REPORT. Antimicrobial activity of various antibiotics against non-invasive clinical isolates of Streptococcus pneumoniae

Transcription

1 212 REPORT In vitro study to investigate the antimicrobial activity of various antibiotics against non-invasive clinical isolates of Streptococcus pneumoniae collected in Belgium during winter Antimicrobial activity of various antibiotics against non-invasive clinical isolates of Streptococcus pneumoniae

2 REPORT In vitro study to investigate the antimicrobial activity of various antibiotics against non-invasive clinical isolates of Streptococcus pneumoniae collected in Belgium during winter (SP212). INVESTIGATOR R. Vanhoof Scientific Institute of Public Health Unit of Antibiotic Research Service Bacterial Diseases Operational Direction Communicable and Infectious Diseases Engelandstraat Brussels. The 29 th of September 212 ISSN: D/212/255/47

3 2

4 CONTENTS INTRODUCTION... 5 MATERIAL AND METHODS... 7 PARTICIPATING LABORATORIES... 8 ANTIMICROBIAL AGENTS SUSCEPTIBILITY TESTING... 8 INTERPRETATION OF THE RESULTS... 9 CAPSULAR TYPING... 9 STATISTICAL ANALYSIS RESULTS POPULATION PARAMETERS...14 Isolates Age distribution Sample type of the isolates Type and Gender of the isolates Geographic origin of the isolates INTRINSIC ANTIBIOTIC ACTIVITY ANTIBIOTIC RESISTANCE RATES ANTIBIOTIC RESISTANCE RATES AND POPULATION PARAMETERS Resistance and Age Resistance and Sampling site Resistance and Geographic origin Resistance and Type of isolate and Gender Resistance Phenotypes Penicillin Resistance and capsular types COMPARISON WITH FORMER SURVEYS EVOLUTION OF RESISTANCE RATES General overview of the resistance rate Beta-lactam resistance rates Fluoroquinolones resistance rates Macrolide and Tetracycline resistance rates EVOLUTION OF MIC DISTRIBUTIONS... 4 Evolution of Beta-lactam MIC distributions Evolution of Fluoroquinolone MIC distributions Evolution of MDR Cross Resistance

5 4

6 INTRODUCTION. 5

7 S. pneumoniae, with its high degree of morbidity and its considerable rate of mortality, is one of the most commonly identified pathogens both in community and hospital acquired infections. It is not only the causative agent for upper respiratory tract infections but also for a number of important invasive infections such as septicaemia, pneumonia and meningitis. The appearance of resistant strains, in which both the de novo acquisition of new genetic material and the clonal spread of resistant isolates are implied, can be an incriminating factor in the treatment and outcome of pneumococcal disease. In 1967, the first penicillin-insusceptible S. pneumoniae strain was reported from Australia and since then antibiotic resistance in this micro-organism became a problem of paramount importance. High rates of insusceptibility to penicillin and other related and unrelated compounds have been reported world-wide. Over the years, penicillin insusceptible isolates became more and more concomitantly resistant to other classes of antimicrobials. In Belgium, the first two penicillin resistant isolates were reported by Vanhoof et al in The situation had been evolving less spectacular but since the beginning of the nineties both the Antibiotic research Unit of the Institute Pasteur of Brussels (actually part of the Institute of Public health) and the National Reference Centre have been reporting a slow but steadily increase in penicillin insusceptibility in clinical isolates with a peak in 21. An important decrease in resistance was noted in the following years. In these former surveillance studies, we reported a level of penicillin insusceptibility based on breakpoints proposed by EUCAST of 12.1%, 12.2%, 14.8%, 19.8%, 15.1%, 14.6%, 15.2%, 12.8%, 12.1 %, 11.6%, 1.2%, 9.2% and 1.1% for winter 1995, 1997, 1999, 21, 23, 24, 25, 26, 27, 28, 29, 21 and 211 respectively. The worldwide reported increase of antibiotic resistance among S. pneumoniae together with the concomitant development of co-resistance between various unrelated classes of antimicrobials constitutes a problem of paramount importance. However, the clinical relevance of the impact of resistance on the clinical outcome remains a controversial topic. Furthermore, the epidemiology of antibiotic resistance can be influenced by various factors and important variations can even be found in restricted geographic areas due to differences in antibiotic policies, secular changes and clonal shifts in the bacterial population, demographic and geographic parameters. These findings underline the necessity for continuous national and international surveillance of resistance. Therefore we organized this 13 th collaborative surveillance study, to monitor possible changes in antibiotic resistance in non-invasive clinical isolates of S. pneumoniae collected by 15 participating centres. 6

