Transdermal testosterone delivery in castrated Yucatan minipigs: pharmacokinetics and metabolism

Journal of Controlled Release 52 (1998) 89–98

Transdermal testosterone delivery in castrated Yucatan minipigs: pharmacokinetics and metabolism Qing-Feng Xing, Senshang Lin, Yie W. Chien* Controlled Drug-Delivery Research Center, College of Pharmacy, Rutgers University, Piscataway, NJ 08854, USA Received 7 July 1997; received in revised form 19 September 1997; accepted 1 October 1997

Abstract The feasibility of using the castrated Yucatan minipig as a hypogonadal animal model to investigate the transdermal controlled systemic delivery of testosterone was studied. During a 24 h application of a testosterone transdermal delivery device (T-TDD), serial blood samples were withdrawn from the minipigs, without anesthesia, at predetermined time intervals and the plasma concentrations of testosterone as well as its major metabolites, dihydrotestosterone and estradiol, were assayed by radioimmunoassay. The compartmental pharmacokinetic modeling analysis of the plasma profiles of total testosterone indicated that as much as 92% of the total testosterone dose released from the T-TDD had been delivered transdermally into the systemic circulation during the initial rapid input period (the first 11 h of the application), while only 8% was delivered during the slow input period (up to 23 h). Good correlation was observed between the in vivo input doses [1.9 (60.2), 4.8 (60.2) and 6.4 (60.5) mg / day], determined by the Wagner-Nelson equation, and the daily doses released [1.96 (0.2), 4.7 (60.2) and 6.6 (60.5) mg / day, respectively, for 1, 2, and 3 units of T-TDD]. While the in vivo rate of input in the castrated minipigs was observed to be similar to that in hypogonadal men treated with the T-TDD during the first 8 h period, the input rate was found to be slower during the last 12 h. The agreement could suggest that the mechanism for the transdermal systemic delivery of testosterone in the castrated minipig could be similar to that in the hypogonadal men. However, the plasma testosterone profiles attained in the castrated minipigs were observed to be similar to, but slightly lower than that in the hypogonadal men reported in the literature. The DCmax (baseline normalized peak plasma concentration) and DCavg (baseline normalized average plasma concentration) data in the castrated minipigs were 40 and 44%, respectively, of that in hypogonadal men. The |2.4 fold lower values in DCmax and DCavg data could result from the difference in the clearance rate of testosterone which is |2.8 fold higher in minipigs than in the human. Despite the difference in clearance rate, the castrated minipigs could be a suitable large animal model for studying the pharmacokinetics of testosterone delivered transdermally in humans with hypogonadism.  1998 Elsevier Science B.V. Keywords: Testosterone; Transdermal; Miniature swine; Pharmacokinetics; Metabolism

1. Introduction In recent decades, transdermal drug delivery has been a active field of biomedical research with rapid *Corresponding author.

development [1]. Since the 1980s, several kinds of transdermal delivery drugs have been marketed in the USA [2]. However, most of them were not evaluated in a suitable large animal model for pharmacokinetic analysis before clinical study. The physiological similarity of swine to humans,

0168-3659 / 98 / $19.00  1998 Elsevier Science B.V. All rights reserved. PII S0168-3659( 97 )00190-9

90

Q.-F. Xing et al. / Journal of Controlled Release 52 (1998) 89 – 98

coupled with the development of effective and gentle handling techniques, have led to increased use of swine as a reliable large animal model for studying pharmacokinetic and pharmacodynamic properties of various drugs under trial [3,4]. It has the potential to provide a good prediction of clinical performance. Furthermore, some studies demonstrated that the swine’s skin permeation is more comparable to that of humans than other animals such as rats, rabbits and dogs [5]. The Yucatan minipig is even more similar to humans, for it has sparse hair, making it easy to put patches tightly on the skin. In this investigation, a castrated Yucatan minipigs model was used to evaluate the pharmacokinetics and metabolism of testosterone administered transdermally. The results obtained were compared with the clinical data reported in the literature [6–8] to establish animal–human correlation.

