Peter de Man

Intestinal colonization with enterobacteriaceae in Pakistani and Swedish hospital-delivered infants.

ABSTRACT. Rectal cultures from Swedish and Pakistani hospital‐delivered newborn infants were analysed regarding the early aquisition of enterobacteria. Swedish infants were delivered vaginally, Pakistani infants were delivered either vaginally or by caesarean section. The Swedish infants were all breast‐fed, whereas breastfeeding was incomplete and often started late among the Pakistani infants. Both groups of Pakistani infants were more rapidly colonized with enterobacteria than were the Swedish infants. Cultures from Swedish infants seldom yielded more than one kind of enterobacteria; E. coli and Klebsiefla were most frequently isolated. E. coli dominated in both Pakistani groups, but especially caesarean section delivered infants were in addition often colonized with Proteus, Klebsiella, Enterobacter or Citrobacterspecies. Breastfeeding from the first day of life reduced colonization with Klebsiella/Enterobacter/Citrobucter.The results suggest that environmental exposure, delivery mode and early feeding habits all influence the early intestinal colonization with enterobacteria.

Bacterial attachment as a predictor of renal abnormalities in boys with urinary tract infection

The development of renal scarring was analyzed prospectively in 241 boys with their first known episode of symptomatic urinary tract infection (140 acute pyelonephritis, 61 acute cystitis, and 40 nonspecific). Of 197 boys undergoing urography, 22 (11%) had scars; 20 were in the pyelonephritis group. Vesicoureteral reflux occurred in 81% of those with scarring, compared with 20% of those without scarring. The bacteria causing the first episode of urinary tract infection in each patient were saved, and Escherichia coli organisms were characterized for the expression of both galactose-alpha (1----4)galactose-beta (Gal-Gal)-specific adhesins and pap homologous DNA. Scarring occurred in 41% and other renal abnormalities in 11% of boys infected with bacteria that did not bind Gal-Gal (Gal-Gal negative), compared with 5% and 1%, respectively, in those infected with Gal-Gal-binding strains (Gal-Gal-positive) (relative risk 8.3; 95% confidence limits 3.3 to 20.4; p less than 0.001). That boys infected with Gal-Gal-negative strains more often had reflux did not explain the increased risk for renal scarring in this group. The possibility that the phenotypically negative strains could be induced to express Gal-Gal adhesions in vivo was excluded by dot blot analysis, which showed the absence of pap homologous DNA in all but one of the Gal-Gal-negative strains. The results suggest that the absence of Gal-Gal-specific adhesins in E. coli can be used as an indicator of risk for renal scarring and the need for radiologic examination.

Enterobacter Species in a Pediatric Hospital: Horizontal Transfer or Selection in Individual Patients?

Enterobacter species were studied longitudinally in a childrens hospital. In total, 287 Enterobacter isolates were obtained from 171 children in 15 different wards (from March 1995 through April 1997). Strains were typed by random amplified polymorphic DNA and pulsed-field gel electrophoresis, which were concordant in outcome. In total, 97 DNA types and 199 colonization events were identified. A predominant clone was isolated 111 times from 62 children; another clone was isolated 19 times from 10 patients. These clones caused 36% of all colonizations. In 34% of the children, Enterobacter clones were found in 2-4 patients. The remaining colonizations were due to unique Enterobacter isolates. A large proportion of the Enterobacter strains was acquired through cross-transmission. This finding contrasts with the prevailing opinion that resistant Enterobacter strains are selected primarily from the patients own gut flora.

Bacterial adherence as a virulence factor in urinary tract infection

Escherichia coli (E. coli) causes >90% of urinary tract infections, UTI, in childhood. The capacity to adhere to urinary tract epithelial cells characterizes E. coli strains that cause acute pyelonephritis. Adherence of uropathogenic E. coli is the result of a specific interaction between bacterial adhesins and glycolipid receptors on the host cells, especially the globoseries of glycolipids which share the Galactose α 1 ‐>4Galactoseβ disaccharide (Galα 1 ‐>4Galβ). In childhood UTI, Galα 1 ‐>4Galβ‐binding bacteria caused significantly higher body temperature, C‐reactive protein (CRP), erythrocyte sedimentation rate (ESR), and pyuria, and lower renal concentrating capacity, than E. coli lacking this specificity. The Galα1‐>4Galβ‐binding bacteria thus appeared to be more potent inducers of inflammation than other strains. Since inflammation may lead to tissue damage we examined the relationship of infection with Galα1‐>4Galβ‐positive bacteria to renal scarring. The frequency of renal scarring was 5% in boys with Galα1‐>4Galβ‐positive and 40% in boys with Galα1‐>4Galβ‐negative E. coli. Bacterial binding to Galα1‐>4Galβ can be detected with a commercially available test reagent. This reagent can thus be used as an effective predictor of risk for renal scarring. Interleukin‐6 (IL‐6) is a pyrogen and inducer of the acute phase reactants. It was shown to be produced locally in the urinary tract, in response to UTI, and to spread systemically. Mucosal challenge with dead bacteria was sufficient to induce the IL‐6 response. Circulating IL‐6, and/or IL‐1 and tumor necrosis factor could explain the fever, as well as increased ESR and CRP found in association with acute symptomatic UTI.

Bacterial and host determinants of renal scarring

This review summarizes recent work examining the interaction between host and parasite in recurrent urinary tract infection (UTI) and renal scarring. Virulence in uropathogenic E. coli has been defined by the severity of acute disease. Isolates from patients with acute pyelonephritic strains differ from those causing asymptomatic bacteriuria by multiple traits which contribute to virulence, and which are coexpressed in a non‐random manner. The single marker most characteristic for the pyelonephritogenic clones is bacterial adherence to uroepithelial cells binding specifically to the disaccharide Galα 1–4 Galβ within the globoseries of glycolipids. The notion that the most severe consequence of acute pyelonephritis, i.e. renal scarring, was caused by the most virulent clones, was contradicted by comparison of pyelonephritic strains isolated from children with and without scarring. The virulent clones were significantly less frequent in patients with renal scarring (22%) than in patients with recurrent pyelonephritis not developing renal scars (62%). In view of the unexpected inverse association of bacterial virulence with renal scarring lack of Galα 1–4 Galβ binding capacity of E. coli strains was found to predict the risk for renal scarring among boys with first‐time acute pyelonephritis. Vesicoureteric reflux (VUR) is widely accepted as a host determinant of susceptibility to pyelonephritis and renal scarring. In our study the frequency of renal scarring was 57% among girls with VUR as compared to 8% of those without. The reflux alone did however, not explain the selection of bacteria of low virulence. Individuals prone to UTI and renal scarring were found to be a genetically selected subgroup of the general population. A correlation between P1 blood group phenotype and susceptibility to UTI and between blood group non‐secretor state and renal scarring was found. The mechanisms behind these relationships need to be defined. The bacterial and host parameters combined indicate that host parameters are essential for the tendency to develop renal scarring after acute pyelonephritis.

Ciclosporin-Dependent, nu-Independent, Mucosal Interleukin 6 Response to Gram-Negative Bacteria

Regulation of the mucosal inflammatory response to Gram‐negative bacteria was analysed. The interleukin 6 (IL‐6) secretion, influx of polymorphonuclear leucocytes into urine, and bacterial clearance from the kidneys were compared between Balb/c (nu/nu) and nu/± mice, with and without ciclosporin (CsA) treatment. There was no significant influence of the nu genotype on any of the host responses measured. CsA pretreatment significantly decreased II‐6 secretion in both nu/nu and nu/± mice, but did not affect bacterial clearance or the leucocyte response in any mouse strain tested. Tissue damage, in addition to bacterial infection, resulted in significantly higher levels of IL‐6 than bacterial infection alone. Tissue‐damaged mice were significantly less likely to clear the bacterial infection than their non‐damaged counterparts, but there was no significant difference in the leucocyte response. CsA pretreatment did not significantly reduce the levels of IL‐6 in the tissue‐damaged mice. These results demonstrate that the mucosal inflammatory response to Gram‐negative infection, including IL‐6 secretion, is nu‐independent, and that bacterial infection alone or in combination with tissue damage induce IL‐6 secretion by two different pathways.

Nosocomial Mycobacterium bovis–Bacille Calmette-Guérin Infections Due to Contamination of Chemotherapeutics: Case Finding and Route of Transmission

We studied nosocomial infections due to Mycobacterium bovis bacille Calmette-Guérin (BCG) Onco-TICE bacteria, transmitted by contamination of medication prepared in BCG Onco-TICE-contaminated hoods in the pharmacy, in 5 immunocompromised patients at 3 hospitals. The BCG strains cultured from the patients had the same DNA profile as the BCG Onco-TICE strain used for bladder instillation. To prevent these infections, a change from open to closed preparation was made; strictly separated preparation in time of BCG Onco-TICE instillation and chemotherapy was enforced, the biological safety cabinet was disinfected between preparations, and gloves were changed between preparations.