8 MATERIAL AND METHODS. 7

9 Participating Laboratories. This study was conducted in collaboration with 15 selected clinical centres in Belgium. The aim of this eight annual surveillance study was to obtain information on the level and evolution of antibiotic resistance in Belgian isolates of S. pneumoniae. The isolates of Streptococcus pneumoniae prospectively collected were obtained from the following centres: (A = laboratory code) Hôpital de la Citadelle, Liège. (Dr. M. Carpentier), (B) Laboratoire Cebiodi, Bruxelles (Dr. B. Mulongo), (C) Clinique Universitaire de Mont- Godinne, Yvoir (Dr. Y. Glupczynski), (D) Clinique St. Joseph, Arlon (Dr. J.S. Goffinet), (E) C.H.U. André Vésale, Montignies-le-Tilleul (Dr. D. Govaerts), (F) Hôpital Princesse Paola, Marche-en-Famenne (Dr. Ph. Lefèvre), (G) Medisch Centrum Huisartsen, Leuven (Apoth. M. Lontie), (H) Virga-Jesseziekenhuis, Hasselt (Dr. R. Cartuyvels), (I) Hôpital de Jolimont, Haine-St.-Paul (Dr. F. Meunier), (J) CHR Hôpital de Warquignies, Boussu (Dr. I. Philippart), (K) H. Hartziekenhuis, Roeselare (Dr. E. De Laere), (L) A.Z. Stuivenberg, Antwerpen ( Dr. K. Camps), (M) A.Z. St. Jan, Brugge (Dr. E. Nulens), (N) A.Z. Jan Palfijn, Gent (Dr. L. Ide) and (O) Imeldaziekenhuis, Bonheiden (Dr. J. Frans). Antimicrobial agents. The following antibiotics were tested in the study and were provided as laboratory preparations with known potency: clavulanic acid, cefuroxime, ceftazidime (GlaxoSmithKline), cefepime (Bristol Myers Squibb), cefotaxime, levofloxacin, ofloxacin and telithromycin (Sanofi Aventis), ciprofloxacin and moxifloxacin (Bayer). Amoxicillin, ampicillin, azithromycine, cefaclor, clindamycine, erythromycin, imipenem, penicillin G and tetracycline were obtained from a commercial source (Sigma). Amoxicillin/clavulanic acid was tested in a 2:1 ratio. All antibiotics were tested for 16 serial twofold dilutions (.1 32 µg/ml). Susceptibility testing. The Minimal Inhibitory Concentrations (MICs) were determined by broth microdilution as recommended by the CLSI. All isolates underwent a slide agglutination (Slidex pneumo Kit TM, BioMérieux) and an Optochine test (OPTO-F, BioMérieux) before 8

10 MIC testing. All the isolates were also tested for the LytA gene by PCR. S. pneumoniae ATCC 49619, S. pneumoniae TPN881 (internal control isolate) and Staphylococcus aureus NCTC (ß-lactamase positive to validate the clavulanate component of amoxicillin/clavulanate) were included as quality control organisms in each series. Interpretation of the results. Interpretation of the results is based on breakpoints provided by the EUCAST ( (See table MM1 for EUCAST breakpoints). Capsular typing. The penicillin insusceptible isolates were typed by the National Reference Centre by using the Quellung reaction with sera from the Staten Serum institute (Copenhagen, Denmark). Statistical analysis. The Chi-square test, with or without Yates correction, for two independent samples was used for the statistical evaluation of the results. The level of significance was set at.5. 9

11 Table MM1: EUCAST Breakpoints: S. pneumoniae Compound S I R Beta-Lactam Penicillin Ampicillin Amoxicillin Amox/Clav Cefaclor Cefuroxime Cefuroxime-axetil Cefotaxime (noninv) Cefepime Ceftazidime Cefonicid Cefprozil Cefpodoxime Ertapenem Imipenem 2-4 Fluoroquinolones Ciprofloxacin Gemifloxacin Grepafloxacin Levofloxacin 2-4 Moxifloxacin.5-1 Ofloxacin Trovafloxacin MLS/Ketolides Azithromycin Clarithromycin Clindamycin.5-1 Erythromycin Miokamycin Telithromycin Tetracyclines Tetracycline

12 ABBREVIATIONS USED FOR ANTIBIOTICS COMPOUND Abbr Abbr COMPOUND Amoxicillin AMX AMC Amoxicillin/Clavulanate Amoxicillin/Clavulanate AMC AMP Ampicillin Ampicillin AMP AMX Amoxicillin Azithromycin AZI AZI Azithromycin Cefaclor CFC CFC Cefaclor Cefepime CPM CPM Cefepime Cefotaxime CTX CRX Cefuroxime Ceftazidime CTZ CRXax Cefuroxime-axetil Cefuroxime CRX CTX Cefotaxime Cefuroxime-axetil CRXax CTZ Ceftazidime Ciprofloxacin CIP CIP Ciprofloxacin Clindamycin CLI CLI Clindamycin Erythromycin ERY ERY Erythromycin Imipenem IMI IMI Imipenem Levofloxacin LEV LEV Levofloxacin Moxifloxacin MOX MOX Moxifloxacin Ofloxacin OFL OFL Ofloxacin Penicillin G PEN PEN Penicillin G Telithromycin TEL TEL Telithromycin Tetracycline TET TET Tetracycline 11

13 12

14 RESULTS. 13

15 Population parameters Isolates. In total, 351 documented isolates of S. pneumoniae were included in the study for further analysis. Age distribution. Twelve point three percent (43/351) of the isolates for which age was available were from children (age 15 years) with 35/43 or 81.4 % from children under 5 years of age, while 87.7 % (38/351) were from adults with 28/38 or 67.5 % from adults with age 6 years. Age showed a bimodal distribution (Fig. 1) with a first peak between and 5 years (1.%) and a second broad peak between 56 and 85 year (58.7 %). Fig. 1: Age distribution (SP212) The mean age of the study population was 57. years. The mean age per collection centre (Fig. 2) varied from 37.1 years (Centre B: Laboratoire Cebiodi, Bruxelles) to 68.7 years (Centre L: A.Z. Stuivenberg, Antwerpen). The mean age differed significantly between the North and South (62.5 y versus 54.7 y;.1>p>.1), the North and Brussels (62.5 y versus 37.7y; P<.1) and the South and Brussels (54.7y versus 37.7y;.1>P>.1) (Fig. 3). The 14

