Pharmacokinetic/pharmacodynamic predictions and clinical outcomes of patients with augmented renal clearance and Pseudomonas aeruginosa bacteremia and/or pneumonia treated with extended infusion cefepime versus extended infusion piperacillin/tazobactam.

AIM: We sought to correlate pharmacokinetic (PK)/pharmacodynamic (PD) predictions of antibacterial efficacy and clinical outcomes in patients with augmented renal clearance (ARC) and Pseudomonas aeruginosa bacteremia or pneumonia treated with extended infusion cefepime or piperacillin/tazobactam.
MATERIALS AND METHODS: Cefepime (2 g every 8 h) and piperacillin/tazobactam (4.5 g every 8 h) were administered over 4 h after a loading dose infused over 30 min, and minimum inhibitory concentration was determined by E-test. Published population PK evaluations in critically ill patients were used, and PD analyses were conducted using estimated patient-specific PK parameters and known minimum inhibitory concentration values for P. aeruginosa. Concentration-time profiles were generated every 6 min using first-dose drug exposure estimates including a loading infusion, and free concentration above the minimum inhibitory concentration (fT> MIC) was estimated. Clinical cure was defined as resolution of signs and symptoms attributable to P. aeruginosa infection without need for escalation of antimicrobial.
RESULTS: One hundred and two patients were included (36 cefepime and 66 piperacillin/tazobactam). The two groups of patients had similar age, serum creatinine, weight, and creatinine clearance. The majority of patients required intensive care unit care (63.9% vs. 63.6%) and most had pneumonia (61%). The fT>MIC (93.6 [69.9-100] vs. 57.2 [47.6-72.4], P < 0.001) and clinical cure (91.7% vs. 74.2%, P = 0.039) were significantly higher in cefepime group, whereas mortality (8.3% vs. 22.7%, P = 0.1) and infection-related mortality (0% vs. 2%, P = 0.54) were similar.
CONCLUSIONS: Patients with ARC and P. aeruginosa pneumonia and/or bacteremia who received extended-infusion cefepime achieved higher fT>MIC and clinical cure than those receiving extended infusion piperacillin/tazobactam. Copyright:

INTRODUCTION

Antibiotic resistance is perhaps the greatest current threat to public health worldwide.[1] The increasing prevalence of resistant Gram-negative bacteria is particularly worrisome as they are responsible for significant morbidity and mortality and have limited effective treatment options.[2] Among these difficult-to-treat Gram-negative pathogens, Pseudomonas aeruginosa is considered a serious worldwide threat by the Centers for Disease Control and Prevention.[345] The estimated mortality associated with P. aeruginosa infections ranges from 18% to 60% with an estimated cost of $20,000–80,000/infection.[3678910] P. aeruginosa often possesses multiple resistance mechanisms leading to elevated minimum inhibitory concentrations (MICs), further limiting the already scarce effective treatment options.[111213]

In addition to pathogen-specific issues, patient-related factors can alter antibiotic pharmacokinetics (PKs) and pharmacodynamics (PDs), affecting clinical outcomes. Among these factors, augmented renal clearance (ARC) has been shown to lead to subtherapeutic β-lactam concentrations, suboptimal achievement of desired PK/PD targets (%free concentration [fT>MIC]), and worse clinical outcomes in patients with infections due to a variety of Gram-negative pathogens.[1415161718] Employing extended infusions of antipseudomonal β-lactams such as cefepime and piperacillin/tazobactam may help optimize their PK/PD and improve clinical outcomes in patients with ARC and P. aeruginosa infections. Although there are limited PK/PD data with extended infusion cefepime, use of extended infusion piperacillin/tazobactam is associated with reduced attainment of PK/PD target and decreased clinical cure in patients with ARC than without.[1819]

The objective of this study was to compare PK/PD predictions and clinical cure in patients with ARC treated with extended infusion cefepime or extended infusion piperacillin/tazobactam for P. aeruginosa bacteremia and/or pneumonia. We hypothesize that those who attained PK/PD thresholds would obtain clinical cure.

MATERIALS AND METHODS

Study design

This was a retrospective study of adult hospitalized patients who received extended infusion cefepime or extended infusion piperacillin/tazobactam for bacteremia and/or pneumonia due to P. aeruginosa from January 2013 to September 2016. The study was approved by the institutional review board in accordance with the ethical standard set forth in the Helsinki Declaration of 1975, and a waiver of consent was granted (IRB number 2016E0590). During the study period, cefepime was administered as loading dose of 2 g infused over 30 min followed by 2 g every 8 h infused over 4 h. Piperacillin/tazobactam was infused as a loading dose of 4.5 g infused over 30 min, followed by 4.5 g every 8 h infused over 4 h. Piperacillin/tazobactam was the preferred β-lactam when P. aeruginosa was suspected, and cefepime was second line but preferred for febrile neutropenia, in patients with penicillin allergy, or based on previous cultures and susceptibilities.

