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 Ototoxicity in Neonates

The assessment of hearing in infants is challenging. Behavioural tests (e.g. visual reinforcement audiometry) are the gold standard, but cannot be performed reliably in children below ~8 months.[67] The methods currently employed to quantify auditory function in neonates are otoacoustic emissions (OAEs) and auditory brainstem responses (ABRs). Serial diagnostic OAE or ABR testing has been used to monitor ototoxicity in children receiving aminoglycosides.[68–71] There are no equivalent reports describing serial OAE or ABR following vancomycin use. Diagnostic OAE or ABR is labour-intensive and difficult to interpret in premature infants (postmenstrual age <34 weeks) due to immaturity of the cochlea and central auditory pathways.[72–78]

Universal newborn hearing screening (UNHS) uses a combination of automated OAE and ABR, and is designed to identify severe permanent hearing impairment in term neonates. Standardized UNHS programmes are now being implemented in the United States and Europe.[79,80] UNHS is a single-point assessment and is not designed to detect and monitor subtle high-frequency hearing loss that typically occurs in drug-induced ototoxicity.

Preclinical Studies

The association between vancomycin use and hearing impairment is controversial. As in the case of nephrotoxicity, the mechanism of vancomycin-induced ototoxicity is thought to involve dose-dependent intracellular oxidative damage, which leads to the loss of cochlear sensory hair cells resulting in high-frequency hearing loss.[81] Ototoxicity has not been consistently demonstrated with vancomycin in animal models.[25,82–84] Studies in guinea pigs found no evidence of vancomycin ototoxicity, but found that vancomycin increases the probability of ototoxicity, measured by OAE, when coadministered with gentamicin.[85]

Clinical Studies

The majority of vancomycin-associated ototoxicity was reported early in the drug's use in patients treated with indeterminate doses of an impure fermentation product who often received concomitant therapy with other severely ototoxic agents.[86,87] Overall, the available recent literature on vancomycin-associated ototoxicity is heterogeneous and in many cases causation is uncertain.

Adult Studies

Observational studies have suggested that the risk of ototoxicity may be increased with vancomycin doses at least 4 g daily in adults. In a retrospective study of 89 adults receiving more than 14 days of high-dose vancomycin, Forouzesh et al.[88] reported high-frequency hearing loss identified by pure-tone audiometry in 10 patients (12%), a rate significantly higher than reported at standard doses. However, the study included patients receiving concomitant aminoglycosides and diuretics, and identified no significant difference in mean trough vancomycin concentrations in patients with and without abnormal audiometry, suggesting the phenomenon was unlikely to be dose-dependent.

Paediatric and Neonatal Studies

Ototoxicity is infrequently reported in neonates and children treated with vancomycin in general.[89] Buckingham et al.,[90] however, reported a high incidence of hearing loss in children with pneumococcal meningitis treated with vancomycin. Patients aged 5–20 years (n = 109) received vancomycin alongside a third-generation cephalosporin. Over half of those surviving had significant permanent hearing loss (n = 37, 55%). Of note, hearing loss was independently associated with vancomycin administration within 2 h of presentation. Whether these findings apply to other populations or reflect the specific pathophysiology of pneumococcal meningitis is uncertain. The findings do, however, highlight the need to consider population and disease-specific factors when reporting drug toxicity.

In neonates, reporting of vancomycin-associated ototoxicity comes almost exclusively from routine UNHS. De Hoog et al.[91] studied 625 neonates over 2 years from a single neonatal intensive care unit in the Netherlands. Exposure to vancomycin (alone or in combination with tobramycin or furosemide) was not associated with a significant rise in screening failure rates. Similarly, in a retrospective study of more than 7000 infants, Gopel et al.[92] found no association between vancomycin exposure and UNHS failure rates. Vella-Brincat et al.[93] described UNHS outcomes over 5 years from a single centre in New Zealand. The cohort included 41 neonates who received vancomycin, and found higher failure rates (n = 6; 22%), compared with those not exposed to vancomycin (n = 85/1,233; 7%). The significance of these findings is uncertain given the lack of systematic follow-up and many potential confounding factors.