REVIEW. and 2 Department of Microbiology, Hôpital AP-HP Pitié Salpêtrière, Paris, France

Transcription

1 REVIEW Extended-spectrum b-lactamases in long-term-care facilities Marie-Hélène Nicolas-Chanoine 1, Vincent Jarlier 2 and La Collégiale de Bactériologie-Virologie-Hygiène Hospitalière de l Assistance Publique, Hôpitaux de Paris, France * 1 Department of Microbiology, Hôpital AP-HP Beaujon, Clichy, Faculté de Médecine D. Diderot, Paris and 2 Department of Microbiology, Hôpital AP-HP Pitié Salpêtrière, Paris, France ABSTRACT Although the first reports on extended-spectrum b-lactamase (ESBL)-producing isolates in long-termcare facilities (LTCFs) appeared 1 years ago, there are still scanty data on this topic. A long-term survey starting in 1993 by the microbial laboratories of the Assistance Publique Hopitaux de Paris and covering 21 beds, 7 of them in LTCFs, indicated that the incidence of ESBL-producing isolates 1 hospitalisation days in LTCFs increased from.7 in 1996 to.28 in 25. Escherichia coli accounted for 8% of ESBL-positive isolates in 25, whereas it accounted for <45% in 21. This rise in E. coli with ESBLs reflected clonal spread, as found elsewhere, with CTX-M types now the predominant enzyme types. Keywords CTX-M-type, Escherichia coli, extended-spectrum b-lactamases, long-term-care facilities, long-term survey, review Clin Microbiol Infect 28; 14 (Suppl. 1): INTRODUCTION Corresponding author and reprint requests: M.-H. Nicolas- Chanoine, Hôpital Beaujon, Service de Microbiologie, 1 Bld du Général Leclerc, 9211 Clichy, France mhn.chanoine@bjn.aphp.fr *See list of contributors in Acknowledgements section. Articles focusing on extended-spectrum b-lactamase (ESBL)-producing iaceae isolates in long-term-care facilities (LTCFs) or nursing homes are few in number [1 9]. The present article is, to our knowledge, the first to report a long-term survey of ESBL-producing isolates identified in clinical specimens sampled from patients of LTCFs. The microbiologists of the Assistance Publique Hôpitaux de Paris (AP-HP), which includes 47 hospitals with a total of 21 beds, have analysed the ESBL-producing iaceae isolates collected at 39 short-term-care facilities (STCFs, 14 beds) and the eight LTCFs (7 beds) during 2-month periods of each year since ESBL-producing iaceae isolates were detected using the double-disk synergy test [1] as routinely applied in AP-HP microbiology laboratories and interpreted according to the recommendations of the French Antibiotic Committee [11]. Clinical and antibiotic susceptibility data were also collected for each isolate. RESULTS Incidence of ESBL-producing iaceae in AP-HP STCFs and LTCFs Fig. 1 shows the total number of ESBL-producing iaceae isolates collected each year over the study period and the incidence density per 1 hospitalisation days (HD) in all AP-HP STCFs and LTCFs from 1996 to 25. The total number of ESBL-producing iaceae isolates was four times higher in 25 than in 1996, while the incidence density was globally stable until 21 in both STCFs (between.15 and.2 1 HD) and LTCFs (between.5 and.9 1 HD). It then increased gradually to reach.35 1 HD in STCFs in 25, whereas it suddenly increased in LTCFs in 23, reaching.28 1 HD in 26. Differences between ESBL-producing iaceae isolates from STCFs and LTCFs Between 21 and 25, the proportions of oxytoca, Proteus mirabilis,

