- Email: [email protected]
Albumin concentration significantly impacts on free teicoplanin plasma concentrations in non-critically ill patients with chronic bone sepsis
G Model ANTAGE-4536; No. of Pages 5
ARTICLE IN PRESS International Journal of Antimicrobial Agents xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
International Journal of Antimicrobial Agents journal homepage: http://www.elsevier.com/locate/ijantimicag
Short Communication
Albumin concentration significantly impacts on free teicoplanin plasma concentrations in non-critically ill patients with chronic bone sepsis夽 A.J. Brink a,∗ , G.A. Richards b , E.E.G. Lautenbach c , N. Rapeport d , V. Schillack e , L. van Niekerk f , J. Lipman g,h , J.A. Roberts g,h,i a
Ampath National Laboratory Services, Milpark Hospital, 9 Guild Road, Parktown, 2193 Johannesburg, South Africa Department of Critical Care, Charlotte Maxeke Johannesburg Academic Hospital, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa c Department of Orthopaedic Surgery, Milpark Hospital, 9 Guild Road, Parktown, Johannesburg, South Africa d Department of Medicine, Milpark Hospital, 9 Guild Road, Parktown, Johannesburg, South Africa e Analytical Toxicology Laboratory Services, George, South Africa f Department of Esoteric Sciences, Ampath National Laboratory Services, National Referral Laboratory, Centurion, South Africa g Burns, Trauma and Critical Care Research Centre, The University of Queensland, Brisbane, QLD, Australia h Department of Intensive Care Medicine, Royal Brisbane and Women’s Hospital, Herston, Brisbane, QLD, Australia i Pharmacy Department, Royal Brisbane and Women’s Hospital, Herston, Brisbane, QLD, Australia b
a r t i c l e
i n f o
Article history: Received 18 November 2014 Accepted 27 January 2015 Keywords: Teicoplanin Protein binding Free Hypoalbuminaemia Population pharmacokinetics TDM
a b s t r a c t The impact of decreased serum albumin concentrations on free antibiotic concentrations in non-critically ill patients is poorly described. This study aimed to describe the pharmacokinetics of a high-dose regimen of teicoplanin, a highly protein-bound antibiotic, in non-critically ill patients with hypoalbuminaemia. Ten patients with chronic bone sepsis and decreased serum albumin concentrations (<35 g/L) receiving teicoplanin 12 mg/kg 12-hourly intravenously for 48 h followed by 12 mg/kg once daily were enrolled. Surgical debridement was performed on Day 3. Samples of venous blood were collected pre-infusion and post-infusion during the first 4 days of therapy. Total and free teicoplanin concentrations were assayed using validated chromatographic methods. The median serum albumin concentration for the cohort was 18 (IQR 15–24) g/L. After 48 h, the median (IQR) free trough (fCmin ) and total trough (tCmin ) concentrations were 2.90 (2.67–3.47) mg/L and 15.54 (10.28–19.12) mg/L, respectively, although trough concentrations declined thereafter. Clearance of the free concentrations was significantly high relative to the total fraction at 38.6 (IQR 29.9–47.8) L/h and 7.0 (IQR 6.8–9.8) L/h, respectively (P < 0.001). Multiple linear regression analysis demonstrated that whereas total teicoplanin concentration did not impact on free concentrations (P = 0.174), albumin concentration did (P < 0.001). This study confirms the significant impact of hypoalbuminaemia on free concentrations of teicoplanin in non-critically ill patients, similar to that in critically ill patients. Furthermore, the poor correlation with total teicoplanin concentration suggests that therapeutic drug monitoring of free concentrations should be used in these patients. © 2015 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
1. Introduction It has become increasingly clear that the free (unbound) antibiotic concentration is responsible for the pharmacological effect and
夽 Data from this study were partially presented at the 32nd International Symposium on Intensive Care and Emergency Medicine, 20–23 March 2012, Brussels, Belgium [In: Meeting Abstract Critical Care 2012;16(Suppl 1):P69]. ∗ Corresponding author. Tel.: +27 11 853 6540; fax: +27 12 682 2459. E-mail address: [email protected] (A.J. Brink).
a better understanding of this can enhance the accuracy of therapy and potentially improve clinical outcomes [1]. This might be of particular relevance for highly protein-bound antibiotics such as teicoplanin (90–95% bound), especially in critically ill patients where hypoalbuminaemia is frequent and as a consequence the volume of distribution (V) and clearance (CL) of the unbound drug are increased [2]. These pharmacokinetic (PK) changes could result in suboptimal teicoplanin exposures and may necessitate dose adjustments to ensure that therapeutic exposures are achieved [2]. In this regard, Mimoz et al. utilising a high-dose regimen [12 mg/kg every 12 h (q12 h) for 48 h, followed by 12 mg/kg once daily] in critically ill
http://dx.doi.org/10.1016/j.ijantimicag.2015.01.015 0924-8579/© 2015 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
Please cite this article in press as: Brink AJ, et al. Albumin concentration significantly impacts on free teicoplanin plasma concentrations in non-critically ill patients with chronic bone sepsis. Int J Antimicrob Agents (2015), http://dx.doi.org/10.1016/j.ijantimicag.2015.01.015
G Model ANTAGE-4536; No. of Pages 5
ARTICLE IN PRESS A.J. Brink et al. / International Journal of Antimicrobial Agents xxx (2015) xxx–xxx
2
patients with ventilator-associated pneumonia and severe hypoalbuminaemia (median albumin concentration 16.1 g/L) observed variations in the free fraction of teicoplanin ranging from 8% to 42%, but unfortunately did not correlate these with serum albumin concentrations [3]. This highlights three important shortcomings of available data on the pharmacokinetics of unbound teicoplanin. First, unbound concentrations are often not measured either clinically or for academic reasons [4]. Second, in the rare circumstances that unbound concentrations are determined, patients generally have not had serum albumin concentrations that were sufficiently low to alter the free teicoplanin pharmacokinetics significantly [5]. Third, no PK data are available describing the effect of hypoalbuminaemia on unbound concentration when using dosing regimens for noncritically ill patients with ‘deep-seated’ infections such as those of bone and prostheses, where the target total trough concentration (tCmin ) is ≥20 mg/L [6,7]. One study did correlate clinical outcome and teicoplanin tCmin in patients (n = 35) with bone, joint and vascular-access infections treated with a high-dose regimen, but albumin concentrations were not determined [8]. Another study with a large cohort of bone and joint infections, which was designed to develop dosing guidelines for teicoplanin in the outpatient setting, also used a high-dose regimen but measured total trough concentrations rather than free concentrations in patients with normal albumin concentrations [4]. Data from the critical care literature suggest that hypoalbuminaemia is likely to alter teicoplanin pharmacokinetics significantly and, in particular, free concentrations [6]. To address the above deficiencies in the literature, we aimed to investigate the impact of decreased serum albumin concentrations on free teicoplanin concentrations in non-critically ill patients using a high-dose teicoplanin regimen for chronic osteomyelitis or septic arthritis.