2. Materials and methods

2.1. Materials Testosterone transdermal delivery devices (TTDDs, Androderm  , Smithkline Beecham Pharmaceuticals, Philadelphia, PA) were obtained commercially from local pharmaceutical suppliers. Testosterone, as the reference standard, was purchased from Sigma Chemical Co., (St. Louis, MO). Acetonitrile and methanol used for the HPLC were purchased from Fisher Scientific (Fair Lawn, NJ).

2.2. Animals Three male Yucatan miniature swine were purchased from Charles River Laboratories (Wilmington, MA). The animals were housed individually in a pig pen (approximately 3.5 by 7.0 ft) and had free access to fresh water. The animals were fed a standard, commercial pig diet twice daily ad libitum and exposed to automated 12 h lighting cycles. All synchronizers, including the feeding schedule, temperature (68-708F) and relative humidity (50%), etc. were fixed. Initially, considerable time was spent hand feeding and handling each pig to acclimatise it to its surroundings.

2.3. Castrated minipig model The animals were castrated surgically at 2 months of age and allowed to recover from the surgery for at least another month before the studies. The plasma testosterone levels were assayed before and after the castration to evaluate the results of castration.

2.4. Preparation of animals On the day of the experiment, the minipig was prepared as described elsewhere [9]. In brief, each minipig was put in an upright position and slightly restrained in a sling (Charles River Laboratories, Wilmington, MA). One nonthrombogenic PE-10 tubing (non-radiopaque polyethylene micro-tubing, Clay Adams, Division of Becton Dickinson, Parsippany, NJ), coated on internal surface with tridodecylmethylammonium chloride-heparin complex (Polysciences, Inc., Warrington, PA), was cannulated into the vein of the ear. The cannulated tubing allowed easy serial blood-sampling during the study.

2.5. Animal study Following blood sampling to establish the baseline level, T-TDDs were applied on the dorsal region of the swine for a period of 24 h. The system contains 12.2 mg of testosterone USP, dissolved in an alcohol based gel and has a 7.5 cm 2 of central drug delivery reservoir surrounded by a peripheral adhesive area [7]. The swine was restrained in the sling for several periods (8–12 h each) over the entire course of the experiment. However, during regular meal-time periods (|2 h), the swine was released from the sling and allowed food intake. Serial blood-samples (0.5– 1.0 ml each) were withdrawn via the PE-10 tubing at predetermined time intervals for a period of 28 h. The patency of PE-10 tubing was maintained by heparin lock flush solution (Abbott Hospital, Inc., North Chicago, IL) during the blood sampling period. The blood samples were centrifuged immediately following collection and plasma samples were each separated and stored in a freezer at 2208C until the radioimmunoassay of testosterone and its metabolites. A recovery period of one to two weeks was

Q.-F. Xing et al. / Journal of Controlled Release 52 (1998) 89 – 98

allowed between the experiments with the same minipig. Each of minipigs received triplicate experiments of three different transdermal doses on nine occasions, and the sequence of administration to the minipig was randomized. A total of ten experiments (triplicate of three doses plus one control) was performed using the same swine, and the study comprised a total of 30 experiments. The mean (6S.D.) body weight of swine were 24 (68), 20 (65) and 21 (66) kg, respectively, for 1, 2 and 3 units, and 11 (62) and 29 (61) kg, respectively, at the time of the beginning and the end of the study.

2.6. Radioimmunoassay Plasma total and free testosterone as well as dihydrotestosterone (DHT) and estradiol (E2), its major metabolites, were measured by radioimmunoassay using the kits purchased from Diagnostic Systems Laboratories Inc., (Webster, TX). The total testosterone concentration was determined for all samples, while the concentrations of free testosterone, DHT, and E2, were determined additionally for samples obtained from the experiments in which 3 units of testosterone system were applied.

