Overall, mupirocin resistance seems to have increased since October 2019. It remains at low prevalence (below 2%) in the LHUB-ULB population, whereas it reaches 5.6% in AML. Focusing on the paediatric population, a worrying trend is observed for the last 5 years: the prevalence of mupirocin resistance reached more than 5% for the period October 2022-September 2023 and was almost systematically associated with resistance to fusidic acid (64/70, 91.4%). Focusing on children during Q3, the impetigo season [9], revealed an even more alarming trend, with mupirocin resistance rates ranging from 8.9–11.9% and an almost systematic association with fusidic acid resistance (26/28, 92.9%). Based on analysis of the isolates received by the NRC since October 2017, 78.4% of the mupirocin and fusidic acid coresistant isolates were confirmed to be related to EEFIC. Whole-genome sequencing analysis of isolates collected during the last 2 years support this trend, with 29/33 clonal isolates (87.9%) sharing the following genetic characteristics: ST121, resistance to fusidic acid conferred by fusB, resistance to mupirocin conferred by mupA, and the presence of the virulence genes eta and etb. In a previous study on subpopulations of S. aureus CC121, Kurt et al. showed that isolates sampled from superficial infections (including staphylococcal scalded skin syndrome, bullous impetigo, exfoliative dermatitis, and conjunctivitis) clustered in the same clade, including EEFIC [15]. Interestingly, wgMLST cluster analysis showed that these strains are closely related to the reference genome of an MSSA isolated in Greece, which shares the same characteristics, while the other reference genomes from SSTI-related isolates are more distantly related. However, in contrast to a previous study in Greece describing the emergence of a similar ST121 clone [7], none of them carried the lukS/lukF genes. Notably, 1/33 of the strains is ST15 with a t084 spa type and carrying etb. ST15 strains carrying exfoliatin genes have been observed quite frequently in staphylococcal scalded skin syndrome in France [16].
Treatment of impetigo is often based on the use of topical antibiotics for localised forms, such as fusidic acid cream in Belgium and the Netherlands [5, 6] or mupirocin in France [8]. This option is preferred for localised disease because it has been shown to be at least as effective as oral treatment, with fewer side effects [17, 18]. Therefore, the emergence and spread of an MSSA clone coresistant to these two antibiotics is a cause for concern. Other possible topical treatments are retapamulin [17, 19] and ozenoxacin [20]. However, in the European Union, marketing authorisation for retapamulin was withdrawn by the European Commission on 25 February 2017 at the request of the manufacturer. At the time of manuscript writing, ozenoxacin was only available in Italy, Portugal and Spain [21]. Interestingly, the emerging clone described in the present study frequently carried aadD and cat(pC194), which confer resistance to amikacin and chloramphenicol, respectively. Although these antibiotics are not used in Belgium, they may be used as topical treatments. Acquisition of these resistance genes is likely a selective advantage for strains with a propensity for superficial skin infections. On the other hand, the efficacy of disinfection as sole treatment for impetigo seems to be inferior to that of topical antibiotics [17], even though recommendations in the United Kingdom favour the single use of antiseptics [22]. However, the development of resistance to topical antibiotics may change this consideration, and further studies are needed to determine the best interventions for localised impetigo.
Impetigo is known to affect children in summer [9]. Indeed, in a previous surveillance focused on S. aureus-related SSTIs, we showed that EEFIC was more prevalent in children in late summer [4]. At that time, resistance to mupirocin was anecdotal. However, one of the limitations of this surveillance was that it was conducted during the summer immediately after onset of the COVID-19 crisis. The social distancing and restrictions that gradually came into force during this period, combined with the workload of the clinical laboratories during the study period, certainly introduced biases. The end of the COVID-19-related restrictions and the increase in mupirocin-resistant strains arriving at the NRC led us to perform the present study using immediately available retrospective laboratory data. As strains are sent to the NRC by clinical laboratories on a voluntary basis, NRC data may not always reflect the true epidemiological situation. To determine the accuracy of the emergence of a new clone, we collected epidemiological data from two large laboratories in Belgium (LHUB-ULB and AML). In a previous study focusing on respiratory viruses, LHUB-ULB was demonstrated to act as a sensor for the whole country due to the large number of analyses and its central location within Belgium [23]. Inclusion of AML allowed the addition of data from a different type of laboratory (LHUB-ULB is more of a hospital laboratory, whereas AML, located in Flanders, mainly collects samples directly from general practitioners and outpatient sampling centres). However, additional data from a laboratory in the third region of Belgium (Wallonia) could have made the data completely exhaustive. Setting up an actual prospective surveillance would have been more accurate but would have caused a delay in communicating these results. Nevertheless, such a strategy has limitations: there are no data about the representativeness of AML and LHUB-ULB in Belgium for impetigo. Furthermore, the proportion of skin lesions that are sampled before treatment is unknown, and the total number of skin samples sent to laboratories per year is not known, as these data were not collected by health insurance providers. This bias is limited by the overall number of strains included (32,503 MSSA over 10 years) but may lead to some imprecision when focusing on a specific population during a specific time period, as we did in this study for children during the third quarter of the year. The epidemiological differences observed between AML and LHUB-ULB could be due to regional differences and/or differences between the patients sampled (case mix). The impact of the latter is likely mitigated by the fact that we specifically selected samples from outpatients. Nevertheless, AML collects samples from the federal agency for reception of asylum seekers (FEDASIL), and identification of epidemiological clusters at the level of the clinical laboratory is not easy because the context is rarely provided by the prescribers for outpatients. Regional epidemiological differences may exist. For example, in the most recent surveillance study on the EEFIC in Belgium [4], all EEFICs were from Flanders, and almost half of them (47.4%) were located in the province of Antwerp, where AML is located. From October 2013 to September 2023, the majority of MSSA strains coresistant to mupirocin and fusidic acid analysed by the NRC in this study were from Flanders, followed by Brussels and Wallonia. Beyond an actual regional difference, it is also possible that laboratories and practitioners in Flanders are more aware of the current epidemiological situation due to the EEFIC clusters in 2018 that led to our previous surveillance; in contrast, these clusters were not described in Wallonia.
Although very common, impetigo is likely an understudied pathology [24]. Indeed, there is currently no organised surveillance of impetigo in Belgium. As a consequence, epidemiological data rely on laboratory data, as in this study. At the laboratory level, clinical information on the type of infection is often lacking. We attempted to overcome these problems in our study by including skin samples from outpatients only and by focusing on the more likely target population for impetigo: children sampled in late summer (Q3). Additionally, uncomplicated impetigo may not be sampled. This may lead to bias towards complicated skin infections that should not be treated with topical antibiotics. Nevertheless, this study clearly demonstrated the emergence of a clone combining virulence and resistance. Periodic surveillance is needed to better monitor the true epidemiology of impetigo and adapt treatment recommendations accordingly. In addition, this surveillance may provide useful information for developing intervention trials for impetigo patients.
The present study highlights the emergence and spread of a clonal ST121 mupirocin-resistant clone diverging from EEFIC (M-EEFIC), which has acquired the mupA gene in addition to its virulence (eta and etb) and resistance (fusB) genes. This clone could have a selective advantage in Belgium and similar European countries where there is no other alternative topical treatment. The emergence of this new clone, combined with the overall high prevalence of fusidic acid resistance, should lead to reconsideration of the recommendations for first-line treatment of impetigo in Belgium and introduction of active surveillance of impetigo to better assess the situation.