Cytochrome P450 (P450¹) constitutes a multigene family of hemoproteins responsible for the metabolism of numerous xenobiotics including therapeutic drugs, environmental chemicals, and dietary constitutes, as well as endogenous substrates such as steroids and bile acids. Among the human P450 proteins, CYP3A is the most abundant cytochrome P450 isozyme, and approximately 50% of all currently used drugs are substrates for CYP3A (Watkins, 1994; Wilkinson, 1996; Streetman et al., 2000). Because of wide interindividual variations in the CYP3A activity in humans, the prospective use of CYP3A phenotyping or evaluating in vivo CYP3A activity has recently been an attractive approach to prediction of optimal dosage range to improve the therapeutic outcome and to minimize adverse effects (Dahl, 2002; Zaigler et al., 2000).

CYP3A enzyme catalyzes the oxidation at C-6β of cortisol (F) to form 6β-hydroxycortisol (6β-OHF). The metabolite 6β-OHF is then excreted as an unconjugated form in urine. Park (1981) and Saenger (1983) reported that urinary excretion of 6β-OHF should be an index of enzyme activity of the hepatic mixed-function oxygenases. Thereafter, urinary excretion ratio 6β-OHF/F has extensively been used as a suitable noninvasive index for evaluating the in vivo CYP3A activity to investigate drug-drug interactions (Ged et al., 1989; Wilkinson, 1996; Monsarrat et al., 1998; Pichard-Garcia et al., 2000). The urinary ratio has also been used for the elucidation of the statistical correlation of CYP3A activity with genetic and/or environmental factors, disease states such as hepatic disease, cancer, and hypertension, and ethnic differences, etc. (Watkins, 1994; Lin et al., 1999; Zheng et al., 2001). Although numerous studies have described successful utilization of the urinary ratio 6β-OHF/F, the ratio does not always correlate with the disposition of CYP3A substrate drugs, including typical phenotyping probe drugs such as [¹⁴C]-erythromycin and midazolam (Hunt et al., 1992; Watkins et al., 1992; Kinirons et al., 1993, 1999). The usefulness of the urinary ratio 6β-OHF/F as an appropriate index for the in vivo CYP3A activity has therefore been the subject of controversy. There is need for novel approaches to be developed to accurately interpret how the urinary 6β-OHF/F ratio would be used to evaluate the in vivo CYP3A activity.

The present study was undertaken to stress the importance of applying stable isotope methodology to evaluate 6β-hydroxylation of endogenous cortisol as a measure of CYP3A. It is suggested that endogenous cortisol 6β-hydroxylation clearance can be used as a reliable index for the in vivo CYP3A phenotyping. It is also demonstrated that the urinary ratio 6β-OHF/F varies with renal clearance of cortisol, and the factors contradictory to the use of urinary ratio 6β-OHF/F as an index for in vivo CYP3A activity are discussed (Furuta et al., 2001).

Materials and Methods

Chemicals and Reagents. 6β-OHF and F were purchased from Steraloids (Newport, RI) and Sigma-Aldrich (St. Louis, MO), respectively. 6β-[1,1,19,19,19-²H₅]Hydroxycortisol (6β-OHF-d₅, 91 atom %²H), [1,1,19,19,19-²H₅]cortisol (cortisol-d₅, 98 atom %²H), and [1,2,4,19-¹³C₄]cortisol (cortisol-¹³C₄, 97 atom %¹³C) were synthesized in this laboratory (Furuta et al., 2000b).

Administration of Stable Isotopically Labeled Cortisol. Three healthy adult subjects (subject A, aged 25 years, male, 68 kg in weight; subject B, aged 23 years, male, 57 kg; and subject C, aged 25 years, male, 65 kg) participated in this study. The subjects received a single 5-mg oral dose of cortisol-d₅ (subjects A and B) or a single 3-mg oral dose of cortisol-¹³C₄ (subject C) in the morning (at 10:00 AM) after 12 h of overnight fasting (Kasuya et al., 1998, 2002). No individual was receiving any medication. Blood samples (5-10 ml) were obtained at 10 min before and at 5, 10, 15, 20, 30, 40, 60, 80, 100, 120, 150, 180, 210, 240, 270, 300, 330, 360, 390, 420, 450, and 480 min after dosing. The heparinized blood was collected in glass tubes and centrifuged. Urine samples were obtained at timed periods of 2 h (0-2, 2-4, 4-6, and 6-8 h in subjects A and C, and 0-2, 2-4, and 4-8 h in subject B), and the volume and pH of the urine samples were noted. The plasma and urine samples were stored at -20°C until analysis. Plasma concentrations of labeled cortisol and urinary excreted amounts of labeled cortisol and its metabolite 6β-OHF were analyzed by gas chromatography-mass spectrometry simultaneously with their endogenous counterparts (Furuta et al., 2000a,c).

Administration of the Macrolide Antibiotic Clarithromycin. Clarithromycin (200 mg) was administered to a healthy volunteer (subject D, aged 25 years, male, 80 kg in weight) every 12 h at 10:00 AM and 10:00 PM for 5 days and finally at 10:00 AM on day 6. The subject was not receiving any other medication. Blood samples (5 ml) were obtained every hour from 10:00 AM to 8:00 PM on days 0 (predose), 4, 13, and 68. On days 1 and 6, blood samples were obtained every 30 min from 10:00 AM to 3:00 PM and then every hour from 3:00 to 8:00 PM after dosing clarithromycin at 10:00 AM. On days 2, 3, and 5, blood samples were obtained once at 10:00 AM and then every 2 h from 3:00 to 8:00 PM. Urine samples were obtained at a timed period of 2 h (0-2, 2-4, 4-6, 6-8, 8-10, and 10-12 h). Urinary concentrations of cortisol and 6β-OHF were determined simultaneously by high-performance liquid chromatography (manuscript in preparation), and plasma concentrations of cortisol were determined by gas chromatography-mass spectrometry (Furuta et al., 2000a).

The studies were approved by the Tokyo University of Pharmacy and Life Science Human Subjects Review Board. Written informed consent was obtained from the subjects.

Results and Discussion

The present study used stable isotope methodology to provide evidence for the validity of 6β-hydroxylation clearance of endogenous cortisol as a new reliable index for in vivo CYP3A phenotyping in humans. We also identified important factors contradictory to the use of a conventional index of the urinary ratio 6β-OHF/F to evaluate in vivo CYP3A activity. Three healthy male volunteers received a single oral dose of either 5 mg of cortisol-d₅ (subjects A and B) or 3 mg of cortisol-¹³C₄ (subject C). Blood samples were obtained before and during a period of 8 h after dosing at 10:00 AM. Urine samples were collected at timed periods (0 -2, 2-4, 4-6, and 6-8 h in subjects A and C) and (0-2, 2-4, and 4-8 h in subject B). Plasma concentrations of the labeled (exogenous) cortisol and urinary excreted amounts of the labeled cortisol and 6β-hydroxycortisol (unconjugated form) were analyzed by a stable isotope dilution mass spectrometric method simultaneously with their endogenous counterparts. This enabled the accurate determination of the fractional metabolic clearance specific for the 6β-hydroxylation of labeled cortisol (CL_m(6β)-exo) and to examine how the exogenous clearance (CL_m(6β)-exo) thus obtained was related to the endogenous 6β-hydroxylation clearance (CL_m(6β)-endo) and the urinary ratio of endogenous 6β-OHF/F.

CL_m(6β)-exo was calculated in eq. 1 as the amount of urinary excreted 6β-hydroxycortisol (X_(6β)-exo) during the whole urine collection period divided by the area under the 0- to 8-h concentration-time curve of cortisol (AUC_(F)-exo) (Furuta et al., 1996, 2001):

The CL_m(6β)-exo values were 8.77 ml/min (subject A), 3.94 ml/min (subject B), and 4.75 ml/min (subject C), whereas the urinary ratios of exogenous 6β-OHF/F (= X_(6β)-exo/X_(F)-exo) during the urine collection period were 3.94 (subject A), 8.71 (subject B), and 6.31 (subject C). Similarly, CL_m(6β)-endo was also estimated by eq. 2. wherein X_(6β)-endo is urinary excretion amount of endogenous 6β-hydroxycortisol, and AUC_(F)-endo is area under the curve of endogenous cortisol.

The values for CL_m(6β)-endo were 2.94 ml/min (subject A), 1.69 ml/min (subject B), and 2.25 ml/min (subject C), which were lower than the corresponding exogenous CL_m(6β)-exo values of individual subjects, due most likely to the increased AUC value caused by continuous and episodic adrenal cortisol secretion. The urinary ratios of endogenous 6β-OHF/F (= X_(6β)-endo/X_(F)-endo) were 3.02 (subject A), 5.54 (subject B), and 4.85 (subject C). The endogenous 6β-OHF/F was also lower than the corresponding exogenous 6β-OHF/F in each individual subject.

Figure 1, a and b, compares the endogenous CL_m(6β)-endo with exogenous CL_m(6β)-exo calculated in each urine collection period in the three subjects. There was a good correlation between the endogenous and exogenous 6β-hydroxylation clearances [r = 0.7733 (Fig. 1b), 0.9112, and 0.9534 (Fig. 1a) for 2-h, 4-h, and 6- or 8-h urine collection periods, respectively]. This strongly suggests that CL_m(6β)-endo can be used as an appropriate index for phenotyping the in vivo CYP3A activity. One limitation for use of the proposed endogenous fractional metabolic clearance as an index is that it does not provide a phenotypic measure of CYP3A activity in the intestine and reflects systemic clearance for the hepatic and kidney CYP3A activities (Watkins, 1994).

The following transformation of eq. 1 into eq. 3 indicates that the urinary ratio 6β-OHF/F (= X_(6β)-exo/X_(F)-exo) is a function of two independent parameters of metabolic clearance for the 6β-hydroxylation (CL_m(6β)-exo) and renal clearance of exogenous cortisol (CL_r(F)-exo) (Furuta et al., 2001). The urinary ratio 6β-OHF/F (= X_(6β)-exo/X_(F)-exo) can then be expressed as: From this equation, it is obvious that the urinary ratio 6β-OHF/F is valid as the index reflecting the in vivo CYP3A activity only when there are no intra- and interindividual variations in CL_r(F)-exo.

When CL_r(F)-exo was calculated as the amount of urinary cortisol (X_(F)-exo) excreted during the whole urine collection period divided by the area under the 0- to 8-h concentration-time curve of cortisol (AUC_(F)-exo), it was found that there was an approximately 5-fold interindividual variability (2.23 ml/min for subject A, 0.45 ml/min for subject B, and 0.75 ml/min for subject C). Furthermore, from the results of CL_r(F)-exo calculated in each urine collection period within the same subject, intraindividual variabilities in the CL_r(F)-exo were 4.6-fold for subject A (0.62-2.85 ml/min), 3.6-fold for subject B (0.17-0.61 ml/min), and 2.1-fold for subject C (0.40-0.82 ml/min).

Inter- and intraindividual variations as observed in CL_r(F)-exo in this study support the idea that the urinary ratio of endogenous 6β-OHF/F does not accurately reflect the in vivo CYP3A activity. This is evidenced from the present observation that the order of the urinary 6β-OHF/F ratio (X_(6β)-exo/X_(F)-exo) in the three subjects tested (B > C > A) was not consistent with that of the exogenous CL_m(6β)-exo values (A > C > B). It should be emphasized again that the urinary ratio 6β-OHF/F is valid as the index for in vivo CYP3A activity only when the renal clearance of cortisol (CL_r(F)) is consistent as discussed above (see eq. 4). Even within the same subject, there were large variations (2.1- to 4.6-fold) in the CL_r(F)-exo. This must have led to the poor correlation between the endogenous 6β-OHF/F and the exogenous cortisol 6β-hydroxylation clearance (Fig. 1c). Since the magnitude of variability in the renal cortisol clearance as well as the episodic cortisol secretion from adrenal glands largely affect the endogenous 6β-OHF/F, it is reasonable to conclude that the urinary 6β-OHF/F ratio is not an appropriate index for CYP3A phenotyping.

