REPORT OF THE NATIONAL EPIDEMIOLOGIC SURVEY OF ENTEROBACTER AEROGENES IN BELGIAN HOSPITALS IN

REPORT OF THE NATIONAL EPIDEMIOLOGIC SURVEY OF ENTEROBACTER AEROGENES IN BELGIAN HOSPITALS IN 1996-98. Y. De Gheldre 1, M.J. Struele 1, C. Nonhoff 1, N. Maes 1, P. De Mol 2, H. Chetoui 2, Y. Glupczyki 3, C. Sion 3, O. Ronveaux 4, M. Vaneechoutte 5, and the Groupement pour le Dépistage, l Etude et la Prévention des Infectio Hospitalières-Groep ter Opsporing, Studie en Preventie van de Infecties in de Ziekenhuizen (GDEPIH-GOSPIZ). Laboratoire de Microbiologie 1, Hôpital Erasme, Université Libre de Bruxelles, Bruxelles, Laboratoire de Microbiologie 2, Centre Hospitalier Universitaire Sart Tilman, Liège, Laboratoire de Microbiologie 3, Hôpital Mont Godinne, Université Catholique de Louvain, Yvoir, Ititut Scientifique de la Santé Publique Louis Pasteur 4, Bruxelles, Laboratorium voor Bakteriologie en Virologie 5, Universitair Ziekenhuis, Gent. INTRODUCTION Over the last 5 years, E. aerogenes has gradually increased in frequency as agent of nosocomial infection in several countries. During the same period, multi-resistant E. aerogenes strai have been associated with outbreaks among hospitalized patients, especially in Belgium (2-3) and France (1). Data from the national surveillance survey of Hospital Infectio and from a retrospective study on the epidemiology of E. aerogenes isolates in Belgian hospitals, showed a significant increases between 1994 and 1995 of the incidence of E. aerogenes bloodstream infectio in ICU patients (4). Likewise, the resistance frequency of this organism to third generation cephalospori, fluoroquinolones and imipenem was also observed. OBJECTIVES OF THE SURVEY This national survey was carried out between 1996 and 1998 to evaluate the incidence and antimicrobial susceptibility of nosocomial strai of E. aerogenes in Belgian hospitals in 1996-7 in comparison to 1994-5, and to study the genotypic dissemination, geographic dissemination of epidemic genotype and antibiotic resistance patter of isolates collected during 1997-8. MATERIALS AND METHODS Descriptive epidemiologic survey: data were collected by mailing a retrospective questionnaire to 39 hospital laboratories which were selected on the basis of their voluntary willingness to participate or because they already took part to a previous similar survey carried out in 1994-95 (4). This questionnaire was aimed to evaluate local prevalence and incidence of E. aerogenes clinical isolates recovered from routine culture during 1996 and during the first semester of 1997. Prospective microbiologic survey : the same set of 39 acute care hospitals were asked to send 10 coecutive non-duplicate E. aerogenes isolates from patients hospitalized during the period 1997 first semester 1998. Molecular identification of E. aerogenes : all isolates were identified to the species level by trna- PCR analysis, as previously described (7). Susceptibility testing : minimal inhibitory concentratio (MIC) were determined by the agar dilution method following the NCCLS recommendatio (M7-A4, 1997) for 10 antimicrobials including temocillin, piperacillin-tazobactam, ceftazidime, cefepime, gentamicin, isepamicin, amikacin, imipenem, meropenem, and ciprofloxacin. Double-disk synergy test : screening for extended spectrum beta-lactamase (ESBL) production was performed by determining a potentiation of the zones of inhibition around ceftazidime (30µg) and/ or cefepime (30µg) disks by placing clavulanic acid disks (20 µg amoxicillin, 10 µg clavulanic acid) at a distance of 20 mm (ceftazidime) or 30 mm (cefepime) apart. 1

