Vancomycin Toxicity in Neonates: A Review of the Evidence

Jodi M. Lestner; Louise F. Hill; Paul T. Heath; Mike Sharland


Curr Opin Infect Dis. 2016;29(3):237-247. 

In This Article


Quantification of Nephrotoxicity

Kidney function is difficult to define in the first month of life when renal physiology and fluid balance evolve rapidly. Serum creatinine (SCr) concentrations reflect maternal renal function in the first 48 h after birth.[11,12] In addition, preterm infants experience a rise in SCr due to tubular reabsorption, making this an imprecise proxy for glomerular filtration.[13] Clinical reports also often omit details of the laboratory methods used to quantify SCr (Jaffé colorimetry, enzymatic quantification or isotope dilution mass spectroscopy), which can have an intertest variability of up to 25% in neonates.[14,15] Despite these limitations, most studies define renal impairment as an increase in SCr.[16–19] Definitions for children and neonates have been reported based on several classification systems, all of which were originally developed and validated in adults.[20,21] Modified neonatal criteria exclude urine output or use a higher cutoff (<1.0 ml/kg/h) in order to account for the nonoliguric renal dysfunction that occurs in neonates due to fluid redistribution and impaired absorption in the immature renal tubule.[22,23]

Preclinical Studies

Murine models have demonstrated that vancomycin has an affinity for biological membranes and accumulates in renal tissue.[24,25] The energy-dependent transport mechanisms within the tubular epithelium render the kidney highly susceptible to toxin-induced cellular injury. Vancomycin enhances cellular adenosine triphosphate concentrations and stimulates oxygen consumption, resulting in oxidative phosphorylation.[26–31] However, oxidative stress of sufficient magnitude to cause clinically detectable renal impairment has not been demonstrated in animals, even at supratherapeutic doses.[32] Histopathological changes and biochemical evidence of nephrotoxicity have been observed in an uninfected murine model in which high-dose vancomycin was coadministered with tobramycin.[25]

Clinical Studies

The linear relationship between SCr and vancomycin exposure reflects the drug's primary route of elimination through renal excretion. As a result, a decrease in renal function from any cause will increase serum vancomycin concentrations, a fact that confounds the establishment of exposure–toxicity relationships.[33] Many studies that describe vancomycin-associated nephrotoxicity are conducted in patients otherwise at risk of renal impairment, such as those requiring intensive care and frequently include those receiving other potentially nephrotoxic drugs (e.g. nonsteroidal anti-inflammatory drugs, diuretics and aminoglycoside antibiotics). However, irrespective of the difficulty in defining causal relationships, acute kidney injury from any cause is associated with a significantly increased risk of mortality, hospital length of stay and cost, even after adjustment for age, sex, chronic kidney disease and comorbidities at admission.[34]

Adult Studies

Nephrotoxicity in patients receiving vancomycin has been more systematically reported in adults than in neonates and children. The reported incidence of nephrotoxicity in adults receiving vancomycin is 10–15%.[35] In a recent meta-analysis, van Hal et al.[36] reviewed 15 studies in which vancomycin was administered via intermittent infusion to adults above 18 years of age. Randomized trials and observational studies in which data on renal function and trough vancomycin concentrations were available were included. The analysis found trough concentrations more than 15 μg/ml to be independently associated with an increased odds of nephrotoxicity (odds ratio, 2.67; 95% confidence interval, 1.95–3.65), and also suggested an incremental increase in nephrotoxicity risk with vancomycin administration beyond 7 days. The majority of reported cases were self-resolving, with prolonged impairment requiring short-term dialysis occurring in 3% of cases. Similarly, a retrospective cohort study of 246 adults receiving intermittent-dosing of vancomycin for durations of more than 48 h found toxicity to be correlated with exposure (based on dose). Patients receiving at least 4 g daily (n = 26) had a significantly higher incidence of nephrotoxicity compared with those receiving less than 4 g daily (34.6 vs. 10.9%; P = 0.001).[18]

Concomitant aminoglycoside use has consistently been shown to increase the risk of nephrotoxicity in adults receiving vancomycin. In 2007, Fowler et al.[37] reported a seminal open-label randomized trial of daptomycin vs. either antistaphylococcal penicillin or vancomycin and low-dose gentamicin in 236 patients with S. aureus bacteremia. The highest proportion of renal impairment occurred in patients receiving vancomycin and gentamicin (20.4%), although this was not significantly higher than with penicillin and gentamicin (18.6%).

Furthermore, the study found that the highest incidence of renal impairment in patients receiving vancomycin occurred later in treatment, peaking at 14–28 days.

Paediatric and Neonatal Studies

Vancomycin administration with intermittent dosing and in accordance with published guidance has been shown to result in subtherapeutic trough concentrations in up to half of treated neonates.[4] Furthermore, antimicrobial point prevalence surveys that describe prescribing practices have demonstrated that significant variation in prescribing of vancomycin in neonates exists in clinical practice, with a large proportion of neonates receiving daily doses of vancomycin that are significantly below even the most conservative current recommendations.[38] Table 2 [3,23,33,39–44,45,46–66] outlines the clinical studies that have reported nephrotoxicity associated with vancomycin use in neonates and infants. The majority of studies identified are heterogeneous observational or opportunistic pharmacokinetic studies based on routine therapeutic drug monitoring and are insufficiently powered to detect toxicity.

As with adults, cases of renal impairment are more frequently reported in neonates and children receiving concomitant nephrotoxic drugs.[19,35] Eight of 10 studies describing concomitant nephrotoxic drugs alongside vancomycin therapy demonstrated evidence of renal impairment, which was mild and transient in all reported cases (see Table 2). Other reports have not found vancomycin to be an independent predictor of nephrotoxicity. Constance et al.,[23] for example, found no significant difference in the proportion of neonates developing nephrotoxicity in those receiving vancomycin and gentamicin (12 of 533, 2.2%) vs. those receiving gentamicin alone (seven of 533, 1.3%). Logistic regression demonstrated that while positive blood culture, low birth weight, patent ductus arteriosus, concomitant nonsteroidal anti-inflammatory drug use and illness severity were all independent risk factors for nephrotoxicity, vancomycin in conjunction with gentamicin was not. Six studies report rates of nephrotoxicity in neonates receiving vancomycin alone. A transient rise in SCr, microproteinuria and elevated N-acetyl-[beta]-D-glucosamine was reported in only three, two and one patient, respectively (see Table 2).[51,54,59–61]

Exposure–toxicity relationships have not been clearly defined for vancomycin in neonates. Two case reports have described clinical outcomes in four neonates receiving accidental overdoses of up to 10-fold the maximum recommended dose.[51,59] All four patients developed transient renal impairment, though all had normal renal function at 6 months. It is noteworthy that these cases all involved single or brief exposures that were managed with immediate vancomycin withdrawal, and so these findings do not address the potential risk of cumulative exposure.

A number of recent studies have described the administration of vancomycin via continuous infusion in neonates. Continuous infusion has the theoretical advantage of maintaining constant plasma concentrations, meaning that overall drug exposure can be increased without a rise in peak concentrations. Studies investigating continuous vancomycin infusions in neonates have, to date, been based on studies that defined toxicity differently and involved small patient cohorts that received varied dosing regimens. A loading dose of 7–15 mg/kg was given in three studies. Collectively, the results of these studies suggest that continuous infusion may result in a higher proportion of patients achieving target concentrations between 15 and 20 μg/ml. Reported rates of nephrotoxicity do not differ significantly from those in patients receiving intermittent dosing, and there is currently no evidence to suggest the use of a loading dose that increases the risk of nephrotoxicity.[3,45,47,60] Continuous infusions may, however, be impractical in the neonatal population where venous access is limited, and may lead to periods without effective antibiotic cover if access is lost. There have, as yet, been no systematic studies comparing the safety and efficacy of continuous infusion with high-dose intermittent regimens that specifically target trough concentrations of more than 15 μg/ml.