Meanwhile, Back at the Farm . . .

Abstracts & Commentary

Synopsis: Antibiotic-resistant Enterococcus faecium and Salmonella spp are widespread in meat obtained from US supermarkets. The high prevalence of resistance is the likely consequence of the practice of adding antimicrobials to animal feed.

Sources: McDonald LC, et al. N Engl J Med. 2001;345:1155-1160; Sorensen TL, et al. N Engl J Med. 2001;345:1161-1166; White DG, et al. N Engl J Med. 2001;345:1147-1154.

Virginiamycin is a streptogramin antibiotic commonly used in animal feeds in the United States as a growth promoter. Because resistance to virginiamycin is associated with resistance to other streptogramins, widespread use of virginiamycin could lead to the development of quinupristin-dalfopristin resistance among the bacterial flora of farm animals. Resistance to the latter agent has been identified among turkeys fed virginiamycin.1

In order to assess potential human exposure to quinupristin-dalfopristin-resistant Enterococcus faecium, McDonald and colleagues undertook a multistate survey of chickens purchased in Georgia, Maryland, Minnesota, and Oregon. Using selective media, they sampled 407 chickens from 26 stores. In addition, they examined 334 human stool samples from specimens submitted to clinical laboratories for routine culture. A total of 58% of chicken carcasses yielded quinupristin-dalfopristin E faecium. Resistant isolates were identified in all 4 regions, ranging from a low of 17% in Minnesota to 87% in Oregon. Only 3 (1%) of the human stool specimens were positive for resistant E faecium.

In order to assess the ability of resistant E faecium to colonize the human intestinal tract, Sorensen and colleagues fed 107 CFU of 3 different strains of E faecium of animal origin to healthy volunteers in a randomized, double-blind study. One group received 2 strains resistant to glycopeptides, one group received a strain resistant to streptogramins, and a third group received a strain susceptible to both classes of agents. Stools were cultured on selective media on days 0 through 6, 14, and 35. All subjects receiving the test strains excreted resistant E faecium through day 6; stools were cleared of the test strains in all but 1 patient at day 14 and in all patients at day 35.

In the same issue of the New England Journal of Medicine, White and colleagues reported sampling of ground chicken, turkey, beef, and pork obtained from 3 supermarkets for Salmonella. Of 200 samples tested, 20% yielded Salmonella, representing 13 different serotypes. Eighty-four percent of isolates were resistant to at least 1 antimicrobial; 53% were resistant to 3 or more. Five isolates of Salmonella enterica serotype agona were resistant to 9 antimicrobials; 2 isolates of serotype typhimurium were resistant to 12 agents. Prevalence of resistance to selected agents is displayed in the table.

Resistance to Selected Antimicrobials of Salmonella Isolates from Ground Meat

Agent Resistant (n = 45)
Ampicillin 27%
Ceftriaxone 16%
Gentamicin 4%
TMP/SX 18%
Ciprofloxacin 0%

Comment by Robert Muder, MD

Antimicrobial resistance is widespread among bacterial isolates of agricultural animals, and the likely cause is the practice of adding antibiotics to animal feed in order to enhance growth. This is not a new phenomenon. For example, avoparcin, a glycopeptide similar in structure to vancomycin, was widely used as a growth promoter in Europe until 1997, when it was banned due to the widespread presence of vancomycin-resistant enterococci in farm animals.

The 3 articles presented here provide troubling evidence that resistant bacteria are developing as a result of agricultural practices and these bacteria are a potential threat to human health. In the United States, virginiamycin, a streptogramin antibiotic, is widely used in animal feeds. As a consequence, streptogramin resistance, and—in particular—quinupristin-dalfopristin resistance, appears to be widespread among E faecium strains isolated from poultry. Although streptogramin resistance among human stool isolates of E faecium is currently rare, the study of Sorensen et al demonstrates that ingestion of resistant E faecium in the amount likely to be encountered in food results in at least transient multiplication and carriage of the strain in the human intestine. Although carriage in healthy subjects lasted about a week, it should be noted that none of the volunteers had underlying conditions associated with prolonged carriage of resistant E faecium, such as coadministered antibiotics or abdominal surgery. Given the enormous reservoir of resistant enterococci that exists in the food supply, establishment of persistent colonization in susceptible humans seems likely.

The potential for widespread dissemination of quinupristin-dalfopristin resistance among human isolates of E faecium is of tremendous concern. Since many strains of nosocomially acquired E faecium are resistant to vancomycin, therapeutic options are limited. Spread of streptogramin resistance from animals into the human population would further diminish the number of potentially effective therapies.

White et al demonstrate that multiple antibiotic-resistant Salmonella are widely distributed in the food supply. Given that there are more than 1 million cases of salmonellosis occuring annually, and that at least 80% of these are food borne, this is a significant threat to public health. A number of strains are resistant to 3rd-generation cephalosporins; this is most likely due to the agricultural use of ceftiofur, a cephalosporin licensed for use in livestock. A plasmid-mediated beta-lactamase mediating resistance to both ceftiofur and ceftriaxone was identified among the Salmonella isolates studied. Ceftriaxone is considered by many to be the drug of choice in the treatment of salmonellosis in children; this agent may be "lost" if cephalosporin resistance increases among Salmonella isolates.

Each year in the United States, 25 million pounds of antimicrobials are given to animals for nontherapeutic purposes, primarily for growth promotion. In contrast, "only" 3 million pounds are given to humans.2 Current agricultural practices clearly promote antibiotic resistance among the bacterial flora of farm animals. Spread of antimicrobial resistance from animals to humans can occur either through direct spread of resistant strains, or by transfer of resistance determinants from animal strains to human strains. One wonders about the ultimate effectiveness of our attempts to control the spread of resistant bacteria within hospitals when there appears to be an enormous reservoir of resistance in the food supply. Clearly, the nontherapeutic use of antimicrobials in agriculture must be curtailed.


1. Welton LA, et al. Antimicrob Agents Chemother. 1998; 42:705-708.

2. Gorbach SL. N Engl J Med. 2001;345:1202-1203.

Dr. Muder, Hospital Epidemiologist, Pittsburgh VA Medical Center, is Associate Editor of Infectious Disease Alert.