Neonatal late-onset sepsis (LOS) remains a major clinical problem in intensive care for preterm infants. Vancomycin is widely used to treat Gram-positive LOS in neonates, but optimal pharmacokinetic/pharmacodynamic targets in very-low-birth-weight (VLBW) infants are not well established. This multicenter observational study sought to identify vancomycin exposure thresholds associated with early clinical and microbiologic effectiveness in VLBW neonates with LOS, using clinical and microbiological outcomes measured within the first 72 hours of therapy.
The investigation enrolled VLBW neonates who received vancomycin for LOS and had at least one steady-state trough concentration measured. Patients were classified by outcome during the first 72 hours of vancomycin therapy into two groups: early clinical and microbiological cure (ECMC) and late clinical and microbiological cure. The primary exposure metrics assessed were area under the concentration–time curve over 24 hours (AUC24) and steady-state trough concentrations. The investigators used classification and regression tree analysis and receiver operating characteristic (ROC) curve analysis to identify vancomycin exposure thresholds predictive of ECMC.
Thirty-six VLBW neonates with confirmed Gram-positive LOS were included. Of these, 27 patients (75%) achieved ECMC within 72 hours. A statistically significant correlation was observed between AUC24 and trough concentrations (R2 = 0.568; P < 0.001), indicating a moderate relationship between the two exposure metrics in this cohort.
Comparing groups, both AUC24 and trough concentrations were significantly higher among those who achieved ECMC. Reported median (interquartile range) values were:
Using classification and regression tree and ROC approaches, the investigators identified an AUC24/MIC threshold of ≥253 mg·h/L as the optimal exposure associated with ECMC. This threshold demonstrated better discriminative performance than trough concentrations; reported area under the ROC curves were 0.813 for the AUC24/MIC threshold versus 0.726 for trough concentrations, with the authors reporting corresponding P values of 0.005 and 0.045 respectively.
The analysis suggests that achieving an AUC24/MIC ≥253 mg·h/L is associated with higher likelihood of early clinical and microbiologic cure in VLBW neonates with Gram-positive LOS. While trough concentrations correlated with AUC24 and were higher in early responders, the AUC24/MIC metric provided superior discrimination in this dataset. The findings support consideration of AUC-guided exposure assessment rather than reliance on trough levels alone when evaluating vancomycin effectiveness for early response in this specific neonatal population.
For clinicians caring for VLBW infants with LOS, the study provides evidence that higher vancomycin exposure is associated with earlier clinical and microbiologic resolution. Specifically, targeting an AUC24/MIC ≥253 mg·h/L may improve the probability of early cure compared with strategies based only on trough concentrations. Because AUC24 correlates with trough, trough monitoring can provide partial information about exposure, but AUC-guided dosing and therapeutic drug monitoring that estimate AUC24 may offer better predictive performance for early response.
Practically, the data suggest that when feasible, estimating or calculating AUC24 (and considering MIC where available) should be preferred for exposure assessment in VLBW neonates receiving vancomycin for Gram-positive LOS. Clinicians should balance the goal of achieving this exposure threshold with considerations of safety and local susceptibility patterns.
This study was a multicenter observational analysis of 36 VLBW neonates; the sample size is limited. The authors emphasize that the identified AUC24/MIC ≥253 mg·h/L threshold requires prospective validation in larger cohorts before broad clinical adoption. The abstract does not report detailed patient-level characteristics, dosing regimens, MIC distributions, or safety outcomes beyond the presented exposure–response relationships, and those details were not provided in the source abstract. Therefore, clinicians should interpret the threshold in the context of local practice, microbial susceptibilities, and available therapeutic drug monitoring capabilities.
In this cohort of VLBW neonates with Gram-positive late-onset sepsis, both AUC24 and trough concentrations correlated with early clinical and microbiologic cure, but an AUC24/MIC ≥253 mg·h/L showed superior discriminative performance for predicting early response. The authors conclude that AUC-guided exposure assessment may be a better predictor of early treatment success than trough concentrations in this population, and they call for prospective validation of the exposure threshold in larger studies.
(Conflict of interest statement: the authors reported no funding or conflicts of interest.)