91

2.8. Pharmacokinetic analysis 2.8.1. Noncompartmental parameters analysis Cmax and t max are, respectively, the peak plasma concentration and the time at which Cmax has been attained during the dosing period. The area under the plasma concentration-time curves (AUC) was calculated by the trapezoidal rule for a period from 0 to 24 or 28 h. 2.8.2. Compartmental parameters analysis A general dynamic model developed for describing the plasma profile of drug concentration following transdermal delivery [10] can be applied in this study for data analysis. The absorption and disposition of testosterone in the body can be illustrated by a one-compartment model represented in Fig. 1. The following multiple constant inputs are assumed: the initial rapid input (Fri ) of testosterone released from the T-TDD upon application to the skin and the

2.7. Assay of residual testosterone in used patches Residual testosterone contents in all the patches used during the animal study were first extracted into methanol and then analyzed by HPLC. A HPLC assembly by Millipore Corporation (Marlborough, MA) was used. It consisted of a HPLC pump (Waters, Model 590), UV detector (Spectroflow 733 Kratos), automatic sample processor (Waters, Model 712), integrator (Spectra physics 4270) and the column which was HP ODS. C18 hypersil (5 mm, 200 by 4.6 mm). The detector wavelength was set at 241 nm, while a ratio of acetonitrile:water (60:40) was used as the mobile phase. Under these conditions, a well-separated peak was detected at the retention time of 5.2 min with a sensitivity of 0.1 mg / ml. Meanwhile, the total testosterone contents in unused patches with the same manufacturer’s lot number were also analyzed as a parallel control of the extraction procedure.

Fig. 1. Schematic illustration of one-compartment pharmacokinetic open model for transdermal delivery of testosterone from a T-TDD and mathematical representation where Fri is the initial rapid input to the central compartment; Fsi , the maintenance input to the central compartment; Cp , drug concentration in the central compartment; V, the volume of distribution, Cl, the rate of clearance.

92

Q.-F. Xing et al. / Journal of Controlled Release 52 (1998) 89 – 98

parallel maintenance input (Fsi ) of testosterone across the skin which occurred at a slower rate. The following differential equation was derived using the model described in Fig. 1 to analyze the concentration profiles of testosterone in the central compartment after transdermal delivery. dCp (Fri 1 Fsi 2 Cl Cp ) ]] 5 ]]]]]] dt V

(1)

where Dr 2 Dsi Fri 5 ]]] t ri

(1A)

Dsi Fsi 5 ] t si

(1B)

Prior to the model fitting, the values of Dr (dose released throughout the period of TDD application) were derived by subtracting the total drug content in the unused T-TDD with the residual content in the used T-TDD. The values of V (volume of distribution), Cl (plasma clearance), Dsi (dose delivered at a slow input period), t ri (length of time for the rapid input period) and t si (length of time for the slow input period) were obtained following the model fitting by using the PCNonlin program [11]. Meanwhile, the values of t 1 / 2 (elimination half-life), AUC, FRC ri (fraction of the total amount released during rapid input period) were also determined as the secondary parameters. Mean values and standard deviations of the pharmacokinetic parameters together with the coefficients of determined, r 2 values, and the randomness of the weighted residues were examined to determine the goodness of fit.

2.9. In vivo input kinetic analysis The in vivo testosterone input kinetics, I (t), were calculated for each individual study from the testosterone profiles, Cp (t), measured during the study, using the Wagner–Nelson equation [12]: Cp (t) I(t) 5 Cl []] 1 AUC 0→t ] k

(2)

The values of k (elimination rate constant) and Cl (clearance) were taken from the results of compartmental pharmacokinetic model fitting in each data set

of the individual studies, while the values of AUC 0→t were derived by the trapezoidal rule from 0 to t.

2.10. Statistical analysis The statistical significance of the differences among the various values of pharmacokinetic parameters following the transdermal delivery of various doses of testosterone from the T-TDD were compared by an ANOVA test at the P50.05 level.

