PDF
ePub
Citation
Print
INTRODUCTION
Vancomycin is a glycopeptide antimicrobial with an important therapeutic role in treating invasive methicillin-resistant Staphylococcus aureus (MRSA) infection in children in community-associated (CA) and healthcare-associated (HA) settings (1). Because of the significant burden of MRSA infection in hospitals and the community, it is important to use vancomycin appropriately to ensure optimal drug exposure. Though some authors question the usefulness of therapeutic drug monitoring (TDM) of vancomycin and warn of unnecessary hospital costs, appropriate TDM is acknowledged as the most powerful method of adjusting vancomycin use in MRSA bacteremia (2). Studies of the pharmacokinetics and pharmacodynamics (PK/PD) of vancomycin have advocated that a ratio of the area under the curve to the minimum inhibitory concentration (AUC/MIC) of ≥ 400 is optimal for achieving clinical effectiveness in adults (3). This is frequently complemented by a recommended serum vancomycin Ctrough of > 15 µg/mL when the MIC is 1 µg/mL; to avoid the emergence of resistance, it should at least be maintained above 10 µg/mL (4). However, these recommendations are mainly based on guidelines supported by adult data and may not extrapolate to young children.
Despite extensive use of vancomycin in children, information about the optimal regimen to achieve PK/PD targets in the pediatric population remains limited (5). Recent PK/PD studies suggest that routine aggressive dosing may be unnecessary in pediatric invasive MRSA infections, because a Ctrough of 7–10 μg/mL at a dose of 15 mg/kg every 6 hours predicted achievement of AUC/MIC > 400 in > 90% of children infected by MRSA with MIC ≤ 1 μg/mL (6). In addition, the relationship between the Ctrough and AUC in neonates is similar to those in children regardless of gestational age and kidney function (7). Therefore, higher trough concentrations of 15 to 20 µg/mL are likely to be unnecessary in children and neonates based on AUC/MIC considerations (67). Meanwhile vancomycin treatment failure in MRSA bacteremia is most common in premature infants and immunocompromised individuals, even though vancomycin trough serum concentrations ≥ 15 µg/mL are achieved (8).
The aims of this study were to determine whether initial Ctrough could be used as a practical parameter for predicting clinical and microbiological outcomes with the cut-off value at 10 µg/mL, which is the minimum level avoiding the emergence of hetero-resistant vancomycin-intermediate S. aureus (4), and anticipating achievement of AUC/MIC > 400 in pediatric MRSA infection by pharmacokinetic modeling (67).
MATERIALS AND METHODS
Study populations
This retrospective study was conducted at Asan Medical Center Children’s Hospital, which is a 252 bed university-affiliated tertiary hospital located in Seoul, Korea; it has a pediatric intensive care unit (PICU), neonatal intensive care unit (NICU), hematology/oncology ward, surgical ward, and general pediatric wards.
Children aged under 18 years with culture-confirmed MRSA bacteremia, who were hospitalized in our institute during the study period from January 2010 to April 2014, were eligible for enrollment. Other focal infections without bloodstream infection were not included. Patients were included only if they received vancomycin for at least 48 hours and their initial Ctrough had been checked within ≤ 96 hours of start of treatment. Corresponding clinical data including demographic profiles, diagnoses, primary sites of infection, underlying diseases, durations of hospital stay, intensive care unit (ICU) stays, need for mechanical ventilation, and HA risk factors were abstracted retrospectively from electronic medical records. Only the first MRSA isolate detected during a single clinical episode occurring within a 4-week period was included in the analysis, and duplicates from the same patient were excluded. Patients were included regardless of their underlying medical conditions such as congenital heart diseases, genetic/metabolic diseases, hemato-oncological diseases, chronic lung diseases, and neurological disorders. However, patients on renal replacement therapy such as hemodialysis, or neonates admitted to the nursery or NICU were excluded.
Initial dosages of vancomycin (mg/kg/day), which was determined by the reference of ‘Pediatric and Neonatal Dosage Handbook’ (9), and the resulting serum vancomycin concentration data were collected. The first trough sample before the 4th or 5th dose was obtained within 30 minutes prior to the next dose of vancomycin.