16 percentage of isolates obtained from children was significantly lower in the North (5.%) than in the South (17.%) and Brussels (26.9%) (.1>P>.1 and P<.1 respectively) (Fig. 4). Fig. 2: Mean age per laboratory (SP212) A B C D E F G H I J K L M N O Mean Fig.3: Mean age by region (SP212) ****N ***S ***N BRUSS NORTH SOUTH NAT 15

17 Fig.4: Percentage of Children by region (SP212) % ***S ****B BRUSS NORTH SOUTH NATIONAL Fig. 5: Evolution of mean age and % of children ( 5y) %Ch 5y 26 % mean age Figure 5 depicts the evolution of the percentage of children 5 years and the evolution of the mean age. A high percentage of children 5 years can have an impact on the resistance level of certain antibiotics. We noticed a significant decrease in the proportion of children (- 15y) from 26 (25.6%) to 212 (12.3%). In the same period the presence of adults 6y increased from 49.7% to 59.3% (mean age from 47.9y vs. 57y). 16

18 Sample type of the isolates. Isolates from sputum represented 83.2 % (292/351) of the specimens, 1. % (35/351) were from nasal swab, 3.1 % (11/351) from throat, 2.8 % (1/351) from sinus and.9 % (3/351) from pus. Overall, 83.2 % the isolates were from lower respiratory tract (LRT) specimens and 16.8 % from upper respiratory tract (URT). Isolates from sputum were significantly more present in patients of the age group 6 years (67.1 %) when compared to the other age groups: -5 years (1.4 %), 6-15 years (1.7 %) and years (29.8 %) (P<.1). Isolates from sputum were also more present in the age group years than in the age groups -5 years and 6-15 years (P<.1). Overall, this means that Sputum samples were significantly more present in adults than in children (96.9 % versus 3.1 %; P <.1). Consequently, isolates from upper respiratory tract origin were significantly more present in children than in adults (74. % versus 26. %; P<.1). Type and Gender of the isolates. Isolates from hospitalised patients represented 7.9 % (249/351) of the isolates while 28.8 % (11/351) were from ambulatory patients. Sixty four point one percent (225/351) of the isolates were from male patients. Geographic origin of the isolates. Isolates from the Southern part of the country represented 46.7 % (164/351) while 45.9 % (161/351) and 7.4 % (26/351) of the isolates came from patients from the Northern part and Brussels respectively. Provincial distribution is shown in Fig. 6. Fig.6: Provincial distribution of the isolates (SP212) % AP HA WV LG NA VB LX BS LI OV BW 17

19 (AP: Antwerpen, HA: Hainaut,, WV: West-Vlaanderen, LG: Liège,, NA: Namur, VB: Vlaams Brabant, LX: Luxembourg; BS: Brussels, LI: Limburg, OV: Oost Vlaanderen, BW: Brabant Wallon,) 18

20 Intrinsic antibiotic activity. The highest intrinsic activity on a weight basis was found for imipenem (MIC 5, or MIC for 5% of the isolates tested, of.8 µg/ml) followed by penicillin G, amoxicillin, amoxicillin/clavulanic acid and telithromycin (MIC 5.15 µg/ml), ampicillin cefotaxime, ceftazidime and cefepime, (MIC 5.3 µg/ml), cefuroxime, cefuroxime-axetil, moxifloxacin, erythromycin and clindamycin (MIC 5.6 µg/ml), azithromycin (MIC 5,12 µg/ml), tetracycline (MIC 5.25 µg/ml), cefaclor (MIC 5.5 µg/ml), levofloxacin and ciprofloxacin (MIC 5 1. µg/ml) and ofloxacin (MIC 5 2. µg/ml) (Table 1). The highest level of susceptibility following EUCAST breakpoints was found for imipenem (1%) followed by moxifloxacin (99.7%), telithromycine (98.9%), cefepime (97.2%), cefotaxime and levofloxacin (96.9%), amoxicillin, amoxicillin/clavulanic acid (92.9%), ampicillin (88.9%), cefuroxime (87.5%), penicillin (85.5%), clindamycine (84.9%) and cefuroxime-axetil (82.3%). The lowest degree of susceptibility was found for tetracycline (7.7%), erythromycin and azithromycine (68.4%). The susceptibility rate for cefaclor, ciprofloxacin and ofloxacin was 2.6%,.3% and % respectively (Table 2). There are no separate EUCAST breakpoints for ceftazidime. Using the EUCAST breakpoints of cefotaxime, the susceptibility level of ceftazidime is 9.6% 19

21 Table 1 : Susceptibility of 351 isolates of S. PNEUMONIAE to various antimicrobial agents. (SP212) ( 5/9%) Antibiotic MIC distribution (µg/ml) Penicillin Ampicillin Amoxicillin Amoxicillin/cla vulanate Cefaclor Cefuroxime Cefuroxime-ax Cefotaxime Ceftazidime Cefepime Imipenem Ciprofloxacin Levofloxacin Moxifloxacin Ofloxacin Erythromycin Azithromycin Telithromycin Clindamycin Tetracycline

