Article Text

This article has a correction. Please see:

Download PDFPDF
Genotypic and phenotypic characterisation of invasive Streptococcus pneumoniae isolates from Hungary, and coverage of the conjugate vaccines
  1. Orsolya Dobay1,
  2. Ágnes Ungvári1,
  3. Szilvia Kardos1,
  4. Katalin Kristóf1,2,
  5. Edit Hajdú3,
  6. Judit Szabó4,
  7. Márta Knausz5,
  8. Erzsébet Nagy3,
  9. Ferenc Rozgonyi1,6,
  10. Sebastian G B Amyes7,
  11. Károly Nagy1
  1. 1Institute of Medical Microbiology, Semmelweis University, Budapest, Hungary
  2. 2Division of Clinical Microbiology Diagnostics, Central Diagnostic Laboratory, Semmelweis University, Budapest, Hungary
  3. 3Department of Clinical Microbiology, Faculty of Medicine, University of Szeged, Szeged, Hungary
  4. 4Department of Medical Microbiology, Medical and Health Centre, University of Debrecen, Hungary
  5. 5Aladár Petz County Teaching Hospital, Győr, Hungary
  6. 6Microbiology Laboratory, Department of Dermatology, Venereology and Dermatooncology, Semmelweis University, Budapest, Hungary
  7. 7Centre for Infectious Diseases, University of Edinburgh, Edinburgh, UK
  1. Correspondence to Sebastian G B Amyes, University of Edinburgh, Room SU.312, Centre for Infections Diseases, The Chancellor's Building, 49 Little France Crescent, Edinburgh EH16 4SB, UK; s.g.b.amyes{at}


Background and aims The 7-valent conjugate pneumococcal vaccine (Prevenar) was introduced as a recommended (but not yet obligatory) vaccine in Hungary in April 2009 and there was a sharp increase in the number of children vaccinated. Hence there is an urgent need for in-depth epidemiological data on invasive pneumococci before vaccination becomes widespread. Such a study has never been done before in Hungary.

Methods 144 pneumococcal isolates, obtained from invasive infections or pneumonia, were collected from eight Hungarian diagnostic laboratories between 2000 and 2008. After confirmation of species identity, their susceptibilities to nine antibiotics were determined by Etest and agar dilution method. The serotypes and pulsed-field gel electrophoresis genotypes of the strains were also determined.

Results In this cohort, most of the isolates were from patients at the extreme of life. Only 1.4% of the strains were resistant to penicillin, but nearly 40% were resistant to erythromycin (mainly due to erm(B) gene). Higher incidences of resistance were found in the very young and very old. The most prevalent serotypes in the cohort in descending order were 14, 6A, 6, 6B, 23F, 3, 19F and 11A.

Conclusions Results showed a similar but not identical profile to previously examined strains causing pulmonary infections in Hungary. The serotypes could be correlated to patient groups. Furthermore, there were examples of serotype switching in strains showing identical genotype but different serotype. The study also shows a good coverage by the conjugate vaccines over the invasive pneumococcal strains in Hungary based on the detected serotypes.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Streptococcus pneumoniae (the pneumococcus) is one of the most important human pathogenic bacteria responsible for the death of approximately 1 million children worldwide every year.1 It is the leading cause of community-acquired pneumonia, and also often causes meningitis, sepsis, sinusitis or otitis media. Based on its polysaccharide capsule, 92 different serotypes are known. Currently there are two vaccines against pneumococcal infections: the polysaccharide vaccine, containing 23 serotypes (Pneumovax); and the conjugate vaccines (Prevenar, Synflorix and Prevenar 13); in the latter the capsular polysaccharides conjugated to a carrier protein also elicit an immune response in children <2 years old, the most affected age group.2 3

Prevenar was first introduced in 2000 in the USA and then in Europe.2 Two later conjugate vaccines—a 10-valent vaccine by GlaxoSmithKline (Synflorix),4 and a 13-valent vaccine by Pfizer/Wyeth (Prevenar 13)—were introduced in 2009.5 However, these vaccines contain only 7, 10 and 13 serotypes, respectively (4, 6B, 9V, 14, 18C, 19F, 23F + 1, 5, 7F + 3, 6A, 19A); therefore, it is essential to survey the prevalent serotypes within a country, to establish the vaccine coverage.

Prevenar was recently (April 2009) included as the recommended vaccine in Hungary's national vaccination programme6 for children <2 years old. It was substituted with Prevenar 13 from June 2010. According to the national epidemiological data, a high percentage (82.6%) of small children were vaccinated in this last period,7 so it was judicious to conduct a retrospective epidemiological analysis of invasive pneumococcal infections (IPDs) in Hungary from the pre-vaccination era.

Materials and methods

Bacterial strains

Pneumococcal isolates (144) were obtained from five major cities throughout Hungary—Szeged, Debrecen, Győr, Budapest and Salgótarján—between 2000 and 2008. One isolate was taken from each patient: 48.6% of the strains (n=70) derived from classic sterile sites (eg, haemoculture, liquor or biopsy); and 51.4% were from patients with pneumonia (from bronchoalveolar lavage or ventilation tube (n=37), and from sputum or nasal swab (n=37)). The latter isolates were included as the EU licence of Prevenar now embraces the active immunisation of children <5 years old against pneumonia and acute otitis media.5 The identity of the isolates was confirmed by optochin sensitivity, colony morphology and PCR detection of the autolysin gene (lytA).8 9

Susceptibility testing

Antibiotic susceptibility testing was performed by Etest (AB-Biodisc, Solna, Sweden) and the agar dilution method on Mueller–Hinton (Oxoid, Basingstoke, UK) agar plates supplemented with 5% defibrinated blood, at 37°C in 5% CO2. The antibiotics were purchased from Sigma (Budapest, Hungary) or the appropriate manufacturer. Susceptibility was determined by breakpoints recommended by the BSAC,10 except for clindamycin, where the EUCAST breakpoints were used.11 The control strain was S pneumoniae ATCC 49619.


Serotyping was performed by a combined protocol using conventional serotyping (antisera) and a PCR-based method.12 13


Genotyping was performed by pulsed-field gel electrophoresis (PFGE). Chromosomal DNA was prepared as described previously14: for digestion, SmaI (Sigma, Budapest) was used, and separation was done with the following pulse times: 5–15 s (block 1) and 15–60 s (block 2), for 21 h at 14°C. The PFGE profiles were analysed with the BioNumerics software version 2.5 (Applied Maths, Sint-Martens-Latem, Belgium).

Macrolide resistance determinants

The erm and mef macrolide resistance genes were detected by PCR.15 The mef(A) and mef(E) were differentiated from each other by BamHI digestion and by sequencing.


Patient statistics

The age profile of 135 patients shows a typical twin-peak distribution, most patients being either <5 years or 50–60 years old (figure 1). The least affected age group was those aged 5–40 years, but surprisingly, these patients had higher incidences of pneumonia (figure 2). Among 132 patients, 62.1% were male and 37.9% were female. Males also dominated the very young (<2 years) and old (70–80 years) age groups, with 15/20 and 12/16 cases, respectively.

Figure 1

Age distribution of 135 patients with pneumococcal infections.

Figure 2

Types of infections of 135 patients in the different age categories. Black columns, invasive infections; white columns, pneumonia.

Antibiotic susceptibility results

Although high level penicillin resistance (MIC ≥2 mg/l) was very low, intermediate resistance was high (37.6%) (table 1). Erythromycin resistance was also high (38.5%); however, telithromycin showed excellent efficacy. Resistance to levofloxacin and moxifloxacin was 1.4% and 3.5%, respectively. The strains were fully sensitive to vancomycin and imipenem.

Table 1

Antibiotic susceptibility of the 144 pneumococcal isolates

Reduced penicillin and macrolide susceptibility was significantly higher at the extremities of age. Penicillin insusceptibility was 62.1% and 53.3% in the age groups <5 years and >70 years, respectively, whereas it was only 24.6% in the age group 25–70 years. Similarly, erythromycin insusceptibility was 64.9% and 52.6% among the <5 years and >70 years age groups, respectively (83.3% among those >80 years old), but only 24.6% in the age group 15–70 years.

The pneumococci isolated from strictly sterile sites had higher erythromycin resistance than those from pneumonia (45.7% versus 31.0%), but their penicillin susceptibility was very similar. There was no difference correlating with age or gender.

Macrolide resistant isolates

The 58 erythromycin insusceptible isolates (MIC ≥0.5 mg/l) were tested for the presence of the erm(B), erm(A), erm(TR) and mef genes. In 46 isolates (79.3%) we found the erm(B) gene, in 8 isolates (12.9%) the mef gene and in 2 isolates both genes. The erm+ strains came from diverse serotypes which had high MICs (≥256 mg/l) of both erythromycin and clindamycin. Of the 8 mef+ strains, 7 had mef(E) and only one had mef(A) (figure 3). Also the double positive strains had the mef(E). Although low-level erythromycin resistance and clindamycin sensitivity, the classic M phenotype, is usually linked with the Mef efflux pump, there were nine such isolates harbouring the erm(B) gene. The penicillin susceptibility of the erm+ and mef+ strains was indistinguishable but there was a significant difference in their telithromycin susceptibility, with relative higher MICs in mef+ compared with erm+ isolates (0.36 and 0.04 mg/l, respectively).

Figure 3

Pulsed-field gel electrophoresis patterns of the 8 mef+ strains.


Most (92.4%) of the isolates were successfully serotyped (133 strains) (table 2). The individual serotypes often showed characteristic association with age or gender, invasiveness, antibiotic sensitivity or genotype; for example, the isolates of serotype 3 (n=9) all belonged to the same PFGE clone, and were almost all isolated from invasive samples of older adults, mostly males. These strains were exceptionally sensitive to both penicillin (average MIC=0.014 mg/l) and erythromycin (MIC=0.128 mg/l). Similarly sensitive were the serotype 4 and 7 isolates, also deriving from invasive specimens of adults.

Table 2

Distribution of serotypes of 133 pneumococcal isolates.

The 25 serotype 14 isolates derived mainly from strictly invasive specimens, and were predominantly from males and from children (mean age 8.4 years). The vast majority of the isolates were resistant to macrolides, and half of them had raised penicillin MICs (0.75–2 mg/l). The isolates formed several distinct PFGE clones, which mirrored the differences in the resistance levels (figure 4).

Figure 4

Distinct pulsed-field gel electrophoresis clones among the serotype 14 isolates.

Almost all serogroup 19 strains were macrolide resistant; 19F isolates derived from tracheal or bronchial specimens of adults, while the 19A isolates were mostly from haemoculture. They were genetically very diverse, and only one serotype 19A strain, with the highest penicillin MIC (but only 4 mg/l) belonged to the notorious Hungarian 19A clone.16 Serogroup 6 isolates mainly came from adult tracheal or bronchial specimens. Their penicillin MICs were rather low (0.008–0.5 mg/l), but while the 6B strains were dominantly sensitive to macrolides, the 6A strains had less susceptibility. The 6B isolates were very diverse by PFGE, but there were a few small clones among the 6A isolates.

We have seen a few examples for serotype switching, that is isolates with identical PFGE pattern but different serotypes (figure 5). These happened in both Szeged or Debrecen and the strains were isolated within 1 or 2 years. On two occasions the switch occurred from serotype 14 to another (11A or 6), and on one occasion from serotype 6 to serotype 14. The role of the Prevenar vaccine as the driving force for the switch is unlikely, even in the 14→11A switch.

Figure 5

Examples for serotype switching.


Antibiotic susceptibility

This is the first in-depth epidemiological study of Hungarian invasive pneumococci, though limited information is available from the National Centre for Epidemiology17 and the European Antimicrobial Resistance Surveillance System database.18 The last available data are from 2008 and 2009, respectively.

Our current results show low penicillin resistance (R, 1–2%) but high intermediate level rates (I, 37–38%), and very high erythromycin resistance (37–41%). We observed similar results in our previous study of non-invasive pneumococci.9 These data indicate that resistance in Hungary to these two drugs is relatively stable and not increasing. Among the macrolide resistant isolates, erm(B) gene is still the leading mechanism.

