Panton-Valentine Leukocidin Genes Are Associated With Enhanced Inflammatory Response and Local Disease in Acute Hematogenous Staphylococcus aureus Osteomyelitis in Children
BACKGROUND.Staphylococcus aureus strains carrying the genes encoding Panton-Valentine leukocidin (pvl-positive [pvl+]) are associated with more febrile days and higher complication rates of osteomyelitis in children than are pvl-negative (pvl−) strains.
OBJECTIVES. Selected clinical, laboratory, and radiographic findings in children with osteomyelitis caused by pvl+ and pvl−S aureus strains were compared.
METHODS. The demographics, selected clinical features, laboratory values, and radiographic findings of children with community-acquired S aureus osteomyelitis prospectively identified at Texas Children's Hospital between August 2001 and July 2004 were reviewed. Polymerase chain reaction was performed to detect the genes for pvl (luk-S-PV and luk-F-PV) and fibronectin-binding protein (fnbB) in S aureus isolates. χ2, 2-sample t test, and multiple logistic regression were used for statistical analysis.
RESULTS. Methicillin-susceptible and methicillin-resistant S aureus (MSSA and MRSA, respectively) caused osteomyelitis in 33 and 56 children, respectively. Twenty-six isolates were pvl− (26 MSSA), 59 were pvl+ (3 MSSA, 56 MRSA), and 4 were not available for analysis (4 MSSA). On univariate analysis, patients with pvl+S aureus isolates had significantly higher erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) level both at presentation and as a maximum value during hospitalization and were more likely to have a blood culture positive for S aureus during their admission. Patients with pvl+S aureus isolates were significantly more likely to have concomitant myositis or pyomyositis compared with patients with pvl−S aureus isolates on MRI. In a multivariate analysis pvl remained significantly associated with ESR and CRP levels at presentation and blood culture positive for S aureus. pvl+ status and younger age were associated with myositis on MRI.
CONCLUSIONS. Osteomyelitis caused by pvl+S aureus strains were associated with more severe local disease and a greater systemic inflammatory response compared with osteomyelitis caused by pvl−S aureus.
Staphylococcus aureus is the major cause of acute hematogenous osteomyelitis (AHO) in otherwise healthy children, accounting for 70% to 90% of cases.1,2 Community-acquired (CA) methicillin-resistant S aureus (MRSA) infections are now common in many areas of the United States.3–9 At Texas Children's Hospital (TCH), MRSA accounted for 76% of CA S aureus isolates from August 1, 2003, to July 31, 2004.8 Although the spectrum of illnesses caused by CA-MRSA remains primarily skin and soft tissue infections, similar to that of CA methicillin-susceptible S aureus (MSSA), over the past decade CA-MRSA has been increasingly responsible for invasive infections with substantial morbidity and mortality in children.5–9 From August 1, 2001, to July 31, 2004, at TCH, 117 CA-MRSA isolates were recovered from children with invasive infections, the most common of which was osteomyelitis.8
Multiple virulence factors have been identified as important to the pathogenesis of S aureus infections. One such factor is Panton-Valentine leukocidin (PVL), a bicomponent cytotoxin that destroys leukocytes by pore-forming activity and is encoded by pvl genes, luk-S-PV and luk-F-PV.10,11 PVL is a member of the synergohymenotropic family of exotoxins that destroy leukocytes by creating pores in the cell membrane. In vivo studies have shown that PVL injected intradermally into rabbits caused severe inflammatory lesions including capillary dilation, chemotaxis, polymorphonuclear infiltration, polymorphonuclear karyorrhexis, and skin and tissue necrosis.12 Clinically, PVL has been linked to a severe form of necrotizing pneumonia and necrotic skin lesions such as furuncles and carbuncles.10,11,13–15
The pvl genes are more common in CA-MRSA isolates compared with CA-MSSA isolates,7,13,16,17 and a recent molecular study at TCH showed the presence of pvl in almost all of the CA-MRSA strains.18 More than 90% of the CA-MRSA isolates at TCH belong to 1 clone, identical or closely related to the USA300 clone (Centers for Disease Control and Prevention)18,19 that currently predominates in many areas of the United States.18–21 The predominant (USA300) clone at TCH carries pvl as well as fnbB encoding S aureus fibronectin-binding factor.18 Fibronectin-binding proteins FnBPA and FnBPB, encoded by fnbA and fnbB, are involved in the adherence of S aureus to fibronectin and have been associated with enhanced bacterial adhesion and invasion.22–25
Although there is no evidence that the presence of PVL genes confers a greater risk of AHO, we have reported that children with osteomyelitis caused by S aureus isolates carrying the pvl genes (pvl+) experienced more febrile days and had a greater complication rate than patients with pvl-negative (pvl−) isolates.7 To our knowledge, other clinical features of children with AHO caused by pvl+ versus pvl−S aureus have not been characterized. In this study, we compared selected laboratory and radiographic findings in children with AHO caused by pvl+ and pvl−S aureus and also investigated the role of fnbB as an independent factor contributing to these findings.
