Staphylococcus aureus (SA) is the leading cause of bone and joint infections (BJI).1 Community-acquired SA (CA-SA) acute osteomyelitis, pyomyositis and septic arthritis arise most commonly from hematogenous seeding of bacteria into bone, muscle and joints in previously healthy children without recognized preceding infections or risk factors.2 In the past 20 years, an increased number of severe cases of CA-SA BJI have been reported.3,4
It has been suggested that SA methicillin resistance could lead to an increase in illness severity, implicating, in particular, the USA300 methicillin-resistant clone.5 Other host and pathogen factors related to severity have been studied, including Panton-Valentine leukocidin (PVL), a pore-forming toxin produced by some strains of SA. There is some evidence suggesting its presence is indeed related to disease severity, but there are also data questioning the importance of PVL in pathogenesis of severe infections.6,7,8
The main aims of this prospective study were to define the clinical and microbiologic characteristics of CA-SA BJI in children and to analyze host and pathogen factors related to disease severity.
MATERIAL AND METHODS
This was a prospective multicenter European study performed during a 2-year period (October 1, 2012–September 30, 2014) using a standardized database which incorporated epidemiologic (sex, age, country), microbiologic (site of isolation, antibiotic resistance, PVL presence) and clinical data [presence of risk factors for severe infection, signs and symptoms on admission, antibiotics and surgical treatment, admission to pediatric intensive care unit (PICU) and ventilatory or hemodynamic support].
All children 1 day to 16 years of age with CA-SA invasive infections meeting the criteria (clinical infection as previously defined9 with SA isolation from a normally sterile body site, obtained at the emergency room or up to 48 hours after admission) were enrolled at 13 European centers of 7 different countries (Spain, Lithuania, Israel, Italy, Germany, Greece and Romania). General data of this group (n = 152) has already been reported.10 For the present study, only children with musculoskeletal infections were selected. Data were collected by the main investigator of each site from patients, parents and medical files. BJI were defined as musculoskeletal clinical infections with isolation of SA from synovial fluid, bone aspirate, muscular or bone abscesses or blood samples, which were collected in all cases when infection was suspected.
Only community-acquired infections were included, defined as a positive culture obtained at hospital admission or up to 48 hours later. Severe infection was defined as those requiring the admission to PICU secondary to hemodynamic instability or respiratory failure, as considered by the attending clinician or death.
Biologic samples were processed at each local laboratory with the identification of SA and antimicrobial testing performed by standard procedures as per Clinical and Laboratory Standards Institute. Samples were sent to a central laboratory (Microbiology Department, Hospital Universitario 12 de Octubre, Madrid, Spain) where methicillin-resistant SA (MRSA) isolates, previously identified, underwent confirmatory polymerase chain reaction analysis for the mecA gene. The presence of PVL genes (lukS-PV and lukF-PV) was determined by polymerase chain reaction, in all isolates collected, at the central laboratory, as previously described.11
All continuous variables were expressed as mean ± SD. For univariate analysis, dichotomous variables were analyzed by the χ2 test or by Fisher exact test; continuous variables were analyzed by Student t test. Odds ratios (ORs) were calculated with a 95% CI. Multivariable logistic regression analysis was performed to assess independent associations. The model was validated by Hosmer-Lemeshow test, and the simultaneous entry method was used. Independent variables included in this analysis were those observed to be statistically significant at the univariate analysis and previously reported to be associated in the medical literature, and the number of variables was limited to 4. Only clinical and microbiologic variables were included to assess risk factors. Adjusted ORs (aOR) were calculated by Exp(b) value. A P value of < 0.05 was considered to be statistically significant. Analyses were performed by SPSS 20 (IBM SPSS, Chicago, IL). This study received the approval of the 12 de Octubre Hospital’s clinical research ethics committee.
A total of 85 children (51 boys; 60%) with CA-SA BJI (62 osteomyelitis, 21 arthritis and 2 pyomyositis) were prospectively included in this study from the following countries: Spain (36), Lithuania (30), Israel (8), Greece (4), Germany (4), Italy (2) and Romania (1). Mean age was 8.7 ± 4.8 years.
Medical History and Clinical Outcome
More than half of our patients had risk factors for BJI (36% had recent trauma, and 16% had cutaneous infection). Bloodstream infection was detected in 42 patients (49%). Eleven patients (13%) had a severe infection and were admitted to PICU due to need of hemodynamic (8/11) or respiratory support (3/11); none of them died. Some of the children admitted to PICU developed septic shock (6/11) or pulmonary thromboembolism (3/11) as complications.
All but 2 patients received β-lactam antimicrobials; cloxacillin was given in 41% as empiric or definitive therapy targeting SA. Glycopeptides (mostly vancomycin) were used in 18%. Twelve patients (14%), only one of them presenting with severe infection (8%), received a protein synthesis inhibitor antibiotic (clindamycin or linezolid) during the first 48 hours of admission. No intravenous immunoglobulin was given.
At least one surgical procedure was performed in 60 children (71%), and 25 (29%) needed more than one intervention. One patient developed chronic osteomyelitis in the follow up, and there were no relevant sequelae reported.
PVL genes were detected in 17% of the isolates (12 of 70 strains available), and methicillin resistance was detected in 6% (5/85, 2 of them carrying PVL genes). Additionally, 8 isolates (9%) were resistant to erythromycin, 4 (5%) to clindamycin and 2 to linezolid. No resistance to trimethoprim/sulfamethoxazole was found.
Factors Related to Severity
Severe and nonsevere infections were compared (Table 1). There were no differences regarding age, gender or predisposing conditions between the groups. Time of symptoms before admission was similar between groups. Higher C-reactive protein values on admission and PVL presence were predictors of severe infection. Leukopenia at admission (white blood cell count < 3000/mm3) was found in one single patient.
In comparing PVL-positive with PVL-negative infections, patients with PVL-positive infections had longer mean of hospital stay (24.3 ± 16.1 days vs. 18.0 ± 12.0 days; P = 0.04) and more need for surgical interventions (mean 1.72 ± 0.4 vs. 0.96 ± 0.1 days; P = 0.04).
No differences in demographic data were observed comparing MRSA infections with methicillin-susceptible SA infections, and methicillin resistance was not related to an increased need for surgery (1.2 vs. 1.1 surgeries; P = 0.9) or longer hospital stay (18 vs. 27 days; P = 0.18).
The multivariate analysis identified PVL-positive isolate [aOR, 12.5 (1.75–89); P = 0.01] as the only factor independently associated with severe outcome (Table 1).
Some of the risk factors of bad clinical outcome in CA-SA BJI in children are still to be determined. BJI are considered potential severe infections, confirming our observations in which 11 patients (13%), most of them previously healthy, required admission to PICU as a result of hemodynamic or respiratory failure from pulmonary thromboembolism or septic shock, uncommon but well-described complications of these infections.
PVL isolates in the United States are linked to the rising of USA300 clone, that harbors PVL genes in a high proportion (SCCmec type IV and the toxin gene were detected in 98% of MRSA isolated in skin and soft tissue infections in 2004),12 and to a lesser extent in other areas, where its incidence may be increasing in the last years.13 Our results showed that 17% (n = 12) of the isolates available were PVL-positive, most of them reported in Lithuania (8 cases) and Spain (2 cases). Although a previous European article mentions ≈2% PVL prevalence in SA isolates,14 a significant higher prevalence of PVL-positive strains in skin and soft tissue infections has also been described by our group in Spain (12/46 patients, 26.1%).15
In our work, PVL presence was found to be the only independent risk factor for severe infection in multivariate analysis (aOR, 12.5; P = 0.01). There have been recent studies trying to determine the role of the toxin in musculoskeletal infections in children, where some areas remain uncertain.16 When analyzing clinical outcome, we found PVL presence to be also associated with a higher need for surgical procedures and longer hospital stay. This is consistent with a series reported by Bocchini et al17 where PVL presence was associated with more severe local disease and a greater systemic inflammatory response in osteomyelitis caused by SA in children. Dohin et al,18 in comparing a cohort of 14 children infected with PVL-positive SA BJI with a similar group of PVL-negative strains infections, also found that PVL-positive cases had longer mean hospital stay (45 vs. 13 days; P < 0.001), had more prolonged antibiotic treatment (48 vs. 11.3 days; P < 0.001) and needed more surgical drainage (2.8 vs. 0.52 surgeries; P = 0.002). An association between PVL presence and a higher complication rate in BJI was also reported by Martinez-Aguilar et al,4 with no differences between MRSA and methicillin-susceptible SA infections.
In our study, MRSA caused less than 6% (n = 5) of the musculoskeletal SA infections. MRSA clones were not typed, and only the 2 of them carrying PVL genes caused a severe infection. Although it has been suggested that methicillin resistance may be associated with severe infections, we have not found it related to a longer hospitalization or need of more drainage procedures.
Our study shows that in Europe, the use of protein synthesis inhibitor antibiotics or intravenous immunoglobuline, as recommended in some guidelines and review articles,19 is not very common. According to these results, there is more evidence that clindamycin/linezolid may be considered as adjunctive treatment in a severe SA infection.
There are some limitations in this study. Not all the isolates were studied for PVL presence; however, PVL was tested in a high proportion (82%) of cases, and results would not be significantly affected. Due the very small number of MRSA cases, MRSA association with severity could have been underestimated. Also, MRSA strains were not typed. The major strengths of the study were the prospective, multicenter, multi-country nature of it, and the use of a standardized database for the collection of relevant data.
For all the effort in recruiting patients and their participation in the study, we thank Carlota Montagnani, Oana Falup-Pecurariu, Astra Vitkauskiene, Amadéu Gené, Raya Kudinsky, Emilia Cercenado, A. Lepe, Antonia Makri, Irina Kulecnikova, H.I. Huppertz, P. García-Hierro, Liviu Muntean, Giedre Trumpulyte, Clàudia Fortuny, Assaf Rokney, Mar Santos-Sebastián, O. Neth, Marianna Petra, S. Guillén, Luis Galli, Laura Bleotu and Esther Viedma.
1. Kavanagh N, Ryan EJ, Widaa A, et al. Staphylococcal osteomyelitis
: disease progression, treatment challenges, and future directions. Clin Microbiol Rev. 2018;31:e00084–17.
2. Peltola H, Pääkkönen M. Acute osteomyelitis
. N Engl J Med. 2014;370:352–360.
3. Mitha A, Boutry N, Nectoux E, et al. Community-acquired bone and joint infections in children
: a 1-year prospective epidemiological study. Arch Dis Child. 2015;100:126–129.
4. Martinez-Aguilar G, Hulten AAK, Hammerman W, Mason EO Jr, Kaplan SL. Community-acquired, methicillin-resistant and methicillin-susceptible Staphylococcus aureus
musculoskeletal infections in children
. Pediatr Infect Dis J. 2004;23:701–706.
5. Saavedra-Lozano J, Mejías A, Ahmad N, et al. Changing trends in acute osteomyelitis
: impact of methicillin-resistant Staphylococcus aureus
infections. J Pediatr Orthop. 2008;28:569–575.
6. Shallcross LJ, Fragaszy E, Johnson AM, Hayward AC. The role of the Panton-Valentine leucocidin toxin in staphylococcal disease: a systematic review and meta-analysis. Lancet Infect Dis. 2012;13:43e54.
7. Cunnington A, Brick T, Cooper M, et al. Severe invasive Panton-Valentine leucocidin positive Staphylococcus aureus
infections in children
in London, UK. J Infect. 2009;59:28–36.
8. Gonzalez BE, Martinez-Aguilar G, Hulten KG, et al. Severe staphylococcal sepsis in adolescents in the era of community-acquired methicillin-resistant Staphylococcus aureus
. Pediatrics. 2005;115:642–648.
9. Calvo C, Núñez E, Camacho M, et al. Epidemiology and management of acute uncomplicated septic arthritis
: spanish multicenter Study. Pediatr Infect Dis J. 2016;35:1288–1293.
10. Gijón M, Bellusci M, Petraitiene B, et al. Factors associated with severity in invasive community-acquired Staphylococcus aureus
infections in children
: a prospective European multicentre study. Clin Microbiol Infect. 2016;22:643.e1–643.e6.
11. Vannuffel P, Gigi J, Ezzedine H, et al. Specific detection of methicillin-resistant Staphylococcus species by multiplex PCR. J Clin Microbiol. 1995;33:2864–2867.
12. Moran G, Krishnadasa A, Gorwitz RJ, et al. Methicillin-resistant S. aureus
infections among patients in the emergency department. N Engl J Med. 2006;355:666e74.
13. van der Mee-Marquet N, Poisson DM, Lavigne JP, Francia T, Tristan A, Vandenesch F, Quentin R, Bertrand X. The incidence of Staphylococcus aureus
ST8-USA300 among French pediatric inpatients is rising. Eur J Clin Microbiol Infect Dis. 2015;34:935–942.
14. Holmes A, Ganner M, McGuane S, Pitt TL, Cookson BD, Kearns AM. Staphylococcus aureus
isolates carrying Panton-Valentine leukocidin genes in England and Wales: frequency, characterisation and association with clinical disease. J Clin Microbiol. 2005;43:2384–2390.
15. Daskalaki M, Rojo P, Marin-Ferrer M, et al. Panton-Valentine leukocidine positive Staphylococcus aureus
skin and soft tissue infections among children
in an emergency department in Madrid, Spain. Clin Microbiol Infect. 2010;16:74e7.
16. Ritz N, Curtis N. The role of Panton-Valentine leukocidin in Staphylococcus aureus
musculoskeletal infections in children
. Pediatr Infect Dis J. 2012;31:514–518.
17. Bocchini CE, Hulten KG, Mason EO Jr, Gonzalez BE, Hammerman WA, Kaplan SL. Panton-Valentine leukocidin genes are associated with enhanced inflammatory response and local disease in acute hematogenous Staphylococcus aureus osteomyelitis
. Pediatrics. 2006;117:433–440.
18. Dohin B, Gillet Y, Kohler R, et al. Pediatric bone and joint infections caused by Panton-Valentine leukocidin-positive Staphylococcus aureus
. Pediatr Infect Dis J. 2007;26:1042–1048.
19. Gillet Y, Dumitrescu O, Tristan A, et al. Pragmatic management of Panton-Valentine leukocidin–associated staphylococcal diseases. Int J Antimicrob Agents. 2011;38:457–464.