Definitions
Fever was defined as any temperature ≥ 38.0°C. The primary focus of infection was defined as the culture-positive site and/or a clinically evident site of infection concomitant with bacteremia, and central line-associated blood stream infection (CLABSI) was defined according to the Centers for Disease Control and Prevention (CDC) guidelines (10). Recurrent MRSA bacteremia was defined as MRSA regrowth on blood cultures after at least one culture-negative month (11). Co-infection was defined as the isolation of an organism in addition to MRSA from the same initial blood culture, or clinical or laboratory evidence of viral infection at the time of isolation of MRSA.
Strain identification and antimicrobial susceptibility testing
Isolates were identified and the antimicrobial susceptibilities of the S. aureus isolates were decided using a MicroScan WalkAway 96-Combo Pos 28 panels (Siemens, West Sacramento, CA, USA). This machine contained six vancomycin wells with concentrations ranging from 0.5 to 16.0 μg/mL and the maximum MIC value within the range was selected to determine MIC50 and MIC90. Antimicrobial susceptibility testing data included 8 antibiotics; gentamycin, ciprofloxacin, trimethoprim/sulfamethoxazole, rifampin, tetracyclin, clindamycin, linezolid, and vancomycin. MRSA isolate of intermediate resistance or full resistance were defined as resistant. The in vitro macrolide-lincosamide-streptogramin B (MLSB)-inducible phenotype was detected by the D-zone test (double-disk diffusion test). Multidrug-resistance (MDR) was defined as acquired non-susceptibility to at least one agent in each of three or more antimicrobial categories (12).
Clinical/microbiological outcomes
To evaluate the clinical and microbiological outcomes according to Ctrough level, the children with MRSA bacteremia were divided into those with initial vancomycin Ctrough < 10 μg/mL and ≥ 10 μg/mL, respectively. As measures of clinical outcome we included resolution of fever after 48 hours of vancomycin use, recurrent MRSA bacteremia, and 30-day all-cause fatality. To evaluate microbiological outcomes, we compared time to negative conversion of blood culture and presence of persistent bacteremia at 48–72 hours of vancomycin administration.
Statistical analysis
While continuous variables were compared using the independent t-tests or analysis of variance (ANOVA), the χ2 test or Fisher’s exact test were used for categorical variables. Risk factors for 30-day mortality or persistent bacteremia at 48 hours were investigated by logistic regression analysis. P values were 2-sided and considered significant at P < 0.05 using SPSS 18.0 (SPSS Inc., Chicago, IL, USA).
RESULTS
Patient characteristics
During the study period from January 2010 to April 2014, a total of 61 cases of MRSA bacteremia occurred in our institute among individuals < 18 years old. Of these, 15 were excluded for the following reasons; 12 patients required hemodialysis or NICU stay, 2 in whom vancomycin Ctrough was not monitored, and one treated with teicoplanin. Finally, 46 episodes of MRSA bacteremia were included in the analysis and the demographic and clinical characteristics of the patients are shown in Table 1. Mean age of patients was 22.0 ± 46.9 months (range, 0–17 years old) and 82.6% (38 out of 46) were aged under 24 months; 54.3% (25 of 46) were male. All were HA MRSA cases and occurred in children with underlying medical conditions, the majority with congenital heart disease. Of the cases, 41.3% were primary MRSA bacteremia while the others involved definite focal infections including CLABSI (41.3%, 19/46), pneumonia (15.2%, 7/46), ventriculoperitoneal shunt infection (4.3%, 2/46) and surgical site infection (4.3%, 2/46); none of them were bone and joint infections or skin and soft tissue infections. Fever was the most frequent initial symptom (34/46, 73.9%) and 16 (34.8%) patients presented with unstable vital signs including hypotension, bradycardia and respiratory difficulty. However, non-specific findings such as vomiting or nausea were observed as an initial symptom in 5 (10.9%) patients, and 1 (2.2%) patient had no clinical symptoms other than a finding of elevated C-reactive protein (CRP). Mean duration of fever to initial positive culture was 1.9 ± 2.4 days.
Table 1
Characteristics of children with MRSA bacteremia according to the initial vancomycin trough concentration (Ctrough)
MRSA = methicillin-resistant Staphylococcus aureus, SD = standard deviation, ICU = intensive care unit, WBC = white blood cell, CRP = C-reactive protein, BUN = blood urea nitrogen, CLABSI, central line-associated blood stream infection.
*In each indicated age group, continuous variables were compared using the independent t-test. Binominal variables were compared using the χ2 test; †Others includes surgical conditions such as congenital diaphragmatic hernia, congenital megacolon, hypospadias, jejunal atresia, tracheoesophageal fistula, omphalocele and other endocrine disease such as hypothyroidism, and pseudohypoaldoteronism; ‡Co-infections included respiratory viral infections such as rhinovirus (n = 2) and adenovirus (n = 1); §Invasive devices included vascular catheters (n = 26), tracheostomy (n = 3), gastrostomy (n = 3), ventriculoperitoneal shunt (n = 2), and colonostomy (n = 1); ∥Unstable vital signs means any of the signs of bradycardia, respiratory difficulty, and hypotension, or requirement of vasopressor; ¶Others includes ventriculo-peritoneal shunt infections (n = 2) and superficial surgical site infections (n = 2).
Clinical manifestations according to vancomycin Ctrough
Initial average dose of vancomycin was 40.6 ± 7.9 mg/kg/day (range, 21.4–63.1 mg/kg/day). The initial Ctrough was 8.0 ± 5.4 μg/mL (range, 2.3–25.6 μg/mL), and only 4 (8.7%) patients achieved the Ctrough ≥ 15 μg/mL. Numbers (%) of children according to range of initial Ctrough were as follows; Ctrough < 5 μg/mL, 14 (30.4%), ≥ 5 μg/mL to < 10 μg/mL, 21 (45.7%), and ≥ 10 μg/mL to < 15 μg/mL, 7 (15.2%).
The patients belonging to the group with initial Ctrough < 10 μg/mL were older (26.6 ± 52.9 months vs. 7.3 ± 8.3 months ; P = 0.045), had a shorter hospital stay prior to MRSA bacteremia (17.2 ± 37.9 days vs. 106.5 ± 179.0 days; P = 0.005) and showed relatively milder clinical courses ; initial presentation with unstable vital signs, ICU stay and requirement for mechanical ventilation were less frequent in the group with Ctrough < 10 μg/mL (22.9% vs. 72.7%, P = 0.002; 45.7% vs. 90.0%, P = 0.013; 42.9% vs. 80.0%, P = 0.038, respectively) (Table 1). However, no significant differences were observed between the groups with initial vancomycin Ctrough < 10 μg/mL and ≥ 10 μg/mL, in the other aspects including presence of underlying disease, primary focus of infection and whether the primary focus was removed or not.
Clinical and microbiological outcomes
Clinical outcomes associated with MRSA bacteremia were analyzed in only 45 cases because one patient was transferred to a care hospital within 14 days from the onset of MRSA bacteremia (Table 2). Severe diseases requiring ICU stay, mechanical ventilation and/or resulting in death were observed in 57.8% (26/45); all-cause 30-day fatality was 11.1% (5/45) with average duration from onset of MRSA bacteremia to death of 11.8 days (range, 2–20 days). Nephrotoxicity defined as an increase in serum creatinine of 0.5 mg/dL from baseline occurred in one patient with initial Ctrough of 6.6 μg/mL; none of the 4 children with high vancomycin troughs of > 15 μg/mL developed nephrotoxicity.
Table 2
Clinical and microbiological outcomes of children with MRSA bacteremia according to initial vancomycin trough concentration (Ctrough) and vancomcyin MIC
MRSA = methicillin-resistant Staphylococcus aureus, MIC = minimum inhibitory concentration.
