Statistics from Altmetric.com
Vancomycin is a widely used drug in children. However, it is a relatively old antibiotic, and, as such, has very little in the way of data to back up how we currently use it in paediatrics. There has been a renewed interest in vancomycin dosing in adults in recent years, with concerns that established dosing regimes are inadequate. This has lead to recommendations for higher trough vancomycin levels to improve therapeutic outcome. There appears to have been increasing unease within the paediatric world, with a recent flurry of publications looking at whether we achieve the target trough levels that are recommended for adult patients. This article will review some of the recent literature and discuss how we might need to move forwards.
History of vancomycin
Vancomycin is a glycopeptide antibiotic which inhibits bacterial cell wall synthesis through inhibition of peptidoglycan synthesis. It is bactericidal against most streptococci and staphylococci, however, it is bacteriostatic against Enterococcus species.1
Vancomycin was derived from a naturally occurring substance isolated from Streptomyces orientalis, first demonstrated to have antibacterial properties in 1952. Vancomycin was quickly demonstrated to be safe and effective, successfully treating serious staphylococcal infections.2 However, not long after these reports of successful therapy were published, reports of ototoxicity began to appear, resulting in the recommendation to monitor serum vancomycin levels.3 Early recommendations were for trough vancomycin levels of 5–10 mg/L and peak levels of 30–40 mg/L, with adult dosing regimes of 1 g every 12 h.4 In 1958, the first synthetic penicillin, methicillin, was approved in America and vancomycin fell out of favour somewhat, due to concerns about toxicity, particularly ototoxicty and nephrotoxicty.5 With the emergence of methicillin-resistant Staphylococcus aureus (MRSA), interest in vancomycin returned and it is now commonly used for MRSA skin, soft tissue and invasive infections, along with treatment for coagulase-negative staphylococci, particularly central venous catheter infections.
Do we know how much vancomycin is needed to be clinically effective?
Because vancomycin was developed in the 1950s, the initial dosing regimes were developed before the advent of detailed pharmacodynamics studies. The information that is available is mainly from in vitro and animal studies, with very little data on paediatric dosing. The underlying premise for therapeutic levels of vancomycin is the need to have a serum concentration at a level which is a multiple of the minimum inhibitory concentration (MIC) of the target organism, while avoiding adverse effects. Effective treatment with vancomycin has been shown to be time dependent, but concentration independent, and the area under the concentration curve (AUC) (figure 1) divided by the MIC is the primary predictor of efficacy in treatment.6 According to 2009 consensus recommendations from the Infectious Diseases Society of America (IDSA), the American Society of Health System Pharmacists, and the Society of Infectious Diseases Pharmacists, an AUC/MIC of 400 is an appropriate target to achieve a successful outcome when treating MRSA.6 ,7 These recommendations were made following review of in vitro studies, animal data and some limited human data; however, they were designed for adult patients.
In recent years, there have been concerns about treatment failures in MRSA infections in adults with organisms with higher MICs, as well as the development of Vancomycin Intermediate Susceptible S aureus (VISA). In 2006, the Clinical and Laboratory Standards Institute lowered the breakpoint for S aureus susceptibility from 4 to 2 mg/L,8 recognising that those organisms with higher MICs are not effectively treated with vancomycin. There has also been concern about the emergence of inducible, heterogenous VISA (hVISA) strains, with MICs in the susceptible range (0.5–2 mg/L), but in patients where standard treatment has failed.9
In children, there is the question of which organisms we are treating and their MICs. Much of the literature is based on treatment of MRSA with an MIC of 1 mg/L, which may not be applicable on the neonatal or paediatric ward. Many coagulase-negative staphylococci and 40% of MRSA in Europe have vancomycin MIC of 2 mg/L.10 This has implications for the AUC24 required for therapeutic success being much higher than levels required for organisms with an MIC 1 mg/L.
How does AUC/MIC correlate to trough vancomycin levels?
