To the Editor—The recent increase of extended-spectrum β-lactamase (ESBL)–producing Klebsiella pneumoniae (Kp) and the emergence of carbapenemase-producing Kp in Portuguese clinical settings parallels epidemiological trends described in other countries.Reference Rodrigues, Machado, Ramos, Peixe and Novais ^{1 – 4} Moreover, Kp isolates causing hospital infection often correspond to the patient’s own colonizing strains, stressing the need to survey fecal carriage of multidrug-resistant (MDR) Kp in patients from different clinical settings.Reference Martin, Cao and Brisse ⁵ Long-term care facilities (LTCFs) are fundamental institutions in contemporary healthcare services, mainly assisting elderly people who, due to frequent hospitalizations, recurrent antibiotic consumption, and communal living, are at a high risk of infection by MDR bacteria.Reference Aschbacher, Pagani and Confalonieri ⁶ Different studies among European LTCFs residents have reported high rates of colonization by CTX-M-15-producing Escherichia coli (Ec) B2-ST131, but little is known regarding the prevalence and diversity of other MDR Enterobacteriaceae species (and particularly Kp) colonizing LTCFs residents.Reference Aschbacher, Pagani and Confalonieri ^{6 ,} Reference Gonçalves, Cecílio and Ferreira ⁷ The aim of this study was to assess the fecal carriage rate and epidemiological features of non-Ec Enterobacteriaceae isolates resistant to extended-spectrum β-lactams among Portuguese LTCF residents.

Fecal samples from residents (n=47) at LTCF 1 (25 beds; n=25 samples) and LTCF 2 (40 beds; n=22 samples) in northern Portugal were collected during July 2015 and January 2016, respectively. Demographic and clinical characteristics of the residents are summarized in Online Supplemental Table S1. Rectal swabs were suspended in 2 mL saline solution, and 0.2 mL were seeded on chromID ESBL (bioMèrieux, Marcy-l'Étoile, France) and chromID Carba SMART plates (bioMèrieux, Marcy-l'Étoile, France) directly and after a preenrichment step in 10 mL of trypticase soy broth containing a 10-μg meropenem disk, and incubation at 37°C for 18 hours for the screening of ESBL and/or plasmid-mediated AmpC and carbapenemase producers at 37°C for 24–48 hours, respectively. ^{8 ,} Reference Rodrigues, Machado, Fernandes, Peixe and Novais ⁹ All presumptive non-Ec Enterobacteriaceae (representative morphotypes per plate) were selected for further characterization. ESBL/plasmid–mediated AmpC or carbapenemase production was confirmed by phenotypic or Blue-Carba tests, respectively, polymerase chain reaction (PCR), and sequencing.Reference Rodrigues, Machado, Fernandes, Peixe and Novais ⁹ Susceptibility testing to non-β-lactams antibiotics was performed by the disk diffusion ¹⁰ and bacterial identification was confirmed by species-specific PCR, and/or matrix-assisted laser-desorption/ionization–time-of-flight mass spectroscopy. Clonal relatedness among Kp isolates was evaluated by Fourier transform infrared (FTIR) spectroscopy, multilocus sequence typing (MLST), and wzi capsular typing.Reference Diancourt, Passet, Verhoef, Grimont and Brisse ^{11 –} Reference Sousa, Novais, Magalhães, Lopes and Peixe ¹³

Air samples (250 L) from different common indoor (3 from LTCF 1 and 1 from LTCF 2) and outdoor spaces were also collected using an MAS100 (Merck Millipore, Germany) air sampler to assess microbiological air quality. The different culture medium plates (ie, plate count agar [PCA], chromID ESBL, and chromID Carba SMART) used were incubated at 25°C for 72 hours (PCA, for fungi quantification) or at 37°C for 48 hours (PCA and selective medium, for bacteria quantification). Microbiological air quality (expressed in CFU/m³) was categorized according to the Portuguese law. ¹⁴

