N-trimethylsilyl-trifluoroacetamide was provided by Macherey-Nagel (Düren, Germany). Ammonium iodide and 2-mercaptoethanol were obtained from Merck. All other reagents were of analytical grade. sample collection Six healthy male Spanish volunteers (subjects 1 6; age, 27.2 2.1 years; weight, 73.4 4.0 kg; height, 1.75 0.03 m, mean SD) received a single oral 120-mg dose of TU (Androxon, three 40-mg capsules; Organon). Venous blood (10 ml) was collected from the antecubital vein before and after administration. Sampling times were 0.5h( 0.5 h) before administration, at administration (0 h; 0900 in the morning), and 0.5, 1, 1.5, 2, 3, 4, 6, 8, 10, and 12 h after administration. Aliquots of 20 L of fresh blood were spotted onto filter paper (Whatman 41), and then air dried, sealed in plastic bags, and stored at 20 C until analysis. The remaining heparinized whole blood was centrifuged for preparing plasma samples. To check the stability of steroids in blood spots dried on filter paper, aliquots of blood spot samples (stored at 20 C) from each of the six volunteers obtained 1 h after the oral TU administration were removed from the freezer and kept at room temperature for 1 week. These aliquots were analyzed for steroid profiles simultaneously with the corresponding aliquots stored at 20 C from the time of collection. After a wash-out period of 3 months, the same six subjects received a single intramuscular dose of Testoviron Depot 100 [25 mg of testosterone propionate (TP) plus 110 mg of testosterone enanthate (TE); Schering AB]. Blood spots and plasma samples were collected and stored as described above. The sampling times were on days 3 (0900), 2 (0900), 1 (0900) before administration; at 0 h (0900), 6 h (1500), and 12 h (2100) after administration (on day 0), on days 1 and 2 at 0 h (0900) and 8 h (1700) on days 3 and 4; and at 0 h (0900) on days 5, 6, 7, 8, 9, 11, 13, 15, 17, and 19 after administration. Approval for the above-described clinical trials was granted by the local ethics committee (CEIC no. 94/467) and by the Spanish Health Ministry (DGFPS no. 95/75). sample preparation Dried blood spots. For the extraction of nonconjugated steroids from blood spots on filter paper, two dried blood spots were cut with scissors in small pieces into a tube, and 2.5 ml of sodium phosphate buffer (0.2 mol/l, ph 7) was added. T-D 3 and Et-D 4 were used as internal standards at final concentrations of 0.9 and 5 g/l, respectively, in buffer. Immediately after the subsequent addition of 100 L of 3 mol/l potassium hydroxide, the samples were extracted twice with 5 ml of n-hexane:ethyl acetate (7:3, by volume), using a rocking mixer for 20 min to extract nonconjugated steroids. It has been verified that under these conditions there is no hydrolysis of testosterone esters, which could increase the actual amount of testosterone initially present in the sample. The organic phases were pipetted out into a tube, mixed, and washed with 1 ml of 50 ml/l acetic acid and then with 1 ml of distilled water. After the solvent was evaporated, the residue was kept in a desiccator maintained at 60 C and 60 kpa (phosphorous pentoxide was used as desiccant, which was changed when formation of phosphoric acid was quite notable) at least 30 min before derivatization (see below). For the extraction of glucuronide-conjugated steroids, the extracted potassium hydroxide-water phase indicated above was neutralized with 1 mol/l hydrochloric acid, incubated with 6Uof -glucuronidase at 37 C overnight after a second addition of the same amounts of the same (T-D 3 and Et-D 4 ) internal standards, and then extracted with t-butyl methyl ether after the aqueous solution was adjusted to approximately ph 10 with 50 g/l potassium carbonate. After the solvent was evaporated, the residue was kept in a desiccator at least 30 min before derivatization. The dried extracts were derivatized for gas chromatography mass spectrometry (GC-MS) analysis. Trimethylsilyl (TMS) derivatives of steroids in both extracts were prepared by dissolving the dried residues in 50 L of a mixture containing1lofn-methyl- N-trimethylsilyl-trifluoroacetamide, 2 g of ammonium iodide, and 6 ml of 2-mercaptoethanol and heating at 60 C for 30 min. Each solution (1 2 L) was directly analyzed by GC-MS. Endogenous steroids (testosterone, androsterone, and etiocholanolone) were quantified using the responses of the target compounds relative to the internal standards, calculated from an extracted sample of distilled water containing 1.2 g/l testosterone, 20 g/l androsterone, and 20 g/l etiocholanolone [extracted with hexane-ethyl acetate (7:3, by volume)]. Plasma. A 1-mL plasma sample was diluted with 1.5 ml of sodium phosphate buffer (0.2 mol/l, ph 7) and subjected to a procedure similar to that for the analysis for urine in

Clinical Chemistry 46, No. 4, 2000 517 sports drug testing (6), with modifications. Briefly, the diluted plasma sample was loaded on a Detectabuse TM column pretreated with methanol and water. The column was washed with water, and the steroids were eluted with methanol. The solvent was evaporated, and the residue was reconstituted in 1 ml of sodium phosphate buffer (0.2 mol/l, ph 7). Nonconjugated steroids (containing testosterone and 17 OHP) were extracted with 5 ml of t-butyl methyl ether, evaporated with nitrogen, and kept in a desiccator until derivatization. The remaining aqueous phase (containing TG, AG, and EtG) was then hydrolyzed with6uof -glucuronidase and extracted using the same procedure that was used for the blood-spot samples. Quantification of steroids (including 17 OHP) in both fractions of plasma was performed as for the blood spots samples, with modifications that included a water calibration mixture containing 5 g/l 17 OHP, 12 g/l testosterone, 200 g/l androsterone and 200 g/l etiocholanolone, and monitoring the tris-tms derivative of 17 OHP in the nonconjugated fraction extracted from plasma. Pooled plasma samples and pooled plasma samples supplemented with testosterone, TG, AG, and 17 OHP were run in each analytical batch as qualitycontrol samples. The intra- and interassay variabilities (CVs) were from 2.7% (for 12.7 nmol/l testosterone) to 15% (for 0.82 nmol/l TG). Accuracy (relative error) was from 0.8% (for 29.6 nmol/l EtG) to 15.1% (for 135 nmol/l AG). Luteinizing hormone (LH) was analyzed by microparticle enzyme immunoassay (Abbott) directly in plasma according to the manufacturer s instructions. gc-ms analysis A Hewlett-Packard 5890 II GC model fitted with a HP 7673A autosampler was connected to a HP 5970 massselective detector. The separation was carried out using a methyl silicone fused-silica capillary column [HP Ultra-1; 17 m 0.2 mm (i.d.); film thickness, 0.11 m] with the following oven temperature program: initial temperature of 180 C, ramped to 230 C at a rate of 3.0 C/min, then to 310 C at a rate of 40 C/min, and held for 3 min. Helium was used as carrier gas with a flow rate of 0.8 ml/min (measured at 180 C). The injector (operated in 10:1 split mode) and the interface were maintained at 280 C. The mass spectrometer was operated in selectedion monitoring acquisition mode with one or more ions selected for each substance (m/z 432, 434, and 434 for bis-tms derivatives of testosterone, androsterone and etiocholanolone, respectively; m/z 435 and 438 for bis-tms derivatives of T-D 3 and Et-D 4, respectively; m/z 546 for tris-tms derivative of 17 OHP). statistical analysis The concentrations in blood collected on filter paper and in plasma, and their ratios (log-transformed) after administration of TU were compared to their basal values by ANOVA for testing of significant differences. Results blood steroid profile analyzed in dried blood spots The nonconjugated testosterone measured in dried blood spots after oral administration of TU did not change significantly. Instead TG, which had an extremely low basal concentration in dried blood spots (much lower than nonconjugated testosterone), increased substantially after oral administration of TU in all of the subjects (Table 1) except one (subject 3). Results for the abnormal subject are discussed separately. Other main metabolites of testosterone, such as AG and EtG, were also significantly increased in these five subjects. Thus, hydrolysis of the dried blood sample plays an important role in the interpretation of results. TG, AG, and EtG increased significantly (compared with their basal concentrations) up to 8, 8, and 12 h, respectively (Fig. 1). Because of its extremely low basal concentration, TG showed the highest relative increase. Accordingly, for this report, the concept of steroid profile will refer specifically to those concentrations of testosterone, TG, AG, and EtG. Table 1. Time course of blood steroid concentrations (mean SE; n 5) and TG/T ratio after a single oral 120-mg dose of TU. Time, h Testosterone, nmol/l TG, nmol/l AG, nmol/l EtG, nmol/l TG/T 0.5 17.6 2.5 4.1 0.7 111 16 73.4 16.0 0.23 0.03 0 17.8 3.8 2.5 1.2 115 15 72.7 18.5 0.10 0.01 0.5 19.2 2.6 93.2 42.5 a 352 85 a 329 96 a 4.9 2.0 a 1 21.0 3.3 310 89 a 2279 502 a 1814 605 a 14.2 2.2 a 1.5 18.7 2.5 241 73 a 2576 225 a 1457 354 a 12.0 2.5 a 2 19.4 3.9 127 39 a 1871 179 a 883 219 a 6.2 1.3 a 3 16.3 2.6 43.7 13.1 a 1001 99 a 403 100 a 2.7 0.8 a 4 17.6 3.5 19.4 7.2 a 503 84 a 217 59 a 0.97 0.26 a 6 15.5 2.8 13.2 3.0 a 261 37 a 154 34 a 0.82 0.09 a 8 13.6 3.4 5.8 1.0 a 177 27 a 159 47 a 0.57 0.19 a 10 12.5 1.9 2.4 0.8 111 13 107 28 a 0.24 0.09 12 14.7 2.7 2.2 0.7 111 16 107 27 a 0.19 0.07 a P 0.05 compared with basal values.

518 Peng et al.: Dried Blood Spots and Testosterone Abuse at room temperature were 5.9%, 2.8%, 6.9%, and 4.8%, respectively. Fig. 1. Mean relative increases ([increase/basal concentration] 100; n 5) of TG, AG, and EtG obtained from blood spots dried on filter paper after oral administration of 120 mg of TU. y axis is in logarithmic scale. correlation with plasma steroid profile The above-mentioned steroids and their glucuronides were analyzed in plasma samples obtained at the same time. Similar patterns between blood collected on filter paper and plasma were observed, with slightly higher values obtained in blood (Fig. 2). stability of steroid glucuronides in dried blood spots It is known that androgenic anabolic steroids, such as testosterone and 17 OHP, are stable in dried blood spots (7, 8). The stability of the glucuronides of some steroids in dried blood spots stored for 1 week at room temperature, compared with aliquots maintained in a freezer ( 20 C), was verified (Fig. 3). All four steroids measured (TG, testosterone, AG, and EtG) were sufficiently stable during the study period. The mean relative changes in TG, testosterone, AG, and EtG for the six subjects after 1 week specificity of tg/testosterone as a marker of oral administration To check whether the ratio of TG to testosterone (TG/T) in dried blood spots could be a specific marker for oral vs intramuscular administration of testosterone, dried blood-spot samples from the same six subjects, who received intramuscular injections with a single combined dose of 25 mg of TP plus 110 mg of TE were partially analyzed for steroid profiles. In the first step, all of the dried blood spots from one subject (subject 2) after intramuscular administration were studied. Testosterone, AG, and EtG, but not TG, were clearly increased after intramuscular administration. The highest increase of nonconjugated testosterone was found in those blood spots on filter paper collected between 1.25 and 2.25 days after intramuscular administration of TP plus TE, which is in good agreement with the pharmacokinetic profile known for the same type of preparation and obtained by either RIA (9) or GC-MS (10). In the next step, samples obtained 1 day before ( 1 day), and 2.25 and 19 days after ( 2.25 and 19 days) intramuscular administration from each subject were selected for a more comprehensive analysis. For all the six subjects, nonconjugated testosterone at 2.25 days was increased (relative increases, 100 200%) compared with its basal concentration, whereas no meaningful changes of TG were observed. However, clear increases of AG and EtG were also observed in the blood collected 2.25 days after intramuscular dosing of testosterone esters. The results are shown in Fig. 4. The results obtained from corresponding plasma samples are also presented. When comparing these results with the changes in steroid profiles observed in blood after oral intake of testosterone, testosterone glucuronida- Fig. 2. Mean concentrations (n 5) of testosterone, TG, AG, and EtG in blood spots and plasma after oral administration of 120 mg of TU. y axis is in logarithmic scale. f, testosterone in dried blood spots;, testosterone in plasma samples; Œ, TG in dried blood spots;, TGin plasma samples;, AG in dried blood spots;, AG in plasma samples; F, EtG in dried blood spots; E, EtG in plasma samples. Conc., concentration.

Clinical Chemistry 46, No. 4, 2000 519 Fig. 3. Stability of TG, nonconjugated testosterone, AG, and EtG in blood spots dried on filter paper after oral administration of 120 mg of TU. One aliquot was kept at 20 C in freezer ( ); another aliquot was kept at room temperature for 1 week (f). T, nonconjugated testosterone; sub, subject; Conc., concentration. tion in human blood appears to be a specific marker for oral testosterone intake. other markers (tg/lh, tg/17 ohp, and tg/eg) Recently, plasma T/LH and T/17 OHP have been suggested as sensitive markers of testosterone administration, based on the observation that lower concentrations of plasma LH and 17 OHP could be induced by intramuscular administration of exogenous testosterone (11 13). However, a single oral administration of TU seems unlikely to cause significant suppression of LH secretion and steroidogenesis. LH concentrations showed no meaningful changes during 0 24 h post administration, whereas 17 OHP showed only a small reduction. As shown in Table 2, plasma TG/17 OHP was affected, but mainly because of an increase in TG rather than a decrease in 17 OHP. In addition, plasma TG/17 OHP and TG/LH were significantly increased above their basal concentrations until 10 and 6 h, respectively, post administration. TG/EG in urine was 6 in all subjects but subject 3 in the period 0 4 h, and in three subjects in the period 4 8 h post administration (14). outlier subject One subject (subject 3; Table 3) did not show significant changes of TG in plasma but did show a transient (1 2 h) increase in nonconjugated testosterone after TU administration. The AG and EtG concentrations in his dried blood spots on filter paper did present remarkable increases. It has been verified that his urinary TG/EG ratio remained 2 during the whole controlled study (14). Discussion The clear increase of TG measured in dried blood spots of five of six subjects after oral administration of TU appears to be an efficient marker for exogenous intake of testosterone. Regarding the extremely low basal concentration of TG and relative higher basal concentration of testosterone, the TG/T ratio is a sensitive marker for oral TU administration. In the late 1970s, when oral TU preparations were studied as potential new supplements for hypogonadal deficiency or fertility control, clinical endocrinologists monitored extensively the effect of oral TU on the plasma concentration of nonconjugated testosterone after oral TU administration (15 17). It was reported that after a single oral dose of TU, there were significant increases of nonconjugated testosterone in both saliva and serum, as detected by RIA (18, 19), which seems in apparent disagreement with the absence of significant changes of nonconjugated testosterone in plasma and dried blood spots observed in the present study by GC-MS. It is

520 Peng et al.: Dried Blood Spots and Testosterone Abuse Table 2. Time course of some plasma markers (mean SE; n 5) after a single oral 120-mg dose of TU. Time, h TG/T TG/17 OHP TG/LH 0.5 0.20 0.09 0.55 0.03 0.44 0.05 0 0.20 0.08 0.76 0.09 0.46 0.07 0.5 4.7 1.1 a 36.2 12.6 a 29.3 7.8 a 1 10.6 1.7 a 101 40 a 52.7 18.9 a 1.5 8.8 2.8 a 43.8 11.7 a 33.9 11.3 a 2 4.3 1.7 a 25.9 4.8 a 11.6 4.6 a 3 1.4 0.3 a 9.0 2.1 a 4.6 1.8 a 4 0.64 0.10 a 6.0 1.2 a 2.2 0.5 a 6 0.52 0.28 a 2.7 1.3 a 0.75 0.14 a 8 0.31 0.10 a 1.9 0.4 a 0.71 0.14 10 0.27 0.12 a 1.6 0.3 a 0.42 0.07 12 0.18 0.07 1.3 0.3 0.47 0.13 a P 0.05 compared with basal values. Fig. 4. Mean dried blood-spot (A) and plasma (B) steroid profile [testosterone, TG, AG, and EtG; n 6] after administration of one intramuscular combined dose of 25 mg of TP plus 110 mg of TE. u, 1 day before administration ( 1 day); f, 2.25 days after administration ( 2.25 days);, 19 days after administration ( 19 days). T, testosterone; Conc., concentration. feasible that some cross-reactivity ( 0.2%) (18) of the RIA antiserum used in those previous studies for TU could produce an apparent increase of nonconjugated testosterone, because the unchanged TU is also extracted by the solvent extraction method used in those studies. In fact, the existence of substantial amounts of TU in plasma after oral TU administration is known (20) and was confirmed in those subjects (14). The additional thin-layer chromatographic clean-up of nonconjugated testosterone