- Email: [email protected]
tazobactam or cefepime in the intensive care unit
Accepted Manuscript Comparison of acute kidney injury risk associated with vancomycin and concomitant piperacillin/tazobactam or cefepime in the intensive care unit
Mitchell S. Buckley, Nicole C. Hartsock, Andrew J. Berry, Dale S. Bikin, Emily C. Richards, Melanie J. Yerondopoulos, Emir Kobic, Laura M. Wicks, Drayton A. Hammond PII: DOI: Reference:
S0883-9441(18)31024-4 doi:10.1016/j.jcrc.2018.08.007 YJCRC 52999
To appear in:
Journal of Critical Care
Please cite this article as: Mitchell S. Buckley, Nicole C. Hartsock, Andrew J. Berry, Dale S. Bikin, Emily C. Richards, Melanie J. Yerondopoulos, Emir Kobic, Laura M. Wicks, Drayton A. Hammond , Comparison of acute kidney injury risk associated with vancomycin and concomitant piperacillin/tazobactam or cefepime in the intensive care unit. Yjcrc (2018), doi:10.1016/j.jcrc.2018.08.007
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Comparison of Acute Kidney Injury Risk Associated with Vancomycin and Concomitant Piperacillin/Tazobactam or Cefepime in the Intensive Care Unit Running Head: Renal injury with vancomycin and beta-lactams
Mitchell S. Buckley, PharmD, FCCP, FASHP, FCCM, BCCCP,1 Nicole C. Hartsock, PharmD, MBA,1
IP
T
Andrew J. Berry, PharmD, BCPS,1 Dale S. Bikin, PharmD,1 Emily C. Richards, PharmD, BCPS,1 Melanie J. Yerondopoulos, PharmD, BCCCP,1 Emir Kobic, PharmD,1 Laura M. Wicks, PharmD, BCPS1;
US
CR
and Drayton A. Hammond, PharmD, MBA, BCPS, BCCCP2
Department of Pharmacy, Banner – University Medical Center Phoenix, Phoenix, AZ
1
2
AN
Department of Pharmacy, Rush University Medical Center, Chicago, IL
M
Keywords: acute kidney injury, cefepime, critical care, intensive care unit, nephrotoxicity, piperacillin
Abstract Word Count: 199
ED
tazobactam, vancomycin
PT
Word Count (excluding abstract, tables, references, figures): 3812
CE
Acknowledgements: The authors would like to thank Dr. Sumit K. Agarwal, MBBS, MBA, and Dr. Richard D. Gerkin, MD, at Banner University Medical Center Phoenix for assistance in data
AC
acquisition and statistical analysis. Financial support: No financial and material support was provided for this research. Conflict of interest: No conflicts of interest are reported by the authors pertaining to this article. Poster Presentation: This original research was submitted as an abstract for a poster presentation at the Society of Critical Care Medicine Annual Congress in February 2019 Correspondence:
ACCEPTED MANUSCRIPT Mitchell S. Buckley, PharmD, FASHP, FCCM, FCCP, BCCCP, Clinical Pharmacy Specialist, Banner – University Medical Center Phoenix, Department of Pharmacy, 1111 E. McDowell Road, Phoenix, AZ 85006, email: [email protected]
Abstract
T
Purpose
CR
compared to vancomycin and cefepime (FEP) in critically ill patients.
IP
The objective of this study was to evaluate AKI incidence with concomitant vancomycin and PTZ
US
Materials and Methods
AN
A retrospective, cohort study was conducted in adult critically ill patients from January 1, 2014 to December 31, 2017. The primary aim was to compare the incidence of AKI during concomitant therapy
M
or until hospital discharge. Secondary analyses included AKI severity, time to AKI as well as recovery,
ED
and clinical outcomes.
PT
Results
Overall, 333 patients were evaluated. The AKI rate in the vancomycin/PTZ group and
CE
vancomycin/FEP group were similar (19.5% vs. 17.3%, respectively, p=0.612). Renal replacement therapy was initiated in 10.0% and 3.8% administered vancomycin/PTZ and vancomycin/FEP groups,
AC
respectively (p=0.04). Multivariate regression found vancomycin/PTZ was not associated with an increased risk of developing AKI although the presence of shock was identified as an independent risk factor (odds ratio, 3.22; 95% CI, 1.66-6.26). No significant differences in hospital or ICU length of stay or in-hospital mortality were observed between study groups.
Conclusions Concomitant PTZ and vancomycin in ICU patients was not associated with an increased risk of
ACCEPTED MANUSCRIPT developing AKI compared to FEP and vancomycin combinations. More patients administered vancomycin/PTZ received RRT.
Introduction Acute kidney injury (AKI) is a common complication experienced in critically ill patients and
T
associated with deleterious outcomes. The incidence of AKI in the intensive care unit (ICU) is about 30-
IP
40% with some reports suggesting rates as high as 60% [1-4]. Sepsis, septic shock, and multiple organ
CR
dysfunction syndrome are most frequently implicated as the cause of AKI in the ICU.[3, 5, 6] Several risk factors have been identified including advanced age, severity of illness, higher baseline serum creatinine
US
(SCr), and nephrotoxic drug exposure [1]. Mortality, ICU length of stay, and mechanical ventilation
AN
duration are increased in ICU patients who develop AKI [4, 7]. The mortality risk incrementally increases with the severity of renal failure and death rates can exceed 50% in patients receiving renal replacement
M
therapy (RRT) [4, 8]. Adverse long-term outcomes include the development of chronic kidney disease as well as end-stage renal failure, higher risk of cardiovascular events, and reduced quality of life [8]. The
ED
economic impact is substantial with an additional cost of $28,000 to $56,035 per ICU patient with AKI
PT
requiring RRT compared to those without dialysis [4, 8]. Empiric antimicrobial therapy consisting of vancomycin with a broad-spectrum beta-lactam for
CE
methicillin-resistant Staphylococcus aureus and Pseudomonas aeruginosa coverage, respectively, are frequently utilized in ICU patients [9]. Vancomycin-associated nephrotoxicity has been widely
AC
recognized over the past several decades, with rates ranging from 5% to 43% depending on risk factors.[10] Increased vancomycin exposure (serum trough concentrations >15 mcg/mL, total daily doses >4 g, and area under the curve between 700-1300 mg/hr/L) may raise the risk of nephrotoxicity, especially in ICU patients [10]. Also, piperacillin-tazobactam (PTZ) has been associated with delayed AKI recovery in the critically ill, irrespective of vancomycin and other beta-lactams [11]. Interestingly, rapid reversal of AKI has been observed upon discontinuation of PTZ, while this fast renal recovery phenomenon was not found with other beta-lactams [11].
ACCEPTED MANUSCRIPT Recently, numerous reports evaluating the risk of AKI with concomitant vancomycin and PTZ have been published [12-22]. The combination of vancomycin with PTZ has been associated with higher AKI rates compared to either agent alone [21-24]. Several studies have also demonstrated a significant risk of nephrotoxicity with concomitant vancomycin and PTZ therapy compared to vancomycin with
T
cefepime (FEP) [12, 13, 15-18, 20, 25]. However, these findings were inconsistent, particularly in
IP
critically ill patients [14, 16]. Limitations of previously published trials include significant heterogeneity
CR
of the study patient population, AKI definition (creatinine-based diagnostic criteria without UOP), and inclusion criteria [12, 13, 15-18, 20, 25]. The incremental risk of AKI with PTZ over FEP, in conjunction
US
with vancomycin, remains controversial given these discordant results. Therefore, the purpose of this study was to evaluate the incidence of AKI with concomitant vancomycin and PTZ compared to
AN
vancomycin and FEP in ICU patients.
M
Materials and Methods
ED
Patients and Study Design
This was a retrospective, cohort study conducted at a three large (>500 beds) hospitals within the
PT
same health-system (two community and one university hospital in Phoenix, AZ). Patients concomitantly administered PTZ or FEP with vancomycin during their ICU admission were evaluated over a 4-year
CE
period (January 1, 2014 to December 31, 2017). Inclusion criteria consisted of the following: (1) ≥18
AC
years of age; (2) initiation of concomitant therapy was ≤48 hours after ICU admission; (3) ≥48 consecutive hours of combination therapy during the patient’s ICU stay; (4) baseline creatinine clearance (CrCl) >30 mL/min; and (5) at least one SCr ≤24 hours of ICU admission. Patients with combination therapy (i.e., either PTZ or FEP with intravenous vancomycin) were included if the beta-lactam was administered within 48 hours of vancomycin irrespective of the antibiotic initiated first. Patients were excluded if they had any of the following: (1) vancomycin/PTZ or vancomycin/FEP combinations were initiated in hospitalized patients prior to ICU admission; (2) both PTZ and FEP were administered during
ACCEPTED MANUSCRIPT the same ICU admission; (3) end-stage renal disease; (4) any RRT prior to combination therapy; or (5) combination therapy discontinued and restarted within 7 days. Clinical pharmacist management of renal dosed antibiotics and/or requiring therapeutic drug monitoring (vancomycin and aminoglycosides) was standardized through the use of approved protocols at
T
the study institutions. Vancomycin loading doses were incorporated upon commencement of therapy
IP
using a weight-based approach (20-30 mg/kg) of actual body weight (ABW) rounded to the nearest 250
CR
mg increment up to a maximum of 2000 mg. Vancomycin maintenance dosing regimens were initiated by the pharmacist based on the patients ABW and renal function (CrCl or mode of RRT). Clinical
US
pharmacists were expected to obtain an initial vancomycin serum trough concentration no later than prior to the fifth maintenance dose with a goal target concentration of 10-20 mcg/mL based on the indication
AN
(e.g., 15-20 mcg/mL for pneumonia). However, the clinical pharmacists could obtain vancomycin concentrations at an earlier time, including post-loading dose, in patients with severely impaired or
M
rapidly fluctuating renal function as well as obesity. Follow-up vancomycin serum concentrations were
ED
recommended within seven days of a previous serum concentration for patients receiving longer durations of therapy. However, the pharmacist could order a serum concentration at any time prior to the 7-day
PT
limit to monitor therapy based on the clinical scenario (e.g., changing renal function). Furthermore, the
CE
clinical pharmacist dosed FEP and PTZ based on renal function or RRT pursuant to a standardized protocol. The study sites implemented extended-interval dosing for PTZ prior to the study period. The
AC
standard of practice for PTZ dosing involves 4.5 g loading dose to be administered to patients over 30 minutes upon initiation of therapy followed by subsequent doses of 3.375 g administered over 4-hour intervals. The frequency of dosing is every 8 hours in patients with CrCl ≥20 mL/min or receiving continuous RRT (CRRT). Patients with CrCl <20 mL/min, intermittent hemodialysis, or sustained lowefficiency dialysis were dosed every 12 hours. Cefepime was also renally dosed in patients with CrCl <60 mL/min with multiple acceptable doses and frequencies ranging from 500 mg daily to 2 g every 8 hours. The dosing strategy was customized for each patient, taking into account several variables including renal
ACCEPTED MANUSCRIPT function, location of infection, patient weight, and severity of illness. Also, the only mode of CRRT adopted by the study sites was continuous veno-venous hemodiafiltration. Clinical pharmacists reviewed appropriate vancomycin, PTZ, and FEP dosing at least once daily.