3. Results and discussion

3.1. Castrated minipig model The mean (6S.D.) testosterone level before the castration was 625 (6169) ng / dl, while none could be detected after the castration in the three minipigs studied. The results indicate that the influences of endogenous testosterone during the pharmacokinetic studies in castrated minipigs were limited to a minimum. Therefore, it is a good animal model for pharmacokinetic analysis of testosterone as it is comparable to the situation in hypogonadal men.

3.2. Animal studies The plasma profiles of total testosterone, together with the fit of data following the transdermal administration of testosterone at various doses in the same group of castrated minipigs, are compared graphically in Fig. 2. The results indicate that the plasma concentrations of testosterone all increase rapidly and reach their respective Cmax around 9 h following the administration. Thereafter, they gradually decline during the 24 h of the application period and then remarkably drop to the baseline within 4 h after the patches’ removal. Moreover, the results also indicate, as expected, that the value of Cmax increased with increasing doses of testosterone, and the value of t max was independent of dose. Furthermore, no detectable amount of testosterone could be found in the control study. On the other hand, the plasma-levels were adequately fitted by nonlinear least-squares regression analysis to a one-compartment open model with multiple input rates. The mean (6S.E.M.) values of

Q.-F. Xing et al. / Journal of Controlled Release 52 (1998) 89 – 98

Fig. 2. Plasma profiles of total testosterone in the castrated Yucatan minipigs (n59, triplicate experiments in each of three pigs) and fitting of the experimental data (solid line) obtained following the 24 h transdermal application of T-TDDs: one unit (solid circle), two units (solid square), and three units (solid triangle) in comparison with the controls (n53, single experiment in each of three pigs).

coefficients and Akaike’s information criterion (AIC) were 0.922 (60.016), 0.924 (60.012) and 0.924 (60.009) and 7.3 (62.3), 20.0 (3.1), and 26.8 (62.4), respectively, for 1, 2 and 3 units.

3.3. Pharmacokinetic analysis Pharmacokinetic parameters obtained from noncompartmental analysis of plasma total testosterone profiles, displayed in Fig. 2, are given in Table 1. The mean (6S.E.M.) values of Cmax and t max were 118 (611), 205 (612) and 285 (621) ng / dl and 8.7 (61.1), 7.3 (60.7), and 10.0 (61.0) h, respectively, for 1, 2 and 3 units. The differences in Cmax values are statistically significant (P,0.05) among the doses applied, but there is no difference in the values of t max . The results suggest that the value of Cmax increased as increasing the doses of testosterone, while there was no effect on t max . The mean (6SE)

93

values of AUC0 →28 were 1359 (6140), 2840 (6223), and 3900 (6297) ng / dlxh, respectively, for 1, 2, and 3 units. The differences in AUC0 →28 values are also statistically significant (P,0.05) among the doses applied. The results indicate, as expect, that the value of AUC increased as increasing the doses of testosterone applied. In summary, the results of noncompartmental analysis conclude that the absorption of testosterone in castrated swine model is significantly increased as the dose increased. Pharmacokinetic parameters obtained from compartmental analysis of plasma total testosterone profiles are outlined in Table 1. The mean (6S.E.M.) values of V and Cl were 1040 (6106), 1027 (699) and 1001 (687) l and 128 (610), 151 (612), and 143 (610) l / h, respectively, for 1, 2 and 3 units. There is no difference in the values of V and Cl among the various doses applied. Moreover, results in the values of t ri and t si suggest that these parameters were independent on the dose over the range of 1–3 units, while results in the values of Dsi and AUC, which increased as the dose increased, indicate that dose proportionality was observed over the dose range studied. Although the value of t 1 / 2 of 1 unit (246 min) was higher than that of 2 and 3 units (122 and 136, respectively), there is no statistical significant difference in t 1 / 2 among the various doses applied. In addition, after taking the consideration of two worst fit data sets in 1 unit group, the value of t 1 / 2 (117 min) became similar to the others. Although Dsi appeared dose-dependent, however, after normalized by released amount of testosterone (i.e. FRC ri ), there is, as expect, no statistical significant difference in FRC ri . Overall results from modeling analysis suggest that around 92% of total released testosterone was delivered into the systemic circulation during the first 11 h of the initial rapid input period, and only around 8% of that was used as a maintenance dose during the total period of 23 h for the slow input. Moreover, the absorption and disposition of testosterone administered transdermally in the castrated minipig model can be expressed by the model represented above (Fig. 1).