22 Antibiotic resistance rates. Decreased susceptibility (I+R) to penicillin was 14.5 % [13.7 % intermediate (I =.12-2 µg/ml) and.8 % high-level (R 4µg/mL)], ampicillin 11.1 % [8.3 % I (= 1-2 µg/ml) and 2.8 % R ( 4µg/mL)], amoxicillin ± clavulanic acid 7.1 % [6.8 % I (= 1 2 µg/ml) and.3 % R ( 4 µg/ml)], cefaclor 97.4 % [7.4 % I (=.6.5 µg/ml) and 27. % R ( 1 µg/ml)], cefuroxime 12.5 % [.6 % I (= 1 µg/ml) and 11.9 % R ( 2 µg/ml)], cefuroximeaxetil 17.7 % [5.1 % I (=.5 µg/ml) and 12.6 % R ( 1 µg/ml)], cefotaxime 3.1 % [2.8 % I (= 1-2 µg/ml) and.3 % (R 4 µg/ml)], cefepime 2.8 % [2.5 % I (= 2 µg/ml) and.3 % (R 4 µg/ml)], imipenem % [R 4 µg/ml], ciprofloxacin 99.7 % [91.7 % I (=.2-2 µg/ml) and 8. % R ( 4 µg/ml)], levofloxacin 3.1 % [R ( 4 µg/ml)], moxifloxacin.3 % [R 1 µg/ml], ofloxacin 1 % [98.6 % I (=.25-4 µg/ml) and 1.4 % R ( 8 µg/ml)], erythromycin 31.6 % [.9 % I (=.5 µg/ml) and 3.7 % R ( 1 µg/ml)], azithromycin 31.6 % [.9 % I (=.5 µg/ml) and 3.7 % R ( 1 µg/ml)], telithromycin 1.1 % [.9 % I (=.5 µg/ml) and.3 % (R 1 µg/ml)], clindamycin 15.1 % [R 1 µg/ml)] and tetracycline 29.3 % [3.1 % I (= 2 µg/ml) and % R ( 4 µg/ml)] (Table 2). There are no separate breakpoints for ceftazidime. Using the breakpoints for cefotaxime, we have found the following resistance rates: 9.4 % with 8.8 % I and.6 % R. 21

23 Table 2: Susceptibility rates following EUCAST of 351 non-invasive isolates of S. pneumoniae (SP212) Susceptibility Rates following EUCAST (in %) 1 Antibiotics Susceptible Non Intermediate Resistant susceptible Penicillin Ampicillin Amoxicillin Amoxicillin/clavulanate Cefaclor Cefuroxime Cefuroxime-axetil Cefotaxime Ceftazidime NA 3 NA 3 NA 3 NA 3 Cefepime Imipenem 1 Ciprofloxacin Levofloxacin Moxifloxacin Ofloxacin Erythromycin Azithromycin Telithromycin Clindamycin Tetracycline Minor differences in % are due to rounding off 2 Cefuroxime-axetil: oral form of Cefuroxime 3 NA: breakpoints not available 22

24 Table 3: MIC values of ß-Lactams for various penicillin susceptibility categories following EUCAST (212). ß-Lactam PEN S (n = 3) PEN I/R (n = 51) MIC5 MIC9 MIC5 MIC9 % S Penicillin Ampicillin Amoxicillin Amoxicillin/Clavulanate Cefaclor Cefuroxime Cefuroxime-axetil Cefotaxime Ceftazidime NA Cefepime Imipenem Oral form of cefuroxime MICs of all ß-lactams rose with those of penicillin (Table 3). Overall, ceftazidime and was equally active against the penicillin insusceptible isolates. Imipenem was generally three doubling dilutions more potent on a weight basis, while amoxicillin, amoxicillin/clavulanate, cefotaxime and cefepime were generally one doubling dilution more potent. Ampicillin was one dilution less active. Cefuroxime and Cefuroxime-axetil were two dilutions less active while cefaclor was six dilutions less active. Cross-resistance between penicillin and the other ß-lactams was not complete. The results indicate that 8.4 % of the penicillin-insusceptible isolates remained susceptible to cefepime, 78.4 % to cefotaxime and 51. % to amoxicillin ± clavulanate. All penicillin-insusceptible isolates remained fully susceptible to imipenem. 23

25 Antibiotic resistance rates and population parameters. Resistance and Age. Isolates showing resistance to an antibiotic (IR-isolates) were more present in children (48.8 %) than in the adults (38. %) These differences between the various ages groups as far as the presence of IR-isolates are concerned were not significant. Data are summarized in Table 4 and Figure 7. Resistance to penicillin, erythromycin and tetracycline were higher in children than in adults though the only significant difference was found for tetracycline (44.2 % versus 27.3 %;.5>P>.2). Penicillin resistance in the age group -5 year (25.7 %) was also significantly higher than in adults (13.3 %;.5>P>.2). Differences between age group 6-15 y and other age groups were found for penicillin (age group 16-59y, 12.5 % versus 18. %;.5>P>.2), erythromycin (age group 6y, 37.1 % versus 27.4%,.2>P>.1) and tetracycline (age group 6y, 75. % versus 24. %;.1>P>.1). Table 4: % of isolates in different age groups with resistance to indicated antibiotics (SP212) age Pen Lev Ery Tet IR S Children Adults total

26 Fig.7: Resistance Rates: Children vs Adults (SP212) % Pen Lev Ery Tet* IR Children Adults Resistance and Sampling site. Table 5: Resistance rates (%) in URT and LRT isolates (212) Pen Lev Ery Tet IR S URT LRT TOT Fig.8: Resistance rates in URT and LRT (212) % URT LRT Pen Lev Ery Tet IR 25

