Clinical and Experimental Nephrology

, Volume 17, Issue 5, pp 743–749

Excess fluid distribution affects tacrolimus absorption in peritoneal dialysis patients

Authors

    • Division of Nephrology and Dialysis, Department of Cardiorenal and Cerebrovascular Medicine, Faculty of MedicineKagawa University
  • Masashi Inui
    • Department of Urology, Faculty of MedicineKagawa University
  • Hideyasu Kiyomoto
    • Division of Nephrology, Endocrinology and Vascular Medicine, Department of MedicineTohoku University Graduate School of Medicine
  • Kumiko Moriwaki
    • Division of Nephrology and Dialysis, Department of Cardiorenal and Cerebrovascular Medicine, Faculty of MedicineKagawa University
  • Taiga Hara
    • Division of Nephrology and Dialysis, Department of Cardiorenal and Cerebrovascular Medicine, Faculty of MedicineKagawa University
  • Kazunori Yamaguchi
    • Department of PharmacyKagawa University Hospital
  • Noriyasu Fukuoka
    • Department of PharmacyKagawa University Hospital
  • Kazuko Banno
    • Shiga Pharmaceutical Association Proof Center
  • Akira Nishiyama
    • Department of Pharmacology, Faculty of MedicineKagawa University
  • Yoshiyuki Kakehi
    • Department of Urology, Faculty of MedicineKagawa University
  • Masakazu Kohno
    • Division of Nephrology and Dialysis, Department of Cardiorenal and Cerebrovascular Medicine, Faculty of MedicineKagawa University
Original Article

DOI: 10.1007/s10157-012-0764-6

Cite this article as:
Sofue, T., Inui, M., Kiyomoto, H. et al. Clin Exp Nephrol (2013) 17: 743. doi:10.1007/s10157-012-0764-6

Abstract

Background

Excess fluid distribution is a common disorder in peritoneal dialysis (PD) patients. Tacrolimus malabsorption may also occur in PD patients, and may lead to acute allograft rejection after transplantation. The purpose of this study was to evaluate the relationship between tacrolimus pharmacokinetics and excess fluid distribution according to pre-transplant dialysis modality.

Methods

We retrospectively analyzed 41 adult living-donor kidney transplantations, including nine PD patients and 32 hemodialysis (HD) patients. We examined tacrolimus pharmacokinetics in the peri-operative period and determined the association between the tacrolimus absorption rate and body weight reduction. The absorption efficacy of tacrolimus was evaluated as the dose-normalized tacrolimus absorption rate. Tacrolimus concentrations in PD effluent were measured by high-performance liquid chromatography.

Results

The tacrolimus absorption rate on the day before kidney transplantation tended to be lower in PD patients than in HD patients; however, the rate improved after kidney transplantation and was similar in both groups of patients. The peak tacrolimus concentration time was later in PD patients than in HD patients. The body weight reduction after kidney transplantation was greater in PD patients than in HD patients, and was significantly associated with the change in tacrolimus absorption rate (p = 0.04, r = 0.32). Only 0.002 % of the oral tacrolimus dose was removed by PD itself.

Conclusion

Excess fluid distribution in PD patients appears to contribute to tacrolimus malabsorption rather than PD itself. We should consider the risk of tacrolimus malabsorption in patients with possible excess fluid distribution, particularly in PD patients.

Keywords

Peritoneal dialysisExcess fluid distributionTacrolimus pharmacokineticsLiving-donor kidney transplantation

Introduction

Several renal replacement therapies are available, including hemodialysis (HD), peritoneal dialysis (PD) and kidney transplantation. Kidney transplantation is the preferred strategy for patients with end-stage renal disease for several reasons, including significant improvements in quality of life, better patient survival and decreased medical expenditure as compared with dialysis [13]. PD patients are generally younger than HD patients and the transfer rate to kidney transplantation is higher in PD patients than in HD patients [4].

