Antibiotic susceptibility of Staphylococcus aureus isolated from skin lesions in children. A retrospective analysis from a tertiary care Italian pediatric hospital

Antibiotic susceptibility of S. aureus was retrospectively assessed in 1833 strains isolated from skin lesions observed in an Italian tertiary care hospital. Methicillin resistance was more frequent in outpatients than in inpatients (18% vs. 14%, p = 0.04) as well as resistance to cotrimoxazole (8% vs. 4.1%, p < 0.001). Resistance to ampicillin was 99% in both groups, while for clindamycin it was 11% and 14%, respectively. Among topical antibiotics fusidic acid showed the better resistance profile (3%). Antibiotic resistance in pediatric skin infection in outpatients could represent a therapeutic problem in Italy.


Introduction
Staphylococcus aureus is the most common cause of skin infections with an increasing prevalence of antibiotic resistance and consequent restriction of treatment options. 1 To our knowledge, data about resistance profile of S. aureus in skin lesions in children are overall poor, mostly in Italy.
We evaluated antibiotic susceptibility of S. aureus isolated from cutaneous lesions in patients followed up at a tertiary care Italian pediatric hospital.

Materials and methods
The IRCCS Istituto Giannina Gaslini -Genoa is a tertiary care children's hospital in Northern Italy serving as local pediatric hospital for the Genoa area but representing also a referral hospital nationwide and for many foreign countries. All S. aureus strains isolated from skin lesions in patients <18 years from January 2013 to June 2019 were extracted from the Laboratory of Microbiology database. In case of multiple isolations of the same strain in the same patient only the first isolate was considered. Subsequently data were anonymized, according to Gaslini Institute data protection policy based on European Union Data Protection Rules. The need for informed consent was waived given the retrospective nature of the study and the anonymization of the data.
S. aureus identification was confirmed by MALDI-TOF technology (Vitek MS, BioMerieux, France), according to the manufacturer's instructions. For each strain the susceptibility to the following antibiotics was recorded: ampicillin, methicillin, ciprofloxacin, clindamycin, cotrimoxazole, fusidic acid, mupirocin, gentamicin. The choice was made starting from the availability of results from automated tests (Vitek MS, BioMerieux, France) used routinely in our Insitute for pathogens isolated from skin lesions, and considering methicillin resistance as a standard, ampicillin as a surrogate for the presence of penicillinase and ciprofloxacin, clindamycin and cotrimoxazole as alternative drugs with an available oral formulation. Fusidic acid, mupirocin, gentamicin were also included since in Italy they have topical formulations and are frequently sold as out of the counter drugs.
The interpretation of the results was based on The European Committee on Antimicrobial Susceptibility Testing (EUCAST) breakpoint criteria. 2 Since the study was based on a laboratory database no information was available on demographic and clinical data; only ward of admission was available, that was used to classify patients as outpatients if sampled in Emergency Department, Dermatology Unit and Infectious Diseases Unit, considering the hospital protocols for hospital admission. Samples from all other wards were considered as from inpatients.

Statistical analysis
Data were reported as absolute numbers and proportions with 95% Confidence Interval (95% CI). Differences among proportions were evaluated by chi-square test with Yates correction or Fisher's exact test, when appropriate. A p value 0.05 was considered as statistically significant. Analysis was performed by means of the web platform Social Science Statistics. https://www.socscistatistics.com/.

Results
During the study period a total of 1833 S. aureus strains were extracted, 430 (23.4%) classified as from outpatients. Table 1 reports on antibiotic susceptibility among isolated strains stratified in outpatients or inpatients. The proportion of methicillin-resistant S. aureus (MRSA) was significantly higher in outpatients (18% vs. 14%, p ¼ 0.04, chi-square test). Ampicillin resistance was 99% in both strata and also resistance to clindamycin or ciprofloxacin was similar in the two groups. On the contrary, resistance to cotrimoxazole was significantly higher in outpatients (p < 0.001, chi-square test).
As for drugs with topical formulations, resistance to gentamycin and fusidic acid was similar in inpatients and outpatients, while for inpatients resistance to mupirocin was significantly higher (p ¼ 0.0309, chi-square test), even if many strains were not tested for susceptibility (70% in both groups) to this drug.

Discussion
In this study we describe the profile of antimicrobial susceptibility of S. aureus strains isolated from skin lesions in a pediatric tertiary care center in Italy. To our knowledge, no other Italian data on this topic are available.
The first observation we can drive is that even if the prevalence of MRSA we observed was lower than in other pediatric studies from other countries, 3-5 the proportion of this pathogen on skin lesions was significantly higher in outpatients compared with inpatients. However, since 95% CI of the proportions are imbricated it is probable that this difference is poorly clinically relevant but represents mainly a statistical effect. 6 Since no other pediatric Italian data are available we cannot make any national comparison, but it must be noted that European data (including also adults) report a MRSA prevalence of 15.1%, with a North (0%) to South (29%) increasing gradient, 7 that is consistent with our observations.
A recent Italian consensus on the treatment of skin infections in pediatrics 8 indicated clindamycin for first line treatment in communities where proportions of MRSA is >10% if clindamycin resistance is <10%. Considering the epidemiology of clindamycin resistant strains (11% in outpatients and 14% in inpatients), in our setting these criteria are not met and therefore clindamycin could not represent a possible choice for empirical therapy before susceptibility tests availability. We do not know if this observation can be generalized to other Italian regions, but this observation should represent a warning. In spite of the significantly higher proportion of strains resistant to cotrimoxazole in outpatients (8% vs. 4.1%), as already described, 9 in our study resistance to cotrimoxazole was lower than the 10% threshold proposed for other drugs to be used for first line therapy of skin lesions 8 and therefore it could represent an acceptable first choice, pending susceptibility tests.
The use of topical antibiotics has been associated with selection of resistant strains. 10 In our study data on resistance to mupirocin could be poorly reliable because of the high proportion (70%) of untested strains. Fusidic acid showed a very low proportion of resistance (about 3%) but also in this case near 1/3 of strains were not tested. Finally, topical gentamycin is a highly used over the counter topical antibiotic in Italy, and indeed the resistance we observed was not negligible (9-12%). Our data suggest that fusidic acid could be the better topical choice, at least in our epidemiological condition.
This study suffers also from important limitations, mainly related to the fact that it is derived from a laboratory database. Indeed, we do not have data on demographics and clinical conditions of the patients at time of sampling. Moreover, we do not know any previous antibiotic treatments or hospitalizations, facts that may have contributed to the development of antibiotic resistance. Despite these important shortcomings, the study gives information about an important item such as antimicrobial resistance in pediatric staphylococcal skin infections in a geographic area where epidemiological data are lacking.

Compliance with ethical standards
No funding was received for development of this manuscript.