Population Pharmacokinetic Model-Based Evaluation of Standard Dosing Regimens for Cefuroxime Used in Coronary Artery Bypass Graft Surgery with Cardiopulmonary Bypass

ABSTRACT

The purpose of this study was to investigate the population pharmacokinetics (PK) of cefuroxime in patients undergoing coronary artery bypass graft (CABG) surgery. In this observational pharmacokinetic study, multiple blood samples were collected over a 48-h interval of intravenous cefuroxime administration. The samples were analyzed by using a validated high-performance liquid chromatography (HPLC) method. Population pharmacokinetic models were developed using Monolix (version 4.4) software. Pharmacokinetic-pharmacodynamic (PD) simulations were performed to explore the ability of different dosage regimens to achieve the pharmacodynamic targets. A total of 468 blood samples from 78 patients were analyzed. The PK for cefuroxime were best described by a two-compartment model with between-subject variability on clearance, the volume of distribution of the central compartment, and the volume of distribution of the peripheral compartment. The clearance of cefuroxime was related to creatinine clearance (CL_CR). Dosing simulations showed that standard dosing regimens of 1.5 g could achieve the PK-PD target of the percentage of the time that the free concentration is maintained above the MIC during a dosing interval (fT_MIC) of 65% for an MIC of 8 mg/liter in patients with a CL_CR of 30, 60, or 90 ml/min, whereas this dosing regimen failed to achieve the PK-PD target in patients with a CL_CR of ≥125 ml/min. In conclusion, administration of standard doses of 1.5 g three times daily provided adequate antibiotic prophylaxis in patients undergoing CABG surgery. Lower doses failed to achieve the PK-PD target. Patients with high CL_CR values required either higher doses or shorter intervals of cefuroxime dosing. On the other hand, lower doses (1 g three times daily) produced adequate target attainment for patients with low CL_CR values (≤30 ml/min).

INTRODUCTION

Coronary artery bypass graft (CABG) surgery is considered a major therapeutic approach in the treatment of coronary artery disease. During this procedure, a cardiopulmonary bypass (CPB) by extracorporeal circulation can temporarily replace the function of the heart and lungs. Several complications can occur after CABG surgery, including wound infections. Patients who have undergone this major surgery are at risk for certain severe postoperative infections. It has been reported that the incidence of sternal surgical site infections after CABG surgery ranges from 0.9 to 20% (1). The incidence of deep sternal infections, such as mediastinitis, ranges from 1 to 2% (1). Furthermore, the 1-year mortality rate increases in patients with deep sternal infections, reaching approximately 22% (2). Staphylococcus aureus (44%) and coagulase-negative staphylococci (15%) are considered the most common pathogens responsible for postoperative infections (1).

There is strong evidence that using antibiotic prophylaxis significantly lowers the incidence of postoperative infections. Cephalosporins have a favorable spectrum, low toxicity, and good tissue penetration and are considered the first choice as antibiotic prophylaxis in cardiac surgery (3, 4). Cefuroxime is a second-generation cephalosporin. It possesses characteristics that make it the best choice for antibiotic prophylaxis, such as its safety, low cost, good efficacy against the most common intraoperative microorganisms, and ability to be maintained both in plasma and in tissues throughout surgery (5, 6). Cefuroxime, like other β-lactam antibiotics, is a time-dependent antibiotic. The antibacterial activity of cefuroxime is related to the percentage of the time that the free concentration is maintained above the MIC during a dosing interval (fT_MIC). For cephalosporins, data from animal models have shown that a target fT_MIC of 40% is required for bacteriostasis activity at 24 h and that an fT_MIC of 60% to 70% correlates with near-maximal bactericidal activity at 24 h (7–9). fT_MIC could be used as a pharmacokinetic (PK)-pharmacodynamic (PD) target and as a surrogate measure to predict successful microbiological or clinical outcomes for antibiotics (9, 10). On the basis of these PK-PD targets, dosing simulations can predict the probability of target attainment (PTA) at various MICs.

The PK of cefuroxime during CABG surgery with CPB have been widely investigated (5, 6, 11–15). However, cefuroxime use as a prophylactic agent has not been evaluated by coupling population pharmacokinetic modeling and Monte Carlo simulations to determine the PTA at various MICs. Therefore, the objectives of this study were to develop a population PK model of cefuroxime in patients who underwent CABG surgery and to investigate whether PK-PD targets are achieved with current dosing strategies, as well as to investigate the relevance of alternative dosing regimens and strategies.