T
Data Collection
IP
After Institutional Review Board approval, patients meeting study criteria were identified through
CR
the electronic medical record database. Patients were assigned to one of two cohorts based upon administration of PTZ or FEP. Patient demographic and pertinent clinical data collected were age, gender,
US
height, actual and ideal body weight, comorbidities, SCr, known or suspected locations of infection, RRT requirements (hemodialysis, CRRT, or sustained low efficiency dialysis), and concomitant nephrotoxic
AN
agents. The potential concomitant nephrotoxic agents evaluated were aminoglycosides, amphotericin B (conventional or lipid formulation), angiotensin converting enzyme inhibitors, angiotensin II receptor
M
blockers, calcineurin inhibitors, colistin, contrast dye, loop diuretics, foscarnet, nonsteroidal anti-
ED
inflammatory drugs, and acyclovir. Aerosolized antimicrobials (e.g., colistin, tobramycin, and amphotericin B) were not considered as potentially nephrotoxic exposures. Vancomycin, PTZ, and FEP
PT
doses administered as well as duration of therapy were collected. In addition, vancomycin serum
CE
concentrations and timing in relation to administration were assessed. Serum creatinine was evaluated using an estimated CrCl via the Cockroft-Gault formula [26]. The highest SCr value reported for each
AC
patient day in the medical record was used to calculate CrCl. Other non-drug related risk factors possibly associated with AKI in the ICU setting were also collected, such as the presence of shock, need for mechanical ventilation, end-stage liver disease, and chronic kidney disease [5].
Definitions Acute kidney injury was defined using the Risk, Injury, Failure, Loss, End Stage Renal Disease (RIFLE) criteria.[27] The RIFLE criteria defines Risk, Injury, and Failure for AKI as follows: SCr
ACCEPTED MANUSCRIPT increase of 1.5 times baseline, glomerular filtration rate (GFR) reduction >25%, or UOP <0.5mL/kg/hour by 6 hours; SCr increase of 2 times baseline, GFR decrease >50%, or UOP <0.5mL/kg/hour by 12 hours; and SCr increase of 3 times baseline, GFR reduction >75%, UOP <0.3mL/kg/hour by 24 hours or anuria over 12 hours, respectively.[27] Each patient’s GFR was estimated using CrCl. Renal function was
T
assessed daily using SCr, CrCl, and UOP while the patient was in the ICU. Patients experiencing AKI
IP
were categorized using the RIFLE (Risk, Injury, or Failure) criteria. The worst-case scenario for each patient’s UOP or SCr were used to determine the RIFLE stage to assign the degree of AKI. For example,
CR
if a patient’s UOP aligns with the “injury” stage of the criteria, while the SCr suggests “risk,” then the
US
patient was classified as “injury” since it is the more severe AKI classification. Patients initiated on combination therapy in the ICU and subsequently transferred to the general ward were only assessed by
AN
GFR or SCr for the RIFLE classification criteria since hourly UOP was not typically measured and reported in general ward patients at the study institutions. Recovery of AKI was defined as a return of SCr
M
to <0.3 mg/dL above baseline <3 days (“fast”), 3-7 days (“intermediate”), and > 7 days (“slow”) based on
ED
a modified version of a previously published report [28]. Other definitions used in this study were “baseline” and “shock.” “Baseline” was defined as the
PT
first patient day on which vancomycin with FEP or vancomycin with PTZ were concomitantly
CE
administered. For instances in which the second adjunctive antimicrobial agent was not initiated on the same day, baseline was considered as the time when both were administered within the same day
AC
irrespective of which agent was started first. The presence of “shock” was determined as the use of continuous infusion vasoactive support (norepinephrine, dopamine, epinephrine, phenylephrine, vasopressin) for at least 1 hour, mean arterial pressure <60 mm Hg for at least 1 hour, or serum lactate concentration ≥4 mmol/L.
Study Outcomes and Statistical Analysis The primary endpoint was the incidence of AKI during concomitant antimicrobial therapy and up
ACCEPTED MANUSCRIPT to hospital discharge. Secondary analyses were AKI severity rates, days to AKI recovery, rates of requiring RRT, and duration of RRT. Clinical outcomes including in-hospital mortality and hospital and ICU length of stay were also compared. A sample size of 300 patients (200 and 100 in the vancomycin/PTZ and vancomycin/FEP groups, respectively) was needed to achieve 80% power in
T
detecting ≥15% increase in the incidence of AKI associated with vancomycin and PTZ compared to
IP
vancomycin and FEP assuming AKI rates of 30% and 15%, respectively. These estimates were based on
CR
previously published reports in ICU patients [14, 16, 25]. Patients from a convenience sample in the vancomycin and PTZ group were consecutively selected from the most recent administration date until
US
the minimum number of patients needed to meet the required sample size were included. Fisher’s exact test was used for categorical variables and Student’s t-test for continuous data. Variables potentially
AN
associated with AKI were further evaluated using a stepwise multivariate regression analysis to determine independent risk factors significantly associated with AKI. Although a p-value cut-off point of 0.1 was
M
established, antimicrobial combinations and vasoactive agents were included based on investigators’
ED
preference. A p-value of <0.05 was considered statistically significant. Statistical analyses were
PT
performed using IBM SPSS Statistics for Mac version 24.0 (IBM, Armonk, NY, 10504).
CE
Results
A total of 3861 patients were identified through the electronic medical record during the study
AC
period and screened for inclusion. Overall, 333 patients met inclusion criteria and were included in the analysis with 200 (66.7%) patients in the vancomycin/PTZ group and 133 (33.3%) patients in the vancomycin/FEP group (Figure 1). The baseline characteristics between study groups were similar except for a few comorbid conditions, including hepatic dysfunction, chronic obstructive pulmonary disease, and hypertension (Table 1). Baseline severities of illness were similar between groups with respect to SOFA score, invasive mechanical ventilation rates, and incidence of patients requiring vasoactive support (Table 1). The vast majority (92.0%) of all study patients received ≥1 concomitant nephrotoxic agents without
ACCEPTED MANUSCRIPT any differences between study cohorts for any specific nephrotoxin (Table 1). No significant differences in the duration of concomitant therapy were found between vancomycin/PTZ (5.1 ± 3.8 days) and vancomycin/FEP (5.8 ± 5.2 day) (p=0.185). The mean total daily dose of vancomycin was similar in the PTZ and FEP groups (2200.3 ± 990.9 and 2271.9 ± 978.3 mg, respectively, p=0.517). The median initial
T
vancomycin serum concentration in the vancomycin/PTZ and vancomycin/FEP groups were 13.5 (10.5-
IP
18.6) and 13.1 (9.8-17.7) mcg/mL, respectively, p=0.369. Furthermore, the mean beta-lactam dose
CR
administered in the FEP cohort was 4.0 ± 1.6 g/day, while the PTZ cohort was 11 ± 1.6 g/day. The overall rate of AKI observed among all study patients was 18.6% (n=62). The AKI rate in the
US
vancomycin/PTZ group compared to the vancomycin/FEP group was not significantly different (19.5% vs. 17.3%, respectively, p=0.612). Although no difference was found in the mean time to AKI between
AN
study groups, significantly more patients receiving vancomycin/PTZ compared to the vancomycin/FEP group experienced early AKI (≤3 days) after initiating concomitant antimicrobial therapy. However, the
M
peak incidence of AKI onset occurred ≤24 hours from baseline with 66.7% (n=26) in the
ED
vancomycin/PTZ group compared to 65.2% (n=15) (p=0.858). Overall, about 63% of all patients experiencing AKI progressed to a RIFLE classification of “Failure” without any differences in severity
PT
between study groups. Also, the rate of patients requiring RRT for AKI was two-fold higher with
CE
vancomycin/PTZ than vancomycin/FEP (p=0.035). Recovery patterns for AKI were not significantly different between study groups with most patients (88.7%) failing to recovery to baseline renal function.
(Table 2).
AC
Clinical outcomes including length of stay (overall hospitalization and ICU) and mortality were similar
The choice of beta-lactam in combination with vancomycin was not an independent risk factor for developing AKI (Table 3). Patients were more likely to develop AKI with the presence of any type of shock, higher SOFA score, or requiring at least four vasopressor agents at baseline in the univariate analysis. However, multiple variable regression analysis found shock (any type) was 3-times as likely to develop AKI, while significance was lost for SOFA score and four vasoactive medications. Concomitant
ACCEPTED MANUSCRIPT antimicrobial therapy duration trended towards a significantly increased risk of renal dysfunction. Multiple regression analysis also found lower odds of AKI in male gender patients and those individuals requiring only one catecholamine agent at baseline. None of the adjunctive nephrotoxic agents were
T
found to predict AKI.
IP
Discussion
CR
The aim of this study was to investigate the risk of developing AKI between two common broadspectrum combination antimicrobial strategies (vancomycin and PTZ vs. vancomycin and FEP) in the
US
ICU. To our knowledge, this was the first multicenter study in ICU patients to describe AKI characteristics (occurrence, onset, severity, RRT rates, and recovery patterns). Concomitant vancomycin
AN
and PTZ did not increase the risk of AKI in critically ill patients. Similar characteristics between the combination groups, including AKI onset, severity, and recovery patterns, were observed. Patients
M
developing AKI in this study may have largely been attributed to non-drug causes. The vast majority of
ED
AKI identified in both groups occurred about 24 hours after antimicrobial combination therapy initiation, which might be too soon for antibiotic exposure to suggest medication-induced causes. Furthermore,
PT
shock was identified as a risk factor for AKI on multivariate analysis, while no association was found
CE
with vancomycin/PTZ or other known nephrotoxic medications. Although the administration of RRT was more common in patients receiving vancomycin/PTZ over the FEP combination, this finding should be
AC
considered with caution due to the small number of patients in both study groups. The clinical implication of this study implies providers may not need to avoid concomitant vancomycin and PTZ merely based on concerns of increasing AKI risk in ICU patients. Our study failed to demonstrate an increased risk of AKI with vancomycin/PTZ compared to vancomycin/FEP, which is consistent with previously published reports in ICU patients [14, 16, 25]. Recent meta-analyses have reported significantly higher AKI rates associated with vancomycin/PTZ compared to vancomycin combinations with either FEP or carbapenem [16, 25]. However, a subgroup
ACCEPTED MANUSCRIPT analysis consisting of pooled data in critically ill patients found no significant difference in AKI rates between vancomycin/PTZ compared to vancomycin with either FEP or carbapenem (OR=1.32, 0.73-2.38) [16]. Hammond et al, conducted a study in an ICU population evaluating AKI rates and found no differences between vancomycin/PTZ (32.7%) and vancomycin/FEP groups (28.8%) (p=0.647) [14].
T
Several aspects of their study design (inclusion criteria, duration of the assessment period, etc) were
IP
similar to ours. Although both studies failed to demonstrate a significant difference in AKI rates between
CR
groups, a few dissimilarities may explain the lower rates observed in our study. Vancomycin dosing was primarily managed by clinical pharmacists at our study sites, while their institution was not reflective of a
US
formal pharmacy-managed service. Subsequently, the mean initial vancomycin serum trough concentration in both groups was lower in our study (<15 mcg/mL) compared to theirs (>15 mcg/mL).