3.4. In vivo input kinetics In order to understand the in vivo input mecha-

94

Q.-F. Xing et al. / Journal of Controlled Release 52 (1998) 89 – 98

Table 1 Pharmacokinetic parameters after transdermal administration of testosterone delivered from T-TDD in castrated Yucatan minipigs (n59, triplicate experiments in each of 3 pigs) Parameter

Testosterone (units) 2

3

1a

I. Noncompartmental pharmacokinetics Cmax (ng / dl)b,c 188 (11)d e t max (h) 8.7 (1.1) AUC 0→28 (ng / dlxh)f,c 1359 (140)

205 (12) 7.3 (0.7) 2840 (223)

285 (21) 10.0 (1.0) 3900 (297)

124 (13) 9.4 (1.2) 1404 (177)

II. Compartmental pharmacokinetics V (l)g 1040 (106) Cl (l / h)g 128 (10) t ri (h)g 10.2 (0.8) t si (h)g 22.6 (0.8) Dsi (mg)g,c 137 (20) AUC 0→` (ng / dlxh)h,c 1562 (139) t 1 / 2 (min)h 246 (106) FRC ri (%)h 93 (1)

1027 (99) 151 (12) 10.7 (1.0) 22.7 (0.7) 390 (56) 3333 (244) 122 (16) 92 (1)

1001 (87) 143 (10) 12.4 (0.4) 23.1 (0.9) 496 (57) 662 (312) 136 (12) 92 (1)

986 (118) 126 (13) 10.9 (0.8) 22.2 (0.9) 127 (19) 1596 (173) 117 (13) 93 (1)

1

a

The results were obtained from 7 best fitting data sets. b Cmax , maximum concentration observed during the dosing period. c The result of ANOVA indicated a significant difference (P,0.05) among the units applied. d Data was presented as mean (S.E.M.). e t max , time to reach the maximum concentration (Cmax ). f AUC, area under the plasma concentration-time curves calculated by trapezoidal rule. gV, volume of distribution, Cl, clearance, Dsi , dose of slow input, t ri , length of time for rapid input, and t si , length of time for slow input, were obtained from model fitting as primary parameters in PCNonlin. h t 1 / 2 , elimination half-life, AUC, and FRC ri , fraction of total dose released during rapid input, were obtained from model fitting using secondary parameter option in PCNonlin.

nism, the profiles of cumulative in vivo input kinetics, following a 24 h application of various doses in castrated minipigs, were derived by a Wagner-Nelson analysis, and compared in Fig. 3. Results indicated a biphasic pattern, characterized by a rapid input phase during the first 8 h, together with a slow input phase during the last 12 h. It further confirms the assumption, described in Fig. 1, of multiple constant inputs of testosterone released from the T-TDD administered transdermally. The cumulative input amounts at 24 h were 1.9 (60.2), 4.8 (60.2) and 6.4 (60.5) mg for 1, 2 and 3 units, respectively. As expected, the total amount of the input increases with increasing doses applied. Moreover, they were consistent with the amount of testosterone released, calculated from the residual analysis in used units. The mean (6S.E.M.) values of testosterone released during the 24 h applications were 1.9 (60.2), 4.7 (60.2) and 6.6 (60.5) mg for 1, 2 and 3 units, respectively. Results conclude that the close correspondence (almost 100%) was achieved between the in vivo input kinetics, derived by the Wagner–Nelson analysis, and total amount released.