2 112 Clinical Microbiology and Infection, Volume 14, Supplement 1, January 28 Incidence density cloacae and pneumoniae with ESBLs were 1%, in AP-HP LTCFs (Fig. 2). The proportion of aerogenes isolates with ESBLs significantly decreased from 25% to 5%, whereas that of Escherichia coli isolates dramatically increased, accounting for 8% of ESBL-producing iaceae from LTCFs by 25. The proportion of E. coli isolates among ESBL producers also increased between 21 and 25 in STCFs (Fig. 3), whereas that of E. aerogenes isolates decreased, and those of K. oxytoca, P. mirabilis and K. pneumoniae isolates remained stable. However, the species proportions were different between STCFs and LTCFs, and in 25, the proportions were 5% and 8%, respectively for E. coli, 25% and 1%, respectively, for K. pneumoniae, and 12% and 5%, respectively, for E. cloacae. Other % All settings STC LTC Year Isolate N Fig. 1. Incidence density 1 hospitalisation days of extended-spectrum b-lactamase-producing isolates in all Assistance Publique Hôpitaux de Paris (AP-HP) settings, in AP-HP short-term-care facilities and AP-HP long-termcare facilities ( ). The number of isolates collected each year is indicated at the bottom. Escherichia coli pneumoniae aerogenes cloacae Proteus mirabilis oxytoca year Fig. 2. Prevalence (%) of extended-spectrum b-lactamaseproducing isolates by species in Assistance Publique Hôpitaux de Paris long-term-care facilities (21 25). % Escherichia coli pneumoniae aerogenes cloacae year Proteus mirabilis oxytoca Fig. 3. Prevalence (%) of extended-spectrum b-lactamaseproducing isolates by species in Assistance Publique Hôpitaux de Paris short-term-care facilities (21 25). differences were observed between LTCFs and STCFs. The first difference concerned the delay between patient admission and the time for ESBL positivity. This delay exceeded 3 days for the great majority of the patients hospitalised in LTCFs between 23 and 25, whereas an ESBL-producing isolate was obtained within the first 2 days of hospitalisation for a third of the ESBL-positive patients in STCFs. This feature was particularly apparent for E. coli; 1 out of the 28 ESBL-positive isolates collected between 23 and 25 in STCFs were isolated within the first 2 days of hospitalisation. The second difference concerned the types of clinical specimen. In LTCFs, 89% of the isolates were from urine, whereas this proportion was only 5% in STCFs, where 3% of ESBL producers were isolated from deep specimens (blood, pus and the lower respiratory tract). The third difference concerned antibiotic susceptibility; the LTCF isolates were less often susceptible to quinolones than those from STCFs, with 5% vs. 32% being susceptible to ciprofloxacin (Table 1). Molecular epidemiology of ESBL-producing E. coli isolates in AP-HP LTCFs In 21 and 22, i.e., before the general increase in ESBL-producing isolates in the AP-HP LTCFs, one of these facilities, a 65-bed long-term-care hospital referred to as SP was confronted with the emergence and spread of three virulent clonally related strains of E. coli producing CTX- M-15 enzyme [6]. The three strains, which displayed the same randomly amplified polymorphic DNA (RAPD) profile, were resistant to

3 Nicolas-Chanoine et al. ESBL in long-term-care facilities 113 fluoroquinolones and susceptible to co-trimoxazole, but differed from each other in their aminoglycoside and tetracycline susceptibilities. One lineage, which was first isolated on 2 October 21, was resistant to aminoglycosides, except for gentamicin, and was then called GEN S. The second lineage, which emerged 22 days later in October 21, was resistant to all aminoglycosides and was called GEN R. The third lineage, which emerged in April 22, was susceptible to aminoglycosides and was called AMG