AN
Vancomycin serum concentrations >15 mcg/mL have been associated as a risk factor for developing AKI [29]. Most of our ICU patients admitted during the study period failed to meet inclusion criteria due to
M
multisystem organ dysfunction including AKI. Therefore, the acuity of illness in our ICU population
ED
meeting study inclusion criteria may have been lower as evidenced by our SOFA scores, incidence of shock, proportion of patient requiring vasoactive support and invasive mechanical ventilation, mean
PT
length of stay, and mortality rates. Another important difference was the adopted definition of AKI
CE
utilized in our respective studies (AKIN vs. RIFLE), which may have resulted in inconsistencies for classifying the incidence AKI [30-33]. Although our study included patients administered extended-
AC
infusion PTZ rather than traditional intermittent administration, the beta-lactam infusion strategy has not been shown to be an independent risk factor for developing AKI [34, 35]. Several other studies have investigated the rate of AKI with vancomycin/PTZ compared to vancomycin/FEP in hospitalized patients with conflicting results. The overall incidence of AKI in these reports largely varied with >50% of these reports suggesting an association between increased AKI risk and vancomycin/PTZ utilization [12, 13, 15, 17, 18, 20]. Unfortunately, a direct comparison among these studies remains challenging given the significant heterogeneity of the study populations, AKI definition,
ACCEPTED MANUSCRIPT and inclusion criteria. Two studies included critically ill patients within their respective study cohorts, but did not separately report ICU data [13, 18]. Most of these studies relied solely on SCr (absolute value thresholds, magnitude of change, and/or CrCl estimates) as the primary method to define AKI [12, 13, 15, 17, 18, 20]. Therefore, it is important to recognize the limitation of creatinine-based diagnostic AKI
T
criteria, including delays in SCr concentration changes following the initial kidney damage, disease- and
IP
drug-induced fluctuations on SCr concentrations (production and/or clearance) independent of the
CR
underlying renal function, “baseline” SCr definitions, and the impact of fluid status (volume overload and dehydration) on SCr [6, 36]. These factors may explain the variability in the incidence of AKI among
US
these reports despite using validated criteria (AKIN, RIFLE, KDIGO) [13, 15, 17, 18, 20]. Although evaluating UOP may offer some advantages over SCr, UOP alone to classify AKI is not without
AN
limitations as well [36]. Our study utilized both SCr and UOP to identify AKI and stages of severity. Few studies evaluating vancomycin/PTZ and vancomycin/FEP have described AKI
M
characteristics and patterns of time to onset, severity, duration, need for RRT, and recovery patterns.
ED
Three studies assessed time to AKI, which all found an increased AKI risk with the vancomycin/PTZ combination [13, 18, 20]. Although one study did not show any difference in time to AKI between
PT
groups, two found AKI occurred more rapidly upon antibiotic initiation with the addition of PTZ (3-5
CE
days) compared to FEP (5-8 days) [16]. This is in contrast to our findings that AKI onset was not significantly different between groups, but equally occurred much earlier (<3 days) in both antimicrobial
AC
groups. A recent meta-analysis also found no significant difference in time to AKI between patients receiving vancomycin/PTZ versus vancomycin/FEP [16]. Furthermore, the severity stage associated with AKI in previous studies was different than our findings [13, 15, 18, 20]. We found no difference in AKI severity stages using RIFLE between study groups with over 50% of all AKI classified as “failure”. The most common AKI severity categories in other studies were “risk” and “stage I” for RIFLE and AKIN, respectively [13, 15, 18, 20]. Interestingly, none of the published studies found any significant difference between the rate of patients requiring RRT or the duration of AKI between study groups [13, 14, 18].
ACCEPTED MANUSCRIPT Although our study found no difference in duration of AKI between groups, we did find a significantly higher rate of RRT in patients administered vancomycin/PTZ. However, the higher rate of RRT in patients administered vancomycin/PTZ observed in our study did not exceed the estimated incidence of using RRT for AKI in an ICU setting [3]. Gomes et al, found a high proportion of patients overall failed
T
to fully recover renal function with vancomycin/PTZ (59.0%) and vancomycin/FEP (78.6%) during the
IP
evaluation period (p=0.190) [13]. We also found an unacceptably high rate exceeding 88% of all patients
CR
experiencing AKI (n=62) did not fully recover renal function back to baseline with rates similar between vancomycin/PTZ and vancomycin/FEP (92.3% and 82.6%, respectively). It is important to emphasize the
US
definition of recovery in our study, which was SCr returning ≤0.3 mg/dL of baseline. It is possible patients experienced functional recovery of their kidneys not reflective of their SCr during or after the
AN
evaluation period.
This study has several limitations. First, the retrospective study design may have introduced
M
confounding variables leading to bias and spurious findings. However, groups were well balanced and
ED
multivariate analysis was performed to increase the robustness as well as corroborate our findings. Also, the small sample size may not have been sufficient to observe any study group differences. The rate of
PT
AKI observed in both study groups was lower than anticipated for our sample size calculation a priori. Nonetheless, it is equally important to recognize the limitations of power analysis utility in retrospective
CE
studies. Furthermore, the rigorous inclusion and exclusion criteria may limit this study’s generalizability
AC
for higher acuity ICU patients or those with baseline CrCl ≤30 mL/min. Unfortunately, despite our best efforts to exclude patients experiencing AKI prior to initiation of antimicrobials, we acknowledge some of these patients may have been included in our analysis. The initial kidney insult may have preceded any manifestations in SCr and/or UOP changes upon evaluation to meet our exclusion criteria. Lastly, accurately identifying drug-induced AKI is challenging in the critically ill considering potential etiologies are likely complex and multifactorial.
ACCEPTED MANUSCRIPT Conclusions Critically ill patients who received concomitant vancomycin/PTZ therapy were not at an increased risk of AKI compared to vancomycin/FEP. Irrespective of the beta-lactam choice with vancomycin, this study found no differences in the AKI incidence, severity, time to onset, duration, or
T
recovery patterns. However, RRT was initiated more frequently in patients administered
IP
vancomycin/PTZ. Although the association between concomitant vancomycin/PTZ and increased AKI
not be warranted in critically ill patients.
Cartin-Ceba R, Kashiouris M, Plataki M, Kor DJ, Gajic O, Casey ET. Risk factors for
AN
[1]
US
References
CR
risk remains debatable, these findings suggest a change in clinical practice to avoid this combination may
development of acute kidney injury in critically ill patients: a systematic review and meta-
[2]
M
analysis of observational studies. Crit Care Res Pract 2012;2012:691013. Srisawat N, Sileanu FE, Murugan R, Bellomod R, Calzavacca P, Cartin-Ceba R, et al. Variation
[3]
PT
Nephrol 2015;41(1):81-8.
ED
in risk and mortality of acute kidney injury in critically ill patients: a multicenter study. Am J
Hoste EA, Bagshaw SM, Bellomo R, Cely CM, Colman R, Cruz DN, et al. Epidemiology of
CE
acute kidney injury in critically ill patients: the multinational AKI-EPI study. Intensive Care Med 2015;41(8):1411-23.
Hoste EA, Schurgers M. Epidemiology of acute kidney injury: how big is the problem? Crit Care
AC
[4]
Med 2008;36(4 Suppl):S146-51. [5]
Uchino S, Kellum JA, Bellomo R, Doig GS, Morimatsu H, Morgera S, et al. Acute renal failure in critically ill patients: a multinational, multicenter study. Jama 2005;294(7):813-8.
[6]
Ostermann M, Chang RW. Acute kidney injury in the intensive care unit according to RIFLE. Crit Care Med 2007;35(8):1837-43; quiz 52.
[7]
Vieira JM, Jr., Castro I, Curvello-Neto A, Demarzo S, Caruso P, Pastore L, Jr., et al. Effect of
ACCEPTED MANUSCRIPT acute kidney injury on weaning from mechanical ventilation in critically ill patients. Crit Care Med 2007;35(1):184-91. [8]
Silver SA, Chertow GM. The Economic Consequences of Acute Kidney Injury. Nephron 2017;137(4):297-301. Kalil AC, Metersky ML, Klompas M, Muscedere J, Sweeney DA, Palmer LB, et al. Management
T
[9]
IP
of Adults With Hospital-acquired and Ventilator-associated Pneumonia: 2016 Clinical Practice
CR
Guidelines by the Infectious Diseases Society of America and the American Thoracic Society. Clin Infect Dis 2016;63(5):e61-e111.
Carreno JJ, Kenney RM, Lomaestro B. Vancomycin-associated renal dysfunction: where are we now? Pharmacotherapy 2014;34(12):1259-68.
Jensen JU, Hein L, Lundgren B, Bestle MH, Mohr T, Andersen MH, et al. Kidney failure related
AN
[11]
US
[10]
to broad-spectrum antibiotics in critically ill patients: secondary end point results from a 1200
Davies SW, Efird JT, Guidry CA, Dietch ZC, Willis RN, Shah PM, et al. Top Guns: The
ED
[12]
M
patient randomised trial. BMJ Open 2012;2(2):e000635.
"Maverick" and "Goose" of Empiric Therapy. Surg Infect (Larchmt) 2016;17(1):38-47. Gomes DM, Smotherman C, Birch A, Dupree L, Della Vecchia BJ, Kraemer DF, et al.
PT
[13]
CE
Comparison of acute kidney injury during treatment with vancomycin in combination with piperacillin-tazobactam or cefepime. Pharmacotherapy 2014;34(7):662-9. Hammond DA, Smith MN, Painter JT, Meena NK, Lusardi K. Comparative Incidence of Acute
AC
[14]
Kidney Injury in Critically Ill Patients Receiving Vancomycin with Concomitant PiperacillinTazobactam or Cefepime: A Retrospective Cohort Study. Pharmacotherapy 2016;36(5):463-71. [15]
Jeon N, Staley B, Klinker KP, Hincapie Castillo J, Winterstein AG. Acute kidney injury risk associated with piperacillin/tazobactam compared with cefepime during vancomycin therapy in hospitalised patients: a cohort study stratified by baseline kidney function. Int J Antimicrob Agents 2017;50(1):63-7.
ACCEPTED MANUSCRIPT [16]
Luther MK, Timbrook TT, Caffrey AR, Dosa D, Lodise TP, LaPlante KL. Vancomycin Plus Piperacillin-Tazobactam and Acute Kidney Injury in Adults: A Systematic Review and MetaAnalysis. Crit Care Med 2018;46(1):12-20.
[17]
Moenster RP, Linneman TW, Finnegan PM, Hand S, Thomas Z, McDonald JR. Acute renal
T
failure associated with vancomycin and beta-lactams for the treatment of osteomyelitis in
IP
diabetics: piperacillin-tazobactam as compared with cefepime. Clin Microbiol Infect
[18]
CR
2014;20(6):O384-9.
Navalkele B, Pogue JM, Karino S, Nishan B, Salim M, Solanki S, et al. Risk of Acute Kidney
US
Injury in Patients on Concomitant Vancomycin and Piperacillin-Tazobactam Compared to Those on Vancomycin and Cefepime. Clin Infect Dis 2017;64(2):116-23. Peyko V, Smalley S, Cohen H. Prospective Comparison of Acute Kidney Injury During
AN
[19]
Treatment With the Combination of Piperacillin-Tazobactam and Vancomycin Versus the
Rutter WC, Cox JN, Martin CA, Burgess DR, Burgess DS. Nephrotoxicity during Vancomycin
ED
[20]
M
Combination of Cefepime or Meropenem and Vancomycin. J Pharm Pract 2017;30(2):209-13.
Therapy in Combination with Piperacillin-Tazobactam or Cefepime. Antimicrob Agents
Burgess LD, Drew RH. Comparison of the incidence of vancomycin-induced nephrotoxicity in
CE
[21]
PT
Chemother 2017;61(2).
hospitalized patients with and without concomitant piperacillin-tazobactam. Pharmacotherapy
[22]
AC
2014;34(7):670-6.
Rutter WC, Burgess DR, Talbert JC, Burgess DS. Acute kidney injury in patients treated with vancomycin and piperacillin-tazobactam: A retrospective cohort analysis. J Hosp Med 2017;12(2):77-82.