27 There were only minor differences in resistance rates between isolates from upper respiratory tract isolates and those from lower respiratory tract isolates. The difference were statistically not significant. Resistance and Geographic origin. Fig.9: Resistance rates by region (SP212) % North South Bruss Pen Lev* Ery*** Tet IR* In general, the presence of isolates with decreased antibiotic susceptibility (IR isolates) revealed to be significantly lower in the North than in the South (34.2 % versus 45.1 %;.5>P>.2) but did not differ significantly from the rate found in Brussels (34.6 %). Concerning the resistance rates of the individual compounds, significant differences were found for levofloxacin and erythromycin. For levofloxacin and erythromycin the difference concerned North versus South: levofloxacin = 6.2 % versus.6 %;.5>P>.2, erythromycin = 24.8 % versus 39. %;.1>P>.1). These differences are probably highly influenced by the number of isolates obtained from the different regions and the presence of isolates from children in the different regions. 26

28 Resistance and Type of isolate and Gender. Table 6 : Resistance rates (%) following gender and type of isolate (212) Pen Lev Ery Tet IR S Gender Male Female Type AMB HOSP AMB:Ambulatory; HOS:Hospitalized. Table 6 summarizes the resistance rates in relation to gender and type of isolate. No significant differences were found for gender. A significant difference was found between hospitalized and ambulatory patients for penicillin (21.8 % versus 11.6%;.2>P>.1) and tetracycline (37.6 % versus 26.1 %;.5>P>.2). Resistance Phenotypes. The most common resistance phenotypes (Table 7) were insusceptibility to Erythromycin-Tetracycline (14. %), Penicillin-Erythromycin-Tetracycline (1. %) followed by isolated insusceptibility to Erythromycin (5.1%) and Tetracycline (4. %). Insusceptibility to one compound was present in 12.5 % of the isolates. Two- and threefold resistance was found in 15.4 % and 1.3 % of the isolates respectively. Fourfold resistance was present in 1.1 % of the isolates (Fig. 1). 27

29 Table 7: Distribution of the Penicillin-Levofloxacine-Erythromycin- Tetracycline susceptibility phenotypes (212). Penicillin-Levofloxacin-Erythromycin- Number (%) Tetracycline- phenotype Susceptible 213 (6.7) Ery-Tet 49 (14.) Pen-Ery-Tet 35 (1.) Ery 18 (5.1) Tet 14 (4.) Pen 7 (2.) Lev 5 (1.4) Pen-Ery 4 (1.1) Pen-Lev-Ery-Tet 4 (1.1) Pen-Lev 1 (.3) Lev-Ery-Tet 1 (.3) Fig. 1: Distribution of Resistance Phenotypes (SP212) % S 1AB 2AB 3AB 4AB 1.1 Table 8: Comparison of mono resistance and multi drug resistance (211) PEN LEV ERY TET IR Mono Resistance Multi Drug Resistance

30 Multidrug resistance (MDR or Resistance 2 compounds) was present in 26.8 % of the isolates. Table 8 provides data on the comparison of mono-resistance to a particular compound and MDR with that particular compound. For example, in the actual population of strains, 2. % of the penicillin insusceptible isolates show mono-resistance to his compound, while 12.5 % of the penicillin insusceptible isolates harbour a combined resistance with other compounds. Fig.11: MDR in the different regions (SP212) % North South Bruss MDR was in general more present in isolates obtained from the South than in isolates from the North or from Brussels, though these differences were statistically not significant (Fig. 11). Penicillin Resistance and capsular types. The most important capsular types in penicillin-insusceptible isolates were capsular types 19A (36.3 %) and 15A (23.5 %). 29

31 Table 9: Capsular types found in Penicillin Non-susceptible isolates (SP212) Capsular Type 19A* 15A 14* 35B 6A* 11A 15B 19F* 29 6B* N % *Isolates present in the PCV-13 vaccine Interestingly, 56.9% of the penicillin-insusceptible isolates (6A, 6B, 14, 19A, 19F) belonged to PCV-13 vaccine isolates. These isolates were more present in adults than in children (79.3% versus 2.7%; not significant) and were almost equally present in the North (44.8%) and South (48.3%). The PCV-13 isolates had a significant higher geometric mean MIC than the other capsular types (1.27 versus.622;.1>p>.1). 3

32 COMPARISON WITH FORMER SURVEYS. 31

33 Evolution of resistance rates. General overview of the resistance rate Table 1 : Non-susceptibility rates (I+R/R) obtained in the various surveys following EUCAST breakpoints Compounds PEN 12.1/ / / / / / / / / 11.6/ 1.2/.5 9.2/.5 1.1/ /.8 AMP 7.1/ / / / / / / / / /.2 8.2/ /.5 8.4/ /2.8 AMX 6.4/ 8.5/ 12.3/ /2. 8.4/2.6 8./ /.9 6.5/.5 6.5/.2 5.6/ 4.6/.2 3.5/.3 4.6/.3 7.1/.3 AMC 7.1/ 8.5/ 11./ / / / /.9 6.3/.5 6.5/.2 5.6/ 4.6/.2 3.5/.3 4.6/.3 7.1/.3 CFC 98.6/22.9 1/ /29. 1/ / / / / / / / / / /27. CRX 7.9/ / / / / / / / / / / / / /11.9 CRXax 8.6/ / / / / / / / / / / / / /12.6 CTX 6.4/ 7.9/ 12.9/ / / / /.9 6.5/.5 6.1/ 6,5/ 8./.5 5.4/.3 4.3/ 3.1/.3 CTZ* NT NT NT NT NT 11.1/ / / / 7.1/ 9.4/.5 5.9/.3 8.4/ 9.4/.6 CPM NT NT NT NT NT 6.4/ /.9 2.1/ 1.9/ 2./ 2.4/.5 1.6/.3 3.5/ 2.8/.3 IMI** CIP 96.4/ 1/3. 1/ / / / / / / / / / / /8. LEV** MOX** OFL 1/ 1/2.4 1/2.6 1/ / / / / / /.4 1/ / /2.2 1/1.4 ERY 2./ / / / / / / / / / / / / /3.7 AZI NT NT NT 3.4/ ,6/ / / / / / / / /3.7 TEL NT NT NT NT 4.1/ /.7 1.8/.2.9/.5 1.9/1..7/.2 2.4/ /.8 1.1/.5 1.1/.3 CLI** NT NT TET 27.1/ / / / / / / / / / / / / /26.2 *There are no separate EUCAST breakpoints for CTZ. For comparison, the CTX breakpoints are used **Only one figure (total non-susceptibility) is given for compounds for which no resistance is found (IMI) and for compounds without an intermediate breakpoint (LEV, MOX, CLI) Table 1 summarizes the resistance rates for the different antimicrobial compounds tested in the various surveillance studies conducted between 1995 and 212. Data are based on EUCAST breakpoints. There are no separate breakpoints for Ceftazidime. The ceftazidime data presented in this table are obtained by using the breakpoints for cefotaxime. 32

34 Beta-lactam resistance rates. The graphic representation of the evolution of resistance to beta-lactam antibiotics is shown in Fig. 12. Penicillin G peaked in 21 (19.8 %) and there is a clear downwards tendency in resistance to notice in the period after 23 to 21. A discrete increase in non susceptibility was noted in the 211 survey (1.1%). The actual resistance rate is 14.5%. The increase in Penicillin insusceptibility from 9.2% in 21 to 14.5% in 212 was statistically significant (.5>P>.2). Amoxicillin and cefuroxime followed this evolution. The increase in amoxicillin non-susceptibility in the same period (21-212) from 3.5% to 7.1% was also significant (.1>P>.1). The cefuroxime increase from 7.8% to, 12.5% was not significant (the Chi² value of is near the threshold of significance for P=.5). Resistance to Imipenem is not found. Resistance to Imipenem is considered to be an exceptional event by EUCAST Fig. 12: Evolution of resistance in Beta-Lactams (EUCAST) % PEN AMX CRX CTX IMI 33

35 Table 11: Resistance rates of Penicillin G in the various surveys depending on the category of infection Dosis Breakpoint % of non-susceptibility to PenG S NS Infections other than Meningitis Pneumoniae 1.2g x g x 4 or 1.2g x g x Fig. 13: % isolates with I and R for PEN (EUCAST) I R The absolute number of isolates showing high level resistance (MIC 4 µg/ml) to Penicillin G peaked in 23 and decreased subsequently since 24 to a very low level. Since 25, the rate of high level resistance was lower than 1.% (Fig 13). In the present study, the rate of high level resistance was.8 which is comparable with the levels found in the period The percentage of high level resistance in the non susceptible population varied from a maximum of 32.5 % in 23 to % in 27 and 28. In the actual survey, 5.9% of the non-susceptible isolates showed high level resistance to Penicillin. This indicates that the intermediate isolates (.12 µg/ml to 2. µg/ml) always outnumbered the isolates with high level resistance to Penicillin G. 34

36 Fig. 14: % of isolates with I and R for AMX (EUCAST) I R Figure 14 shows that high level resistance for amoxicillin occurs rarely. The rate of high level resistance remained lower than 1% since 25. Fluoroquinolones resistance rates. EUCAST does not consider wild type isolates of S. pneumoniae to be susceptible to Ciprofloxacin or Ofloxacin. Therefore, the majority of the isolates are categorized as intermediate (Fig. 15). 35

37 Fig. 15: Evolution of resistance in Fluoroquinolones (EUCAST) % 5 CIP 4 LEV 3 MOX ; , Fig. 16: Evolution of resistance to MOX and LEV (EUCAST) % LEV MOX Levofloxacin had its peak of resistance in 23 (3.3%) and showed a downward tendency to a level lower than 1.%. In the actual survey, we noticed a non significant increase to 3.1%. The resistance rate of Moxifloxacin remained low between % and 1.3% (Fig. 16). 36

38 Macrolide and Tetracycline resistance rates. Fig. 17: Evolution of resistance to MLS and TET (EUCAST) % ERY TEL TET Rates for Macrolides (MLS) and Tetracycline were fluctuating in time but were always high. Resistance rates for Telithromycin remained low (Fig. 17). High level resistance in Erythromycin is provoked by the presence of ribosomal methylase encode by the erm gene. Fig. 18: MLS Resistance genes % erm mef As can been seen in Fig. 18, the erm gene is the most important MLS resistance gene present in the Belgian isolates. Its presence varied from 76.6 % in 1995 to 94. % in 23. The actual prevalence of erm is 82. %. The mefa gene is responsible for the efflux of certain macrolides. In general, the prevalence of mefa is low in Belgian isolates. In the survey period, we found a prevalence varying from 7.9 % in 1997 to 27.2 % in 29. In the 212 study, this figure was 23.4 % which is approximately at the same level as in the period Fig. 19 shows the presence of the MLS resistance mechanisms. It can be seen that 37

