Abstract
While previous studies have examined the association between vancomycin (VAN) exposure and MIC with regard to outcomes in methicillin-resistant Staphylococcus aureus bacteremia (MRSA-B), none have explored if a relationship exists with the VAN minimum bactericidal concentration (MBC). The objective of this study was to evaluate the VAN 24-h area under the curve (AUC24)/MBC ratio as a pharmacodynamic predictor of mortality. This retrospective cohort study included patients treated with VAN for MRSA-B with the primary outcome of 30-day all-cause mortality. Data collected included patient demographics, comorbidities, antimicrobial treatment data, therapeutic drug levels, and laboratory and microbiological data. Vancomycin MICs and MBCs were determined by Etest (MIC only) and broth microdilution (BMD). The vancomycin AUC24 was determined by pharmacokinetic maximum a posteriori probability Bayesian (MAP-Bayesian) analysis. The most significant breakpoint for 30-day mortality was determined by classification and regression tree (CART) analysis. The association between pharmacodynamic parameters (VAN AUC24/MICBMD, VAN AUC24/MICEtest, and AUC24/MBCBMD) and mortality were determined by χ2 and multivariable Poisson regression. Overall mortality in this cohort (n = 53) was 20.8% (n = 11/53), and all corresponding MRSA blood isolates were VAN susceptible (MIC range, 0.5 to 2 μg/ml; MIC50, 1 μg/ml; MIC90, 1 μg/ml). The CART-derived breakpoints for mortality were 176 (VAN AUC24/MBC) and 334 (VAN AUC24/MICBMD). In multivariable analysis, the association between a VAN AUC24/MBC of ≥176 and survival persisted, but VAN AUC24/MICBMD values (≥334 or ≥400) were not associated with improved mortality. In conclusion, VAN AUC24/MBC was a more important predictor of 30-day mortality than VAN AUC24/MIC for MRSA-B.
INTRODUCTION
The incidence of methicillin-resistant Staphylococcus aureus (MRSA) infections in health care and community settings has increased in recent years (1). Vancomycin (VAN) has been a mainstay in the armamentarium against MRSA for nearly half a century, but changes in susceptibility patterns have presented new treatment challenges (2). The gradual reduction in susceptibility of MRSA isolates to VAN occurring over the last 2 decades has been associated with increased treatment failure and mortality (3, 4). While the relationship between VAN MIC and clinical outcomes in MRSA bacteremia (MRSA-B) is well established, the importance of VAN minimum bactericidal concentration (MBC) is less clear (5).
Multiple studies have highlighted the importance of bactericidal therapy in treating S. aureus bacteremia (SAB). Previous research has demonstrated that in vitro 24-h bactericidal activity by VAN is associated with a decrease in the duration of MRSA-B clinically compared to in vitro bacteriostatic activity (6). Additionally, decreased in vitro VAN killing activity over 72 h also has been associated with an increase in 30-day mortality in patients with MRSA-B treated with VAN (7). Sakoulas et al. reported an association between elevated VAN MIC and treatment failure but also noted that reduced bactericidal activity had a similar effect (8). Lastly, multiple studies have demonstrated improved clinical outcomes with β-lactam therapy compared to those with VAN for methicillin-susceptible SAB, which may be partially attributed to the superior killing activity of β-lactams compared to that of VAN (9, 10). Thus, there is evidence to suggest that VAN bactericidal activity plays an influential role in clinical outcomes of SAB, including MRSA-B.
Consensus recommendations support a VAN area under the curve (AUC)/MIC ratio of ≥400 as a pharmacodynamic predictor of therapeutic effectiveness for MRSA infections (11). An AUC/MIC ratio of ≥421 also has been associated with clinical success in MRSA-B (12). Due to the trend toward improved outcomes with increased bactericidal activity of VAN, we hypothesized that VAN exposure as a function of MBC rather than MIC may be a more important predictor of VAN effectiveness in the treatment of MRSA-B. Therefore, the objectives of this study were to (i) evaluate the association between pharmacodynamic parameters (VAN AUC24/MIC ratio determined by broth microdilution [MICBMD], VAN AUC24/MIC ratio determined by Etest [MICEtest], and VAN AUC24/MBC ratio) and 30-day mortality and (ii) quantify the VAN AUC24/MBC threshold associated with an increased probability of mortality in a cohort of hospitalized patients with MRSA-B.
(This work was presented in part at the 54th Interscience Conference on Antimicrobial Agents and Chemotherapy, Washington, DC, 8 September 2014.)
MATERIALS AND METHODS
Study population and data sources.