[23]
McQueen KE, Clark DW. Does Combination Therapy With Vancomycin and PiperacillinTazobactam Increase the Risk of Nephrotoxicity Versus Vancomycin Alone in Pediatric Patients? J Pediatr Pharmacol Ther 2016;21(4):332-8.
ACCEPTED MANUSCRIPT [24]
Kim T, Kandiah S, Patel M, Rab S, Wong J, Xue W, et al. Risk factors for kidney injury during vancomycin and piperacillin/tazobactam administration, including increased odds of injury with combination therapy. BMC Res Notes 2015;8:579.
[25]
Hammond DA, Smith MN, Li C, Hayes SM, Lusardi K, Bookstaver PB. Systematic Review and
T
Meta-Analysis of Acute Kidney Injury Associated with Concomitant Vancomycin and
Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron
CR
[26]
IP
Piperacillin/tazobactam. Clin Infect Dis 2017;64(5):666-74.
1976;16(1):31-41.
Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P, Acute Dialysis Quality Initiative w.
US
[27]
Acute renal failure - definition, outcome measures, animal models, fluid therapy and information
AN
technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 2004;8(4):R204-12. Heung M, Steffick DE, Zivin K, Gillespie BW, Banerjee T, Hsu CY, et al. Acute Kidney Injury
M
[28]
ED
Recovery Pattern and Subsequent Risk of CKD: An Analysis of Veterans Health Administration Data. Am J Kidney Dis 2016;67(5):742-52. van Hal SJ, Paterson DL, Lodise TP. Systematic review and meta-analysis of vancomycin-
PT
[29]
CE
induced nephrotoxicity associated with dosing schedules that maintain troughs between 15 and 20 milligrams per liter. Antimicrob Agents Chemother 2013;57(2):734-44. Bagshaw SM, George C, Bellomo R, Committe ADM. A comparison of the RIFLE and AKIN
AC
[30]
criteria for acute kidney injury in critically ill patients. Nephrol Dial Transplant 2008;23(5):156974. [31]
Chang CH, Lin CY, Tian YC, Jenq CC, Chang MY, Chen YC, et al. Acute kidney injury classification: comparison of AKIN and RIFLE criteria. Shock 2010;33(3):247-52.
[32]
Lopes JA, Fernandes P, Jorge S, Goncalves S, Alvarez A, Costa e Silva Z, et al. Acute kidney injury in intensive care unit patients: a comparison between the RIFLE and the Acute Kidney
ACCEPTED MANUSCRIPT Injury Network classifications. Crit Care 2008;12(4):R110. [33]
Pereira M, Rodrigues N, Godinho I, Gameiro J, Neves M, Gouveia J, et al. Acute kidney injury in patients with severe sepsis or septic shock: a comparison between the 'Risk, Injury, Failure, Loss of kidney function, End-stage kidney disease' (RIFLE), Acute Kidney Injury Network (AKIN)
T
and Kidney Disease: Improving Global Outcomes (KDIGO) classifications. Clin Kidney J
Cotner SE, Rutter WC, Burgess DR, Wallace KL, Martin CA, Burgess DS. Influence of beta-
CR
[34]
IP
2017;10(3):332-40.
Lactam Infusion Strategy on Acute Kidney Injury. Antimicrob Agents Chemother 2017;61(10). Karino S, Kaye KS, Navalkele B, Nishan B, Salim M, Solanki S, et al. Epidemiology of Acute
US
[35]
Kidney Injury among Patients Receiving Concomitant Vancomycin and Piperacillin-Tazobactam:
AN
Opportunities for Antimicrobial Stewardship. Antimicrob Agents Chemother 2016;60(6):374350.
CE
PT
ED
Biochem Rev 2016;37(4):153-75.
M
Makris K, Spanou L. Acute Kidney Injury: Diagnostic Approaches and Controversies. Clin
AC
[36]
ACCEPTED MANUSCRIPT Table 1. Patient demographic and clinical characteristics
PT
CE
AC
p value 0.93 0.60 0.33 0.35 0.27 0.33 0.10 0.83 0.44
42 (31.6%) 4 (3.0%) 18 (13.5%)
0.86 0.30 0.07
IP
T
Vancomycin + Cefepime (n=133) 57.1 ± 15.6 84 (63.2%) 4.9 ± 3.1 169.7 ± 10.5 86.2 ± 33.9 62.5 ± 9.7 46 (34.6%) 17 (12.8%) 19 (14.3%)
CR
US
65 (32.5%) 12 (6.0%) 15 (7.5%)
0.06 100 (75.2%) 26 (19.6%) 6 (4.5%) 1 (0.8%) 0 (0.0%)
68 (34.0%) 51 (25.5%) 69 (34.5%) 41 (20.5%) 40 (20.0%) 1 (0.5%) 6 (3.0%) 24 (12.0%)
49 (36.8%) 32 (24.1%) 64 (48.1%) 40 (30.1%) 13 (9.8%) 1 (0.8%) 7 (5.3%) 25 (18.8%)
0.60 0.77 0.01 0.05 0.01 1.00 0.39 0.09
26 (13.0%) 2 (1.0%) 55 (27.5%) 4 (2.0%) 0 (0.0%) 124 (62.0%) 0 (0.0%) 29 (14.5%) 138 (69.0%) 7 (3.5%)
27 (20.3%) 4 (3.0%) 32 (24.1%) 6 (4.5%) 0 (0.0%) 85 (63.9%) 0 (0.0%) 21 (15.8%) 76 (57.1%) 4 (3.0%)
0.07 0.22 0.48 0.21 N/A 0.72 N/A 0.75 0.03 1.00
AN
138 (69.0%) 38 (19.0%) 12 (6.0%) 6 (3.0%) 6 (3.0%)
M
ED
Variable Age, years Male gender SOFA score Height, cm TBW, kg IBW, kg Mechanical ventilation at baseline MAP <60 mmHg at baseline Serum lactate ≥4 mmol/L at baseline Type of shock Vasodilatory Cardiogenic Other Number of vasopressor agents at baseline None 1 2 3 4 Comorbidities Diabetes mellitus Heart failure Hypertension COPD Hepatic dysfunction HIV/AIDS Transplant Cancer Concomitant nephrotoxic agents Aminoglycoside Amphotericin ACEI/ARB Calcineurin Inhibitor Colistin Diuretic Foscarnet NSAID Contrast Acyclovir (intravenous)
Vancomycin + Piperacillin/Tazobactam (n=200) 56.9 ± 16.9 132 (66.0%) 5.1 ± 3.8 179.6 ± 122.1 82.4 ± 28.0 71.8 ± 110.6 87 (43.5%) 24 (12.0%) 35 (17.5%)
Data presented as mean ± SD or n (%). AIDS = acquired immune deficiency syndrome; ACEI = angiotensin-converting-enzyme inhibitor; ARB = angiotensin-II receptor blocker; COPD =
ACCEPTED MANUSCRIPT chronic obstructive pulmonary disease; HIV = human immunodeficiency virus; IBW = ideal body weight; MAP = mean atrial pressure; NSAID = nonsteroidal anti-inflammatory drug; SOFA = Sequential Organ Failure Assessment; TBW = total body weight.
AC
CE
PT
ED
M
AN
US
CR
IP
T
:
ACCEPTED MANUSCRIPT Table 2. Clinical outcomes between study groups
34 (87.2%) 1 (2.6%)
18 (78.3%) 5 (21.7%)
8 (20.5%) 7 (17.9%) 24 (61.5%)
5 (21.7%) 3 (13.0%) 15 (65.2%)
IP CR
0.03 -0.88 ----
1 (2.6%) 1 (2.6%) 1 (2.6%) 36 (92.3%)
3 (13.0%) 0 (0.0%) 1 (4.3%) 19 (82.6%)
14.4 ± 8.8 9.1 ± 6.9 26 (13.0%)
15.6 ± 11.8 10.5 ± 10.2 20 (15.0%)
0.31 0.18 0.60
AN
5 (3.8%) 17.2 ± 9.7
p value 0.61 0.65
0.04 0.10 0.34 -----
US
20 (10.0%) 10.2 ± 7.7
T
Vancomycin + Cefepime (n=133) 23 (17.3%) 67.9 ± 83.9
M
Variable AKI Incidence Time to AKI from baseline, hours AKI onset from baseline Early (≤3 days) Late (>5 days) AKI severity Risk Injury Failure Renal replacement therapy Incidence Duration of therapy, days AKI recovery patterns Fast (<3 days) Intermediate (3-7 days) Slow (>7 days) No recovery during hospital stay Length of stay Hospital, days ICU, days Mortality
Vancomycin + Piperacillin/Tazobactam (n=200) 39 (19.5%) 54.6 ± 93.2
AC
CE
PT
ED
Data presented as mean ± SD or n (%). AKI = acute kidney injury; ICU = intensive care unit.
ACCEPTED MANUSCRIPT Table 3. Univariate and Multivariate Logistic Regression Analysis Predicting Acute Kidney Injury
AC
Multivariate 95% CI
OR
0.86
Reference 0.49-1.53
0.61
Reference 0.99
0.54-1.82
0.98
1.07
1.00-1.16
0.05
1.07
0.98-1.17
0.12
0.99 0.59 1.41
0.97-1.01 0.34-1.04 0.81-2.46
0.20 0.07 0.22
0.54
0.30-0.98
0.042
1.73
0.82-3.68
0.15
1.31
0.64-2.66
2.70 1.10
1.53-4.75 1.02-1.19
3.22 1.06
1.66-6.26 0.95-1.17
0.01 0.31
0.34 1.14 2.18 4.93
Reference 0.14-0.85 0.37-3.47 0.45-10.60 0.83-29.43
0.02 0.82 0.34 0.08
p value
US
IP
T
p value
CR
OR
AN
0.46
ED
M
0.01 0.01
0.69 2.40 3.60 9.61
Reference 0.30-1.55 0.85-6.77 0.78-16.71 1.70-54.23
0.37 0.10 0.10 0.01
1.02 0.88 0.48 1.31 1.68 0.93 2.60
0.48-2.16 0.47-1.67 0.06-3.84 0.73-2.35 0.83-3.39 0.534-1.65 0.74-9.18
0.96 0.70 0.49 0.37 0.15 0.80 0.14
PT
CE
Variable Antimicrobial therapy Combination group Vancomycin + PTZ Vancomycin + cefepime Duration of concomitant therapy (per day increase) Patient characteristics Age Male gender Mechanical ventilation at baseline MAP <60 mmHg at baseline Serum lactate ≥4 mmol/L at baseline Shock (any type) SOFA score (each incremental 1 point score increase) Number of vasopressor agents at baseline None 1 2 3 4 Concomitant nephrotoxic agents Aminoglycoside ACEI/ARB Calcineurin Inhibitor Diuretic NSAID Contrast Acyclovir (intravenous)
Univariate 95% CI
ACEI = angiotensin-converting-enzyme inhibitor; ARB = angiotensin-II receptor blocker; CI = confidence interval MAP = mean atrial pressure; NSAID = nonsteroidal anti-inflammatory drug; OR = odds ratio; SOFA = Sequential Organ Failure Assessment.