S. Strains GEN S and AMG S spread into the different units and subunits of the LTCF, whereas strain GEN R spread only in the unit where it was first isolated. Overall, during the study period (October 21 to October 22), 87 new patients had urinary tract colonisation or infection due to strain GEN S, while there were 12 episodes due to strain GEN R and 13 due to strain AMG S. All three clonal ESBL producers were shown to belong to the phylogenetic group B2, which is one of the two predominant groups observed among extra-intestinal pathogenic E. coli isolates (the other group being D) and harbours the same virulence factor genes. Interestingly, the number of virulence factor genes was low, at just three (aer, fuya and irp2), which is uncommon for pathogenic E. coli strains from phylogenetic group B2, although apparently common among those that are resistant to fluoroquinolones [12 14]. As shown in Fig. 4, the % of non-susceptible isolates year Fig. 4. Consumption of fluoroquinolones (m), ceftriaxone (d), and co-trimoxazole ( ), expressed as the defined daily doses (DDDs) per 1 hospital days in the long-term-care facility SP and the proportion of urinary Escherichia coli isolates resistant to these antibiotics in 2, 21 and 22 [6]. The bars represent data for E. coli isolates not susceptible to fluoroquinolones (bars), ceftriaxone (striped bars) and co-trimoxazole (grey bars) ddd/1 hospitalisation-days emergence and spread of the lineages resulted in an increased percentage of E. coli isolates not susceptible to fluroquinolones and ceftriaxone, although still susceptible to co-trimoxazole. This feature was accompanied by a decrease in the consumption of fluroquinolones, an increase in that of co-trimoxazole, but no change in that of ceftriaxone (Fig. 4). Kassis-Chikhani et al. [9] described a similar outbreak in one of the three long-term-care units of the AP-HP STCF, referred to as TN. The index case was detected in October 21. Subsequently, 26 among the 47 patients in that unit were colonised or infected, over a period of 27 months, by a phylogenetic group B2 strain that produced the CTX-M-15 ESBL. Once again, this had very few virulence factor genes, and was resistant to fluoroquinolones but susceptible to co-trimoxazole. The curve of Fig. 1 clearly shows that more than the two AP-HP LTCFs TN and SP reported ESBL-producing isolates in 25. To obtain more insight into the wider distribution of ESBL-producing isolates, especially E. coli, it was decided to type the E. coli isolates collected in 25 by RAPD, as previously described [6]. Some examples of the results are presented in Figs 5 and 6, in order to illustrate the general features of the ESBL producers from LTCFs in Paris. The 14 ESBLproducing E. coli isolates from the LTCF CF (Fig. 5) displayed five distinct profiles, two comprising single isolates and three corresponding to clusters: cluster 1 with three isolates, cluster 2 with seven isolates, and cluster 3 with two isolates. For the LTCF CR (Fig. 6), three distinct M NC Fig. 5. Randomly amplified polymorphic DNA profiles of 14 extended-spectrum b-lactamase-producing Escherichia coli isolates from the long-term-care facilities CF (east southern suburb of Paris). Numbers at the top correspond to the profile type. M, weight marker; NC, negative control.