Measuring AUC is not a part of routine clinical practice, however, measurement of trough antibiotic levels is commonly used. The 2009 consensus guidelines determined that, in adult patients, to achieve an AUC:MIC of 400, for an organism with an MIC of 1 mg/L, a trough vancomycin level of 15 mg/L is required. They also recognised that recommended dosing regimes in adults at the time were probably inadequate to achieve this trough level. These 2009 guidelines did not address the issue of paediatric dosing.
It was not until 2011, when further guidelines were published by the IDSA regarding treatment of MRSA infections, that there were any recommendations for the use of vancomycin in children. These 2011 guidelines recommended a dosing regime of 15 mg/kg every 6 h in children, aiming for trough levels of 15–20 mg/L in serious infections. However, they recognised that limited information was available to make these recommendations, grading the evidence as B-III (moderate evidence to support recommendation, from expert opinion and descriptive studies only).11 This dosing regime contrasts with the 15 mg/kg 8 h regime that is currently recommended in the UK British National Formulary (BNF) for children (BNFc). However, the BNFc still suggests similar trough levels, 10–15 or 15–20 mg/L for ‘less sensitive strains’ of MRSA. This highlights the lack of clarity on this issue, with two different dosing regimes aiming for the same trough levels! It also provides no information about how trough levels correlate with AUC/MIC in children.
What dose of vancomycin is needed in children?
The increasing interest in the issue of vancomycin dosing in children in recent years has generated a number of retrospective reviews of practice. These highlight the wide variation in dosing regimes and therapeutic trough levels used by different institutions, at different time periods (summarised in table 1). Many studies have looked at the attainment of vancomycin trough levels at a range of 15–20 mg/L in children, as this is the required level to achieve an AUC:MIC>400 in adults.6 These show that few children achieve these trough vancomycin levels. One of the larger studies, published in 2011, was carried out by Eiland et al,12 who reviewed 435 trough vancomycin levels from 295 children over 5 years. Over time, vancomycin doses increased, associated with an increase in recommended trough levels from 2005 to 2008 compared with 2009 to 2010 (5–15 to 10–20 mg/L, respectively). Mean total daily dose increased from 48 to 59 mg/kg/day. Despite this increase in dose, only 49% of the samples from 2009/2010 reached therapeutic trough levels. Using the data gathered they developed a linear equation for vancomycin dosing and calculated that a total daily dose of 70 mg/kg would be required to achieve a trough of 10 mg/L and a dose of 85 mg/kg to achieve a trough of 15 mg/L.12 However, when these values were further evaluated by Goutelle et al,1,3 using pharmacokinetic modelling, they found only 40% would achieve a trough of 10 mg/L at 70 mg/kg/day and only 50% would achieve a trough of 15 mg/L at 85 mg/kg/day.
Frymoyer et al compared two different dosing regimes, 45 and 60 mg/kg/day, at two different time points in 182 children. This followed a hospital decision to increase their dosing regime in 2008. Despite this, the mean initial trough level only increased from 7 to 9 mcg/mL, and only 14% of those who received the higher dose achieved initial trough levels in the range 15–20 mcg/mL.14
Miller et al1,5 studied 232 vancomycin trough levels from 187 patients. Their primary objective was to look at differences in ability to achieve a therapeutic level in normal weight and obese children using a variety of different dosing regimes, including 6, 8 and 12 h dosing. There was no difference in the percentage achieving initial troughs in the range 5–15 mg/L in the obese and normal weight groups. However, they also highlighted that only 22% achieved troughs in the range 10–20 mg/L.
Gordon et al1,6 found that dosing frequency affected trough vancomycin concentration, despite total daily dose achieving a potentially therapeutic AUC. No children achieved troughs in the target range 15–20 mg/L on initial sampling, but further analysis of repeat samples demonstrated that 43% children receiving 6 h dosing achieved target levels after a median of 8 days. They also noted that children under the age of 6 years had significantly lower vancomycin levels despite similar daily doses.
