Critically ill patients are often complicated with acute renal failure and thus are hemodynamically unstable. Continuous renal replacement therapy (CRRT) has proven to be a convenient extracorporeal technique to treat renal failure and subsequent fluid overload in critically ill patients. However, there is no uniformity in the pharmacokinetics of drugs used during CRRT because the procedure is performed with the use of many different combinations of dialysate flows (QD) and ultrafiltrate flows (QF).
Critically ill patients often have various severe infections and need to be intravenously administered antibiotics with a broad spectrum. For these patients, panipenem/beta Mipron (PAPM/BP) is often intravenously administered because it demonstrates good clinical and bacteriological efficacy.1 Panipenem is carbapenem that has a broad spectrum of activity covering Gram-negative and Gram-positive aerobic and anaerobic bacteria.1,2 Panipenem is administered with BP, an organic anion tubular transport inhibitor with very low toxicity that inhibits the active transport of PAPM in the renal cortex, thereby reducing the nephrotoxic potential of PAPM.1 For adult patients with normal renal function, the recommended dosage of PAPM/BP is 0.5 g for a period of at least 30 minutes every 12 hours.1 This dosage may be safely increased to 1 g for a period of at least 60 minutes every 12 hours.1
The influences of CRRT on the pharmacokinetics of PAPM/BP are therefore suspected to depend on QD and QF. The aims of this study were thus as follows: 1) to estimate total PAPM/BP clearance during CRRT, based on QD, QF and renal function of patients, and 2) to determine the appropriate dose regimens of PAPM/BP, based on the total PAPM/BP clearance in critically ill patients during CRRT.
Materials and Methods
In Vitro Study
An automated CRRT system (ACH-10, Asahi Medical Co., Tokyo, Japan) was used. A CRRT circuit was set up using a cellulose triacetate hollow-fiber 1.1 m2 hemofilter (UT-110, Nipro, Osaka, Japan) and filled with 5% bovine serum albumin (BSA) (Figure 1). The BSA was circulated for 10 minutes to adhere proteins to the circuit. Thirty milligrams each of PAPM and BP was mixed with 200 ml of 5% BSA in a reservoir. The CRRT conditions were as follows: the BSA flow was fixed at 150 ml/min; QD was defined from 0, 1, and 2 l/h; QF was defined from 0, 1, and 2 l/h, independent of QD; normal saline was used as a dialysate and also served as a replacement fluid infused postdilusionally with an equal amount of QF. We took samples from the prehemofilter and posthemofilter. Samples from the filtrates were also taken (Figure 1). The sampling times were 15, 30, 60, and 120 minutes after the start of CRRT. All samples were mixed with an equal amount of 1 M MOPS buffer (pH 7.0) and frozen at –80°C until analysis. These studies were performed repeatedly for eight different combinations of QD and QF.
This study was performed in the Intensive Care Unit of Hokkaido University Hospital. Four patients with acute renal failure, who were being treated by CRRT and receiving PAPM/BP intravenously, were studied. Approval for the study was obtained from the ethics committee of our institution, and written informed consent was obtained from each patient's next of kin. Table 1 shows the patients' detailed clinical backgrounds. The creatinine clearances (CLcre) were measured on the basis of serum and urine creatinine levels, the urine volume for 24 hours, the body weight, and the height of the patients.
In patient 1, 0.5 g of PAPM/BP was intravenously administered during a 60-minute period every 12 hours, 1.0 g during a 60-minute period every 12 hours, and 1.0 g during a 60-minute period every 8 hours. These schedules of PAPM/BP administration were made to investigate the outline of the drug's pharmacokinetics. When 1.0 g of PAPM/BP was intravenously administered over a period of 60 minutes every 8 hours in patient 1, an appropriate concentration of PAPM was obtained. Consequently, this administration schedule was used for all patients. Table 1 shows the PAPM/BP dose regimens.
Vascular access was obtained by inserting a double-lumen catheter (10 F, Mahurkar, Quinton Instruments, Bothell, WA, USA) into a femoral vein. An automated CRRT system (JUN600, UBE, Tokyo, Japan) was used. The same hemofilter (UT-110, Nipro, Osaka, Japan) as that used in the in vitro study was used. The dialysate and replacement fluid were Sublood-A (sodium 140, potassium 2.0, calcium 1.75, magnesium 0.5, chloride 111, bicarbonate 35, acetate 3.5, and glucose 5.51 mmol/l; Fuso, Osaka, Japan). The CRRT conditions were as follows: The blood circuit pumped a constant blood flow rate of 120 ml/min; QF was defined as 1.0 l/h in all patients; QD was infused in a countercurrent at rates of 1 l/h in addition to continuing hemofiltration. The replacement fluid was infused after dilution as clinically indicated.
Samples were collected >48 hours after the start of PAPM/BP administration and CRRT. The sampling points were referred to as prehemofilter and posthemofilter. Samples were also taken from the filtrates. Sets of three samples were taken before the start of the next PAPM/BP administration, at 1, 1.5, 2, 4, and 8 hours after the start of the drug administration. The blood samples were promptly centrifuged and plasma was separated. The samples from the plasma and filtrate were mixed with an equal amount of 1 M MOPS buffer (pH 7.0) and frozen at –80 °C until analysis.
Analysis of PAPM and BP
The concentrations of PAPM and BP in the samples were determined by the high-performance liquid chromatography (HPLC) method. Plasma and filtrate samples (100 μl) were combined with 50 μl of 20 mmol MOPS (pH 7.0) and 200 μl of MeOH in a 1.5 ml Eppendorf tube. The samples were vortexed and centrifuged at 12,100 g at 4°C for 20 minutes. Twenty microliters of the layers was injected into the HPLC system. A liquid chromatograph (L-7110, HITACHI, Tokyo, Japan) and a reversed-phase column (L-column ODS, 4.6 nm inner diameter × 150 nm) were used. PAPM and BP were detected at 296 nm and at 240 nm, respectively, by using a variable wavelength ultraviolet monitor (L-7405, HITACHI, Tokyo, Japan). Protein-bound fractions of PAPM and BP were measured by using the Centrafree Micropartition Device (Millipore, Bedford, MA, USA), with each drug at concentrations of 12.5, 25, 50, and 100 μg/ml in 5% BSA and human plasma.
A pharmacokinetic analysis was performed with the use of the nonlinear least-squares regression program (MULTI).3 The parameters were calculated by a two-compartment open model with a constant rate of infusion. The plasma concentration-time data were fitted to the following equation:
where C1 is the plasma concentration of PAPM or BP, D is the dose of the drug, K21 is the rate constant from the peripheral compartment to the central compartment, and V1 is the volume of the central compartment. The area under the plasma concentration-time curve (AUC) was calculated by using the trapezoidal rule. The clearances in vivo (CLvivo) and by CRRT (CLCRRT) were calculated as follows:
where CD+F is the concentration of the drug in dialysate and filtrate and CPre is concentration of the drug at prehemofilter. Because PAPM and BP are low molecular substances, QD and QF influence similarly to CLCRRT.4
The plasma concentration-time curves of PAPM were simulated by using the pharmacokinetics parameters that were determined in this clinical study. Based on the predictive PAPM CLtot, appropriate dose regimens of PAPM/BP were determined to maintain an effective plasma concentration level throughout the dosing intervals. The effective plasma concentration of PAPM was determined to be 4 μg/ml as a breakpoint minimum inhibitory concentration (MIC).5–7
The StatView 5.0 statistical software package (SAS Institute Inc., Cary, NC, USA) was used for all statistical calculation analyses. Predictive clearances were obtained by using a simple linear regression. Comparisons between the paired groups were made by using Student's paired t test. A p value of <0.05 was considered to be statistically significant. All data were expressed as medians (minimum minus maximum) or mean ± standard deviation.
In Vitro Study
Protein-bound fractions of PAPM in 5% BSA and human plasma were 8.4 (6.3 to 10.6)% and 7.4 (7.2 to 8.2)%, respectively. Protein-bound fractions of BP in 5% BSA and human plasma were 55 (53 to 58)% and 52 (50 to 55)%.
Figure 2 shows the relations between CLCRRT and QD+QF of PAPM and BP. The predictive PAPM and BP CLCRRT were obtained by interpolation into simple linear regression of each CLCRRT against QD+QF closely correlates with the experimental data as follows (Figure 2):
Previous reports8,9 showed a relation between CLcre and PAPM CLvivo in patients with various renal functions. Based on results of previous reports,8,9 the predictive clearance of PAPM in vivo (PAPM CLvivo) was obtained by interpolation into simple linear regression of PAPM CLvivo against CLcre closely correlates with the data of previous reports as follows (Figure 3).
The predictive total PAPM clearance (PAPM CLtot) in a patient with acute renal failure during CRRT was calculated as follows:
The plasma concentration-time curves of PAPM at 0.5 g and 1 g PAPM/BP every 12 hours in patient 1 are shown in Figure 4. The peak plasma concentrations of PAPM after the intravenous infusion of 0.5 g and 1 g PAPM/BP over a period of 1 hour were 23.2 μg/ml and 42.9 μg/ml, respectively.
The plasma concentrations of PAPM and BP at 1 g PAPM/BP every 8 hours are fitted to a two-compartment model and presented in Figure 5. After the intravenous infusion of 1 g PAPM/BP over a period of 1 hour, the peak and trough plasma concentrations of PAPM were 36.8 (25.1 to 48.8) μg/ml and 4.3 (1.8 to 5.2) μg/ml, respectively, and the peak and trough plasma concentrations of BP were 69.6 (54.0 to 85.5) μg/ml and 30.6 (16.8 to 49.8) μg/ml, respectively. Table 2 shows the pharmacokinetic parameters of PAPM and BP for each patient. The predicted total clearances of PAPM were almost equal to the actual measured total clearances.
Based on the predictive PAPM CLtot, appropriate dose regimens of PAPM/BP to maintain an effective plasma concentration level throughout the dosing intervals were determined as follows: PAPM CLtot < 80 (ml/min) 0.5 g every 12 hours or 1.0 g every 15 hours PAPM CLtot 80 to 120 (ml/min) 0.5 g every 8 hours or 1.0 g every 12 hours PAPM CLtot 120 to 160 (ml/min) 0.5 g every 6 hours or 1.0 g every 8 hours
The simulated plasma concentration-time curves of PAPM are shown in Figure 6.
Critically ill patients often undergo CRRT to treat acute renal failure and a subsequent fluid overload. Because CRRT is performed under various conditions, it is difficult to generalize the pharmacokinetics of certain drugs during CRRT. For example, although Giles et al.10 investigated the pharmacokinetics of meropenem in patients during CRRT, they indicated the clearance in their study to be different from that reported in a previous study.11 This difference was caused by dissimilarities of CRRT. The pharmacokinetics of drugs during CRRT depend on the following conditions: 1) the pore size and adsorption ability of the hemofilter membrane; 2) the molecular size and protein binding fraction of the drug; 3) QD and QF in CRRT.4,12
The hemofilter used in this study was a cellulose tri-acetate membrane with large pores and without drug absorption. In CRRT, high-flux membranes with large pores and no drug absorption are recommended. We use the hemofilter not only for the in vitro study but also in clinical settings.
The protein-bound fraction of a drug influences its CLCRRT.12 A drug having a high protein-bound fraction shows a lower CLCRRT than that of another drug having a low protein-bound fraction when the two drugs have the same molecular sizes.12 In the in vitro study to establish PAPM and BP CLCRRT, we used BSA instead of human plasma. Protein-bound fractions of PAPM and BP in BSA were the same as those in the human plasma. Based on this premise, the results of in vitro study could be applied in clinical settings. Predictive PAPM CLCRRT indicated in this study is applied to various conditions of CRRT in clinical settings because the formula was based on the various settings of the CRRT circuit model. In a patient with renal failure, the amount of PAPM elimination decreased in correlation with the patient's CLcre8,9 because PAPM is mainly eliminated by the kidneys.1 We estimated PAPM CLvivo on the basis of the CLcre (i.e., PAPM CLvivo (ml/min) = 1.2 × CLcre + 66.5). However, PAPM is not eliminated by the kidneys alone. In the predictive formula of PAPM CLvivo, a fixed number, such as 66.5, may thus indicate the nonrenal elimination rate. In a patient during CRRT, PAPM CLtot is the sum of the clearance of the patient and CRRT. In this study, we established PAPM CLtot applied to a patient with various renal functions during CRRT with various conditions.
The percentage of PAPM CLCRRT in PAPM CLtot was below 30% in the four patients (Table 2). The CRRTs were performed on conditions that the amount of QD and QF were about 33 ml/min (2 l/h) in this study. However, the amount of QD and QF does not have an upper limit. Recent studies13,14 showed that high-volume replacement CRRT improved the prognosis of patients with acute renal failure. The influences of CLCRRT on CLtot increase when the high-volume replacement CRRT is performed for a patient with severe acute renal failure.
Panipenem is a carbapenem antibiotic with a broad spectrum of activity against many common pathogens.1,2 It is often administered to critically ill patients with severe infections in intensive care units. In β-lactam antibiotics such as PAPM, the most appropriate surrogate marker for predicting the outcome is the duration in which the concentration of the drug in plasma exceeds the MIC.15 Although PAPM demonstrated a postantibiotic effect in vitro,1 the clinical significance of this effect has not yet been evaluated. Therefore, it is desirable to maintain a concentration above the MIC throughout the dosing interval. The clinical breakpoint MIC of carbapenem has been shown in several previous reports.5–7 The Japan Society of Antimicrobial Agents showed the breakpoint MIC to be 2 μg/ml in pneumonia and 1 μg/ml in sepsis.6 The National Committee for Clinical Laboratory Standards determined the breakpoint MIC to be 4 μg/ml.5 Based on these findings, in the present study the breakpoint MIC of PAPM was thus determined to be 4 μg/ml.5–7 In pharmacokinetic simulation, the doses and intervals in PAPM administration were adjusted to maintain the breakpoint MIC throughout the dosing intervals (Figure 6).
Beta Mipron, an organic anion transport inhibitor, is administered with PAPM to reduce the renal toxicity of PAPM.1 The physicochemical properties of BP include a low molecular weight (193 Da) the same as that of PAPM (339 Da) and high degree of binding to plasma proteins (73%) in contrast to PAPM (7%).1 Because of these physicochemical properties, BP is not eliminated at the same rate as PAPM by CRRT, and it accumulates in a patient with renal failure during CRRT (Figure 5). Concentration ratio of PAPM/BP significantly decreases also. However, a previous report showed that BP had neither toxicity nor any side effects.16 The accumulated concentration of BP is sufficient to reduce the renal toxicity of PAPM. We also did not observe any side effects of BP accumulation or decrease of concentration ratio of PAPM/BP during the study period when we monitored several laboratory data and systemic findings.
This pilot study has several potential limitations. First, when the predictive formula of PAPM CLvivo was obtained, we used results of previous reports8,9; however, our critically ill patients may have had more severe renal failure than that of previous reports8,9 or different organ failures. Second, the clinical CRRT study was performed under similar conditions, although the predictive formula of PAPM CLCRRT was established by many experimental data in various conditions. Third, our clinical study was a pilot study. Therefore, a larger, more precise clinical study is needed to confirm the accuracy of our predictive formulas and the results of a simulation study.
We established a predictive formula of PAPM CLtot applied to a patient with various renal functions during CRRT with various conditions. On the basis of PAPM CLtot, we arrived at the recommended doses and intervals in PAPM administration to achieve effective concentrations.
When PAPM/BP is administered during CRRT, PAPM CLtot should be calculated based on the formula PAPM CLtot (ml/min) = (1.2 CLcre + 66.5) + 0.86 (QD + QF)
Next, the appropriate dosages and intervals should be selected on the basis of the calculated PAPM CLtot: PAPM CLtot < 80 (ml/min) 0.5 g every 12 hours or 1 g every 15 hours PAPM CLtot 80 to 120 (ml/min) 0.5 g every 8 hours or 1 g every 12 hours PAPM CLtot 120 to 160 (ml/min) 0.5 g every 6 hours or 1 g every 8 hours.
1.Goa KL, Noble S: Panipenem/betamipron. Drugs
63: 913–925, 2003.
2.Shimada J, Kawahara Y: Overview of a new carbapenem, panipenem/betamipron. Drugs Exp Clin Res
20: 241–245, 1994.
3.Yamaoka K, Tanigawara Y, Nakagawa T, et al:
A pharmacokinetic analysis program (multi) for microcomputer. J Pharmacobiodyn
4: 879–885, 1981.
4.Clark WR, Ronco C: CRRT efficiency and efficacy in relation to solute size. Kidney Int Suppl
72: S3–S7, 1999.
5.NCCLS: Performance standard for microbial susceptibility testing. Wayne
6.Saito A, Inamatsu T, Okada J, et al:
Clinical breakpoints in pulmonary infections and sepsis: new antimicrobial agents and supplemental information for some agents already released. J Infect Chemother
5: 223–226, 1999.
7.Unal S, Masterton R, Goossens H: Bacteriemia in Europe: antimicrobial susceptibility data from the MYSTIC surveillance programme. Int J Antimicrob Agents
23: 155–163, 2004.
8.Aoki N, Usuda Y, Koda Y, et al:
Clinical pharmacology and efficacy of panipenem/betamipron [in Japanese]. Chemotherapy
39: 372–384, 1991.
9.Nakashima M, Uematsu T, Kanamaru M, et al:
Phase I study of panipenem/betamipron I: single dose study [in Japanese]. Chemotherapy
39: 242–264, 1991.
10.Giles LJ, Jennings AC, Thomson AH, et al:
Pharmacokinetics of meropenem in intensive care unit patients receiving continuous veno-venous hemofiltration or hemodiafiltration. Crit Care Med
28: 632–637, 2000.
11.Krueger WA, Schroeder TH, Hutchison M, et al:
Pharmacokinetics of meropenem in critically ill patients with acute renal failure treated by continuous hemodiafiltration. Antimicrob Agents Chemother
42: 2421–244, 1998.
12.Bohler J, Donauer J, Keller F: Pharmacokinetic principles during continuous renal replacement therapy: drugs and dosage. Kidney Int Suppl
72: S24–S28, 1999.
13.Ronco C, Bellomo R, Homel P, et al:
Effects of different doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: a prospective randomised trial. Lancet
356: 26–30, 2000.
14.Cole L, Bellomo R, Journois D, et al:
High-volume haemofiltration in human septic shock. Intensive Care Med
27: 978–986, 2001.
15.Hyatt JM, McKinnon PS, Zimmer GS, et al:
The importance of pharmacokinetic/pharmacodynamic surrogate markers to outcome: focus on antibacterial agents. Clin Pharmacokinet
28: 143–160, 1995.
Copyright © 2006 by the American Society for Artificial Internal Organs
16.Naganuma H, Tokiwa H, Hirouchi Y, et al:
Nephroprotective effect and its mechanism of betamipron (1): relationships of renal transport. Chemotherapy
39: 166–177, 1991.