ACCEPTED MANUSCRIPT Highlights
The addition of piperacillin/tazobactram to vancomycin regimens in critically ill patients may not increase the risk of acute kidney injury
Shock was identified as an independent risk factor for acute kidney injury, while
IP
More patients required renal replacement therapy with concomitant
CE
PT
ED
M
AN
US
CR
piperacillin/tazobactram and vancomycin
AC
T
antimicrobial combinations as well as other nephrotoxic agents were not
Figure 1
Mitchell S. Buckley, Nicole C. Hartsock, Andrew J. Berry, Dale S. Bikin, Emily C. Richards, Melanie J. Yerondopoulos, Emir Kobic, Laura M. Wicks, Drayton A. Hammond PII: DOI: Reference:
S0883-9441(18)31024-4 doi:10.1016/j.jcrc.2018.08.007 YJCRC 52999
To appear in:
Journal of Critical Care
Please cite this article as: Mitchell S. Buckley, Nicole C. Hartsock, Andrew J. Berry, Dale S. Bikin, Emily C. Richards, Melanie J. Yerondopoulos, Emir Kobic, Laura M. Wicks, Drayton A. Hammond , Comparison of acute kidney injury risk associated with vancomycin and concomitant piperacillin/tazobactam or cefepime in the intensive care unit. Yjcrc (2018), doi:10.1016/j.jcrc.2018.08.007
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Comparison of Acute Kidney Injury Risk Associated with Vancomycin and Concomitant Piperacillin/Tazobactam or Cefepime in the Intensive Care Unit Running Head: Renal injury with vancomycin and beta-lactams
Mitchell S. Buckley, PharmD, FCCP, FASHP, FCCM, BCCCP,1 Nicole C. Hartsock, PharmD, MBA,1
IP
T
Andrew J. Berry, PharmD, BCPS,1 Dale S. Bikin, PharmD,1 Emily C. Richards, PharmD, BCPS,1 Melanie J. Yerondopoulos, PharmD, BCCCP,1 Emir Kobic, PharmD,1 Laura M. Wicks, PharmD, BCPS1;
US
CR
and Drayton A. Hammond, PharmD, MBA, BCPS, BCCCP2
Department of Pharmacy, Banner – University Medical Center Phoenix, Phoenix, AZ
1
2
AN
Department of Pharmacy, Rush University Medical Center, Chicago, IL
M
Keywords: acute kidney injury, cefepime, critical care, intensive care unit, nephrotoxicity, piperacillin
Abstract Word Count: 199
ED
tazobactam, vancomycin
PT
Word Count (excluding abstract, tables, references, figures): 3812
CE
Acknowledgements: The authors would like to thank Dr. Sumit K. Agarwal, MBBS, MBA, and Dr. Richard D. Gerkin, MD, at Banner University Medical Center Phoenix for assistance in data
AC
acquisition and statistical analysis. Financial support: No financial and material support was provided for this research. Conflict of interest: No conflicts of interest are reported by the authors pertaining to this article. Poster Presentation: This original research was submitted as an abstract for a poster presentation at the Society of Critical Care Medicine Annual Congress in February 2019 Correspondence:
ACCEPTED MANUSCRIPT Mitchell S. Buckley, PharmD, FASHP, FCCM, FCCP, BCCCP, Clinical Pharmacy Specialist, Banner – University Medical Center Phoenix, Department of Pharmacy, 1111 E. McDowell Road, Phoenix, AZ 85006, email: [email protected]
Abstract
T
Purpose
CR
compared to vancomycin and cefepime (FEP) in critically ill patients.
IP
The objective of this study was to evaluate AKI incidence with concomitant vancomycin and PTZ
US
Materials and Methods
AN
A retrospective, cohort study was conducted in adult critically ill patients from January 1, 2014 to December 31, 2017. The primary aim was to compare the incidence of AKI during concomitant therapy
M
or until hospital discharge. Secondary analyses included AKI severity, time to AKI as well as recovery,
ED
and clinical outcomes.
PT
Results
Overall, 333 patients were evaluated. The AKI rate in the vancomycin/PTZ group and
CE
vancomycin/FEP group were similar (19.5% vs. 17.3%, respectively, p=0.612). Renal replacement therapy was initiated in 10.0% and 3.8% administered vancomycin/PTZ and vancomycin/FEP groups,
AC
respectively (p=0.04). Multivariate regression found vancomycin/PTZ was not associated with an increased risk of developing AKI although the presence of shock was identified as an independent risk factor (odds ratio, 3.22; 95% CI, 1.66-6.26). No significant differences in hospital or ICU length of stay or in-hospital mortality were observed between study groups.
Conclusions Concomitant PTZ and vancomycin in ICU patients was not associated with an increased risk of
ACCEPTED MANUSCRIPT developing AKI compared to FEP and vancomycin combinations. More patients administered vancomycin/PTZ received RRT.
Introduction Acute kidney injury (AKI) is a common complication experienced in critically ill patients and
T
associated with deleterious outcomes. The incidence of AKI in the intensive care unit (ICU) is about 30-
IP
40% with some reports suggesting rates as high as 60% [1-4]. Sepsis, septic shock, and multiple organ
CR
dysfunction syndrome are most frequently implicated as the cause of AKI in the ICU.[3, 5, 6] Several risk factors have been identified including advanced age, severity of illness, higher baseline serum creatinine
US
(SCr), and nephrotoxic drug exposure [1]. Mortality, ICU length of stay, and mechanical ventilation
AN
duration are increased in ICU patients who develop AKI [4, 7]. The mortality risk incrementally increases with the severity of renal failure and death rates can exceed 50% in patients receiving renal replacement
M
therapy (RRT) [4, 8]. Adverse long-term outcomes include the development of chronic kidney disease as well as end-stage renal failure, higher risk of cardiovascular events, and reduced quality of life [8]. The
ED
economic impact is substantial with an additional cost of $28,000 to $56,035 per ICU patient with AKI
PT
requiring RRT compared to those without dialysis [4, 8]. Empiric antimicrobial therapy consisting of vancomycin with a broad-spectrum beta-lactam for
CE
methicillin-resistant Staphylococcus aureus and Pseudomonas aeruginosa coverage, respectively, are frequently utilized in ICU patients [9]. Vancomycin-associated nephrotoxicity has been widely
AC
recognized over the past several decades, with rates ranging from 5% to 43% depending on risk factors.[10] Increased vancomycin exposure (serum trough concentrations >15 mcg/mL, total daily doses >4 g, and area under the curve between 700-1300 mg/hr/L) may raise the risk of nephrotoxicity, especially in ICU patients [10]. Also, piperacillin-tazobactam (PTZ) has been associated with delayed AKI recovery in the critically ill, irrespective of vancomycin and other beta-lactams [11]. Interestingly, rapid reversal of AKI has been observed upon discontinuation of PTZ, while this fast renal recovery phenomenon was not found with other beta-lactams [11].
ACCEPTED MANUSCRIPT Recently, numerous reports evaluating the risk of AKI with concomitant vancomycin and PTZ have been published [12-22]. The combination of vancomycin with PTZ has been associated with higher AKI rates compared to either agent alone [21-24]. Several studies have also demonstrated a significant risk of nephrotoxicity with concomitant vancomycin and PTZ therapy compared to vancomycin with
T
cefepime (FEP) [12, 13, 15-18, 20, 25]. However, these findings were inconsistent, particularly in
IP
critically ill patients [14, 16]. Limitations of previously published trials include significant heterogeneity
CR
of the study patient population, AKI definition (creatinine-based diagnostic criteria without UOP), and inclusion criteria [12, 13, 15-18, 20, 25]. The incremental risk of AKI with PTZ over FEP, in conjunction
US
with vancomycin, remains controversial given these discordant results. Therefore, the purpose of this study was to evaluate the incidence of AKI with concomitant vancomycin and PTZ compared to
AN
vancomycin and FEP in ICU patients.
M
Materials and Methods
ED
Patients and Study Design
This was a retrospective, cohort study conducted at a three large (>500 beds) hospitals within the
PT
same health-system (two community and one university hospital in Phoenix, AZ). Patients concomitantly administered PTZ or FEP with vancomycin during their ICU admission were evaluated over a 4-year
CE
period (January 1, 2014 to December 31, 2017). Inclusion criteria consisted of the following: (1) ≥18
AC
years of age; (2) initiation of concomitant therapy was ≤48 hours after ICU admission; (3) ≥48 consecutive hours of combination therapy during the patient’s ICU stay; (4) baseline creatinine clearance (CrCl) >30 mL/min; and (5) at least one SCr ≤24 hours of ICU admission. Patients with combination therapy (i.e., either PTZ or FEP with intravenous vancomycin) were included if the beta-lactam was administered within 48 hours of vancomycin irrespective of the antibiotic initiated first. Patients were excluded if they had any of the following: (1) vancomycin/PTZ or vancomycin/FEP combinations were initiated in hospitalized patients prior to ICU admission; (2) both PTZ and FEP were administered during
ACCEPTED MANUSCRIPT the same ICU admission; (3) end-stage renal disease; (4) any RRT prior to combination therapy; or (5) combination therapy discontinued and restarted within 7 days. Clinical pharmacist management of renal dosed antibiotics and/or requiring therapeutic drug monitoring (vancomycin and aminoglycosides) was standardized through the use of approved protocols at
T
the study institutions. Vancomycin loading doses were incorporated upon commencement of therapy
IP
using a weight-based approach (20-30 mg/kg) of actual body weight (ABW) rounded to the nearest 250
CR
mg increment up to a maximum of 2000 mg. Vancomycin maintenance dosing regimens were initiated by the pharmacist based on the patients ABW and renal function (CrCl or mode of RRT). Clinical
US
pharmacists were expected to obtain an initial vancomycin serum trough concentration no later than prior to the fifth maintenance dose with a goal target concentration of 10-20 mcg/mL based on the indication
AN
(e.g., 15-20 mcg/mL for pneumonia). However, the clinical pharmacists could obtain vancomycin concentrations at an earlier time, including post-loading dose, in patients with severely impaired or
M
rapidly fluctuating renal function as well as obesity. Follow-up vancomycin serum concentrations were
ED
recommended within seven days of a previous serum concentration for patients receiving longer durations of therapy. However, the pharmacist could order a serum concentration at any time prior to the 7-day
PT
limit to monitor therapy based on the clinical scenario (e.g., changing renal function). Furthermore, the
CE
clinical pharmacist dosed FEP and PTZ based on renal function or RRT pursuant to a standardized protocol. The study sites implemented extended-interval dosing for PTZ prior to the study period. The
AC
standard of practice for PTZ dosing involves 4.5 g loading dose to be administered to patients over 30 minutes upon initiation of therapy followed by subsequent doses of 3.375 g administered over 4-hour intervals. The frequency of dosing is every 8 hours in patients with CrCl ≥20 mL/min or receiving continuous RRT (CRRT). Patients with CrCl <20 mL/min, intermittent hemodialysis, or sustained lowefficiency dialysis were dosed every 12 hours. Cefepime was also renally dosed in patients with CrCl <60 mL/min with multiple acceptable doses and frequencies ranging from 500 mg daily to 2 g every 8 hours. The dosing strategy was customized for each patient, taking into account several variables including renal
ACCEPTED MANUSCRIPT function, location of infection, patient weight, and severity of illness. Also, the only mode of CRRT adopted by the study sites was continuous veno-venous hemodiafiltration. Clinical pharmacists reviewed appropriate vancomycin, PTZ, and FEP dosing at least once daily.