A high proportion of fecal samples (19 of 47; 40.4%) was positive for non-Ec Enterobacteriaceae producing ESBL (29.8%) or plasmid-mediated AmpC (10.6%). Despite some epidemiological differences, similar colonization rates were observed in both institutions: 44% in LTCF 1 and 36% in LTCF 2 (Table 1). Notably, in 53% of the samples, we also identified ESBL-producing Ec, leading to an overall rate of fecal carriage with ESBL producers of 81% (data not shown). ESBL carriage was significantly associated with the gender, length of stay, and residents of shared rooms, whereas plasmid-mediated AmpC carriage was only significantly associated with consumption of β-lactams in the previous 3 months (Online Supplemental Table S1). Air quality was within the established standards only at LTCF 1, although in both institutions no growth was detected on selective media. The colonization rates by ESBL-producing non-Ec Enterobacteriaceae (29.8%) were significantly higher than those observed in these species among LTCFs and nursing home residents years ago in Portugal in 2008–2012 (~6%) or in LTCFs in other European countries in 2012–2013 (~8%).Reference Aschbacher, Pagani and Confalonieri ⁶ Despite the low sample size, this extraordinary increase (~5-fold) is worrisome in this at-risk population; it is probably influenced by the recent global expansion of MDR Kp isolates in Portuguese clinical institutions. ⁴ Carbapenemase-producing Enterobacteriaceae were not detected even though they have been increasingly identified among clinical isolates in our country.Reference Rodrigues, Bavlovič, Machado, Amorim, Peixe and Novais ^{2 ,} Reference Manageiro, Ferreira and Almeida ³

TABLE 1 Epidemiological Data of ESBL- or Plasmid-Mediated AmpC-Producing Non-E. coli Isolates Identified in Residents From Portuguese Long-Term Care Facilities (LTCFs)

NA, not applicable; NI, not identified; F, female; M, male; H, hospital; LTCF, long-term care facility; S, shared room; I, individual room; Kp, K. pneumoniae; Ea, E. aerogenes; Ent., Enterobacter spp.; Pm, Proteus mirabilis; ST, sequence type; AMK, amikacin; CIP, ciprofloxacin; CLO, chloramphenicol; GEN, gentamicin; KAN, kanamycin; NAL, nalidixic acid; NET, netilmicin; STR, streptomycin; SUL, sulphonamides; SXT, sulfamethoxazole; TET, tetracycline; TMP, trimethoprim; TOB, tobramycin.

^a Variability among isolates is shown in parenthesis.

^b Single locus variant (SLV) of ST405.

CTX-M-15 (16 isolates; 14 samples) and DHA-1 (6 isolates; 5 samples) were, respectively, the only ESBL and plasmid-mediated AmpC variants identified (Table 1). In LTCF 1, CTX-M-15 was predominant (n=12 isolates in 91% of the samples), especially among Kp ST15 lineages (7 wzi19-K19, 3 wzi24-K24). A DHA-1-producing ST11 (wzi75) Kp isolate was sporadically detected (Table 1). Both ST15 Kp clonal lineages (wzi19, wzi24) or ST11-wzi75 have been frequently identified among isolates producing CTX-M-15, KPC-3 or DHA-1 in Portuguese clinical institutions.Reference Rodrigues, Machado, Ramos, Peixe and Novais ^{1 –} Reference Manageiro, Ferreira and Almeida ³ Conversely, in LTCF 2, CTX-M-15 (4 isolates in 50% of the samples) was detected in 2 Kp clones (2 ST348-wzi94 and 1 ST15-wzi24) and 1 Enterobacter aerogenes, whereas DHA-1 (n=5 isolates in 50% of the samples) was identified in diverse species and Kp lineages (Table 1). Notably, the cocolonization by isolates from the same or different species harboring the same enzyme suggests in vivo transfer of bla gene.

Our study reveals alarming (and increased) colonization rates by MDR non-Ec Enterobacteriaceae among LTCFs residents. These rates represent a challenge for these institutions in terms of antibiotic stewardship and infection prevention and control, and they highlight the risk of further dissemination to healthcare professionals, residents’ families, and the community in general. Most of the Kp lineages (ST11 [wzi75], ST15 [wzi19/wzi24], and ST348 [wzi94]) identified are the same lineages involved in infections in clinical institutions from the same region, evidencing a higher risk of infection among this vulnerable population. The variability observed on the acquired antibiotic resistance genes among colonization/infection isolates in our settings suggests an open genome of such lineages for plasmid exchange.