39 resistance to MLS is mainly due to the presence of the isolated erm gene. However, in 29 an increase in the prevalence of isolated mefa and the combination mefa+erm was noticed. In the 211 survey, the combination of mef+erm decreased to level of 5% (level of 21). In the atual study, the level was 7.2%. Fig. 19: MLS Resistance mechanisms 1 % erm mef erm+mef Phenotypically, erm resistance (MLS B = Resistance to Macrolides, Lincosamides and Streptogramin B) is characterized by a high level resistance to erythromycin. MLS B resistance can be inducible or constitutive. Inducible MLS B resistance is characterized by high level resistance to erythromycin and susceptibility or intermediate resistance to Clindamycin. Isolates with high level resistance to erythromycin and clindamycin belong to the inducible or constitutive type of MLS B resistance. The mef type of resistance encodes low level resistance to 14 and 15 membered macrolides. These isolates are susceptible to clindamycin. year Table 12: Type of resistance mechanism in Macrolide Non susceptible isolates gene N MIC-ERY MIC-CLI MIC-TEL Range MIC5 geomic Range MIC5 geomic Range MIC5 geomic TYPE erm (92) MLS-B Induc MLS-B Induc/Const mef Efflux erm+mef MLS-B erm (83) MLS-B Induc MLS-B Induc/Const mef Efflux erm+mef * MLS-B In the present study (Table 12), 83 isolates harboured the single erm gene. Thirty seven one of these isolates (44.6 %) showed an inducible type of MLS B resistance with a MIC 5 for erythromycin of 64 µg/ml (range.5-64 µg/ml) and a MIC 5 for clindamycin of.25 µg/ml (range.3.5 µg/ml). The remaining erm positive isolates (55.4 %) can be 38

40 classified as MLS B inducible/constitutive and had a MIC 5 for erythromycin of 64 µg/ml (range µg/ml) and a MIC 5 for clindamycin of 64 µg/ml (range 1 64 µg/ml). The efflux isolates with the mef phenotype had a MIC 5 for erythromycin of 2 µg/ml (range.5-8 µg/ml) and a MIC 5 for clindamycin of.6 µg/ml (range µg/ml). Isolates harbouring the combination of both genes showed a phenotype corresponding to the MLS B resistance phenotype: MIC 5 for erythromycin of 64 µg/ml (range 4-64 µg/ml) and a MIC 5 for clindamycin of 32 µg/ml (range.6 64 µg/ml). 39

41 Evolution of MIC distributions. Evolution of Beta-lactam MIC distributions. Fig. 2: MIC distribution of PEN % ,1,2,3,6,12,25, Fig. 2 shows clearly a bimodal distribution of the MIC for Penicillin G. This was also the case for the other beta-lactam antibiotics. This bimodal distribution was found in every survey. The bimodal character indicates that there exists a distinct modus for a susceptible population and another modus for the non-susceptible population. The Modus of the 212 population was.15 µg/ml as it was the case since 24. This indicates that the population did not shift to higher MIC values. Fig. 21 shows the evolution of the MIC5 (expressed in µg/ml) values in Penicillin susceptible and Penicillin non-susceptible isolates collected during the various surveillance studies. The MIC5 can be considered as the indicator for intrinsic activity. The lower the value, the more active the compound is on a given population. A population becomes less sensitive to a certain compound when the MIC5 value shifts to the right side of the curve (i.e. the higher values). As it can be seen from this figure, the MIC5 increased importantly from 1995 (.8 µg/ml) to 23 (.3 µg/ml), indicating that in this period the penicillin susceptible population became less susceptible. From 24, the MIC5 was positioned at a lower value i.e..15 µg/ml (with the exception of.8 µg/ml in 27). The MIC5 value for the non-susceptible did not vary significantly during the years. This means that the shift of 4

42 the modus in the total population is due to changes in the susceptible population rather than to changes in the non-susceptible strains. The MIC5 values were positioned within the PEN-I breakpoints Fig. 21: Evolution of the MIC5 in PEN-S and PEN-NS isolates (EUCAST) PEN-S PEN-I PEN-R Evolution of Fluoroquinolone MIC distributions. % Fig. 22: MIC distribution of CIP ,3,6,12,25,

43 % ,3,6,12,25 Fig.23: MIC Distribution of LEV, % ,1,2 Fig. 24: MIC distribution of MOX,3,6,12,25, % ,12,25 Fig. 25: MIC distribution of OFL,

44 Figures 22 to 25 show the MIC distributions of Ciprofloxacin (CIP), Levofloxacin (LEV), Moxifloxacin (MOX) and Ofloxacin (OFL). As can been seen from these figures, the modus of the populations in the different years and for the different compounds did not fluctuate significantly. Furthermore, these compounds did not have a bimodal distribution. There distributions remained very Gaussian during the years. From figures 26, 27 and 28 it can be seen that for all the fluoroquinolones (Ofloxacine included but not shown as it is completely comparable to ciprofloxacin) the MIC5 value in both the susceptible and non-susceptible population did not shift in the various surveys. This indicates that the intrinsic activity of these compounds remained fairly stable during the period Fig. 26: Evolution of the MIC5 in CIP-S and CIP-NS isolates (EUCAST) S I R CIP-S CIP-NS MIC5 43

45 Fig. 27: Evolution of the MIC5 in LEV-S and LEV-NS isolates (EUCAST) LEV-S LEV-NS MIC5 Fig. 28: Evolution of the MIC5 in MOX-S and MOX-R isolates (EUCAST) MOX-S MOX-NS MIC5 Evolution of MDR. For the follow-up of resistance and MDR, four target compounds representative for different classes of antibiotics were chosen. These 4 compounds were Penicillin (PEN), Levofloxacin (LEV), Erythromycin (ERY) and Tetracycline (TET). Fig. 29 shows the evolution of MDR (resistance 2 compounds) throughout the study period In general, the MDR was more important than the resistance to one single compound. 44

46 Fig. 29: Evolution of MDR: % MDR Mono Figures 3 to 33 give the comparison of mono-resistance to one single target compound with MDR for this particular compound. For PEN, ERY and TET the MDR was always more important than the mono-resistance. For LEV, this difference was not present. Fig.3: PEN-monoR versus PEN-MDR Mono MDR Fig.31: LEV-monoR versus LEV-MDR Mono MDR 45

47 Fig.32: ERY-monoR versus ERY-MDR Mono MDR Fig.33: TET-monoR versus TET-MDR Mono MDR Cross Resistance. By categorizing isolates on the basis of susceptibility and non-susceptibility, we have found that cross resistance between related and unrelated compounds can be an important item (Figures 34 to 41). Resistance to beta-lactams (PEN, AMX, CRX) is always highly associated with resistance to other non related compounds (LEV, MOX, ERY, TEL, TET) (P values are indicated in the figures: **** P<.1; ***.1>P>.1; **.2>P>.1; *.5>P>.2). Levofloxacine resistance is highly associated with beta-lactam resistance (PEN, CRX:****; AMX:***) and resistance to Moxifloxacine (MOX:****). Furthermore, Moxifloxacin resistance were not significantly more present in isolates resistant to erythromycin, telithromycin and tetracycline. Finally, there was a significant level of cross resistance between erythromycin, telithromycin and tetracycline: ERY-R versus TEL-R, TET- R (****), TEL-R versus ERY-R and TET-R (****) and TET-R versus ERY-R and TEL-R (****). 46

48 Fig. 34: Mean Levels of Cross-Resistance between PEN- S and PEN-R isolates ( ) % AMX**** CRX**** LEV** MOX**** ERY**** TEL**** TET**** PEN-S PEN-R Fig. 35: Mean Levels of Cross-Resistance between AMX- S and AMX-R isolates ( ) % PEN**** CRX**** LEV**** MOX**** ERY**** TEL**** TET**** AMX-S AMX-R Fig. 36: Mean Levels of Cross-Resistance between CRX- S and CRX-R isolates ( ) % PEN**** AMX**** LEV**** MOX**** ERY**** TEL**** TET**** CRX-S CRX-R 47

49 Fig. 37: Mean Levels of Cross-Resistance between LEV-S and LEV-R isolates ( ) 1 8 % PEN**** AMX*** CRX**** MOX**** ERY* TEL** TET 41.7 LEV-S LEV-R Fig. 38: Mean Levels of Cross-Resistance between MOX-S and MOX-R isolates ( ) % PEN*** AMX**** CRX**** LEV**** ERY TEL TET MOX-S MOX-R Fig. 39: Mean Levels of Cross-Resistance between Ery-S and Ery-R isolates ( ) % PEN**** AMX**** CRX**** LEV* MOX TEL**** TET**** Ery-S Ery-R 48

50 Fig. 4: Mean Levels of Cross-Resistance between TEL-S and TEL-R isolates ( ) % PEN**** AMX**** CRX** LEV*** MOX ERY**** TET**** TEL-S TEL-R Fig. 41: Mean Levels of Cross-Resistance between TET-S and TET-R isolates ( ) % PEN**** AMX**** CRX**** LEV MOX ERY**** TEL**** TET-S TET-R 49

51 Tabel 13: Non-susceptibility rates of indicated antimicrobials in S. pneumoniae populations susceptible or non-susceptible (S/NS) to the various compounds ( ) S.pneumoniae Antimicrobial compounds 2 populations 1 PEN AMX CRX LEV MOX ERY TEL TET PEN-S/NS - /54..1/ /3.7.3/ /7.9.8/ /66.9 AMX-S/NS 6.4/1-4.3/ /3.9.3/ / / /69.2 CRX-S/NS 2.6/99.6.1/ /4.2.3/ / / /69.2 LEV-S/NS 12.7/ / /3.6 - / / / /41.7 MOX-S/NS 12.9/4. 6.9/ /4. 1.1/1-31.2/ / /4. ERY-S/NS 5.5/ / / /2.1.4/.5 - / /78.7 TEL-S/NS 11.9/5. 6.1/ / /5.9.4/ /1-29.7/91.2 TET-S/NS 6.3/ / / /2..4/.5 9.6/78.8.2/ Subpopulations categorized by susceptibility (S) or non-susceptibility(ns) to indicated compounds NS 2 PEN: Penicillin; AMX: Amoxicillin; CRX: Cefuroxime; LEV: Levofloxacin; MOX: Moxifloxacin; ERY: Erythromycin; TEL: Telithromycin; TET: Tetracyclne, Table 13 summarizes all these data in a numerical way. For explaining the reading of the table the first line PEN-S/NS is used as an example. On this first line we can see that in PEN-S isolates Amoxicillin resistance is % in PEN-S isolates and 54.% in OEN-NS isolates, for cefuroxime these figures are.1% CRX-resistance in PEN-S isolates and 83.8% CRX-resistance in PEN-NS isolates. The shading of data indicates that there is no significant difference between the 2 resistance rates. 5

52 Investigator Dr R. Vanhoof T F raymond.vanhoof@wiv-isp.be General address Juliette Wytsmanstreet Brussels Belgium T F Editorial address Engelandstreet Brussels Belgium T F info@wiv-isp.be Operational Direction Communicable and Infectious Diseases Unit of Antibiotic Research La Science au service de la Santé Publique, de la Sécurité de la chaîne alimentaire et de l'environnement. Wetenschap ten dienste van Volksgezondheid, Veiligheid van de Voedselketen en Leefmilieu. Editeur responsable Dr Johan Peeters Directeur général Editor in chief Dr Johan Peeters, General Director ISSN: D/212/255/47 N de dépôt : D/28/255/34