Search

This study investigated the molecular epidemiology of a clonal outbreak of multidrug-resistant Acinetobacter baumannii that occurred between June 2003 and June 2004 in a tertiary-care hospital in Naples, Italy. A. baumannii was isolated from 74 patients, of whom 38 were infected and 36 were colonised. Thirty-three patients had ventilator-associated pneumonia, three had hospital-acquired pneumonia, and two had sepsis. Genotypic analysis of 45 available A. baumannii isolates revealed two distinct pulsed-field gel electrophoresis (PFGE) patterns. Of these, PFGE pattern 1 was represented by isolates from 44 patients and was identical to that of an epidemic A. baumannii clone isolated in another hospital of Naples during 2002. All A. baumannii isolates of PFGE type 1 showed identical multiresistant antibiotypes, characterised by resistance to all antimicrobial agents tested, including carbapenems, with the exception of colistin. In these isolates, inhibition of OXA enzymes by 200 mM NaCl reduced the imipenem MIC by up to four-fold. Molecular analysis of antimicrobial resistance genes showed that all A. baumannii isolates of PFGE type 1 harboured a class 1 integron containing the aacA4, orfX and bla(OXA-20) gene cassettes, an ampC gene and a bla(OXA-51)-like allele. Moreover, a bla(OXA-58)-like gene surrounded by the regulatory elements ISAba2 and ISAba3 was identified in a 30-kb plasmid from A. baumannii isolates of PFGE type 1, but not PFGE type 2. Thus, selection of a single A. baumannii clone producing an OXA-58-type carbapenem-hydrolysing oxacillinase was responsible for the increase in the number of A. baumannii infections that occurred in this hospital.

Streptococcus bovis is being recognised increasingly as a cause of infective endocarditis, and has also been associated with underlying gastrointestinal malignancy. This study evaluated the molecular epidemiology of S. bovis isolates responsible for endocarditis or bacteraemia in Italian patients between January 1990 and August 2003. S. bovis isolates were classified on the basis of their biochemical profiles, antimicrobial susceptibilities and genotypes. Of 25 isolates studied, 20 were S. bovis I and five were S. bovis II. Seven biochemical profiles were identified. Pulsed-field gel electrophoresis (PFGE) analysis identified 22 profiles that differed by at least two DNA fragments and showed a similarity of < 87%. Most PFGE patterns represented single isolates that differed in antimicrobial susceptibility, but three PFGE types were observed, with identical profiles and antibiotypes, in isolates from two different patients. S. bovis I and II isolates grouped into two distinct genetic clusters (I and II) with a similarity coefficient of 38%. Two sub-clusters (Ia and Ib), with a similarity coefficient of 47%, included 17 S. bovis I isolates with similar biochemical profiles (15 with biotype A, and two with biotype B), but different resistance phenotypes. Based on the phenotypic and genotypic heterogeneity of the isolates, it is postulated that the increase in S. bovis endocarditis in this geographical area might have been caused by the selection of sporadic endemic clones from the endogenous intestinal flora.

The effects of amphotericin B (AmB) on the hepatic excretory function and the colloid clearance capacity were investigated in the perfused rat liver. AmB at 5 or 10 microM caused dose-dependent reductions in bile and perfusate flow rates and in biliary bile acid (BA) excretion. BA concentration in bile tended to increase, due to a prominent reduction in bile water induced by the drug. At 5 microM, AmB also caused an increase in [14C]sucrose clearance by the liver and a release of hepatocytic enzymes into the perfusate. These alterations were not related to the decrease in the perfusate flow induced by AmB. In addition, the drug, at 5 microM, caused a significant decrease in the colloidal carbon clearance by the liver. In this case also, the effect was independent of the reduction in the perfusate flow induced by the drug. The toxic effects of AmB on the rat liver could be interpreted as a derangement of the cell membrane functional integrity, which causes cholestasis, enzyme leakage and an impairment of the reticuloendothelial system function. This latter effect deserves careful evaluation of its clinical implications.

The possibility that endotoxin pretreatment could prevent the hepatotoxic effects of erythromycin estolate (EE) was investigated using the isolated perfused rat liver. The addition of E. coli endotoxin (25 micrograms/ml) to the perfusate, 30 min prior to EE administration at 150 or 200 microM, significantly ameliorated the decreases in bile and perfusate flow caused by either concentrations of the drug in control liver preparations. This phenomenon was also studied using liver isolated from rats pretreated in vivo with endotoxin for three days. In these preparations, EE at both concentrations did not alter bile flow and caused reductions of perfusate flow which were far less than those observed in untreated control livers. Furthermore, in livers from endotoxin-treated rats EE induced less reduction of bile acid excretion and, at 150 microM, it did not increase the bile to perfusate ratio of sucrose seen in control preparations after the drug, which may be an expression of altered hepatocytic membrane permeability. Since it is known that both endotoxin and EE interact with membranes, it is suggested that the "protective" effects of endotoxin may occur at the membrane level.