Beta-lactamase detection and identification: isoelectric focusing (IEF) was performed as previously described on polyacrylamide gels to presumptively identify the beta-lactamase type according to pi (5). Arbitrarily primed-pcr (AP-PCR) typing: bacterial DNA was amplified by PCR with two primers (Primer 5 and M13) using the Ready-To-Go PCR Beads (Pharmacia Biotech) and DNA fragments were separated on agarose gels. AP-PCR profiles were compared using the Pearson correlation product moment coefficient. Pulsed field gel electrphoresis (PFGE) typing: to validate results of PCR typing, PFGE analysis was performed on samples from predominant PCR types. DNA preparation and cleavage with XbaI was performed as previously described (2). Normalized PFGE profiles were compared using the Pearson correlation product moment coefficient. Epidemiological analysis: an epidemic AP-PCR type of E. aerogenes was defined as a type recovered in more than one hospital. The geographic distribution of isolates by PCR type was mapped according to the city where the hospital is located. RESULTS Descriptive epidemiologic survey : thirty-nine hospitals (20 % of Belgian acute care ititutio) completed the questionnaire, of which 21 supplied data which could be analysed. Middle-size (400-599 beds) and larger ititutio (>600 beds) had a higher respoe rate (52 % and 58%, respectively) than the smaller hospitals (12%). Regional distribution showed 16 participants each from Wallonia and Flanders, and 7 from the Brussels area. No significant increase was observed between 1996 and 1997 neither in the relative frequency of isolation of Enterobacter spp. within the Enterobacteriaceae nor in the proportion of E. aerogenes within the Enterobacter genus (Table 1). Table 1. Evolution of the relative frequency and incidence of isolation of E. aerogenes from clinical specime between 1996 and 1997 in Belgian hospitals (n=21) Indicator Study year p 1996 1997 No. Enterobacter spp / No. Enterobacteriaceae 3,395/26,671 1,895/15,154 Pooled Mean (%) 12.7 12.5 Median (%) 11.2 10.9 Range (%) 5.5-30.7 6.7-23.8 No. E. aerogenes / No. Enterobacter spp 1,632/3,395 988/1,895 Pooled Mean (%) 48.1 52.1 Median (%) 50 54.6 Range (%) 22.1-82.7 14.8-79.1 No. E. aerogenes / No. admissio 1,632/30,0082 988/153,325 Pooled Mean/ 1000 admissio 5.2 6.6 <0.01 Median/ 1000 admissio 3.3 4.2 <0.01 Range/ 1000 admissio 0.6-16.3 0.3-21.2 2

In contrast the median incidence of E. aerogenes isolates rose from 3.3 to 4.2/ 1000 admissio between 1996 and 97 (p<0.01). A higher proportion of E. aerogenes strai within the Enterobacter genus was found in hospitals from the Walloon region as compared to the other regio (Table 2). Table 2. Evolution by region of the relative frequency and incidence of E. aerogenes from clinical specime between 1996 and 1997 in Belgian hospitals (n=21) Region No. hospitals No. E. aerogenes / No. Enterobacter spp (%) Mean incidence of E. aerogenes colonization/ 1,000 admissio 1996 1997 1996 1997 p 1996 1997 p Flanders 7 7 276 (42) 175 (46) 3.4 4.1 Brussels 6 6 712 (41) 357 (41) 5 5.4 Wallonia 7 7 644 (66) 456 (72) 9.3 12.4 <0.001 Prospective microbiologic survey: Overall, 274 viable strai originating from clinical specime were sent from 31 participating centres. Phenotypic identification could be confirmed for 265 (96.7%) of the isolates. E. aerogenes strai were recovered typically from elderly patients (mean age 69 years [range 11-96]), admitted to ICUs (22%), medecine (26%), surgery (18%), geriatry (11%). Specimen from the urinary and respiratory tracts represented respectively 45% and 32% of the isolates source; 9% of isolates were recovered from wound swabs and 3% from blood cultures. Molecular identification: 259 of 262 (99%) strai were identified as E. aerogenes. Three strai could not be identified by this technique. AP-PCR typing: E. aerogenes isolates (n=260) clustered in 25 distinct PCR types, of which 12 were epidemic (Fig. 1). Figure 1. Dendrogram of % similarity of AP-PCR profiles of E. aerogenes Belgian isolates (==260). 10 % similarity (Pearson) AP-PCR type No. of strai 20 30 40 50 60 70 80 90 100 (%) No. of hospitals (%) BE 1 94 (36) 21 (72) BE 3 16 (6) 11 (38) BE 4 5 (2) 3 (10) BE 5 4 (1) 4 (14) BE 6 4 (1) 3 (10) BE 7 3 (1) 3 (10) BE 8 2 (1) 2 (7) BE 9 2 (1) 2 (7) BE 2 100 (38) 25 (86) BE 10 6 (2) 4 (1) 3

