In 2007, from randomly selected farms in Belgium, we sampled 5 laying hens from each of 10 farms and 5 broiler chickens from each of 14 farms. One broiler farm was sampled twice (4 months apart, from different flocks in the same house, leaving 1 production round unsampled). Samples were taken from the cloaca and nasal cavity of these 50 laying hens and 75 broiler chickens.

Samples were first incubated in a brain–heart infusion broth supplemented with nalidixic acid and colistin, each at a concentration of 10 µg/mL. After overnight incubation at 37°C, 1 µL of this broth was streaked onto an MRSA ChromID plate (bioMérieux, Marcy l’Etoile, France) and incubated for 24–48 h at 37°C. To differentiate phenotypes of S. aureus, we purified colonies that showed typical growth on MRSA Chrom-ID plates by transferring them to modified Baird-Parker agar (7) and Columbia blood agar (Oxoid, Hampshire, UK) for DNAse and catalase testing.

The phenotypically identified MRSA strains were then confirmed by 16S rRNA-mecA-nuc triplex PCR as previously described (8). For all strains, the spa type was determined, and for 3 strains, multilocus sequence typing (MLST) was performed as described (9). Relatedness with other spa types from porcine origin (10) was determined by using Ridom SpaServer software version 1.3 (Ridom GmbH; Würzburg, Germany; www.ridom.de/spa-server).

Disk susceptibility of the strains was tested by using the Kirby-Bauer disk-diffusion method. Clinical Laboratory Standards Institute guidelines (M31-A3) were followed for inoculum standardization. After plates were incubated for 18 h, inhibition zones were measured in millimeters and interpreted according to Neo-Sensitabs manufacturer’s instructions (http://rosco.dk). S. aureus ATCC 25923 was included for internal quality control.

MRSA was not isolated from any laying hen samples. This finding may indicate that MRSA is absent or present only in low numbers in laying hens, possibly because of the limited use of antimicrobial drugs in these animals. Use of certain antimicrobial drugs in human hospitals has been shown to be a risk factor for acquiring MRSA infection, especially when the chosen treatment is inappropriate or insufficient (11). Antimicrobial-drug use may also be a risk factor for MRSA colonization of animals. The antimicrobial drugs used in the flocks included in this study were tylosin, amoxicillin, trimethoprim-sulfamethoxazole, lincomycin, tetracycline, and colistin.

MRSA was isolated from 8 broiler chickens from 2 of the 14 farms sampled. Low prevalence in poultry has also been found by Kitai et al. (6) and Lee (5), although they sampled chicken carcasses from slaughterhouses and did not find any livestock-associated strains. Given our relatively small sample size, our data did not permit us to estimate the within- and between-flock prevalence.

In the MRSA-positive flocks, the number of positive samples varied between 1/5 (20%) and 5/5 (100%). From the 1 MRSA-positive farm that was sampled twice, MRSA was isolated on both occasions. This finding indicates that MRSA may persist on a farm and colonize future flocks. MRSA was found in nearly equal numbers from the nares samples and the cloaca samples. Of the 8 MRSA-positive animals (16 samples), MRSA was found in all samples except for 1 cloacal swab, for a total of 15 MRSA isolations.

Susceptibility testing showed that all 15 isolated strains were resistant to erythromycin, kanamycin, tobramycin, lincomycin, tylosin, tetracycline, and trimethoprim. All strains were susceptible to chloramphenicol, ciprofloxacin, linezolid, mupirocin, quinopristin-dalfopristin, rifampin, and sulfonamides.

Molecular typing showed that the strains all belonged to spa type t1456 of the livestock-associated ST398, which is typically not typeable by pulsed-field gel electrophoresis. To our knowledge, this spa type has not been found in other animal species (11,12). Its relatedness to other spa types isolated from pigs is shown in the Table. A shortening of variable number tandem repeat composition seems to be present in spa type t1456. Whether t1456 is a clone typically associated with poultry, or specifically broiler chickens, and whether it is spreading internationally needs further investigation.

We confirmed the presence of MRSA in broiler chickens, but we were unable to find it in laying hens. All isolates belonged to 1 spa type, t1456, and thus differed from the other strains belonging to ST398 isolated from other animal species in Belgium and abroad. Whether this spa type is typically associated with poultry still needs to be confirmed. More detailed data are also needed to gain further insight in the true within- and between-flock prevalence of MRSA in poultry and its evolution over time.