T
Data Collection
IP
After Institutional Review Board approval, patients meeting study criteria were identified through
CR
the electronic medical record database. Patients were assigned to one of two cohorts based upon administration of PTZ or FEP. Patient demographic and pertinent clinical data collected were age, gender,
US
height, actual and ideal body weight, comorbidities, SCr, known or suspected locations of infection, RRT requirements (hemodialysis, CRRT, or sustained low efficiency dialysis), and concomitant nephrotoxic
AN
agents. The potential concomitant nephrotoxic agents evaluated were aminoglycosides, amphotericin B (conventional or lipid formulation), angiotensin converting enzyme inhibitors, angiotensin II receptor
M
blockers, calcineurin inhibitors, colistin, contrast dye, loop diuretics, foscarnet, nonsteroidal anti-
ED
inflammatory drugs, and acyclovir. Aerosolized antimicrobials (e.g., colistin, tobramycin, and amphotericin B) were not considered as potentially nephrotoxic exposures. Vancomycin, PTZ, and FEP
PT
doses administered as well as duration of therapy were collected. In addition, vancomycin serum
CE
concentrations and timing in relation to administration were assessed. Serum creatinine was evaluated using an estimated CrCl via the Cockroft-Gault formula [26]. The highest SCr value reported for each
AC
patient day in the medical record was used to calculate CrCl. Other non-drug related risk factors possibly associated with AKI in the ICU setting were also collected, such as the presence of shock, need for mechanical ventilation, end-stage liver disease, and chronic kidney disease [5].
Definitions Acute kidney injury was defined using the Risk, Injury, Failure, Loss, End Stage Renal Disease (RIFLE) criteria.[27] The RIFLE criteria defines Risk, Injury, and Failure for AKI as follows: SCr
ACCEPTED MANUSCRIPT increase of 1.5 times baseline, glomerular filtration rate (GFR) reduction >25%, or UOP <0.5mL/kg/hour by 6 hours; SCr increase of 2 times baseline, GFR decrease >50%, or UOP <0.5mL/kg/hour by 12 hours; and SCr increase of 3 times baseline, GFR reduction >75%, UOP <0.3mL/kg/hour by 24 hours or anuria over 12 hours, respectively.[27] Each patient’s GFR was estimated using CrCl. Renal function was
T
assessed daily using SCr, CrCl, and UOP while the patient was in the ICU. Patients experiencing AKI
IP
were categorized using the RIFLE (Risk, Injury, or Failure) criteria. The worst-case scenario for each patient’s UOP or SCr were used to determine the RIFLE stage to assign the degree of AKI. For example,
CR
if a patient’s UOP aligns with the “injury” stage of the criteria, while the SCr suggests “risk,” then the
US
patient was classified as “injury” since it is the more severe AKI classification. Patients initiated on combination therapy in the ICU and subsequently transferred to the general ward were only assessed by
AN
GFR or SCr for the RIFLE classification criteria since hourly UOP was not typically measured and reported in general ward patients at the study institutions. Recovery of AKI was defined as a return of SCr
M
to <0.3 mg/dL above baseline <3 days (“fast”), 3-7 days (“intermediate”), and > 7 days (“slow”) based on
ED
a modified version of a previously published report [28]. Other definitions used in this study were “baseline” and “shock.” “Baseline” was defined as the
PT
first patient day on which vancomycin with FEP or vancomycin with PTZ were concomitantly
CE
administered. For instances in which the second adjunctive antimicrobial agent was not initiated on the same day, baseline was considered as the time when both were administered within the same day
AC
irrespective of which agent was started first. The presence of “shock” was determined as the use of continuous infusion vasoactive support (norepinephrine, dopamine, epinephrine, phenylephrine, vasopressin) for at least 1 hour, mean arterial pressure <60 mm Hg for at least 1 hour, or serum lactate concentration ≥4 mmol/L.
Study Outcomes and Statistical Analysis The primary endpoint was the incidence of AKI during concomitant antimicrobial therapy and up
ACCEPTED MANUSCRIPT to hospital discharge. Secondary analyses were AKI severity rates, days to AKI recovery, rates of requiring RRT, and duration of RRT. Clinical outcomes including in-hospital mortality and hospital and ICU length of stay were also compared. A sample size of 300 patients (200 and 100 in the vancomycin/PTZ and vancomycin/FEP groups, respectively) was needed to achieve 80% power in
T
detecting ≥15% increase in the incidence of AKI associated with vancomycin and PTZ compared to
IP
vancomycin and FEP assuming AKI rates of 30% and 15%, respectively. These estimates were based on
CR
previously published reports in ICU patients [14, 16, 25]. Patients from a convenience sample in the vancomycin and PTZ group were consecutively selected from the most recent administration date until
US
the minimum number of patients needed to meet the required sample size were included. Fisher’s exact test was used for categorical variables and Student’s t-test for continuous data. Variables potentially
AN
associated with AKI were further evaluated using a stepwise multivariate regression analysis to determine independent risk factors significantly associated with AKI. Although a p-value cut-off point of 0.1 was
M
established, antimicrobial combinations and vasoactive agents were included based on investigators’
ED
preference. A p-value of <0.05 was considered statistically significant. Statistical analyses were
PT
performed using IBM SPSS Statistics for Mac version 24.0 (IBM, Armonk, NY, 10504).
CE
Results
A total of 3861 patients were identified through the electronic medical record during the study
AC
period and screened for inclusion. Overall, 333 patients met inclusion criteria and were included in the analysis with 200 (66.7%) patients in the vancomycin/PTZ group and 133 (33.3%) patients in the vancomycin/FEP group (Figure 1). The baseline characteristics between study groups were similar except for a few comorbid conditions, including hepatic dysfunction, chronic obstructive pulmonary disease, and hypertension (Table 1). Baseline severities of illness were similar between groups with respect to SOFA score, invasive mechanical ventilation rates, and incidence of patients requiring vasoactive support (Table 1). The vast majority (92.0%) of all study patients received ≥1 concomitant nephrotoxic agents without
ACCEPTED MANUSCRIPT any differences between study cohorts for any specific nephrotoxin (Table 1). No significant differences in the duration of concomitant therapy were found between vancomycin/PTZ (5.1 ± 3.8 days) and vancomycin/FEP (5.8 ± 5.2 day) (p=0.185). The mean total daily dose of vancomycin was similar in the PTZ and FEP groups (2200.3 ± 990.9 and 2271.9 ± 978.3 mg, respectively, p=0.517). The median initial
T
vancomycin serum concentration in the vancomycin/PTZ and vancomycin/FEP groups were 13.5 (10.5-
IP
18.6) and 13.1 (9.8-17.7) mcg/mL, respectively, p=0.369. Furthermore, the mean beta-lactam dose
CR
administered in the FEP cohort was 4.0 ± 1.6 g/day, while the PTZ cohort was 11 ± 1.6 g/day. The overall rate of AKI observed among all study patients was 18.6% (n=62). The AKI rate in the
US
vancomycin/PTZ group compared to the vancomycin/FEP group was not significantly different (19.5% vs. 17.3%, respectively, p=0.612). Although no difference was found in the mean time to AKI between
AN
study groups, significantly more patients receiving vancomycin/PTZ compared to the vancomycin/FEP group experienced early AKI (≤3 days) after initiating concomitant antimicrobial therapy. However, the
M
peak incidence of AKI onset occurred ≤24 hours from baseline with 66.7% (n=26) in the
ED
vancomycin/PTZ group compared to 65.2% (n=15) (p=0.858). Overall, about 63% of all patients experiencing AKI progressed to a RIFLE classification of “Failure” without any differences in severity
PT
between study groups. Also, the rate of patients requiring RRT for AKI was two-fold higher with
CE
vancomycin/PTZ than vancomycin/FEP (p=0.035). Recovery patterns for AKI were not significantly different between study groups with most patients (88.7%) failing to recovery to baseline renal function.
(Table 2).
AC
Clinical outcomes including length of stay (overall hospitalization and ICU) and mortality were similar
The choice of beta-lactam in combination with vancomycin was not an independent risk factor for developing AKI (Table 3). Patients were more likely to develop AKI with the presence of any type of shock, higher SOFA score, or requiring at least four vasopressor agents at baseline in the univariate analysis. However, multiple variable regression analysis found shock (any type) was 3-times as likely to develop AKI, while significance was lost for SOFA score and four vasoactive medications. Concomitant
ACCEPTED MANUSCRIPT antimicrobial therapy duration trended towards a significantly increased risk of renal dysfunction. Multiple regression analysis also found lower odds of AKI in male gender patients and those individuals requiring only one catecholamine agent at baseline. None of the adjunctive nephrotoxic agents were
T
found to predict AKI.
IP
Discussion
CR
The aim of this study was to investigate the risk of developing AKI between two common broadspectrum combination antimicrobial strategies (vancomycin and PTZ vs. vancomycin and FEP) in the
US
ICU. To our knowledge, this was the first multicenter study in ICU patients to describe AKI characteristics (occurrence, onset, severity, RRT rates, and recovery patterns). Concomitant vancomycin
AN
and PTZ did not increase the risk of AKI in critically ill patients. Similar characteristics between the combination groups, including AKI onset, severity, and recovery patterns, were observed. Patients
M
developing AKI in this study may have largely been attributed to non-drug causes. The vast majority of
ED
AKI identified in both groups occurred about 24 hours after antimicrobial combination therapy initiation, which might be too soon for antibiotic exposure to suggest medication-induced causes. Furthermore,
PT
shock was identified as a risk factor for AKI on multivariate analysis, while no association was found
CE
with vancomycin/PTZ or other known nephrotoxic medications. Although the administration of RRT was more common in patients receiving vancomycin/PTZ over the FEP combination, this finding should be
AC
considered with caution due to the small number of patients in both study groups. The clinical implication of this study implies providers may not need to avoid concomitant vancomycin and PTZ merely based on concerns of increasing AKI risk in ICU patients. Our study failed to demonstrate an increased risk of AKI with vancomycin/PTZ compared to vancomycin/FEP, which is consistent with previously published reports in ICU patients [14, 16, 25]. Recent meta-analyses have reported significantly higher AKI rates associated with vancomycin/PTZ compared to vancomycin combinations with either FEP or carbapenem [16, 25]. However, a subgroup
ACCEPTED MANUSCRIPT analysis consisting of pooled data in critically ill patients found no significant difference in AKI rates between vancomycin/PTZ compared to vancomycin with either FEP or carbapenem (OR=1.32, 0.73-2.38) [16]. Hammond et al, conducted a study in an ICU population evaluating AKI rates and found no differences between vancomycin/PTZ (32.7%) and vancomycin/FEP groups (28.8%) (p=0.647) [14].
T
Several aspects of their study design (inclusion criteria, duration of the assessment period, etc) were
IP
similar to ours. Although both studies failed to demonstrate a significant difference in AKI rates between
CR
groups, a few dissimilarities may explain the lower rates observed in our study. Vancomycin dosing was primarily managed by clinical pharmacists at our study sites, while their institution was not reflective of a
US
formal pharmacy-managed service. Subsequently, the mean initial vancomycin serum trough concentration in both groups was lower in our study (<15 mcg/mL) compared to theirs (>15 mcg/mL).
