The occurrence and the further spread of high-level glycopeptide-resistant, vanA-positive Enterococcus faecium strains outside of hospitals have been investigated. We could isolate such bacteria directly from thawing liquids of commercially produced frozen poultry (chickens, turkeys; no further data on previous feeding with avoparcin were available). In 5 of 13 samples of raw minced meat of pigs originating from 13 different butcher's shops, glycopeptide-resistant E. faecium (VanA type) could be detected after overnight broth cultivation of these samples. No glycopeptide-resistant enterococci could be isolated from meat samples of chickens that were fed without avoparcin. VanA type E. faecium strains were also identified in 12 fecal samples recovered from 100 nonhospitalized humans in the rural area of Saxony-Anhalt federal county. These results suggest a possible role of the food chain in the spread of glycopeptide-resistant E. faecium. Molecular typing (macrorestriction and multilocus enzyme analysis) reveal a wide dissemination of the vanA gene among strains of different ecological origins.

The use of the glycopeptide antibiotic avoparcin (AVO) as a feed additive in animal husbandry of many European countries led in 1994-1995 to frequent isolation of VanA-type vancomycin-resistant enterococci (VRE) from commercially produced animal foodstuffs as well as from fecal samples of nonhospitalized persons in Germany (Saxony-Anhalt state). However, at the end of 1997, a decreasing number of such VRE was detected in frozen and fresh poultry meat (chickens and turkeys) from German producers. At this point in time, AVO had been discontinued in animal husbandry for more than 2 and one-half years in Denmark/Norway, nearly 2 years in Germany, and about 8-9 months in all countries of the European Community and Switzerland, respectively. VRE were then only detected in very low concentrations in one-quarter of the poultry meat samples (eight of 31, originating from 18 distinct German producers and bought in 12 different supermarkets). A decline of VRE prevalence was also observed in the gut flora of healthy persons (VRE carriers) in the same region (Saxony-Anhalt state, Germany), having fallen from 12% (12/100) in 1994 when AVO was being used to 6% (6/100) in 1996 and 3% (13/400) in 1997 after it was discontinued. These results likely indicate the importance of antibiotic selective pressure by glycopeptides such as AVO for the presence of VRE in animal meat products from commercial animal husbandry. Additionally, it underlines the role of animal products for the spread of resistant bacteria and transferable resistance genes to humans in the community.

SUMMARYA summer outbreak of severe gastroenteritis followed by haemolytic uraemic syndrome (HUS) and thrombotic thrombocytopenic purpura in a nursery school and kindergarten is described. Sandwiches prepared with green butter made with contaminated parsley were the likely vehicle of infection. The parsley originated from an organic garden in which manure of pig origin was used instead of artificial fertilizers, Cornally identical VerotoxinogenicCitrobacter freundiiwere found as causative agents of HUS and gastroenteritis and were also detected on the parsley.

The semisynthetic streptogramin combination quinupristin/dalfopristin (Synercid) is a promising alternative for treatment of infections due to multiply resistant gram-positive bacteria including vancomycin-resistant Enterococcus faecium. Resistance is mediated by acetyltransferases SatA (VatD) or SatG (VatE). Recent papers have indicated a possible link between the use of the streptogramin virginiamycin S/M as a feed additive in commercial animal husbandry and a selection of quinupristin/dalfopristin-resistant E. faecium (QDRE). We screened manure samples from two different turkey farms and from six different pig farms (using virginiamycin), samples from a sewage water treatment plant, 24 broiler carcasses, 10 pork samples, and 200 stool samples of nonhospitalized humans for QDRE. Our strain culture collection of hospital E. faecium isolates from the last 2 years was also reviewed for QDRE. All manure and sewage samples were positive for QDRE, as well as 11 from broiler carcasses (46%), 1 from pork (10%), and 28 from human stool specimens (14%). Thirty-six hospital isolates of E. faecium exhibited resistance to quinupristin/dalfopristin. In 141 QDRE of different origin satA (vatD) and satG (vatE) genes were detected (seven isolates from humans with an unknown resistance mechanism). Streptogramin resistance determinants were tansferable in filtermating experiments for 5 of 10 satA (vatD) and 9 of 22 satG (vatE) isolates. Different EcoRI patterns of satG (vatE) plasmids and corresponding hybridizations of the satG (vatE) gene indicated nonhomologous resistance plasmids in isolates of different origin. The results of this study indicate a common gene pool for streptogramin resistance in E. faecium of different ecological origin. A selection of QDRE using the streptogramin virginiamycin S/M as a feed additive and a spread of the resistance via the food chain to humans is probable.

No abstract