RESULTS

Sample collection and patient demographics.

In total, 468 blood samples from 78 patients enrolled in this study were analyzed; 59 patients (76%) were males. The baseline demographics and general clinical characteristics of the patients used for model building are shown in Table 1. The average ± standard deviation (SD) age and weight in the population were 54.2 ± 13.2 years and 76.7 ± 14.7 kg, respectively. The average ± SD creatinine clearance (CL_CR) estimated by the Cockcroft-Gault formula was 78.5 ± 24 ml/min.

TABLE 1.

Summary of patient characteristics

Characteristic	Mean (SD)	Range
Age	54.2 (13.2)	18–80
Wt (kg)	76.7 (14.7)	41–111.8
BMIa	28.6 (5.2)	16.6–43.2
Serum creatinine concn (μmol/liter)	85 (29.7)	41–245
Albumin concn (g/liter)	35.6 (4.4)	24–44
CLCRb (ml/min)	78.5 (24)	28.3–125
Sex (% male/%female)	76/24

^aBMI, body mass index.

^bCL_CR was estimated by the Cockcroft-Gault formula.

Population pharmacokinetics.

Visual inspection of the cefuroxime raw data supported a two-compartment model, which was confirmed by the goodness-of-fit (GOF) plots and a statistically significant drop in the objective function value (OFV) with respect to that of the one-compartment model. The pharmacokinetic model was parameterized in terms of clearance (CL), the volume of distribution the central compartment (V₁), the volume of distribution of the peripheral compartment (V₂), and the intercompartmental clearance (Q). A combined-error model was the most accurate for residual and interpatient variability. After covariate testing, CL_CR was the only covariate showing a correlation with cefuroxime CL. The inclusion of CL_CR statistically improved the base model by decreasing the OFV by 19 points and decreasing the between-subject variability of CL by approximately 23%. Therefore, CL_CR was included in the model, whereas other covariates, which displayed no correlation with the pharmacokinetic parameters, were not investigated further. The values of the parameters for the final models are summarized in Table 2.

TABLE 2.

Population PK model estimates for cefuroxime following intravenous infusion^a

Parameter	Base model		Final model
	Estimate	RSE (%)	Estimate	RSE (%)	Shrinkage (%)
CLb (liters/h)	2.23	17	3.43	16
V1 (liters)	2.99	19	3.88	12
Q (liters/h)	16.9	30	22.2	31
V2 (liters)	6.45	8	5.7	6
IIV for CL (%)	58.3	12	45.2	8	ηsh = 6
IIV for V1 (%)	69.5	15	61.3	12	ηsh = 8
IIV for Q (%)	57.3	11	56.2	11	ηsh = 2
IIV for V2 (%)	25.3	24	21.8	19	ηsh = 5
Residual errors
a	6.17	13	6.45	12
b	0.093	10	0.092	10	εsh = 14

^aIIV, interindividual variability, expressed as the coefficient of variation; RSE, relative standard error; η_sh, shrinkage value for a parameter; ε_sh, shrinkage value for the residual error.

^bCL = 3.43 · (CL_CR/78.5)^0.56.

Model evaluation.

Figure 1 shows the GOF plots for the cefuroxime final covariate model. The RSE values (in percent) shown in Table 2 reveal that all parameters were precisely estimated. Additionally, visual inspection of the visual predictive check (VPC) in Fig. 2 reveals a good correlation between the percentile intervals obtained by simulation in the final model and the observed data. Both Fig. 1 and 2 show that the final pharmacokinetic model describes the measured concentrations adequately. This model was used to simulate all subsequent dosing scenarios.

FIG 1.

Goodness-of-fit plots obtained from the final model for cefuroxime. (Right) Individual predictions of cefuroxime concentrations versus observed concentrations; (left) population predictions of cefuroxime concentrations versus observed concentrations.

FIG 2.

Visual predictive check (VPC) for cefuroxime concentration versus time based on 1,000 Monte Carlo simulations. The solid green lines represent the 10th, 50th, and 90th percentiles of the observed data. The shaded regions represent the 90% confidence intervals around the 10th, 50th, and 90th percentiles of the simulated data. The blue circles are observed concentrations. emp prctile, empirical percentile; prctile out., outlier dots; P.I out., outlier area; P.I, percentile interval.

Monte Carlo simulations.

