ARTICLE |
a Department of Pediatrics, Baylor College of Medicine, Houston, Texas
b Department of Neonatalogy
c Infectious Diseases Service, Texas Children's Hospital, Houston, Texas
| ABSTRACT |
|---|
|
|
|---|
OBJECTIVE. Our goal is to describe the clinical presentation of neonatal community-acquired S aureus disease and provide molecular analyses of the infecting isolates.
PATIENTS AND METHODS. We retrospectively reviewed the demographics and hospital course of term and near-term previously healthy neonates,
30 days of age, with community-acquired S aureus infections presenting after nursery discharge between August 2001 and March 2005 at Texas Children's Hospital. Prospectively collected isolates were characterized by pulsed-field gel electrophoresis, staphylococcal cassette chromosome mec type, and the presence of PVL genes.
RESULTS. Of 89 S aureus infections, 61 were methicillin-resistant S aureus; S aureus infections increased each year. Methicillin-resistant S aureus infections increased from 10 of 20 to 30 of 36 infections from 2002 to 2004. Most subjects, 65 of 89, were male. Symptoms began at 7 to 12 days of age for 26 of 45 male infants with methicillin-resistant S aureus. Most infections, 77 of 89, involved skin and soft tissue; 28 of 61 methicillin-resistant S aureus versus 7 of 28 methicillin-susceptible S aureus infections required drainage. Invasive manifestations included shock, musculoskeletal and urinary tract infection, perinephric abscess, bacteremia, empyema/lung abscess, and a death. Maternal S aureus or skin-infection history occurred with 13 of 61 methicillin-resistant S aureus versus 1 of 28 methicillin-susceptible S aureus infections. The predominant community clone, USA300 (PVL genes +), accounted for 55 of 57 methicillin-resistant S aureus and 3 of 25 methicillin-susceptible S aureus isolates.
CONCLUSIONS. Community-acquired methicillin-resistant S aureus is a substantial and increasing proportion of S aureus infections in previously healthy neonates. Male infants 7 to 12 days of age are affected most often. Neonatal community-acquired S aureus infection may be associated with concurrent maternal infection. USA300 is the predominant clone among these neonatal isolates in our region.
Key Words: community-acquired Staphylococcus aureus methicillin resistance infant newborn
Abbreviations: CAcommunity-acquired MRSAmethicillin-resistant S aureus SCCmecstaphylococcal cassette chromosome mec TCHTexas Children's Hospital PVLPanton-Valentine leukocidin PFGEpulsed-field gel electrophoresis CSFcerebrospinal fluid MSSAmethicillin-susceptible S aureus
Staphylococcus aureus frequently causes pediatric skin and soft tissue as well as invasive life-threatening infections.1 The number of community-acquired (CA) methicillin-resistant S aureus (MRSA) infections at Texas Children's Hospital (TCH) in Houston, Texas, has continued to increase since 2000. These children have no traditional risk factors for infection such as recent hospitalization, surgery, or underlying conditions.2,3 Although CA MRSAtype isolates have been documented in nosocomially transmitted infections in the NICU, we noted an unexpected peak of infections in healthy neonates presenting after nursery discharge when reviewing our experience with CA S aureus.4 This separate and increasing group of infections may result from a risk factor(s) specific to neonates and delivery.
Genetic variation affects the clinical manifestations of different S aureus clones. The staphylococcal cassette chromosome mec (SCCmec) region contains antibiotic-resistance genes; CA MRSA isolates carry smaller SCCmec types (IV and V) with fewer resistance genes than hospital-acquired strains.5 The accessory gene regulator region, agr, controls cellular metabolism and virulence factors including many toxins; agr type II has been associated with intermediate vancomycin resistance.6,7 An important virulence factor, Panton-Valentine leukocidin (PVL), encoded by the luk-S-PV and luk-F-PV genes, is a pore-forming exotoxin associated with necrotizing pneumonia and necrotic skin lesions in humans and severe inflammation in animal models.811 Previous studies have demonstrated that the USA300-ST 8 clone (PVL genes +, agr type I, SCCmec type IV) represents the majority of CA S aureus isolates in Houston and other regions in the United States.1215 In this report we describe community S aureus disease in previously healthy term or near-term neonates and characterize the isolated S aureus strains by molecular methods.
| PATIENTS AND METHODS |
|---|
|
|
|---|
Patients
30 days of age seen at TCH between August 1, 2001, and March 30, 2005, were selected from the database. By medical chart review, we identified neonates born at
36 weeks' gestation who were previously healthy and treated for S aureus infection presenting after nursery discharge. "Previously healthy" was defined as no hospitalizations since birth and no surgeries except circumcision. A data-collection form was developed to include more detailed information regarding the maternal history, clinical presentation and manifestations, management, and outcome. Data were entered into a separate database.
