Incidence of glycopeptide hetero-intermediate Staphylococcus aureus strains in Maltese hospitals


Corresponding author and reprint requests: M. A. Borg, Infection Control Unit, St Luke's Hospital, Guardamangia MSD 07, Malta


The incidence of hetero-intermediate glycopeptide susceptibility among Staphylococcus aureus isolates in Malta, a country with a high incidence of methicillin resistance, was studied by screening 454 non-repetitive S. aureus isolates on teicoplanin-supplemented agar plates, followed by Etests and genotypic studies. All strains were susceptible to vancomycin, but four (0.88%) exhibited teicoplanin MICs of > 12 mg/L. High methicillin-resistant S.aureus endemicity was not an accurate predictor of the emergence of non-susceptibility to glycopeptides.

Strains of Staphylococcus aureus with reduced susceptibility to the glycopeptides vancomycin and teicoplanin were first reported from Japan in 1997 [1]. Further reports from Japan have indicated that these strains of S. aureus, with vancomycin MICs of < 4 mg/L, may include resistant sub-population variants, occurring at a frequency of approximately 1 × 10−6[2]. The clinical significance of these hetero-resistant isolates remains a matter of considerable debate and it is not recommended to report them as glycopeptide-intermediate S. aureus (GISA) in clinical laboratory reports [3]. Nevertheless, strains of glycopeptide hetero-intermediate S. aureus (hGISA) have been associated with therapeutic failure [4]. It has also been postulated that hGISA strains may be precursors for the development of the GISA phenotype [2]. Therefore, their quantification may offer an insight into the possibility of emergence of GISA in a particular geographical location [5].

Malta is an island country in the Mediterranean with a population of 400 000, and is considered to have a high endemicity of methicillin-resistant S. aureus (MRSA), with > 45% of S. aureus isolates from blood cultures being reported as methicillin-resistant. No clinical cases of GISA have ever been reported. Consecutive non-repetitive S. aureus isolates (n = 454) were examined over a 9-month period between January and September 2003 at the Microbiology Laboratory of St Luke's Hospital, the only tertiary-care hospital on the island. Glycopeptide susceptibility testing followed the technical procedure recommended by the European Antimicrobial Resistance Surveillance System [6], based on the study of Walsh et al.[7] and the method of screening for GISA adopted by the French Microbiology Society [8]. Screening was undertaken by growing the bacteria overnight in brain–heart infusion broth (Oxoid, Basingstoke, UK); 10 µL of the stationary-phase culture was then inoculated on to a Mueller–Hinton agar (Oxoid) plate containing teicoplanin 5 mg/L. The plates were incubated for 48 h at 37 ± 0.5°C. Growth of two or more colonies was deemed to be a positive result, whereupon the colonies were analysed further by the Etest method.

The inoculum for Etests was prepared in brain–heart infusion broth (BBL, Cockeysville, MD, USA) to a density of 2× McFarland standards. Aliquots of 200 µL were spread evenly with a swab on two 90-mm brain–heart infusion agar (Oxoid) plates. The plates were then left to stand to allow the inoculum to soak into the plate, after which Etest strips for teicoplanin and vancomycin were applied according to the manufacturer's instructions (AB Biodisk, Solna, Sweden). The plates were read after 48 h at 37 ± 0.5°C, with particular attention paid to any ‘micro’ colonies or small colony variants growing in the Etest ellipse. If colonies were found growing at cut-off levels of ≥ 12 mg/L for the teicoplanin plate, or ≥ 8 mg/L for both the teicoplanin and vancomycin plates [7], the isolates were then analysed by pulsed-field gel electrophoresis typing and 16S−23S rRNA intergenic spacer polymorphism analysis [9,10].

Of the 454 isolates tested, 14 (3.1%) yielded an average of 24 CFU/plate (range: 2–100 CFU/plate) on the glycopeptide screening plate. Vancomycin MICs for these 14 isolates were 1.5–4 mg/L; however, only four (0.88%) isolates were defined as glycopeptide non-susceptible on the basis of a teicoplanin MIC of > 12 mg/L. Two isolates had been submitted from within St Luke's Hospital itself (renal and intensive care units), while the other two originated from a geriatric rehabilitation hospital and a district general hospital, respectively (Table 1). All four of these isolates were non-susceptible to methicillin and penicillin. This is not surprising, as resistance to glycopeptides in S. aureus is found predominantly among MRSA isolates [11]. The isolate designated Malta-3 was also resistant to cefuroxime, erythromycin and ciprofloxacin. The three other isolates had wider antimicrobial resistance profiles, including resistance to lincomycins and aminoglycosides.

Table 1.  Isolates from Malta with the glycopeptide hetero-intermediate resistance phenotype (hGISA)
SourceResistance profile (disk tests)PFGE profileEtest MIC (mg/L)
  1. PFGE, pulsed-field gel electrophoresis.

  2. Pn, penicillin (1-µg disk); Mt, methicillin (5 µg); Cx, cefuroxime (30 µg); Er, erythromycin (5 µg); Cp, ciprofloxacin (1 µg); Cl, clindamycin (2 µg); Km, kanamycin (30 µg); Tb, tobramycin (10 µg); No, neomycin (10 µg); Fd, fusidic acid (10 µg); Te, tetracycline (10 µg); St, streptomycin (10 µg); Gn, gentamicin (10 µg).

Dialysis; tertiary hospitalPnMtCxErClGnCpKmStNoTbMalta-1a316
District general hospitalPnMtCxErClGnKmStTbMalta-1b412
Geriatric hospitalPnMtCxErClCpKmNoTbMalta 2312
Intensive care unit; tertiary hospitalPnMtFdTeKmStNoMalta-3316

The four isolates were grouped into three clusters by pulsed-field gel electrophoresis. Malta-3 was a clonal variant (four bands difference) of the most prevalent community MRSA clone (MLST type 80) in Europe, and carried the PVL toxin gene associated with this clone (results not shown) [12]. The two variants (one band difference) of the Malta-1 clone were similar to one of the major international epidemic MRSA clonal complexes (CC5) by pulsed-field gel electrophoresis and 16S−23S rRNA intergenic spacer polymorphism analysis. The first Japanese GISA isolate and five of the USA GISA isolates also belonged to this clonal complex [13].

Malta is among the countries with the highest incidence of MRSA in Europe [6]. Studies in other Mediterranean countries with high MRSA levels have reported varying GISA and h-GISA rates, ranging from 65% in Spain [14] to 1.1% in Italy [11] and 0.6% in France [15]. The rate of 0.88% identified in Malta supports the hypothesis that the mere prevalence of methicillin resistance in a particular geographical area is not an accurate predictor of the likelihood of glycopeptide non-susceptibility.

The results of the present study differed from those reported in other Mediterranean countries, in that all teicoplanin hetero-intermediate strains isolated in Malta were susceptible to vancomycin. Exposure to glycopeptides is recognised as a risk factor for the emergence of resistance in S. aureus[16]. Teicoplanin is the most popular glycopeptide used in Malta, with annual defined daily doses more than five-fold greater than those for vancomycin. Further studies are needed to establish whether the predominant use of teicoplanin in a setting with a high MRSA prevalence and substantial glycopeptide consumption has any impact on reducing the emergence of vancomycin resistance in S. aureus.