The situation in the case of the fluoroquinolones is somewhat different. In this study we detected nearly three times higher resistance to moxifloxacin than to levofloxacin, but this can be explained with the BSAC resistance breakpoints (R >2 mg/l for levofloxacin and R >0.5 mg/l for moxifloxacin).10

In summary, there was a strong correlation between age and resistance, with higher penicillin and macrolide resistance rates in the very young and rather old age groups. It is well known that the burden of pneumococcal diseases is greater at the extremities of life, and especially in the young, the incidences of resistance are higher. It is interesting to note from this study that some serotypes are frequently represented in isolates from patients under the age of 5 and are found again in isolates from patients older than 60. In particular, the common serotype 14 fell into this category as it is almost unknown in patients between the ages of 5 and 60. Only a small difference in erythromycin resistance was observed between the strictly invasive and pneumonia isolates, otherwise they had similar susceptibility results.

Serotypes and vaccine coverage

There have been two previous studies in Hungary, which provided pneumococcal serotyping data. In 2003, we determined the serotypes of 112 non-invasive isolates.9 In another study in 2009, serotypes of 82 IPD isolates, mostly from children, were determined.19 Serotypes 3, 14 and 6A were found to be the most frequent. We can draw the following general conclusions: (i) serogroup 9 seems to be much rarer among the invasive isolates; (ii) serotype 3 was very frequent among the invasive isolates in both studies; and (iii) serotype 19F was never found in haemoculture or liquor.

The most important question is how Prevenar, the other new conjugate vaccines and Pneumovax compare against the severe pneumococcal diseases in Hungary, based on the serotyping data. According to our data, the additional serotypes covered by Synflorix (1, 5 and 7F) were very rare among Hungarian isolates, whereas the other three serotypes represented in Prevenar 13 (3, 6A and 19A) were important. Although in contrast to earlier reports from Hungary,20 but in agreement with recent findings,9 15 19 serotype 19A does not rank among the most frequent types in Hungary, it comprises a successful clone with higher resistance to penicillin and macrolides. Serotype 6A is, on the other hand, one of the most prevalent types in the country. Although Prevenar contains 6B, and not 6A, the presence of 6B provides sufficient protection against infections caused by 6A.21 This cross-protection does not exist between 19F and 19A,21 therefore inclusion of 19A appears very important. Taking these facts into account, we have calculated the vaccine coverage rates over the isolates of this study (table 3). Prevenar would cover 71.4% of all isolates. While it has a better coverage over the lower respiratory tract isolates than the strictly invasive ones, this difference disappears in the case of Prevenar 13, due to invasive infections caused by serotype 3. The 23 serotypes included in Pneumovax would naturally provide a much higher coverage (97.7%), as only two rare serotypes found in this study (13 and 42) are not part of this vaccine. Based on our results, we can conclude that use of the conjugate vaccines could provide good protection, and the 13-valent vaccine in particular, has a high coverage in Hungary.

Table 3

Coverage (%) of the three conjugate vaccines over 133 isolates of this study

Take-home messages

  • The pneumococcal conjugate vaccines provide a good coverage in Hungary, therefore their inclusion in the routine vaccination scheme is advisable.

  • The penicillin and erythromycin insusceptibility is significantly higher in the very young and very old age groups than in the adult population.

  • Serotype 3 is very prevalent in invasive pneumococcal infections, but rare in pneumonia.

  • Serotype 19F was never found in haemoculture or liquor, while 19A was mostly isolated from IPDs.

  • Serotype 19A does not rank among the frequent types in Hungary.


These results were presented in part at the 18th European Congress of Clinical Microbiology and Infectious Diseases (ECCMID), Barcelona, Spain, April 2008 (poster P1710).We wish to thank Drs Gyula Kispál, Nóra Szabó, Dimitrinka Boriszova and Károly Csiszár, for providing strains for this study. We would like to thank Mónika Kapás for her assistance in the laboratory work.



  • Funding This work was financially supported by the Hungarian National Scientific Research Fund (OTKA), grant numbers F61665 and PD75660.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles

  • PostScript
    BMJ Publishing Group Ltd and Association of Clinical Pathologists