Journal of Infection (2014) xx, 1e6
www.elsevierhealth.com/journals/jinf
Determination of vancomycin trough level in serum and cerebrospinal fluid of patients with acute community-acquired meningitis: A prospective study Shervin Shokouhi a, Ilad Alavi Darazam a,b,* a
Infectious Diseases and Tropical Medicine Research Centre, Shahid Beheshti University of Medical Sciences, Tehran, Iran b NRITLD, National Research Institute of Tuberculosis and Lung Diseases, Shahid Beheshti University of Medical Sciences, Tehran, Iran Accepted 20 June 2014 Available online - - -
KEYWORDS Meningitis; Acute; Human; Vancomycin; Chromatography; High pressure liquid; Vancomycin/ administration/dosage; Vancomycin/ pharmacokinetics; Vancomycin/ therapeutic use; Drug monitoring/ methods
Summary Background: Penetration and concentration of vancomycin is still an elusive and complex issue particularly in Cerebrospinal Fluid (CSF). The aim of this study was to clarify the penetration of this antimicrobial agent in CSF during meningeal inflammation. Methods: In a prospective study, adult patients, with clinical and CSF analysis compatible with acute meningitis, who received vancomycin (15 mg/kg loading and 30 mg/kg daily maintenance dose) with ceftriaxone (4 gr/daily) were enrolled. CSF analysis including vancomycin trough levels before the fourth maintenance dose and during the 8e10th days of treatment, and simultaneous serum levels were performed by High-Pressure Liquid Chromatography (HPLC). Results: Twenty-seven patients (18 men, 9 women; mean age of 39.4 14.7) were enrolled. The first serum trough level of vancomycin was 13.82 1.28 mg/l. The mean of corresponding trough level in CSF was 11.2 1.41 mg/l. The serum and CSF trough levels revealed positive linear correlation (r: 0.60) and was significant at the 0.01 level (P: 0.004). The penetration CSF/serum ratio was 0.811 0.082 (coefficient of variation: 10.1%). The second trough levels of serum and CSF in (14 patients) vancomycin were 13.32 1.02 and 10.64 1.21, respectively. The serum and CSF trough levels revealed positive linear correlation (r: 0.71). The serum and CSF concentrations revealed no variation compared to the first trough levels.
* Corresponding author. Loghman Hospital, Kamali Ave., Tehran, Iran. Tel.: þ98 914 149 1958; fax: þ98 21 55416170. E-mail address: [email protected] (I. Alavi Darazam). http://dx.doi.org/10.1016/j.jinf.2014.06.010 0163-4453/ª 2014 The British Infection Association. Published by Elsevier Ltd. All rights reserved. Please cite this article in press as: Shokouhi S, Alavi Darazam I, Determination of vancomycin trough level in serum and cerebrospinal fluid of patients with acute community-acquired meningitis: A prospective study, J Infect (2014), http://dx.doi.org/10.1016/ j.jinf.2014.06.010
2
S. Shokouhi, I. Alavi Darazam Conclusion: Vancomycin has appropriate concentration in CSF during the treatment of meningitis and do not decrease along with the alleviation of meningeal inflammation in spite of concerns in this regard. ª 2014 The British Infection Association. Published by Elsevier Ltd. All rights reserved.
Introduction Acute meningitis, particularly with a bacterial cause, is a well-known life threatening condition. It is a true medical emergency and requires prompt diagnosis and treatment. Recently, the emergence of resistant pathogens has influenced the medical strategies and selection of antimicrobial agents. Vancomycin and cephalosporins are currently the antibiotics of choice for empirical therapy of acute meningitis. Vancomycin is a large glycopeptide with a high molecular weight.1 Its oral absorption is not significant, and the main route of elimination is renal excretion.2 The pharmacokinetic profile of vancomycin is complex and there are still many unresolved issues. In patients with normal glomerular filtration rate, vancomycin seems to have a maximum distribution phase of 1 h, a volume of distribution of 0.4e1 l/kg, and an elimination half-life of 6e12 h3 The range of protein binding is estimated to be 10e50%.3 Constitutionally, vancomycin is a timedependent antibiotic, and the activity is influenced by several factors including the distribution, the inoculum size, and the protein-binding effects. Vancomycin penetrates into most of the body spaces with variable concentrations, depending, to some extent, on the degree of inflammation.3 Since the Cerebrospinal Fluid (CSF) penetration of hydrophilic antibiotics including penicillins, cephalosporins, or vancomycin is strongly influenced by meningeal infiammation, there are some concerns over the administration of high-dose steroids.4,5 Vancomycin was reported to penetrate into the inflamed meninges, with CSF concentrations of 6.4e11.1 mg/l and CSF-to-serum ratios of 0.36e0.48.3 In addition to the anti-inflammatory agents (steroids), abating of the meningeal inflammation during the treatment may also influence the CSF penetration of antibiotics. In patients with intact meninges, vancomycin penetration into the CSF was reported to be fairly low, as it was shown by a CSF concentrations of 0e3.45 mg/l, with corresponding CSF-to-serum ratios of 0e0.18.3 There have been very few human studies on the pharmacodynamics of vancomycin, most of them with inconclusive findings in determining parameters reliably predicting patients’ outcomes. The majority of these studies have involved relatively small populations of patients with a variety of infection types. The therapeutic range often quoted for vancomycin monitoring is peak and trough serum concentrations of 30e40 and 5e10 mg/ l, respectively.3 However, it should be noted that the practice of routine monitoring and adjusting of serum vancomycin concentrations has been the subject of intense debate for many years.6 The controversy has resulted from conflicting evidence on the use of serum vancomycin concentrations for predicting drug-induced toxicity
and also as a measure of effectiveness in treating infections.6 Since vancomycin is a time-dependent antibiotic and there are practical issues associated with determining a precise peak serum concentration, most clinicians have abandoned the routine practice of determining peak serum concentrations and relied solely on monitoring trough serum concentrations for this antibiotic.3 Owing to the lack of data and presence of substantial differences among the previous studies regarding the setting, patients, and applied methods for therapeutic drug monitoring, this study was conducted to determine CSF vancomycin concentrations in patients with acute meningitis and its connection with serum levels of the drug during the therapy.
Materials and methods This prospective study was carried out in Loghman Hospital, Tehran, Iran in 2012 and 2013. All patients aged 18e60 years who were admitted to our center with clinical presentations suggestive of acute meningitis including fever (>38 C), headache, nausea and/or vomiting, meningeal irritation signs, and nuchal rigidity with or without alternation in consciousness were recruited to the study. For all of the patients, the time between initiation and stabilization of the signs and symptoms was less than 24 h. The clinical presentation of the patients was associated with pleocytosis (polymorphonuclear (PMN) cell count >1 mm3), with or without hypoglycorrhachia (CSF glucose 50 mg/dl) or bacteria in direct gram smear. All of the patients had normal creatinine clearance (>90 ml/min) based on estimation equation (CockrofteGault formula).7 They received empirically vancomycin (15 mg/kg loading dose followed by 30 mg/kg daily maintenance divided into two doses, infused intravenously within 1 h) and ceftriaxone (2 g, twice daily, intravenously) without dexamethasone. Patients with other treatment regimens or those who underwent escalation or deescalation of antimicrobial agents were excluded from the study. Other exclusion or withdrawal criteria were as follow: primary or secondary immune deficiency including HIV infection of any CD4 count, patients with solid or hematopoietic transplants, patients receiving steroids or other anti-inflammatory drugs, exclusion of acute meningitis during the therapy, initiation of antibiotics before admission for recent manifestations, hospitalization during the last three months before admission, and disinclination of the patients to stay in the study. To determine vancomycin trough level during the therapy, CSF and serum samples were obtained from the study patients 15e30 min prior to vancomycin infusion on the day 4 and between the days 8e10 of the treatment.
Please cite this article in press as: Shokouhi S, Alavi Darazam I, Determination of vancomycin trough level in serum and cerebrospinal fluid of patients with acute community-acquired meningitis: A prospective study, J Infect (2014), http://dx.doi.org/10.1016/ j.jinf.2014.06.010
Vancomycin trough level in CSF Vancomycin concentration was measured by High Performance Liquid Chromatography (Agilent Technologies, Inc.). Sample preparation was based on protein precipitation followed by ultrafiltration. In order to minimize differential modulation of ionization by matrix constituents extended chromatographic separation was applied leading to a retention time of 9.8 min for the analyte. Measurement was done by HPLC-ESI-MS/MS. For internal standardization the derivative vancomycineglycin (ISTD) prepared by chemical synthesis was used, HPLC conditions ensured coelution of ISTD with the analyte. The correlation between serum and CSF trough level of vancomycin was evaluated by Pearson’s r coefficient. Paired-sample t-test was applied to assess the variations between the first and second CSF and serum trough levels. Data were presented as mean standard deviation. P value less than 0.05 was considered to be statistically significant. SPSS version 16 (Chicago, IL) was used for statistical analysis. The purpose and procedures of this study were explained to the patients, or to their proxies in case of patients’ incompetence, and then formal informed consents were obtained. This study was registered in the Infectious Diseases and Tropical Medicine Research Centre affiliated to Shahid Beheshti University of Medical Sciences and approved by ethical committee of the university (official approval number: 90-1-105-8150-8344, letter dated 8 April 2012).
Results Twenty-seven patients (18 men and 9 women; mean age 39.4 14.7 years) with presumptive clinical diagnosis of acute meningitis associated with pleocytosis and/or hypoglycorrhachia were included. The CSF leakage was the underlying cause of meningitis in three patients following trauma, but there was no other significant underlying factor in the remaining patients. The mean (SD) leucocyte count, glucose and protein concentrations in the CSF were 2525/mm3, 58.3 39.3 mg/dl and 124.15 179.2 mg/dl, respectively. The trough level of vancomycin, 15e30 min before the fourth maintenance dose, was 13.82 1.28 mg/l in serum and 11.2 1.41 mg/l in CSF. The serum and CSF trough level displayed a positive linear correlation (r Z 0.60, P Z 0.004, Fig. 1). The CSF-to-serum concentration ratio was 0.811 0.082 (coefficient of variation: 10.1%). Analysis of CSF trough level based on the elements of meningeal inflammation showed no significant correlation with the leucocyte count, glucose and protein concentrations. The mean creatinine concentration and clearance were 0.99 0.22 mg/dl and >90 ml/min, respectively; and none of the patients had a creatinine concentration more than 1.4 mg/dl. No significant correlation was found between serum creatinine levels and serum or CSF vancomycin trough level (P Z NS). After 2e3 days of the treatment, 24 patients showed defervescence and amelioration of initial symptoms and signs including loss of consciousness, meningeal irritation signs and nausea. However, 3 of the patients had CSF vancomycin concentrations under the 8.3 mg/l and
3
Figure 1 Correlation between the first serum and cerebrospinal fluid trough level (r: 0.60, P: 0.001), 27 samples. The filled black dot shows two similar data.
