Sustained high prevalence of pneumococcal serotype 1 in paediatric parapneumonic empyema in southern Spain from 2005 to 2009

Transcription

1 ORIGINAL ARTICLE BACTERIOLOGY Sustained high prevalence of pneumococcal serotype 1 in paediatric parapneumonic empyema in southern Spain from 2005 to 2009 I. Obando 1, M. S. Camacho-Lovillo 1, A. Porras 2, M. A. Gandía-González 2, A. Molinos 1, I. Vazquez-Barba 2, B. Morillo-Gutierrez 1, O. W. Neth 1 and D. Tarrago 3 1) Department of Paediatric Infectious Diseases and Immunology, Hospital Infantil Universitario Virgen del Rocío, 2) Department of Microbiology, Hospital Universitario Virgen del Rocío, Sevilla and 3) Microbiological Reference Laboratory, Majadahonda, Madrid, Spain Abstract The epidemiology and microbiological characteristics of paediatric parapneumonic empyema (PPE) before the introduction of the new generation of conjugate pneumococcal vaccines (10-valent and 13-valent) are described. All patients <14 years old admitted to a tertiary paediatric hospital with a diagnosis of PPE were prospectively enrolled from January 2005 to December Pneumococcal serotyping of culture-negative pleural fluid samples was performed using a multiplex real-time PCR assay. Overall, 219 patients had PPE. Incidence rates for PPE remained stable during the study period with a not significant increase in 2009 compared with 2005 (p 0.13), and were temporally associated with higher circulation of pandemic influenza A H1N1 during the last quarter in our population (p 0.001). Pneumococci were detected in 72% of culture-positive and 79% of culture-negative samples. Serotypes were determined in 104 PPE cases. Serotype 1 was the most prevalent serotype identified (42%) followed by serotypes 7F (20%), 3 (16%), 19A (8%) and 5 (7%). Serotype distribution remained similar during all time periods. Pneumococcal serotype 1 remained the most common cause of PPE during the 5- year study. The new generation of pneumococcal conjugate vaccines offers potential serotype coverage of 73% (10-valent) and 99% (13- valent) in the population studied suffering from PPE. Continuous epidemiological and molecular studies are paramount to monitor the impact of pneumococcal vaccines on the epidemiology of PPE. Keywords: Influenza A virus (H1N1), pleural paraneumonic empyema, Streptococcus pneumoniae Original Submission: 7 March 2011; Revised Submission: 14 April 2011; Accepted: 7 July 2011 Editor: J.-L. Mainardi Article published online: 13 July 2011 Clin Microbiol Infect 2012; 18: /j x Corresponding author: I. Obando, Department of Paediatric Infectious Diseases and Immunology, Hospital Infantil Universitario Virgen del Rocío, Londres 98, Sevila, Spain IOSANTAELLA@telefonica.net Introduction Pleural parapneumonic empyema (PPE) often requires surgical intervention; it is associated with prolonged hospitalization and a risk of long-term morbidity [1]. Since 2000 an increase in the incidence of PPE has been observed in our population as well as in other countries with Streptococcus pneumoniae being the most commonly isolated micro-organism [2 5]. The reasons for this epidemiological change are not clearly understood, partly because of the low sensitivity in the diagnosis of this pathology using conventional microbiological culture techniques. In a transverse multicentre study we previously analysed the epidemiology of PPE in Spain using multiple molecular techniques [5]. Pneumococcal infection was proven in 79% of cases with positive microbiological culture and in 84% of pleural fluid samples; out of which nearly 90% were caused by serotypes not included in the conjugated heptavalent pneumococcal vaccine. Serotype 1 was responsible for pneumococcal PPE in nearly 50% and similar observations have been reported in other European countries as well as in certain areas of the USA [2,4]. The introduction of serotype 1 clone ST306, first detected in Spain in 1998, and its rapid expansion from 2000, correlated with an increment of PPE incidence [5,6]. The highly invasive pneumococcal serotype 1 is able to produce epidemic waves of variable duration and Clinical Microbiology and Infection ª2011 European Society of Clinical Microbiology and Infectious Diseases

2 764 Clinical Microbiology and Infection, Volume 18 Number 8, August 2012 CMI magnitude, which can markedly influence the epidemiology of PPE on a geographic and temporal level [7 9]. The new generation of conjugated pneumococcal vaccines (10-valent and 13-valent) has an optimized formulation that theoretically covers the serotypes responsible for PPE during the last years (serotypes included in the seven-valent vaccine: 4, 6B, 9V, 14, 18C, 19F and 23F; additional serotypes are 1, 5 and 7F in the 10-valent plus 3, 6A and 19A in the 13-valent vaccine) [10]. The objective of this study was to describe the epidemiology and microbiological characteristics of PPE in the last years before the introduction of the new generation of conjugated pneumococcal vaccines. Furthermore, we investigated whether there have been changes in the distribution of pneumococcal serotypes causing PPE, in particular serotype 1, to improve the understanding of the current epidemic wave. Finally, the influence of the pandemic influenza H1N1 on PPE was described. Patients and Methods Setting and study population All patients <14 years old admitted to Hospital Universitario Infantil Virgen del Rocio (HUIVR) in Sevilla with a diagnosis of PPE were prospectively enrolled from January 2005 through to December This is a tertiary paediatric hospital that served a referral population of children in Ascertainment and management of PPE cases PPE was defined according to published criteria [2]. Requirements for thoracocentesis, thoracoscopy and antimicrobial therapy were determined according to usual clinical practice and to the discretion of the attending physician. Patients with PPE not requiring thoracocentesis or surgical decortication were excluded, as were cases with pleural fluid analysis consistent with a transudate. Thoracocentesis was performed by paediatric surgeons, except for acutely ill patients who were urgently tapped in the emergency department or the intensive care unit. Written informed consent was obtained from the parent or legal guardian of participating children before thoracocentesis. Pleural fluid specimens were sent for routine microbiological culture and biochemical analysis; the remaining fluid was frozen at )80 C for further molecular testing. Clinical, demographic and outcome data were collected using a standardized case-report form. Microbiological analysis The PPE was defined according to conventional microbiological culture results as (i) culture-positive cases with organism identified from blood, pleural fluid or gastric washing (for Mycobacterium tuberculosis only), and (ii) culture-negative cases when no organisms could be isolated in any of those samples. Pneumococcal isolates from culture-positive cases were further serotyped using the Quellung method, performed by the Pneumococcal Reference Laboratory (Majadahonda). Pneumococcal identification and serotyping of culture-negative cases were performed retrospectively using PCR methodology as previously described [5,11]. Briefly, culture-negative pleural fluid specimens were analysed using a pneumolysin gene PCR assay and a confirmatory real-time PCR assay for the detection of the S. pneumoniae wzg gene. Determination of pneumococcal serotypes was performed using ten groups of real-time PCR typing assays covering 24 serotypes (1, 3, 4, 5, 6A, 6B, 7F/A, 8, 9V/A/N/L, 14, 15B/C, 18C/B, 19A, 19F/B/C, 23F and 23A). In the last quarter of 2009 co-infection with pandemic influenza A virus (H1N1) was investigated using primers and conditions used by the Centers for Disease Control (Atlanta, GA) [12]. Current Clinical and Laboratory Standards Institute (CLSI, Wayne, PA) breakpoints [13] were considered for non-susceptibility (intermediate plus resistance) interpretation: parenteral penicillin ( 4 g/ml) and, erythromycin ( 0.5 g/ml). Statistical analysis Statistical analysis was performed using SPSS for Windows (version 17.0). Descriptive analyses were performed using medians, interquartile ranges (IQR), frequency distributions and percentages. Incidence rates of PPE were calculated for each year using the frequency of Accident and Emergency patients, identified from an administrative database, as denominator. Z-test was used to compare these rates at the 95% confidence level. Comparison of frequency of PPE cases during the last quarter in 2009 with the historical average was performed using one sample Student s t-test. Analysis of categorical variables was carried out using the chi-square test and the Fisher s exact test, as appropriate. All p-values were two-tailed, and the level of significance was set at Results Demographic characteristics and outcome From 2005 to 2009 a total of 219 children diagnosed with PPE were admitted. Median age and duration of clinical symptoms before admission were 44 months (IQR 32 70) and 4 days (IQR 3 6), respectively. All but 22 (10%) of the children received antibiotic therapy before thoracocentesis. Approximately one-third of the patients (35%) were referred from other hospitals. Information regarding PCV7 vaccination status was available in 197 children (90%); of which 91 (46%)

3 CMI Obando et al. Sustained high prevalence of pneumococcal serotype 1 in paediatric parapneumonic empyema 765 had received at least one dose. Video-assisted thoracoscopy was performed in 138/219 patients (63%) of whom 25 (18%) were secondary after clinical failure of thoracocentesis and chest tube insertion. Median length of hospitalization was 9 days (IQR 6 13); 66 patients (30%) required management in the Intensive Care unit. There were no fatalities. Frequency of disease The incidence of PPE in 2005 was 60/ children attending the paediatric Accident + Emergency department; this was followed by a non-significant decline of cases during 2006 and Incidence then increased steadily to a total of 82/ in 2009 (p 0.13 compared with 2005) as seen in Fig. 1. The frequency of PPE cases was significantly higher in the last quarter of 2009 compared with the mean frequency in (29 versus 13.5, range 11 17; p 0.001) with S. pneumoniae being the cause in 89% (Fig. 2). This coincided with the highest prevalence of pandemic influenza A (H1N1). An unusual cluster of six cases was observed in September 2009, compared with a cumulative frequency of one case in during the same month. Microbiological analysis and distribution of pneumococcal serotypes A micro-organism in blood and pleural fluid culture was identified in 47/219 (21%) children out of whom S. pneumoniae was isolated in 34 (72%) patients (Table 1). Other organisms included Streptococcus pyogenes and M. tuberculosis (n = 4 each), Staphylococcus aureus (n = 3) and Haemophilus influenzae non-typable (n = 2). In patients with a negative culture result additional PCR for S. pneumoniae from pleural fluid was performed in 118/172 (69%) and the results were positive in 93/118 (79%). All pneumococcal culture isolates Incidence (cases/ A + E patients) Year Annual frequency A + E patients FIG. 1. Annual incidence of parapneumonic empyema at the HUIVR Sevilla from 2005 to st quarter TABLE 1. Microbiological characteristics of all children (n = 219) with parapneumonic empyema Number of patients (%) Culture positive (n = 47) Streptococcus pneumoniae 34 (72) 93 (79) Streptococcus pyogenes 4 (9) NA Mycobacterium tuberculosis 4 (9) NA Staphylococcus aureus 3 (6) NA Haemophilus influenzae non-typable 2 (4) NA NA, not applicable. 2nd quarter * 2009 vs , p = rd quarter* FIG. 2. Quarterly frequencies of parapneumonic empyema cases between 2005 and *2009 versus , p PCR positive (n = 118) were sensitive to penicillin. Parenteral penicillin and erythromycin non-susceptibility rates were 0% and 3%, respectively. From 127 children diagnosed with S. pneumoniae PPE (34 using culture and 93 using molecular PCR methodology), 104 samples were successfully serotyped. A total of ten pneumococcal serotypes were identified; serotype 1 was isolated in 39 (42%) patients, followed by serotypes 7F (n = 21), 3 (n = 17), 19A (n = 8), 5 (n = 7) and 14 (n = 3); serotypes 9V, 19F, 12A and 6A were isolated from one patient each and this is summarized in Table 2. All the PCV7 strains were isolated from non-immunized patients. There was no significant difference regarding the distribution of those serotypes between January 2005 to June 2007 and July 2007 to December Co-infection with pandemic influenza A virus (H1N1) During the last quarter of 2009, 22/29 patients diagnosed with PPE were also tested for co-infection with pandemic influenza A virus (H1N1) using real-time PCR, 8/22 children were found to be co-infected. In these eight patients S. pneumoniae

4 766 Clinical Microbiology and Infection, Volume 18 Number 8, August 2012 CMI TABLE 2. Distribution of pneumococcal serotypes in children with parapneumonic empyema Number of patients (%) Serotype January 2005 to June 2007 (n = 53) July 2007 to December 2009 (n = 51) Total (n = 104) p-value a 1 23 (43) 21 (41) 44 (42) F 14 (26) 7 (14) 21 (20) (15) 9 (18) 17 (16) A 2 (4) 6 (12) 8 (8) (4) 5 (10) 7 (7) (4) 1 (2) 3 (3) 1 9V 0 1 (2) 1 (1) 1 19F 1 (2) 0 1 (1) 1 12A 0 1 (2) 1 (1) 1 6A 1 (2) 0 1 (1) 1 a For comparison of time periods: 1 January 2005 to 31 June 2007 versus 1 July 2007 to 31 December and methicillin-resistant Staphylococcus aureus (MRSA; USA300 clone) were isolated in six and one case, respectively. Discussion This study investigates the epidemiology of PPE in a tertiary centre with high incidence of this pathology. The use of molecular techniques in addition to conventional microbial cultures resulted in an increased identification of pathogens in this study group. In addition pneumococcal serotypes causing PPE over the last 5 years were determined. These data will provide useful baseline information for future studies of the effectiveness in PPE of the new generation of conjugated pneumococcal vaccines. After a continuous increase of PPE from 1998 to 2005, incidence plateaued subsequently and increased by 38% in 2009 compared with 2005; which was probably related to the circulation of the pandemic influenza A virus (H1N1) [5]. In this study, PCR from pleural fluid was only used to detect S. pneumoniae. The incidence of potential pathogens other than S. pneumoniae was not investigated in culture-negative pleural fluid. Even though some research has used broad-range 16S rdna or multiplex PCR to increase the spectrum of micro-organisms causing PPE, its potential use in our setting is limited because the vast majority of PPE are the result of S. pneumoniae [14,15]. However, this might change in the future, as has been the case in the USA, where community-associated MRSA has become the commonest pathogen causing PPE [16]. To date community-associated MRSA is still an unusual finding in our population group but close surveillance is important because of its epidemic and virulent potential [17,18]. There was a temporal association between the pandemic influenza A virus (H1N1) and PPE in our population group. The frequency of children with PPE was, compared with previous years, significantly higher during the last quarter of 2009, correlating with the highest prevalence of H1N1 cases [19]. An unusual cluster of children with PPE was admitted to the hospital in September 2009 coinciding with increased circulation of the H1N1 virus. During this month, PPE is usually very rare because viral infections are uncommon at this time because school activities in southern Spain start in late September. During the last quarter of 2009, 36% of children with PPE had evidence of co-infection with influenza A virus (H1N1). Similar observations were made in other geographic areas, supporting our findings [20,21]. All pneumococcal isolates were susceptible to parenteral penicillin. This finding coincides with the significant decrease in penicillin non-susceptibility observed in Spain in recent years, probably because of the introduction of PCV7 and a reduction in antibiotic consumption [22]. In addition, highly invasive serotypes tend to be over-represented in PPE cases and these serotypes are not generally associated with antibiotic resistance. Pneumococcal serotypes isolated from PPE in this study are covered by conjugated heptavalent, and the new generation 10-valent and 13-valent pneumococcal vaccines in 5%, 73% and 99%, respectively. Throughout the study serotype 1 was persistently predominant and responsible for over 40% of PPE. Therefore, serotype 1 being highly relevant in the epidemiology of this disease must be considered as a priority candidate for vaccine strategies. However, there are uncertainties regarding the potential impact of the new generation of conjugated vaccines and the current circulating epidemic serotype 1. PPE caused by serotype 1 is more common in older children compared with other serotypes; so shortterm vaccine effectiveness will depend mainly on herd immunity. Colonization rates of serotype 1 and immunization rates in our geographic area are generally low, so one might not expect a significant impact on serotype 1 carriage circulation [5,7]. In this respect only 11% of children with serotype 1

5 CMI Obando et al. Sustained high prevalence of pneumococcal serotype 1 in paediatric parapneumonic empyema 767 infection in this study were aged <2 years (vaccine target population in routine vaccine schedule without catch-up) whereas 52% were >5 years. In addition, independent of vaccine pressure, temporal fluctuations of serotype 1 prevalence can result in a marked decline because of its own clonal dynamics, as was recently the case in Utah [15]. To further understand the above and the impact of vaccines on the global incidence of PPE as well as the distribution of pneumococcal serotypes, we will continue to survey the epidemiology of PPE using molecular technology. There are several limitations to this study. First, data were collected from a single centre and results may differ in other geographic locations. However, the epidemiology of PPE has been consistent over the last years in our country with no differences in the relevance of pneumococcal disease and the distribution of pneumococcal serotypes [5,23,24]. Second, seasonal changes in climatic parameters and, apart from the pandemic influenza A, virus circulation were not studied; factors that have been shown to influence the incidence of pneumococcal PPE [25,26]. Third, molecular serotyping in our study was limited to the detection of 24 serotypes; therefore it cannot be ruled out that some unusual serotypes not included in the screening caused PPE have not been detected [11]. Finally, the distribution of pneumococcal serotypes in uncomplicated pneumonia was not investigated and might have been different [2]. In conclusion, pneumococcal serotype 1 was mainly and persistently responsible for the high incidence of PPE since 2005 in our centre. Theoretically the new generation of conjugate pneumococcal vaccine (10-valent and 13-valent) offers high protection against the serotypes isolated from children with PPE during the study period. Prevalence of PPE was significantly influenced by the pandemic influenza A (H1N1) virus. Ongoing epidemiology surveillance using molecular techniques is necessary to understand the influence of these vaccines on the epidemiology of PPE in the future. Acknowledgements We thank J. A. Leon, M. Anchoriz and D. Falcon for their collaboration in collecting samples from PPE patients and A. Fenoll from CNM (Majadahonda) for performing serotyping in pneumococcal isolates. Funding This work was supported by grant PI 07/0110 from the Fondo de Investigaciones sanitarias. Transparency Declaration Ignacio Obando has received lecture fees from Pfizer and GlaxoSmithKline Biologicals and has participated in an advisory meeting for Pfizer. Other authors declare no conflicts of interests. References 1. Tan TQ, Mason EO Jr, Wald ER et al. Clinical characteristics of children with complicated pneumonia caused by Streptococcus pneumoniae. Pediatrics 2002; 110: Byington CL, Spencer LY, Johnson TA et al. An epidemiological investigation of a sustained high rate of pediatric parapneumonic empyema risk factors and microbiological associations. Clin Infect Dis 2002; 34: Hsieh YC, Hsueh PR, Lee PI, Lee CY, Huang LM. Clinical manifestations and molecular epidemiology of necrotizing pneumonia and empyema caused by Streptococcus pneumoniae in children in Taiwan. Clin Infect Dis 2004; 38: Eastham KM, Freeman R, Kearns AM et al. Clinical features, aetiology and outcome of empyema in children in the north east of England. Thorax 2004; 59: Obando I, Muñoz-Almagro C, Arroyo LA et al. Pediatric parapneumonic empyema, Spain. Emerg Infect Dis 2008; 14: Marimon JM, Ercibengoa M, Alonso M, Zubizarreta M, Pérez-Trallero E. Clonal structure and 21-year evolution of Streptococcus pneumoniae serotype 1 isolates in northern Spain. Clin Microbiol Infect 2009; 15: Henriques Normark B, Kalin M, Ortqvist A et al. Dynamics of penicillin-susceptible clones in invasive pneumococcal disease. J Infect Dis 2001; 184: Hausdorff WP. The roles of pneumococcal serotypes 1 and 5 in paediatric invasive disease. Vaccine 2007; 25: Harboe ZB, Benfield TL, Valentiner-Branth P et al. Temporal trends in invasive pneumococcal disease and pneumococcal serotypes over 7 decades. Clin Infect Dis 2010; 50: Grijalva CG, Pelton SI. A second-generation pneumococcal conjugate vaccine for prevention of pneumococcal diseases in children. Curr Opin Pediatr 2011; 23: Tarragó D, Fenoll A, Sánchez D et al. Identification of pneumococcal serotypes from culture-negative clinical specimens by novel multiplex real-time PCR. Clin Microbiol Infect 2008; 14: Center for Disease Control and prevention. Surveillance for pediatrics deaths associated with 2009 pandemic influenza A (H1N1) virus infection United States, April August MMWR Morb Mortal Wkly Rep 2009; 58: Clinical and Laboratory Standards Institute. Performance standards for antimicrobial susceptibility testing: twenty-first informational supplement M100-S21. Wayne, PA, USA: CLSI, Saglani S, Harris KA, Wallis C, Hartley JC. Empyema: the use of broad range 16S rdna PCR for pathogen detection. Arch Dis Child 2005; 90: Blaschke AJ, Heyrend C, Byington CL et al. Molecular analysis improves pathogen identification and epidemiologic study of pediatric parapneumonic empyema. Pediatr Infect Dis J 2011; 30: Schultz KD, Fan LL, Pinsky J et al. The changing face of pleural empyemas in children: epidemiology and management. Pediatrics 2004; 113:

6 768 Clinical Microbiology and Infection, Volume 18 Number 8, August 2012 CMI 17. Obando I, Valderrabanos ES, Millan JA, Neth OW. Necrotising pneumonia due to influenza A (H1N1) and community-acquired methicillin-resistant Staphylococcus aureus clone USA300: successful management of the first documented paediatric case. Arch Dis Child 2010; 95: Frick MA, Moraga-Llop FA, Bartolomé R et al. Infecciones por Staphylococcus aureus resistente a meticilina adquirido en la comunidad en niños. Enferm Infecc Microbiol Clin 2010; 28: Mayoral Cortés JM, Ruiz Fernández J, Pachón Díaz J et al. Infección por el virus pandémico (H1N1) 2009 en Andalucía. Rev Esp Salud Publica 2010; 84: See H, Blondé R, Mariani P et al. Increased incidence of parapneumonic empyema in children at a French pediatric tertiary care center during the 2009 influenza A (H1N1) virus pandemic. Pediatr Infect Dis J 2010; 29: Ampofo K, Herbener A, Blaschke AJ et al. Association of 2009 pandemic influenza A (H1N1) infection and increased hospitalization with parapneumonic empyema in children in Utah. Pediatr Infect Dis J 2010; 29: Pérez-Trallero E, Martín-Herrero JE, Mazón A et al. Spanish Surveillance Group for Respiratory Pathogens. Antimicrobial resistance among respiratory pathogens in Spain: latest data and changes over 11 years ( to ). Antimicrob Agents Chemother 2010; 54: Deiros Bronte L, Baquero-Artigao F, García-Miguel MJ, Hernández González N, Peña García P, del Castillo Martín F. Derrame pleural paraneumónico: revisión de11años. An Pediatr 2006; 64: Picazo J, Ruiz-Contreras J, Casado-Flores J et al. Laboratory-based, 2- year surveillance of pediatric parapneumonic pneumococcal empyema following heptavalent pneumococcal conjugate vaccine universal vaccination in Madrid. Pediatr Infect Dis J 2011; 30: Tvedebrink T, Lundbye-Christensen S, Thomsen RW, Dethlefsen C, Schønheyder HC. Seasonal changes in climatic parameters and their relationship with the incidence of pneumococcal bacteraemia in Denmark. Clin Microbiol Infect 2008; 14: Ampofo K, Bender J, Sheng X et al. Seasonal invasive pneumococcal disease in children: role of preceding respiratory viral infection. Pediatrics 2008; 122: