![]() |
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Nephrology and Renal Transplantation, University Hospitals, Leuven, Belgium
(Received July 16, 2004; accepted September 15, 2004)
| Abstract |
|---|
|
|
|---|
We therefore conducted a prospective pharmacokinetic study in stable renal recipients on tacrolimus therapy, simultaneously measuring the rate of gastric emptying through a carbon-14-octanoic acid breath test, while at the same time quantifying drug exposure by AUC blood concentration sampling. The aim of the study was to determine the concurrent effect of gastric emptying on the rate and extent of tacrolimus absorption in stable renal recipients.
| Materials and Methods |
|---|
|
|
|---|
Inclusion and Exclusion Criteria. Patients who had received a single cadaveric donor kidney and were clinically stable were eligible for inclusion. Exclusion criteria were medical or surgical hepatic and gastrointestinal disorders, including active peptic ulcer disease and diabetes mellitus, that could interfere with the absorption, distribution, metabolism, or excretion of tacrolimus. Patients had to be free from biopsy-proven acute rejection for at least 6 months before the study, and any acute illness in the last 6 weeks was a contraindication for enrollment. Drugs known to affect gastric emptying were prohibited by the study protocol. Patients with a history of noncompliance were also excluded, as were recipients known to have a current drug, nicotine, or alcohol addiction. Finally, since this was a primary pharmacokinetic study, all substances documented to have a significant clinical effect on the absorption, distribution, metabolism, and excretion of tacrolimus were prohibited (Christians et al., 2002
). If patients required one of the latter drugs as maintenance therapy, they were excluded from the study. Approval was obtained from the ethics committee from the University of Leuven, Faculty of Medicine, and each patient gave his/her written informed consent.
Safety Parameters. Standard safety evaluation during this study included physical examination with systolic and diastolic blood pressure measurement, body weight, vital signs, and laboratory tests. The use of any concomitant medication was noted. Renal allograft function was assessed using serum creatinine determinations and creatinine clearance calculated by the Cockcroft-Gault formula.
Pharmacokinetic and Gastric Emptying Studies. Tacrolimus blood samplings were performed while, simultaneously, gastric emptying rate was measured using a radiolabeled octanoic acid breath test (Maes et al., 1994
) for all patients who consented. These measurements were done at 12 (n = 24) or 24 (n = 26) months post-transplantation. Patients had to adhere to an overnight fast for at least 10 h, and the morning dose of tacrolimus was ingested at the start of the test meal (see below), 12 h after the previous dose.
The abbreviated 4-h tacrolimus blood-sampling profile consisted of consecutive blood samples taken through an intravenous catheter, from predose (time point zero: C0) up until 4 h postdosing. Concomitant medication that could interfere with the absorption and metabolism of tacrolimus, including over-the-counter drugs, and the validity of the dose-interval were double-checked. Whole blood tacrolimus concentrations were determined using a microparticulate enzyme immunoassay (Tacrolimus II MEIA/IMx analyzer; Abbott Diagnostics, Abbott Park, IL).
Gastric emptying of a standard mixed solid-liquid meal was measured by means of the combined 14C-octanoic acid/13C-glycine gastric emptying breath test (Maes et al., 1994
). In short, a test meal composed of an egg omelette labeled with 74 kBq of 14C-octanoic acid (PerkinElmer Life and Analytical Sciences, Boston, MA), 60 g of white bread, and 150 ml of water labeled with 91 mg of 13C-glycine (Isotec, Miamisburg, OH) was consumed in less than 10 min. The meal consisted of 14 g of proteins, 26 g of carbohydrates, and 9 g (18.4%) of fat. The 13CO2 and 14CO2 excretion in breath was subsequently measured every 15 min for a total duration of 4 h and expressed as percentage of dose per hour to calculate a gastric emptying coefficient (GEC), a half-emptying time (t1/2), and a solid lag phase (tlag) as described in detail elsewhere (Maes et al., 1994
). A correlation exists between GEC, t1/2, and tlag determined by the breath test technique and radioscintigraphic techniques (respectively, r = 0.88, r = 0.92, and r = 0.89; p < 0.0001) (Maes et al., 1994
). The gastric emptying parameters calculated from the 14C-octanoic acid data, reflecting the solid-phase gastric emptying rate, were used for further analysis and are reported here. No additional information was obtained from measuring the 13CO2 excretion in breath (13C-glycine reflecting the liquid phase of gastric emptying) and, therefore, these data are not shown.
At all time points during the study, the investigators were blinded for the results of the tacrolimus pharmacokinetic studies and the gastric emptying data; adjustments of tacrolimus dose were made strictly based on single predose 12-h trough blood concentrations.
Clinical Interpretation of 14C-Octanoic Acid Gastric Emptying Breath Test Results. All results of the 14C-octanoic acid gastric emptying breath test were independently assessed by either of two investigators who were not involved in the study and who remained blinded at all times for the results of the pharmacokinetic studies and the clinical characteristics of the patients as well as their drug therapy. Gastric emptying was defined as delayed based on visual inspection of the 14CO2 breath excretion curve, a gastric half emptying time (t1/2) of more than 75 min, and/or a GEC value below 3.3 in comparison with gastric emptying data obtained in healthy volunteers described in detail elsewhere (Maes et al., 1994
, 1997
). To control for a possible interpretation bias, the GEC, t1/2, and tlag were compared post hoc between those recipients determined by the blinded investigators as having delayed gastric emptying (n = 24) and patients determined as having normal gastric emptying (n = 26). All gastric emptying parameters (GEC, t1/2, tlag) differed highly significantly between both groups (GEC, 3.82 ± 0.44 versus 3.28 ± 0.32, p < 0.0001; t1/2, 56.8 ± 16.3 min versus 102.4 ± 21.5 min, p < 0.0001; tlag, 28.7 ± 12.9 min versus 57.7 ± 18.1 min, p < 0.0001).
Determination of Pharmacokinetic Parameters. Pharmacokinetic modeling was performed using WinNonlin 3.2 Pro software (Pharsight, Mountain View, CA) and an SAS 8.02 statistical program (SAS Institute, Cary, NC). Model-independent pharmacokinetic parameters for tacrolimus were calculated and dose-corrected when appropriate. The corresponding AUC0-12h was calculated from the abbreviated 4-h AUC, using an algorithm previously validated in de novo renal recipients (Kuypers et al., 2004a
) that explained 96% of the variance in AUC0-12h with a mean percentage prediction error of -0.57 ± 5.6% (range, -13.6% to +15.1%) and a mean absolute prediction error of 4.7 ± 3.2% (range, 0.1% to 15.1%). Maximum blood concentration (Cmax), predose trough blood concentration (C0), and time to reach maximum blood concentration (tmax) were determined. An estimate of total steady-state body clearance was obtained from the calculated dose-interval AUC and tacrolimus dose.
Statistical Analysis. A sample size of 44 recipients was calculated (
= 0.05 and power of 80%) assuming that delayed gastric emptying would result in a clinically relevant decrease of tacrolimus exposure by 30% and taking into account a 10% dropout rate.
Distributions for continuous data were evaluated (Kolmogorov-Smirnov), and, consequently, parametric tests and nonparametric tests were applied when appropriate. Data are always expressed as mean ± standard deviation (S.D.) except when stated differentially (median + range). Nonparametric statistics were used (Kruskall-Wallis and Wilcoxon rank sum tests; SAS 8.2 software) for comparison. Simple regression analysis (Pearson's and Kendall's tau) for correlation of gastric emptying data with pharmacokinetic parameters, and demographic and laboratory variables, were performed as appropriate. Multiple linear regression analysis of pharmacokinetic parameters as dependent variables was applied with backward elimination and stepwise selection models for the different gastric emptying parameters and clinical variables. A p value < 0.05 was considered statistically significant.
| Results |
|---|
|
|
|---|
|
Gastric Emptying Parameters and Tacrolimus Pharmacokinetics. Gastric half emptying time (t1/2) significantly correlated with time to reach maximum blood tacrolimus concentration (r2 = 0.30; p < 0.0001). In patients with delayed gastric emptying, gastric half emptying time (t1/2) correlated significantly with time to reach maximum blood tacrolimus concentration (r2 = 0.18; p = 0.03), albeit weaker than for patients with normal gastric emptying (r2 = 0.37; p = 0.0009) (Fig. 1). Gastric half emptying time did not correlate significantly with any of the other dose-uncorrected or dose-corrected tacrolimus pharmacokinetic exposure parameters including AUC0-4h, AUC0-12h, Cmax, C0, oral clearance, and weight-corrected daily dose. The GEC, reflecting the overall gastric emptying rate for solids, showed a weak but significant inverse correlation with tmax (r2 = 0.14; p = 0.007) (Fig. 2). Again, GEC did not correlate with other tacrolimus exposure parameters. The lag phase for solids (tlag) was found not to correlate with tacrolimus exposure or with tmax. Using multiple linear regression analysis (backward and stepwise elimination model), with tmax as dependent variable and gastric emptying parameters as explanatory variables, only t1/2 was withheld as significant variable in the model.
|
|
The mean rate of gastric emptying strongly correlated with the simultaneously measured mean blood tacrolimus concentration over the first 4 h after oral drug administration (r2 = 0.96; p < 0.0001). This significant correlation was present both in recipients with (r2 = 0.81; p = 0.0003) and without (r2 = 0.97; p < 0.0001) delayed gastric emptying, albeit weaker in the former group (Fig. 3). The 14CO2 excretion curve in recipients with delayed gastric emptying, reflecting gastric emptying rate, clearly demonstrates a diminished ascending slope of the curve, a diminished and prolonged descending part, with a lower peak excretion rate of 14CO2, appearing significantly later (time to peak excretion rate, 1.90 ± 0.28 h versus 1.42 ± 0.22 h; p < 0.0001) in comparison with patients without delayed emptying. The blood tacrolimus concentration over time very closely follows a pattern similar to that of the gastric emptying rate in both patient groups. Comparison of tacrolimus pharmacokinetic parameters between stable patients with (n = 24) and without (n = 26) delayed gastric emptying confirms that the maximum blood tacrolimus concentration is reached significantly more slowly in the former group (tmax, 2 ± 1 h versus 1.48 ± 0.68 h; p = 0.04), whereas the extent of tacrolimus absorption and exposure is not different, as reflected by Cmax, AUC, or C0 (Table 2). The dose of tacrolimus necessary to obtain comparable drug exposure did not differ between groups, even after correction for body weight. No significant correlation was detected between gastric emptying parameters (GEC, t1/2, tlag) and recipient age, weight, allograft function, or albumin and hemoglobin concentration; neither could an effect of gender be demonstrated (data not shown).
|
|
| Discussion |
|---|
|
|
|---|
Tacrolimus, due to its macrolide structure, exhibits intrinsic prokinetic properties, as opposed to cyclosporine, as was clearly demonstrated in renal and lung transplant recipients (Maes et al., 1999
; Verleden et al., 2002
). It is possible that extreme delays in gastric emptying rate, encountered, as for example, in some cases of severe diabetic gastroparesis, were not detected in the present study because primary diabetic patients were excluded from the trial and the prokinetic effects of tacrolimus on gastric motility could have masked (ameliorated) strongly delayed emptying in some patients. In fact, three patients in the present study developed post-transplantation diabetes mellitus, but their gastric emptying parameters did not differ significantly from those of the nondiabetic patients; nor were there any significant differences in tacrolimus exposure parameters (data not shown). When we considered only those recipients with a strong delay in gastric emptying, in the upper quartile of gastric half emptying times, again, no differences in tacrolimus exposure (AUC0-12h, Cmax) could be established (data not shown). Of course, very early after successful transplantation, gastric motility might be even more extensively disturbed than in chronic stable recipients, possibly resulting in more marked alterations of tacrolimus absorption.
The effect of diurnal variation in tacrolimus exposure (Tada et al., 2003
) and ethnicity (Mancinelli et al., 2001
) could not have played a role in the current study since all measurements were performed in the morning, after an overnight fast, and all participants were of Caucasian origin. Also, concomitant drugs that could have influenced gastric emptying were excluded per protocol and double-checked on every visit. It is therefore unlikely that concomitant medication would have caused a systematic bias. Mycophenolate mofetil, administered routinely in this study, does not affect gastric emptying, as was demonstrated in renal recipients (Maes et al., 2003
), and could therefore not have altered our findings.
The ingestion of tacrolimus together with a low-fat test meal might of course have influenced our results but, at the same time, probably reflects more accurately real life. It is known that the exact timing of tacrolimus intake in relation to a meal is important in terms of tmax and Cmax but not for total drug exposure (Kimikawa et al., 2001
). In contrast, in healthy volunteers, timing of a single 5-mg tacrolimus dose immediately after a meal does affect drug exposure significantly (Bekersky et al., 2001b
). Extrapolating pharmacokinetic data from healthy volunteers and pretransplantation (uremic) circumstances to chronic stable transplant recipients is, however, prone to error (Satoh et al., 2001
). Furthermore, the fat content of a meal also influences tacrolimus absorption, at least in healthy volunteers (Bekersky et al., 2001a
). In the current study, tacrolimus was administered exactly timed with a standardized test meal containing 18% fat, and this methodology was strictly adhered to throughout the study for all participants. In contrast to van Duijnhoven et al. (1998
), who used a high-fat (43%) meal in diabetic uremic transplant candidates, we could not demonstrate an effect of the low-fat test meal on the extent of tacrolimus absorption (Cmax) in patients with delayed gastric emptying compared with patients with normal gastric emptying. Similarly, Christiaans et al. (1998
) could not find a significant effect of a high-fat meal on tacrolimus absorption in renal recipients at different time points after transplantation.
Considering tacrolimus as a poorly water-soluble drug with low dissolution characteristics in its current formulation (Yamashita et al., 2003
), it could be argued on theoretical grounds that the gastric emptying rate per se will have less impact on drug absorption (Kaus et al., 1999
). Furthermore, the intrinsic prokinetic effect of tacrolimus, as opposed to cyclosporin A (Maes et al., 1999
), could ameliorate mild to moderate delays in gastric emptying to such a degree that the extent of drug absorption remains unaffected. Nevertheless, it appears that tacrolimus absorption and subsequent exposure in stable recipients is mainly determined at more distal parts of the gastrointestinal tract (jejunum, ileum, colon) and regulated through the activity of P-glycoprotein and CYP3A4 in the gut wall and liver (Tuteja et al., 2001
; Shimomura et al., 2002
; Tamura et al., 2003
).
In conclusion, despite the fact that there exists a strong correlation between the rate of gastric emptying in stable renal allograft recipients and the timing of tacrolimus absorption from the gut, gastric emptying rate does not affect the total extent of drug absorption and is therefore not responsible for significant alterations in drug exposure, even in situations of delayed gastric emptying.
| Footnotes |
|---|
ABBREVIATIONS: AUC, area under the concentration-time curve; GEC, gastric emptying coefficient; tlag, solid lag phase; t1/2, half emptying time.
Address correspondence to: Dr. Dirk R. J. Kuypers, Department of Nephrology and Renal Transplantation, University Hospitals Leuven, Herestraat 49, B-3000 Leuven, Belgium. E-mail: Dirk.Kuypers{at}uz.kuleuven.ac.be
| References |
|---|
|
|
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||