remained febrile with no remarkable improvement in consciousness. Evaluation of delayed response ruled out probable complications. One of these patients received improperly a higher dosage of vancomycin relative to his body weight, giving rise to increased serum and CSF trough level of 13.5 and 10.5 mg/l, respectively (the initial tough levels were 7.5 and 6.8 mg/l). He improved two days later and became afebrile. The remaining two patients were treated with no change in the initial vancomycin dosage, and finally responded to the treatment regimen approximately after one week. During the treatment phase, only one patient experienced increases of serum creatinine levels that was resolved by parenteral fluid resuscitation, and none of the patients had subjective hearing loss (see Table 1). Fourteen patients underwent another serum and CSF analysis for vancomycin levels between the days 8e10 of vancomycin treatment, 15e30 min before the next dose. The mean trough level of vancomycin in serum and CSF were 13.32 1.02 and 10.64 1.21 mg/l, respectively. A positive linear correlation existed between the serum and CSF trough level of the drug (r Z 0.71, P Z 0.004, Fig. 2). The serum and CSF concentrations of vancomycin showed no significant variation on days 8e10 of treatment relative to the day 4. The CSF-to-serum ratio was 0.79 0.063 (coefficient of variation: 8.3%). Although, the second CSF analysis showed decreases in the leukocyte count and protein concentrations and increases in the glucose levels compared to the first analysis, but none of these changes were statistically significant (P Z NS). There was no correlation between these CSF elements and the vancomycin concentrations on the second analysis. At the end of the treatment course, all of the patients appeared to be in the normal status, except for one. A 49year-old opiate dependent man, without underlying comorbidity, experienced sudden death on the day 7 of the treatment, in spite of dramatic improvements after 2 days. The cause of death remained unresolved due to no consent of his family to autopsy (see Table 2).
Please cite this article in press as: Shokouhi S, Alavi Darazam I, Determination of vancomycin trough level in serum and cerebrospinal fluid of patients with acute community-acquired meningitis: A prospective study, J Infect (2014), http://dx.doi.org/10.1016/ j.jinf.2014.06.010
4
S. Shokouhi, I. Alavi Darazam Table 1
The summary of CSF analysis and corresponding serum level of vancomycin levels.
LAB analysis
The leukocyte The glucose The protein count of CSF of CSF (mg/dl) of CSF (cell/mm3) (mg/dl)
Before the 4th dose 2525 8860 During 8e10th days 235 561
58.27 39.3 51.32 29.1
Serum trough level CSF trough (mg/l) level (mg/l)
124.15 179.2 13.82 1.28 68.88 37.4 13.32 1.02
Discussion The main finding of this study was the relative stability of the vancomycin concentration in CSF of the patients with acute meningitis over the treatment course. The diagnosis of acute meningitis was based on the clinical presentation and CSF analysis that showed the inflammatory process of meninges. However, the etiology and microbiology of acute meningitis in the study subjects and their effects on the therapeutic drug monitoring were not evaluated in this study. Nevertheless, it should be noted that the main concern about the vancomycin penetration into the CSF has been related to meningeal inflammation regardless of the etiological considerations. Therefore, we focused only on meningeal inflammation and its effect on CSF vancomycin concentrations. Despite abundant information about the monitoring of vancomycin levels in serum, particularly since 1980s, data on the pharmacokinetics of vancomycin in the CSF are lacking and limited to a few studies with small sample sizes. In 2009, the Infectious Diseases Society of America (IDSA) and the Society of Infectious Diseases Pharmacists published a consensus review suggesting therapeutic monitoring of vancomycin in adult patients and maintaining a total vancomycin serum trough concentrations of 15e20 mg/l for complicated infections such as bacteremia, endocarditis, osteomyelitis, meningitis, and hospitalacquired pneumonia caused by Staphylococcus aureus.6 However, there was no clear background for this recommendation. Moreover, in this paper, the concentration of
Figure 2 Correlation between the second serum and cerebrospinal fluid trough level (r: 0.71, P: 0.004), 14 samples.
CSF/serum trough ratio
11.2 1.41 0.811 0.082 10.64 1.21 0.79 0.063
vancomycin in the CSF was referred to a single study that originally had no data about CSF vancomycin.6 However, this was the only authentic consensus review about vancomycin monitoring and dosage so far. The vancomycin levels in the serum and CSF samples was measured using HPLC method. It is also possible to measure vancomycin concentrations by the Fluorescent Polarization Immunoassays (FPIA), but there are concerns about the efficacy and accuracy of this method due to overestimation.8,9 On the other hand, the HPLC method has high sensitivity and specificity, but it is time consuming and generally labor-intensive.10 The HPLC measurements results was reported to be either higher or lower than the immunoassay measurements in different times, albeit non-significant.9 Initially, we used the FPIA method to analyze vancomycin trough levels, but interestingly, the results were very unacceptable, particularly for CSF (data not shown). This prompted us to change the assay protocol to the HPLC method. On the basis of our experience, we assume that one of the most important sources of variations in studies about vancomycin levels might be due to the different measurement methods. Currently, the British National Formulary (BNF) recommendations on vancomycin trough concentrations are 10e15 mg/l, and 15e20 mg/l for more resistant strains. Similar levels have also been suggested for pneumonia and meningitis, whereas continuous infusion of vancomycin to achieve an average steady state concentration of 15e25 mg/l has been proposed for critically ill patients.11 In a study conducted in ICU setting on 13 patients with vancomycin sensitive infections, the effect of continuous infusions of the drug on its serum levels was evaluated by daily measurements of the vancomycin with FPIA, and efforts was made to maintain its serum values between 20 and 30 mg/l. Their results showed that patients with meningitis had maximal CSF vancomycin levels ranging from 5.7 to 19.0 mg/l (mean, 11.1 4.9 mg/l), with a mean CSF-to-serum ratio of 48%. On the other hand, patients with pathologies other than meningitis had maximal CSF levels ranging from 4.89 to 2.42 mg/l (mean, 3.45 1.11 mg/l), with a mean CSF-to-serum ratio of 18%. These data indicate a more effective CSF penetration of the drug in infiamed meningeal tissues. The CSF-toserum concentration ratio was twice as high for the group of patients with meningitis as for the other group (48 se J. and co-workers reversus 18%).10 Additionally, Albane ported that the clinical and bacteriological efficacy of treatment by continuous vancomycin infusions was high. They recommended the administration of high doses of vancomycin at a mean of 62 mg/kg/day to obtain stable concentrations of 25e30 mg/l in serum which ensure CSF levels of 6e19 mg/l.12
Please cite this article in press as: Shokouhi S, Alavi Darazam I, Determination of vancomycin trough level in serum and cerebrospinal fluid of patients with acute community-acquired meningitis: A prospective study, J Infect (2014), http://dx.doi.org/10.1016/ j.jinf.2014.06.010
Vancomycin trough level in CSF Table 2
1 2 3 4 5 6 7 8 9 10 11 12 13 14
5
Summary of the characteristics of fourteen patients.
Gender
1st Serum trough (mg/l)
1st CSF trough (mg/l)
2nd Serum trough (mg/l)
2nd CSF trough (mg/l)
1st CSF/serum trough
2nd CSF/serum trough
Male Female Male Male Male Male Male Male Male Female Female Male Male Male
15.5 13.9 13.5 13.5 12.5 13.5 15.0 11.6 13.8 13.2 13.4 15.9 12.6 10.9
12.6 9.9 13.0 11.2 9.2 10.8 8.3 10.3 11.9 11.9 9.8 11.7 10.3 7.5
14.6 12.7 14.6 14.2 13.5 13.2 14.2 13.9 13.2 13.8 13.4 11.6 11.9 11.7
13.5 11.9 11.5 10.9 10.9 10.8 10.8 10.7 10.5 10.3 9.7 9.5 9.3 8.6
1.23 1.4 1.03 1.20 1.35 1.25 1.80 1.12 1.16 1.11 1.37 1.36 1.22 1.45
1.08 1.07 1.27 1.30 1.24 1.22 1.31 1.3 1.26 1.34 1.38 1.22 1.28 1.36
The first clinical trial to show the relations between serum and CSF levels of vancomycin in patients with bacterial meningitis was carried out by Ricard et al. in 2007 wherein vancomycin concentrations were measured by FPIA in the serum and CSF samples obtained on the day 2 or 3 of the antimicrobial therapy. They showed that adequate levels of vancomycin in the CSF was obtained by administration of high-dose systemic vancomycin in patients receiving adjunctive dexamethasone treatment (mean, 7.9 mg/l, or 30% of the serum concentrations). The mean serum concentration of vancomycin was 25.2 mg/l (ranging from 14.2 to 39.0 mg/l). Moreover, a significant correlation was found between vancomycin concentrations in the serum and CSF (r Z 0.68). They also found that vancomycin penetration into the CSF was connected with CSF protein levels a marker of increased bloodebrain barrier permeability caused by meningeal inflammation.4 In our study, patients did not receive steroids as an adjunct, and the dose of vancomycin was lower. However, the levels of vancomycin proved to be higher. Various sampling times and different assay methods may explain these variations. In a consensus review, Thompson et al. recommended that in patients with CNS shunt infections the usual intravenous regimen of 60 mg/kg/day should be divided and administered every 6 h. They also suggested that intraventricular vancomycin should be considered for most patients, starting with 10 mg/day. Additionally, CSF vancomycin concentrations should be monitored and dose adjustments should be made as needed to maintain CSF trough vancomycin concentrations between 5 and 20 mg/ l.13 Although, these results are based on post-surgical infections and all of patients have had CNS intervention and meningeal device. In a trial, simultaneous serum and CSF samples were obtained at 4 h of the therapy to determine the minimal levels of vancomycin maintained in the CSF throughout the initial period of treatment. Additional CSF and serum sampling were also performed at the end of therapy on day 10, 30e60 min following the last vancomycin infusion to determine the final CSF values of leukocyte count, protein
and glucose concentrations, peak vancomycin levels, and bactericidal titers. CSF leukocyte count was 0e51 per mm3 (median, 9), and protein and glucose concentrations in the CSF were 0.30e0.84 (median, 0.49) g/l and 46‒64 (median, 54) mg/dl, respectively. Peak vancomycin levels in CSF and serum were undetectable to 2.6 (median, 1.9) mg/ml and 15‒25 (median, 18.5) mg/ml, respectively. Peak CSF bactericidal titers ranged from
www.elsevierhealth.com/journals/jinf
Determination of vancomycin trough level in serum and cerebrospinal fluid of patients with acute community-acquired meningitis: A prospective study Shervin Shokouhi a, Ilad Alavi Darazam a,b,* a
Infectious Diseases and Tropical Medicine Research Centre, Shahid Beheshti University of Medical Sciences, Tehran, Iran b NRITLD, National Research Institute of Tuberculosis and Lung Diseases, Shahid Beheshti University of Medical Sciences, Tehran, Iran Accepted 20 June 2014 Available online - - -
KEYWORDS Meningitis; Acute; Human; Vancomycin; Chromatography; High pressure liquid; Vancomycin/ administration/dosage; Vancomycin/ pharmacokinetics; Vancomycin/ therapeutic use; Drug monitoring/ methods
Summary Background: Penetration and concentration of vancomycin is still an elusive and complex issue particularly in Cerebrospinal Fluid (CSF). The aim of this study was to clarify the penetration of this antimicrobial agent in CSF during meningeal inflammation. Methods: In a prospective study, adult patients, with clinical and CSF analysis compatible with acute meningitis, who received vancomycin (15 mg/kg loading and 30 mg/kg daily maintenance dose) with ceftriaxone (4 gr/daily) were enrolled. CSF analysis including vancomycin trough levels before the fourth maintenance dose and during the 8e10th days of treatment, and simultaneous serum levels were performed by High-Pressure Liquid Chromatography (HPLC). Results: Twenty-seven patients (18 men, 9 women; mean age of 39.4 14.7) were enrolled. The first serum trough level of vancomycin was 13.82 1.28 mg/l. The mean of corresponding trough level in CSF was 11.2 1.41 mg/l. The serum and CSF trough levels revealed positive linear correlation (r: 0.60) and was significant at the 0.01 level (P: 0.004). The penetration CSF/serum ratio was 0.811 0.082 (coefficient of variation: 10.1%). The second trough levels of serum and CSF in (14 patients) vancomycin were 13.32 1.02 and 10.64 1.21, respectively. The serum and CSF trough levels revealed positive linear correlation (r: 0.71). The serum and CSF concentrations revealed no variation compared to the first trough levels.
* Corresponding author. Loghman Hospital, Kamali Ave., Tehran, Iran. Tel.: þ98 914 149 1958; fax: þ98 21 55416170. E-mail address: [email protected] (I. Alavi Darazam). http://dx.doi.org/10.1016/j.jinf.2014.06.010 0163-4453/ª 2014 The British Infection Association. Published by Elsevier Ltd. All rights reserved. Please cite this article in press as: Shokouhi S, Alavi Darazam I, Determination of vancomycin trough level in serum and cerebrospinal fluid of patients with acute community-acquired meningitis: A prospective study, J Infect (2014), http://dx.doi.org/10.1016/ j.jinf.2014.06.010
2
S. Shokouhi, I. Alavi Darazam Conclusion: Vancomycin has appropriate concentration in CSF during the treatment of meningitis and do not decrease along with the alleviation of meningeal inflammation in spite of concerns in this regard. ª 2014 The British Infection Association. Published by Elsevier Ltd. All rights reserved.
Introduction Acute meningitis, particularly with a bacterial cause, is a well-known life threatening condition. It is a true medical emergency and requires prompt diagnosis and treatment. Recently, the emergence of resistant pathogens has influenced the medical strategies and selection of antimicrobial agents. Vancomycin and cephalosporins are currently the antibiotics of choice for empirical therapy of acute meningitis. Vancomycin is a large glycopeptide with a high molecular weight.1 Its oral absorption is not significant, and the main route of elimination is renal excretion.2 The pharmacokinetic profile of vancomycin is complex and there are still many unresolved issues. In patients with normal glomerular filtration rate, vancomycin seems to have a maximum distribution phase of 1 h, a volume of distribution of 0.4e1 l/kg, and an elimination half-life of 6e12 h3 The range of protein binding is estimated to be 10e50%.3 Constitutionally, vancomycin is a timedependent antibiotic, and the activity is influenced by several factors including the distribution, the inoculum size, and the protein-binding effects. Vancomycin penetrates into most of the body spaces with variable concentrations, depending, to some extent, on the degree of inflammation.3 Since the Cerebrospinal Fluid (CSF) penetration of hydrophilic antibiotics including penicillins, cephalosporins, or vancomycin is strongly influenced by meningeal infiammation, there are some concerns over the administration of high-dose steroids.4,5 Vancomycin was reported to penetrate into the inflamed meninges, with CSF concentrations of 6.4e11.1 mg/l and CSF-to-serum ratios of 0.36e0.48.3 In addition to the anti-inflammatory agents (steroids), abating of the meningeal inflammation during the treatment may also influence the CSF penetration of antibiotics. In patients with intact meninges, vancomycin penetration into the CSF was reported to be fairly low, as it was shown by a CSF concentrations of 0e3.45 mg/l, with corresponding CSF-to-serum ratios of 0e0.18.3 There have been very few human studies on the pharmacodynamics of vancomycin, most of them with inconclusive findings in determining parameters reliably predicting patients’ outcomes. The majority of these studies have involved relatively small populations of patients with a variety of infection types. The therapeutic range often quoted for vancomycin monitoring is peak and trough serum concentrations of 30e40 and 5e10 mg/ l, respectively.3 However, it should be noted that the practice of routine monitoring and adjusting of serum vancomycin concentrations has been the subject of intense debate for many years.6 The controversy has resulted from conflicting evidence on the use of serum vancomycin concentrations for predicting drug-induced toxicity
and also as a measure of effectiveness in treating infections.6 Since vancomycin is a time-dependent antibiotic and there are practical issues associated with determining a precise peak serum concentration, most clinicians have abandoned the routine practice of determining peak serum concentrations and relied solely on monitoring trough serum concentrations for this antibiotic.3 Owing to the lack of data and presence of substantial differences among the previous studies regarding the setting, patients, and applied methods for therapeutic drug monitoring, this study was conducted to determine CSF vancomycin concentrations in patients with acute meningitis and its connection with serum levels of the drug during the therapy.