DNA Isolation and Molecular Characterization
S aureus DNA was prepared from all S aureus strains as described previously.12 Pulsed-field gel electrophoresis (PFGE) was performed by using a modified procedure based on the method of Murray et al.17,18 Interstrain relationships were determined as described previously.18,19 The SCCmec type was determined by using the methods of Okuma et al5 and Mishaan et al.12 The polymerase chain reaction for PVL genes (luk-S-PV and luk-F-PV) was performed as described elsewhere.12 The agr types were analyzed by polymerase chain reaction using published primers.20 Positive controls were USA300 (agr type I), N315 (agr type II), and TCH358 (agr type III). Amplicons representing each agr type were confirmed by sequencing and BLAST analysis (available at www.ncbi.nlm.nih.gov/BLAST).
Statistical Analysis
Statistical analysis was performed by using
2 or
2 for trend for dichotomous variables or Student's t test for comparison of means by using GraphPad Prism 4.0 (GraphPad, San Diego, CA). Analyses were 2 tailed, and a P value of <.05 was considered significant.
| CASE REPORT |
|---|
|
|
|---|
Chest radiograph and ultrasound showed a large complex right pleural effusion with septations and internal debris surrounding an area of lung consolidation resulting in mediastinal shift (Fig 1 A and B). Blood and cerebrospinal fluid (CSF) were obtained for culture. A thoracentesis revealed pleural fluid with Gram-positive cocci in clusters on Gram-stain. The infant was admitted to intensive care on oxygen and treated with ampicillin, gentamicin, and vancomycin.
|
The final chest tube was removed on day-of-age 31. Follow-up chest computed tomography on day-of-age 44 showed right-sided atelectasis and consolidation as well as fluid pockets in the right upper lobe suspicious for lung necrosis. The empyema was decreased in size (Fig 1C). After showing normal results from a nitroblue tetrazolium test, he was discharged on day-of-age 47.
| RESULTS |
|---|
|
|
|---|
Antibiotic Susceptibility
Twenty-eight isolates were susceptible to oxacillin by disk diffusion, and 61 were resistant. The number of S aureus isolates recovered from previously healthy infants increased each year resulting entirely from an increase in the absolute number of MRSA isolates. The proportion of MRSA isolates increased from 10 (50%) of 20 to 30 (83%) of 36 infections from 2002 to 2004 (P = .01), whereas the absolute number of methicillin-susceptible S aureus (MSSA) isolates remained steady or decreased slightly. Eight isolates showed clindamycin resistance; 4 (50%) were inducibly resistant according to the D-test.
Demographics and Epidemiology
Symptom onset peaked in the male neonates infected with MRSA between 7 and 12 days of age, with 58% of the infections occurring during this time (vs 25% for MSSA; P = .02). No peak time of symptom onset was seen in female infants or males with MSSA (Fig 2). Both MRSA and MSSA infections, however, occurred predominantly in males (74% vs 71%).
|
|
|
Sixty infants had lumbar punctures as part of their evaluation of serious bacterial infection. Although no CSF cultures grew S aureus, CSF pleocytosis was treated for at least 14 days in 2 patients (183 white blood cells per mm3 [85% monocytes]; 32 white blood cells per mm3 [92% monocytes]) because they received antibiotics before their lumbar punctures. The association with S aureus infection was not certain.
Of the 12 invasive infections (including all patients with bacteremia [n = 4], urinary tract infection [n = 3], and treated CSF pleocytosis [n = 2], as well as the only death), 10 occurred in male infants. The child who died was found at autopsy to have disseminated herpes simplex virus 1 and MRSA infection that resulted in liver failure, hyperammonemia, and coagulopathy. S aureus was isolated from the patient's blood (premortem and postmortem), lung, trachea, larynx, esophagus, adrenal gland, and liver.
The 3 other children with bacteremia had skin and soft tissue infections including orbital cellulitis with myositis, dactylitis, and sacral abscess. Two of 4 bacteremia isolates were MRSA. Both of the MSSA-infected patients required observation in a level 2 or 3 nursery at birth for transient tachypnea of the newborn and had symptom onset near the end of the first month of age. Both of the MRSA infections occurred on day-of-age 9, and neither child had level 2 or 3 nursery exposure at birth.
A greater proportion of infants with MRSA infections underwent drainage procedures (44% vs 25%) and placement of central venous catheters (10% vs 4%) than those with MSSA infections. However, these differences were not statistically significant.