We used Monte Carlo simulations to evaluate different dosing regimens of cefuroxime. The probability of target attainment for different dosing regimens of cefuroxime is presented in Fig. 3. The standard dose of 1.5 g cefuroxime three times daily resulted in adequate target attainment in patients with CL_CRs of 30, 60, and 90 ml/min up to an MIC of 8 mg/liter. This dosing regimen led to a more than 90% probability of target attainment for an MIC of 8 mg/liter, whereas this standard regimen resulted in inadequate target attainment for patients with CL_CRs of ≥125 ml/min. Administration of 1.5 g two times daily produced adequate target attainment in patients with CL_CRs of 30 and 60 ml/min but was inadequate in patients with CL_CRs of ≥90 ml/min. Administration of 1 g cefuroxime two times daily produced adequate target attainment only in patients with low CL_CRs (<30 ml/min). A dose of 1 g cefuroxime three times daily was not enough to produce adequate target attainment in all patients, with the PTA being <90%.

FIG 3.

Probability of target attainment analysis for cefuroxime (fT_MIC, 65%) for various dosing regimens for CL_CRs of 30, 60, 90, and 125 ml/min at different MIC values.

DISCUSSION

During CABG surgery, contamination of the surgical site can occur at any time during the operation period. The risk is highest during the intraoperative period, where there is exposure of tissue and entry of microorganisms into the surgical site is possible. There are several factors that could increase the risk of infection. These may be endogenous, such as advanced age, obesity, and diabetes, or exogenous, such as breaking of the aseptic barrier, inadequate hand hygiene, wound classification, the duration of the procedure, and surgical technique (16–18). Surgical site infections (SSIs) following CABG surgery, particularly from serious infections, such as deep sternal infections, have implications resulting in significant increases in both morbidity and mortality and placement of substantial burdens on both patients and health care systems. It has been reported that the total length of hospitalization for patients with SSIs after CABG surgery is significantly longer than that for patients without SSIs (19). A number of infection control strategies can effectively be applied to reduce the risk of SSIs, including appropriate antibiotic prophylaxis, effective patient skin preparation, and surveillance systems to monitor and report SSIs after specific procedures with the provision of feedback of appropriate data to surgeons and hospitals (20).

The clinical situation for patients undergoing CABG surgery is compromised, and the risk of developing infection is high, with the health of the patients being further compromised if infection does occur. Thus, establishment of the best regimen in antibiotic prophylaxis is very important, although it is not an easy task due to the heterogeneity of patients and uncertainty regarding the infectious pathogens.

According to Society of Thoracic Surgeons practice guidelines, a β-lactam antibiotic is indicated as the single antibiotic of choice for standard cardiac surgical prophylaxis in populations that do not have a high incidence of methicillin-resistant Staphylococcus aureus (MRSA), and the duration of routine postoperative administration of prophylactic antibiotics should be no longer than 48 h (3, 21). Cefuroxime is considered the first choice as antibiotic prophylaxis in cardiac surgery because of its favorable pharmacokinetics and safety profile.

The population PK of cefuroxime as a prophylactic antibiotic in cardiac surgery have not been evaluated previously, and studies involving population modeling for this drug are limited (22–25). In this study, we evaluated the PK profile of cefuroxime in patients undergoing CABG surgery using population modeling and applied the proposed models to Monte Carlo simulations to predict how patient characteristics influence the adequacy of the prophylaxis.

The plasma concentration-time curve for cefuroxime was fitted by a two-compartment model, as shown in other studies (22–25). Cefuroxime is eliminated mainly by renal excretion, and we found that its CL was related to CL_CR, but with substantial interindividual variability, which was consistent with findings observed in other studies (22–25). Our study showed that only CL_CR could be included in the final covariate model, as none of the other patient characteristics significantly influenced the model, and, therefore, they could not be included. The estimated values of the pharmacokinetic parameters of cefuroxime were 3.43 liters/h, 3.88 liters, 22.2 liters/h, and 5.7 liters for CL, V₁, Q, and V₂, respectively. Our values are slightly lower than those reported in other studies, most likely because of the differences between study populations.

Previous studies have reported that the use of CPB during CABG may have a significant effect on the PK of administered drugs, including the cephalosporins, due to, for example, hemodilution, hypotension, hypothermia, and alterations of protein binding, which could lead to suboptimal antibiotic prophylaxis (12, 13, 26). Aalbers et al. (5) evaluated the effect of CPB on unbound cefuroxime plasma concentrations during CABG. They reported that, during CPB, cefuroxime clearance decreased an average of 11% compared to the clearance before CPB, the volume of distribution remained unchanged, and no hemodilution effect was observed (5). They concluded that CPB during CABG surgery is not a risk factor for inadequate unbound cefuroxime concentrations and does not require dose modifications. However, risk factors, such as renal function, age, and weight, could be the reason for insufficient cefuroxime concentrations (5).