Materials and methods This prospective study was carried out in Loghman Hospital, Tehran, Iran in 2012 and 2013. All patients aged 18e60 years who were admitted to our center with clinical presentations suggestive of acute meningitis including fever (>38 C), headache, nausea and/or vomiting, meningeal irritation signs, and nuchal rigidity with or without alternation in consciousness were recruited to the study. For all of the patients, the time between initiation and stabilization of the signs and symptoms was less than 24 h. The clinical presentation of the patients was associated with pleocytosis (polymorphonuclear (PMN) cell count >1 mm3), with or without hypoglycorrhachia (CSF glucose 50 mg/dl) or bacteria in direct gram smear. All of the patients had normal creatinine clearance (>90 ml/min) based on estimation equation (CockrofteGault formula).7 They received empirically vancomycin (15 mg/kg loading dose followed by 30 mg/kg daily maintenance divided into two doses, infused intravenously within 1 h) and ceftriaxone (2 g, twice daily, intravenously) without dexamethasone. Patients with other treatment regimens or those who underwent escalation or deescalation of antimicrobial agents were excluded from the study. Other exclusion or withdrawal criteria were as follow: primary or secondary immune deficiency including HIV infection of any CD4 count, patients with solid or hematopoietic transplants, patients receiving steroids or other anti-inflammatory drugs, exclusion of acute meningitis during the therapy, initiation of antibiotics before admission for recent manifestations, hospitalization during the last three months before admission, and disinclination of the patients to stay in the study. To determine vancomycin trough level during the therapy, CSF and serum samples were obtained from the study patients 15e30 min prior to vancomycin infusion on the day 4 and between the days 8e10 of the treatment.
Please cite this article in press as: Shokouhi S, Alavi Darazam I, Determination of vancomycin trough level in serum and cerebrospinal fluid of patients with acute community-acquired meningitis: A prospective study, J Infect (2014), http://dx.doi.org/10.1016/ j.jinf.2014.06.010
Vancomycin trough level in CSF Vancomycin concentration was measured by High Performance Liquid Chromatography (Agilent Technologies, Inc.). Sample preparation was based on protein precipitation followed by ultrafiltration. In order to minimize differential modulation of ionization by matrix constituents extended chromatographic separation was applied leading to a retention time of 9.8 min for the analyte. Measurement was done by HPLC-ESI-MS/MS. For internal standardization the derivative vancomycineglycin (ISTD) prepared by chemical synthesis was used, HPLC conditions ensured coelution of ISTD with the analyte. The correlation between serum and CSF trough level of vancomycin was evaluated by Pearson’s r coefficient. Paired-sample t-test was applied to assess the variations between the first and second CSF and serum trough levels. Data were presented as mean standard deviation. P value less than 0.05 was considered to be statistically significant. SPSS version 16 (Chicago, IL) was used for statistical analysis. The purpose and procedures of this study were explained to the patients, or to their proxies in case of patients’ incompetence, and then formal informed consents were obtained. This study was registered in the Infectious Diseases and Tropical Medicine Research Centre affiliated to Shahid Beheshti University of Medical Sciences and approved by ethical committee of the university (official approval number: 90-1-105-8150-8344, letter dated 8 April 2012).
Results Twenty-seven patients (18 men and 9 women; mean age 39.4 14.7 years) with presumptive clinical diagnosis of acute meningitis associated with pleocytosis and/or hypoglycorrhachia were included. The CSF leakage was the underlying cause of meningitis in three patients following trauma, but there was no other significant underlying factor in the remaining patients. The mean (SD) leucocyte count, glucose and protein concentrations in the CSF were 2525/mm3, 58.3 39.3 mg/dl and 124.15 179.2 mg/dl, respectively. The trough level of vancomycin, 15e30 min before the fourth maintenance dose, was 13.82 1.28 mg/l in serum and 11.2 1.41 mg/l in CSF. The serum and CSF trough level displayed a positive linear correlation (r Z 0.60, P Z 0.004, Fig. 1). The CSF-to-serum concentration ratio was 0.811 0.082 (coefficient of variation: 10.1%). Analysis of CSF trough level based on the elements of meningeal inflammation showed no significant correlation with the leucocyte count, glucose and protein concentrations. The mean creatinine concentration and clearance were 0.99 0.22 mg/dl and >90 ml/min, respectively; and none of the patients had a creatinine concentration more than 1.4 mg/dl. No significant correlation was found between serum creatinine levels and serum or CSF vancomycin trough level (P Z NS). After 2e3 days of the treatment, 24 patients showed defervescence and amelioration of initial symptoms and signs including loss of consciousness, meningeal irritation signs and nausea. However, 3 of the patients had CSF vancomycin concentrations under the 8.3 mg/l and
3
Figure 1 Correlation between the first serum and cerebrospinal fluid trough level (r: 0.60, P: 0.001), 27 samples. The filled black dot shows two similar data.