Fig. 3. The in vivo input kinetic profiles (open symbols), derived by the Wagner-Nelson analysis, and the total released amount (solid symbols), calculated from residual analysis of the used units, of testosterone delivered from the 24 h transdermal application of T-TDDs at three dosage strengths on the castrated Yucatan minipigs (n59, triplicate experiments in each of three pigs).

Q.-F. Xing et al. / Journal of Controlled Release 52 (1998) 89 – 98

3.5. Metabolites The mean (6S.E.M.) plasma level profiles of free and total testosterone, after 24 h transdermal delivery of testosterone (three units, as shown in Fig. 2), are compared in Fig. 4A. A close correspondence was observed between the profiles of free and total testosterone. The mean (6S.E.M.) plasma level profiles of DHT and E2 are displayed in Fig. 4B and 4C. The results in Fig. 4B indicate that a similar increasing trend was observed in both profiles of

95

total testosterone and DHT, although the DHT levels of pre- and post-application of T-TDD were observed to be higher than those of total testosterone. It suggests that the basal level of DHT might be come from other resources than from the metabolism of testosterone in the castrated Yucatan minipig model. Moreover, no clear relationship of increasing trend could be observed between E2 and total testosterone (Fig. 4C), which could result from low ratio of testosterone being metabolized to E2 (200:1). The comparison of pharmacokinetic parameters

Fig. 4. Comparative plasma profiles of total and free testosterone (A) as well as dihydrotestosterone (B) and estradiol (C), the major metabolites of testosterone, following the 24 h transdermal delivery of testosterone from 3 units of T-TDDs in the castrated Yucatan minipigs (n59, triplicate experiments in each of three pigs).

96

Q.-F. Xing et al. / Journal of Controlled Release 52 (1998) 89 – 98

Table 2 Comparison of pharmacokinetic parameters of testosterone and its metabolites in castrated minipigs after transdermal administration of testosterone (n59, triplicate experiments in each of three pigs) Testosterone / Metabolite

Mean (S.E.M.)

Ratio of TT

Total testosterone (TT) Cmax (ng / dl)a AUC 0→24 (ng / dlxh)b

85 (21) 3990 (297)

1 1

Free testosterone (FT) Cmax (ng / dl) AUC 0→24 (ng / dlxh)

0.9 (0.1) 12.0 (1.4)

0.0032 0.0032

Dihydrotestosterone (DHT) Cmax (ng / dl) AUC 0→24 (ng / dlxh)

28 (3) 414 (41)

0.100 0.104

3 (0.4) 29 (4)

0.010 0.007

Estradiol (E2) Cmax (ng / dl) AUC 0→24 (ng / dlxh) a

Cmax , maximum concentration observed during the dosing period. AUC, area under the plasma concentration-time curves calculated by trapezoidal rule.

b

(i.e., Cmax and AUC) of testosterone and its metabolites in castrated minipigs after transdermal administration of testosterone are outlined in Table 2, which indicate that the ratios of FT / TT, DHT / TT and E2 / TT attained in castrated minipigs were |0.003, |0.102 and |0.009, respectively.

3.6. Comparison of testosterone between minipig and human Testosterone administered transdermally has been studied and evaluated in hypogonadal men [6–8]. Therefore, the mean (6S.D.) plasma profiles of total testosterone, after 24 h transdermal applications of 2 T-TDDs in hypogonadal men were obtained from the literature [8] and compared with that obtained from the castrated Yucatan minipigs in Fig. 5. In addition, the values of pharmacokinetic parameters were calculated from the plasma profiles of total testosterone obtained from hypogonadal men and castrated minipigs, as displayed in Fig. 5, and are compared in Table 3. The results from Fig. 5 indicate that plasma testosterone profiles obtained in castrated minipigs were found to show similar trends, but with lower levels, to that in hypogonadal men. The agreement suggests that the transdermal absorption mechanism of testosterone in the castrated minipig

Fig. 5. Comparison between the plasma profiles of total testosterone attained by the 24 h transdermal delivery of testosterone from 2 units of T-TDDs in the castrated Yucatan minipigs (n59) and that in the hypogonadal men (n529) reported in literature ( [8]).