This was a retrospective study of patients treated for MRSA-B at the University of Kansas Hospital, a tertiary-care academic medical center, from September 2012 through June 2014. Inclusion criteria were (i) age of ≥18 years, (ii) positive blood culture for MRSA, and (iii) intravenous treatment with VAN. Patients were excluded if they were on hemodialysis, had polymicrobial bacteremia at onset, or were treated with VAN for <48 h. For patients with multiple clinical MRSA blood isolates during the study period, only the first isolate and corresponding clinical case of MRSA-B were evaluated. Clinical data were collected by retrospective chart review. Variables abstracted included baseline patient demographics, setting of bacteremia onset (community versus hospital-acquired [admission within ≥72 h]), comorbidities, Charlson comorbidity index, Pitt bacteremia score (PBS), acute physiology and chronic health evaluation II (APACHE II) score, laboratory data (including VAN therapeutic drug monitoring), microbiology data, antimicrobial treatment data (including previous VAN exposure within 30 days), and vital signs. The focus of MRSA-B was determined as documented by a treating physician and categorized according to mortality risk (13). Immunosuppression was defined as chronic steroid (equivalent to ≥20 mg prednisone) or antineoplastic use, neutropenia, or leukopenia. The primary clinical outcome was 30-day all-cause mortality. This study was approved by the University of Kansas Medical Center institutional review board.
AUC determination.
Vancomycin AUC values were estimated with ADAPT 5 software (University of Southern California, Los Angeles, CA, USA). Using data from a previously published 2-compartment VAN model, the mean parameter vector and full covariance matrix were embedded into the PRIOR subroutine of ADAPT 5 (14, 15). For each individual patient, the pharmacokinetic parameter values were estimated using the maximum a posteriori probability (MAP) procedure. After the MAP-Bayesian step, values of the AUC for 0 to 72 h (AUC0–72) and AUC0–96 were estimated for each patient. For the purposes of these analyses, the AUC at 72 to 96 h (AUC24) of therapy was used, because this is the 24-h exposure variable that most closely resembles a steady-state AUC (16). The vancomycin AUC24 was calculated as the difference in integrated AUC0–72 and AUC0–96 values.
Microbiological analysis. (i) Bacterial strains.
Clinical S. aureus blood isolates from MRSA-B cases encountered during the study period were stored at −70°C and passed for 3 consecutive days on tryptic soy agar (TSA) to ensure uniform metabolic activity prior to testing.
(ii) Antimicrobials.
Vancomycin was commercially purchased from Sigma-Aldrich Co. (St. Louis, MO, USA). Stock solutions of VAN were prepared daily.
(iv) Media.
Mueller-Hinton II broth (Becton-Dickinson, Sparks, MD, USA) was used as media for susceptibility testing. Brain heart infusion agar (Becton-Dickinson, Sparks, MD, USA) was used as the medium for the Etest. Tryptic soy agar (Becton-Dickinson, Sparks, MD, USA) was used for colony counts.
(v) Susceptibility testing.
Vancomycin MICs were determined by broth microdilution (BMD) at concentrations of 0.125 to 64 μg/ml at a standard inoculum of approximately 1 × 106 CFU/ml according to Clinical and Laboratory Standards Institute (CLSI) guidelines (17). The vancomycin MIC also was determined for all isolates by Etest according to manufacturer recommendations (bioMérieux, Marcy l'Etoile, France). Staphylococcus aureus ATCC 25923 was used as the control strain, and each test was performed in duplicate to ensure reproducibility.
(vi) MBC determination.
Vancomycin MBCs were determined by the microdilution method according to CLSI guidelines (18). To account for antibiotic carryover, 100-μl aliquots of all wells with no visible growth after 24 h of incubation at 35°C (i.e., above the MICBMD endpoint) were streaked onto TSA, allowed to visibly dry at room temperature, and cross-streaked using a sterile cotton-tipped swab (18, 19). The vancomycin MBC was defined as the lowest concentration with ≥99.9% killing after 24 h of incubation at 35°C. Vancomycin tolerance was defined as a MBC/MIC ratio of ≥32.
Statistical analysis.
Categorical variables were compared by χ2 or two-tailed Fisher's exact test, and continuous variables were compared by t test or Mann-Whitney U test. The AUC24 variable was divided by MICBMD, MICEtest, and MBC, respectively, to compute the VAN pharmacodynamic variables of interest. Classification and regression tree (CART) analyses were used to identify the breakpoints in these pharmacodynamic parameters associated with an increased probability of the primary outcome (30-day mortality). A secondary analysis examining the relationship between VAN exposure within the preceding 30 days and VAN MICBMD and MBC was performed using the Cochran-Armitage trend test.
The association between VAN pharmacodynamic parameters and clinical outcomes was evaluated using backward stepwise Poisson regression with CART-derived breakpoints forced into respective models. Variables which were associated with the pharmacodynamic parameter of interest or 30-day mortality (P < 0.2) were entered into the model, and only those that confounded the relationship between the pharmacodynamic breakpoint and 30-day mortality (≥10% change in the associated risk ratio) were retained in the final parsimonious model. The association between the guideline concordant VAN AUC24/MICBMD target of ≥400 and 30-day mortality also was evaluated in this cohort (11). Time-to-event analyses were conducted for 30-day mortality using the Kaplan-Meier method and log-rank test. Survivors were right-censored at the end of treatment with VAN, and backward stepwise Cox regression was performed. Statistical analyses were performed using SAS (version 9.2; SAS Institute, Inc., Cary, NC, USA), and a two-tailed P value of <0.05 was considered statistically significant for all tests.
RESULTS
During the study period, a total of 100 patients with MRSA-B who were treated with VAN were identified. Patients were excluded due to treatment with VAN for <48 h (n = 13) and due to hemodialysis (n = 34). The final analysis included a total of 53 patients. All corresponding clinical MRSA blood isolates were VAN susceptible (MIC range, 0.5 to 2 μg/ml; MIC50, 1 μg/ml; MIC90, 1 μg/ml). Among these isolates, 13/53 (24.5%) were VAN tolerant (MBC range, 0.5 to 64 μg/ml; MBC50, 1 μg/ml; MBC90, 64 μg/ml). Additionally, 25/53 (47.2%) of clinical MRSA isolates had dissociated MICBMD and MBC values.
Overall, 30-day mortality was 20.8% (n = 11/53). Factors associated with 30-day mortality among MRSA-B patients treated with VAN are included in Table 1. In univariable analysis, mean age, an age of ≥85 years, previous VAN exposure within 30 days, intensive care unit (ICU) admission, high-risk infection focus, malignancy, median Charlson comorbidity index, PBS, and APACHE II score were significantly associated with increased 30-day mortality. Conversely, a medium-risk focus of infection was associated with decreased 30-day mortality.
TABLE 1.
Factors associated with 30-day mortality among MRSA-B patients treated with VAN
| Factora | Value for patients with: |
P value | |
|---|---|---|---|
| No mortality (n = 42) | Mortality (n = 11) | ||
| Age (yr), means ± SD | 58.5 ± 15.9 | 70.6 ± 16.6 | 0.031 |
| ≥65 yr, n (%) | 17 (40.5) | 7 (63.6) | 0.170 |
| ≥85 yr, n (%) | 1 (2.4) | 3 (27.3) | 0.025 |
| Female gender, n (%) | 15 (35.7) | 3 (27.3) | 0.730 |
| Previous VAN exposure within 30 days, n (%) | 4 (9.5) | 5 (45.5) | 0.031 |
| Hospital-acquired MRSA-B, n (%) | 12 (28.6) | 6 (54.5) | 0.105 |
| ICU, n (%) | 10 (23.8) | 6 (54.5) | 0.048 |
| Sepsis, n (%) | 27 (64.3) | 9 (81.8) | 0.267 |
| Septic shock, n (%) | 6 (14.3) | 2 (18.2) | 0.665 |
| Immunosuppression, n (%) | 10 (23.8) | 4 (36.4) | 0.453 |
| MRSA-B focus | |||
| High risk, n (%) | 10 (23.8) | 8 (72.7) | 0.002 |
| Medium risk, n (%) | 26 (61.9) | 2 (18.2) | 0.016 |
| Low risk, n (%) | 6 (14.3) | 1 (9.1) | 1.000 |
| MRSA bacteriuria, n (%) | 4 (9.5) | 1 (9.1) | 1.000 |
| Diabetes mellitus, n (%) | 15 (35.7) | 4 (36.4) | 0.968 |
| Alcoholism, n (%) | 4 (9.5) | 1 (9.1) | 1.000 |
| Cirrhosis, n (%) | 3 (7.1) | 1 (9.1) | 1.000 |
| Congestive heart failure, n (%) | 3 (7.1) | 2 (18.2) | 0.275 |
| Malignancy, n (%) | 8 (19.0) | 7 (63.6) | 0.003 |
| HIV positive, n (%) | 1 (2.4) | 0 (0.0) | 1.000 |
| Chronic obstructive pulmonary disease, n (%) | 5 (11.9) | 2 (18.2) | 0.626 |
| Charlson comorbidity index, median (IQR) | 6 (3–7) | 7 (6–9) | 0.017 |
| PBS, median (IQR) | 1 (0–3) | 4 (1–6) | 0.028 |
| PBS ≥ 4, n (%) | 7 (16.7) | 6 (54.5) | 0.009 |
| APACHE II score, median (IQR) | 11 (7–14) | 17 (14–25) | 0.002 |
IQR, interquartile range.
In this cohort, the mean duration of VAN therapy was 8.73 days (standard deviations [SD], 5.23 days). The mean VAN AUC24 at steady state [AUC24(ss)] was 681 mg · h/liter (SD, 331 mg · h/liter). The median VAN AUC24(ss) was 661.7 mg · h/liter (interquartile range, 440.6 to 834.1 mg · h/liter). The CART-derived 30-day mortality breakpoints were 334 for VAN AUC24/MICBMD ratio and 176 for VAN AUC24/MBC ratio. Classification and regression tree analysis yielded no optimal breakpoint for the association between AUC24/MICEtest and 30-day mortality, so no further analyses were conducted for this parameter. In univariable analysis, mortality was significantly lower when the VAN AUC24/MICBMD was ≥334 (7.7% [n = 2/26] versus 33.3% [n = 9/27]; P = 0.021) and when the VAN AUC24/MICBMD was ≥400 (14.3% [n = 6/42] versus 45.5% [n = 5/11]; P = 0.037). Mortality also was significantly lower when the VAN AUC24/MBC was ≥176 (9.4% [n = 3/32] versus 38.1% [n = 8/21]; P = 0.017). Comparisons of baseline patient characteristics by CART-derived breakpoints are displayed in Tables 2 and 3. As shown, there were many significant differences between groups with regard to these characteristics. In the secondary trend analysis, prior VAN exposure within 30 days was associated with a linear increase in VAN MICBMD (P = 0.033) and VAN MBC (P = 0.001).
TABLE 2.
Patient characteristics by CART-derived AUC24/MBC breakpoint for VAN in MRSA-B
| Characteristica | Value by AUC24/MBC category |
P value | |
|---|---|---|---|
| <176 (n = 21) | ≥176 (n = 32) | ||
| Age (years), means ± SD | 67.6 ± 15.5 | 55.2 ± 15.9 | 0.007 |
| ≥65 yr, n (%) | 13 (61.9) | 11 (34.4) | 0.049 |
| ≥85 yr, n (%) | 4 (19.0) | 0 (0.0) | 0.020 |
| Female gender, n (%) | 6 (28.6) | 12 (37.5) | 0.502 |
| Previous VAN exposure within 30 days, n (%) | 7 (33.3) | 2 (6.2) | 0.021 |
| Hospital-acquired MRSA-B, n (%) | 11 (52.4) | 7 (21.9) | 0.022 |
| ICU, n (%) | 8 (38.1) | 8 (25.0) | 0.310 |
| Sepsis, n (%) | 16 (76.2) | 20 (62.5) | 0.296 |
| Septic shock, n (%) | 6 (28.6) | 2 (6.2) | 0.047 |
| Immunosuppression, n (%) | 4 (19.0) | 10 (31.2) | 0.324 |
| MRSA-B focus, n (%) | |||
| High risk | 10 (47.6) | 8 (25.0) | 0.089 |
| Medium risk | 7 (33.3) | 21 (65.6) | 0.021 |
| Low risk | 4 (19.0) | 3 (9.4) | 0.415 |
| MRSA bacteriuria, n (%) | 2 (9.5) | 2 (9.4) | 1.000 |
| Diabetes mellitus, n (%) | 7 (33.3) | 12 (37.5) | 0.757 |
| Alcoholism, n (%) | 1 (4.8) | 4 (12.5) | 0.637 |
| Cirrhosis, n (%) | 0 (0.0) | 4 (12.5) | 0.143 |
| Congestive heart failure, n (%) | 5 (23.8) | 0 (0.0) | 0.007 |
| Malignancy, n (%) | 7 (33.3) | 8 (25.0) | 0.546 |
| HIV positive, n (%) | 0 (0.0) | 1 (3.1) | 1.000 |
| Chronic obstructive pulmonary disease, n (%) | 4 (19.0) | 3 (9.4) | 0.415 |
| Charlson comorbidity index, median (IQR) | 7 (5–9) | 6 (3–7) | 0.058 |
| PBS, median (IQR) | 2 (1–4) | 2 (0–4) | 0.528 |
| PBS ≥ 4, n (%) | 5 (23.8) | 8 (25.0) | 0.922 |
| APACHE II, median (IQR) | 13 (9–18) | 10 (7–15) | 0.084 |
IQR, interquartile range.
TABLE 3.
Patient characteristics by CART-derived AUC24/MICBMD breakpoint for VAN in MRSA-Ba
| Characteristic | Value by AUC24/MICBMD category |
P value | |
|---|---|---|---|
| <334 (n = 27) | ≥334 (n = 26) | ||
| Age (yr), means ± SD | 63.5 ± 19.1 | 56.5 ± 13.5 | 0.136 |
| ≥65 yr, n (%) | 14 (51.9) | 10 (38.5) | 0.328 |
| ≥85 yr, n (%) | 4 (14.8) | 0 (0.0) | 0.111 |
| Female gender, n (%) | 8 (29.6) | 10 (38.5) | 0.497 |
| Previous VAN exposure within 30 days, n (%) | 8 (29.6) | 1 (3.8) | 0.024 |
| Hospital-acquired MRSA-B, n (%) | 12 (44.4) | 6 (23.1) | 0.101 |
| ICU, n (%) | 10 (37.0) | 6 (23.1) | 0.268 |
| Sepsis, n (%) | 20 (74.1) | 16 (61.5) | 0.328 |
| Septic shock, n (%) | 7 (25.9) | 1 (3.8) | 0.050 |
| Immunosuppression, n (%) | 6 (22.2) | 8 (30.8) | 0.480 |
| MRSA-B focus, n (%) | |||
| High risk | 12 (44.4) | 6 (23.1) | 0.101 |
| Medium risk | 11 (40.7) | 17 (65.4) | 0.072 |
| Low risk | 4 (14.8) | 3 (11.5) | 0.725 |
| MRSA bacteriuria, n (%) | 2 (7.4) | 3 (11.5) | 0.669 |
| Diabetes mellitus, n (%) | 9 (33.3) | 10 (38.5) | 0.697 |
| Alcoholism, n (%) | 1 (3.7) | 4 (15.4) | 0.192 |
| Cirrhosis, n (%) | 0 (0.0) | 4 (15.4) | 0.051 |
| Congestive heart failure, n (%) | 5 (18.5) | 0 (0.0) | 0.051 |
| Malignancy, n (%) | 9 (33.3) | 6 (23.1) | 0.407 |
| HIV positive, n (%) | 1 (3.7) | 0 (0.0) | 1.000 |
| Chronic obstructive pulmonary disease, n (%) | 4 (14.8) | 3 (11.5) | 0.725 |
| Charlson comorbidity index, median (IQR) | 7 (5–9) | 5 (3–7) | 0.099 |
| PBS, median (IQR) | 2 (1–4) | 2 (0–3) | 0.238 |
| PBS ≥ 4, n (%) | 7 (25.9) | 6 (23.1) | 0.810 |
| APACHE II, median (IQR) | 13 (6–17) | 11 (8–14) | 0.232 |
IQR, interquartile range.