*Continuous variables were compared by the independent t-tests or analysis of variance (ANOVA), and the [INSERT FIGURE 001]2 test or Fisher’s exact test were used for categorical variables; †Microbiological outcome analysis was done in the available cases because follow-up blood cultures within 72 hours were missed in some cases; ‡Recurrent MRSA infection was defined as MRSA regrowth after at least one month of culture-negativity; §Because some cases presented without fever at the onset of MRSA bacteremia, only febrile cases are included; ∥One patient, who was transferred to care hospital within 14 days from the onset of MRSA bacteremia, could not be included in this analysis.
As follow-up blood culture at 48–72 hours was missed in some cases, 36 of the 46 cases could be included in the analysis of initial microbiological outcomes at 48–72 hours of vancomycin administration; persistent bacteremia at 48–72 hours was observed in 63.9% (23/36) of these patients (Table 2). Average time to negative conversion for the first time was 5.3 ± 4.5 days. MRSA bacteremia recurred in 7 (15.2%) patients after at least one culture-negative month; 3 were CLABSI and 4 had primary bacteremia without a definite focus. All of the recurrent CLABSI cases had a history of salvage therapy without removal of the infected central venous catheter. Mean recurrence interval was 80 days (range, 28–260 days).
Impact of vancomycin Ctrough on clinical and microbiological outcomes
There was no statistically significant difference between the two groups with initial vancomycin Ctrough < 10 μg/mL and ≥ 10 μg/mL, in terms of clinical outcome including 30-day mortality and recurrence (P = 0.899, and P = 0.754, respectively) (Table 2). However, persistent bacteremia at 48 hours was observed more frequently in the group with initial Ctrough < 10 μg/mL (77.8%, compared to 33.3%; P = 0.032), and these patients had a tendency to require longer time for negative conversion (5.3 days, compared to 3.5 days; P = 0.236). Although univariate logistic regression analysis suggested initial Ctrough < 10 μg/mL was associated with persistent bacteremia at 48 hours (odds ratio [OR], 7.00; 95% confidence interval [CI], 1.02–47.97), multivariate logistic regression analysis could not identify statistically significant predictors for persistent bacteremia at 48–72 hours and 30-day fatality (Table 3).
Table 3
Risk factors for persistent bacteremia and 30-day fatality in pediatric MRSA bacteremia
Values are presented as odds ratio (95% confidence interval).
MRSA = methicillin-resistant Staphylococcus aureus, ICU = intensive care unit.
*Univariate logistic regression analysis was carried out including the factors that were significantly different depending on initial Ctrough as presented in Table 1.†Multivariate logistic regression was done adjusting for the any risk factors included in the univariate analysis. ‡All of the fatal cases required mechanical ventilation.
In a subgroup analysis of children aged under 24 months, which included 82.6% (38/46) of the study population, 73.7% (28/38) had an initial Ctrough < 10 μg/mL; all-cause 30-day fatality was 8.1% (3/37). Persistent bacteremia at 48 hours was observed more frequently when initial Ctrough was below 10 μg/mL (78.3%, compared to 33.3% in the group with Ctrough ≥ 10 μg/mL; P = 0.034).
Antimicrobial susceptibilities
Among the 46 MRSA isolates, the multi-drug resistance rate was 46.9% and 69.6% were resistant to erythromycin. With an erythromycin-inducible clindamycin resistance rate of 88.9% (8/9), the overall rate of resistance to clindamycin was 52.2% (24/46). Frequencies of resistance to trimethoprim/sulfamethoxazole and rifampin were 2.2% (1/46) and 4.3% (2/46), respectively. The vancomycin MIC50 and MIC90 were 1.0 and 2.0 μg/mL, respectively. MRSA isolates with vancomycin MIC < 1.0 μg/mL, 1.0 μg/mL, and > 1.0 μg/mL comprised 26.1% (12/46), 58.7% (27/46) and 15.2% (7/46), respectively, and no MRSA isolates with vancomcyin MIC > 2.0 μg/mL were detected.
The clinical and microbiological outcomes of the children with MRSA bacteremia according to the vancomycin MIC are shown in Table 2, however, vancomycin MIC did not significantly influence clinical and microbiologic outcomes of MRSA bacteremia in this study.