2. Patients and methods This single-centre, prospective, open-label, post-authorisation study was performed at Milpark Hospital, a private referral hospital in Johannesburg, South Africa. Ethical approval was obtained from the local ethics committee of Milpark Hospital. The study was conducted following the guidelines of the Declaration of Helsinki. Written informed consent was obtained either from the patient or, when appropriate, from their closest relative.
recruitment and during the 4-day study period. Urine creatinine concentrations were not measured. 2.2. Blood sampling The investigations took place during the first 4 days of therapy. Samples were collected 15 min before and 30, 120 and 720 min after each teicoplanin administration. Following centrifugation at 3000 rpm for 10 min at 4 ◦ C, plasma was removed and transferred into separate tubes. Two fractions were kept frozen at −20 ◦ C whilst being transferred to the Ampath National Reference Laboratory (Centurion, South Africa) for analysis of total and free concentrations. The other three tubes were used for onsite measurement of biochemistry and biomarkers. 2.3. Drug assay Plasma teicoplanin concentrations were determined at the Department of Esoteric Sciences at Ampath National Reference Laboratory by a validated high-performance liquid chromatography (HPLC) method as described by Roberts et al. [6]. The non-protein-bound fraction of teicoplanin was determined by HPLC after filtration with a Centrifree® device [6]. Standards and controls were prepared in blank plasma. The assay had a linear range of 0.3–26.6 mg/L, a correlation coefficient (r2 ) of 0.9984, a limit of detection of 0.40 mg/L and a limit of quantification of 1.36 mg/L. The accuracy (95% confidence level) was determined as 2.0 ± 0.2 and 20.0 ± 2.3 at the nominal concentration with a precision <5% coefficient of variation (n = 10). 2.4. Pharmacokinetic analysis Non-compartmental PK analysis was performed using previously described methods [9]. The following covariates were evaluated for effects on free teicoplanin concentrations: total teicoplanin concentration; serum albumin concentration; age; MDRD creatinine clearance; and serum creatinine concentration. Following identification of significant covariates in univariate testing (P < 0.2), a multiple linear regression model was constructed to determine the primary determinants of subtherapeutic trough concentrations, which was defined as a total trough concentration (tCmin ) <20 mg/L. Goodness of fit of the model was assessed by the Hosmer–Lemeshow statistic. All statistical analyses employed IBM SPSS Statistics for Windows v.19 (IBM Corp., Armonk, NY). A P-value of <0.05 was considered statistically significant.
2.1. Selection of patients
3. Results
Inclusion criteria were adult patients (aged >18 years) referred for surgical debridement of chronic, deep-seated, Gram-positive bone or joint infections with decreased serum albumin concentrations (<35 g/L; normal range 35–52 g/L) who at the discretion of the physician and/or orthopaedic surgeon were deemed to require high-dose teicoplanin (Targocid® ; Sanofi-Aventis, Midrand, South Africa) regimens for ≥4 days (loading dose of 12 mg/kg q12 h intravenously for 48 h, followed by 12 mg/kg once daily over 30 min). Only patients who had debridement performed on Day 3 of antibiotic therapy were enrolled. Patients who were acutely ill, had suspected glycopeptide allergy, were pregnant or who had moderate-to-severe renal dysfunction [arbitrarily defined as an estimated glomerular filtration rate (eGFR) <59 mL/min/1.73 m2 using the Modification of Diet in Renal Disease (MDRD) equation for GFR calculation] were excluded from enrolment. Chronicity of osteomyelitis or septic arthritis was defined as duration of >4 weeks. Hypoalbuminaemia was defined as a serum albumin level <25 g/L [2]. Demographic characteristics were recorded at
3.1. Patient characteristics Ten patients (eight male and two female) with confirmed Grampositive, chronic, deep-seated bone or joint infections referred to Milpark Hospital for surgical debridement were recruited over a 4-month period. Their demographic data are given in Table 1. All patients had decreased albumin concentrations (<35 g/L) at recruitment. Despite two patients not actually having hypoalbuminaemia as per the definition of <25 g/L (i.e. concentrations of 27 g/L and 34 g/L), the median [interquartile range (IQR)] for the cohort was 18 (15–24) g/L. Small changes in serum creatinine concentrations occurred particularly perioperatively; on Day 3, the median value was 84.0 (65.5–106.0) mol/L. 3.2. Pharmacokinetic profile Teicoplanin free trough concentrations (fCmin ) and tCmin over the 4-day study period are given in Table 2. After 48 h, the
Please cite this article in press as: Brink AJ, et al. Albumin concentration significantly impacts on free teicoplanin plasma concentrations in non-critically ill patients with chronic bone sepsis. Int J Antimicrob Agents (2015), http://dx.doi.org/10.1016/j.ijantimicag.2015.01.015
G Model
ARTICLE IN PRESS
ANTAGE-4536; No. of Pages 5
A.J. Brink et al. / International Journal of Antimicrobial Agents xxx (2015) xxx–xxx Table 1 Demographic characteristics of patients (n = 10) before first teicoplanin administration and perioperatively. Parameter
Median
IQR
Normal value
Age (years) Weight (kg) Height (m) Duration of symptoms (days) Albumin (g/L) WBC count (109 /L) CRP (mg/L) ESR (n = 4) (mm/h) Iron (mol/L) Transferrin (n = 9) (g/L) Saturation (%) Duration of surgery (h)a Blood transfusions (mL) (Days 3–4)a Post-operative high-care admission (h)a Serum creatinine (mol/L)a Day 1 Day 2 Day 3 (n = 9) Day 4 (n = 9) eGFR (MDRD) (mL/min/1.73 m2 )a Day 1 Day 2 Day 3 (n = 9) Day 4 (n = 9)
56 80 1.73 103 18.0 (L) 10.0 (H) 42 (H) 18.5 (H) 5.7 (L) 2.0 (L) 9.0 (L) 3.0 1250 72
44–69 67–98 1.69–1.85 56–201 15–24 5.8–13.2 13–95 10.5–22.50 3.9–9.95 1.85–2.45 7–21 2–4 500–2250 72–120
N/A N/A N/A N/A 35–52 3.92–9.88 <5 0–15 11.6–31.3 2.2–3.7 20–50 N/A N/A N/A 64–104
84.50 86.50 84.0 84.0
72–90 68–89 65.5–106 65.5–96.5
78.5 81 76 86
63 to >90 59 to >90 52.5 to >90 59–88.5
>60
IQR, interquartile range; N/A, not applicable; L, low; WBC, white blood cell; H, high; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; eGFR, estimated glomerular filtration rate; MDRD, Modification of Diet in Renal Disease. a Perioperative patient characteristics.
median (IQR) fCmin and tCmin concentrations were 2.90 (2.67–3.47) mg/L and 15.54 (10.28–19.12) mg/L, respectively; however, both began to decline once the dose was reduced to 12 mg/kg once daily. The median protein binding ranged from 66.3% to 80.8% between Day 1 and Day 4, respectively. The PK parameter estimates of free and total teicoplanin are shown in Table 3. The total drug area under the concentration–time curve from 0 to 12 h (AUC0–12 h ) increased from 137.9 mg h/L on Day 1 to 203.4 mg h/L on Day 4. Clearance (CL) of the free (unbound) concentrations was extremely high relative to the total fraction at 38.6 (IQR 29.9–47.8) L/h and 7.0 (IQR 6.8–9.8) L/h, respectively (P < 0.001). Results of the multiple linear regression model for free teicoplanin demonstrated that total concentration [odds ratio (OR) = 0.015, 95% confidence interval (CI) 0.003–0.052; P = 0.174] did not impact statistically significantly on free concentrations, whereas serum creatinine concentration (OR = 0.013, 95% CI 0.001–0.025; P = 0.034) and serum albumin concentration (OR = 0.120, 95% CI 0.078–0.161; P < 0.001) did. High interpatient variations were found in the free teicoplanin fraction, whilst higher free fractions were seen in patients with lower albumin concentrations (data not shown).