Maternal History of Infection
Twenty-one percent (13 of 61) of the patients with MRSA infection compared with 4% (1 of 28) of those with MSSA infections had a mother with a history of skin infection, which was noted in the infant's medical chart (P = .06) (Table 3). Of 89 mothers, 12 had a skin infection at the time of their infant's presentation. For some of these mothers, this was their first significant skin infection. Two other mothers had skin and soft tissue infections during pregnancy or delivery. The mother of one infant had sepsis and thrombophlebitis and required intensive care. Most of these concurrent infections in mother-infant pairs occurred within the first 2 weeks after delivery. In this retrospective chart review, the bacterial cause of the maternal infections was not available in most cases. For 3 mothers, the infecting bacteria were identified as MSSA, MRSA, and Staphylococcus. Twenty-one percent of all the patients had a family member with a history of past or concurrent infection (Table 1).
|
Birth Hospitalization and Underlying Conditions
Thirteen patients received therapy beyond normal newborn care during their birth hospitalization. Examples of therapy received include antibiotics for 48 to 72 hours (n = 3), intravenous fluid for treatment of hypoglycemia (n = 3) and hyponatremia (n = 1), phototherapy (n = 6), and observation for transient tachypnea of the newborn or brief respiratory distress (n = 4), hypotonia, and hypothermia (n = 1). The proportion of MRSA to MSSA infections was similar to the group overall. All statistical analyses were repeated excluding this group of neonates, and all previously significant results remained significant.
Patients were classified as previously healthy on the basis of information available to the treating physicians at the time of admission. Two patients, however, during their evaluation were diagnosed with an underlying condition: Menke's kinky-hair syndrome and spina bifida occulta.
Circumcision
A similar circumcision rate was observed for MRSA-infected (60%) and MSSA-infected (53%) male infants. The circumcision rate was not significantly increased in male infants with infection onset at 7 to 12 days of age, the major cluster suggesting a specific exposure, compared with other ages. We did not find a greater proportion of isolated groin pustulosis infections in male infants with circumcisions compared with uncircumcised infant males. However, males did have significantly more infections that involved the groin than female infants (P < .0001). Only 2 children were evaluated by urology for an infection after circumcision. In one case, the urologist thought the circumcision site was a possible but unlikely source of infection. In the other case, the infection occurred >2 weeks after the initial appointment to remove a PlastiBell that failed to fall off, and no comment was made regarding a relationship between the procedure and the infection.
Molecular Analysis
Fifty-seven MRSA isolates were available for analysis. Forty-two had PFGE patterns consistent with the USA300 clone; 13 were variant strains of this predominant CA MRSA clone in Houston.12 Two strains were unrelated to USA300 and each other. Variant strains differed from USA300 by 1 to 3 bands. All MRSA strains carried genes consistent with SCCmec type IV (SCCmec B and ccr2 genes). All but one isolate amplified the J-1 region specific for subtype IVa. All USA300 strains carried agr type I and the PVL genes. Non-USA300 strains did not carry the PVL genes; 1 isolate carried agr type II, and 1 isolate did not carry agr types I, II, or III.
Of 25 MSSA isolates available for analysis, 3 were related to USA300 and carried the PVL genes, and 2 were USA200 variants. Only 1 non-USA300 strain carried the PVL genes, a USA200 variant. Among the MSSA isolates with unique pulsotypes, we found 2 groups containing 5 and 2 isolates, respectively (Fig 3).
|
| DISCUSSION |
|---|
|
|
|---|
Male predominance of infection is seen not only in our neonatal population but also in 60% to 65% of older children evaluated at TCH with severe or fatal staphylococcal disease.23 The reason for the predisposition of males to infection is unclear. The circumcision rate in our infants was lower than a previously reported rate of 85% in 2 Houston hospitals.24 In addition, we did not find an association between circumcision and type of infection or time of presentation. Several historical studies have found a similar predominance of S aureus infection in male neonates delivered at hospitals with high (>95%) and low (<1%) prevalences of circumcision.2527
The onset of symptoms peaked between 7 and 12 days of age, which likely relates to a specific time of exposure or birth-related risk factors. This hypothesis is supported by our observation of frequent concurrent onset of maternal infection, in some cases as a primary presentation. Because the mother's medical chart was not available for review, our retrospective review of the infant's chart may underestimate the number of mothers with concurrent infection and does not identify the bacterial cause of their infections. These infections may represent a common time of exposure at delivery or during the hospital stay. In a study completed in 1995 before the recognition of CA MRSA in Houston, the obstetrics and gynecology service at Baylor College of Medicine found that 17% of culture-positive wound infections after cesarean delivery resulted from S aureus.28 A more recent study of 57 pregnant patients with CA MRSA at Parkland Hospital in Dallas, Texas, a metropolitan center with disease burden similar to that of Houston, revealed that the second trimester was the most common time of presentation for infection, but 20% presented postpartum. Postpartum presentations included surgical-incision infection (30%), breast mass (40%), and soft tissue infection of the extremities, buttocks, breast, groin, and abdomen (30%). Fifty-eight percent had multiple sites of infection; HIV and diabetes were predisposing conditions in many of these patients. Seventy-seven percent of the patients were multiparous compared with 40% in the uninfected population. The authors speculated that these women had an inherent risk factor for infection: multiple recent visits to outpatient medical facilities for routine pregnancy care.29 Additional studies are needed to evaluate the association of maternal and infant infection and to determine if both the infant and mother are infected by the same bacterial isolate.