4 114 Clinical Microbiology and Infection, Volume 14, Supplement 1, January 28 M A 1 B 1 C 1 NC Fig. 6. Randomly amplified polymorphic DNA profiles of 14 extended-spectrum b-lactamase-producing Escherichia coli isolates for the long-term-care facilities (LTCFs) CR (northern suburb of Paris) and for known epidemic strains A, B and C. Numbers at the top correspond to the types of profile. (A) LTCF SP current epidemic strain. (B) LTCF TN 21 epidemic strain. (C) LTCF SP 21 epidemic strain. M, weight marker, NC, negative control. profiles were displayed among the14 ESBL-producing E. coli isolates collected there, two of them being very close to (profiles 1 and 1 in Fig. 6) but Table 1. Antibiotic susceptibility of extended-spectrum b-lactamase-producing isolates by location (23 25) Antibiotic Total no. of isolates (% susceptible to antibiotic) LTCF STCF Gentamicin 268 (63) 726 (42) Tobramycin 244 (36) 713 (24) Amikacin 262 (58) 728 (56.5) Nalidixic acid 258 (4) 699 (21.7) Ciprofloxacin 25 (5) 71 (32.5) Imipenem 259 (98) 727 (98.2) LTCF, long-term-care facility; STCF, short-term-care facility. clearly distinguishable from the third profile (profile 2). Interestingly, profiles 1 and 1 were very close, or identical, to those of the clonal strains of 21 and 22 in LTCF SP (C in Fig. 6 ), and the long-term-care unit of STCF TN (B in Fig. 6). RAPD profiles were available for the ESBLproducing E. coli isolates from six of the eight AP- HP LTCFs. On the basis of the global analysis of these profiles, it was observed that: (i) clonal strains were more common than unique strains in the AP-HP LTCFs; (ii) the clonal strains in LTCFs SP and TN in 21 were probably the same strain; and (iii) this strain was still present in 25 in six AP-HP LTCFs. Characterisation of the ESBL produced by Entrobacteriacae isolates in AP-HP LTCFs The ESBLs produced by the iaceae in LTCFs CF and CR in 25 were molecularly characterised [6,15]. All producers from facility CF were E. coli, whereas they also included other enterobacterial species in LTCF CR (Table 2). Ten of the 14 E. coli isolates from facility CF produced CTX-M-15 b-lactamase, as did 13 of 14 E. coli isolates from CR. This b-lactamase was also found in the sakazakii isolate identified in LTCF CR (Table 2). Of the four remaining E. coli isolates from LTCF CF, two were not confirmed as ESBL producers and single representatives produced the b-lactamases CTX-M and TEM-14 12, respectively. The last E. coli isolate from LTCF CR produced SHV-12, as did two E. cloacae isolates and one K. pneumoniae isolate, whereas the sole Serratia marcescens isolate produced SHV-5. Table 2. Distribution of extended-spectrum b-lactamase (ESBL)-producing isolates and ESBL type among species in two Assistance Publique Hôpitaux de Paris long-term-care facilities (LTCFs) in 25 LTCF CF (east-south suburb) LTCF CR (north suburb) Species No. ESBL type (isolate no.) No. ESBL type (isolate no.) Escherichia coli 14 CTX-M-15 (1) 14 CTX-M-15 (13) CTX-M (1) SHV-12 (1) TEM (1) Non-BLSE (2) cloacae 2 SHV-12 (2) Serratia marcescens 1 SHV-5 (1) sakazakii 1 CTX-M-15 (1) pneumoniae 1 SHV-12 (1)

5 Nicolas-Chanoine et al. ESBL in long-term-care facilities 115 DISCUSSION The long-term survey ( ) performed by the Collégiale de Bactériologie-Virologie-Hygiène Hospitalière of AP-HP made it possible to monitor a clear increase in ESBL-producing isolates of E. coli and the stability over time of those of K pneumoniae, independently of the care facility type. These results are very similar to those found by Romero et al. [5], who also carried out a longterm survey in Seville ( ). These authors observed an increase in the frequency of ESBLproducing isolates of E. coli and a stable rate of ESBL-producing K. pneumoniae isolates in the area served by the university hospital Virgen Macarena located in the North of