Pregnant Swiss-Webster mice were vaginally inoculated with 5 x 10(4) Escherichia coli K1 strain LH (O75:K1:H3) or C94 (O7:K1:H-). Inhibitor solutions were applied vaginally before delivery and the incidence of bacteremia and surface colonization determined in neonates at 3 days of age. E. coli K1 strain LH resulted in bacteremia in 77% and colonization in 74% of control newborn mice. After topical maternal vaginal D-mannose treatment, bacteremia and colonization were present in 25% of neonates. Topical vaginal application of subinhibitory concentration of gentamicin reduced bacteremia to 23% of neonates. Topical methyl-alpha-D-mannoside and p-nitrophenyl-D-mannoside, however, prevented bacteremia in 100% of newborn mice. A neonatal meningitis strain of E. coli K1 (C94) caused bacteremia in 100% of neonates and was also completely inhibited by methyl-alpha-D-mannoside. This technique of vaginal treatment before delivery may have applicability to human mothers and their infants.

Pregnant Swiss-Webster mice were vaginally inoculated with 10(5) virulent and avirulent serotype III Streptococcus agalactiae and treated 4 days later with topical vaginal inhibitor solutions. Preparations containing lipoteichoic acid (LTA) or glycerophosphate (GP), the repeating linear backbone of LTA, significantly reduced neonatal colonization and bacteremia by the virulent isolate and colonization by the avirulent strain. Similar results were obtained if bacteria were preincubated with LTA or GP at 37 degrees C for 30 min before vaginal inoculation. Human serum albumin (HSA), a known inhibitor of binding of LTA to human fetal epithelial cells, also resulted in reduction in colonization and bacteremia of neonatal mice. However, maternal treatment with a combination of HSA (2%) and GP (1%) completely prevented neonatal colonization and bacteremia without altering the normal aerobic bacterial vaginal flora. These results provide impetus to the development of an alternative means of preventing neonatal group B streptococcal infections in humans without requiring maternal immunization or chemoprophylaxis.

Adherence of group B streptococci to buccal epithelial cells was determined in 81 neonates with different gestational ages. There was a significant curvilinear increase in adherence in premature neonates with decreasing gestational ages. Amniotic fluid did not inhibit attachment of group B streptococci to epithelial cells. Increased adherence of group B streptococci to premature neonates may explain, in part, the increased susceptibility of these infants to disease from this organism.

No abstract

Eighteen patients 2 months to 11 years of age with culture proven bacterial infections were treated with parenteral ticarcillin/clavulanic acid in a noncomparative study. Seven patients had pneumonia, two had tracheobronchitis, three had soft tissue abscess, two had periorbital cellulitis, three had urinary tract infection and one had purulent bursitis. Four of the 18 were bacteremic. Organisms treated included Staphylococcus aureus (6), Pseudomonas aeruginosa (5), Haemophilus influenzae (2), Branhamella catarrhalis (2), Escherichia coli (1), Streptococcus pneumoniae (1), Klebsiella pneumoniae (1), Streptococcus pyogenes (1) and Serratia marcescens (1). Thirteen of 15 (87%) organisms tested were beta-lactamase positive. Therapy was given intravenously in six doses per day at 310 mg/kg. Duration of treatment ranged from 5 to 28 (mean 11) days, with an average time of 4 days to clinical improvement. Seventeen patients (94%) were clinically cured. One patient with recurrent aspiration pneumonia due to mixed infection with multiple gram-negative enteric bacilli failed therapy. Adverse effects were minimal and transient. Notably, mild to moderate thrombocytosis occurred in four (22%) patients that resolved uneventfully. We conclude that ticarcillin/clavulanic acid is safe and effective therapy for serious infections in hospitalized children.