3
Table 3 Pharmacokinetic parameter estimates [median (interquartile range)] of teicoplanin. Total concentrations
Unbound concentrations
Cmax (mg/L)
Day 1 Day 4
20.1 (15.7–22.2) 25.3 (20.4–31.7)
2.6 (2.3–3.1) 3.0 (2.1–4.6)
Cmin (mg/L)
Day 1 Day 4
6.7 (4.0–10.2) 10.3 (8.0–12.5)
2.3 (1.3–2.5) 1.93 (1.04–2.59)
AUC0–12 h (mg h/L) CL (L/h) Vz (L/kg)
Day 1 Day 1 Day 1
137.9 (83.1–168.2) 7.0 (6.8–9.8) 174.1 (101.6–330.5)
28.6 (21.4–33.2) 38.6 (29.9–47.8) 196.6 (195.7–200.8)
Cmax , observed maximum concentration during sampling period; Cmin , observed minimum concentration during sampling period; AUC0–12 h , area under the concentration–time curve during 12-h dosing period; CL, total clearance; Vz , volume of distribution during terminal phase.
4. Discussion To the best of our knowledge, this study is the first to describe the pharmacokinetics of free teicoplanin in non-critically ill patients with hypoalbuminaemia. Hypoalbuminaemia is likely to be relatively common in malnourished or diabetic patients and/or those with chronic infections [2]. In this study, at the time of recruitment the median duration of chronic bone or joint infections was 103 days (IQR 56–201 days) and as a consequence this probably impacted on serum albumin concentrations (median 18 g/L). The data presented in this study provide support for previous studies that have described altered pharmacokinetics of highly protein-bound antibiotics in intensive care unit (ICU) patients with hypoalbuminaemia [2,3,6,10,11]. Interestingly, compared with similar patients with deep-seated infections but normal albumin levels described by Lamont et al. [4], we found a 10fold increase in the median teicoplanin CL (data not shown), which is likely to have resulted from an increased free fraction of teicoplanin that was available for renal clearance in these patients with conserved renal function, and perhaps even augmented renal function. In this study, hypoalbuminaemia (P < 0.001) and not total teicoplanin concentrations (P = 0.174) impacted significantly on free concentrations. As such, it is not clear whether increasing the dose will result in linear increases in free concentrations. However, concentrations of both free and total drug did begin to decline on Day 3 once the dose was reduced to 12 mg/kg once daily (Table 2) and, therefore, 12 mg/kg q12 h should probably be extended beyond 48 h, particularly in patients with large fluid or blood losses during extensive surgery, as was the case here. Despite the statistically significant (P = 0.034) relationship between serum creatinine and free drug concentrations, we feel that this is unlikely to be clinically relevant given the small changes in serum creatinine concentrations that occurred perioperatively. Therapeutic drug monitoring (TDM) of free concentrations to individualise antimicrobial therapy has been recommended for ICU patients [1,2]. It is believed that for antibiotics where TDM targets are based on trough concentrations, the free concentration might be more likely to predict efficacy than the total concentration
Table 2 Teicoplanin free trough concentration (fCmin ) and total trough concentration (tCmin ) observed with 12-hourly dosing (Days 1–2), followed by once-daily dosing (Days 3–4). Day 1
Median (IQR) fCmin (mg/L) Median (IQR) tCmin (mg/L) Median (IQR) protein binding (%)
Day 2
Day 3a
Day 4
Trough 1
Trough 2
Trough 1
Trough 2
Trough 1
Trough 1
0.00 0.00 N/A
2.24 (1.28–2.53) 6.66 (4.02–10.17) 66.3 (54.5–84.1)
2.74 (1.00–2.92) 8.61 (4.12–12.85) 74.8 (52.9–78.6)
2.26 (1.65–2.52) 10.31 (7.52–15.52) 74.2 (61.5–85.7)
2.90 (2.67–3.47) 15.54 (10.28–19.12) 74.8 (68.3–83.6)
1.93 (1.04–2.59) 9.17 (5.63–12.26) 80.8 (66.1–89.2)
IQR, interquartile range; N/A, not applicable. a Surgical debridement.
Please cite this article in press as: Brink AJ, et al. Albumin concentration significantly impacts on free teicoplanin plasma concentrations in non-critically ill patients with chronic bone sepsis. Int J Antimicrob Agents (2015), http://dx.doi.org/10.1016/j.ijantimicag.2015.01.015
G Model ANTAGE-4536; No. of Pages 5 4
ARTICLE IN PRESS A.J. Brink et al. / International Journal of Antimicrobial Agents xxx (2015) xxx–xxx
[1,12]. Hence, to optimise microbiological and clinical outcomes in non-ICU patients with infection and hypoalbuminaemia, the results from this study ‘make the case’ for use of TDM and specifically measurement of free concentrations, probably for all highly protein-bound antibiotics (>80%), including cloxacillin, an agent frequently used in this setting [1]. In fact, Ulldemolins et al. previously documented the impact of hypoalbuminaemia on free flucloxacillin concentrations and the implications thereof, not only for dosing but for alternative administration strategies albeit in ICU patients [11]. As with critically ill patients, it is evident from the results in this chronically ill population that the variability in protein binding and in teicoplanin concentrations makes the prediction of free and total plasma concentrations difficult. For example, one patient in this study had an fCmin of 0 mg/L from Day 2 onwards (data not shown). In an earlier report, a model was proposed to estimate free teicoplanin serum concentrations, however this could not be validated in the Defining Antibiotic Levels in Intensive care unit patients (DALI) teicoplanin study in which highly variable differences between measured and calculated free concentrations were observed [6,13]. This has also recently been shown for -lactams [1]. Hence, the only way to ensure that adequate free concentrations are present and thereby to improve bacteriological and clinical outcomes appears to be by direct measurement of free drug and appropriate dose adjustment. Furthermore, challenges for teicoplanin dosing in these patients that support TDM are that measuring renal function is difficult in these patients and teicoplanin is a renally cleared antibiotic with a terminal half-life much longer than the dosing interval. In addition, the recommended teicoplanin target tCmin for patients with deep-seated infections (≥20 mg/L) was not achieved at steady state in all our patients despite using a high-dose regimen [4,6,7]. Pea et al. have previously shown that using TDM in an antimicrobial stewardship programme significantly improved the likelihood of early optimal teicoplanin exposure [14]. This highlights an important question that was posed by Roberts et al., namely what should the targets for free TDM be [6]? The authors hypothesised that the lower therapeutic range of free plasma teicoplanin trough concentrations should be an fCmin of 1–2 mg/L. This is based on the fact that TDM targets were determined in patients with normal albumin levels and assuming 90% protein binding, the free equivalent range to a tCmin of 10–20 mg/L would be a free trough concentration of 1–2 mg/L. This is the same logic used in recommendations for free phenytoin target concentrations where the total concentration target for phenytoin TDM is divided by 10 using an estimated 10% free fraction [15]. If that is the case, then the higher dosing schedule used in the current study, at least for the first 48 h, appears to be appropriate if free concentrations are used as a surrogate for teicoplanin efficacy as opposed to total trough concentrations, which are currently recommended. Nine of ten patients in the current study had an fCmin > 2 mg/L after 48 h (data not shown). The impact of protein binding on the pharmacokinetics of antibiotics that are moderately (30–70%) or minimally bound (<30%) has yet to be clarified in detail. Wong et al. recently demonstrated marked variability both in free and total concentrations of seven lactam antibiotics in ICU patients irrespective of whether they were highly protein bound or not [1]. Indeed, the variability occurred with other agents active against Gram-positive organisms such as ampicillin, benzylpenicillin, cefazolin and flucloxacillin where protein binding under normal circumstances is 20%, 65%, 80% and 93%, respectively. This was also confirmed to be the case for linezolid (which is usually 31% protein bound) in critically ill patients, particularly if they had decreased albumin concentrations [12].