Patients were included if all the following criteria were met: (i) age >18 years, (ii) ARC defined as CrCl >130 ml/min based on the Cockcroft–Gault equation using actual body weight for patients up to 120% of ideal body weight and adjusted body weight (40% of the difference between actual and ideal body weight added to ideal body weight) for those above 120% of ideal body weight at the time of β-lactam administration, (iii) blood and/or respiratory culture with in vitro susceptibility to cefepime or piperacillin/tazobactam, (iv) cefepime or piperacillin/tazobactam administered within 72 h of culture obtainment, and (v) receipt of cefepime or piperacillin/tazobactam for >48 h. Only the first isolate per patient per study period was included. Patients were excluded if any of the following criteria were met: (i) receipt of concurrent β-lactam therapy with activity against P. aeruginosa within 2 days of the initiation of cefepime or piperacillin/tazobactam therapy or (ii) incarceration.

Pharmacokinetic/pharmacodynamic analysis

Published population PK studies in critically ill patients were used to generate estimated plasma concentrations every 6 min using first-dose drug exposure estimates including loading infusions.[2021] PD analyses were conducted using patient-specific PK parameters and MICs determined via E-test for P. aeruginosa. The PK/PD target attainment was defined as fT>MIC greater than 60% of the dosing interval for cefepime and greater than 50% for piperacillin/tazobactam.[22]

Data and outcomes

Demographic and clinical outcome data were collected from the electronic medical record. Data collected included age, gender, hospital service, Charlson comorbidity index,[23] intensive care unit (ICU) admission, mechanical ventilator status at culture collection, location in the ICU at the time of culture collection, microbiological data, antibiotics administered and treatment duration, infectious disease (ID) consult, and discharge disposition. Microbiological data included all respiratory and blood cultures positive for P. aeruginosa. Susceptibility testing for cefepime and piperacillin/tazobactam was performed using E-test (BioMérieux, Marcy l'Etoile, France) and confirmed using the MicroScan WalkAway System (Beckman Coulter). Treatment data included antibiotics administered for the P. aeruginosa infection. Concomitant therapy with an aminoglycoside, fluoroquinolone, or polymyxin was considered combination therapy if it was administered within 72 h of the positive culture, for >24 h, and P. aeruginosa was susceptible to the agent. The primary outcome was clinical cure. Clinical cure was defined as resolution of signs and symptoms of P. aeruginosa without the need for escalation of antibiotics and was assessed by an ID physician (JAB). An ID physician also determined infection-related mortality. Secondary clinical outcomes included length of hospital and ICU stay, duration of mechanical ventilation for patients with pneumonia, and in-hospital mortality.

Statistical analysis

Sample size was not calculated as there was a paucity of data comparing outcomes in patients with P. aeruginosa infections treated with cefepime or piperacillin/tazobactam let alone in patients with ARC. To minimize bias in this small study, we tried to include all patients who met inclusion and exclusion criteria. Nominal data were presented as percentages and analyzed using Chi-square or Fisher's exact tests. Nonparametric continuous and ordinal data were presented at median (25%–75% percentile) and analyzed using Mann–Whitney U-test. Parametric continuous data were presented as mean + standard deviation and analyzed using Student's t-test. Following univariate analysis of those with clinical success, a multivariable logistic regression was conducted to determine factors independently associated with clinical success. Factors significant at P ≤ 0.2 in the univariate analysis were included into a backward step-wise multiple logistic regression model. The results were reported as adjusted odds ratios (aORs) with a corresponding 95% confidence interval (CI). P <0.05 was considered statistically significant and all tests were two-tailed. SPSS version 21 (IBM, Armonk, NY, USA) was used for all calculations.

RESULTS

A total of 102 patients were included, 36 received extended infusion cefepime and 66 received extended infusion piperacillin/tazobactam [Figure 1]. The groups had similar demographics [Table 1]. Mean serum creatinine (0.55 ± 0.19 mg/dL vs. 0.52 ± 0.14 mg/dL, P = 0.36) and median creatinine clearance (157.2 [138.6–229] vs. 171.1 [147.3–205.1] ml/min, P = 0.70) were similar between groups. The majority of patients in each group were in the ICU at the time of culture collection (63.9% vs. 63.6%, P > 0.99), with 58.3% receiving cefepime requiring mechanical ventilation compared to 48.5% receiving piperacillin/tazobactam (P = 0.41). Similar percentages of patients were treated for pneumonia (52.8% vs. 65.2%, P = 0.29) and bacteremia (50% vs. 37.9%, P = 0.30). The median MIC in the cefepime group was 4 (2–8) mg/L and 8 (4–8) mg/L in the piperacillin/tazobactam group.

Figure 1

Flowchart for evaluation of patients for study inclusion

Table 1

Patient demographics

	Cefepime (n=36)	Piperacillin/tazobactam (n=66)	P
Age (years)	40.5±13.5	45.89±12.6	0.053
Female gender	17 (47.2)	22 (33.3)	0.2
Weight (kg)	75.2 (59.2-93)	74.8 (55.7-89.6)	0.81
Serum creatinine (mg/dL)	0.55±0.19	0.52±0.14	0.36
Creatinine clearance (mL/min)	157.2 (138.6-229)	171.1 (147.3-205.1)	0.7
Charlson comorbidity score	5 (3-6)	6 (4-8)	0.28
Site of infection
Blood	18 (50)	25 (37.9)	0.3
Respiratory	19 (52.8)	43 (65.2)	0.29
Both	1 (2.8)	2 (3)	>0.99
In ICU at time of culture collection	23 (63.9)	42 (63.6)	>0.99
Receipt of mechanical ventilation	21 (58.3)	32 (48.5)	0.41
Duration of mechanical ventilation (days)	11 (5-14)	8 (5-14)	0.82
Concomitant therapy
Aminoglycoside	8 (22.2)	18 (27.3)	0.64
Fluoroquinolone	1 (2.8)	8 (12.1)	0.15
Polymyxin	0	3 (4.5)	0.55

Data are presented as n (%), median (25%–75% interquartile range) or mean±SD. SD: Standard deviation, ICU: Intensive care unit

Overall, the PK/PD target was achieved in 65.7% of patients, including 83.3% of patients treated with cefepime and 60.5% of patients treated with piperacillin/tazobactam (P = 0.16). Patients receiving cefepime achieved a higher median %fT>MIC (93.6 [69.9–100] vs. 57.2 [47.6–72.4], P < 0.001) and clinical cure rate (91.7% vs. 74.2% P = 0.039) compared to those receiving piperacillin/tazobactam. These results are similar to the approximately two-thirds of patients admitted to an ICU where those receiving cefepime experience a statistically higher %fT>MIC (89.3 [61.3–99.4] vs. 55 [47.6–72.4], P < 0.001) and clinical cure (91.3–66.7, P = 0.036). Although hospital length of stay was longer in the cefepime group (22 [19-44] days vs. 15 [10-26] days, P < 0.001), infection-related length of stay was not different (14 [7-16] days vs. 9.5 [5-14] days, P = 0.13, [Table 2]). Overall mortality (8.3% vs. 22.7%, P = 0.10) and infection-related mortality (0% vs. 3%, P = 0.54) were similar between groups.

Table 2

Clinical outcome data for patient with Pseudomonas aeruginosa pneumonia and/or bacteremia

	Cefepime (n=36)	Piperacillin/tazobactam (n=66)	P
Overall mortality	3 (8.3)	15 (22.7)	0.1
Infection- related mortality	0	2 (3)	0.54
Hospital length of stay (days)	22 (19-44)	15 (10-26)	<0.001
Infection-related length of stay (days)	14 (7-16)	9.5 (5-14)	0.13
ICU length of stay (days)	16.5 (10-20.5)	10 (5-19)	0.12

Data are presented as n (%), median (25%–75% interquartile range) or mean±SD. SD: Standard deviation, ICU: Intensive care unit

On univariate analysis, only receipt of cefepime was associated with clinical cure [Table 3]. Factors entered into the multivariable analysis included age, gender, and cefepime administration. The multivariable logistic regression analysis only identified cefepime administration (aOR: 3.8, 95% CI: 1.04–14.1, P = 0.044) as an independent predictor of clinical cure.

Table 3

Univariate analysis of clinical cure

	No clinical cure (n=20)	Clinical cure (n=82)	P
Age (years)	49.5 (42-57)	43.5 (31-55)	0.15
Female gender	4 (20)	35 (42.7)	0.075
Serum creatinine (mg/dL)	0.56 (0.44-0.67)	0.5 (.0.44-0.6)	0.36
Creatinine clearance (mL/min)	168 (139.1-193.5)	170.1 (144.5-212.9)	0.47
Combination antibiotics	7 (35)	29 (35.4)	>0.99
Cefepime administration	3 (15)	33 (40.2)	0.039
Target attainment	14 (70)	62 (75.6)	0.78

Data are presented as n (%), median (25%–75% interquartile range)

DISCUSSION

Patients who received extended infusion cefepime achieved a significantly higher %fT>MIC and clinical cure rate compared to patients who received extended infusion piperacillin/tazobactam, and this relationship persisted after controlling for covariates. For those patients in an ICU, those receiving extended infusion cefepime also achieved a significantly higher %fT>MIC and clinical cure. This was the first study to compare PK/PD predictions and clinical outcomes in patients with ARC receiving cefepime and piperacillin/tazobactam P. aeruginosa bacteremia and/or pneumonia.

This work adds to the growing body of literature emphasizing the importance of achieving antimicrobial PK/PD targets to optimize clinical outcomes, particularly in patients infected with difficult-to-treat pathogens and those with physiological derangements such as ARC. Numerous studies have demonstrated an association between higher %fT>MIC and improved clinical outcomes, with many indicating that a goal of 100% fT>MIC may be ideal in the critically ill.[222425] Patients who received cefepime in our study achieved a %fT>MIC approaching 100% and were more likely to achieve a clinical cure than those who received piperacillin/tazobactam.

Given the recent studies showing increased acute kidney injury with the combination of vancomycin and piperacillin/tazobactam and the potential lack of efficacy of piperacillin/tazobactam in certain high risk patient populations such as febrile neutropenia, cefepime may be the drug of choice for empiric or targeted therapy of P. aeruginosa.[252627] Currently, there is a paucity of data comparing outcomes in patients with P. aeruginosa infections treated with cefepime or piperacillin/tazobactam. Recently, Jacobs et al. conducted at retrospective review of 215 critically ill patients with a creatinine clearance of >120 ml/min measured by 24 h urine and β-lactam therapeutic drug monitoring.[28] All patients had a diagnosis of sepsis or septic shock. Overall, 11 patients received cefepime 2 g every 8 h as a 30-min infusion and 89 patients receiving piperacillin/tazobactam 4.5 g every 6 h as a 30-min infusion. The median creatinine clearance was 179 (148–233) ml/min. Using a MIC of 8 mg/L for cefepime and 16 mg/L for piperacillin/tazobactam, the authors stated that the proportion of patients with insufficient concentrations to treat P. aeruginosa at these breakpoints was 82% for cefepime and 79% for piperacillin/tazobactam. In our study, the median MIC was lower for both cefepime (4 mg/L) and piperacillin/tazobactam (8 mg/L), and extended infusion was administered. Optimizing PK/PD indices may not be as simple as increasing the dose or infusion time as the MIC of the organism is an important factor. While there was a difference in achieving the bactericidal target between patients who received cefepime compared to those who received piperacillin/tazobactam in our study, additional data are needed to define PK/PD targets in patients with ARC.

There are additional data, demonstrating that intermittent piperacillin/tazobactam 4.5 g every 8 h may not be appropriately dosed even in patients without ARC. Andersen et al. conducted a prospective, PK study in 22 patients with sepsis who received piperacillin/tazobactam as an intermittent push.[29] In patients with ARC, the maximum MIC to achieve 90% probability of target attainment for 50% fT>MIC was 2 mg/L and 0.125 mg/L for 100% fT>MIC. The attainment of 50% fT>MIC might be too low of a PK/PD target in patients with ARC. Some have advocated for 100% fT>MIC attainment, especially in critically ill and immunocompromised patients as it is associated with better clinical cure and microbiologic eradication.[25273031] In our study, our median piperacillin/tazobactam MIC against P. aeruginosa was 8 mg/L. All patients received piperacillin/tazobactam 4.5 g every 8 h as an extended infusion; however, 4.5 g every 6 h may be required to optimize PD. At our institution, we do not measure β-lactam concentrations; therefore, we currently measure creatinine clearance based on an 8-h urine collection and adjust β-lactam dosing as appropriate. In patients with ARC, we use a piperacillin/tazobactam regimen of 4.5 g every 6 h administered as an extended infusion. In addition, our surgical ICU preferentially uses cefepime over piperacillin/tazobactam in patients with suspected or confirmed P. aeruginosa bacteremia and/or pneumonia.

There are several limitations to our study. First, this was a small retrospective study at a single institution. The dosing and administration of cefepime and piperacillin/tazobactam reflected the dosing at our institution at the time of the study and may not reflect the current practice. Determination of creatinine clearance was estimated using Cockcroft–Gault formula and not measured. Likewise, we estimated PK parameters using population estimates in critically ill patients as therapeutic drug monitoring of β-lactams is not routinely performed. Finally, the lack of assessment of severity of illness limited logistic regression analyses.

CONCLUSIONS

Critically ill patients with ARC and serious infections due to P. aeruginosa require optimization of pharmacologic therapy and achieving antimicrobial PK/PD targets is crucial to improve clinical outcomes. Our analysis revealed that extended infusion cefepime may more reliably achieve PK/PD endpoints and lead to improved clinical cure compared to extended infusion piperacillin/tazobactam. Additional studies are warranted in evaluating achievement of targeted PK/PD indices in patients with ARC receiving β-lactam therapy.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

Ethical conduct of research

This study was approved by the Institutional Review Board. The authors followed applicable EQUATOR Network (http://www.equator-network.org/) guidelines during the conduct of this research project.