Seville (Spain). This mixture of change and stability concerned both inpatients in different wards of the university hospital and in a nearby chronic-care hospital, and outpatients. The increase in ESBL-producing E. coli isolates started in 1999 for inpatients and in 2 for outpatients, 1 2 years before that in the AP-HP hospitals. By typing the ESBL-producing E. coli isolates collected by the AP-HP LTCFs, we have shown that clonal groups are common across these facilities. Although the literature concerning ESBL-producing isolates in LCTFs or nursing homes is limited, some of the published articles mention the clonal relatedness of the ESBL-producing E. coli isolates collected in these medical centres [2 5]. These data are summarised in Table 3. Therefore, the most original point demonstrated by the study performed at AP-HP is the persistence of the same strain of ESBL-producing E. coli over a long period of time in the same LTCF, and the presence of this strain in different LCTFs at the same time, although the LTCFs concerned are distant from one another. Although this strain belonged to phylogenetic group B2, it harboured very few virulence factor-encoding genes. This latter feature might explain why bacteraemias due to this strain were rare in the patients of the AP-HP LTCFs. Nevertheless, this strain was resistant to amoxycillin, ceftriaxone and fluoroquinolones, antibiotics that are widely used to treat infections in elderly people. Thus, antibiotic selection pressure and the fact that it is very difficult to implement isolation precautions in centres caring for elderly patients may both explain the persistence and the spread of this type of strain in LTCFs. CONTRIBUTOR LIST Participating hospitals Albert Chenevier (C.-J. Soussy, P. Legrand, N. Mangeney, C. Dupeyron); Ambroise Paré, Saint Perrine (B. Heym, F. Espinasse); Antoine Beclère (D. Ingrand, M. Guibert, C. Argentin); Armand Trousseau (A. Gabarg-Chenon, H. Vu-Thien); Beaujon (M.-H. Nicolas-Chanoine, F. Bert); Bichat-Claude Bernard (A. Andremont, L. Armand-Lefevre); Charles Foix-Jean Rostand (V. Jarlier, S. Lefrancois); Charles Richet (C. Abramowitz, L. Bensidhoum, M.-H. Nicolas-Chanoine, L. Noussair); Cochin-Broca, La Rochefoucauld (C. Poyart, H. Poupet); Emile-Roux (C.-J. Soussy, A. Akpabie); Européen Georges Pompidou, Corentin Celton (L. Gutmann, G. Kac, I. Podglagen); Georges Clémenceau (J.-L. Avril, F. Blonde- Cynober, L. Cukier); Henri Mondor (C.-J. Soussy, P. Legrand, L. Desforges); Hôtel Dieu (A. Bouvet, S. Coignard, A. Casetta); Jean Verdier (A. Collignon, I. Poilane); Joffre-Dupuytren (C. Aussel, H. Nebbab-Lechani, N. Dangla); Kremlin Bicêtre Table 3. Previous studies on the clonal relatedness of extended-spectrum b-lactamase-producing Escherichia coli isolates in long-term-care facilities or nursing homes Author (country) Journal (year) Study period Patient no. Patient origin Profile no. clonal group Wiener et al. (USA) Cukier et al. (France) Ma et al. (Japan) Romero et al. (Spain) JAMA (1999) July One nursing home 7 Clonal group only if patients in the same room Path Biol (1995) March 1994 to 24 One geriatric 4 One clonal group for 21 September 1995 department patients J Infect Chemother January 1996 to 17 Three geriatric 3 One clonal group for 15 (22) October 1997 hospitals patients Clin Microbiol January 1995 to Infect (25) June Outpatients STCF LTCF 84 Thirteen clonal groups of two to four isolates with two larger groups. In 2 21: the largest group (n =9 isolates) = LTCF isolates LTCF, long-term-care facility; STCF, short-term-care facility.