4.1. Limitations of the study The teicoplanin minimum inhibitory concentrations of the infecting pathogens were not available for all isolates and therefore the study results do not address pharmacodynamic indices (except for PK targets) and their relation to outcome. However, both bacteriological and clinical outcomes would have been confounded by source control, which would probably be more important than antibiotic therapy in this regard [8]. The small cohort of 10 patients could also be considered to be a limitation as it may have prevented other factors that influence PK variability from being identified. Finally, the recommendations derived from the data analysis cannot be extrapolated to other patient populations with chronic bone or joint infections with concurrent albumin plasma concentrations >34 g/L or to those with moderate or severe renal dysfunction (eGFR < 59 mL/min/1.73 m2 ). 5. Conclusions The results of this study highlight the importance of protein binding and the potential role of TDM in optimising free concentrations of highly protein-bound antibiotics in non-critically ill patients with decreased serum albumin concentrations. However, the clinical and bacteriological implications of subtherapeutic teicoplanin concentrations due to hypoalbuminaemia and the monitoring of free teicoplanin concentrations need to be clarified. Given the uncertainty regarding unpredictable protein binding and the impact on pharmacokinetics, it is likely that direct measurement of free antibiotic concentrations should be preferred. In conclusion, if we are to ensure both efficacy and a sustained suppression of antibiotic resistance, antibiotic exposure must be optimised to robust pharmacokinetic/pharmacodynamic targets in all patients. Furthermore, whatever patient population is studied, hypoalbuminaemia is proving to be one of the most important covariates influencing the interindividual PK variability of proteinbound antibiotics. Funding This work was supported by an investigator grant for determination of teicoplanin concentrations from Sanofi-Aventis (Midrand, Johannesburg, South Africa). The funder had no involvement in conception and designing of the study, neither in the collection, analysis and interpretation of data nor in the writing of the report and in the decision to submit the article for publication. JAR is funded by a Career Development Fellowship from the National Health and Medical Research Council of Australia [APP1048652]. Competing interests AJB and GAR have received research grants from or served on scientific advisory boards for Pfizer, Sanofi-Aventis, Aspen/GSK and MSD. All other authors declare no competing interests. Ethical approval Ethical approval was obtained from the local ethics committee of Milpark Hospital (Johannesburg, South Africa) [02.12.2012]. The study was conducted following the guidelines of the Declaration of Helsinki. Written informed consent was obtained either from the patient or, when appropriate, from their closest relative. References [1] Wong G, Briscoe S, Adnan S, McWhinney A, Ungerer J, Lipman J, et al. Protein binding of -lactam antibiotics in critically ill patients: can we
Please cite this article in press as: Brink AJ, et al. Albumin concentration significantly impacts on free teicoplanin plasma concentrations in non-critically ill patients with chronic bone sepsis. Int J Antimicrob Agents (2015), http://dx.doi.org/10.1016/j.ijantimicag.2015.01.015
G Model ANTAGE-4536; No. of Pages 5
ARTICLE IN PRESS A.J. Brink et al. / International Journal of Antimicrobial Agents xxx (2015) xxx–xxx
[2]
[3]
[4]
[5] [6]
[7]
[8]
successfully predict unbound concentrations? Antimicrob Agents Chemother 2013;57:6165–70. Ulldemolins M, Roberts JA, Rello J, Paterson D, Lipman J. The effects of hypoalbuminaemia on optimizing antibiotic dosing in critically ill patients. Clin Pharmacokinet 2011;50:99–110. Mimoz O, Rolland D, Adoun M, Marchand S, Breilh D, Brumpt I, et al. Steadystate trough serum and epithelial lining fluid concentrations of teicoplanin 12 mg/kg per day in patients with ventilator-associated pneumonia. Intensive Care Med 2006;32:775–9. Lamont E, Seaton RA, Macpherson M, Semple L, Bell E, Thomson AH. Development of teicoplanin dosage guidelines for patients treated within an outpatient parenteral antibiotic therapy (OPAT) programme. J Antimicrob Chemother 2009;64:181–7. Soy D, López E, Ribas J. Teicoplanin population pharmacokinetic analysis in hospitalized patients. Ther Drug Monit 2006;28:737–43. Roberts JA, Stove V, De Waele JJ, Sipinkoski B, McWhinney B, Ungerer JPJ, et al.; DALI Study Authors. Variability in protein binding of teicoplanin and achievement of therapeutic drug monitoring targets in critically ill patients: lessons from the DALI Study. Int J Antimicrob Agents 2014;43:423–30. Brink AJ, Richards GA, Cummins RR, Lambson J. Recommendations to achieve rapid therapeutic teicoplanin plasma concentrations in adult hospitalised patients treated for sepsis. Int J Antimicrob Agents 2008;32:455–8. Greenberg RN. Treatment of bone, joint and vascular-access-associated Grampositive bacterial infections with teicoplanin. Antimicrob Agents Chemother 1990;34:2392–7.
5
[9] Sime FB, Roberts MS, Warner MS, Hahn U, Robertson TA, Yeend S, et al. Altered pharmacokinetics of piperacillin in febrile neutropenic patients with haematological malignancy. Antimicrob Agents Chemother 2014;58: 3533–7. [10] Brink AJ, Richards GA, Schillack V, Kiem S, Schentag J. Pharmacokinetics of once-daily dosing of ertapenem in critically ill patients with severe sepsis. Int J Antimicrob Agents 2009;33:432–6. [11] Ulldemolins M, Roberts JA, Wallis SC, Rello J, Lipman J. Flucloxacillin dosing in critically ill patients with hypoalbuminaemia: special emphasis on unbound pharmacokinetics. J Antimicrob Chemother 2010;65:1771–8. [12] Yagi T, Naito T, Doi M, Nagurac O, Yamada T, Maekawa M, et al. Plasma exposure of free linezolid and its ratio to minimum inhibitory concentration varies in critically ill patients. Int J Antimicrob Agents 2013;42: 329–34. [13] Yano R, Nakamura T, Tsukamoto H, Igarashi T, Goto N, Wakiya Y, et al. Variability in teicoplanin protein binding and its prediction using serum albumin concentrations. Ther Drug Monit 2007;29:399–403. [14] Pea F, Viale P, Pavan F, Tavio M, Poz D, Beltrame A, et al. The effect of multifactorial, multidisciplinary educational interventions on appropriate use of teicoplanin. Int J Antimicrob Agents 2006;27:344–50. [15] Patsalos PN, Berry DJ, Bourgeois BFD, Cloyd JC, Glauser TA, Johannessen SI, et al. Antiepileptic drugs—best practice guidelines for therapeutic drug monitoring: a position paper by the Subcommission on Therapeutic Drug Monitoring, ILAE Commission on Therapeutic Strategies. Epilepsia 2008;49: 1239–76.
