Pharmacokinetics of ruboxistaurin are significantly altered by rifampicin-mediated CYP3A4 induction

Abstract

Aims

The aim of this study was to evaluate the effect of rifampicin co-administration on the pharmacokinetics of ruboxistaurin and its active metabolite, N-desmethyl ruboxistaurin and, in addition, to compare the changes in pharmacokinetics of ruboxistaurin and N-desmethyl ruboxistaurin with the urinary 6β-hydroxycortisol : cortisol ratio. Ruboxistaurin is a specific protein-kinase-C β inhibitor in clinical development for the treatment of diabetic microvascular complications.

Methods

This was a two-period, one-sequence study. Sixteen healthy male subjects completed both study periods. In period one, a single 64 mg oral dose of ruboxistaurin was administered. In period two, 600 mg rifampicin was administered daily for 9 days, during which another single 64 mg ruboxistaurin dose was administered on day 7. Blood samples were collected and assayed for ruboxistaurin and N-desmethyl ruboxistaurin. CYP3A4 induction was assessed by ratios of urinary 6β-hydroxycortisol : cortisol (6β-OHC : C) obtained via 24 h and morning-spot sampling techniques.

Results

Following repeated doses of rifampicin, both the mean C_max and AUC(0,∞) of ruboxistaurin were significantly reduced by approximately 95% (P ≤ 0.001). For the metabolite, the mean C_max decreased by 68% (P ≤ 0.001), and AUC(0,∞) decreased by 77% (P ≤ 0.001). The t_max values did not appear affected. The 6β-OHC : C ratios from both 24 h and morning spot methods increased significantly, consistent with CYP3A4 induction.

Conclusions

The effect of rifampicin co-administration on the exposure of ruboxistaurin is consistent with ruboxistaurin being a substrate of CYP3A4. Therefore, co-administration with known CYP3A4 inducing agents (rifampicin, carbamazepine, phenobarbital, etc.) may decrease the concentrations of ruboxistaurin and N-desmethyl-ruboxistaurin. In this study, 6β OHC : C ratios substantially underestimated the impact of rifampicin on ruboxistaurin.

Introduction

Ruboxistaurin mesylate (LY333531 or ruboxistaurin) is a selective inhibitor of the β 1 and β 2 isoforms of protein kinase C (PKCβ) [1]. The experimental drug is currently in late stage clinical development for the treatment of chronic diabetic microvascular complications such as neuropathy, retinopathy, and macular oedema [2–4]. Abnormal vascular function has been implicated as one mechanism in the pathogenesis of diabetic microvascular complications [5]. However, the mechanism through which hyperglycaemia results in vascular dysfunction has not been well established. Both in vivo and in vitro preclinical evidence suggest that hyperglycaemia-induced activation of PKCβ might underlie the development of microvascular disease into chronic diabetic complications [6–8].

In vitro studies using several techniques demonstrated that CYP3A4 is the primary cytochrome P450 enzyme responsible for the metabolism of ruboxistaurin to its main equipotent metabolite, N-desmethyl ruboxistaurin (compound 338522) [9]. The half-life of ruboxistaurin is approximately 9 h and that of N-desmethyl ruboxistaurin is approximately 16 h, thus allowing for once-daily dosing [10].

CYP3A4 is the most abundant form of cytochrome P450 in humans, accounting for approximately 30% and 70% of the total cytochrome content in the liver and intestinal wall, respectively [11, 12]. CYP3A4 is responsible for the metabolism of a wide range of drugs, including immunosuppressants and calcium channel blockers [13].

The primary objective of this study was to evaluate the effect of rifampicin on the single dose pharmacokinetics of ruboxistaurin and its active metabolite, N-desmethyl ruboxistaurin. Rifampicin is a potent inducer of several cytochrome P450 enzymes including CYP3A4, in both the intestinal wall and the liver [13]. Rifampicin also induces the expression of drug transporters such as P-glycoprotein, multiple drug resistance protein 2 and organic anion-transporting polypeptide 2 (cMOAT, or MRP2) [13].