When simulations were performed for different dosing regimens, we demonstrated that the standard regimen of 1.5 g cefuroxime ensured a 90% PTA (plasma free concentrations are greater than the MIC for at least 65% of the dosing interval) for MICs up to 8 mg/liter in all patients with different CL_CRs except in patients with high CL_CR values (≥125 ml/min). However, the use of 1.5 g two times daily and 1 g two times or three times daily failed to achieve a PTA above 90% for an MIC of 8 mg/liter in most of the simulated scenarios. These findings show that the standard regimen of 1.5 g cefuroxime is not enough for patients with high CL_CRs (≥125 ml/min) and that these patients require either higher doses or shorter intervals of cefuroxime dosing. On the other hand, lower doses (1 g three times daily) produced adequate target attainment for patients with low CL_CRs (≤30 ml/min). It is already known that the MIC for most suspect bacteria is ≤8 mg/liter (27). According to the European Committee on Antimicrobial Susceptibility Testing (EUCAST), the cefuroxime breakpoints for Enterobacteriaceae (Escherichia coli, Klebsiella spp., and Proteus mirabilis), Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis are 8, 0.5, 1, and 4 mg/liter, respectively (27).

Regarding the PK-PD analysis, the PK-PD target that best predicts the activity of cefuroxime is the fT_MIC (9, 10). However, the use of antibiotics in prophylaxis settings supposes a completely new concept, where the risk of infection and life-threatening consequences of deep sternal infections is higher. Therefore, the goal of antibiotic prophylaxis is to achieve unbound plasma and tissue concentrations of cefuroxime that exceed the MICs for the organisms likely to be encountered during the operation, until the time that the surgical wound is closed (28). Consequently, in this study, we used the highest MIC with a PTA of at least 90% as the PK-PD MIC breakpoint.

In this study, we did not measure free drug concentrations or the concentrations at the site of infection. Instead, we measured the total cefuroxime concentrations, and we calculated the amount of serum-unbound drug by scaling the total serum concentration by the literature value for the unbound fraction (29, 30). Previous studies have revealed that the use of this approach is acceptable for drugs with low to moderate levels of protein binding, such as cefuroxime, whereas it is not accurate for drugs with a high protein binding ability (31, 32).

In summary, a population pharmacokinetic model has been developed for cefuroxime used as a prophylactic agent in patients undergoing CABG surgery. The PK of cefuroxime were best described by a two-compartment model, and a relationship between CL and CL_CR was detected. This model is useful to establish dosing recommendations to reduce the incidence of SSIs depending on patient characteristics. The results of the dosing simulations show that dosing regimens of 1 g of cefuroxime administered two or three times daily as a bolus infusion fail to achieve the PK-PD target, whereas dosing regimens of 1.5 g of cefuroxime administered three times daily are more likely to achieve the PK-PD targets in most of the patients but not in patients with CL_CRs of ≥125 ml/min.

MATERIALS AND METHODS

Study design and settings.

A prospective, open-label study was conducted in patients who underwent cardiac surgical procedures at King Fahad Cardiac Surgery, King Saud University Medical City (KSUMC; Riyadh, Saudi Arabia). All study procedures were approved by the Institutional Review Board (IRB) at the hospital and were conducted in accordance with good clinical practice. Written informed consent was obtained from all patients or their legally authorized representatives. Patients aged ≥18 years scheduled to undergo cardiac surgical procedures were included in the study. Creatinine clearance (CL_CR) was estimated for each patient using the Cockcroft-Gault equation (33). Patients were excluded if they were allergic to β-lactam antibiotics, were diagnosed with previous systemic infections, or were given antibiotic therapy in the last 72 h before the surgery.

Drug administration and sampling procedure.

According to the surgical protocol of the hospital, patients received 1.5 g of cefuroxime as an intravenous infusion 30 min to 1 h before skin incision. An extra dose (redosing) was administered if the surgery lasted for more than 4 h by mixing cefuroxime at 1.5 g with the CPB solution. For 48 h after the surgery, the subsequent doses were either 1.5 g of cefuroxime every 12 h or 1 g every 8 h. Blood samples were collected from the radial artery catheter. Six blood samples were collected, as follows: (i) immediately before the skin incision, (ii) at the start of CPB, (iii) 1 h after the start of CPB, (iv) immediately before skin closure, (v) 24 h after administering the first dose, and (vi) 48 h after administering the first dose. Blood samples were collected in EDTA-containing blood collection tubes, stored on ice, mixed, and centrifuged at 5,000 rpm for 10 min directly after the procedure. The resulting plasma was stored at −20°C until analysis, which was performed during the same week.

