| Abstract |
Introduction: Staphylococcus aureus infections cause significant morbidity and mortality in children and adolescents. There is limited data on the characteristics of S. aureus infections requiring hospitalization in childhood. Aim of this study was to investigate the molecular epidemiology and antibiotic resistance of S. aureus clinical isolates from children and adolescents.
Methodology: This cohort study included all S. aureus isolates recovered from children and adolescents younger than 18 years who were admitted to the University Hospital of Heraklion, Crete from 1 January 2015 to 31 December 2018. Medical records were reviewed for demographic and clinical characteristics. Infections were classified into (I) skin and soft tissue infections (SSTIs) such as impetigo, cellulitis, abscess, omphalitis, (II) toxin-mediated disease (TMD) including staphylococcal scalded skin syndrome and toxic shock syndrome and (III) invasive disease (INV) including septicaemia, pneumonia, osteomyelitis, and septic arthritis. Infections with positive cultures obtained from normally sterile fluids, including CSF, blood and synovial fluid were considered invasive infections. Infections and the recovered isolates were classified as community-associated (CA), community-onset healthcare-associated (COHA) and hospital-associated (HA) based on established CDC criteria. To examine possible changes over time, we studied subsets of isolates collected during two periods: 2015–2016 (period A) and 2017–2018 (period B).
The isolates were identified by conventional methods, including colony morphology, Gram staining, standard biochemical tests and by the automated Vitek 2 system (bioMérieux, Marcy l’ Etoile, France). Antimicrobial susceptibilities were interpreted according to the 2018 Clinical and Laboratory Standards Institute (CLSI) criteria. S. aureus ATCC 25923 and S. aureus ATCC 43300 were used as control strains. Isolates were phenotypically classified as methicillin susceptible S. aureus (MSSA) or MRSA based on the cefoxitin disc diffusion test and the latex agglutination test for the detection of PBP2a (bioMerieux). Inducible resistance to clindamycin was tested by d-test as per CLSI guidelines. S. aureus isolates that were resistant to cefoxitin were phenotypically classified as MRSA and verified by PCR for mecA or mecC gene carriage. Genes encoding Panton–Valentine leukocidin (lukS/lukF-PV), toxic shock syndrome toxin-I (tst), exfoliative toxins (eta, etb), fibronectin binding proteins A and B (fnbA, fnbB) were investigated by polymerase chain reactions with specific primers. Isolates were classified into main pulsotypes and subtypes by PFGE of chromosomal DNA SmaI digests performed in a CHEF DR III apparatus (Bio-Rad, Richmond, CA, USA). A dendrogram comparing molecular weights of strains’ DNA fragments was performed by FPQuest software version 4.5. (Bio-Rad Laboratories). Representative isolates of the main PFGE types, including all MRSA and selected MSSA isolates, were further characterized by MLST to sequence types (STs), based on the online MLST database (https://pubmlst.org/organisms/ staphylococcus-aureus). Statistical analysis for categorical variables, chi-square or two-tailed Fisher’s exact test were conducted to calculate p values and odds ratios. For continuous variables, statistical significance of observed associations was evaluated using Mann–Whitney U-test. Antibiotic susceptibility trends and MLST trends were analysed using chi-square for trend. The level of significance was set at p value less than 0.05. Statistical analysis was conducted using GraphPad Prism 9.1.0.
Results: During the 4-year study period, 139 non-duplicated S. aureus isolates were collected from children and adolescents (77 boys, 62 girls) with invasive and non-invasive infections. Sixteen (11.5%) isolates were MRSA and 123 (88.5%) were MSSA. The mean age of patients was 3.8 years (median age; 2.02 years) and it did not differ between children with MSSA and those with MRSA. S. aureus isolates were derived most frequently from skin and soft tissue infections (SSTIs; 94/139, 67.6%), followed by toxin-mediated disease (TMD; 23/139,16.5%, all 23 cases were SSSS) and invasive infections (INV; 22/139,15.8%). Pneumonia (9/22), bacteraemia (4/22) and osteoarticular infections (6/22) were the most common invasive S. aureus infections. MRSA isolates did not differ with regards to gender, length of stay, infection type, PICU admission as compared to MSSA and there was no associated mortality. All isolates recovered from children with TMD were MSSA. Most isolates were classified as CA (102/139; 73.4%), while HA and COHA were less common (20/139;14.4% and 17/139; 12.2%, respectively). A total of 52 isolates belonged to children with comorbidities or prolonged hospital-stay, including neonates (38 isolates), children with cystic fibrosis (seven isolates), children admitted to PICU (six isolates) and one child with malignancy. HA and COHA-isolates were more commonly isolated from these children (OR 2.8, 95%CI 1.27–5.78, p 0.01) and were more commonly associated with invasive disease (OR 4.41, 95%CI 1.71–10.79, p 0.001). Of 139 S. aureus isolates, 93.5% were resistant to penicillin, 35.9% to fusidic acid, 7.2% to mupirocin, 22.3% to tobramycin, 18.7% to erythromycin, 16.5% to clindamycin and 9.3% to tetracycline. Among 26 erythromycin-resistant isolates, prevalence of constitutive (cMLSB), inducible (iMLSB) and MS resistance phenotypes was 53.9, 34.6 and 11.5%, respectively. Resistance to clindamycin and tetracycline was higher among MRSA isolates (p 0.02 and p 0.001, respectively). MRSA and MSSA isolates did not differ regarding cMLSB and iMLSB resistance phenotype (OR 0.20, 95%CI 0.03–1.39, p 0.16). None of the S. aureus isolates were resistant to vancomycin, rifampicin, linezolid or teicoplanin. All phenotypically cefoxitin-resistant isolates were also oxacillin-resistant, carrying mecA and were characterized as MRSA. Fusidic acid resistance was more common in isolates derived from SSTI and TMD infections (OR 7.2, 95%CI 2.42–20.3, p 0.0002). We did not observe significant differences in methicillin resistance among CA and HA isolates (OR 2.78, 95%CI 0.66–12.79, p 0.17). CA-isolates were highly resistant to penicillin compared to healthcare-associated isolates (OR 75.2, 95%CI 22.7–223.4, p <0.0001) and they did not differ regarding resistance to other antimicrobials.
Methicillin resistance declined significantly from 18.3% in period A to 6.3% in period B (p 0.03). A significant increase in fusidic acid and mupirocin resistance was observed (fusidic acid resistance x2 test for trend, P 0.01; mupirocin resistance x2 test for trend, p <0.0001).
All 139 isolates were tested for virulence genes; 13 isolates (9.4%) harboured lukS/lukF-PV, 27 (19.4%) eta, six (4.3%) etb, 13(9.4%) tst, 125 (89.9%) fnbA and 21 (15.1%) fnbB genes. MRSA showed higher detection rates of lukS/lukF-PV and tst genes as compared to MSSA isolates and MSSA showed higher rates of eta gene carriage. Detection rates of eta gene was also higher among isolates derived from TMD as compared to those causing INV/ SSTIs (OR 3.5, 95%CI 1.36–9.29, p 0.01). MSSA isolates showed polyclonality by PFGE and were classified into 20 pulsotypes including one to 40 strains each. However, two predominant pulsotypes were identified: type 1 with ten subtypes (ST121) including 40 strains and type 2 with seven subtypes and 16 strains. Sixteen PFGE types included one to eight MSSA strains each. Two additional MSSA were classified to PFGE types C (ST80) and H (ST225). MRSA belonged to four pulsotypes; type A: four strains (ST30), B: four strains (ST239), C: five strains (ST80) and H: three strains (ST225). Multilocus sequence typing identified five sequence types among the 58 tested isolates. All 23 isolates from children with SSSS were characterized by MLST. Pulsotype 1, ST121 was predominant (13/23, 56.5%, p 0.002). Of 23 SSSS isolates, nine isolates (39.1%) were shown to carry eta gene, three (13%) carried etb, one isolate (4.3%) carried both eta and etb. Isolates harbouring eta or etb genes all belonged to PFGE type1, ST121. In our study, 40 isolates belonged to ST121, all of them were MSSA and CA in their majority (31/40, 77.5%). As compared to other isolates, ST121 was highly resistant to fusidic acid (80%), tobramycin (35%), mupirocin (25%) and clindamycin (25%), rarely carried PVL and tst genes (3/40 and 0/40, respectively) but harboured epidermolysin genes (eta 27/40; etb 6/40; both eta and etb 4/40) and genes facilitating adherence to the epithelium (fnbA 40/40) and causing mainly SSTIs (26/40), SSSS (13/40) and one case of invasive disease. Compared to other pulsotypes, mupirocin resistance and presence of eta/etb genes was exclusively linked to pulsotype 1 (p <0.0001).
Conclusion: In our population, community-associated MSSA was the predominant cause of S. aureus infections characterized by polyclonality, increasing resistance to fusidic acid and mupirocin. PFGE type 1 ST121 clone, harboured exfoliative toxins genes and was associated with rising trends of SSSS.
|
Collections