Please cite this article in press as: Brink AJ, et al. Albumin concentration significantly impacts on free teicoplanin plasma concentrations in non-critically ill patients with chronic bone sepsis. Int J Antimicrob Agents (2015), http://dx.doi.org/10.1016/j.ijantimicag.2015.01.015
ARTICLE IN PRESS International Journal of Antimicrobial Agents xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
International Journal of Antimicrobial Agents journal homepage: http://www.elsevier.com/locate/ijantimicag
Short Communication
Albumin concentration significantly impacts on free teicoplanin plasma concentrations in non-critically ill patients with chronic bone sepsis夽 A.J. Brink a,∗ , G.A. Richards b , E.E.G. Lautenbach c , N. Rapeport d , V. Schillack e , L. van Niekerk f , J. Lipman g,h , J.A. Roberts g,h,i a
Ampath National Laboratory Services, Milpark Hospital, 9 Guild Road, Parktown, 2193 Johannesburg, South Africa Department of Critical Care, Charlotte Maxeke Johannesburg Academic Hospital, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa c Department of Orthopaedic Surgery, Milpark Hospital, 9 Guild Road, Parktown, Johannesburg, South Africa d Department of Medicine, Milpark Hospital, 9 Guild Road, Parktown, Johannesburg, South Africa e Analytical Toxicology Laboratory Services, George, South Africa f Department of Esoteric Sciences, Ampath National Laboratory Services, National Referral Laboratory, Centurion, South Africa g Burns, Trauma and Critical Care Research Centre, The University of Queensland, Brisbane, QLD, Australia h Department of Intensive Care Medicine, Royal Brisbane and Women’s Hospital, Herston, Brisbane, QLD, Australia i Pharmacy Department, Royal Brisbane and Women’s Hospital, Herston, Brisbane, QLD, Australia b
a r t i c l e
i n f o
Article history: Received 18 November 2014 Accepted 27 January 2015 Keywords: Teicoplanin Protein binding Free Hypoalbuminaemia Population pharmacokinetics TDM
a b s t r a c t The impact of decreased serum albumin concentrations on free antibiotic concentrations in non-critically ill patients is poorly described. This study aimed to describe the pharmacokinetics of a high-dose regimen of teicoplanin, a highly protein-bound antibiotic, in non-critically ill patients with hypoalbuminaemia. Ten patients with chronic bone sepsis and decreased serum albumin concentrations (<35 g/L) receiving teicoplanin 12 mg/kg 12-hourly intravenously for 48 h followed by 12 mg/kg once daily were enrolled. Surgical debridement was performed on Day 3. Samples of venous blood were collected pre-infusion and post-infusion during the first 4 days of therapy. Total and free teicoplanin concentrations were assayed using validated chromatographic methods. The median serum albumin concentration for the cohort was 18 (IQR 15–24) g/L. After 48 h, the median (IQR) free trough (fCmin ) and total trough (tCmin ) concentrations were 2.90 (2.67–3.47) mg/L and 15.54 (10.28–19.12) mg/L, respectively, although trough concentrations declined thereafter. Clearance of the free concentrations was significantly high relative to the total fraction at 38.6 (IQR 29.9–47.8) L/h and 7.0 (IQR 6.8–9.8) L/h, respectively (P < 0.001). Multiple linear regression analysis demonstrated that whereas total teicoplanin concentration did not impact on free concentrations (P = 0.174), albumin concentration did (P < 0.001). This study confirms the significant impact of hypoalbuminaemia on free concentrations of teicoplanin in non-critically ill patients, similar to that in critically ill patients. Furthermore, the poor correlation with total teicoplanin concentration suggests that therapeutic drug monitoring of free concentrations should be used in these patients. © 2015 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
1. Introduction It has become increasingly clear that the free (unbound) antibiotic concentration is responsible for the pharmacological effect and
夽 Data from this study were partially presented at the 32nd International Symposium on Intensive Care and Emergency Medicine, 20–23 March 2012, Brussels, Belgium [In: Meeting Abstract Critical Care 2012;16(Suppl 1):P69]. ∗ Corresponding author. Tel.: +27 11 853 6540; fax: +27 12 682 2459. E-mail address: [email protected] (A.J. Brink).
a better understanding of this can enhance the accuracy of therapy and potentially improve clinical outcomes [1]. This might be of particular relevance for highly protein-bound antibiotics such as teicoplanin (90–95% bound), especially in critically ill patients where hypoalbuminaemia is frequent and as a consequence the volume of distribution (V) and clearance (CL) of the unbound drug are increased [2]. These pharmacokinetic (PK) changes could result in suboptimal teicoplanin exposures and may necessitate dose adjustments to ensure that therapeutic exposures are achieved [2]. In this regard, Mimoz et al. utilising a high-dose regimen [12 mg/kg every 12 h (q12 h) for 48 h, followed by 12 mg/kg once daily] in critically ill
http://dx.doi.org/10.1016/j.ijantimicag.2015.01.015 0924-8579/© 2015 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
Please cite this article in press as: Brink AJ, et al. Albumin concentration significantly impacts on free teicoplanin plasma concentrations in non-critically ill patients with chronic bone sepsis. Int J Antimicrob Agents (2015), http://dx.doi.org/10.1016/j.ijantimicag.2015.01.015
G Model ANTAGE-4536; No. of Pages 5
ARTICLE IN PRESS A.J. Brink et al. / International Journal of Antimicrobial Agents xxx (2015) xxx–xxx
2
patients with ventilator-associated pneumonia and severe hypoalbuminaemia (median albumin concentration 16.1 g/L) observed variations in the free fraction of teicoplanin ranging from 8% to 42%, but unfortunately did not correlate these with serum albumin concentrations [3]. This highlights three important shortcomings of available data on the pharmacokinetics of unbound teicoplanin. First, unbound concentrations are often not measured either clinically or for academic reasons [4]. Second, in the rare circumstances that unbound concentrations are determined, patients generally have not had serum albumin concentrations that were sufficiently low to alter the free teicoplanin pharmacokinetics significantly [5]. Third, no PK data are available describing the effect of hypoalbuminaemia on unbound concentration when using dosing regimens for noncritically ill patients with ‘deep-seated’ infections such as those of bone and prostheses, where the target total trough concentration (tCmin ) is ≥20 mg/L [6,7]. One study did correlate clinical outcome and teicoplanin tCmin in patients (n = 35) with bone, joint and vascular-access infections treated with a high-dose regimen, but albumin concentrations were not determined [8]. Another study with a large cohort of bone and joint infections, which was designed to develop dosing guidelines for teicoplanin in the outpatient setting, also used a high-dose regimen but measured total trough concentrations rather than free concentrations in patients with normal albumin concentrations [4]. Data from the critical care literature suggest that hypoalbuminaemia is likely to alter teicoplanin pharmacokinetics significantly and, in particular, free concentrations [6]. To address the above deficiencies in the literature, we aimed to investigate the impact of decreased serum albumin concentrations on free teicoplanin concentrations in non-critically ill patients using a high-dose teicoplanin regimen for chronic osteomyelitis or septic arthritis.
2. Patients and methods This single-centre, prospective, open-label, post-authorisation study was performed at Milpark Hospital, a private referral hospital in Johannesburg, South Africa. Ethical approval was obtained from the local ethics committee of Milpark Hospital. The study was conducted following the guidelines of the Declaration of Helsinki. Written informed consent was obtained either from the patient or, when appropriate, from their closest relative.
recruitment and during the 4-day study period. Urine creatinine concentrations were not measured. 2.2. Blood sampling The investigations took place during the first 4 days of therapy. Samples were collected 15 min before and 30, 120 and 720 min after each teicoplanin administration. Following centrifugation at 3000 rpm for 10 min at 4 ◦ C, plasma was removed and transferred into separate tubes. Two fractions were kept frozen at −20 ◦ C whilst being transferred to the Ampath National Reference Laboratory (Centurion, South Africa) for analysis of total and free concentrations. The other three tubes were used for onsite measurement of biochemistry and biomarkers. 2.3. Drug assay Plasma teicoplanin concentrations were determined at the Department of Esoteric Sciences at Ampath National Reference Laboratory by a validated high-performance liquid chromatography (HPLC) method as described by Roberts et al. [6]. The non-protein-bound fraction of teicoplanin was determined by HPLC after filtration with a Centrifree® device [6]. Standards and controls were prepared in blank plasma. The assay had a linear range of 0.3–26.6 mg/L, a correlation coefficient (r2 ) of 0.9984, a limit of detection of 0.40 mg/L and a limit of quantification of 1.36 mg/L. The accuracy (95% confidence level) was determined as 2.0 ± 0.2 and 20.0 ± 2.3 at the nominal concentration with a precision <5% coefficient of variation (n = 10). 2.4. Pharmacokinetic analysis Non-compartmental PK analysis was performed using previously described methods [9]. The following covariates were evaluated for effects on free teicoplanin concentrations: total teicoplanin concentration; serum albumin concentration; age; MDRD creatinine clearance; and serum creatinine concentration. Following identification of significant covariates in univariate testing (P < 0.2), a multiple linear regression model was constructed to determine the primary determinants of subtherapeutic trough concentrations, which was defined as a total trough concentration (tCmin ) <20 mg/L. Goodness of fit of the model was assessed by the Hosmer–Lemeshow statistic. All statistical analyses employed IBM SPSS Statistics for Windows v.19 (IBM Corp., Armonk, NY). A P-value of <0.05 was considered statistically significant.
2.1. Selection of patients
3. Results
Inclusion criteria were adult patients (aged >18 years) referred for surgical debridement of chronic, deep-seated, Gram-positive bone or joint infections with decreased serum albumin concentrations (<35 g/L; normal range 35–52 g/L) who at the discretion of the physician and/or orthopaedic surgeon were deemed to require high-dose teicoplanin (Targocid® ; Sanofi-Aventis, Midrand, South Africa) regimens for ≥4 days (loading dose of 12 mg/kg q12 h intravenously for 48 h, followed by 12 mg/kg once daily over 30 min). Only patients who had debridement performed on Day 3 of antibiotic therapy were enrolled. Patients who were acutely ill, had suspected glycopeptide allergy, were pregnant or who had moderate-to-severe renal dysfunction [arbitrarily defined as an estimated glomerular filtration rate (eGFR) <59 mL/min/1.73 m2 using the Modification of Diet in Renal Disease (MDRD) equation for GFR calculation] were excluded from enrolment. Chronicity of osteomyelitis or septic arthritis was defined as duration of >4 weeks. Hypoalbuminaemia was defined as a serum albumin level <25 g/L [2]. Demographic characteristics were recorded at
3.1. Patient characteristics Ten patients (eight male and two female) with confirmed Grampositive, chronic, deep-seated bone or joint infections referred to Milpark Hospital for surgical debridement were recruited over a 4-month period. Their demographic data are given in Table 1. All patients had decreased albumin concentrations (<35 g/L) at recruitment. Despite two patients not actually having hypoalbuminaemia as per the definition of <25 g/L (i.e. concentrations of 27 g/L and 34 g/L), the median [interquartile range (IQR)] for the cohort was 18 (15–24) g/L. Small changes in serum creatinine concentrations occurred particularly perioperatively; on Day 3, the median value was 84.0 (65.5–106.0) mol/L. 3.2. Pharmacokinetic profile Teicoplanin free trough concentrations (fCmin ) and tCmin over the 4-day study period are given in Table 2. After 48 h, the
Please cite this article in press as: Brink AJ, et al. Albumin concentration significantly impacts on free teicoplanin plasma concentrations in non-critically ill patients with chronic bone sepsis. Int J Antimicrob Agents (2015), http://dx.doi.org/10.1016/j.ijantimicag.2015.01.015
G Model
ARTICLE IN PRESS
ANTAGE-4536; No. of Pages 5
A.J. Brink et al. / International Journal of Antimicrobial Agents xxx (2015) xxx–xxx Table 1 Demographic characteristics of patients (n = 10) before first teicoplanin administration and perioperatively. Parameter
Median
IQR
Normal value
Age (years) Weight (kg) Height (m) Duration of symptoms (days) Albumin (g/L) WBC count (109 /L) CRP (mg/L) ESR (n = 4) (mm/h) Iron (mol/L) Transferrin (n = 9) (g/L) Saturation (%) Duration of surgery (h)a Blood transfusions (mL) (Days 3–4)a Post-operative high-care admission (h)a Serum creatinine (mol/L)a Day 1 Day 2 Day 3 (n = 9) Day 4 (n = 9) eGFR (MDRD) (mL/min/1.73 m2 )a Day 1 Day 2 Day 3 (n = 9) Day 4 (n = 9)
56 80 1.73 103 18.0 (L) 10.0 (H) 42 (H) 18.5 (H) 5.7 (L) 2.0 (L) 9.0 (L) 3.0 1250 72
44–69 67–98 1.69–1.85 56–201 15–24 5.8–13.2 13–95 10.5–22.50 3.9–9.95 1.85–2.45 7–21 2–4 500–2250 72–120
N/A N/A N/A N/A 35–52 3.92–9.88 <5 0–15 11.6–31.3 2.2–3.7 20–50 N/A N/A N/A 64–104
84.50 86.50 84.0 84.0
72–90 68–89 65.5–106 65.5–96.5
78.5 81 76 86
63 to >90 59 to >90 52.5 to >90 59–88.5
>60
IQR, interquartile range; N/A, not applicable; L, low; WBC, white blood cell; H, high; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; eGFR, estimated glomerular filtration rate; MDRD, Modification of Diet in Renal Disease. a Perioperative patient characteristics.
median (IQR) fCmin and tCmin concentrations were 2.90 (2.67–3.47) mg/L and 15.54 (10.28–19.12) mg/L, respectively; however, both began to decline once the dose was reduced to 12 mg/kg once daily. The median protein binding ranged from 66.3% to 80.8% between Day 1 and Day 4, respectively. The PK parameter estimates of free and total teicoplanin are shown in Table 3. The total drug area under the concentration–time curve from 0 to 12 h (AUC0–12 h ) increased from 137.9 mg h/L on Day 1 to 203.4 mg h/L on Day 4. Clearance (CL) of the free (unbound) concentrations was extremely high relative to the total fraction at 38.6 (IQR 29.9–47.8) L/h and 7.0 (IQR 6.8–9.8) L/h, respectively (P < 0.001). Results of the multiple linear regression model for free teicoplanin demonstrated that total concentration [odds ratio (OR) = 0.015, 95% confidence interval (CI) 0.003–0.052; P = 0.174] did not impact statistically significantly on free concentrations, whereas serum creatinine concentration (OR = 0.013, 95% CI 0.001–0.025; P = 0.034) and serum albumin concentration (OR = 0.120, 95% CI 0.078–0.161; P < 0.001) did. High interpatient variations were found in the free teicoplanin fraction, whilst higher free fractions were seen in patients with lower albumin concentrations (data not shown).