6 116 Clinical Microbiology and Infection, Volume 14, Supplement 1, January 28 (P. Nordmann, N. Fortineau, C. Poy); Lariboisière-Fernand Widal (M.-J. Sanson-Le Pors, L. Raskine); Louis Mourier (C. Branger, M.-L. Joly Guillou, M. Eveillard); Necker-Enfants Malades (P. Berche, J. R. Zahar, A. Ferroni); Paul Brousse (E. Dussaix, N. Kassis, D. Mathieu); Pitié-Salpêtrière (V. Jarlier, J. Robert, R. Bismuth, D. Trystram); Raymond Poincaré (J.-L. Gaillard, E. Ronco, C. Lawrence); René Muret-Bigottini (A. Collignon, I. Durand); Robert Debré (E. Bingen, C. Doit); Rotschild (A. Gabarg-Chenon, H. Vu-Thien, B. Salauze); Saint Antoine (J.-C. Petit, V. Lalande); Saint-Louis (P. Lagrange, J.-L. Donnay); Saint Vincent De Paul (C. Poyard, H. poupet, P. Lebon, N. Boutros, J. Raymond, M. Degrave); San Salvadour (V. Simha, M. F. Lipens); Tenon (G. Arlet); Villemin-Paul Doumer (C. Cattoire). Data manager D. Trystram (Hôpital Pitié-Salpêtrière). Molecular studies W. Sougakoff, L. Drieux, F. Brossier (Hôpital Pitié-Salpêtrière); V. Leflont-Guibout (Hôpital Beaujon). REFERENCES 1. Muller M, McGeer A, Willey BM et al. Outbreaks of multidrug resistant Escherichia coli in long-term care facilities in the Durham, York, and Toronto regions of Ontario, Can Commun Dis Rep 22; 28: Wiener J, Quinn JP, Bradford PA et al. Multiple antibioticresitant and Escherichia coli in nursing homes. JAMA 1999; 281: Cukier L, Lutzler P, Bizien A, Avril JL. Investigation of an epidemic of an extended spectrum beta-lactamase producing Escherichia coli in a geriatric department. Pathol Biol 1999; 47: Ma L, Matsuo H, Ishii Y, Yamaguchi K. Characterization of cefotaxime-resistant Escherichia coli isolates from a nosocomial outbreak at three geriatric hospitals. J Infect Chemother 22; 8: Romero L, Lopez L, Rodriguez-Bano J, Ramon-Hernandez J, Martinez-Martinez L, Pascual A. Long-term study of frequency of Escherichia coli and Kebsiella pneumoniae isolates producing extended-spectrum b-lactamases. Clin Microbiol Infect 25; 11: Leflon-Guibout V, Jurand C, Bonacorsi S et al. Emergence and spread of three clonally related virulent isolates of CTX-M-15-producing Escherichia coli with variable resistance to aminoglycosides and tetracycline in a French geriatric hospital. Antimicrob Agents Chemother 24; 48: Mendelson G, Hait V, Ben-Israel J, Gronich D, Granot E, Raz R. Prevalence and risk factors of extended-spectrum beta-lactamase-producing Escherichia coli and pneumoniae in an Israeli long-term care facility. Eur J Clin Microbiol Infect Dis 25; 24: Einhorn AE, Pharm D, Neuhauser MM et al. Extendedspectrum b-lactamases: frequency, risk factors, and outcomes. Pharmacotherapy 22; 22: Kassis-Chikhani N, Vimont S, Asselat K et al. CTX-M betalactamase-producing Escherichia coli in long-term care facilities, France. Emerg Infect Dis 24; 1: Jarlier V, Nicolas MH, Fournier G, Philippon A. Extended broad-spectrum b-lactamases conferring transferable resistance to newer b-lactam agents in iaceae: hospital prevalence and susceptibility patterns. Rev Infect Dis 1988; 1: Comité de l Antibiogramme de la Société Française de Microbiologie. Communiqué 26http:// 12. Maslow JN, Lautenbach E, Glaze T, Bilker W, Johnson JR. Colonization with extraintestinal pathogenic Escherichia coli among nursing home residents and its relationship to fluoroquinolone resistance. Antimicrob Agents Chemother 24; 48: Horcajada JP, Soto S, Gajewskia X et al. Quinolone-resistant uropathogenic Escherichia coli strains from phylogenetic group B2 have fewer virulence factors than their susceptible counterparts. J Clin Microbiol 25; 43: Drews SJ, Poutanen SM, Mazzulli T et al. Decreased prevalence of virulence factors among ciprofloxacinresistant uropathogenic Escherichia coli isolates. J Clin Microbiol 25; 43: Leflon-Guibout V, Bonacorsi S, Clermont O, Ternat G, Heym B, Nicolas-Chanoine MH. Pyelonephritis caused by multiple clones of Escherichia coli, susceptible and resistant to co-amoxiclav, after a 45 day course of co-amoxiclav. J Antimicrob Chemother 22; 49: