Abstract
Objective
The aim of this study was to measure and evaluate the appropriateness of the actual concentrations of serum and dialysate ceftazidime in Thai continuous ambulatory peritoneal dialysis (CAPD) patients.
Design
Prospective and descriptive study of patients treated following the International Society for Peritoneal Dialysis (ISPD) 2000 recommendation for the empiric therapy of CAPD-related peritonitis.
Setting
Institutional level of clinical care.
Patients
CAPD-related peritonitis patients were diagnosed by dialysate effluent white blood cell count of more than 100/mm3 and polymorphonuclear leukocytes of at least 50%. There were 10 patients, all at least 18 years of age, entered; all completed the study.
Intervention
In accordance with the ISPD 2000 recommendations, the antibiotic regimen comprised continuous intraperitoneal (IP) cefazolin and once-daily IP ceftazidime. Cefazolin was administered as loading and continuous maintenance doses of 500 and 125 mg/L dialysate respectively. Ceftazidime (20 mg/kg body weight) was given IP once daily. Duration of treatment was 96 hours.
Main Outcome Measures
Serum and dialysate effluent samples of the 10 CAPD patients with peritonitis were measured for ceftazidime levels, which were used for the development of pharmacokinetic equations that could predict drug concentrations at any treatment time.
Results
Following ceftazidime administration as in the ISPD 2000 recommendation, serum ceftazidime levels were above 8 μg/mL, the minimum inhibitory concentration (MIC) recommended by NCCLS, throughout 24 hours. Dialysate ceftazidime levels were below the MIC for total periods of 4.19 and 6.26 hours in day 1 and day 4 respectively. The clinical response rate to the empiric regimen was 90%.
Conclusions
Once-daily IP administration of ceftazidime according to the ISPD 2000 recommendation could not provide adequately therapeutic levels of ceftazidime in dialysate throughout 24 hours. Despite this finding and the poor post-antibiotic property of ceftazidime, the empiric regimen including once-daily IP ceftazidime could yield good clinical outcome.
In the interest of preserving residual renal function, in 2000 the International Society for Peritoneal Dialysis (ISPD) changed the ISPD 1996 regimen of empiric therapy in chronic ambulatory peritoneal dialysis (CAPD)-related peritonitis by replacing aminoglycoside with ceftazidime in patients with a urine output of more than 100 mL/day (1,2).
Ceftazidime is in the beta-lactam group of antibiotics, the efficacy of which depends on the drug level being above the minimum inhibitory concentration (MIC). This necessity is further underscored by the fact that ceftazidime has poor post-antibiotic effects (3). Thus, according to the NCCLS 2000, the suggested MIC range of ceftazidime is >8 – 32 μg/mL (4). The advantage of once-daily intraperitoneal (IP) administration of ceftazidime is that reduction in the risk of contamination of the CAPD system, caused by continuous dosing, could be lessened by the decrease in ceftazidime levels during the non-drug–administration cycles.
Heretofore there have been no available pharmacokinetic data for ceftazidime in CAPD patients with peritonitis. Previous pharmacokinetic studies of ceftazidime were conducted in CAPD patients without peritonitis (5-7), the membrane transport characteristic of whom would differ from patients with infection. Whether once-daily IP administration of ceftazidime provides consistently therapeutic drug levels in dialysate effluent throughout 24 hours remains unknown.
The present study was carried out to determine the pharmacokinetics of once-daily IP ceftazidime treatment in patients with CAPD-associated peritonitis.
Patients and Methods
Patients
Pharmacokinetic studies of once-daily IP ceftazidime were performed prospectively in 10 Thai CAPD patients with peritonitis, treated at King Chulalongkorn Memorial Hospital from December 2000 to November 2001. Prospective approval was obtained from the Ethics Committee on Clinical Research, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand. Informed consent was acquired from each patient participating in the study. Patients with CAPD-related peritonitis were included if their dialysate effluent had a white blood cell count of more than 100 cells/mm3 and the percentage of polymorphonuclear leukocytes was at least 50% of all white blood cells. All patients were older than 18 years and had been on a stable regimen of CAPD for at least 3 months. They were dialyzed with 2 L 1.5% dextrose dialysate fluid every 6 hours. Patients were excluded if they had received previous antibiotic treatment within 1 week, had exit-site or tunnel infection, had peritonitis due to other known causes, currently had systemic infection needing other routes of antibiotic administration, or were allergic to cephalosporin.
Methods
Patients fulfilling the study criteria were admitted to the hospital. The empiric therapy was prescribed according to the ISPD 2000 recommendation. The doses of cefazolin administered were 20 mg/kg in the first bag and 125 mg/L in the remaining three bags. Ceftazidime at a dose of 20 mg/kg (1 g in the present study) was added only in the first bag. Pharmacokinetic studies of cefazolin, prescribed following the ISPD suggestion, were performed recently in CAPD patients with peritonitis (8). Thus, in the present work, only pharmacokinetic data of ceftazidime were examined. Prior to the first dialysis bag containing antibiotics, spent dialysate effluent from the preceding exchange was immediately drained and collected for Gram stain and culture using the BACTEC system (Bennex, Shannon, Ireland). Also, pre-study serum samples were drawn for complete blood count and chemical analysis. Urine was collected over 24 hours on the first day of the study.
On days 1 and 4 of the empiric therapy, for pharmacokinetic studies, 5-mL serum and 5-mL dialysate samples were collected at 1, 2, 3, 6, 7, 9, 12, 13, 15, 18, 19, 21, and 24 hours after complete instillation of the first dialysis bag containing ceftazidime into the peritoneal cavity. All serum and dialysate samples were stored at –70°C within 1 hour after collection. High performance liquid chromatography (HPLC) was used to measure ceftazidime concentrations in blood and dialysate samples. Briefly, samples were prepared by protein precipitation with perchloric acid. An aliquot of the supernatant (20 μL) was injected into the HPLC system including the μ Bondapak C18 3.9 × 300 mm column (Waters, Milford, Massachusetts, USA). The mobile phase of this system was a combination of KH2PO4 and methanol in a ratio of 8:12. The flow rate was 1 mm/minute. The column effluent was monitored with an UV detector set at 254 nm. Cefadroxil was used for the internal standard. Linear calibration curves were obtained for ceftazidime over the concentration range of 0 – 500 μg/dL. The intra- and interday coefficients of variation at low, medium, and high concentrations were less than 5%.
When the culture and drug sensitivity results became available, the antibiotics were changed when the pathogens were resistant to both antibiotics.
Pharmacokinetic Analysis
Serum and dialysate effluent ceftazidime concentrations at each period were averaged. The mean values of each time point were used in calculating pharmacokinetic equations.
Pharmacokinetic Equations for Serum Ceftazidime Concentration
In accordance with a pharmacokinetic model of bidirectional transfer and intermittent IP drug administration, a pharmacokinetic equation [Eq. (1)] was used to calculate the mean serum ceftazidime concentration at any time during the treatment period:
Pharmacokinetic Equations for Dialysate Effluent Ceftazidime Concentration
Equations (2), (3), (4), (5) represent pharmacokinetic equations for dialysate effluent ceftazidime concentration:
These pharmacokinetic equations were used to determine mean dialysate effluent ceftazidime level at any time during the study period. Using the RSTRIP II program, version 2.02 (1993) (Micromath Scientific Software, Salt Lake City, Utah, USA), the best-fit equations and equation constant parameters were calculated. Pharmacokinetic equations from the RSTRIP II program can predict ceftazidime concentrations in serum and dialysate effluent at any time during the study day.
All the data presented in this study are expressed as mean ± SD of 10 patients.
Results
Basic Patient Characteristics
Ten CAPD patients (5 males, 5 females; mean age 55.1 ± 14.7 years; mean body weight 53.5 ± 2.7 kg, range 44 – 61 kg) with peritonitis were recruited for the study. The causes of end-stage renal disease were diabetic nephropathy (50%), hypertension (40%), and unknown (10%). The mean duration of peritoneal dialysis treatment was 9.5 ± 2.5 months. Peritoneal equilibration tests showed “low average” results in all patients. Basic serum biochemistry data consisted of blood urea nitrogen 37.1 ± 13.8 mg/dL, creatinine 7.4 ± 2.7 mg/dL, and albumin 3.3 ± 0.6 g/dL. Hemoglobin was 10.4 ± 1.8 g/dL. All patients were nonanuric; mean 24-hour urine volume was 802 ± 198 mL. Mean creatinine clearance, determined by 24-hour urine collection, was 3.1 ± 1.2 mL/minute. No statistically significant differences were observed in parametric data between males and females. In all cases, there was no simultaneously active exit-site infection.
Nine of 10 cases responded excellently, with dialysate effluent white blood cell count less than 10 cells/mm3, to the 4-day period of empiric treatment. The causative organisms of peritonitis were found to be culture-negative (50%), gram-negative bacteria (40%), and gram-positive bacteria (10%). Gram-positive bacteria were Staphylococcus aureus. The gram-negative pathogens comprised Pseudomonas species (50%), Enterobacter species (25%), and Alcaligenes species (25%). In the culture-positive patients, the pathogens were responsive to the empiric regimen in 80% of cases. Two of 3 patients with a ceftazidime-sensitive pathogen had a good early response but relapse infection occurred later.
Pharmacokinetic Data of Serum Ceftazidime Concentration
From Eq. (1) and analysis of mean serum ceftazidime concentration data by the RSTRIP II program, the results from the first day were A1 = 46.86 μg/mL, A2 = 39.71 μg/mL, β = 0.052/hour, and ka = 0.55/hour; on the fourth day, A1 = 46.36 μg/mL, A2 = 26.97 μg/mL, β = 0.044/hour, and ka = 0.43/hour.
Figure 1 represents the mean serum ceftazidime concentrations obtained by direct measurement on the first and the fourth days of treatment. The estimated graph in the figure was obtained from the above Eq. (1) -derived values. Serum ceftazidime reached levels higher than the NCCLS-recommended MIC level (8 μg/mL) at 2.7 minutes after drug administration.
Mean serum ceftazidime concentrations by direct measurement (circles) and estimation by equation (line) during 4 days of treatment. MIC = minimum inhibitory concentration.
All serum ceftazidime levels during the second and third days of treatment were higher than the MIC level throughout the 24 hours (Figure 1).
Pharmacokinetic Data of Dialysate Ceftazidime Concentrations
The results from the first study day were C1 = 477.9 μg/mL, C2 = 24.7 μg/mL, C3 = 19.8 μg/mL, C4 = 14.2 μg/mL, β1 = 0.42/hour, β2 = 0.51/hour, β3 = 0.41/hour, and β4 = 0.38/hour; on the fourth study day, C1 = 469.3 μg/mL, C2 = 21.2 μg/mL, C3 = 16.0 μg/mL, C4 = 11.0 μg/mL, β1 = 0.37/hour, β2 = 0.65/hour, β3 = 0.4/hour, and β4 = 0.33/hour.
The mean dialysate effluent ceftazidime concentrations during the 4 days of treatment are illustrated in Figure 2. The graph lines in the figure were created using the concentrations obtained from direct measurement and also by exploiting the values derived from pharmacokinetic Eqs. (2) to (5).
Mean dialysate effluent ceftazidime concentrations by direct measurement (circles) and estimation by equation (lines) during 4 days of treatment. Shaded areas represent the duration during which ceftazidime levels were below the minimum inhibitory concentration (MIC).
It was seen that dialysate ceftazidime levels were below the MIC level for a period during each of the second, third, and fourth cycles, during which ceftazidime was not administered. These periods of time were 0.76, 1.27, and 2.16 hours, respectively, on the first day, and 0.72, 1.69, and 5.87 hours, respectively, on the fourth day. The total periods were 4.19 (17%) and 6.26 (26%) hours on the first and fourth days respectively. Such findings were also observed on the second and third days (Figure 2).
Discussion
The results of the present study involving CAPD-related peritonitis show that, following the once-daily IP administration of ceftazidime suggested by the ISPD 2000 recommendation, (1) serum ceftazidime levels were above 8 μg/mL, the MIC level recommended by the NCCLS, throughout 24 hours; (2) dialysate ceftazidime concentrations were below the MIC level for 4.19 and 6.29 hours on the first and fourth days respectively; and (3) the response rate to the empiric regimen was 90%.
Despite reduced incidences, peritonitis remains a major cause of hospitalization, technical failure (9-14), and mortality in CAPD patients (10,15-18). Since CAPD-associated peritonitis is a localized infection, the ceftazidime level in dialysate, instead of serum, is essential in eradicating organisms causing peritonitis. By the rule of bidirectional transfer (19), the existence of dialysate ceftazidime during the three remaining non-administration cycles was the result of diffusion of the drug in the serum. It is contended, although not been proved, that peritonitis might enhance peritoneal membrane transport of the drug from the peritoneal cavity to the serum, causing increased serum drug levels, which in turn could lead to higher dialysate drug levels.
All previous pharmacokinetic studies involving intermittent IP administration of ceftazidime were carried out in non-peritonitis CAPD patients (5-7). Furthermore, several of these works were performed by measuring dialysate drug levels only in the administration cycle and at the end of the day. As such, pharmacokinetic data from these reports could not ensure that dialysate ceftazidime levels are adequate throughout the day.
The underlying causes of the decreased dialysate ceftazidime levels observed in the present study remain unclear. As stated earlier, during non-ceftazidime–administered cycles, the presence of the drug in the dialysate is the result of back diffusion of free drug in the serum. Adequacy of free drug in the serum depends on the protein-binding property of the drug and on renal clearance. With respect to the former factor, in the present study, serum ceftazidime levels were several times higher than the MIC throughout 24 hours (Figure 1). Furthermore, the percentage of protein binding of ceftazidime in the serum is only 10% – 17%, which is not sufficient to pre-fix and prevent ceftazidime from diffusing through the peritoneal membrane (3). With respect to elimination of ceftazidime via the kidney, a recent study has revealed that renal clearance of ceftazidime approaches creatinine clearance (5). In the present study, the mean creatinine clearance of the patients was only 3.1 mL/minute.
Taken together, the inadequate dialysate ceftazidime level observed in the current work is not mediated via inadequacy of free ceftazidime level in the serum. Despite the unclear etiology of the insufficient dialysate level, it is obvious that, following administration of a single dose of ceftazidime, there must be an interval during which dialysate ceftazidime levels are below MIC in the non-ceftazidime–administration cycles (Figure 2).
It should be noted that the above circumstance, observed in once-daily IP ceftazidime treatment, is not evident in administration by continuous dosing. To support this view, a recent pharmacokinetic study of continuous IP administration of cefazolin — the drug used with ceftazidime in the ISPD 2000 recommendation — revealed that continuous dosing could provide dialysate cefazolin levels several-fold above the MIC throughout the dwell period (8). Responsiveness to an antibiotic, however, does not depend only on dialysate drug levels. Indeed, recent studies in animal infection models, although not in CAPD-related peritonitis, have shown that concentrations of beta-lactam antibiotics do not need to exceed the MIC for 100% of the dosing interval to have a significant antibacterial effect (20-24). Moreover, clinical studies of several infection types and sites have demonstrated that a beta-lactam concentration higher than the MIC for only 60% – 70% of the time correlated with clinical effectiveness (20). Furthermore, tissue drug levels may be more important than serum or dialysate effluent levels in predicting treatment outcome. There are, however, scarce data regarding such circumstance and this needs further study.
Due to a high percentage of culture-negative cases and, thus, unknown drug sensitivity results for the pathogens, the results of the current pharmacokinetic study cannot be compared with clinical outcome. Despite a small number of cases, the high occurrence of relapse infection (2 of 3 cases) observed in patients with a ceftazidime-sensitive organism underscores the importance of dialysate ceftazidime levels. A longer period below the MIC in subsequent days [6.26 hours (26%) on the fourth day compared with 4.19 hours (17%) on the first day] might increase the opportunity for relapse infection. Reducing inflammation during the early period might reduce peritoneal permeability, leading to slower diffusion of ceftazidime from the blood.
In conclusion, in CAPD-related peritonitis, once-daily IP administration of ceftazidime as recommended by ISPD 2000 did not sustain and could not provide sufficient dialysate ceftazidime levels (above the MIC level of 8 μg/mL suggested by NCCLS) throughout 24 hours. Despite this finding and poor post-antibiotic properties, the empiric regimen containing once-daily IP ceftazidime recommended by ISPD 2000 could provide good clinical outcome. A large-scale study is needed to delineate the correlation between pharmacokinetic data and clinical outcome.
Footnotes
Acknowledgment
A portion of this work was presented at the annual meeting of The American Society of Nephrology, October 30 – November 4, 2002, Philadelphia, Pennsylvania, USA.