Such a combination of repeat duplication followed by point mutation has been described by Kahl et al. (2005). In another case (patient 12), the observed change could be either this website duplication or a deletion (t844 and t953, eight and nine repeats; Table 2). Eight samples were collected from this patient at the same time, six of them containing MRSA t844 and two of them MRSA t953. We cannot determine whether t844

or t953 is the ancestor spa type, i.e. if there was a duplication or deletion of a repeat 12. However, at some point a duplication of repeat 12 must have taken place, making it more likely that t844 is the ancestor. The same argument can be applied to t3677 and t4225 (patient 13) with one or two copies of repeat CTLA-4 inhibiton 31. The

spa type t024 was by far the most common spa type found. This spa type was found in 796 isolates from 213 patients. In addition, six patients had only the Variant spa type t430 that was the most common t024 variant. It was first found on a t024-positive resident of a nursing home and subsequently has only been found on residents from that nursing home. In seven patients, the Ancestor and Variant spa types were found simultaneously in the first MRSA-positive sampling (patients 2, 4, 5, 7, 8, 9 and 12; Table 3). No follow-up data were recorded for patients 2, 9 and 12. In patients 4, 5 and 7, only the Ancestor was found in the latest sample, and one patient (patient 8) had both spa types at both sample times. In five cases (patients 1, 3, 10, 11 and 13) the Variant was recovered between 1 and 20 months after the Ancestor. At the latest sampling, the Ancestor alone or both spa types were recovered in one case each (patient 1 and 3) while the Variant alone was found in three patients (patients 10, 11 and 13) (Table 3). In one

case only, tuclazepam the suggested Variant was the first to be found (patient 6). At the second sampling from this patient, both Ancestor and Variant were recovered (patient 6). We found that the spa repeat region was very stable. Of 319 individuals with more than one occasion of MRSA, only 13 (4%) exhibited spa type changes. Of 1536 isolates, only 30 (2%) had changes in the spa repeat region. In a Swiss study, 10% of healthy carriers exhibited spa type mutations within 1 year (Sakwinska et al., 2010). This number is higher than ours but still in the same range of order. This is in contrast to the study by Kahl et al. (2005) of cystic fibrosis patients with persistent S. aureus airway infections. In that study, spa type alterations were found in MSSA isolates from five of 10 patients (50%) and in 14 of 142 isolates (10%). The mutation rates probably reflect differences in selection pressure. The isolates in our study are from infections as well as long-term carriers. The selection pressures under which the S. aureus existed were probably higher for the cystic fibrosis patients (Kahl et al., 2005).