Antibiotic resistance among enterococci causing endocarditis in the UK: analysis of isolates referred to a reference laboratory

BMJ 1998; 317 doi: (Published 05 September 1998) Cite this as: BMJ 1998;317:629
  1. Alan P Johnson (ajohnson{at}, clinical scientist,
  2. Marina Warner, biomedical scientist,
  3. Neil Woodford, clinical scientist,
  4. David C E Speller, consultant microbiologist,
  5. David M Livermore, clinical scientist.
  1. Antibiotic Reference Unit, Central Public Health Laboratory, London NW9 5HT
  1. Correspondence to: Dr Alan P Johnson

    Enterococci account for 5-15% of cases of bacterial endocarditis.1 They are the most resistant bacteria commonly encountered in this type of infection, which is still associated with a mortality of 20-30%.2 The treatment regimen for enterococcal endocarditis recommended by the British Society for Antimicrobial Chemotherapy and the American Heart Association is a synergistic bactericidal combination of a penicillin or glycopeptide with an aminoglycoside, usually gentamicin or streptomycin. 3 4 However, enterococci can acquire high level resistance to aminoglycosides, which abolishes this synergy.1 Enterococci can also exhibit high level resistance to penicillin or to glycopeptides.1 Our laboratory undertakes testing of bacteria from cases of endocarditis as a routine service; we analysed resistance among isolates from 120 cases of enterococcal endocarditis, received over 27 months.

    Methods and results

    Results of tests for antibiotic susceptibility were analysed for enterococci referred from patients with a clinical diagnosis of endocarditis between January 1995 and March 1998. Isolates exhibiting high level resistance to gentamicin or streptomycin were defined as those where the concentration of antibiotic required to inhibit growth on laboratory media (minimum inhibitory concentration) exceeded 2000 mg/l.1 Resistance to other antibiotics was defined according to criteria specified by the British Society for Antimicrobial Chemotherapy.5

    The isolates, which were from 60 UK hospitals, comprised 106 Enterococcus faecalis, 13 E faecium, and one E avium. The table shows the major resistance characteristics of these isolates. Overall, 26% of isolates had high level resistance to both gentamicin and streptomycin (22% of E faecalis isolates; 62% of E faecium isolates). A further 28 E faecalis isolates showed high level resistance to either gentamicin (7 isolates) or streptomycin (21 isolates); four E faecium isolates and the sole E avium isolate showed high level resistance to streptomycin but not to gentamicin.

    Resistance of enterococci from 120 cases of endocarditis to aminoglycosides and to cell wall active antibiotics

    View this table:

    All the E faecalis isolates remained susceptible to ampicillin (minimum inhibitory concentration 0.5-4 mg/l for 105 isolates and 8 mg/l for 1 isolate), but 6 were resistant to vancomycin, with 2 exhibiting cross resistance to teicoplanin. One isolate resistant to glycopeptides also had high level resistance to both gentamicin and streptomycin. The 13 E faecium isolates were all resistant to ampicillin (>8 mg/l), with 3 also resistant to vancomycin but not teicoplanin. Two of these vancomycin resistant isolates showed high level resistance to gentamicin and streptomycin.


    Treatment for enterococcal endocarditis comprises a bactericidal synergistic combination of a penicillin (usually ampicillin or benzylpenicillin) or glycopeptide with an aminoglycoside, usually gentamicin or streptomycin for at least 4 weeks. 2 3 It is therefore disturbing that 26% of the enterococci in this study were highly resistant to both gentamicin and streptomycin. Aminoglycosides have no therapeutic benefit in infections involving such strains, and unnecessarily expose patients to possible ototoxic or nephrotoxic side effects. A further 28% of the isolates were highly resistant to either gentamicin or streptomycin, emphasising the value of testing both of these compounds in determining appropriate treatment.

    Current guidelines recommend that endocarditis caused by enterococci with high level resistance to aminoglycosides should be treated with high dose amoxycillin or ampicillin for 6-12 weeks. 2 3 However, 11% of the enterococci were E faecium, which typically is resistant to ampicillin. Moreover, amoxycillin or ampicillin would be unsuitable for patients allergic to penicillin. This latter constraint applies to other proposed regimens that combine ampicillin with imipenem or ciprofloxacin. Although glycopeptides may be considered in place of penicillin, the finding of glycopeptide resistance in several isolates, including three of the E faecium isolates, means that their efficacy cannot be guaranteed.

    The picture revealed is disturbing, with frequent resistance to the recommended synergistic combinations. Evaluation in endocarditis of unconventional regimens—for example, ampicillin plus carbapenems, ampicillin plus ciprofloxacin, or ciprofloxacin plus co-trimoxazole—is desirable, although the use of such broad spectrum agents may risk selecting resistance in the body microflora. Also desirable is early evaluation, in endocarditis, of novel narrow spectrum anti-Gram positive agents, such as streptogramins, oxazolidinones, and everninomycins.


    We thank Dr Ty Pitt and his colleagues in the Epidemiological Typing Unit at the Laboratory of Hospital Infection in the Central Public Health Laboratory for confirming the species of the isolates studied.

    Contributors: APJ and DML jointly conceived the idea for the project. APJ analysed the data and drafted and edited the paper. MW performed the antibiotic susceptibility tests and contributed to the writing. DCES, NW, and DML contributed to the discussions on interpretation of the data and writing of the paper.

    Funding: Public Health Laboratory Service.

    Competing financial interest: APJ and DML have received financial support from Rh∘ne-Poulenc Rorer for work on novel streptogramins and for attending conferences.


    View Abstract