AN
Vancomycin serum concentrations >15 mcg/mL have been associated as a risk factor for developing AKI [29]. Most of our ICU patients admitted during the study period failed to meet inclusion criteria due to
M
multisystem organ dysfunction including AKI. Therefore, the acuity of illness in our ICU population
ED
meeting study inclusion criteria may have been lower as evidenced by our SOFA scores, incidence of shock, proportion of patient requiring vasoactive support and invasive mechanical ventilation, mean
PT
length of stay, and mortality rates. Another important difference was the adopted definition of AKI
CE
utilized in our respective studies (AKIN vs. RIFLE), which may have resulted in inconsistencies for classifying the incidence AKI [30-33]. Although our study included patients administered extended-
AC
infusion PTZ rather than traditional intermittent administration, the beta-lactam infusion strategy has not been shown to be an independent risk factor for developing AKI [34, 35]. Several other studies have investigated the rate of AKI with vancomycin/PTZ compared to vancomycin/FEP in hospitalized patients with conflicting results. The overall incidence of AKI in these reports largely varied with >50% of these reports suggesting an association between increased AKI risk and vancomycin/PTZ utilization [12, 13, 15, 17, 18, 20]. Unfortunately, a direct comparison among these studies remains challenging given the significant heterogeneity of the study populations, AKI definition,
ACCEPTED MANUSCRIPT and inclusion criteria. Two studies included critically ill patients within their respective study cohorts, but did not separately report ICU data [13, 18]. Most of these studies relied solely on SCr (absolute value thresholds, magnitude of change, and/or CrCl estimates) as the primary method to define AKI [12, 13, 15, 17, 18, 20]. Therefore, it is important to recognize the limitation of creatinine-based diagnostic AKI
T
criteria, including delays in SCr concentration changes following the initial kidney damage, disease- and
IP
drug-induced fluctuations on SCr concentrations (production and/or clearance) independent of the
CR
underlying renal function, “baseline” SCr definitions, and the impact of fluid status (volume overload and dehydration) on SCr [6, 36]. These factors may explain the variability in the incidence of AKI among
US
these reports despite using validated criteria (AKIN, RIFLE, KDIGO) [13, 15, 17, 18, 20]. Although evaluating UOP may offer some advantages over SCr, UOP alone to classify AKI is not without
AN
limitations as well [36]. Our study utilized both SCr and UOP to identify AKI and stages of severity. Few studies evaluating vancomycin/PTZ and vancomycin/FEP have described AKI
M
characteristics and patterns of time to onset, severity, duration, need for RRT, and recovery patterns.
ED
Three studies assessed time to AKI, which all found an increased AKI risk with the vancomycin/PTZ combination [13, 18, 20]. Although one study did not show any difference in time to AKI between
PT
groups, two found AKI occurred more rapidly upon antibiotic initiation with the addition of PTZ (3-5
CE
days) compared to FEP (5-8 days) [16]. This is in contrast to our findings that AKI onset was not significantly different between groups, but equally occurred much earlier (<3 days) in both antimicrobial
AC
groups. A recent meta-analysis also found no significant difference in time to AKI between patients receiving vancomycin/PTZ versus vancomycin/FEP [16]. Furthermore, the severity stage associated with AKI in previous studies was different than our findings [13, 15, 18, 20]. We found no difference in AKI severity stages using RIFLE between study groups with over 50% of all AKI classified as “failure”. The most common AKI severity categories in other studies were “risk” and “stage I” for RIFLE and AKIN, respectively [13, 15, 18, 20]. Interestingly, none of the published studies found any significant difference between the rate of patients requiring RRT or the duration of AKI between study groups [13, 14, 18].
ACCEPTED MANUSCRIPT Although our study found no difference in duration of AKI between groups, we did find a significantly higher rate of RRT in patients administered vancomycin/PTZ. However, the higher rate of RRT in patients administered vancomycin/PTZ observed in our study did not exceed the estimated incidence of using RRT for AKI in an ICU setting [3]. Gomes et al, found a high proportion of patients overall failed
T
to fully recover renal function with vancomycin/PTZ (59.0%) and vancomycin/FEP (78.6%) during the
IP
evaluation period (p=0.190) [13]. We also found an unacceptably high rate exceeding 88% of all patients
CR
experiencing AKI (n=62) did not fully recover renal function back to baseline with rates similar between vancomycin/PTZ and vancomycin/FEP (92.3% and 82.6%, respectively). It is important to emphasize the
US
definition of recovery in our study, which was SCr returning ≤0.3 mg/dL of baseline. It is possible patients experienced functional recovery of their kidneys not reflective of their SCr during or after the
AN
evaluation period.
This study has several limitations. First, the retrospective study design may have introduced
M
confounding variables leading to bias and spurious findings. However, groups were well balanced and
ED
multivariate analysis was performed to increase the robustness as well as corroborate our findings. Also, the small sample size may not have been sufficient to observe any study group differences. The rate of
PT
AKI observed in both study groups was lower than anticipated for our sample size calculation a priori. Nonetheless, it is equally important to recognize the limitations of power analysis utility in retrospective
CE
studies. Furthermore, the rigorous inclusion and exclusion criteria may limit this study’s generalizability
AC
for higher acuity ICU patients or those with baseline CrCl ≤30 mL/min. Unfortunately, despite our best efforts to exclude patients experiencing AKI prior to initiation of antimicrobials, we acknowledge some of these patients may have been included in our analysis. The initial kidney insult may have preceded any manifestations in SCr and/or UOP changes upon evaluation to meet our exclusion criteria. Lastly, accurately identifying drug-induced AKI is challenging in the critically ill considering potential etiologies are likely complex and multifactorial.
ACCEPTED MANUSCRIPT Conclusions Critically ill patients who received concomitant vancomycin/PTZ therapy were not at an increased risk of AKI compared to vancomycin/FEP. Irrespective of the beta-lactam choice with vancomycin, this study found no differences in the AKI incidence, severity, time to onset, duration, or
T
recovery patterns. However, RRT was initiated more frequently in patients administered
IP
vancomycin/PTZ. Although the association between concomitant vancomycin/PTZ and increased AKI
not be warranted in critically ill patients.
Cartin-Ceba R, Kashiouris M, Plataki M, Kor DJ, Gajic O, Casey ET. Risk factors for
AN
[1]
US
References
CR
risk remains debatable, these findings suggest a change in clinical practice to avoid this combination may
development of acute kidney injury in critically ill patients: a systematic review and meta-
[2]
M
analysis of observational studies. Crit Care Res Pract 2012;2012:691013. Srisawat N, Sileanu FE, Murugan R, Bellomod R, Calzavacca P, Cartin-Ceba R, et al. Variation
[3]
PT
Nephrol 2015;41(1):81-8.
ED
in risk and mortality of acute kidney injury in critically ill patients: a multicenter study. Am J
Hoste EA, Bagshaw SM, Bellomo R, Cely CM, Colman R, Cruz DN, et al. Epidemiology of
CE
acute kidney injury in critically ill patients: the multinational AKI-EPI study. Intensive Care Med 2015;41(8):1411-23.
Hoste EA, Schurgers M. Epidemiology of acute kidney injury: how big is the problem? Crit Care
AC
[4]
Med 2008;36(4 Suppl):S146-51. [5]
Uchino S, Kellum JA, Bellomo R, Doig GS, Morimatsu H, Morgera S, et al. Acute renal failure in critically ill patients: a multinational, multicenter study. Jama 2005;294(7):813-8.
[6]
Ostermann M, Chang RW. Acute kidney injury in the intensive care unit according to RIFLE. Crit Care Med 2007;35(8):1837-43; quiz 52.
[7]
Vieira JM, Jr., Castro I, Curvello-Neto A, Demarzo S, Caruso P, Pastore L, Jr., et al. Effect of
ACCEPTED MANUSCRIPT acute kidney injury on weaning from mechanical ventilation in critically ill patients. Crit Care Med 2007;35(1):184-91. [8]
Silver SA, Chertow GM. The Economic Consequences of Acute Kidney Injury. Nephron 2017;137(4):297-301. Kalil AC, Metersky ML, Klompas M, Muscedere J, Sweeney DA, Palmer LB, et al. Management
T
[9]
IP
of Adults With Hospital-acquired and Ventilator-associated Pneumonia: 2016 Clinical Practice
CR
Guidelines by the Infectious Diseases Society of America and the American Thoracic Society. Clin Infect Dis 2016;63(5):e61-e111.
Carreno JJ, Kenney RM, Lomaestro B. Vancomycin-associated renal dysfunction: where are we now? Pharmacotherapy 2014;34(12):1259-68.
Jensen JU, Hein L, Lundgren B, Bestle MH, Mohr T, Andersen MH, et al. Kidney failure related
AN
[11]
US
[10]
to broad-spectrum antibiotics in critically ill patients: secondary end point results from a 1200
Davies SW, Efird JT, Guidry CA, Dietch ZC, Willis RN, Shah PM, et al. Top Guns: The
ED
[12]
M
patient randomised trial. BMJ Open 2012;2(2):e000635.
"Maverick" and "Goose" of Empiric Therapy. Surg Infect (Larchmt) 2016;17(1):38-47. Gomes DM, Smotherman C, Birch A, Dupree L, Della Vecchia BJ, Kraemer DF, et al.
PT
[13]
CE
Comparison of acute kidney injury during treatment with vancomycin in combination with piperacillin-tazobactam or cefepime. Pharmacotherapy 2014;34(7):662-9. Hammond DA, Smith MN, Painter JT, Meena NK, Lusardi K. Comparative Incidence of Acute
AC
[14]
Kidney Injury in Critically Ill Patients Receiving Vancomycin with Concomitant PiperacillinTazobactam or Cefepime: A Retrospective Cohort Study. Pharmacotherapy 2016;36(5):463-71. [15]
Jeon N, Staley B, Klinker KP, Hincapie Castillo J, Winterstein AG. Acute kidney injury risk associated with piperacillin/tazobactam compared with cefepime during vancomycin therapy in hospitalised patients: a cohort study stratified by baseline kidney function. Int J Antimicrob Agents 2017;50(1):63-7.
ACCEPTED MANUSCRIPT [16]
Luther MK, Timbrook TT, Caffrey AR, Dosa D, Lodise TP, LaPlante KL. Vancomycin Plus Piperacillin-Tazobactam and Acute Kidney Injury in Adults: A Systematic Review and MetaAnalysis. Crit Care Med 2018;46(1):12-20.
[17]
Moenster RP, Linneman TW, Finnegan PM, Hand S, Thomas Z, McDonald JR. Acute renal
T
failure associated with vancomycin and beta-lactams for the treatment of osteomyelitis in
IP
diabetics: piperacillin-tazobactam as compared with cefepime. Clin Microbiol Infect
[18]
CR
2014;20(6):O384-9.
Navalkele B, Pogue JM, Karino S, Nishan B, Salim M, Solanki S, et al. Risk of Acute Kidney
US
Injury in Patients on Concomitant Vancomycin and Piperacillin-Tazobactam Compared to Those on Vancomycin and Cefepime. Clin Infect Dis 2017;64(2):116-23. Peyko V, Smalley S, Cohen H. Prospective Comparison of Acute Kidney Injury During
AN
[19]
Treatment With the Combination of Piperacillin-Tazobactam and Vancomycin Versus the
Rutter WC, Cox JN, Martin CA, Burgess DR, Burgess DS. Nephrotoxicity during Vancomycin
ED
[20]
M
Combination of Cefepime or Meropenem and Vancomycin. J Pharm Pract 2017;30(2):209-13.
Therapy in Combination with Piperacillin-Tazobactam or Cefepime. Antimicrob Agents
Burgess LD, Drew RH. Comparison of the incidence of vancomycin-induced nephrotoxicity in
CE
[21]
PT
Chemother 2017;61(2).
hospitalized patients with and without concomitant piperacillin-tazobactam. Pharmacotherapy
[22]
AC
2014;34(7):670-6.
Rutter WC, Burgess DR, Talbert JC, Burgess DS. Acute kidney injury in patients treated with vancomycin and piperacillin-tazobactam: A retrospective cohort analysis. J Hosp Med 2017;12(2):77-82.
[23]
McQueen KE, Clark DW. Does Combination Therapy With Vancomycin and PiperacillinTazobactam Increase the Risk of Nephrotoxicity Versus Vancomycin Alone in Pediatric Patients? J Pediatr Pharmacol Ther 2016;21(4):332-8.
ACCEPTED MANUSCRIPT [24]
Kim T, Kandiah S, Patel M, Rab S, Wong J, Xue W, et al. Risk factors for kidney injury during vancomycin and piperacillin/tazobactam administration, including increased odds of injury with combination therapy. BMC Res Notes 2015;8:579.
[25]
Hammond DA, Smith MN, Li C, Hayes SM, Lusardi K, Bookstaver PB. Systematic Review and
T
Meta-Analysis of Acute Kidney Injury Associated with Concomitant Vancomycin and
Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron
CR
[26]
IP
Piperacillin/tazobactam. Clin Infect Dis 2017;64(5):666-74.
1976;16(1):31-41.
Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P, Acute Dialysis Quality Initiative w.
US
[27]
Acute renal failure - definition, outcome measures, animal models, fluid therapy and information
AN
technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 2004;8(4):R204-12. Heung M, Steffick DE, Zivin K, Gillespie BW, Banerjee T, Hsu CY, et al. Acute Kidney Injury
M
[28]
ED
Recovery Pattern and Subsequent Risk of CKD: An Analysis of Veterans Health Administration Data. Am J Kidney Dis 2016;67(5):742-52. van Hal SJ, Paterson DL, Lodise TP. Systematic review and meta-analysis of vancomycin-
PT
[29]
CE
induced nephrotoxicity associated with dosing schedules that maintain troughs between 15 and 20 milligrams per liter. Antimicrob Agents Chemother 2013;57(2):734-44. Bagshaw SM, George C, Bellomo R, Committe ADM. A comparison of the RIFLE and AKIN
AC
[30]
criteria for acute kidney injury in critically ill patients. Nephrol Dial Transplant 2008;23(5):156974. [31]
Chang CH, Lin CY, Tian YC, Jenq CC, Chang MY, Chen YC, et al. Acute kidney injury classification: comparison of AKIN and RIFLE criteria. Shock 2010;33(3):247-52.
[32]
Lopes JA, Fernandes P, Jorge S, Goncalves S, Alvarez A, Costa e Silva Z, et al. Acute kidney injury in intensive care unit patients: a comparison between the RIFLE and the Acute Kidney
ACCEPTED MANUSCRIPT Injury Network classifications. Crit Care 2008;12(4):R110. [33]
Pereira M, Rodrigues N, Godinho I, Gameiro J, Neves M, Gouveia J, et al. Acute kidney injury in patients with severe sepsis or septic shock: a comparison between the 'Risk, Injury, Failure, Loss of kidney function, End-stage kidney disease' (RIFLE), Acute Kidney Injury Network (AKIN)
T
and Kidney Disease: Improving Global Outcomes (KDIGO) classifications. Clin Kidney J
Cotner SE, Rutter WC, Burgess DR, Wallace KL, Martin CA, Burgess DS. Influence of beta-
CR
[34]
IP
2017;10(3):332-40.
Lactam Infusion Strategy on Acute Kidney Injury. Antimicrob Agents Chemother 2017;61(10). Karino S, Kaye KS, Navalkele B, Nishan B, Salim M, Solanki S, et al. Epidemiology of Acute
US
[35]
Kidney Injury among Patients Receiving Concomitant Vancomycin and Piperacillin-Tazobactam:
AN
Opportunities for Antimicrobial Stewardship. Antimicrob Agents Chemother 2016;60(6):374350.
CE
PT
ED
Biochem Rev 2016;37(4):153-75.
M
Makris K, Spanou L. Acute Kidney Injury: Diagnostic Approaches and Controversies. Clin
AC
[36]
ACCEPTED MANUSCRIPT Table 1. Patient demographic and clinical characteristics
PT
CE
AC
p value 0.93 0.60 0.33 0.35 0.27 0.33 0.10 0.83 0.44
42 (31.6%) 4 (3.0%) 18 (13.5%)
0.86 0.30 0.07
IP
T
Vancomycin + Cefepime (n=133) 57.1 ± 15.6 84 (63.2%) 4.9 ± 3.1 169.7 ± 10.5 86.2 ± 33.9 62.5 ± 9.7 46 (34.6%) 17 (12.8%) 19 (14.3%)
CR
US
65 (32.5%) 12 (6.0%) 15 (7.5%)
0.06 100 (75.2%) 26 (19.6%) 6 (4.5%) 1 (0.8%) 0 (0.0%)
68 (34.0%) 51 (25.5%) 69 (34.5%) 41 (20.5%) 40 (20.0%) 1 (0.5%) 6 (3.0%) 24 (12.0%)
49 (36.8%) 32 (24.1%) 64 (48.1%) 40 (30.1%) 13 (9.8%) 1 (0.8%) 7 (5.3%) 25 (18.8%)
0.60 0.77 0.01 0.05 0.01 1.00 0.39 0.09
26 (13.0%) 2 (1.0%) 55 (27.5%) 4 (2.0%) 0 (0.0%) 124 (62.0%) 0 (0.0%) 29 (14.5%) 138 (69.0%) 7 (3.5%)
27 (20.3%) 4 (3.0%) 32 (24.1%) 6 (4.5%) 0 (0.0%) 85 (63.9%) 0 (0.0%) 21 (15.8%) 76 (57.1%) 4 (3.0%)
0.07 0.22 0.48 0.21 N/A 0.72 N/A 0.75 0.03 1.00
AN
138 (69.0%) 38 (19.0%) 12 (6.0%) 6 (3.0%) 6 (3.0%)
M
ED
Variable Age, years Male gender SOFA score Height, cm TBW, kg IBW, kg Mechanical ventilation at baseline MAP <60 mmHg at baseline Serum lactate ≥4 mmol/L at baseline Type of shock Vasodilatory Cardiogenic Other Number of vasopressor agents at baseline None 1 2 3 4 Comorbidities Diabetes mellitus Heart failure Hypertension COPD Hepatic dysfunction HIV/AIDS Transplant Cancer Concomitant nephrotoxic agents Aminoglycoside Amphotericin ACEI/ARB Calcineurin Inhibitor Colistin Diuretic Foscarnet NSAID Contrast Acyclovir (intravenous)
Vancomycin + Piperacillin/Tazobactam (n=200) 56.9 ± 16.9 132 (66.0%) 5.1 ± 3.8 179.6 ± 122.1 82.4 ± 28.0 71.8 ± 110.6 87 (43.5%) 24 (12.0%) 35 (17.5%)
Data presented as mean ± SD or n (%). AIDS = acquired immune deficiency syndrome; ACEI = angiotensin-converting-enzyme inhibitor; ARB = angiotensin-II receptor blocker; COPD =
ACCEPTED MANUSCRIPT chronic obstructive pulmonary disease; HIV = human immunodeficiency virus; IBW = ideal body weight; MAP = mean atrial pressure; NSAID = nonsteroidal anti-inflammatory drug; SOFA = Sequential Organ Failure Assessment; TBW = total body weight.
AC
CE
PT
ED
M
AN
US
CR
IP
T
:
ACCEPTED MANUSCRIPT Table 2. Clinical outcomes between study groups
34 (87.2%) 1 (2.6%)
18 (78.3%) 5 (21.7%)
8 (20.5%) 7 (17.9%) 24 (61.5%)
5 (21.7%) 3 (13.0%) 15 (65.2%)
IP CR
0.03 -0.88 ----
1 (2.6%) 1 (2.6%) 1 (2.6%) 36 (92.3%)
3 (13.0%) 0 (0.0%) 1 (4.3%) 19 (82.6%)
14.4 ± 8.8 9.1 ± 6.9 26 (13.0%)
15.6 ± 11.8 10.5 ± 10.2 20 (15.0%)
0.31 0.18 0.60
AN
5 (3.8%) 17.2 ± 9.7
p value 0.61 0.65
0.04 0.10 0.34 -----
US
20 (10.0%) 10.2 ± 7.7
T
Vancomycin + Cefepime (n=133) 23 (17.3%) 67.9 ± 83.9
M
Variable AKI Incidence Time to AKI from baseline, hours AKI onset from baseline Early (≤3 days) Late (>5 days) AKI severity Risk Injury Failure Renal replacement therapy Incidence Duration of therapy, days AKI recovery patterns Fast (<3 days) Intermediate (3-7 days) Slow (>7 days) No recovery during hospital stay Length of stay Hospital, days ICU, days Mortality
Vancomycin + Piperacillin/Tazobactam (n=200) 39 (19.5%) 54.6 ± 93.2
AC
CE
PT
ED
Data presented as mean ± SD or n (%). AKI = acute kidney injury; ICU = intensive care unit.
ACCEPTED MANUSCRIPT Table 3. Univariate and Multivariate Logistic Regression Analysis Predicting Acute Kidney Injury
AC
Multivariate 95% CI
OR
0.86
Reference 0.49-1.53
0.61
Reference 0.99
0.54-1.82
0.98
1.07
1.00-1.16
0.05
1.07
0.98-1.17
0.12
0.99 0.59 1.41
0.97-1.01 0.34-1.04 0.81-2.46
0.20 0.07 0.22
0.54
0.30-0.98
0.042
1.73
0.82-3.68
0.15
1.31
0.64-2.66
2.70 1.10
1.53-4.75 1.02-1.19
3.22 1.06
1.66-6.26 0.95-1.17
0.01 0.31
0.34 1.14 2.18 4.93
Reference 0.14-0.85 0.37-3.47 0.45-10.60 0.83-29.43
0.02 0.82 0.34 0.08
p value
US
IP
T
p value
CR
OR
AN
0.46
ED
M
0.01 0.01
0.69 2.40 3.60 9.61
Reference 0.30-1.55 0.85-6.77 0.78-16.71 1.70-54.23
0.37 0.10 0.10 0.01
1.02 0.88 0.48 1.31 1.68 0.93 2.60
0.48-2.16 0.47-1.67 0.06-3.84 0.73-2.35 0.83-3.39 0.534-1.65 0.74-9.18
0.96 0.70 0.49 0.37 0.15 0.80 0.14
PT
CE
Variable Antimicrobial therapy Combination group Vancomycin + PTZ Vancomycin + cefepime Duration of concomitant therapy (per day increase) Patient characteristics Age Male gender Mechanical ventilation at baseline MAP <60 mmHg at baseline Serum lactate ≥4 mmol/L at baseline Shock (any type) SOFA score (each incremental 1 point score increase) Number of vasopressor agents at baseline None 1 2 3 4 Concomitant nephrotoxic agents Aminoglycoside ACEI/ARB Calcineurin Inhibitor Diuretic NSAID Contrast Acyclovir (intravenous)
Univariate 95% CI
ACEI = angiotensin-converting-enzyme inhibitor; ARB = angiotensin-II receptor blocker; CI = confidence interval MAP = mean atrial pressure; NSAID = nonsteroidal anti-inflammatory drug; OR = odds ratio; SOFA = Sequential Organ Failure Assessment.
ACCEPTED MANUSCRIPT Highlights
The addition of piperacillin/tazobactram to vancomycin regimens in critically ill patients may not increase the risk of acute kidney injury
Shock was identified as an independent risk factor for acute kidney injury, while
IP
More patients required renal replacement therapy with concomitant
CE
PT
ED
M
AN
US
CR
piperacillin/tazobactram and vancomycin
AC
T
antimicrobial combinations as well as other nephrotoxic agents were not
Figure 1