remained febrile with no remarkable improvement in consciousness. Evaluation of delayed response ruled out probable complications. One of these patients received improperly a higher dosage of vancomycin relative to his body weight, giving rise to increased serum and CSF trough level of 13.5 and 10.5 mg/l, respectively (the initial tough levels were 7.5 and 6.8 mg/l). He improved two days later and became afebrile. The remaining two patients were treated with no change in the initial vancomycin dosage, and finally responded to the treatment regimen approximately after one week. During the treatment phase, only one patient experienced increases of serum creatinine levels that was resolved by parenteral fluid resuscitation, and none of the patients had subjective hearing loss (see Table 1). Fourteen patients underwent another serum and CSF analysis for vancomycin levels between the days 8e10 of vancomycin treatment, 15e30 min before the next dose. The mean trough level of vancomycin in serum and CSF were 13.32 1.02 and 10.64 1.21 mg/l, respectively. A positive linear correlation existed between the serum and CSF trough level of the drug (r Z 0.71, P Z 0.004, Fig. 2). The serum and CSF concentrations of vancomycin showed no significant variation on days 8e10 of treatment relative to the day 4. The CSF-to-serum ratio was 0.79 0.063 (coefficient of variation: 8.3%). Although, the second CSF analysis showed decreases in the leukocyte count and protein concentrations and increases in the glucose levels compared to the first analysis, but none of these changes were statistically significant (P Z NS). There was no correlation between these CSF elements and the vancomycin concentrations on the second analysis. At the end of the treatment course, all of the patients appeared to be in the normal status, except for one. A 49year-old opiate dependent man, without underlying comorbidity, experienced sudden death on the day 7 of the treatment, in spite of dramatic improvements after 2 days. The cause of death remained unresolved due to no consent of his family to autopsy (see Table 2).
Please cite this article in press as: Shokouhi S, Alavi Darazam I, Determination of vancomycin trough level in serum and cerebrospinal fluid of patients with acute community-acquired meningitis: A prospective study, J Infect (2014), http://dx.doi.org/10.1016/ j.jinf.2014.06.010
4
S. Shokouhi, I. Alavi Darazam Table 1
The summary of CSF analysis and corresponding serum level of vancomycin levels.
LAB analysis
The leukocyte The glucose The protein count of CSF of CSF (mg/dl) of CSF (cell/mm3) (mg/dl)
Before the 4th dose 2525 8860 During 8e10th days 235 561
58.27 39.3 51.32 29.1
Serum trough level CSF trough (mg/l) level (mg/l)
124.15 179.2 13.82 1.28 68.88 37.4 13.32 1.02
Discussion The main finding of this study was the relative stability of the vancomycin concentration in CSF of the patients with acute meningitis over the treatment course. The diagnosis of acute meningitis was based on the clinical presentation and CSF analysis that showed the inflammatory process of meninges. However, the etiology and microbiology of acute meningitis in the study subjects and their effects on the therapeutic drug monitoring were not evaluated in this study. Nevertheless, it should be noted that the main concern about the vancomycin penetration into the CSF has been related to meningeal inflammation regardless of the etiological considerations. Therefore, we focused only on meningeal inflammation and its effect on CSF vancomycin concentrations. Despite abundant information about the monitoring of vancomycin levels in serum, particularly since 1980s, data on the pharmacokinetics of vancomycin in the CSF are lacking and limited to a few studies with small sample sizes. In 2009, the Infectious Diseases Society of America (IDSA) and the Society of Infectious Diseases Pharmacists published a consensus review suggesting therapeutic monitoring of vancomycin in adult patients and maintaining a total vancomycin serum trough concentrations of 15e20 mg/l for complicated infections such as bacteremia, endocarditis, osteomyelitis, meningitis, and hospitalacquired pneumonia caused by Staphylococcus aureus.6 However, there was no clear background for this recommendation. Moreover, in this paper, the concentration of
Figure 2 Correlation between the second serum and cerebrospinal fluid trough level (r: 0.71, P: 0.004), 14 samples.