These 12 epidemic types were found dispersed from a number of hospitals ranging from 2 to 25. About 75% of the strai clustered in two major types, BE 1 and BE 2. 21% of the strai grouped in 10 epidemic types and 4% in 11 sporadic (unique pattern for each strain) types. BE 1 and BE 2 coisted of 94 (36%) and 100 (38%) strai which were isolated from 21 (68%) hospitals and 25 (81%) hospitals, respectively. BE 1 type was predominant in Wallonia (representing 60% of the strai) whereas BE 2 was predominant in Flanders (representing 48% of the strai) (Fig. 2). In Brussels, these two types were equally represented (each type representing 20% of the strai). Figure 2. Geographic distribution of E. aerogenes isolates belonging to the 2 major epidemic types. BE 1 strai BE 2 strai PFGE typing: analysis of PFGE profiles indicated that strai belonging to BE 1 type clustered distantly from those belonging to BE 2. However, coiderable heterogeneity was observed among profiles within each group. Antimicrobial susceptibility: the MIC 50/ MIC 90 (in µg/ ml) by antimicrobial of 249 strai available for testing were : temocillin:16/ 32; piperacillin-tazobactam: 64/ 256; ceftazidime: 256/ 256; cefepime: 1/ 8; imipenem: 1/ 2; meropenem: 0,06/ 0,12; amikacin: 8/ 16; gentamicin: 1/ 2; isepamicin: 1/ 2; ciprofloxacin: 32/ 64. Resistance to ceftazidime was seen in 88% of isolates and to piperacillin-tazobactam and ciprofloxacin in over 75% of isolates (Table 3). Table 3. Susceptibility of E. aerogenes isolates (n=249) to 10 antimicrobials according to their AP-PCR type. Antimicrobial % susceptible strai with AP-PCR type (No. of strai tested) BE 1 (n=93) BE 2 (n=94) Other epidemic (n=51) Sporadic (n=11) Any (n=249) Temocillin 78 65 80 82 74 Piperacillin-tazobactam 19 21 41 36 25 Ceftazidime 1* 13* 28 27 12 Cefepime 90 94 97 100 93 Amikacin 100 85 100 100 94 Gentamicin 100 100 98 96 96 Isepamicin 100 100 100 100 100 Imipenem 99 98 98 100 98 Meropenem 100 100 100 100 100 Ciprofloxacin 7* 24* 47 82 23 p<0.001, for the proportion of susceptible strai versus other epidemic and sporadic type strai. 4

Occasional strai were resistant to cefepime (7%) and imipenem (2%). All strai were susceptible to meropenem and isepamicin. Production of ESBL was detected among 107 (45%) of 236 E. aerogenes isolates which were tested by double disk synergy test and by IEF. These 107 strai produced a TEM-24 (86% of strai) or a TEM-3 (14% of the strai) enzyme. No single isolate harbored two ESBLs. Strai with detectable ESBLs were significantly more resistant to piperacillin-tazobactam, ceftazidime, cefepime and ciprofloxacin than non-producing strai (Table 4). Table 4. Susceptibility of E. aerogenes isolates (n=236) to 10 antimicrobials according to the production of ESBL. Antimicrobial % susceptible strai with ESBL production (No of strai) yes (n=108) no (n=128) P value Temocillin 71 78 Piperacillin-tazobactam 14 33 <0.001 Ceftazidime 1 21 <0.001 Cefepime 86 98 <0.001 Amikacin 94 94 Gentamicin 96 95 Isepamicin 100 100 Imipenem 97 99 Meropenem 100 99 Ciprofloxacin 2 42 <0.001 MIC of amikacin 4 fold that of gentamicin were seen in 93% of those strai in comparison with 53% of the non ESBL producing strai (p<0.001). ESBLs were present in 91% of BE 1 strai, of which 95% were identified as TEM-24. In contrast, ESBL production was detected in only 21% of BE 2 strai, of which 50% were identified as TEM-24 and 50% as TEM-3. BE 1 and BE 2 strai were significantly more frequently resistant to ceftazidime (p<0.001) and ciprofloxacin (p<0.001) than other strai. BE 1 strai were more frequently resistant than BE 2 strai to ceftazidime and ciprofloxacin