3
Table 3 Pharmacokinetic parameter estimates [median (interquartile range)] of teicoplanin. Total concentrations
Unbound concentrations
Cmax (mg/L)
Day 1 Day 4
20.1 (15.7–22.2) 25.3 (20.4–31.7)
2.6 (2.3–3.1) 3.0 (2.1–4.6)
Cmin (mg/L)
Day 1 Day 4
6.7 (4.0–10.2) 10.3 (8.0–12.5)
2.3 (1.3–2.5) 1.93 (1.04–2.59)
AUC0–12 h (mg h/L) CL (L/h) Vz (L/kg)
Day 1 Day 1 Day 1
137.9 (83.1–168.2) 7.0 (6.8–9.8) 174.1 (101.6–330.5)
28.6 (21.4–33.2) 38.6 (29.9–47.8) 196.6 (195.7–200.8)
Cmax , observed maximum concentration during sampling period; Cmin , observed minimum concentration during sampling period; AUC0–12 h , area under the concentration–time curve during 12-h dosing period; CL, total clearance; Vz , volume of distribution during terminal phase.
4. Discussion To the best of our knowledge, this study is the first to describe the pharmacokinetics of free teicoplanin in non-critically ill patients with hypoalbuminaemia. Hypoalbuminaemia is likely to be relatively common in malnourished or diabetic patients and/or those with chronic infections [2]. In this study, at the time of recruitment the median duration of chronic bone or joint infections was 103 days (IQR 56–201 days) and as a consequence this probably impacted on serum albumin concentrations (median 18 g/L). The data presented in this study provide support for previous studies that have described altered pharmacokinetics of highly protein-bound antibiotics in intensive care unit (ICU) patients with hypoalbuminaemia [2,3,6,10,11]. Interestingly, compared with similar patients with deep-seated infections but normal albumin levels described by Lamont et al. [4], we found a 10fold increase in the median teicoplanin CL (data not shown), which is likely to have resulted from an increased free fraction of teicoplanin that was available for renal clearance in these patients with conserved renal function, and perhaps even augmented renal function. In this study, hypoalbuminaemia (P < 0.001) and not total teicoplanin concentrations (P = 0.174) impacted significantly on free concentrations. As such, it is not clear whether increasing the dose will result in linear increases in free concentrations. However, concentrations of both free and total drug did begin to decline on Day 3 once the dose was reduced to 12 mg/kg once daily (Table 2) and, therefore, 12 mg/kg q12 h should probably be extended beyond 48 h, particularly in patients with large fluid or blood losses during extensive surgery, as was the case here. Despite the statistically significant (P = 0.034) relationship between serum creatinine and free drug concentrations, we feel that this is unlikely to be clinically relevant given the small changes in serum creatinine concentrations that occurred perioperatively. Therapeutic drug monitoring (TDM) of free concentrations to individualise antimicrobial therapy has been recommended for ICU patients [1,2]. It is believed that for antibiotics where TDM targets are based on trough concentrations, the free concentration might be more likely to predict efficacy than the total concentration
Table 2 Teicoplanin free trough concentration (fCmin ) and total trough concentration (tCmin ) observed with 12-hourly dosing (Days 1–2), followed by once-daily dosing (Days 3–4). Day 1
Median (IQR) fCmin (mg/L) Median (IQR) tCmin (mg/L) Median (IQR) protein binding (%)
Day 2
Day 3a
Day 4
Trough 1
Trough 2
Trough 1
Trough 2
Trough 1
Trough 1
0.00 0.00 N/A
2.24 (1.28–2.53) 6.66 (4.02–10.17) 66.3 (54.5–84.1)
2.74 (1.00–2.92) 8.61 (4.12–12.85) 74.8 (52.9–78.6)
2.26 (1.65–2.52) 10.31 (7.52–15.52) 74.2 (61.5–85.7)
2.90 (2.67–3.47) 15.54 (10.28–19.12) 74.8 (68.3–83.6)
1.93 (1.04–2.59) 9.17 (5.63–12.26) 80.8 (66.1–89.2)
IQR, interquartile range; N/A, not applicable. a Surgical debridement.
Please cite this article in press as: Brink AJ, et al. Albumin concentration significantly impacts on free teicoplanin plasma concentrations in non-critically ill patients with chronic bone sepsis. Int J Antimicrob Agents (2015), http://dx.doi.org/10.1016/j.ijantimicag.2015.01.015
G Model ANTAGE-4536; No. of Pages 5 4
ARTICLE IN PRESS A.J. Brink et al. / International Journal of Antimicrobial Agents xxx (2015) xxx–xxx
[1,12]. Hence, to optimise microbiological and clinical outcomes in non-ICU patients with infection and hypoalbuminaemia, the results from this study ‘make the case’ for use of TDM and specifically measurement of free concentrations, probably for all highly protein-bound antibiotics (>80%), including cloxacillin, an agent frequently used in this setting [1]. In fact, Ulldemolins et al. previously documented the impact of hypoalbuminaemia on free flucloxacillin concentrations and the implications thereof, not only for dosing but for alternative administration strategies albeit in ICU patients [11]. As with critically ill patients, it is evident from the results in this chronically ill population that the variability in protein binding and in teicoplanin concentrations makes the prediction of free and total plasma concentrations difficult. For example, one patient in this study had an fCmin of 0 mg/L from Day 2 onwards (data not shown). In an earlier report, a model was proposed to estimate free teicoplanin serum concentrations, however this could not be validated in the Defining Antibiotic Levels in Intensive care unit patients (DALI) teicoplanin study in which highly variable differences between measured and calculated free concentrations were observed [6,13]. This has also recently been shown for -lactams [1]. Hence, the only way to ensure that adequate free concentrations are present and thereby to improve bacteriological and clinical outcomes appears to be by direct measurement of free drug and appropriate dose adjustment. Furthermore, challenges for teicoplanin dosing in these patients that support TDM are that measuring renal function is difficult in these patients and teicoplanin is a renally cleared antibiotic with a terminal half-life much longer than the dosing interval. In addition, the recommended teicoplanin target tCmin for patients with deep-seated infections (≥20 mg/L) was not achieved at steady state in all our patients despite using a high-dose regimen [4,6,7]. Pea et al. have previously shown that using TDM in an antimicrobial stewardship programme significantly improved the likelihood of early optimal teicoplanin exposure [14]. This highlights an important question that was posed by Roberts et al., namely what should the targets for free TDM be [6]? The authors hypothesised that the lower therapeutic range of free plasma teicoplanin trough concentrations should be an fCmin of 1–2 mg/L. This is based on the fact that TDM targets were determined in patients with normal albumin levels and assuming 90% protein binding, the free equivalent range to a tCmin of 10–20 mg/L would be a free trough concentration of 1–2 mg/L. This is the same logic used in recommendations for free phenytoin target concentrations where the total concentration target for phenytoin TDM is divided by 10 using an estimated 10% free fraction [15]. If that is the case, then the higher dosing schedule used in the current study, at least for the first 48 h, appears to be appropriate if free concentrations are used as a surrogate for teicoplanin efficacy as opposed to total trough concentrations, which are currently recommended. Nine of ten patients in the current study had an fCmin > 2 mg/L after 48 h (data not shown). The impact of protein binding on the pharmacokinetics of antibiotics that are moderately (30–70%) or minimally bound (<30%) has yet to be clarified in detail. Wong et al. recently demonstrated marked variability both in free and total concentrations of seven lactam antibiotics in ICU patients irrespective of whether they were highly protein bound or not [1]. Indeed, the variability occurred with other agents active against Gram-positive organisms such as ampicillin, benzylpenicillin, cefazolin and flucloxacillin where protein binding under normal circumstances is 20%, 65%, 80% and 93%, respectively. This was also confirmed to be the case for linezolid (which is usually 31% protein bound) in critically ill patients, particularly if they had decreased albumin concentrations [12].