CSF/serum trough ratio
11.2 1.41 0.811 0.082 10.64 1.21 0.79 0.063
vancomycin in the CSF was referred to a single study that originally had no data about CSF vancomycin.6 However, this was the only authentic consensus review about vancomycin monitoring and dosage so far. The vancomycin levels in the serum and CSF samples was measured using HPLC method. It is also possible to measure vancomycin concentrations by the Fluorescent Polarization Immunoassays (FPIA), but there are concerns about the efficacy and accuracy of this method due to overestimation.8,9 On the other hand, the HPLC method has high sensitivity and specificity, but it is time consuming and generally labor-intensive.10 The HPLC measurements results was reported to be either higher or lower than the immunoassay measurements in different times, albeit non-significant.9 Initially, we used the FPIA method to analyze vancomycin trough levels, but interestingly, the results were very unacceptable, particularly for CSF (data not shown). This prompted us to change the assay protocol to the HPLC method. On the basis of our experience, we assume that one of the most important sources of variations in studies about vancomycin levels might be due to the different measurement methods. Currently, the British National Formulary (BNF) recommendations on vancomycin trough concentrations are 10e15 mg/l, and 15e20 mg/l for more resistant strains. Similar levels have also been suggested for pneumonia and meningitis, whereas continuous infusion of vancomycin to achieve an average steady state concentration of 15e25 mg/l has been proposed for critically ill patients.11 In a study conducted in ICU setting on 13 patients with vancomycin sensitive infections, the effect of continuous infusions of the drug on its serum levels was evaluated by daily measurements of the vancomycin with FPIA, and efforts was made to maintain its serum values between 20 and 30 mg/l. Their results showed that patients with meningitis had maximal CSF vancomycin levels ranging from 5.7 to 19.0 mg/l (mean, 11.1 4.9 mg/l), with a mean CSF-to-serum ratio of 48%. On the other hand, patients with pathologies other than meningitis had maximal CSF levels ranging from 4.89 to 2.42 mg/l (mean, 3.45 1.11 mg/l), with a mean CSF-to-serum ratio of 18%. These data indicate a more effective CSF penetration of the drug in infiamed meningeal tissues. The CSF-toserum concentration ratio was twice as high for the group of patients with meningitis as for the other group (48 se J. and co-workers reversus 18%).10 Additionally, Albane ported that the clinical and bacteriological efficacy of treatment by continuous vancomycin infusions was high. They recommended the administration of high doses of vancomycin at a mean of 62 mg/kg/day to obtain stable concentrations of 25e30 mg/l in serum which ensure CSF levels of 6e19 mg/l.12
Please cite this article in press as: Shokouhi S, Alavi Darazam I, Determination of vancomycin trough level in serum and cerebrospinal fluid of patients with acute community-acquired meningitis: A prospective study, J Infect (2014), http://dx.doi.org/10.1016/ j.jinf.2014.06.010
Vancomycin trough level in CSF Table 2
1 2 3 4 5 6 7 8 9 10 11 12 13 14
5
Summary of the characteristics of fourteen patients.
Gender
1st Serum trough (mg/l)
1st CSF trough (mg/l)
2nd Serum trough (mg/l)
2nd CSF trough (mg/l)
1st CSF/serum trough
2nd CSF/serum trough
Male Female Male Male Male Male Male Male Male Female Female Male Male Male
15.5 13.9 13.5 13.5 12.5 13.5 15.0 11.6 13.8 13.2 13.4 15.9 12.6 10.9
12.6 9.9 13.0 11.2 9.2 10.8 8.3 10.3 11.9 11.9 9.8 11.7 10.3 7.5
14.6 12.7 14.6 14.2 13.5 13.2 14.2 13.9 13.2 13.8 13.4 11.6 11.9 11.7
13.5 11.9 11.5 10.9 10.9 10.8 10.8 10.7 10.5 10.3 9.7 9.5 9.3 8.6
1.23 1.4 1.03 1.20 1.35 1.25 1.80 1.12 1.16 1.11 1.37 1.36 1.22 1.45
1.08 1.07 1.27 1.30 1.24 1.22 1.31 1.3 1.26 1.34 1.38 1.22 1.28 1.36
The first clinical trial to show the relations between serum and CSF levels of vancomycin in patients with bacterial meningitis was carried out by Ricard et al. in 2007 wherein vancomycin concentrations were measured by FPIA in the serum and CSF samples obtained on the day 2 or 3 of the antimicrobial therapy. They showed that adequate levels of vancomycin in the CSF was obtained by administration of high-dose systemic vancomycin in patients receiving adjunctive dexamethasone treatment (mean, 7.9 mg/l, or 30% of the serum concentrations). The mean serum concentration of vancomycin was 25.2 mg/l (ranging from 14.2 to 39.0 mg/l). Moreover, a significant correlation was found between vancomycin concentrations in the serum and CSF (r Z 0.68). They also found that vancomycin penetration into the CSF was connected with CSF protein levels a marker of increased bloodebrain barrier permeability caused by meningeal inflammation.4 In our study, patients did not receive steroids as an adjunct, and the dose of vancomycin was lower. However, the levels of vancomycin proved to be higher. Various sampling times and different assay methods may explain these variations. In a consensus review, Thompson et al. recommended that in patients with CNS shunt infections the usual intravenous regimen of 60 mg/kg/day should be divided and administered every 6 h. They also suggested that intraventricular vancomycin should be considered for most patients, starting with 10 mg/day. Additionally, CSF vancomycin concentrations should be monitored and dose adjustments should be made as needed to maintain CSF trough vancomycin concentrations between 5 and 20 mg/ l.13 Although, these results are based on post-surgical infections and all of patients have had CNS intervention and meningeal device. In a trial, simultaneous serum and CSF samples were obtained at 4 h of the therapy to determine the minimal levels of vancomycin maintained in the CSF throughout the initial period of treatment. Additional CSF and serum sampling were also performed at the end of therapy on day 10, 30e60 min following the last vancomycin infusion to determine the final CSF values of leukocyte count, protein
and glucose concentrations, peak vancomycin levels, and bactericidal titers. CSF leukocyte count was 0e51 per mm3 (median, 9), and protein and glucose concentrations in the CSF were 0.30e0.84 (median, 0.49) g/l and 46‒64 (median, 54) mg/dl, respectively. Peak vancomycin levels in CSF and serum were undetectable to 2.6 (median, 1.9) mg/ml and 15‒25 (median, 18.5) mg/ml, respectively. Peak CSF bactericidal titers ranged from