Having established that the majority of children do not reach vancomycin trough levels of 15–20 mg/L, it would be reasonable to assume we are falling short of where we need to be with paediatric dosing to achieve an AUC:MIC>400. However, a number of recent articles have called this into question. Within the recent paediatric literature, there have been some attempts to directly measure AUC:MIC rather than trough vancomycin levels. Silva et al1,7 published a small series of haematology/oncology patients in 2012. Using an MIC of 1 mg/L, they found a mean dose of 81 mg/kg/day was required to achieve an AUC:MIC>400. Doses of 40–60 mg/kg/day only achieved AUC:MIC>400 in 50% of patients. Trough vancomycin concentrations>15 mg/L had a 100% positive predictive value, but 71% negative predictive value for having AUC:MIC>400, suggesting lower trough levels may also achieve a therapeutic AUC:MIC.
Chhim et al18 attempted to calculate AUC for children on vancomycin. This was a complex process involving estimation of creatinine clearance with the Schwartz equation which was then used to estimate vancomycin clearance using further mathematical formulae. A 24 h AUC (AUC24) was then calculated from vancomycin daily dose/vancomycin clearance. They were able to calculate AUC24 for 190 patients of the 200 receiving either 40 or 60 mg/kg/day vancomycin. A lack of documented height prevented calculation for the remaining 10 patients. Only 17% and 40% of patients achieved AUC24 >400 with 40 and 60 mg/kg/day dosing respectively (equivalent to AUC:MIC >400 if MIC of the organism is 1 mg/L). They also demonstrated a poor correlation between trough levels and AUC24.
Le et al have compared AUC:MIC with trough vancomycin levels. They took 1660 samples from 702 patients from two large American children's hospitals. Using Monte Carlo simulations they were able to demonstrate that a total daily dose of 60 mg/kg/day would achieve AUC:MIC >400 in 75% cases in children over 12 years of age. However, children under the age of 12 years of age required higher doses, with 70 mg/kg/day achieving the desired MIC:AUC three-quarters of the time. Interestingly, they calculated that an AUC:MIC 400 was only equivalent to a trough vancomycin level of 8–9 mg/L.19 This raises the question of whether we should be targeting the same trough levels as recommended for adults. Recently, further pharmacokinetic modelling and simulation has been undertaken by Frymoyer et al.20 They performed analysis for doses of 15 mg/kg 6 h, 15 mg/kg 8 h and 20 mg/kg 8 h. They also evaluated MICs of 0.5, 1.0 and 2.0 mg/L. Using an assumed MIC of 1.0, dosing of 15 mg/kg 6 h 90% cases achieved an AUC:MIC >400 with trough levels of 7–10 mg/L. For 20 mg/kg 8 h a trough level of 6–8 mg/L was predictive of AUC:MIC >400. However, for 15 mg/kg 8 h trough levels of 8–10 mg/L were needed for the target AUC:MIC. When a lower MIC (0.5 mg/L) was analysed, 100% of cases achieved AUC:MIC >400 with trough levels greater than 5 mg/L when using 15 mg/kg vancomycin 6 h. However, with an MIC 2 mg/L very few cases achieved AUC:MIC >400 with 6 h dosing, even when trough levels of 15–20 mg/L were achieved. This suggests that aiming for higher trough levels may not be necessary if the MIC of the organism being treated is 1 mg/L or less. However, it was purely a modelling study and cannot be extrapolated to children where vancomycin clearance may be altered, for example, in young children, obese children and those with impaired renal function.
Will increasing doses lead to increased toxicity?