and in the hypogonadal men could be similar. Moreover, results in Table 3 suggest that the mean (6S.D.) values of Cmax and Cavg obtained from hypogonadal men were observed to be higher than that from castrated minipigs [753 (6276) vs. 205 (636) and 498 (6169) vs. 116 (627) ng / dl, respectively], while no difference was found in t max . In view of the difference in Cmin between minipigs and human, normalization of Cmax and Cavg was considered and therefore the values of DCmax and DCavg were calculated and compared in Table 3. By comparing the values between DCmax and Cmax as well as DCavg and Cavg , the ratios of minipig / human were increased to 0.40 from 0.27 and to 0.44 from 0.23, respectively. In other words, around 40% of correlation was observed, based on DCmax and DCavg , between castrated minipigs and hypogonadal men. Furthermore, it is interested to note that the testosterone clearance rate obtained in castrated minipigs (36196828 l / day) were quite different (|2.8 fold higher than human) from that in the hypogonadal human (13046464 l / day). It may be

Q.-F. Xing et al. / Journal of Controlled Release 52 (1998) 89 – 98

97

Table 3 Comparison of testosterone pharmacokinetics between hypogonadal men and castrated male minipigs following the 24 h transdermal delivery of testosterone from 2 units of T-TDDs Parameter

Human a

Minipig

b

Cmax (ng / dl) Cavg (ng / dl)c t max (h)d Cmin (ng / dl)e DCmax (ng / dl)f DCavg (ng / dl)g Cl (l / day)h Rel d (mg)i

Ratio

Mean (S.D.)

N

Mean(S.D.)

N

(Minipig / Human)

205 (36) 116 (27) 7.3 (2.0) 4 (1) 201 112 3619 (828) 4.7 (0.2)

9 9 9 9 9 9 9 9

753 (276) 498 (169) 7.9 (2.2) 246 (120) 507 252 1304 (464) 6.4 (0.5)

56 56 56 56 56 56 49 6

0.27 0.23 0.92 0.40 0.44 2.78 0.73

a

Data from [8]. b Cmax , maximum concentration observed during the dosing period. c Cavg , average concentration (AUC / 24 h). d t max , time to reach the maximum concentration (Cmax ). e Cmin , minimum concentration observed during the dosing period. f DCmax , baseline (i.e. Cmin ) normalized maximum concentration. g DCavg , baseline (i.e. Cmin ) normalized average concentration. h Cl, clearance. i Rel d , total amount released; human data from [6].

the reason that the values of DCmax and DCavg in the castrated minipig were 2.4 fold lower (i.e. around 40% of minipig / human ratio) than that in the hypogonadal human. In vivo kinetic input profiles together with the

residue analysis of testosterone, following the 24 h transdermal delivery of testosterone from 2 units of T-TDDs, in hypogonadal men were obtained from the literature [6] and compared with that from castrated Yucatan minipigs in Fig. 6. The results in Fig. 6 suggest that the in vivo kinetics obtained in minipigs were similar to those in humans during the first 8 h, but with a lower input rate during the last 12 h. This may be further evidence that the lower values of DCmax and DCavg found in the minipig as compared to the human resulted from the lower amount of testosterone input in the minipig compared to that in the human.

3.7. Comparison of metabolites of testosterone between minipig and human The value of FT / TT ratio (0.03%) in Table 2 is lower than that reported in the human (1-3%) [13], while the ratios of DHT / TT and E2 / TT (|0.10 and |0.009, respectively), attained in castrated minipigs, are consistent with those reported in the human (|0.10 and |0.005, respectively) [6–8].

Fig. 6. Comparison of in vivo input kinetic profiles (open symbols), derived by the Wagner-Nelson analysis, and the total released amount (solid symbols), calculated from the residue analysis of the used T-TDDs, of testosterone delivered from the 24 h transdermal delivery of testosterone from 2 units of T-TDDs between the castrated Yucatan minipigs (n59) and the hypogonadal men (n56) reported in the literature [6].

4. Conclusion Despite the difference in clearance rate, agreement was observed in the plasma testosterone profiles and in vivo input kinetics between castrated minipigs and hypogonadal men. Therefore, castrated minipigs

98

Q.-F. Xing et al. / Journal of Controlled Release 52 (1998) 89 – 98

could be a suitable large animal model for studying the potential pharmacokinetics and metabolism of testosterone delivered transdermally in humans.

Acknowledgements This series of studies were supported by grants from World Health Organization (WHO) Contraceptive Research and Development Program, Geneva, Switzerland. Qing-Feng Xing is a visiting scholar from Liaoning Research Institute for Family Planning sponsored by the World Health Organization.

References [1] G.W. Cleary, Transdermal delivery systems: a medical rationale, in: V.P. Shah, H.I. Maibach (Eds.), Topical Drug Bioavailability, Bioequivalence, and Penetration, Plenum Press, New York, 1993, pp. 17–68. [2] Physicians’ Desk Reference, 50th ed., Medical Economics Company Inc., Oradell, NJ, 1996. [3] R.L. Oberle, H. Das, S.L. Wong, K.K.H. Chan, R.J. Sawchuk, Pharmacokinetics and metabolism of diclofenac sodium in yucatan miniature pigs, Pharm. Res. 11(5) (1994) 698–703. [4] D. Leveque, I.D. Peter, J. Salmon, Y. Salmon, H. Elkhaili, G. Kaltenbach, F. Jehl, The catheterized yucatan micropig as an animal model for experimental pharmacokinetics of anticancer drugs (Abstract), Proc. Annu. Meet. Am. Assoc. Cancer Res. 36 (1995) A2136.

[5] T. Kurihara-Bergstrom, M. Woodworth, S. Feisullin, P. Beall, Characterization of the yucatan miniature pig skin and small intestine for pharmaceutical applications, Lab. Anim. Sci. 36(4) (1986) 396–399. [6] N.A. Mazer, W.E. Heiber, J.F. Moellmer, A.W. Meikle, J.D. Stringham, S.W. Sanders, K.G. Tolman, W.D. Odell, Enhanced transdermal delivery of testosterone: a new physiological approach for androgen replacement in hypogonadal men, J. Control. Release 19(3) (1992) 347–362. [7] N.A. Mazer, S.W. Sanders, C.D. Ebert, A.W. Meikle, Mimicking the circadian pattern of testosterone and metabolite levels with an enhanced transdermal delivery system, in: R. Gurney, H.E. Junjinger, N. Peppas, (Eds.), Pulsatile Drug Delivery: Current Applications and Future Trends, Stuttgart, Wissenshaftliche Verlagsgesellschaft mbH, 1993, pp. 73–97. [8] Androderm  (testosterone transdermal system) product information, SmithKline Beecham Pharmaceuticals, Philadelphia, PA, 1995. [9] S. Lin, Y.W. Chien, A rapid and simple technique for serial / continuous sampling of blood and intravenous administration in conscious swine. Contemp. Topics Lab. Anim. Sci. 36 (1997) 58–61. [10] J. Gabrielsson, D. Weiner (Eds.), Pharmacokinetic / Pharmacodynamic Data Analysis: Concepts and Applications, Swedish Pharmaceutical Press, Stockholm, Sweden, 1994, pp. 411–416. [11] PCNONLIN User Guide, 4th ed., Statistical Consultants, Inc., Lexington, KY, 1992. [12] J.G. Wagner, E. Nelson, Percent absorbed time plots derived from blood level and / or urinary excretion data, J. Pharm. Sci. 52(6) (1963) 610–611. [13] A. Vermeulen, T. Stoica, L. Verdonck, The apparent free testosterone concentration, an index of androgenicity, J. Clin. Endocrinol. Metab. 33(5) (1971) 759–767.