The secondary objective of the study was to compare the changes in pharmacokinetic parameters of ruboxistaurin and N-desmethyl ruboxistaurin with the ratio of urinary 6β-hydroxycortisol to cortisol (6β-OHC : C). Cortisol is predominantly metabolized to 6β-hydroxycortisol [14] and 6β-OHC : C is a well-recognized endogenous marker for CYP3A4 activity [15, 16]. Traditionally, this measurement of the urinary 6β-OHC : C ratio has relied on a 24 h urine collection [15, 16]. A morning spontaneously voided urine (or morning spot urine) provides a more convenient method of assessing the 6β-OHC : C ratio. Results from studies where both collection methods have been used have suggested that there is a correlation between the morning spot and 24 h urinary measurements of the 6β-OHC : C ratio [17–20]. In this study, both the 24 h urine collection and the morning spot urinary measurement were obtained to evaluate further whether the latter could serve as a suitable substitute for the 24 h urine collection in assessing CYP3A4 activity.

Methods

Subjects

A total of 17 healthy Chinese male volunteers (age 21–27 years; weight 58–86 kg; BMI 19–26 kg m⁻²) were enrolled in this study after giving their written consent. Sixteen of the 17 subjects completed the study according to the protocol. One subject withdrew from the study after completing period 1. Data from all completed study periods were included in the pharmacokinetic and statistical analyses. Based on an estimated 23% intrasubject coefficient of variation from previous studies [10], 16 subjects completing the study would provide at least 80% power to show, at the 5% level, a 20% difference in ruboxistaurin AUC(0,∞) in the presence and absence of CYP3A4 induction.

Study design

The study protocol was approved by the National University Hospital Research and Ethics Committee and performed in compliance with the Declaration of Helsinki.

An open label, one-sequence study design with two periods was used. A one-sequence design with rifampicin dosing in the second period was chosen due to the possible long carry-over effects of rifampicin. In the first period, each subject received a single oral dose of 64 mg (4 × 16 mg tablets) of ruboxistaurin approximately 5 min after a standardized high fat breakfast. Ruboxistaurin absorption is enhanced by food [10]. After a washout interval of 4–14 days after the first period, each subject was dosed orally with 600 mg rifampicin (2 × 300 mg capsules) daily for 9 days in the second period. On the 7th day of the second period, rifampicin was dosed before the standardized high-fat breakfast, followed by a single 64 mg dose of ruboxistaurin after the meal. Previous studies have shown that full induction of drug-metabolizing enzymes is reached with 600 mg day⁻¹ rifampicin administered for about 7 days [13]. Rifampicin dosing was continued until the 9th day of the second period to maintain the maximum induction during the pharmacokinetic sampling period.

A single 64 mg dose of ruboxistaurin was used in this study. This dose is twice the anticipated therapeutic dose that is undergoing evaluation in large-scale clinical trials and was used since the plasma concentrations of ruboxistaurin were expected to be reduced by rifampicin treatment.

Blood sampling and urine sampling

Sequential venous blood samples (6 ml) were collected via a venous cannula or direct venepuncture prior to and at 0.5, 1, 1.5, 2, 2.5, 3, 4, 6, 8, 10, 12, 16, 24, 36, 48 and 72 h following ruboxistaurin administration. After collection, blood samples were kept at 4 °C prior to centrifugation and the plasma samples were stored at −20 °C until time of analysis. A 24 h complete urine collection was obtained during the 24 h preceding each ruboxistaurin dosing. Ruboxistaurin was administered after completing the urine collection to avoid the possibility of ruboxistaurin interfering with urinary cortisol assays. A morning spot urine sample was taken from the first morning void at the end of the 24 h collection period. The remainder of the first morning void was then pooled with the rest of the 24 h urine collection and the volume measured.

Analytical assay

Ruboxistaurin and N-desmethyl ruboxistaurin were measured in human plasma by turbo ion spray liquid chromatographic method using tandem mass spectrometry detection (LC/MS-MS) over the concentration range 0.5 ng ml⁻¹ to 150 ng ml⁻¹ using a method developed, validated, and used by Advanced BioAnalytical Services Inc., Ithaca, NY, USA. This method utilized ([²H 6] ruboxistaurin) as an internal standard. Heparinized human plasma was thawed, internal standard added, and analytes extracted by solid-phase extraction using a polypropylene 96-well solid phase extraction (SPE) block (isolute carboxypropyl acid stationary phase) procedure. Samples were eluted from the SPE block with 1% formic acid in methanol, evaporated to dryness under nitrogen at 45 °C, and reconstituted in 100 µl 75 : 25 methanol : 10 mm ammonium formate, pH 3.6. Samples were centrifuged at 3200 rev min⁻¹ for 5 min prior to injection into a Shimadzu HPLC system. Samples were applied to the column (Genesis CN, 4 µm, 50 mm × 2.1 mm) at ambient temperature, and the analytes were quantified by mass spectral analysis using a Sciex API 3000 triple quadrupole mass spectrometer and turbo ion spray in the positive mode. Validation intra-assay and interassay precision data (relative standard deviation) for the validation samples ranged from 0.87 to 17.46% for ruboxistaurin and from 1.26 to 12.93% for N-desmethyl ruboxistaurin. The validation intra-assay and interassay accuracy (relative error) for the validation samples ranged from −2.67% to 6.92% (97% to 107% accuracy) for ruboxistaurin and from −8.52 to 15.80% (91% to 116% accuracy) for N-desmethyl ruboxistaurin. Sample stability has been demonstrated up to 290 days at −20 °C in heparinized human plasma.

Urine samples were analyzed for cortisol and 6β-hydroxycortisol using a validated radio-immunoassay method. The method for the quantification of cortisol in urine was validated at PPD Development (Richmond, VA, USA) over the concentration range of 1.0–32.0 µg dl⁻¹. The samples were analyzed at PPD Development. The analytes were extracted from human urine by liquid-liquid extraction into dichloromethane for cortisol and by solid phase extraction with an Oasis HLB extraction cartridge for 6β-hydroxycortisol. Validation intra-assay and interassay precision data (relative standard deviation) for the validation samples ranged from 5.33 to 11.5% for cortisol and 7.15% to 10.7% for 6β-hydroxycortisol. The validation intra-assay and interassay accuracy (relative error) for the validation samples ranged from −3.51% to 5.04% for cortisol and 0.0664% to 3.13% for 6β-hydroxycortisol (measured as percentage difference from theoretical).

Pharmacokinetic analysis

In the analysis, concentrations of ruboxistaurin and its active metabolite, N-desmethyl ruboxistaurin were converted to their molar units to allow for comparisons of their relative amounts. All pharmacokinetic parameters were calculated by noncompartmental analyses from each subject's plasma data at actual sampling times using WinNonlin Professional Edition Version 2.1 (Pharsight Corporation, Mountain View, CA, USA). Maximum plasma concentrations (C_max) and the time to reach this concentration (t_max) were determined for ruboxistaurin and N-desmethyl ruboxistaurin directly from the original data. The terminal elimination rate constants (λ_z) were estimated by a least-squares method from the terminal log-linear phases of the plasma concentration-time curves visually identified in each profile. The apparent terminal elimination half-lives (t_1/2) were calculated from the equation t_1/2 = ln2/λ_z. The values of area under concentration-time curves from time of dose administration to the time of the last measurable concentration (AUC(0,t_last)) and through infinity (AUC(0,∞)) were calculated by linear trapezoidal rule for the rising phase and log-linear trapezoidal rule for the descending phase.

Statistical analysis

All pharmacokinetic parameters, except t_max, were natural log-transformed. Procedure Mixed of SAS System® for Windows, Version 6.12 (SAS Institute, Cary, NC) was used for analysing the effect of rifampicin. The model took into account the fixed effect of rifampicin administration and the random subject effect. Least-square geometric means were computed and compared in terms of ratio of means. The t_max values were compared using the Wilcoxon's signed rank test. Twenty-four hour and morning 6β-OHC : C ratios before and after induction by rifampicin were compared using a mixed effects analysis of variance (anova) model. The model included time (morning or 24 h) and rifampicin (before or after) and their interaction as fixed effects, and subject as random effect. The relationships between the parameters, urinary 6β-OHC : C ratio and AUC(0,∞) of ruboxistaurin, before and after induction with rifampicin were investigated. These were examined by a linear regression model for the ratios of these parameters after vs. before rifampicin induction. A P value of less than 0.05 was regarded as statistically significant.

Results

Pharmacokinetics of ruboxistaurin and N-desmethyl ruboxistaurin

Rifampicin significantly decreased the plasma concentrations of both ruboxistaurin and its metabolite N-desmethyl ruboxistaurin (Figure 1). The mean AUC(0,t_last) and AUC(0,∞) of ruboxistaurin declined by 96% (90% CI, P value: (95%, 97%), P ≤ 0.001). For N-desmethyl ruboxistaurin, the mean AUCs decreased by approximately 78% ((74%, 81%), P ≤ 0.001). The C_max values were also markedly reduced for both ruboxistaurin by 95%, ((94%, 96%), P ≤ 0.001) and N-desmethyl ruboxistaurin by 68% ((63%, 73%), P ≤ 0.001)) after rifampicin treatment. Reductions of AUC(0,t_last) and C_max were observed in all subjects (Figure 2). Rifampicin also reduced the mean estimated t_1/2 of ruboxistaurin from 13.8 h to 2.28 h, and that of N-desmethyl ruboxistaurin from 20.6 h to 13.7 h. However, t_max values for both compounds remained relatively constant at about 2 h. Apparent clearance (CL/F) was increased on average by more than 23-fold and apparent volume of distribution (V_z/F) was increased more than 4-fold (Table 1).

Figure 1.

Plasma concentrations (mean ± SEM) of ruboxistaurin (upper panel) and metabolite N-desmethyl ruboxistaurin (lower panel) in 16 healthy subjects after administration of 64 mg ruboxistaurin before (closed symbols) and after (open symbols) induction with rifampicin. Insets depict the same data on a semilogarithmic scale

Figure 2.

Individual peak plasma concentration (C_max) values and area under the concentration-time curve (AUC(0,t_last)) of ruboxistaurin and metabolite N-desmethyl ruboxistaurin in healthy subjects after single dose oral administration of 64 mg ruboxistaurin before (without rifampicin) and after (with rifampicin) induction with rifampicin

Table 1.

Summary of ruboxistaurin and N-desmethyl ruboxistaurin pharmacokinetic parameters before and after induction with rifampicin

	Ruboxistaurin			N-desmethyl ruboxistaurin
Parameter	Without rifampicin	With rifampicin	Ratio† of geometric means (90% CI)	Without rifampicin	With rifampicin	Ratio† of geometric means (90% CI)
n	17	16	–	17	16	–
Cmax	200	10.7	0.049	128	42.1	0.316
(nmol l−1)b	(63.3)	(5.68)	(0.041, 0.059)*	(37.6)	(16.4)	(0.271, 0.368)*
AUC(0,tlast)	1216	53.4	0.041	1640	373	0.222
(nmol l−1 h)b	(420)	(28.0)	(0.034, 0.050)*	(551)	(121)	(0.193, 0.256)*
AUC(0,∞)	1364	60.4	0.042	1831	406	0.219
(nmol l−1 h)b	(429)	(28.0)	(0.035–0.051)*	(600)	(122)	(0.189, 0.253)*
t1/2b	13.8	2.28	0.186	20.6	13.7	0.658
(h)	(10.1)	(0.82)	(0.141, 0.244)*	(7.09)	(3.87)	(0.562, 0.769)*
tmaxa	2.00	2.00	0.417c	2.50	2.00	0.374c
(h)	(1.00–4.00)	(0.500, 6.00)		(1.50, 4.00)	(0.500, 6.00)
CL/Fb	109	2639	23.5	NC	NC	NC
(l h−1)	(29.8)	(996)	(20.0, 28.5)*
VZ/Fb	2107	8718	4.35	NC	NC	NC
(l)	(1819)	(4548)	(3.23, 5.88)*

^at_max data are given as the median values with ranges

^bvalues are shown as the arithmetic mean (+ SD.)

^cP values based on Wilcoxon Signed Rank test

^†ratio of geometric means after: before rifampicin induction; P value ≤ 0.001; CI = confidence interval; NC = not calculated.

6β-hydroxycortisol : cortisol excretion ratios

The individual cross-plots of urinary 6β-OHC : C ratios calculated via 24 h and morning spot urine collection methods are represented in Figure 3. Several subjects had concentrations of cortisol below the level of quantification (BQL) and so the 6β-OHC : C excretion ratio in these subjects could not be assessed. For analysis of the 24 h urine collection method, 12 of 17 subjects had calculable values of the 6β-OHC : C ratio in period 1, 12 of 16 subjects had calculable values of the 6β-OHC : C ratio in period 2 during induction by rifampicin, and nine subjects had calculable 6β-OHC : C ratios for both study periods. All calculable ratios from both period 1-and period 2 are represented in Figure 3, and were used in the analysis of the extent of induction. The exclusion of the BQL measurements of 6β-OHC : C ratios from analysis would probably not have biased the analysis as there were comparable numbers of unquantifiable cortisol concentrations both pre- and post-rifampicin induction. For the morning spot-urine method, 13 of 17 subjects and 11 of 16 subjects had calculable 6β-OHC : C ratios when ruboxistaurin was administered alone and after rifampicin induction, respectively.

Figure 3.

Individual and mean (± SD) 6β-hydroxycortisol : cortisol ratio in urine samples collected via 24 h () and morning spot (), collection techniques before (−) and after (+) rifampicin induction

Rifampicin significantly increased the urinary 6β-OHC : C ratios in both the 24 h (average ratio increased from 2.15 to 5.66 [90% CI for difference (2.14, 4.88), P ≤ 0.001]) and the morning spot (average ratio increased from 1.72 to 3.62 [90% CI for difference (0.53, 3.28), P = 0.025]) urine collection methods, consistent with induction of CYP3A4. However, in contrast to the large reduction in plasma AUC of ruboxistaurin, both mean and individual changes in the 6β-OHC : C ratios show that the increases in the 6β-OHC : C ratio for both 24 h and morning spot urine were less pronounced (Figure 3). Indeed, for individual changes in urinary cortisol ratio, one subject showed a decrease in 6β-OHC : C ratio after rifampicin induction using the 24 h collection method and two subjects showed a decrease in the ratio after induction using the morning spot technique (Figure 3). There were no significant differences in the 6β-OHC : C ratios between the two urine collection methods prior to rifampicin induction. However the 24 h urine collection method produced a significantly higher ratio as compared with the morning spot technique (P = 0.014) after rifampicin induction. There was no correlation of 6β-OHC : C ratios (both before and after rifampicin induction) between the 24 h and morning spot urine collection techniques (data not shown).

Relationship between 6β-hydroxycortisol : cortisol ratio and pharmacokinetic parameters

The scatter plots of the nine measurable changes in urinary 6β-OHC : C ratios calculated via 24 h urine collection method and the changes in ruboxistaurin and its metabolite AUC are represented in Figure 4. As anticipated from the induction with rifampicin, the increase in 6β-OHC : C ratios, derived from both the 24 h and morning spot urine collection techniques, were accompanied by an increase in the apparent clearance of ruboxistaurin, and reductions in its C_max and AUC (Table 1). However, the correlation between the changes in 6β-OHC : C ratios from the 24 h collection methods and the changes in ruboxistaurin and N-desmethyl ruboxistaurin AUC(0, ∞) were not significant (Figure 4). Similar results were obtained from correlations between the changes in ruboxistaurin and N-desmethyl ruboxistaurin AUC(0, ∞) and the changes in 6β-OHC : C ratios from the morning spot technique (data not shown).

Figure 4.

Scatter plots of after/before rifampicin induction 6β-OHC : C ratios vs. after/before rifampicin induction AUC(0, ∞) after of ruboxistaurin (upper panel) and N-desmethyl ruboxistaurin (lower panel), calculated from 24 h collection technique

The safety findings in this study were consistent with previous ruboxistaurin studies, in that there were no clinically significant adverse events.

Discussion

In this study, the exposure (AUC and C_max) of ruboxistaurin and its active metabolite, N-desmethyl ruboxistaurin, were reduced by a very large extent following co-administration of rifampicin. A likely mechanism for the interaction between ruboxistaurin and rifampicin is the induction of CYP3A4 activity by rifampicin. Results from in vitro studies using several routine methods demonstrated that CYP3A4 is the primary route of ruboxistaurin metabolism [9]. The observed effects of rifampicin induction on the exposure of ruboxistaurin are consistent with other studies of CYP3A4 substrates such as midazolam and nifedipine whose AUCs are reduced by more than 90% after rifampicin induction [21–23].

In vitro studies have shown that ruboxistaurin metabolism is also mediated by CYP2D6. However, the in vitro contribution of CYP2D6 was very minor in relation to CYP3A4, with the N-desmethyl ruboxistaurin being formed by CYP3A4 at a rate 57-fold greater than by CYP2D6 [9]. Furthermore, current evidence suggests that rifampicin produces no clinically significant induction of CYP2D6 [13].

It is unclear if the induction of ruboxistaurin metabolism is solely related to induction of CYP3A4 within the liver, the gut, or both. If the increase in conversion of ruboxistaurin to N-desmethyl ruboxistaurin takes place in the gut (with absorption of both ruboxistaurin and N-desmethyl ruboxistaurin), the reduction in AUC(0,∞) of ruboxistaurin after rifampicin induction may be partly attributable to a reduced bioavailability of ruboxistaurin. Although not specifically addressed in this study, studies using midazolam, a well characterized and specific CYP3A4 substrate, to investigate the separate contributions of intestinal and hepatic CYP3A4 to its metabolism have shown that intestinal CYP3A4 contributes significantly to the increased metabolism of midazolam after CYP3A4 induction [24, 25]. These studies indirectly support the idea that intestinal CYP3A4 may have had a significant role in the metabolism of ruboxistaurin, as it did in the midazolam studies.

Also, the mean concentration-time curves in Figure 1 suggest that ruboxistaurin follows two-compartment pharmacokinetics. The estimated mean half-life reported in the post-rifampicin induction period may be associated more with the λ₁ (distribution) phase, as a λ_z (elimination) phase could not be observed in many profiles with most ruboxistaurin concentrations below the lower limit of quantification after 12 h. Hence it is not clear if the reductions in ruboxistaurin concentrations with rifampicin are due to an increase in clearance, a decrease in bioavailability, or a combination of both effects.

Although the metabolism of the metabolite N-desmethyl ruboxistaurin has yet to be elucidated, its shorter terminal half-life and reduced plasma AUC after rifampicin treatment suggest that it may also be metabolized by cytochromes P450 induced by rifampicin. However, rifampicin is a ‘general’ inducing agent and the clearance of ruboxistaurin and N-desmethyl ruboxistaurin by other pathways cannot be completely ruled out [13]. While P-glycoprotein (Pgp) mediated transport, for example, is also induced by rifampicin [13, 26–28], studies have shown that ruboxistaurin is not a substrate for this transporter [10]. Similar studies have not been performed for N-desmethyl ruboxistaurin, but given it structural similarity to ruboxistaurin, Pgp transport of N-desmethyl ruboxistaurin is unlikely.

The observed increases in 6β-OHC : C ratios after rifampicin induction using both the 24 h and morning spot collection are in agreement with other studies where treatment with rifampicin consistently increases urinary 6β-OHC : C ratios by up to 10 fold [15, 18, 19, 29–31]. The findings in the current study support the induction of CYP3A4 by rifampicin. Again, cortisol is also a substrate of Pgp and so metabolism through this and other pathways cannot be ruled out based on rifampicin's general inducing properties, as noted above [32]. Interestingly, the absolute values of mean 6β-OHC : C ratios seen in this study are lower than those reported in the other studies. The reason(s) for this difference is(are) unclear. In general, the human metabolism of cortisol is very complex, there is great variability and interindividual differences in the reported values of 6β-OHC : C ratios across studies, which may be due to one or a combination of factors such as different analytical techniques employed, physiological differences due to gender, age, hormonal status, ethnicity, genetic factors and genetic polymorphisms [14]. While there is evidence from a study in females that the 6β-OHC : C ratio may be lower in Chinese compared with Caucasians [33], it is unlikely that the lower 6β-OHC : C ratios reported in this study are due to ethnicity of the subject population, as Soon et al. recently reported lower baseline 6β-OHC : C ratios among Caucasians compared with Chinese subjects using a similar assay technique to the one used in this study [25].

There was a lack of correlation in this study between changes in urinary 6β-OHC : C ratio and pharmacokinetics of ruboxistaurin and its metabolite N-desmethyl ruboxistaurin, similar to previous studies that failed to find significant correlations between changes in urinary 6β-OHC : C and pharmacokinetic profiles of CYP3A4 substrates, such as cyclosporin [34], erythromycin [35] and midazolam [25, 36]. The reason for the noncorrelation in these studies is unknown. One limitation of this assessment is the small sample size in the urinary 6β-OHC : C ratio cohort. Previous studies have suggested that the 6β-OHC : C ratio may not be suitable for predicting the magnitude of the effect of CYP3A4 induction for an orally administered drug that is also subject to gut metabolism during first pass or enterohepatic cycling [5, 14]. In this study, large changes in plasma ruboxistaurin AUCs are not reflected in the 6β-OHC : C ratios obtained from both 24 h and spot urine collection methods and supporting the fact that the 6β-OHC : C ratio cannot be used to predict the magnitude of CYP3A4 induction. Therefore both techniques are not reliable quantitative measures of CYP3A4 induction. Furthermore, no correlation was found between the two collection methods, suggesting that there is inherent variability between different techniques.

In conclusion, the effect of rifampicin on the exposure of ruboxistaurin is consistent with ruboxistaurin being a substrate of CYP3A4. Therefore, co-administration with known CYP3A4 and inducing agents (rifampicin, carbamazepine, phenobarbital, etc.) may decrease the concentrations of ruboxistaurin and N-desmethyl-ruboxistaurin. In this study, urinary 6β-OHC : C ratios substantially underestimated the impact of rifampicin on ruboxistaurin.