preceding the RIA assay in some of these previous studies (18), however, seems to preclude such an easy explanation. Differences in the health status of the subjects in the present study (healthy) and the previous study (patients) might be relevant to the differences between the results, perhaps because the differences in health status affected the preferential absorption route of TU (see below). Although no definitive explanation could be given for the above-mentioned apparently conflicting results obtained by RIA and GC-MS techniques, the changes in nonconjugated testosterone and TG in plasma in the present study do agree with the results obtained in a controlled study by Dehennin and Pérès (21), who used a GC-MS analytical technique to monitor seven healthy subjects who received a single oral combined dose of 40 mg of TU plus 1.5 mg of epitestosterone undecanoate. Moreover, a sharp increase of TG compared with nonconjugated testosterone in the plasma of one healthy subject after a single oral dose of 80 mg of TU was observed by Giagulli et al. (22), using RIA assay after paper chromatographic purification. It is generally considered that esterification of testosterone with undecanoic acid shifts the route of absorption from the portal vein to the lymph system (23) and that TU enters the blood via the thoracic duct. However, the remarkable increase of TG in plasma after oral TU administration observed here suggests that a substantial percentage of TU may be absorbed via the portal vein, probably losing the undecanoate group in the intestine, whereas the remaining nonmetabolized TU could be absorbed via the lymphatic system, as suggested by Coert et al. (24) in a controlled study of rats after oral TU administration. This hypothesis is supported by the observation that plasma TG was sharply increased post Table 3. Time course of blood steroid concentrations from subject 3 after a single oral 120-mg dose of TU. Time, h Testosterone, nmol/l TG, nmol/l AG, nmol/l EtG, nmol/l TG/T 0.5 11.4 5.7 46.3 39.5 0.50 0 12.6 3.8 54.7 38.5 0.10 0.5 8.9 6.1 253 233 0.68 1 17.1 4.2 1276 815 0.25 1.5 18.3 3.4 974 476 0.18 2 19.4 7.0 1330 725 0.36 3 13.6 6.7 889 338 0.49 4 8.7 3.7 349 140 0.43 6 10.9 1.7 130 68.6 0.16 8 8.5 4.9 102 57.4 0.58 10 8.3 5.0 117 52.8 0.60 12 10.8 4.1 57.5 45.0 0.38

Clinical Chemistry 46, No. 4, 2000 521 administration, with the mean peak concentration being reached at 1 h after the dosing, whereas unchanged TU was found in plasma of all the six subjects receiving oral TU, with the mean peak concentration appearing at 3 h after the dosing (14). AG and EtG were increased in blood of all six subjects studied either after oral administration of TU or after intramuscular administration of combined TP plus TE. Thus, increases of AG and EtG in blood, compared with their basal concentrations, seem to be indicative of exogenous testosterone intake, regardless of the administration route. Given the relevance of glucuronidation of testosterone, androsterone, and etiocholanolone, analysis with and without hydrolysis of blood and plasma samples appears necessary for obtaining the proper steroid profile after exogenous testosterone intake, especially after oral administration. However, one subject showed no increase in TG but a very short increase in nonconjugated testosterone, and longer-lasting clear increases of AG and EtG after oral intake of TU. This subject did show a similar total recovery of testosterone excreted in urine after TU administration compared with the other five subjects (data not shown), as calculated by the addition of its main urinary metabolites, thus ruling out the possibility of poor gastrointestinal absorption of TU. Some other possibilities for his abnormal kinetic results could be a faster metabolism of testosterone to androsterone and etiocholanolone, a specific enzyme deficiency for testosterone glucuronidation in splanchnic organs, or again, a different balance between the portal vein vs lymphatic absorption, as mentioned above. At present, blood sampling is not a common practice in sports drug testing, although expectations for its future use are strong, especially for detecting the administration of peptide hormones (25). In the present study, the steroid profile (testosterone, TG, AG, and EtG) measured in blood spots dried on filter paper correlated well with that obtained from plasma, with the advantages of slightly higher concentrations being found in dried whole blood, which suggest that blood cells may retain small portions of testosterone and its glucuronide-conjugated metabolites. Thus, dried blood spots on filter paper could be an optimal alternative to plasma for assessing individual steroid profiles, with the advantages of the small volume of blood needed, easy collection by fingerprick or earprick, and less invasiveness. Blood TG/T obtained from dried blood spots was increased until 8 h after administration, whereas other plasma markers, TG/ 17 OHP and TG/LH, were significantly increased until 10 and 6 h, respectively, all confirming the signs of administration of testosterone. It has also been verified in the present study that steroids in dried blood spots are quite stable and that the samples can be stored easily and transported from the field to laboratories. Additionally, blood spots may be subjected to individual DNA fingerprinting, which may offer definitive proof of the authenticity of the sample. Taking into account the abovementioned merits of blood spots dried on filter paper as a sampling technique, it could be feasible that a database of basal blood steroid concentrations for each athlete be established and filed as historical fingerprints for doping control. Thus, changes in blood concentrations of TG, AG, and EtG, compared with historical basal concentrations for each particular athlete, would be easily detectable in dried blood spots. The existence of intact testosterone esters, which also can be detected in blood spots dried on filter paper, may offer additional unequivocal confirmation of testosterone intake. In fact, traces of the unchanged testosterone ester of orally administrated TU have been detected in some of these dried blood-spot samples, as reported previously (14). In conclusion, the ability of dried blood spots to provide reliable information on the glucuronidation of testosterone and its main metabolites (androsterone and etiocholanolone), which can be obtained by analyzing only one or two dried blood spots on filter paper, make this sample material a good alternative to plasma as a complement to urinalysis for detecting testosterone abuse. However, it should be kept in mind that the present proposed markers are derived from a small group of subjects. For the wide application of blood TG/T, AG, and EtG to detect testosterone abuse in doping tests, relevant population reference values and intraindividual variations should be taken into consideration and be studied carefully. The usefulness of easily obtained capillary blood (from a fingerprick or earprick), compared with drops of venous blood (studied in the present work), should also be verified for widespread applicability of the proposed approach. We thank the Spanish and Catalonian research administrations (projects FIS 96/1050, CICYT SAF 97-068, and CIRIT DOGC 2320) for financial support. S-H. Peng also thanks the Spanish Ministry of Science and Education for a postdoctoral fellowship. We thank Pere Roset, Marta Mas, and Esther Menoyo for their work in the clinical trials and some assistance in statistical analysis. References 1. Donike M, Barwald KR, Klostermann K, Schänzer W, Zimmermann J. Nachweis von exogenen Testosteron. In: Heck H, Hollmann W, Liesen H, Rost R, eds. Sport: Leistung und Gesundheit. Köln: Deutscher Arzte Verlag, 1983:293 300. 2. Geyer H, Schänzer W, Mareck-Engelke U, Donike M. Factors influencing steroid profile. In: Donike M, Geyer H, Gotzmann A, Mareck-Engelke U, eds. Recent advances in doping analysis (3): proceedings of the Manfred Donike Workshop on Dope Analysis. Köln: Sport und Buch Strauss, 1996:95 113. 3. de la Torre X, Segura J, Yang ZY, Ye L, Wu M. Testosterone detection in different ethnic group. In: Schänzer W, Geyer H, Gotzman A, Mareck-Engelke U, eds. Recent advances in doping

522 Peng et al.: Dried Blood Spots and Testosterone Abuse analysis (4): proceedings of the Manfred Donike Workshop on Dope Analysis. Köln: Sport und Buch Strauss, 1997:71 89. 4. Wright F, Lafarge JP, Antréassian J, Lagoguey M, Péres G. Long term study of steroid and peptidic hormones in the plasma of healthy young men under controlled testosterone undecanoate therapy. In: Hemmersbach P, Birkeland KI, eds. Blood samples for doping control. Oslo, Norway: Pensumtjeneste, 1994:65 74. 5. Horning H, Geyer H, Machnik M, Schänzer W, Hilkert A, Oesselmann J. Detection of exogenous testosterone by 13 C/ 12 C analysis. In: Schänzer W, Geyer H, Gotzmann A, Mareck-Engelke U, eds. Recent advances in doping analysis (4): proceedings of the Manfred Donike Workshop on Dope Analysis. Köln: Sport und Buch Strauss, 1997:275 83. 6. Segura J, Pascual JA, Ventura R, Ustaran JI, Cuevas A, Gonzalez R. International cooperation in analytical chemistry: experience of antidoping control at XI Pan American Games. Clin Chem 1993; 39:836 45. 7. Pang S, Murphey W, Levine LS, Spence DA, Leon A, Lafranchi S, et al. A pilot newborn screening for congenital adrenal hyperplasia in Alaska. J Clin Endocrinol Metab 1982;55:413 20. 8. Howe CJ, Handelsman DJ. Use of filter paper for sample collection and transport in steroid pharmacology. Clin Chem 1997;43: 1408 15. 9. Behre HM, Nieschlag E. Comparative pharmacokinetics of androgen preparations: application of computer analysis and simulation. In: Nieschlag E, Behre HM, eds. Testosterone action, deficiency, substitution. Berlin: Springer-Verlag, 1990:115 35. 10. de la Torre X, Segura J, Polettini A. Detection of testosterone esters in human plasma by GC/MS and GC/MS/MS. In: Donike M, Geyer H, Gotzmann A, Mareck-Engelke U, eds. Recent advances in doping analysis (2): proceedings of the Manfred Donike Workshop on Dope Analysis. Köln: Sport und Buch Strauss, 1995:59 80. 11. Kicman AT, Brooks RV, Collyer SC, Cowan DA, Nanjee NN, Southan GJ, et al. Criteria to indicate testosterone administration. Br J Sports Med 1990;24:53 64. 12. Carlström K, Palonek E, Garle M, Oftebro H, Stanghelle J, Björkhem I. Detection of testosterone administration by increased ratio between serum concentrations of testosterone and 17 hydroxyprogesterone. Clin Chem 1992;38:1779 84. 13. Palonek E, Gottlieb C, Garle M, Björkhem I, Carlström K. Serum and urinary markers of exogenous testosterone administration. J Steroid Biochem Mol Biol 1995;55:121 7. 14. Peng SH, Segura J, Farré M, González JC, de la Torre X. Overall biological markers of oral testosterone undecanoate misuse. In: Schänzer W, Geyer H, Gotzman A, Mareck-Engelke U, eds. Recent advances in doping analysis (6): proceedings of the Manfred Donike Workshop on Dope Analysis. Köln: Sport un Buch Strauss, 1999:185 203. 15. Nieschlag E, Hoogen H, Bolk M, Schuster H, Wickings EJ. Clinical trial with testosterone undecanoate for male fertility control. Contraception 1978;18:607 14. 16. Geere G, Jones J, Atherden SM, Grant DB. Plasma androgens after a single oral dose of testosterone undecanoate. Arch Dis Child 1980;55:218 20. 17. Maisy NM, Bingham J, Marks V, English J, Chakraborty J. Clinical efficacy of testosterone undecanoate in male hypogonadism. Clin Endocrinol 1981;14:625 9. 18. Nieschlag E, Mauss J, Coert A, Kicovic P. Plasma androgen levels in men after oral administration of testosterone or testosterone undecanoate. Acta Endocrinol 1975;79:366 74. 19. Schürmeyer T, Wickings EJ, Freischem CW, Nieschlag E. Saliva and serum testosterone following oral testosterone undecanoate administration in normal and hypogonadal men. Acta Endocrinol 1983;102:456 62. 20. de la Torre X, Segura J, Polettini, Montagna M. Detection of testosterone esters in human plasma. J Mass Spectrom 1995; 30:1393 404. 21. Dehennin L, Pérès G. Plasma and urinary markers of oral testosterone misuse by healthy men in presence of masking epitestosterone administration. Int J Sports Med 1996;17:315 9. 22. Giagulli VA, Verdonck L, Desltpere JP, Giorgino R, Vermeulen A. Comparison of plasma androgen glucuronide levels after percutaneous or peroral androgen treatment in men: evidence for important splanchnic contribution to plasma 17 -hydroxyandrogen glucuronides. J Clin Endocrinol Metab 1993;76:429 31. 23. Horst HJ, Höftge WF, Dennis M, Coert A, Geelen J, Voigt KD. Lymphatic absorption and metabolism of orally administered testosterone undecanoate in man. Klin Wochenschr 1975;54: 875 9. 24. Coert A, Geelen J, de Visser J, van der Vies J. The pharmacology an metabolism of testosterone undecanoate (TU), a new orally active androgen. Acta Endocrinol 1975;79:789 800. 25. Segura J. Sports. In: Karch SB, ed. Drug abuse handbook. Boca Raton, FL: CRC Press, 1998:641 720.