Every bug counts : neonatal colonization and infection of Gram-negative bacilli - aspects on antibiotic resistance in Sweden and Ecuador

Abstract: Neonatal infections caused by extended-spectrum beta-lactamase (ESBL)-producing Gramnegative bacilli are associated with high morbidity and mortality. In order to perform targeted preventive interventions against the impacts of antibiotic resistance in neonates, it is crucial to study colonization, infection and the spread of EPE-bacteria. Methods and results Study I. A prospective observational study where the proportion of intestinal colonization with EPE, their resistance pattern and risk factors of EPE-colonization were assessed in a neonatal intensive care unit in Ecuador. In total, 56% of the neonates were colonized with EPE. The strains found were ESBL-E. coli (ESBL-EC, 89%) and ESBL-K. pneumoniae (ESBL-KP, 11 %) and the main risk factor for colonization was length of hospital stay. Two of the isolated clones were epidemic and known to disseminate carbapenem-hydrolysing beta-lactamases. These results underline the necessity of implementing colonization surveillance and improved hygiene standards in this and similar neonatal care settings. Study II. A prospective cohort study where we analysed the ESBL-KP isolates in a neonatal outbreak of EPE and determined the duration of intestinal colonization of EPE in affected neonates (n= 14). The intestinal relative abundance of EPE was determined in each carrier. One fourth of the neonates were still carriers of EPE, two years after a NICU outbreak. The median length of colonization was 12.5 months. The low virulent but highly resistant ESBLKP strain ST101 persisted in 2/13 children. One patient was colonized with ESBL-EC at five years of age. No infant suffered from an EPE-infection during the 5-year follow-up. Study III. In this study, the findings in study 2 were extended by characterizing the resistance encoding plasmids using optical DNA mapping (ODM) combined with Cas9- assisted identification of resistance genes. The method detected two plasmids; one small (80 kb) and one large plasmid (220 kb) in all ESBL-KP isolates. We found that the blaCTX-M-15 gene was located on the small plasmid and that this plasmid was stable in the ESBL-KP clone for two years. There was an unrelated acquisition of an ESBL-EC strain, contradicting plasmid transfer between KP and EC in this outbreak. ODM is a promising and rapid tool for surveillance and infection control in clinical settings. Study IV. In a population-based retrospectively matched cohort study, we investigated the incidence, mortality and bacterial characteristics of neonatal sepsis caused by Gram-negative bacilli (GNB) sepsis in Stockholm County during a 11-year period. The primary outcome was death before discharge and secondary outcomes were death within 5 and 30 days after sepsis onset. The cumulative incidence of GNB-sepsis was 0.35 cases per 1,000 live born and for early onset sepsis (EOS) 0.11 and late onset sepsis (LOS) 0.24 respectively. Case fatality rate was 5/33 (15%) in GNB-EOS and 26/74 (35%) in GNB LOS. Neonates with GNB-LOS were 3.9 times more likely (adjusted OR, 95% CI: 1.6-9.4) to die before discharge compared to the uninfected matched control group. Overall, 6.5% (7/107) of the isolates were multidrug-resistant. The incidence of both GNB-EOS and GNB-LOS was lower than reported in previous studies from other high-income countries comparable settings. Mortality in GNB-LOS remains high. The occurrence of acquired antibiotic resistance was low and did not change substantially over time. Conclusions A high proportion (56%) of the neonates at a NICU in a tertiary hospital in Ecuador, became colonized by EPE. These EPE included two clones of ESBL-KP that are known to disseminate carbapenemases. Infants that were colonized during an EPE-outbreak at two NICUs in Stockholm carried the same ESBL-KP for 26 months. There was no plasmid transfer between bacteria during the outbreak. During a 11-year period in Stockholm, Gramnegative sepsis was rare but a major risk for neonatal mortality.