These attempts to directly measure or model AUC/MIC demonstrate the complexity of vancomycin dosing in children. For organisms with vancomycin MICs of 1 mg/L or less, a trough level of 15–20 mg/L may be excessive, however, for those infections caused by organisms with higher MICs, these levels may be required. This raises the potential for increased toxicity with higher vancomycin doses. Initial concerns about vancomycin toxicity, which lead to the drive to measure levels, were associated with the impure form, or ‘Mississippi mud’ as it was known.21 With better development techniques, a purer form of vancomycin resulted in less toxicity. However, nephrotoxicity continues to be problematic in adults, and concerns have been raised with the recent move to higher doses and higher trough levels. A recent American single centre study found an overall rate of nephrotoxicity of 14% in 167 children treated with vancomycin, despite the hospital having a vancomycin-dosing regime which varied with age and serum creatinine (30–60 mg/kg/day).22 However, nephrotoxicity was significantly more common in those with high troughs (mean >15 mg/L across all vancomycin treatment), when compared with those with lower troughs (mean <15 mg/L across whole treatment) at 28% compared with 7%. There was no statistically significant difference in mean dose at the time that toxicity occurred, 46.6 mg/kg/day compared with 49.6 mg/kg/day, in the low trough and high trough groups. Toxicity occurred within the first week of vancomycin therapy rather than during prolonged therapy. Factors associated with an increased risk of nephrotoxicity were concomitant use of furosemide and intensive care admission. No patients needed to stop vancomycin as a result of nephrotoxicity. No patients required dialysis; 46% had returned to baseline serum creatinine levels by the end of therapy and 75% had serum creatinine levels close to baseline at discharge from hospital. It is reassuring that no child appears to have suffered significant lasting damage, however, if we are going to move to using higher vancomycin doses to achieve higher trough levels, we will need to be aware of the potential for increasing numbers of children to have some degree of renal dysfunction.
What about a loading dose and/or continuous infusions of vancomycin?
Other areas of interest in adult medicine, which may be beneficial in the paediatric population, are the use of a loading dose and continuous infusions. The use of a loading dose to rapidly achieve a therapeutic trough level, is recommended in seriously ill adult patients in the IDSA guidelines.6 There is very limited data regarding the use of a loading dose in children. The study by Gordon et al1,6 did include a small number of patients who received a loading dose, but this was not consistent across the population studied. A continuous infusion might be beneficial in providing greater time at a concentration above the MIC of the organism being treated and could make both administration and monitoring easier, provided there is adequate venous access. These issues are also of interest in the neonatal population, and there have been some recent publications suggesting the use of a loading dose and then continuous infusion for neonates to achieve better vancomycin levels.23 ,24 This needs further investigation in children to establish the correct loading dose and total daily dose required in the infusion to achieve optimal vancomycin levels.
It is clear that we do not yet have a full understanding of the best dosing regime for vancomycin in children. In fact, we do not really know what levels we are aiming for or the best way to administer it. We can say that we need to achieve AUC/MIC>400. Previously recommended dosing regimens, for example, 40–45 mg/kg/day, are too conservative and produce very low trough levels and inadequate AUC:MIC, unless the infection being treated is caused by an organism with a very low MIC (<1 mg/L). Much of the recent literature would suggest a move towards at least 60 mg/kg/day to improve trough levels, as has been recommended in the IDSA guidelines.11 Whether we need to be aiming for trough vancomycin levels of 15–20 mg/L, as in adult patients, remains questionable and depends on the MIC of the organism being treated. Further work is required to clarify this, particularly in light of the work by Le et al19 and Frymoyer et al20 who have both demonstrated that it may be possible to achieve an AUC:MIC>400 with trough levels below 10 mg/L if the MIC is 1 mg/L. This also highlights the need to understand the infections that vancomycin is being used to treat, for example, MRSA or coagulase negative staphylococci, and the MICs of these organisms. The use of pharmacokinetic modelling studies can contribute to the discussion about vancomycin dosing. However, they cannot provide all the answers, and further trials will be necessary in children with a wide range of ages, weights and renal function. The drive to obtain trough levels in the same range as required for adults also raises potential concerns about increased toxicity which needs further investigation and monitoring.
Further, large-scale studies are vital to determine the most appropriate dosing of vancomycin in children of different ages, in terms of total daily dose and whether this is provided in intermittent doses or continuous infusion. While awaiting these, it is perhaps time to consider moving to 15 mg/kg 6 h as a standard starting regime for vancomycin. It is also vital, for the individual patient, to determine the MIC of the organism being treated, as this may give some guidance about suitable trough levels to be aimed for. There is currently little evidence to guide the use of loading doses or continuous vancomycin infusions in children, but these may be considered in individual cases.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.