Excess fluid distribution is a common disorder in PD patients because of the lower levels of salt and water removal, and the dependency for residual renal function [58]. Of note, an earlier study using bioimpedance measurement demonstrated that total body water was greater in PD patients than in HD patients before HD, although plasma ANP levels were similar in PD patients and HD patients before HD [7]. Therefore, excess fluid distribution is considered to be a major cause of intestinal edema, which influences the blood concentration of intestinal absorptiveness drugs, such as calcineurin inhibitors (CNI) [9].

Tacrolimus, a CNI, is widely used as an immunosuppressant in kidney transplantation and has improved allograft outcomes [10]. However, the therapeutic dose range of tacrolimus is narrow; under-dosing is associated with organ rejection and overdosing may be toxic [11, 12]. Tacrolimus is also associated with broad inter-individual pharmacokinetic variability, which makes determination of the initial dose difficult. The inter-individual differences in the bioavailability of tacrolimus are at least partly due to differences in its absorption, and may be influenced by meals, genotypes of cytochrome P (CYP) and multidrug resistance-1 (MDR1) in the intestine [13, 14], and potential interactions with other medicines, such as proton pump inhibitors (PPI) and calcium channel blockers (CCB) [15, 16].

A previous paper reported that tacrolimus absorption was inadequate in PD patients [17]. However, it is still unclear whether dialysis modality and excess fluid distribution can adversely affect tacrolimus absorption. Furthermore, because of the large molecular weight and high protein-binding rate of tacrolimus, it is important to determine whether PD itself can remove tacrolimus into the PD effluent. However, to the best of our knowledge, no reports have described whether or not tacrolimus is removed into the PD effluent. Moreover, the pharmacokinetic properties of tacrolimus in PD patients have not yet been determined. Therefore, the purpose of this study was to better understand whether there are differences in the relationship between tacrolimus absorption and excess fluid distribution among pre-transplant dialysis modalities.

Patients and methods

Patients

We retrospectively analyzed 41 adult patients who underwent living-donor kidney transplantation with a tacrolimus-based immunosuppressant regimen (Prograf™; Astellas Pharma Inc., Tokyo, Japan) at Kagawa University Hospital between August 2003 and June 2010. Recipients of deceased-donor kidney transplantation, preemptive kidney transplantation or kidney transplantation in combination with cyclosporine or prolonged-release tacrolimus (Graceptor™; Astellas Pharma Inc.) were excluded from this study. The protocol, patient information sheet and informed consent form were reviewed and approved by the Ethics Committee of Kagawa University [18].

Study design

The 41 recipients were classified into two subgroups according to their pre-transplant dialysis modality, as those who received PD (n = 9) or HD (n = 32) before kidney transplantation. Immunosuppressive regimens were based on four drugs, which were used in combination, including tacrolimus, mycophenolate mofetil, methylprednisolone and basiliximab. The type of CNI was not switched during the study. Patients with clear edema and cardiomegaly on chest X-ray images were defined as having clinical excess fluid distribution [19]. The body weight (BW) reduction after kidney transplantation was calculated using the following formula: BW reduction (%) = [BW on the day before transplantation − BW at 14 days after transplantation]/BW on the day before transplantation × 100. BW in PD patients was measured once the abdominal cavity was in an empty state. PD was terminated at the time of transplantation by removing the PD catheter. Acute allograft rejection (AR) was defined using the Banff criteria [20] as the presence of AR or a borderline change on post-transplant allograft biopsies.

Blood tacrolimus concentrations

Blood tacrolimus concentrations were determined using a microparticle enzyme immunoassay [MEIA] (IMx™; Abbott Laboratories; Abbott Park, IL, USA) [21]. Tacrolimus administration was started 7 days before kidney transplantation at an initial dosage of 0.1 mg/kg BW. Tacrolimus was orally administered twice daily at 09:00 and 21:00 hours (approximately 2 h after meals). The trough tacrolimus blood concentration (C0) was determined as the blood tacrolimus concentration measured immediately before the administration of tacrolimus, and was measured several times before transplantation. The tacrolimus dose was adjusted to achieve minimum C0 levels of 8 ng/mL by the day before transplantation or 10 ng/mL at 14 days after transplantation. The day before transplantation, whole blood samples were collected just before and at 0.5, 1, 2, 4, 6 and 8 h after oral tacrolimus administration at 09:00 hours, and shortly before the evening dose (i.e., 12 h). The area under the blood concentration–time curve (AUC)0–12 was calculated using the linear trapezoidal rule. The BW-normalized tacrolimus dose (0.1 mg/kg) was calculated using the following formula: 0.1 × tacrolimus dosage (mg)/the patient’s BW (kg). The absorption efficacy of tacrolimus was evaluated as the dose-normalized tacrolimus absorption rate, as previously described [22], and was calculated using the following formula: C0 (ng/mL)/BW-normalized tacrolimus dose (0.1 mg/kg). The change in tacrolimus absorption rate was calculated using the following formula: value on day 14 after transplantation − the value before transplantation/the value before transplantation × 100 (%). PD effluent in the abdominal cavity was not exchanged during the AUC measurement.

Tacrolimus concentration in the PD effluent

The concentrations of tacrolimus and its metabolic products 13-o-demethylated tacrolimus (M-I) and 31-o-demethylated tacrolimus (M-II) in PD effluent were measured by high-performance liquid chromatography (API-4000 LC/MS/MS system; Applied Biosystems/MDS Sciex, Toronto, Canada) using an Inertsil ODS-4 2.1 mm × 100 mm (3 μm) column (GL Sciences, Tokyo, Japan), as previously described [23, 24]. Glucose-based Dianeal-N™ PD-4 (1.5 %, 1500 mL; 1.36 % glucose; Baxter Healthcare, Round Lake, IL, USA) was used for PD effluent analyses. Two days before transplantation, the PD effluent was obtained after a 4-h dwell from one non-diabetic 41-year-old woman who was undergoing PD with Dianeal-N™ PD-4 (1.5 %). The patient gave informed consent for this procedure. The PD fluid was exchanged three times between 09:00 and 21:00 hours. All three bags of PD effluent were analyzed to determine the total amount of tacrolimus removed by PD. The three bags were collected at 13:00 (bag 1), 17:00 (bag 2) and 21:00 (bag 3) hours. From 7 days before transplantation, this patient was administered 3 mg tacrolimus twice daily at 09:00 and 21:00 hours, including on the day of PD effluent collection. C4 was determined as the blood tacrolimus concentration measured 4 h after tacrolimus administration. The rate of tacrolimus removal by PD was determined using the following formula: total amount of tacrolimus removed/by total daily tacrolimus dose.

Statistical analysis

Statistical analyses were performed using SPSS software version 20.0 for Windows (IBM Japan, Tokyo, Japan). Values are presented as means and standard deviation. Values of p < 0.05 were considered statistically significant. Clinical variables were compared between groups using the chi-squared test for categorical variables or Student’s t test for continuous variables.

Results

Comparison of pre-transplant recipient characteristics between dialysis modalities

The comparison of pre-transplant recipient characteristics between dialysis modalities is shown in Table 1. The recipient characteristics, including age, prevalence of hypertension, prevalence of diabetes, duration of dialysis, ABO-incompatible living-donor kidney transplantation, body mass index, plasma albumin level and hematocrit were similar between the two modalities. None of the recipients had obvious edema or cardiomegaly at the time of transplantation. The proportions of PD and HD patients using PPIs or CCBs in the perioperative period were also similar. The most commonly used PPIs were lansoprazole or rabeprazole; none of the patients used omeprazole.
Table 1

Pre-transplant recipient characteristics

 

PD patients

HD patients

p value

Number

9

32

 

 Age (years), mean (SD)

45 (3)

47 (2)

0.67

 Male, n (%)

4 (44)

26 (81)

0.04*

 Prevalence of hypertension, n (%)

4 (44)

23 (72)

0.23

 Prevalence of diabetes, n (%)

1 (7)

11 (33)

0.24

 Duration of dialysis (month), mean (SD)

49 (11)

53 (15)

0.88

 ABOIKT, n (%)

3 (43)

14 (44)

1.00

 Body mass index (kg/m2), mean (SD)

21 (1)

22 (1)

0.49

 Plasma albumin (g/dL), mean (SD)

3.6 (0.3)

3.8 (0.1)

0.25

 Hematocrit (%), mean (SD)

31 (4)

32 (5)

0.53

 Administration of PPI, n (%)

9 (100)

30 (94)

1.00

 Administration of CCB, n (%)

4 (44)

17 (53)

0.72

PD peritoneal dialysis, HD hemodialysis, SD standard deviation, ABOIKT ABO-incompatible kidney transplantation, PPI proton pump inhibitors, CCB calcium channel blockers

p < 0.05

Pharmacokinetic properties of tacrolimus in the peri-operative period

The pharmacokinetic properties of tacrolimus in PD and HD patients in the peri-operative period are shown in Fig. 1. On the day before transplantation, the C0 for tacrolimus tended to be lower in PD patients than in HD patients (Fig. 1b), although the BW-normalized dose was similar in both groups (Fig. 1a); however, the difference in C0 was not statistically significant. The tacrolimus absorption rate on the day before transplantation also tended to be lower in PD patients than in HD patients, although this was not statistically significant (Fig. 1c). The tacrolimus absorption rate improved in PD patients at 14 days after transplantation, reaching a level similar to that in HD patients. The change in tacrolimus absorption rate was significantly greater in PD patients than in HD patients (221 ± 47 vs. 148 ± 13 %, p = 0.04). Overall, 44 % of PD patients and 59 % of HD patients had adequate C0 tacrolimus concentrations on the day before transplantation, increasing to 89 and 72 %, respectively at 14 days after transplantation (Fig. 1d).
https://static-content.springer.com/image/art%3A10.1007%2Fs10157-012-0764-6/MediaObjects/10157_2012_764_Fig1_HTML.gif
Fig. 1

Tacrolimus pharmacokinetics at day −1 and day 14 in patients undergoing PD or HD. a BW-normalized dose, b C0, c absorption rate and d proportion of patients with adequate C0. HD hemodialysis, PD peritoneal dialysis, BW body weight

The tacrolimus AUC0–12 profiles on the day before transplantation are shown in Fig. 2. The peak tacrolimus concentration was delayed in PD patients compared with that in HD patients (Fig. 2a). However, the maximum concentration, C6 and C8 of tacrolimus were similar between the PD and HD patients. The AUC0–12 of tacrolimus was also similar in both groups of patients (Fig. 2b).
https://static-content.springer.com/image/art%3A10.1007%2Fs10157-012-0764-6/MediaObjects/10157_2012_764_Fig2_HTML.gif
Fig. 2

Tacrolimus AUC0–12 profiles on the day before transplantation. a Serial tacrolimus concentrations in patients undergoing PD or HD. b Tacrolimus AUC0–12 in patients undergoing PD or HD. AUC0–12 area under the curve for 0–12 h, HD hemodialysis, PD peritoneal dialysis, TAC tacrolimus

BW reduction and tacrolimus absorption rate after kidney transplantation

Although no obvious excess fluid distribution was observed in the PD patients before transplantation, the BW reduction after kidney transplantation was significantly higher in PD patients than in HD patients (Fig. 3a). The association between BW reduction and the change in tacrolimus absorption rate in all patients is shown in Fig. 3b. BW reduction after kidney transplantation was associated with the change in tacrolimus absorption rate (p = 0.04, r = 0.32). The association between BW reduction and the change in tacrolimus absorption rate is shown in Fig. 3c, d for PD and HD patients, respectively. The association was slightly stronger in PD patients than in HD patients, although it was not statistically significant in either group (r = 0.49, p = 0.18 vs. r = 0.18, p = 0.33).
https://static-content.springer.com/image/art%3A10.1007%2Fs10157-012-0764-6/MediaObjects/10157_2012_764_Fig3_HTML.gif
Fig. 3

a BW reduction after kidney transplantation in patients undergoing PD or HD. Association between BW reduction and the change in tacrolimus absorption rate in all patients (b), in PD patients (c), and in HD patients (d). *p < 0.05 versus HD patients. BW body weight, HD hemodialysis, PD peritoneal dialysis, TAC tacrolimus

Concentrations of tacrolimus in PD effluent

The concentrations of tacrolimus in the PD effluent and blood samples from a non-diabetic 41-year-old woman are shown in Table 2. Although her blood tacrolimus concentrations measured in terms of C0, C4 and AUC0–12, reached therapeutic levels, the concentrations of tacrolimus, M-I and M-II in the PD effluent after a 4-h dwell were extremely low. The total amount of tacrolimus present in the PD effluent was 135 ng, corresponding to 0.002 % of the oral dose.
Table 2

Blood and PD effluent concentrations of tacrolimus and its metabolites

 

Blood

PD effluent

TAC

TAC

M-I

M-II

C0 (ng/mL)

12.4

   

Blank fluid (ng/mL)

 

<0.03

<0.03

<0.03

C4 (ng/mL)

27.9

   

Bag 1 (ng/mL)

 

0.03

0.04

<0.03

Bag 2 (ng/mL)

 

0.03

0.03

<0.03

Bag 3 (ng/mL)

 

<0.03

<0.03

<0.03

AUC0–12 (ng h mg−1)

212.0

   

Total amount of TAC removed (ng)

 

135

140

0

Rate of TAC removal (%)

 

0.002

  

TAC tacrolimus, M-I 13-o-demethylated tacrolimus, M-II 31-o-demethylated tacrolimus, C0 trough tacrolimus blood concentration, C4 blood tacrolimus concentration measured 4 h after tacrolimus administration, AUC0–12 area under the blood concentration–time curve for 0–12 h

Transplant prognoses

Pre-transplant dialysis modality did not affect allograft outcomes, including patient and graft survival rates, or allograft function (data not shown). Additionally, the rate of AR was not affected by pre-transplant dialysis modality (Fig. 4). PD-related complications, such as encapsulating peritoneal sclerosis or ascites retention after transplantation, were not observed in any of the PD patients during this study.
https://static-content.springer.com/image/art%3A10.1007%2Fs10157-012-0764-6/MediaObjects/10157_2012_764_Fig4_HTML.gif
Fig. 4

Comparison of the acute allograft rejection rate between patients undergoing PD or HD. AR acute allograft rejection, HD hemodialysis, PD peritoneal dialysis

Discussion

The clinical introduction of tacrolimus has contributed to improvements in graft survival following organ transplantation [10]. However, the broad inter-individual pharmacokinetic variability of tacrolimus remains an important problem in the field of organ transplantation. Furthermore, tacrolimus malabsorption at the time of transplantation is a major cause of subsequent AR. In the present study, the tacrolimus absorption rate at the time of transplantation tended to be lower in PD patients than in HD patients, similar to that reported in an earlier study [17]. To identify the possible causes of tacrolimus absorption insufficiency in PD patients, we determined whether tacrolimus was removed from the blood into the PD effluent. In the present study, the blood tacrolimus concentrations were measured using an MEIA method in samples containing red blood cells. Because of the absence of red blood cells in PD effluent, we used tandem mass spectrometry to measure the tacrolimus concentrations in PD effluent. Using this approach, we showed that an extremely small amount of tacrolimus was removed into the PD effluent. Therefore, PD itself does not appear to be the main cause of tacrolimus malabsorption in PD patients.

Excess fluid distribution is a common disorder associated with PD [58]. In the present study, excess fluid distribution was not clinically apparent in any of the PD patients, but their BW decreased immediately after transplantation. Although the change in BW is influenced by muscle breakdown or nutrient intake, we considered that the short-term changes in BW after transplantation strongly reflect changes in body fluid. Accordingly, we considered that PD patients had ‘subclinical’ excess fluid distribution before transplantation. We also showed an association between BW reduction and improvement in tacrolimus absorption rate. This indicates that tacrolimus absorption improved in PD patients following improvements in excess fluid distribution. From these results, we considered that excess fluid distribution contributes to tacrolimus malabsorption.

It should be considered that intestinal edema is strongly influenced by excess fluid distribution. Because tacrolimus absorption is primarily controlled by efflux pumps and enzymes belonging to the CYP 450 family in the jejunum, intestinal edema may be a cause of tacrolimus malabsorption [2527]. However, intestinal edema is usually subclinical and cannot be detected during conventional clinical examinations. Nevertheless, we should consider the risk of tacrolimus malabsorption in patients with possible excess fluid distribution, particularly PD patients. Notably, however, tacrolimus malabsorption also occurred in HD patients with excess fluid distribution.

Based on our present results, the major characteristics of tacrolimus pharmacokinetics in PD patients were an inadequate dose-normalized absorption rate and the maximum drug concentration time was delayed. These characteristics are primarily due to excess fluid distribution, which affects tacrolimus absorption in the jejunum. Furthermore, the decrease in tacrolimus concentrations from the maximum concentration to C6 was quicker in PD patients than in HD patients. PD patients were considered to have a larger distribution volume because of their excess fluid distribution compared with HD patients. Therefore, the distribution rate of tacrolimus is expected to be greater in PD patients. This may contribute to the rapid decrease in the tacrolimus concentration from the maximum concentration to C6 in PD patients.

Importantly, the rate of acute rejection was similar between PD patients and HD patients. This means that frequent monitoring of the blood tacrolimus concentration is important to maintain the tacrolimus concentration within an appropriate therapeutic range. In the present study, we assessed the tacrolimus absorption rate (ng mL−1/0.1 mg kg−1) to determine the absorption efficacy, as previously described [22]. Practical use of this index may help us to determine its absorption efficacy in individual patients.

Because of the small number of patients and the retrospective design of this study, there are some limitations that should be discussed. First, the tacrolimus content in PD effluent was only measured in one patient. Second, we did not confirm the presence of excess fluid distribution in PD patients. Although we defined the change in BW as a marker for excess fluid distribution, studies using bioimpedance measurement are needed to confirm our results. Third, there was no significant difference in the tacrolimus absorption rate before transplantation between the PD and HD patients. Fourth, we could not determine the confounding effects of CYP and MRD1 genotypes on the tacrolimus pharmacokinetics. Further studies involving a larger number of patients and prospective studies are needed to confirm our results.

Conclusions

In summary, we found some associations between subclinical excess fluid distribution and tacrolimus malabsorption. On the other hand, PD transferred negligible amounts of tacrolimus into the PD effluent. By accounting for possible tacrolimus malabsorption in patients with potential subclinical excess fluid distribution, we should adjust the tacrolimus dosing regimen to maintain adequate therapeutic concentrations.

Acknowledgments

We wish to thank Ms. Yoshiko Fujita for performing laboratory tests and Ms. Chigusa Nakai for preparing the clinical data. We also thank Astellas Pharma Inc. (Tokyo, Japan) for supplying 13-o-demethylated tacrolimus, 31-o-demethylated tacrolimus and tacrolimus. This work was supported in part by a Grant-in-aid for scientific research from the Ministry of Education, Culture, Sports, Science and Technology of Japan (No. 24791653 to Tadashi Sofue).

Conflict of interest

The authors have declared that no conflicts of interest exist.

Copyright information

© Japanese Society of Nephrology 2012