The association between pharmacodynamic parameters and 30-day mortality were evaluated in multivariable Poisson regression analyses (Table 4). As shown, the association between a VAN AUC24/MBC ratio of <176 and 30-day mortality persisted after adjusting for age, malignancy, PBS, and cirrhosis (Table 4, model 1). After adjusting for confounding factors, VAN AUC24/MICBMD ratios of <334 and <400 were not significantly associated with mortality in this cohort (Table 4, models 2 and 3).
TABLE 4.
Poisson regression models for 30-day mortality by pharmacodynamic parameter among MRSA-B patients
| Model and characteristic | Adjusted risk ratio (95% CI) | P value |
|---|---|---|
| Model 1 | ||
| VAN AUC24/MBC < 176 | 3.68 (1.17–11.56) | 0.025 |
| Age ≥85 yr | 4.24 (1.75–10.23) | 0.001 |
| Malignancy | 5.62 (2.31–13.65) | <0.001 |
| PBS ≥ 4 | 3.63 (1.48–8.67) | 0.005 |
| Cirrhosis | 6.33 (0.60–67.38) | 0.120 |
| Model 2 | ||
| VAN AUC24/MICBMD < 334 | 5.12 (0.85–30.67) | 0.074 |
| Malignancy | 3.70 (1.25–10.97) | 0.019 |
| MRSA pneumonia | 3.56 (1.13–11.26) | 0.031 |
| Cirrhosis | 9.88 (0.72–34.79) | 0.086 |
| Model 3 | ||
| VAN AUC/MICBMD < 400 | 1.95 (0.92–4.15) | 0.084 |
| Malignancy | 3.89 (1.31–11.53) | 0.014 |
| MRSA pneumonia | 3.59 (1.16–11.11) | 0.026 |
| Cirrhosis | 3.73 (0.39–36.07) | 0.255 |
In Kaplan-Meier analysis, the CART-derived VAN AUC24/MBC breakpoint of ≥176 was significantly associated with improved survival (P = 0.011 by log-rank test). When evaluated by multivariable Cox regression, the association between an AUC24/MBC ratio of ≥176 and improved survival persisted (adjusted hazard ratio [aHR], 7.32; 95% confidence interval [CI], 1.66 to 32.29; P = 0.009) after adjusting for APACHE II score (aHR, 1.09; 95% CI, 1.01 to 1.19; P = 0.042) and PBS of ≥4 (aHR, 6.80, 95% CI, 1.68 to 27.60; P = 0.007).
DISCUSSION
The present investigation aimed to evaluate the relationship between various VAN pharmacodynamic parameters and 30-day mortality in MRSA-B. Primarily, we sought to derive an associated VAN AUC24/MBC breakpoint and evaluate this breakpoint alongside other pharmacodynamic parameters in a series of multivariable analyses. To our knowledge, this is the first study to establish a VAN AUC24/MBC breakpoint for 30-day mortality. Multiple studies have derived VAN AUC/MIC breakpoints, and current consensus recommendations target a VAN AUC/MIC ratio of ≥400 to maximize therapeutic effectiveness, although this goal is only clinically feasible with some MIC values in the susceptible range (11, 12, 16). In the present investigation, a VAN AUC24/MICBMD ratio of ≥334 and VAN AUC24/MICBMD ratio of ≥400 both were significantly associated with improved mortality in univariable analysis. However, the CART-derived VAN AUC24/MBC breakpoint of 176 was the only pharmacodynamic parameter that was significantly associated with improved mortality after adjusting for confounding factors. A vancomycin AUC24/MBC ratio of ≥176 also was associated with improved survival in Kaplan-Meier analysis and multivariable Cox regression.
The association between VAN MIC and clinical outcomes has been evaluated exhaustively in the literature, while little attention has been paid to the potential importance of bactericidal activity (3, 5). Studies that have evaluated VAN bactericidal activity have noted an apparent association between reduced bactericidal activity and worse clinical outcomes (7, 8). Although VAN MIC and MBC values are correlated, it is not uncommon for these values to be either slightly or greatly dissociated (20–22). In the present study, nearly half of corresponding MRSA isolates had some level of dissociation between VAN MICBMD and MBC. Because bactericidal activity is preferred in the treatment of MRSA-B, a VAN AUC24/MIC target theoretically would be less likely to predict therapeutic effectiveness when MIC and MBC values are dissociated (6). Vancomycin AUC24/MBC estimations would allow for the differentiation of these cases and potentially provide a more sensitive predictor of VAN effectiveness. The results of the present study appear to support this logic.
The finding that prior VAN exposure can influence VAN activity as well as clinical outcomes in MRSA-B is consistent with previous research (23, 24). This observation potentially has significant clinical implications in the treatment of MRSA-B, as previous VAN exposure was associated with increased VAN MICBMD and VAN MBC and a lower probability of achieving optimal AUC24/MICBMD and AUC24/MBC ratios in univariable analysis. Therefore, alternatives to VAN may be warranted for treatment of MRSA-B in patients with recent VAN exposure, even if the corresponding isolate tests susceptible.
The utility of the VAN AUC24/MBC ratio as a pharmacodynamic predictor of VAN effectiveness likely is dependent on the prevalence of dissociated VAN MIC and MBC values, including VAN tolerance. The prevalence of VAN tolerance is approximately 20% among MRSA blood isolates, although tolerance is as high as 43% in some institutions (21). Therefore, the proportion of VAN tolerant isolates observed in the present study is consistent with previous research and likely similar to tolerance rates typically found in U.S. hospitals. In institutions where VAN tolerance is more common, the difference between AUC24/MIC and AUC24/MBC targets likely would be even more profound. It should be noted that AUC72–96 was used in this study as a measure of 24-h VAN exposure. The decision to select VAN AUC72–96 was informed by the literature (16). Furthermore, the vast majority of patients will have achieved steady-state concentrations of VAN and most clinicians will have commenced VAN therapeutic drug monitoring by this time point. Unfortunately, the literature is not clear regarding the optimal VAN AUC time interval for predicting clinical outcomes, and this question should be addressed in future studies (16, 25–27).
The present investigation has a number of noteworthy limitations. Most notably, this was a retrospective analysis and suffers from the limitations of such a design. Although the VAN AUC24/MICBMD ratio was not predictive of 30-day mortality in this cohort after adjusting for other factors, there is a concern for type II error due to the relatively small number of patients included. It is important to consider that the utility of the VAN AUC24/MBC ratio as a pharmacodynamic predictor of VAN effectiveness likely will depend on institution-specific VAN tolerance rates, which may not be known. Previous research has demonstrated that other phenotypic characteristics of MRSA, such as biofilm formation and susceptibility to host defense cationic peptide, play an influential role in clinical outcomes of MRSA-B (28). Although the purpose of this study was to investigate the relationship between VAN AUC24/MBC ratios and clinical outcomes, it is important to consider the possible influence of unmeasured phenotypic and genotypic microbiological characteristics (28, 29). Host defense cationic peptide phenotypes have been associated with reduced susceptibility to vancomycin and defective autolytic mechanisms (30). These phenotypic changes, in addition to enhanced biofilm production, have been associated with poor clinical outcomes in MRSA-B, including persistent infection (20, 28, 30). Additionally, the cause of death was not known and we were unable to ascertain attributable mortality. Important strengths of this study were the utilization of a method for MBC testing which appropriately accounted for antibiotic carryover, use of a MAP-Bayesian approach to determine VAN AUC24, and application of CART analysis to define a novel VAN AUC24/MBC breakpoint for 30-day mortality.
In summary, the VAN AUC24/MBC was a more important predictor of 30-day mortality than the VAN AUC24/MICBMD in a cohort of patients with MRSA-B. The VAN AUC24/MBC breakpoint most predictive of survival was ≥176. Overall, the results of this study add to the growing body of literature underscoring the importance of VAN bactericidal activity in the treatment of MRSA-B. Alternatives to VAN may be warranted in cases of MRSA-B where the achievement of optimal AUC24/MBC is unlikely, such as in patients with VAN-tolerant clinical isolates or recent VAN exposure. This research should be replicated in a larger analysis and at institutions with various rates of VAN tolerance.
REFERENCES
- 1.Popovich KJ, Weinstein RA, Hota B. 2008. Are community-associated methicillin-resistant Staphylococcus aureus (MRSA) strains replacing traditional nosocomial MRSA strains? Clin Infect Dis 46:787–794. doi: 10.1086/528716. [DOI] [PubMed] [Google Scholar]
- 2.Brink AJ. 2012. Does resistance in severe infections caused by methicillin-resistant Staphylococcus aureus give you the “creeps”? Curr Opin Crit Care 18:451–459. doi: 10.1097/MCC.0b013e3283578968. [DOI] [PubMed] [Google Scholar]
- 3.Mavros MN, Tansarli GS, Vardakas KZ, Rafailidis PI, Karageorgopoulos DE, Falagas ME. 2012. Impact of vancomycin minimum inhibitory concentration on clinical outcomes of patients with vancomycin-susceptible Staphylococcus aureus infections: a meta-analysis and meta-regression. Int J Antimicrob Agents 40:496–509. doi: 10.1016/j.ijantimicag.2012.07.023. [DOI] [PubMed] [Google Scholar]
- 4.Jacob JT, DiazGranados CA. 2013. High vancomycin minimum inhibitory concentration and clinical outcomes in adults with methicillin-resistant Staphylococcus aureus infections: a meta-analysis. Int J Infect Dis 17:e93–e100. doi: 10.1016/j.ijid.2012.08.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lodise TP, Graves J, Evans A, Graffunder E, Helmecke M, Lomaestro BM, Stellrecht K. 2008. Relationship between vancomycin MIC and failure among patients with methicillin-resistant Staphylococcus aureus bacteremia treated with vancomycin. Antimicrob Agents Chemother 52:3315–3320. doi: 10.1128/AAC.00113-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Moise PA, Sakoulas G, Forrest A, Schentag JJ. 2007. Vancomycin in vitro bactericidal activity and its relationship to efficacy in clearance of methicillin-resistant Staphylococcus aureus bacteremia. Antimicrob Agents Chemother 51:2582–2586. doi: 10.1128/AAC.00939-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Miyazaki M, Takata T, Yoshimura H, Matsunaga A, Ohta D, Ishikura H, Futo M, Hara S, Kamimura H, Tamura K, Ngo D, Tsuji BT. 2011. Vancomycin bactericidal activity as a predictor of 30-day mortality in patients with methicillin-resistant Staphylococcus aureus bacteremia. Antimicrob Agents Chemother 55:1819–1820. doi: 10.1128/AAC.01536-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Sakoulas G, Moise-Broder PA, Schentag J, Forrest A, Moellering RC Jr, Eliopoulos GM. 2004. Relationship of MIC and bactericidal activity to efficacy of vancomycin for treatment of methicillin-resistant Staphylococcus aureus bacteremia. J Clin Microbiol 42:2398–2402. doi: 10.1128/JCM.42.6.2398-2402.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Schweizer ML, Furuno JP, Harris AD, Johnson JK, Shardell MD, McGregor JC, Thom KA, Cosgrove SE, Sakoulas G, Perencevich EN. 2011. Comparative effectiveness of nafcillin or cefazolin versus vancomycin in methicillin-susceptible Staphylococcus aureus bacteremia. BMC Infect Dis 11:279. doi: 10.1186/1471-2334-11-279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.McDanel JS, Perencevich EN, Diekema DJ, Herwaldt LA, Smith TC, Chrischilles EA, Dawson JD, Jiang L, Goto M, Schweizer ML. 2015. Comparative effectiveness of beta-lactams versus vancomycin for treatment of methicillin-susceptible Staphylococcus aureus bloodstream infections among 122 hospitals. Clin Infect Dis 61:361–367. doi: 10.1093/cid/civ308. [DOI] [PubMed] [Google Scholar]
- 11.Rybak MJ, Lomaestro BM, Rotschafer JC, Moellering RC Jr, Craig WA, Billeter M, Dalovisio JR, Levine DP. 2009. Therapeutic monitoring of vancomycin in adults summary of consensus recommendations from the American Society of Health-System Pharmacists, the Infectious Diseases Society of America, and the Society of Infectious Diseases Pharmacists. Pharmacotherapy 29:1275–1279. doi: 10.1592/phco.29.11.1275. [DOI] [PubMed] [Google Scholar]
- 12.Kullar R, Davis SL, Levine DP, Rybak MJ. 2011. Impact of vancomycin exposure on outcomes in patients with methicillin-resistant Staphylococcus aureus bacteremia: support for consensus guidelines suggested targets. Clin Infect Dis 52:975–981. doi: 10.1093/cid/cir124. [DOI] [PubMed] [Google Scholar]
- 13.Soriano A, Martinez JA, Mensa J, Marco F, Almela M, Moreno-Martinez A, Sanchez F, Munoz I, Jimenez de Anta MT, Soriano E. 2000. Pathogenic significance of methicillin resistance for patients with Staphylococcus aureus bacteremia. Clin Infect Dis 30:368–373. doi: 10.1086/313650. [DOI] [PubMed] [Google Scholar]
- 14.Pryka RD, Rodvold KA, Garrison M, Rotschafer JC. 1989. Individualizing vancomycin dosage regimens: one- versus two-compartment Bayesian models. Ther Drug Monit 11:450–454. doi: 10.1097/00007691-198907000-00013. [DOI] [PubMed] [Google Scholar]
- 15.Rodvold KA, Pryka RD, Garrison M, Rotschafer JC. 1989. Evaluation of a two-compartment Bayesian forecasting program for predicting vancomycin concentrations. Ther Drug Monit 11:269–275. doi: 10.1097/00007691-198905000-00009. [DOI] [PubMed] [Google Scholar]
- 16.Patel N, Pai MP, Rodvold KA, Lomaestro B, Drusano GL, Lodise TP. 2011. Vancomycin: we can't get there from here. Clin Infect Dis 52:969–974. doi: 10.1093/cid/cir078. [DOI] [PubMed] [Google Scholar]
- 17.Clinical and Laboratory Standards Institute. 2011. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 9th ed Approved standard M7-A9. Clinical and Laboratory Standards Institute, Wayne, PA, USA. [Google Scholar]
- 18.Clinical and Laboratory Standards Institute. 1999. Methods for determining bactericidal activity of antimicrobial agents. Approved guideline M26-A. Clinical and Laboratory Standards Institute, Wayne, PA, USA. [Google Scholar]
- 19.Pelletier LL Jr, Baker CB. 1988. Oxacillin, cephalothin, and vancomycin tube macrodilution MBC result reproducibility and equivalence to MIC results for methicillin-susceptible and reputedly tolerant Staphylococcus aureus isolates. Antimicrob Agents Chemother 32:374–377. doi: 10.1128/AAC.32.3.374. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Rose WE, Fallon M, Moran JJ, Vanderloo JP. 2012. Vancomycin tolerance in methicillin-resistant Staphylococcus aureus: influence of vancomycin, daptomycin, and telavancin on differential resistance gene expression. Antimicrob Agents Chemother 56:4422–4427. doi: 10.1128/AAC.00676-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Sader HS, Jones RN, Rossi KL, Rybak MJ. 2009. Occurrence of vancomycin-tolerant and heterogeneous vancomycin-intermediate strains (hVISA) among Staphylococcus aureus causing bloodstream infections in nine USA hospitals. J Antimicrob Chemother 64:1024–1028. doi: 10.1093/jac/dkp319. [DOI] [PubMed] [Google Scholar]
- 22.Sieradzki K, Tomasz A. 2006. Inhibition of the autolytic system by vancomycin causes mimicry of vancomycin-intermediate Staphylococcus aureus-type resistance, cell concentration dependence of the MIC, and antibiotic tolerance in vancomycin-susceptible S. aureus. Antimicrob Agents Chemother 50:527–533. doi: 10.1128/AAC.50.2.527-533.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Moise PA, Smyth DS, El-Fawal N, Robinson DA, Holden PN, Forrest A, Sakoulas G. 2008. Microbiological effects of prior vancomycin use in patients with methicillin-resistant Staphylococcus aureus bacteraemia. J Antimicrob Chemother 61:85–90. [DOI] [PubMed] [Google Scholar]
- 24.van Hal SJ, Jensen SO, Vaska VL, Espedido BA, Paterson DL, Gosbell IB. 2012. Predictors of mortality in Staphylococcus aureus Bacteremia. Clin Microbiol Rev 25:362–386. doi: 10.1128/CMR.05022-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Lodise TP, Patel N, Lomaestro BM, Rodvold KA, Drusano GL. 2009. Relationship between initial vancomycin concentration-time profile and nephrotoxicity among hospitalized patients. Clin Infect Dis 49:507–514. doi: 10.1086/600884. [DOI] [PubMed] [Google Scholar]
- 26.Lodise TP, Drusano GL, Zasowski E, Dihmess A, Lazariu V, Cosler L, McNutt LA. 2014. Vancomycin exposure in patients with methicillin-resistant Staphylococcus aureus bloodstream infections: how much is enough? Clin Infect Dis 59:666–675. doi: 10.1093/cid/ciu398. [DOI] [PubMed] [Google Scholar]
- 27.Hamada Y, Kuti JL, Nicolau DP. 2015. In vitro pharmacodynamics of vancomycin against methicillin-susceptible and -resistant Staphylococcus aureus: considering the variability in observed tissue exposure. Antimicrob Agents Chemother 60:955–961. doi: 10.1128/AAC.01553-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Seidl K, Bayer AS, Fowler VG Jr, McKinnell JA, Abdel Hady W, Sakoulas G, Yeaman MR, Xiong YQ. 2011. Combinatorial phenotypic signatures distinguish persistent from resolving methicillin-resistant Staphylococcus aureus bacteremia isolates. Antimicrob Agents Chemother 55:575–582. doi: 10.1128/AAC.01028-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Lodise TP, Drusano GL, Lazariu V, El-Fawal N, Evans A, Graffunder E, Stellrecht K, Mendes RE, Jones RN, Cosler L, McNutt LA. 2014. Quantifying the matrix of relationships between reduced vancomycin susceptibility phenotypes and outcomes among patients with MRSA bloodstream infections treated with vancomycin. J Antimicrob Chemother 69:2547–2555. doi: 10.1093/jac/dku135. [DOI] [PubMed] [Google Scholar]
- 30.Sakoulas G, Eliopoulos GM, Fowler VG Jr, Moellering RC Jr, Novick RP, Lucindo N, Yeaman MR, Bayer AS. 2005. Reduced susceptibility of Staphylococcus aureus to vancomycin and platelet microbicidal protein correlates with defective autolysis and loss of accessory gene regulator (agr) function. Antimicrob Agents Chemother 49:2687–2692. doi: 10.1128/AAC.49.7.2687-2692.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]