Patient Identification and Medical Chart Review
From August 2001 to July 2004, all S aureus strains isolated by the TCH clinical microbiology laboratory were collected as part of prospective surveillance. A standardized form including antibiotic susceptibility and patient demographic and diagnostic information was filled out for each S aureus strain, and this information was entered into a computerized patient database. For this study, all patients who were diagnosed with AHO caused by S aureus were selected, and their medical charts were reviewed. The diagnosis of AHO was made via a consistent clinical presentation (patients had no history of penetrating trauma or overlying infection at the site of the osteomyelitis) and radiographic evidence on ≥1 of the following: plain film, computed tomography, radioisotope scan, or MRI. All of the patients had a positive culture from bone biopsy and/or blood culture for S aureus. Only patients with CA S aureus infection were included in this study. The strains were classified as described previously8,18 and in accordance with Centers for Disease Control and Prevention classifications.26 Briefly, CA S aureus infection was defined as a positive culture taken within 48 hours of hospitalization or after 48 hours of hospitalization if the patient's clinical presentation was consistent with CA infection and the culture was obtained from bone cultures later in the admission. Patients with underlying illness or prior surgery at the site of the infection were excluded from this study.
We collected information on patient demographics, clinical presentation, hospital course (including duration of symptoms before hospitalization, antibiotic use before admission, site of infection, surgical treatment, and ICU days), laboratory evaluations (including erythrocyte sedimentation rate [ESR] at presentation [within 24 hours of admission], maximum ESR value during hospitalization, C-reactive protein [CRP] value at presentation, maximum CRP value during hospitalization, white blood cell [WBC] count at admission, absolute neutrophil count [ANC] at admission, and incidence of blood cultures positive for Saureus), and radiographic studies (plain film, radioisotope scan, MRI scan, and computed tomography scan) from the medical charts. The Baylor College of Medicine Institutional Review Board approved this study.
S aureus Identification
Student's t test, χ2 analysis, log-likelihood ratio test, and multivariate analysis using logistic regression were used for statistical analysis. All analyses were 2-tailed, and P < .05 was considered statistically significant. Statistical analysis was performed by using True Epistat, fifth edition (Epistat Services, Richardson, TX).
During the study period of August 2001 to July 2004, 89 patients were identified as having AHO caused by CA S aureus. Thirty-three of these patients had CA-MSSA, and 56 had CA-MRSA. With the exception that the proportion of cases caused by CA-MSSA was greater in Hispanic children and the proportion of cases caused by CA-MRSA infection was greater in black children (P = .007), no significant differences were found in the demographic characteristics between the children with CA-MSSA compared with CA-MRSA osteomyelitis (Table 1).
Twenty-six S aureus isolates were pvl− (26 MSSA, 0 MRSA), and 59 were pvl+ (3 MSSA, 56 MRSA; P < .000001); 4 MSSA isolates were not available for molecular analysis. Related to the differences in distribution of CA-MRSA versus CA-MSSA isolates among the children of different races, pvl+S aureus was less likely to be isolated from Hispanic children but more likely to be isolated from black children (P = .01). The mean age of the children with pvl+ isolates was significantly greater than for those whose isolates were pvl− (P = .046) (Table 2).
Clinical and Laboratory Features of Patients With Osteomyelitis Caused by pvl+ and pvl−S aureus Isolates
Patients with osteomyelitis caused by pvl+S aureus (15.3%) were more likely to have multiple sites of osteomyelitis compared with patients with pvl−S aureus (3.8%), although this difference was not statistically significant. For patients with a single site of infection, patients with pvl+ and pvl− osteomyelitis had similar locations of infection. There was no difference between the pvl+ and pvl− patients with regard to receiving antibiotics before diagnosis or in the duration of symptoms before diagnosis. Surgical interventions (including both aspiration and incision-and-drainage procedures) were performed in a high percentage of patients in both the pvl+ and pvl− patient groups (Table 3).
pvl+S aureus isolates (18.6%) were more likely to cause severe infection that required care in the ICU than were pvl− isolates (3.8%) (P = .01). Patients with pvl+S aureus isolates had a significantly higher ESR and CRP level, both at presentation (P = .0008 and P < .000002, respectively) and as a maximum value during hospitalization (P < .0000001 and P < .0000001, respectively). Likewise, pvl+ patients had higher WBC counts (P = .03) and ANCs (P = .002) on presentation and were more likely to have a blood culture positive for Saureus during their admission (P = .0001) (Table 3).
Radiographic Findings in Patients With Osteomyelitis Caused by pvl+ and pvl−S aureus Isolates
Patients with osteomyelitis caused by pvl+ and pvl−S aureus had similar radiographic studies during their initial diagnostic workup. Seventy-six patients (24 pvl−, 48 pvl+, 4 pvl-unknown) had a plain film, 9 (2 pvl−, 6 pvl+, 1 pvl-unknown) of which revealed early stages of osteomyelitis. Twenty-seven patients (6 pvl−, 18 pvl+, 3 pvl-unknown) had a radioisotope scan, 16 (3 pvl−, 12 pvl+, 1 pvl-unknown) of which were positive for osteomyelitis. Sixty-eight patients (19 pvl−, 45 pvl+, 4 pvl-unknown) had an MRI, 67 (18 pvl−, 45 pvl+, 4 pvl-unknown) of which had a positive study.
Of the patients who had an MRI as part of their initial diagnostic workup, the patients with pvl+S aureus isolates were significantly more likely to have concomitant myositis or pyomyositis (P = .05) and tended to have more subperiosteal or intraosseal abscesses (P = .06) compared with patients with pvl−S aureus isolates (Table 3).
Laboratory and Radiographic Features of Patients With Osteomyelitis Isolated to the Femur, Tibia, or Fibula Caused by pvl+ and pvl−S aureus
Because the patients with infections caused by pvl+S aureus isolates included more children with multifocal disease and more children requiring ICU care compared with the patients with pvl−S aureus, patients with osteomyelitis isolated to the femur, tibia, or fibula who did not require ICU care were analyzed separately. Fifty-one patients had osteomyelitis confined to the femur, tibia, or fibula, 35 of whom had pvl+S aureus and 16 of whom had pvl−S aureus. Four of these patients (all pvl+) required ICU care and were excluded from this analysis.
The patients with pvl+S aureus isolates had a significantly higher ESR (P = .001), CRP level (P = .006), and ANC (P = .004) at presentation, as well as maximum ESR (P < .0000001) and CRP level (P = .0004) during hospitalization. In addition, the patients with pvl+S aureus compared with patients with pvl−S aureus were significantly more likely to have a blood culture positive for Saureus (70.0% vs 12.5%, respectively; P = .0004) and tended to have more subperiosteal/intraosseal abscesses on MRI (72.0% vs 38.5%; P = .1) than the patients with pvl−S aureus isolates. There was no significant difference in the incidence of surrounding myositis on MRI between patients with pvl+ and pvl−S aureus isolates (64.0% vs 61.5%) or in the WBC count on admission (P = .08).
Laboratory and Radiographic Features of Patients With Osteomyelitis Caused by fnbB+ and fnbB−S aureus
Nineteen S aureus isolates were fnbB− (4 MRSA, 15 MSSA; 5 pvl+, 14 pvl−), and 66 isolates were fnbB+ (52 MRSA, 14 MSSA; 54 pvl+, 12 pvl−). Four MSSA isolates were not available for fnbB analysis. MRSA isolates (P = .000006) and pvl+ isolates (P = .00002) were significantly more likely to be fnbB+ than MSSA and pvl− isolates, respectively. There were no demographic differences between patients with fnbB+ and fnbB− isolates. Patients with fnbB+ isolates were not significantly more likely to require ICU care than the patients with fnbB− isolates (16.7% vs 5.3%, respectively).
When the laboratory and radiographic findings of patients with fnbB+S aureus strains were compared with those of the patients with fnbB−S aureus strains, there were no significant differences in the ESR, CRP level, WBC count, or ANC at presentation, the maximum CRP level during admission, incidence of S aureus–positive blood cultures, or radiographic evidence of myositis or subperiosteal/intraosseal abscess. Only the maximum ESR during hospital admission varied significantly by fnbB status, with fnbB+ patients having higher maximum ESR values than fnbB− patients (P = .03) (Table 4).
Age, gender, race, duration of symptoms, prior antibiotic therapy, methicillin susceptibility, pvl status, and fnbB status were included in a multivariate analysis to determine which factors correlated independently with our laboratory and radiographic parameters. Only the presence of pvl remained significantly independently associated with an ESR of >75 mm/hour at presentation (P = .002) and maximum value during admission (P = .000002), CRP level of >10 mg/dL at presentation (P = .0003) and maximum value during admission (P = .00002), and blood culture positive for S aureus (P = .0002). pvl+ status (P = .01) and younger age (P = .02) were associated with myositis on MRI. Methicillin resistance remained significantly associated with subperiosteal/intraosseal abscesses on MRI (P = .01).
PVL is an S aureus exotoxin that has been linked to necrotic skin lesions and severe necrotizing pneumonia.10,11,13–15 In 1999, Lina et al13 reported a study of 172 clinical S aureus isolates, only 37% of which were pvl+. When the isolates were categorized according to disease, 93% of 30 necrotic skin isolates and 85% of 27 primary CA pneumonia isolates were pvl+. In contrast, only 23% of the 13 strains isolated from patients with osteomyelitis were pvl+.13
Although there is no proven association between pvl and development of osteomyelitis, we recently reported a retrospective analysis of children with musculoskeletal S aureus infections, in which children with osteomyelitis caused by pvl+S aureus had more febrile days and were more likely to have a complication such as a deep venous thrombosis or to develop chronic osteomyelitis than children with osteomyelitis caused by pvl−S aureus.7 In our current study, multivariate analysis of several factors including methicillin susceptibility and presence of pvl and fnbB genes again showed an association between pvl (or an unknown pvl-linked factor) and a more severe systemic inflammatory response, S aureus–positive blood cultures, and greater frequency of contiguous myositis/pyomyositis than pvl−S aureus isolates. Thus, PVL or an associated factor likely contributes to the severity of AHO caused by CA S aureus.
In patients who are undergoing treatment for AHO, the ESR and CRP values typically increase until the second or third day of therapy, after which the CRP level falls rapidly within 1 week, which is an indicator of acute response to therapy. The ESR takes weeks to normalize and is often used to gauge length of therapy.1,2,29,30 The WBC count frequently is not elevated in children with AHO.1,2,29,30 In our study, 95.5% and 95.0% of patients had elevated ESR and CRP values, respectively, and 53.9% of patients had an elevated WBC count of >12000/mm3 at admission. The patients with pvl+S aureus isolates had a significantly increased systemic inflammatory response compared with patients with pvl−S aureus, as evidenced by higher ESR, CRP, WBC count, and ANC values at presentation and maximum ESR and CRP level during admission.
Roine et al31 reported that children with higher CRP values on days 1 through 6 of admission and ESR values on days 4 through 7 of admission were more likely to have sequelae at their 1- to 2-month follow-up appointments than children with lower CRP and ESR values. All patients who had sequelae and had an organism identified as responsible for their AHO (21 of 28) had S aureus identified.31 Both the higher CRP and ESR values and the development of sequelae were likely secondary to the development of more invasive disease and could be a consequence of S aureus organisms with increased virulence secondary to a factor such as PVL.
Overall, 30% to 50% of children with AHO caused by any organism have S aureus–positive blood cultures.1 In our previous study, the number of days with persistently positive blood cultures between patients with osteomyelitis caused by pvl+ and pvl−S aureus was not different.7 However, in our current study, patients with pvl+S aureus isolates were significantly (P = .0001) more likely to have an S aureus–positive blood culture during admission compared with children with pvl−S aureus (67.2% vs 19.2%, respectively). Patients who required ICU care and those with multifocal osteomyelitis in our study tended to have pvl+S aureus isolated more commonly than patients who did not require ICU care or who had a single focus of disease. This, along with the increased incidence of S aureus–positive blood cultures, indicates that pvl likely plays some role, or is a marker for some other factor, in the increasing numbers of children who are being reported with severe systemic invasive disease, including septic shock, secondary to S aureus.32
In addition, when we analyzed the subgroup of patients in our study who had osteomyelitis that was isolated to the femur, tibia, or fibula and did not require ICU care, the children with pvl+S aureus isolates remained significantly (P = .0004) more likely to have an S aureus–positive blood culture (70.0% vs 12.5%, respectively), a more severe inflammatory response (including ESR, CRP level, and ANC at presentation and maximum ESR and CRP level during admission), and more extensive local disease (with more subperiosteal/intraosseal abscesses seen on MRI). Thus, the differences between the laboratory and radiographic features of the pvl+ and pvl− patients in this study are not simply secondary to the fact that the pvl+ group contained more patients with multifocal disease and who required ICU care.
We also screened our isolates for the S aureus gene fnbB in this study. A greater proportion of MRSA isolates carry pvl and fnbB compared with MSSA isolates, and both factors have been linked to the increased virulence of CA-MRSA. The products of fnbB mediate S aureus adhesion to epithelial cells and are hypothesized to facilitate internalization of the bacteria into these cells.22–24 When infused into mice, S aureus strains that are fnbB+ are associated with induction of systemic inflammation that is characterized by interleukin-6 secretion, significant weight loss, and mortality.33 Peacock et al23 reported that fnbB was significantly more frequently detected in CA-invasive strains (responsible for endocarditis plus septic arthritis and/or osteomyelitis) compared with carriage isolates (84.4% vs 68.4%, respectively; P = .02).23 The patients in our study did have a high frequency of fnbB+ isolates (77.6% were positive), but the presence of fnbB was only associated with a greater mean maximum ESR during hospitalization in univariate analysis and was not independently associated with a greater systemic inflammatory reaction in our multivariate analysis. Furthermore, patients with fnbB+ isolates were not more likely to have a subperiosteal/intraosseal abscess or surrounding myositis/pyomyositis on MRI, and in our previous study, fnbB+ isolates were not associated with higher complication rates in children with musculoskeletal infections that were caused by S aureus.7
All parameters included in this study (ESR, CRP level, WBC count, and ANC on admission, maximum ESR and CRP level during hospitalization, incidence of blood culture positive for S aureus, and subperiosteal/intraosseal abscess on MRI) except for surrounding myositis/pyomyositis on MRI were significantly associated with methicillin resistance on univariate analysis, which indicates that children with MRSA isolates were more likely to have evidence of increased systemic inflammatory response and more extensive local disease than children with MSSA isolates. However, on multivariate analysis, methicillin resistance only remained significantly associated with subperiosteal/intraosseal abscesses.
Our study supports the hypothesis that pvl is strongly associated with the severity of AHO caused by S aureus in children. Because the USA300 clone that is predominant in our area is present in many parts of the United States,18–21 the results of this study and the significance of pvl are relevant to physicians across the country. If future studies determine that PVL specifically contributes to the pathophysiology of invasive staphylococcal infections, therapies to combat PVL may prove beneficial.34
- Accepted April 20, 2005.
- Address correspondence to Sheldon L. Kaplan, MD, Texas Children's Hospital, Mail Code 3-2371, 6621 Fannin St, Houston, TX 77030. E-mail:
Financial Disclosure: Dr Kaplan has received a grant from Pfizer.
- ↵Krogstad P. Osteomyelitis and septic arthritis. In: Feigin RD, Cherry JD, Demmler GJ, Kaplan SL, eds. Textbook of Pediatric Infectious Diseases. 5th ed. Philadelphia, PA: WB Saunders;2004:713– 736
- ↵Sattler CA, Mason EO Jr, Kaplan SL. Prospective comparison of risk factors and demographic and clinical characteristics of community-acquired, methicillin-resistant versus methicillin-susceptible Staphylococcus aureus infection in children. Pediatr Infect Dis J.2002;21 :910– 917
- ↵Kaplan SL, Hulten KG, Gonzalez BE, et al. Three-year surveillance of community-acquired Staphylococcus aureus infections in children. Clin Infect Dis.2005;40 :1785– 1791
- ↵Dufour P, Gillet Y, Bes M, et al. Community-acquired methicillin-resistant Staphylococcus aureus infections in France: emergence of a single clone that produces Panton-Valentine leukocidin. Clin Infect Dis.2002;35 :819– 824
- ↵Lina G, Piemont Y, Godail-Gamot F, et al. Involvement of Panton-Valentine leukocidin-producing Staphylococcus aureus in primary skin infections and pneumonia. Clin Infect Dis.1999;29 :1128– 1132
- Prevost G, Couppie P, Prevost P, et al. Epidemiological data on Staphylococcus aureus strains producing synergohymenotropic toxins. J Med Microbiol.1995;42 :237– 245
- ↵Francis JS, Doherty MC, Lopatin U, et al. Severe community-onset pneumonia in healthy adults caused by methicillin-resistant Staphylococcus aureus carrying the Panton-Valentine leukocidin genes. Clin Infect Dis.2005;40 :100– 107
- ↵Ellis MW, Hospenthal DR, Dooley DP, Gray PJ, Murray CK. Natural history of community-acquired methicillin-resistant Staphylococcus aureus colonization and infection in soldiers. Clin Infect Dis.2004;39 :971– 979
- ↵Diep BA, Sensabaugh GF, Somboona NS, Carleton HA, Perdreau-Remington F. Widespread skin and soft-tissue infections due to two methicillin-resistant Staphylococcus aureus strains harboring the genes for Panton-Valentine leucocidin. J Clin Microbiol.2004;42 :2080– 2084
- ↵McDougal LK, Steward CD, Killgore GE, Chaitram JM, McAllister SK, Tenover FC. Pulsed-field gel electrophoresis typing of oxacillin-resistant Staphylococcus aureus isolates from the United States: establishing a national database. J Clin Microbiol.2003;41 :5113– 5120
- ↵Nashev D, Toshkova K, Salasia SI, Hassan AA, Lammler C, Zschock M. Distribution of virulence genes of Staphylococcus aureus isolated from stable nasal carriers. FEMS Microbiol Lett.2004;233 :45– 52
- ↵Kloos WE, Bannerman TL. Staphylococcus and Micrococcus. In: Murray PR, Baron EJ, Pfaller MA, Tenover FC, Yolken RH, eds. Manual of Clinical Microbiology. 7th ed. Washington, DC: American Society for Microbiology;1999:262– 282
- ↵Clinical and Laboratory Standards Institute. Performance Standards for Antimicrobial Disk Susceptibility Testing; Fifteenth Informational Supplement. Wayne, PA: Clinical and Laboratory Standards Institute;2005
- ↵Unkila-Kallio L, Kallio MJ, Eskola J, Peltola H. Serum C-reactive protein, erythrocyte sedimentation rate, and white blood cell count in acute hematogenous osteomyelitis of children. Pediatrics.1994;93 :59– 62
- ↵Roine I, Arguedas A, Faingezicht I, Rodriguez F. Early detection of sequela-prone osteomyelitis in children with use of simple clinical and laboratory criteria. Clin Infect Dis.1997;24 :849– 853
- ↵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
- ↵Palmqvist N, Foster T, Fitzgerald JR, Josefsson E, Tarkowski A. Fibronectin-binding proteins and fibrinogen-binding clumping factors play distinct roles in staphylococcal arthritis and systemic inflammation. J Infect Dis.2005;191 :791– 798
- ↵Gauduchon V, Cozon G, Vandenesch F, et al. Neutralization of Staphylococcus aureus Panton Valentine leukocidin by intravenous immunoglobulin in vitro. J Infect Dis.2004;189 :346– 353
- Copyright © 2006 by the American Academy of Pediatrics