4.1. Limitations of the study The teicoplanin minimum inhibitory concentrations of the infecting pathogens were not available for all isolates and therefore the study results do not address pharmacodynamic indices (except for PK targets) and their relation to outcome. However, both bacteriological and clinical outcomes would have been confounded by source control, which would probably be more important than antibiotic therapy in this regard [8]. The small cohort of 10 patients could also be considered to be a limitation as it may have prevented other factors that influence PK variability from being identified. Finally, the recommendations derived from the data analysis cannot be extrapolated to other patient populations with chronic bone or joint infections with concurrent albumin plasma concentrations >34 g/L or to those with moderate or severe renal dysfunction (eGFR < 59 mL/min/1.73 m2 ). 5. Conclusions The results of this study highlight the importance of protein binding and the potential role of TDM in optimising free concentrations of highly protein-bound antibiotics in non-critically ill patients with decreased serum albumin concentrations. However, the clinical and bacteriological implications of subtherapeutic teicoplanin concentrations due to hypoalbuminaemia and the monitoring of free teicoplanin concentrations need to be clarified. Given the uncertainty regarding unpredictable protein binding and the impact on pharmacokinetics, it is likely that direct measurement of free antibiotic concentrations should be preferred. In conclusion, if we are to ensure both efficacy and a sustained suppression of antibiotic resistance, antibiotic exposure must be optimised to robust pharmacokinetic/pharmacodynamic targets in all patients. Furthermore, whatever patient population is studied, hypoalbuminaemia is proving to be one of the most important covariates influencing the interindividual PK variability of proteinbound antibiotics. Funding This work was supported by an investigator grant for determination of teicoplanin concentrations from Sanofi-Aventis (Midrand, Johannesburg, South Africa). The funder had no involvement in conception and designing of the study, neither in the collection, analysis and interpretation of data nor in the writing of the report and in the decision to submit the article for publication. JAR is funded by a Career Development Fellowship from the National Health and Medical Research Council of Australia [APP1048652]. Competing interests AJB and GAR have received research grants from or served on scientific advisory boards for Pfizer, Sanofi-Aventis, Aspen/GSK and MSD. All other authors declare no competing interests. Ethical approval Ethical approval was obtained from the local ethics committee of Milpark Hospital (Johannesburg, South Africa) [02.12.2012]. The study was conducted following the guidelines of the Declaration of Helsinki. Written informed consent was obtained either from the patient or, when appropriate, from their closest relative. References [1] Wong G, Briscoe S, Adnan S, McWhinney A, Ungerer J, Lipman J, et al. Protein binding of -lactam antibiotics in critically ill patients: can we
Please cite this article in press as: Brink AJ, et al. Albumin concentration significantly impacts on free teicoplanin plasma concentrations in non-critically ill patients with chronic bone sepsis. Int J Antimicrob Agents (2015), http://dx.doi.org/10.1016/j.ijantimicag.2015.01.015
G Model ANTAGE-4536; No. of Pages 5
ARTICLE IN PRESS A.J. Brink et al. / International Journal of Antimicrobial Agents xxx (2015) xxx–xxx
[2]
[3]
[4]
[5] [6]
[7]
[8]
successfully predict unbound concentrations? Antimicrob Agents Chemother 2013;57:6165–70. Ulldemolins M, Roberts JA, Rello J, Paterson D, Lipman J. The effects of hypoalbuminaemia on optimizing antibiotic dosing in critically ill patients. Clin Pharmacokinet 2011;50:99–110. Mimoz O, Rolland D, Adoun M, Marchand S, Breilh D, Brumpt I, et al. Steadystate trough serum and epithelial lining fluid concentrations of teicoplanin 12 mg/kg per day in patients with ventilator-associated pneumonia. Intensive Care Med 2006;32:775–9. Lamont E, Seaton RA, Macpherson M, Semple L, Bell E, Thomson AH. Development of teicoplanin dosage guidelines for patients treated within an outpatient parenteral antibiotic therapy (OPAT) programme. J Antimicrob Chemother 2009;64:181–7. Soy D, López E, Ribas J. Teicoplanin population pharmacokinetic analysis in hospitalized patients. Ther Drug Monit 2006;28:737–43. Roberts JA, Stove V, De Waele JJ, Sipinkoski B, McWhinney B, Ungerer JPJ, et al.; DALI Study Authors. Variability in protein binding of teicoplanin and achievement of therapeutic drug monitoring targets in critically ill patients: lessons from the DALI Study. Int J Antimicrob Agents 2014;43:423–30. Brink AJ, Richards GA, Cummins RR, Lambson J. Recommendations to achieve rapid therapeutic teicoplanin plasma concentrations in adult hospitalised patients treated for sepsis. Int J Antimicrob Agents 2008;32:455–8. Greenberg RN. Treatment of bone, joint and vascular-access-associated Grampositive bacterial infections with teicoplanin. Antimicrob Agents Chemother 1990;34:2392–7.
5
[9] Sime FB, Roberts MS, Warner MS, Hahn U, Robertson TA, Yeend S, et al. Altered pharmacokinetics of piperacillin in febrile neutropenic patients with haematological malignancy. Antimicrob Agents Chemother 2014;58: 3533–7. [10] Brink AJ, Richards GA, Schillack V, Kiem S, Schentag J. Pharmacokinetics of once-daily dosing of ertapenem in critically ill patients with severe sepsis. Int J Antimicrob Agents 2009;33:432–6. [11] Ulldemolins M, Roberts JA, Wallis SC, Rello J, Lipman J. Flucloxacillin dosing in critically ill patients with hypoalbuminaemia: special emphasis on unbound pharmacokinetics. J Antimicrob Chemother 2010;65:1771–8. [12] Yagi T, Naito T, Doi M, Nagurac O, Yamada T, Maekawa M, et al. Plasma exposure of free linezolid and its ratio to minimum inhibitory concentration varies in critically ill patients. Int J Antimicrob Agents 2013;42: 329–34. [13] Yano R, Nakamura T, Tsukamoto H, Igarashi T, Goto N, Wakiya Y, et al. Variability in teicoplanin protein binding and its prediction using serum albumin concentrations. Ther Drug Monit 2007;29:399–403. [14] Pea F, Viale P, Pavan F, Tavio M, Poz D, Beltrame A, et al. The effect of multifactorial, multidisciplinary educational interventions on appropriate use of teicoplanin. Int J Antimicrob Agents 2006;27:344–50. [15] Patsalos PN, Berry DJ, Bourgeois BFD, Cloyd JC, Glauser TA, Johannessen SI, et al. Antiepileptic drugs—best practice guidelines for therapeutic drug monitoring: a position paper by the Subcommission on Therapeutic Drug Monitoring, ILAE Commission on Therapeutic Strategies. Epilepsia 2008;49: 1239–76.
Please cite this article in press as: Brink AJ, et al. Albumin concentration significantly impacts on free teicoplanin plasma concentrations in non-critically ill patients with chronic bone sepsis. Int J Antimicrob Agents (2015), http://dx.doi.org/10.1016/j.ijantimicag.2015.01.015