DISCUSSION
Our findings suggest that initial Ctrough < 10 μg/mL could be used as a predictor of microbiological failure at 48 hours of vancomycin therapy for children with MRSA bacteremia, even though multivariate logistic regression analysis could not identify statistical significance. However, initial Ctrough < 10 μg/mL did not significantly influence 30-day mortality or recurrence even when we adjusted for clinical status at presentation. This may be partly due to the better clinical outcomes in pediatric MRSA bacteremia than in adult cases. Indeed, overall 30-day mortality associated with MRSA bacteremia was 11.1% in this study, which was much lower than that in adult patients (13). A recent prospective cohort study reported a 25.6%, 30-day mortality of invasive MRSA infections in a Korean population with median age 64 years (14). It may be pointless to monitor initial vancomycin Ctrough to optimize final clinical outcomes in pediatric MRSA bacteremia, in which, compared to adult cases, relatively favorable outcomes can be expected regardless of initial vancomycin Ctrough.
The AUC of vancomycin is a well-known theoretical PK/PD model reflecting the effectiveness of vancomycin exposure (15). Because it is nearly impossible to measure an accurate AUC in clinical practice especially in young children, it is important to develop an appropriate PK/PD model. In our study, more than 80% of our study population was less than 24 months old and no very appropriate model for calculating AUC was available in this age group; estimation of an AUC using vancomycin clearance and creatinine clearance based on the model developed by Frymoyer et al. (16), was originally developed to predict vancomycin AUC24/MIC in children aged 2 to 12 years. Although a Bayesian model employing the serum creatinine value and one set of trough and peak vancomycin concentrations has been developed to estimate a PK model for vancomycin in young children (1718), our study was conducted retrospectively and the standard of care in our institute involves checking only trough levels during vancomycin therapy.
There is controversy regarding the association of vancomycin MIC and clinical outcomes, in some reports higher vancomycin MIC value had a higher likelihood of treatment failure (19) in others not (2021). In this study, vancomycin MIC did not impact on the clinical and microbiologic outcomes of MRSA bacteremia in young children.
This study had some limitations. First, it was retrospective and included a relatively small number of patients and about one-fourth of the cases of MRSA bacteremia could not be included in the initial microbiological outcome analysis because of missing follow-up blood cultures within 72 hours, even though it is recommended that follow-up blood cultures should be conducted every 48 to 72 hours until a negative result is obtained in our institute. Second, during the study period from January 2010 to April 2014, the initial dose of vancomycin was usually 40 mg/kg/day—with some exceptions such as central nervous system (CNS) infections where a starting dose of 60 mg/kg/day was used based on the Pediatric and Neonatal Dosage Handbook (9)—even though this is suboptimal according to the recent recommendation of a daily dose of 60 mg/kg in children. Since 2012–2013, however, the recommended initial dose of vancomycin was increased from 40 mg/kg/day to 60 mg/kg/day with further modification according to the TDM report in cases of invasive MRSA infection regardless of CNS or non CNS infection. Vancomycin is a well-known age-dependent pharmacokinetic antibiotic, because creatinine clearance has a crucial role in vancomycin pharmacodynamics and the vancomycin clearance of children is 2 or 3 times higher than that of adults (1419). As a consequence, it was hard to obtain the adult target range in most children with the dosage of 40 mg/kg/day. Thirdly, the paucity of appropriate models for calculating AUCs given only the serum trough values of vancomycin in younger children under 2 years old made it difficult to determine whether the PK/PD parameters were comparable to the data obtained in older population groups. However, there have been few clinical studies to determine the impact of the vancomycin Ctrough and MIC on clinical and microbiological outcomes in MRSA bacteremia among young children < 2 years old, and the only available literature being theoretical PK/PD models predicting clinical outcomes in pediatric MRSA infections (622).
In conclusion, initial Ctrough may be a useful TDM parameter in pediatric MRSA bacteremia for reducing early microbiological failure, though it did not impact on final clinical outcomes including 30-day fatality. Further prospective data on vancomycin dosing are needed to find the optimal drug exposure and clarify its impact on clinical outcomes in pediatric populations.
XML Download