Analytical method.

Cefuroxime concentrations in plasma were determined as described previously (11). The method is fully validated and was achieved using an isocratic Prominence Shimadzu high-performance liquid chromatography (HPLC) system (Columbia, MD). The system consisted of a SIL-20AHT autosampler and an LC-20AB pump connected to a Dgu-20A3 degasser. Data acquisition was achieved by using LC Solution software (version 1.22 SP1). Following deproteinization of plasma samples (containing cefazolin as an internal standard) by methanol precipitation, the supernatant was diluted with the mobile phase, consisting of potassium buffer-acetonitrile (85:15, vol/vol). Separation was performed on a Phenomenex Luna C₁₈ column (250 by 4.6 mm [inside diameter]; particle size, 5 μm; Phenomenex, CA). For cefuroxime detection, the Shimadzu UV SPD-20A (Shimadzu, OR) detector was set at 275 nm. All samples were analyzed in duplicate. Under these chromatographic conditions, the total run time was 10 min, with a retention time of 5.6 min for the internal standard and 8 min for cefuroxime. The assay was linear over the concentration range of 0.5 to 200 μg/ml. The intraday and interday coefficients of variation (CV) ranged from 0.81% to 8.33%, and bias ranged from 0.53% to 13.23%.

Population pharmacokinetics.

The population PK model for cefuroxime was developed using Monolix (version 4.4) software. Monolix estimates PK parameters using the stochastic approximation expectation maximization (SAEM) algorithm (34). Initially, we developed the base structural model for cefuroxime. One- and two-compartment systems were evaluated with linear or nonlinear elimination. Pharmacokinetic parameters were assumed to follow a lognormal distribution. For the residual variability, the following error models were tested: constant, proportional, and combined error models. Selection between models was based on the following: (i) the decrease in the minimum of the objective function value (log-likelihood value); (ii) the precision of the parameter estimation, expressed as the relative standard error (RSE [in percent]) and calculated as the ratio between the standard error and the final parameter estimate; (ii) physiological plausibility; and (iv) goodness-of-fit (GOF) plots that included the observed versus predicted concentration, residuals plot, and the visual predictive check (VPC).

After the appropriate base model was established, eight covariates were tested, specifically, age, weight, serum creatinine concentration, CL_CR, gender, height, albumin concentration, and body mass index. For covariate testing, we started by plotting the individual pharmacokinetic parameters versus covariates to screen for potentially significant correlations. Then, we performed a stepwise regression analysis to test the significant covariates identified in step 1 using the log-likelihood ratio test. If a trend between a covariate and PK parameter was found, then it was considered for inclusion in the base model.

Model evaluation.

GOF plots were used as the first indicator of suitability, including the representation of model-based individual predictions (IPRED) and population predictions (PRED) versus the observed concentrations. A VPC was constructed to study the performance of the final model. The VPC was constructed with the 10th, 50th, and 90th percentiles of the observed data.

Monte Carlo simulations.

The final population model was utilized to simulate different dosing regimens for cefuroxime at different CL_CR values. The simulated dosage regimens are presented in Table 3. The CL_CR values simulated were 30, 60, 90, and 125 ml/min. Each Monte Carlo simulation created the time-concentration profile for 1,000 subjects per dosing regimen using the parameters from the final covariate model.

TABLE 3.

Simulated dosage scenarios^a

Dosage regimen scenario	First dose (g)	Subsequent doses
1	1.5	1.5 g every 8 h
2	1.5	1.5 g every 12 h
3	1	1 g every 8 h
4	1	1 g every 12 h

^aThe infusion time was 0.5 h in all scenarios.

Previous studies have shown that a target fT_MIC of 60 to 70% is the near-maximal bactericidal activity for cephalosporins (8, 9). Therefore, the probability of target attainment was obtained by counting the subjects who achieved an fT_MIC of 65%. It has been reported that the protein binding levels for cefuroxime range from 30 to 50% (29, 30); thus, in this study, we assumed an average protein binding of 40% for cefuroxime. Target attainment of 90% was evaluated at different MIC values of 0.5, 1, 2, 4, 8, 16, and 32 mg/liter. All simulations and graphical representations were performed using R statistical software.