The currently accepted definition of CA infection is difficult to apply to this population, because most neonatal patients were exposed to a hospital environment at birth. Despite the selection of previously healthy neonates with no hospitalizations except birth and no surgeries except circumcision, these patients may actually have nosocomially acquired organisms related to recent hospitalization in an area in which CA MRSA is endemic. Nosocomial transmission may also explain the number of first-time maternal infections. Other possible sources of bacteria include the mother or other family members colonized by S aureus.
Unlike the racial predisposition reported in the TCH population of all age groups, there was no statistical predominance of black patients in the MRSA group. We did observe ethnicity to affect the type of infection. Black patients presented with less-severe skin and soft tissue disease than white non-Hispanic infants. To our knowledge, this finding has not been reported previously in other age groups. Neonatal patients may have overriding risk factors for infection, which dilute some effects of ethnicity and intensify others. MRSA invasive infections were also more common than in the TCH population of all ages (13% vs 4%; P = .001), but invasive MSSA infections were not significantly increased (14% vs 8%).2
Other issues yet to be addressed include the optimal length and route of antibiotic administration, the need for lumbar puncture for isolated groin or localized pustulosis in the absence of fever, and the risk of acquiring S aureus as colonizing flora to both the mother and the infant after a lengthened stay (35 days) in the newborn nursery such as after maternal cesarean section or a brief time (4872 hours) in a level 2 or 3 nursery for tachypnea or hypoglycemia. Because some practitioners are already using outpatient oral and topical therapy, shortened courses of treatment, and limited sepsis evaluations, randomized trials are needed to optimally determine the safety and efficacy of different evaluation and treatment protocols in this vulnerable population.
Current literature is scarce regarding appropriate treatment of CA MRSA in otherwise healthy neonates. One major textbook30 recommends hospitalization, parenteral antibiotic therapy, and drainage as indicated for infections beyond superficial skin lesions. Local conditions such as omphalitis without cellulitis or systemic symptoms may be treated with local care. Cellulitis requires the addition of systemic antibiotics, but the authors do not comment on the choice of oral versus parenteral treatment. Constitutional symptoms necessitate a sepsis evaluation and parenteral therapy. Recommendations for bullous impetigo treatment involve local care of lesions but assert that damaged skin may not effectively absorb topical antibiotics. The authors state that oral antibiotics may be appropriate, but do not address antibiotic choice or duration of therapy. Treatment of pustulosis is not addressed. Other major references do not specifically comment on the treatment of CA MRSA infections in the previously healthy neonate as separate from S aureus infections overall.1,31
The USA300-ST 8 clone carrying the PVL genes continues to predominate in Houston and other regions in the United States causing mostly skin and soft tissue infections.1215 Because most of our circulating isolates carry the PVL genes, we did not find an association of these genes with severe infections or pneumonia in this relatively small population of infants <30 days of age.23,32 As methicillin resistance increases and more clinicians treat children with clindamycin, clindamycin resistance may increase in the isolates from neonates as observed in those from older children.2,18,33
| CONCLUSIONS |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
We thank Linda Lambert for technical assistance and Drs Keiichi Hiramatsu and Teruyo Ito (Juntendo University, Tokyo, Japan) for providing control strains.
| FOOTNOTES |
|---|
Address correspondence to Sheldon L. Kaplan, MD, Texas Children's Hospital, Mail Code 3-2371, 6621 Fannin St, Houston, TX 77030. E-mail: skaplan{at}bcm.tmc.edu
Financial Disclosure: Dr Kaplan received a grant from Pfizer for a Staphylococcus aureus surveillance study. The other authors have indicated they have no financial relationships relevant to this article to disclose.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. M. Fortunov, K. G. Hulten, W. A. Hammerman, E. O. Mason Jr, and S. L. Kaplan Evaluation and Treatment of Community-Acquired Staphylococcus aureus Infections in Term and Late-Preterm Previously Healthy Neonates Pediatrics, November 1, 2007; 120(5): 937 - 945. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. N Johnson, R. P Rapp, C. T Nelson, J. Butler, S. Overman, and R. J Kuhn Characterization of Community-Acquired Staphylococcus aureus Infections in Children Ann. Pharmacother., September 1, 2007; 41(9): 1361 - 1367. [Abstract] [Full Text] [PDF] |
||||
Read all P3Rs
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||