Abstract
Purpose:
To determine whether there is an association between higher vancomycin trough concentrations and attainment of a calculated area under the concentration–time curve (AUC)/minimum inhibitory concentration (MIC) ≥400.
Methods:
A retrospective analysis was conducted among vancomycin-treated adult patients with a positive methicillin-resistant Staphylococcus aureus (MRSA) culture. Attainment of a calculated AUC/MIC ≥400 was compared between patients with troughs in the reference range of 15 to 20 mg/L and those with troughs in the following ranges: <10, 10 to 14.9, and >20 mg/L. Nephrotoxicity was assessed as a secondary outcome based on corrected average vancomycin troughs over 10 days of treatment.
Results:
Overall, 226 patients were reviewed and 100 included. Relative to troughs ≥10, patients with vancomycin troughs <10 mg/L were 73% less likely to attain an AUC/MIC ≥400 (odds ratio [OR] 0.27, 95% confidence interval [CI]: 0.01-0.75). No difference was found in the attainment of an AUC/MIC ≥400 in patients with troughs of 10 to 14.9 mg/L and >20 mg/L when compared to patients with troughs of 15 to 20 mg/L. The mean corrected average vancomycin trough was higher in patients developing nephrotoxicity compared to those who did not (19.5 vs 14.5 mg/L, P < .001).
Conclusion:
Achieving vancomycin serum trough concentrations of 15 to 20 mg/L did not result in an increased attainment of the AUC/MIC target relative to troughs of 10 to 14.9 mg/L but may increase nephrotoxicity risk.
Introduction
Vancomycin is currently the drug of choice for infections due to methicillin-resistant Staphylococcus aureus (MRSA) and is historically dosed based on weight with subsequent adjustments aimed at achieving serum trough concentrations between 10 and 20 mg/L. 1 Serum vancomycin trough concentrations (troughs) are used as a surrogate for achieving a ratio of area under the concentration–time curve (AUC) to minimum inhibitory concentration (MIC) of ≥ 400, the pharmacodynamic parameter that has been identified as the primary predictor of vancomycin effectiveness. 1,2 Moise-Broder and colleagues were among the first to associate vancomycin efficacy with maintaining an AUC/MIC ≥400. 2 In their retrospective study of 108 patients, they found that AUC/MIC values predicted time-related clinical and bacteriological success for patients with lower respiratory tract infections caused by MRSA. Numerous studies have supported these findings in other populations and different types of infections. 3 –7 Despite these data supporting AUC/MIC as the preferred target for vancomycin efficacy, consensus guidelines issued by the American Society of Health-System Pharmacists, Infectious Diseases Society of America, and the Society of Infectious Diseases Pharmacists recommend trough monitoring as a convenient surrogate for AUC/MIC, citing an increased probability of attaining an AUC/MIC ≥ 400 with vancomycin troughs in the range of 15 to 20 mg/L. 1,8 Troughs, however, represent a single point on the concentration–time profile and do not accurately predict the total concentration exposure over a 24-hour period, as is represented by the AUC. Additionally, the clinical benefits of maintaining higher vancomycin troughs are not well described, with only 2 studies to date demonstrating improved outcomes with vancomycin troughs between 15 and 20 mg/L. 3,9 Several studies have failed to show a link between maintaining higher trough concentrations in the range of 15 to 20 mg/L and clinical success. 10 –16 Despite this, the consensus guidelines recommend troughs within this range for complicated infections such as bacteremia, endocarditis, osteomyelitis, and hospital-acquired pneumonia caused by S. aureus. 1 No studies to date have assessed whether there is an association between achieving vancomycin trough concentrations within different ranges and attaining an AUC/MIC ≥ 400. Adding to this quandary is the multitude of data suggesting that higher vancomycin troughs are associated with an increased risk of developing acute kidney injury (AKI). 1,17 The primary objective of this study was to determine whether there is an association between different target ranges of vancomycin trough concentrations and attainment of a calculated AUC/MIC of 400 or greater.
Methods
Study Design and Patient Population
A retrospective chart analysis was conducted at a 472-bed, tertiary-care academic medical center and Level I Trauma center. The institutional review board granted exempt status for this study.
All adult patients with a positive methicillin-resistant Staphylococcus aureus (MRSA) culture obtained from any site between November 2013 and January 2015 were identified via microbiology laboratory reports. Patients were assessed consecutively in reverse chronological order until a sample size of 100 patients meeting inclusion criteria was reached. This sample size was determined based on similar sample sizes used in previous pharmacokinetic/pharmacodynamic studies of vancomycin. 2,5,7,12 Patients were included if a steady state vancomycin trough concentration was obtained. Steady state was considered to be achieved provided the trough was taken prior to the fourth dose or later in patients with stable renal function. Inappropriately timed trough concentrations, based on medication administration times or documentation by a clinical pharmacist at the time, were disregarded. Patients with unstable renal function prior to the steady state measurement were excluded, defined as a fluctuation in serum creatinine (SCr) of >0.2 mg/dL. Patients receiving intermittent vancomycin doses based on serum concentrations and those on hemodialysis were also excluded.
Outcome Analysis
The primary outcome was to determine whether there is an association between attaining a vancomycin trough within a specified range and reaching a calculated vancomycin AUC/MIC ≥ 400. Patients were stratified by their initial troughs in the following ranges: <10, 10 to 14.9, 15 to 20, and >20 mg/L. Attainment of the calculated target AUC/MIC≥ 400 was compared among the groups. Target attainment was first compared between patients with troughs <10 mg/L and those with troughs ≥10, as current recommendations suggest that troughs be maintained above 10 mg/L. 1,18 For subsequent comparisons, patients with vancomycin serum concentrations within the 15 to 20 mg/L range were used as the reference group, to which other groups were compared. The secondary outcome was to determine the corrected average vancomycin trough associated with the development of AKI.
Calculations and Definitions
Vancomycin AUC was calculated as the daily dose of vancomycin in milligrams divided by an estimate for vancomycin clearance. In this formula, vancomycin clearance was based on the following previously validated formula: (creatinine clearance [CrCl] × 0.79 + 15.4) × 0.06. 2,19 CrCl was estimated with the Cockcroft-Gault equation, using the SCr measurement on the date of vancomycin initiation. 20 No adjustments were made for SCr, as there is limited evidence to support the rounding of SCr in patients with low SCr. 21 For patients >130% of their ideal body weight (IBW), an adjusted body weight (ABW) equal to IBW + 0.4 (actual body weight − IBW) was used. 21,22 Actual body weight was used for patients less than their IBW, and IBW was used for normal-weight patients. Estimated CrCl (eCrCl) was capped at a clearance of 120 mL/min, as there are data supporting the capping of CrCl, and an inflated eCrCl would significantly impact the calculated AUC, resulting in an underestimation. 22 –25 The calculated AUC was related to the MIC of the patient-specific MRSA isolate. Vancomycin MIC was determined by VITEK-2, the standard at our institution.
For the secondary outcome, AKI was defined as either an increase in SCr by 0.5 mg/dL or a 50% increase from pretreatment levels on consecutive measurements. 26 . A corrected average vancomycin trough was calculated based on the following formula: ([trough 1 × number of days with trough 1] + [trough 2 × number of days with trough 2] + [trough n × number of days with trough n]) / (Total number of days on vancomycin). 12,27 This was calculated out to 10 days or the end of therapy, based on the premise that vancomycin-induced nephrotoxic events typically occur around days 4 to 8 of therapy. 17,28
Data Collection and Statistics
A standardized form was utilized to collect data including patient demographics, SCr measurements, vancomycin trough concentrations, initial vancomycin dosage, dosage adjustments made after the first trough was obtained, and duration of vancomycin therapy. Categorical variables were compared using a chi-square test for independence or Fisher exact test. Normality tests were performed, and continuous variables were evaluated using the Student t test for parametric data or Mann-Whitney U test for nonparametric data. All statistical tests were 2-tailed, and a P value <.05 was considered statistically significant. Calculations were performed using SPSS version 22.0 (SPSS, Inc, Chicago, Illinois).
Results
Patient inclusion data are depicted in Figure 1. Two hundred twenty six patients were identified as having a positive MRSA culture between November 2013 and January 2015, of which 100 met the inclusion criteria. Baseline demographics and vancomycin dosing characteristics for the population are shown in Table 1. A majority of MRSA isolates were obtained from wound and sputum cultures. Ninety-four percent of isolates had an MIC of 1 mg/L, and only 1 isolate had an MIC of 2.0 mg/L. Overall, 42 (42%) of patients met the AUC/MIC target of ≥400, including all isolates with an MIC ≤0.5 mg/L. The single patient with an isolate demonstrating an MIC of 2 mg/L did not meet this target (AUC/MIC = 162.5). As represented in Figure 2, having a trough <10 mg/L was associated with a 73% decreased likelihood of attaining a calculated AUC/MIC ≥ 400 relative to higher troughs (odds ratio [OR] 0.27, 95% confidence interval [CI]: 0.01-0.75). There were no significant differences at baseline between patients within these trough ranges for age, weight, SCr, vancomycin dosing, or eCrCl. Mean AUC (404.3 ± 117.1 vs 353.6 ± 72.4 mg × hr/L, P = .035) and median AUC/MIC (404.5 [IQR 305.5-509.5] vs 367 [IQR 303.5-384.5]; P = .041) were significantly greater when initial vancomycin trough concentrations were ≥10 mg/L.
Patient inclusion.
Baseline Demographics and Vancomycin Dosing Characteristics.a
Abbreviations: AUC, area under the concentration-time curve; MIC, minimum inhibitory concentration; SD, standard deviation; IQR, interquartile range; ICU, intensive care unit; MRSA, methicillin-resistant Staphylococcus aureus.
an = 100.
Distribution of patients meeting an area under the concentration-time curve (AUC)/minimum inhibitory concentration (MIC) ≥400 according to trough concentrations <10 mg/L compared to ≥10 mg/L. Patients with troughs <10 mg/L were significantly less likely to attain an AUC/MIC ≥400 relative to higher troughs (P = .018).
When compared to the reference group of patients with initial troughs between 15 and 20 mg/L, only the group of patients with troughs <10 mg/L differed significantly in attainment of a calculated AUC/MIC ≥400, as shown in Figure 3 (P = .045). A chi-square test for independence indicated no significant association between attaining a calculated AUC/MIC ≥400 and vancomycin troughs when comparing patients with troughs in the range of 10 to 14.9 mg/L to those with troughs of 15 to 20 mg/L (P = .817). There was also no significant difference in target AUC/MIC attainment between patients with troughs of 15 to 20 mg/L and those with troughs above 20 mg/L (P = .410). With the exception of the >20 mg/L group, all groups had similar baseline characteristics to the 15 to 20 mg/L group with regard to age, weight, MIC, SCr, eCrCl, and vancomycin dosing. Patients with troughs of >20 mg/L had significantly lower mean age, higher mean weights, and higher median eCrCl.
Attainment of area under the concentration-time curve (AUC)/minimum inhibitory concentration (MIC) ≥400 according to vancomycin trough range.
Of the 100 patients, 97 had assessable renal function for the secondary outcome. Nine (9.3%) of 97 patients developed AKI within the first 10 days of vancomycin treatment, 4 of which were intensive care unit (ICU) patients. Corrected average vancomycin trough was significantly higher in patients who developed AKI compared to those who did not (19.5 ± 3.6 vs 14.5 ± 4.2 mg/L, P < .001). Mean AUC and other baseline characteristics were not significantly different between the 2 groups.
Discussion
Our analysis found a limited association between serum vancomycin trough concentrations and attaining a calculated AUC/MIC ≥ 400 when troughs are above 10 mg/L. This seems to contradict consensus recommendations for vancomycin monitoring, suggesting that a minimum trough concentration of 15 mg/L would be desirable to improve AUC/MIC target attainment. 1 To our knowledge, previous studies have not evaluated the association between achieving a vancomycin trough concentration within target ranges and attainment of a calculated AUC/MIC ≥ 400. Our results comparing lower troughs to higher troughs within the generally accepted target range of 10 to 20 mg/L indicate that troughs within this range are poorly associated with reaching the target AUC/MIC of ≥400. From a pharmacodynamic stand point, this is expected since troughs represent a single point along the concentration–time profile and do not accurately represent overall exposure to the drug. Additionally, troughs do not take into account the MIC of the pathogen, rendering their measurement less useful when taken alone as a surrogate for treatment success.
In a retrospective cohort study of 127 patients using a similar method of AUC calculation, Ghosh and colleagues assessed the impact of AUC/MIC targets on vancomycin treatment failure in patients with MRSA bacteremia. 4 Of patients with vancomycin troughs between 15 and 20 mg/L, 34% did not achieve the pharmacodynamic target of an AUC/MIC ≥400 compared to 54.3% in our study. Additionally, 64% of patients in their study who attained an AUC/MIC ≥400 had vancomycin trough concentrations of <15 mg/L. 4 In our study, 52.4% (22 of 42) of patients with a calculated AUC/MIC ≥400 had troughs <15 mg/L. In a study by Neely and colleagues, various methods of AUC calculation were used, including multipoint kinetics. 29 They similarly found that about 60% of patients meeting the AUC/MIC target had steady state vancomycin troughs <15 mg/L. 29 Further, pharmacokinetic modeling studies in pediatric patients have suggested that an AUC/MIC ≥400 can be consistently met with trough concentrations <15 mg/L when the MIC is ≤1 mg/L. 30,31 Up to a 3-fold variability in AUC has also been noted with any given trough concentration, further illustrating the disconnect between AUC, AUC/MIC, and vancomycin trough concentrations. 31 Collectively, these results have clinical implications with regard to vancomycin dose adjustments aimed at achieving troughs within the “desired” range of 15 to 20 mg/L. In particular, vancomycin dosage increases made based on a trough concentration may subject the patient to an increased risk of adverse effects without improved efficacy, as the attainment of a calculated AUC/MIC ≥ 400 does not correlate with trough concentrations >15 mg/L.
Consensus guidelines recommend maintaining troughs above 10 mg/L, which is a recommendation consistent with our findings. 1 Previous literature supports targeting troughs above 10 mg/L to suppress the development of resistance in S. aureus isolates, primarily based on a report of a dialysis patient with chronic MRSA bacteremia that failed to develop resistance at vancomycin concentrations >10 mg/L. 18 In vitro, the same isolate developed reduced susceptibility when subjected to lower vancomycin concentrations. Interestingly, data from Tsuji and colleagues suggest that vancomycin resistance may also be driven by suboptimal AUC/MIC. 32 To our knowledge, there are no studies to date evaluating AUC/MIC attainment in the context of a trough concentration to evaluate the suppression of resistance. Our data suggest that there may be clinical implications associated with low vancomycin trough concentrations, as troughs <10 mg/L were associated with a decreased likelihood of reaching an AUC/MIC of ≥400 relative to higher troughs. Therefore, we agree that vancomycin trough concentrations should always be maintained above 10 mg/L.
Supplemental to our primary outcome, we also assessed the development of nephrotoxicity with regard to vancomycin trough concentrations. We calculated a corrected average vancomycin trough for each patient based on the thought that vancomycin-induced nephrotoxicity is a result of sustained exposure to elevated trough concentrations of the drug. A number of previous studies support the argument that elevated vancomycin trough concentrations predict the development of AKI. 12,33 –37 A meta-analysis further confirmed these findings, indicating a correlation between troughs ≥15 mg/L and an increased risk of nephrotoxicity (OR = 2.67,95% CI: 1.95-3.65). 17 Our study is consistent with these findings, noting a significantly higher mean corrected average vancomycin trough in patients developing AKI, though other causes of nephrotoxicity were not assessed.
Limitations
Our study had some limitations worth noting. First, this was a retrospective study that did not assess clinical outcomes, and the number of patients within each trough range was small. Although our sample size was based upon previous studies, the lack of a pre hoc power analysis limits statistical conclusions that can be made from our study. This limitation is particularly applicable to the comparison among higher trough groups, where a statistically significant difference in the attainment of the MIC/AUC target was not found. Despite this limitation, a greater percentage of patients in the 10 to 14.9 mg/L trough group achieved the target AUC–MIC ratio compared to the 15 to 20 mg/L group (P = .817). As such, it is highly unlikely that patients in the 15 to 20 mg/L group would have a statistically higher chance of achieving this target even if more patients were included. Another potential limitation with regard to the retrospective nature of the study was the determination of whether deviations in the trough draw time occurred. This was controlled by reviewing vancomycin administration times as well as the documentation made by clinical pharmacists at the time the trough value was evaluated.
We did not directly measure AUC but rather used a validated equation to calculate it. During the study period, it was not our institution’s standard to collect 2 vancomycin levels relative to the same dose, and a patient-specific vancomycin clearance could not be calculated. In order to control for retrospective bias, quantitative measures were identified and analyzed. Although the equation used in the current study has previously been validated, concerns exist that it may underestimate the actual AUC. However, this was determined to be the most readily and consistently obtainable measure of AUC, given the retrospective nature of the study. It should also be noted that this measure of AUC is calculated based on an estimate of CrCl, which has considerable inherent limitations. A recently published survey of clinical pharmacists demonstrated that there are a number of different ways to calculate CrCl. 38 For this reason, we used a standardized formula to calculate eCrCl across all patients. Estimated CrCl was capped at 120 mL/min for 29% of patients and ABWs were used in an attempt to control for some of the confounding variables that would artificially increase eCrCl and thus negatively impact the probability of attaining an AUC/MIC ≥400. Additionally, patients with unstable renal function were excluded to control for variability in true AUC relative to the time of trough obtainment. Given the retrospective nature of this study and current standards of care, more direct measurements of vancomycin clearance were not available. Therefore, our decision to utilize a validated equation for estimating vancomycin clearance and AUC was appropriate for the design of the study.
Finally, although an evaluation of vancomycin exposure and nephrotoxicity was not the primary objective of the study, we did evaluate this relationship with collected data. This part of the study was limited by the fact that other potential contributors to AKI were not assessed. It is general knowledge that the concomitant use of nephrotoxins strongly predicts the development of nephrotoxicity, and our results are likely to have been affected by such confounding variables. Additionally, it was inherently difficult to determine whether higher trough levels led to AKI or whether renal impairment resulted in the elevated trough levels we observed. Based upon these limitations, our data cannot provide definitive conclusions about the relationship between vancomycin exposure and nephrotoxicity. Instead, our data lend support to existing data that have correlated higher vancomycin trough concentrations with an increased risk of nephrotoxicity.
Future Directions
This study uncovered 2 aspects we aim to investigate in future studies. We are not aware of a study investigating the impact of using an AUC/MIC-guided dosing approach on clinical outcomes and attainment of the AUC/MIC target. Previous studies have indicated that standard vancomycin dosing regimens targeting recommended trough ranges do not reliably result in adequate attainment of the target AUC/MIC. 4,39,40 Our results show that patients receiving similar weight-based doses also had variable attainment of an AUC/MIC ≥ 400. Therefore, obtaining 2-point kinetic measurements in individual patients, including a postdistributional peak and a trough, would better represent a patient-specific concentration–time profile, thus allowing for a more accurate calculation of the AUC to MIC ratio. One caveat to this approach is that AUC calculations based on 2 points have been shown to underestimate true AUC. 31 Therapeutic drug monitoring of vancomycin may be improved using a patient-specific AUC/MIC as a more precise target to predict efficacy, while serum trough concentrations may serve to assess the risk of nephrotoxicity and the development of resistance.
Conclusion
Overall, our results support maintenance of serum vancomycin trough concentrations ≥10 mg/L but do not support the targeting of troughs above 15 mg/L as recommended for complicated infections. 1 When considering the numerous data correlating higher trough concentrations with increased nephrotoxicity risks, our results indicate that there may be inadequate benefit of targeting higher troughs to offset such risks. Future studies are needed to establish the utility of an AUC/MIC-guided vancomycin dosing approach and to identify indication-specific targets in order to improve patient outcomes.
Footnotes
Acknowledgments
The authors thank Luke Probst, PharmD, BCPS, and Scott Riddell, PhD.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Cory Hale, PharmD, received financial support in the form of a grant from the Research and Education Foundation of the New York State Council of Health-system Pharmacists.