In this study, administration of labeled cortisol and the simultaneous measurements of exogenous (labeled) and endogenous cortisol and 6β-hydroxycortisol demonstrated that CL_m(6β)-endo was to be an appropriate index for in vivo CYP3A phenotyping. The proposed CYP3A phenotyping was then applied to the assessment of the inhibitory effects of the macrolide antibiotic clarithromycin on in vivo CYP3A activity. Macrolide antibiotics generally form inactive iron-metabolite complexes with CYP3A and cause a decrease in its catalytic activity. Clarithromycin (200 mg) was administered to a healthy volunteer (subject D) every 12 h at 10:00 AM and 10:00 PM for 6 days. The CL_m(6β)-endo value on the day before clarithromycin administration (day 0, baseline) was 3.47 ± 0.33 ml/min (mean ± S.D.). After the initial administration of clarithromycin at 10:00 AM on day 1, the CL_m(6β)-endo value rapidly decreased from 3.71 ml/min to 1.83 ml/min, and then remained as low as 1.46 ± 0.18 ml/min to 1.74 ± 0.15 ml/min during treatment with clarithromycin (days 2-6), indicating 50% to 58% reduction in the cortisol 6β-hydroxylation clearance (Fig. 2a). After terminating the administration, the values increased to 2.85 ± 0.35 ml/min on day 13 after a 7-day washout period and 3.48 ± 0.26 ml/min on day 68 after a 2-month washout period, respectively (Fig. 2a). Inhibitory effects of clarithromycin on the 6β-hydroxylation clearance of cortisol were obvious.

On the other hand, the time course data of the urinary 6β-OHF/F ratios shown in Fig. 2b lead to erroneous interpretations as to how clarithromycin affected the CYP3A activity. There was no obvious decrease in the urinary 6β-OHF/F ratios during the 6-day clarithromycin treatment, although the ratio values (3.12 ± 0.66) on day 5 were below the baseline range (6.95 ± 1.04). The ratios increased by 1.4- to 2.2-fold of the baseline on days 13 and 68. The urinary 6β-OHF/F ratios therefore do not clarify the inhibition of the 6β-hydroxylation of cortisol by clarithromycin, as evidenced by the reduction of CL_m(6β)-endo (Fig. 2a).

Careful examination of the changes in the 6β-hydroxylation clearance (Fig. 2a) and the renal clearance of endogenous cortisol (Fig. 2c) in each test day explains why the urinary 6β-OHF/F ratio is not an appropriate index for the in vivo CYP3A activity. When the renal clearance of endogenous cortisol (CL_r(F)-endo) was calculated as the amount of urinary cortisol (X_(F)-endo) excreted during the 2-h urine collection period divided by the corresponding 2-h AUC_(F)-endo, there were approximately 2-fold day-to-day and 3-fold within-day variabilities (Fig. 2c). It is difficult to show approximately 50% reduction of the CYP3A activity with clarithromycin treatment by the urinary 6β-OHF/F ratio under the 2- to 3-fold fluctuations in CL_r(F) (Fig. 2c). The fact that the 6β-OHF/F ratios lie within the baseline range during the clarithromycin treatment (Fig. 2b), for example, resulted from the decreased CL_r(F) (Fig. 2c). The decreased renal clearance of cortisol on days 13 and 68 resulted in the increase of the urinary 6β-OHF/F ratio by 1.4- to 2.2-fold of the baseline, since the cortisol 6β-hydroxylation clearance had returned to the baseline level. The 6β-OHF/F ratio can be changed by CYP3A inhibitors or inducers only when the change in the 6β-hydroxylation clearance (CL_m(6β)) by inhibition or induction is much larger than the magnitude of variability in CL_r(F) as seen in eq. 4.

By using a probe drug in phenotyping, the clearance of the drug should provide the best estimate of in vivo catalytic activity of the enzyme of interest (Watkins, 1994). If the probe drug has multiple metabolic pathways, the fractional metabolic clearance corresponding to the pathway of interest should be an appropriate measure (Watkins, 1994). The present study uniquely used stable isotope methodology to evaluate the in vivo 6β-hydroxylation of endogenous cortisol. This is the first example that demonstrates the accurate fractional metabolic clearance specific for the 6β-hydroxylation of cortisol as a measure of the in vivo CYP3A activity in humans (Furuta et al., 2001). Very recently, the clarithromycin-induced inhibition in the in vivo CYP3A activity in Helicobacter pylori-positive patients was assessed by changes in plasma concentrations of lansoprazole concomitantly administered and partial cortisol 6β-hydroxylation clearance (Ushiama et al., 2002). It remained, however, unclear whether endogenous cortisol 6β-hydroxylation clearance would have certain advantages over the traditional urinary index of CYP3A (the urinary 6β-OHF/F ratio) for detecting an inhibitory effect of CYP3A. The present stable isotope methodology has provided direct evidence to suggest that endogenous cortisol 6β-hydroxylation clearance is an appropriate index for the in vivo CYP3A phenotyping in humans. It is also suggested that the urinary ratio of 6β-OHF to cortisol does not always reflect in vivo CYP3A activity.