and more susceptible to amikacin (p<0.001). CONCLUSIONS The prevalence of E.aerogenes within the Enterobacter genus increased in 1996-1997 as compared to with 1994 and 1995 in Flanders and in Wallonia but remained stable in Brussels. Incidence data from this survey indicated a continuing increase of the role E. aerogenes as an important nosocomial pathogen, particularly in Wallonia where its incidence was found to be the highest of the three regio. A majority of E. aerogenes isolates were resistant to ciprofloxacin and to all the beta-lactams tested, except temocillin, cefepime and carbapenems. Over 90% of the strai tested were also susceptible to gentamicin and amikacin, although higher MICs were found for the latter aminoglycoside, a finding coistent with the well described production of AAC (6 )-I by this species in Belgium (6). Approximately half of the strai produced an ESBL, predominantly TEM-24 or less commonly TEM- 3. ESBL producing strai were more resistant to piperacillin-tazobactam, ceftazidime, cefepime and ciprofloxacin than non-producing strai and AAC(6 )-I production was more frequent among the ESBL producing strai. Isolates were distributed predominantly in two major PCR types, BE 1 and 2. BE 1 strai, which were more prevalent in Wallonia, produced a TEM-24 beta-lactamase. The potential of MREA to spread was illustrated by the number of small clusters of epidemic MREA that we observed. Multidrug-resistant E. aerogenes (MREA) did not only occur among ICU admitted patients but also among patients cared for in medical or surgical wards. In conclusion, this study shows that a diversity of antibiotic-resistant E. aerogenes strai have become endemic and that two groups of related MREA strai are widespread in hospitals distributed all over Belgium. These data support the need for screening and isolation of patients carriers of these strai to control their nosocomial spread. Control efforts should be evaluated by the establishment of a local surveillance program and supported by a national surveillance program of MREA in Belgian hospitals. 5

REFERENCES 1. Arpin, C., C. Coze, A.M. Rogues, J.P. Gachie, C. Bebear, and C. Quentin. 1996. Epidemiological study of an outbreak due to multidrug-resistant Enterobacter aerogenes in a medical inteive care unit. J. Clin. Microbiol. 34: 2163-2169. 2. De Gheldre, Y., N. Maes, F. Rost, R. De Ryck, P. Clevenbergh, J.L. Vincent and M.J. Struele. 1997. Molecular epidemiology of an outbreak of multidrug-resistant Enterobacter aerogenes infectio and in vivo emergence of imipenem resistance. J. Clin. Microbiol. 35 : 152-160. 3. Jalaluddin, S., J.M. Devaster, R. Scheen, M. Gerard and J.P. Butzler. 1998. Molecular epidemiological study of nosocomial Enterobacter aerogenes isolates in a Belgian hospital. J. Clin. Microbiol. 36 : 1846-1852. 4. Ronveaux, O., Y. De Gheldre, Y. Glupczyki, M.J. Struele and P. De Mol. 1999. Emergence of E. aerogenes as a major antibiotic-resistant nosocomial pathogen in Belgian hospitals. Clin. Microbiol. Infect. 5: 622-627. 5. Matthew, M., A.M. Harris, M.J. Marshall, and G.W. Ross. 1975. The use of isoelectric focusing for detection and identification of beta-lactamases. J. Gen. Microbiol. 88 :169-178. 6. Vanhoof, R., H.J. Nyssen, E. Van Bossuyt, E. Hannecart-Pokorini and the Aminoglycoside Resistance Study Group. 1999. Aminoglycoside resistance in Gram-negative blood isolates from various hospitals in Belgium and the Grand Duchy of Luxembourg. J. Antimicrob. Chemother. 44: 483-488. 7. Welsh, J., M. McClelland. 1992. PCR-amplified length polymorphisms in trna intergenic spacers for categorizing staphylococci. Mol. Microbiol. 6 : 1673-1680. ACKNOWLEDGMENTS We thank our colleagues from participating laboratories for their dedicated participation to this study. This work was supported by a grant of Bristol Meyers Squibb, Belgium. 6