Original research

Colistin resistance in Escherichia coli and Klebsiella

pneumoniae in humans and backyard animals in
Ecuador
Carlos Bastidas-Caldes,1 Salomé Guerrero-Freire,2 Nimer Ortuño-Gutiérrez,3 Temmy Sunyoto,4
Cícero Armídio Gomes-Dias,5 Maria Soledad Ramírez,6 William Calero-Cáceres,7 Anthony D. Harries,8
Joaquín Rey,9 Jacobus H. de Waard10 and Manuel Calvopiña10

Suggested citation Bastidas-Caldes C, Guerrero-Freire S, Ortuño-Gutiérrez N, Sunyoto T, Gomes-Dias CA, Ramírez MS, et al. Colistin resistance
in Escherichia coli and Klebsiella pneumoniae in humans and backyard animals in Ecuador. 2023;47:e48.
https://doi.org/10.26633/RPSP.2023.48

ABSTRACT Objective. Colistin is an antibiotic of last resort for treating serious Gram-negative bacterial infections. However,
the misuse of colistin, especially as an animal growth promoter, has contributed to increasing antimicrobial
resistance, mediated mainly through plasmid transfer of the mcr-1 gene. This study assessed the prevalence
of phenotypic and molecular colistin resistance in Escherichia coli and Klebsiella pneumoniae in Ecuador in
healthy humans and their chickens and pigs.
Methods. Fecal samples were collected from humans and their chickens and pigs in two rural coastal and
Amazon regions between April and August 2020. Gram-negative bacteria were isolated and identified using
conventional techniques. Phenotypic resistance was determined using the broth microdilution technique, and
the mcr-1 gene was detected using conventional polymerase chain reaction.
Results. A total of 438 fecal samples were obtained from 137 humans, 147 pigs and 154 chickens. The preva-
lence of E. coli isolates was 86.3% (378/438) and K. pneumoniae, 37.4% (164/438). Overall, the mcr-1 gene
was found in 90% (340/378) of E. coli isolates, with higher prevalences found in isolates from coastal regions
(96.5%, 191/198), humans (95.6%, 111/116) and chickens (91.8%, 123/134); for K. pneumoniae, the gene was
found in 19.5% (32/164) of isolates, with equal distribution between regions and hosts. Only four isolates, two
E. coli and two K. pneumoniae, showed phenotypic resistance: mcr-1 was present in both E. coli strains but
absent in the K. pneumoniae strains.
Conclusions. Despite a low prevalence of phenotypic resistance to colistin, the high prevalence of the mcr-1
gene in E. coli is of concern. Ecuador’s ban on using colistin in animal husbandry must be enforced, and con-
tinual monitoring of the situation should be implemented.

Keywords Colistin; Escherichia coli; Klebsiella pneumoniae; humans; animals; drug resistance; genes, MDR; operational
research; Ecuador.

1
One Health Research Group, Biotecnología, Facultad de Ingeniería y Cien- 6
Department of Biological Science, College of Natural Sciences and Mathematics,
cias Aplicadas (FICA), Universidad de las Américas (UDLA), Quito, Ecuador California State University Fullerton, Fullerton, California, United States of America
*   Carlos Bastidas-Caldes, [email protected] 7
Universidad Técnica de Ambato-Resistencia a los Antimicrobianos (UTA RAM)
2
Programa de Doctorado de Ciencias Veterinarias, Facultad de Ciencias Veteri- One Health, Department of Food and Biotechnology Science and Engineering,
narias, Universidad de Buenos Aires, Buenos Aires, Argentina Universidad Técnica de Ambato, Ambato, Ecuador
3
Damien Foundation, Brussels, Belgium 8
International Union Against Tuberculosis and Lung Disease, Paris, France
4
Médecins Sans Frontières Operational Centre Brussels, Luxembourg Opera- 9
Unidad de Patología Infecciosa y Epidemiología, Facultad de Veterinaria, Uni-
tional Research Unit, Luxembourg versidad de Extremadura, Cáceres, Spain
5
Department of Basic Health Sciences, Federal University of Health Sciences of 10
One Health Research Group, Facultad de Ciencias de la Salud, Universidad de
Porto Alegre (UFCSPA), Porto Alegre, Brazil las Américas (UDLA), Quito, Ecuador

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs 3.0 IGO License, which permits use, distribution, and reproduction in any medium, provided the
original work is properly cited. No modifications or commercial use of this article are permitted. In any reproduction of this article there should not be any suggestion that PAHO or this article endorse any specific organization
or products. The use of the PAHO logo is not permitted. This notice should be preserved along with the article’s original URL. Open access logo and text by PLoS, under the Creative Commons Attribution-Share Alike 3.0
Unported license.

Rev Panam Salud Publica 47, 2023 | www.paho.org/journal | https://doi.org/10.26633/RPSP.2023.48 1

Original research Bastidas-Caldes et al. • Colistin resistance in Ecuador

Antimicrobial resistance is recognized as one of the most Ecuador is a country that straddles the equator on South
serious global threats to human health. The emergence of mul- America’s west coast. It has a diverse landscape that includes
tidrug resistance – defined as an organism showing resistance the Amazon jungle, Andean highlands, a coastal region and the
to three or more classes of antibiotics – in Escherichia coli and wildlife-rich Galapagos Islands. Rural hamlets in the coastal
Klebsiella pneumoniae is particularly concerning (1). Both are region (province of Santo Domingo de los Tsáchilas) and the
Gram-negative bacteria that cause serious infections and have Amazon region (province of Pastaza) were included in the
multiple resistance mechanisms. The commonest examples of study. Each hamlet had a population of around 700 inhabitants
these are extended-spectrum β-lactamases, carbapenemases (2) in about 140 households, with a mean household size of 5 per-
and for colistin, the presence of the mcr-1 gene. sons. There were approximately 15 hamlets and farmhouses
A recent study has highlighted the growing threat of resist- in each of the study regions. Most inhabitants earn their liveli-
ance to colistin mediated through mcr genes, with documented hood from agriculture and farming activities, with every family
examples of resistance in animals, humans, food and the usually having chickens, pigs, ducks and some cattle in their
environment (3). Colistin is an antibiotic considered one of the backyard. These hamlets were selected because there was a
medicines of last resort against multidrug-resistant Gram-nega- preliminary report disclosing information about resistance to
tive bacterial infections (4). Colistin is commonly used in animal colistin and cefotaxime in E. coli in backyard animals in these
husbandry for treatment and as a growth promoter in food regions (12).
supplements, either in normal or excessively high doses, thus Human fecal samples were collected and placed in semisolid
increasing the selection for colistin resistance in animals (5). In Cary–Blair transport medium for coliforms (Eiken Chemical,
addition, the effectiveness of antibiotics, including colistin, is Tokyo, Japan). For samples from backyard animals, cloacal
being threatened by their excessive use in veterinary medicine. and rectal swabs were taken from, respectively, chickens and
Indeed, in some countries, the use of colistin has been 600 times pigs, and these were placed in transport medium similar to that
higher in animals than in humans (6). used for humans. These specimens were all kept at 4 °C until
In 2016, Liu et al. identified in China for the first time a plasmid- transportation. Laboratory testing was conducted by the micro-
mediated colistin- resistance gene, a mobile element that con- biology laboratory at the Universidad de las Américas.
tains the mcr-1 gene (7). This gene encodes the expression of
phosphoethanolamine transferase, affecting lipid A, which Phenotypic identification of isolates
confers antibiotic resistance in Enterobacterales (7). Colistin-
resistant Enterobacterales, particularly E. coli and K. pneumoniae, All samples from humans and animals were screened by
carrying the mcr-1 gene have since been reported worldwide, inoculating onto selective CHROMagar COL-APSE chromo-
including in the Americas (in Argentina, Bolivia [Plurinational genic medium (Paris, France), and they were incubated at 37 °C
State of], Brazil, Colombia, Peru and the United States) (8). for 18 hours. Plates were considered positive when at least one
Studies in Latin America have suggested that this spread might typical (coliform) colony had formed (15). Subsequently, the
be associated with the horizontal transfer of colistin-resistance typical colonies were inoculated onto BD DIFCO MacConkey
genes (9). Since the discovery of the mcr-1 gene and its presence Agar medium (Becton Dickinson, Franklin Lakes, New Jersey,
in Enterobacterales plasmids, worldwide surveillance of colis- USA). Suspected colonies were separated and further con-
tin resistance has been strengthened (5). firmed as E. coli or K. pneumoniae through biochemical testing
In 2016, Ortega-Paredes et al. reported the first clinical isolate that included inoculation onto Simmons citrate medium, and
of colistin-resistant E.coli harboring the mcr-1 gene in Ecuador in the triple sugar iron, urea, sulfur indole motility and the methyl
an adolescent with appendicitis (10). Since then, studies of ani- red/Voges–Proskauer tests (all media and solutions were from
mals on rural farms where there is extensive use of colistin as a Becton Dickinson); isolates were preserved for future experi-
growth promoter have shown widespread distribution of colistin- ments. All isolates were preserved in brain heart infusion broth
resistant E.coli, with a high proportion of isolates containing with 10% glycerol; those used in the study were frozen at −20 °C
mcr-1 genes (11–13). With the growing importance of the One and those preserved for future research were frozen at −80 °C.
Health concept, documenting the prevalence of colistin-resistance ATCC (Manassas, Virginia, USA) strains 25922 for E. coli and
genes in both humans and animals in Ecuador is critical to estab- 700603 for K. pneumoniae were used for quality control.
lish a baseline of how far the mcr genes have dispersed (14).
This information will help to reinforce national surveillance and Phenotypic resistance to colistin
guide policies for the rational use of colistin. Such a study will
also provide evidence about the possible cross-transfer of mcr-1 The isolates of E. coli and K. pneumoniae were incubated in
genes between different host species. Therefore, this study aimed brain heart infusion broth for 18 hours at 37 °C for enrichment.
to assess the prevalence of phenotypic colistin resistance and Subsequently, colonies were isolated on nutrient agar and cul-
the presence of mcr-1 in E. coli and K. pneumoniae from humans tivated at 37 °C for 18 hours. These colonies in Muller Hinton
and their backyard animals (chickens and pigs) living in selected broth medium were compared with the 0.5 McFarland standard.
rural communities in Ecuador between April and August 2020. Briefly, the broth microdilution test was carried out in line with
the recommendations of the Clinical and Laboratory Standards
MATERIALS AND METHODS Institute (CLSI standard M100, 2022), in which bacteria diluted
to the 0.5 McFarland standard are cultured together with the
Study design and sampling test antibiotic. Two cut-off points were used: ≤ 2 μg/mL to indi-
cate intermediate resistance and ≥ 4 μg/mL to indicate colistin
This was a cross-sectional study using primary data and con- resistance (CLSI standard M100, 2022). All isolates were then
ducted in rural areas of Ecuador. cultured in brain heart infusion broth for cryopreservation for

2 Rev Panam Salud Publica 47, 2023 | www.paho.org/journal | https://doi.org/10.26633/RPSP.2023.48

Bastidas-Caldes et al. • Colistin resistance in Ecuador Original research

future experiments. The plates were processed in a reader at mcr-1 gene was kindly donated by the Osaka Institute of Public
the recommended wavelength (260 nm). Each isolate was inoc- Health, Japan. Finally, to confirm the variant of the mcr gene,
ulated and subsequently measured in duplicate. the 3130 Genetic Analyzer (Applied Biosystems, Waltham,
Massachusetts, USA) was used for Sanger sequencing on posi-
Bacterial DNA extraction tive amplification products from PCR. The sequences obtained
were aligned with others already described and characterized
DNA was extracted following the Chelex 100 and proteinase in the GenBank database using the Geneious 8.0 (Biomatters,
K methods previously described (16). Briefly, all confirmed iso- Auckland, New Zealand) and MEGA X (Molecular Evolution-
lates of E. coli and K. pneumoniae were resuspended in 200 μL ary Genetics Analysis, https://www.megasoftware.net/docs)
of Chelex 10% (Sigma-Aldrich, St. Louis, Missouri, USA) and 5 bioinformatics programs. A summary of study procedures is
μL of proteinase K (Invitrogen, Waltham, Massachusetts, USA). presented in Figure 1.
Samples were incubated at 56  °C for 60 min. Following this,
the samples were vortexed and centrifuged at 8000 g for 2 min. Sample size calculation
They were then incubated at 96 °C for 20 min. Finally, samples
were centrifuged at 8000 g for 5 min and the supernatant was The sample size was calculated based on the formula
transferred to a clean tube and stored at −20 °C. N  = (Z1− α/2)2p(1 − p)/d2, where Z1 − α/2 is the value of
normal deviation at a 95% confidence level, p is the expected
Molecular detection of mcr frequency of colistin resistance in about 50% of farms, accord-
ing to the reports by Yamamoto et al. (12), with 15% relative
One-step polymerase chain reaction (PCR) was per- precision (d), a design effect of 2 for cluster sampling and a
formed for the mcr-1 gene using the forward primer nonresponse rate of 10%.
5′-GCTACTGATCACCACGCTGT-3′ and the reverse primer Cluster random sampling was adopted to select the host
5′-AGCTGAACATACACGGCACA-3′, giving a product size from which samples were collected. A total of 20 clusters were
of 698 bp. The QIAGEN PCR kit (QIAGEN, Hilden, Germany) selected from the available hamlets. Then 20 households were
was used. The reaction mix for the PCR contained an amplifica- selected from each cluster by simple random sampling. From
tion mixture of 10 µL, with a concentration of 1 µM of multiplex each household that agreed to participate, one adult human
master mix, 0.2 µM of the mixture of primer pairs for the mcr- or one animal was recruited into the study. This process
1 gene, and 1 µM of Q solution, following the manufacturer’s was followed until the desired total sample size of 400 was
instructions for multiplex PCR kits (17, 18). achieved.
PCR was performed in a Mastercycler thermocycler (Eppen-
dorf, Hamburg, Germany) using gradient operation, with Data entry and statistical analysis
the initial denaturation step at 95  °C for 15 min followed by
30 cycles of denaturation at 94 °C for 30 s, annealing at 62 °C Data were extracted from the study forms and laboratory
for 90 s, extension at 72  °C for 90 s and a final extension at records, double entered into a Microsoft Excel 2010 spreadsheet
72  °C for 10 min. PCR products were analyzed using SyBr (Microsoft, Redmond, Washington, USA) and analyzed using
Safe DNA Gel Stain (Invitrogen) in 2% (w/v) agarose gel in EpiData v. 2.2.3.187 (EpiData, Odense, Denmark). The preva-
1X tris–borate–EDTA (ethylenediaminetetraacetic acid) buffer, lence of bacterial isolates and their genotypic and phenotypic
and electrophoresis was performed at 100 V for 40 min in an resistance to colistin were calculated and reported with a 95%
Enduro Gel XL Electrophoresis System (Labnet International, confidence interval (CI). Phenotypic resistance was defined as
Edison, New Jersey, USA). The positive control DNA for the bacterial growth occurring in the broth microdilution test at

FIGURE 1. Flow chart of the study’s methodologya

Sampling and ethics Bacterial isolation and identification Screening for resistance

Broth microdilution
Escherichia coli Confirmation: Red CLSI standard M100
Informed
ed colonies in COL-APSE
consentt a identification using (2022)
COL-APSE
Fecal samples collected

COL-APSE selective Identification in colistin Citrate Phenotypic

chromogenic medium supplemented by resistance
Biochemical identification
Klebsiella pneumoniae

Other inoculation onto TSI

Enterobacterales MacConkey agar medium
Urea PCR for mcr-1
SIM
Isolates positive
MR/VP
for mcr-1

CLSI: Clinical and Laboratory Standards Institute (standard M100, 2022); mcr: mobile colistin resistance gene; MR/VP: methyl red/Voges–Proskauer; PCR: polymerase chain reaction; TSI: triple sugar iron; SIM: sulfur
indole motility.
a
Phenotypic resistance breakpoints followed CLSI standard M100 for 2022. Using microbroth dilution, bacteria are considered to have intermediate resistance to colistin if growth occurs at ≤ 2 µg/µL of the antibiotic
and considered resistant to colistin if growth occurs at ≥ 4 µg/µL.
Source: Figure prepared by the authors for this study.

Rev Panam Salud Publica 47, 2023 | www.paho.org/journal | https://doi.org/10.26633/RPSP.2023.48 3

Original research Bastidas-Caldes et al. • Colistin resistance in Ecuador

≥ 4 μg/mL of colistin and genotypic resistance as the presence TABLE 1. Prevalence of Escherichia coli isolates and the mcr-1
of mcr-1. The prevalence of genotypic and phenotypic resis- gene in fecal samples from humans, pigs and chickens in two
tance in relation to geographical location and different hosts rural provinces of Ecuador, 2020 (N = 438)
was compared using the χ2 test. A P value < 0.05 was considered
statistically significant. Variables No. of Prevalence of confirmed Prevalence of mcr-1
samples isolates gene
n % (95% CI) n % (95% CI)
Ethics
Province      
Approval was obtained from the Ethics Committee for the Pastaza 212 180 84.9 149 82.7
Investigation on Human Beings at Universidad San Francisco (78.0 to 91.7) (74.7 to 90)
de Quito (approval code 2018-110E), in accordance with the Santo 226 198 87.6 191 96.5
Domingo (81.5 to 93.6) (93.1 to 99)
Declaration of Helsinki. Ethics approval was also obtained
from the Union Ethics Advisory Group (EAG approval num- Host
ber 19/21) of the International Union Against Tuberculosis and Human 137 116 84.7 111 95.6
Lung Disease, Paris, France. All participants included in the (76.1 to 93.3) (91.2 to 100)
study signed approved informed consent forms. Chicken 154 134 87.0 123 91.8
(79.5 to 94.5) (85.7 to 98)
Pig 147 128 87.1 106 82.8
RESULTS (79.3 to 94.6) (73.6 to 92)
CI: confidence interval.
Prevalence of E. coli and K. pneumoniae Source: Table prepared by the authors with data from their study.

A total of 438 individual fecal samples were obtained from Prevalence of phenotypic resistance to colistin
rural hamlets of the coastal and the Amazon regions of Ecua-
dor. These included 226 samples from Santo Domingo province Two (0.5%) of the 378 E. coli isolates showed statistically sig-
in the coastal region (72 from humans, 78 from pigs, 76 from nificant phenotypic resistance to colistin (Table 3). Two (1.2%)
chickens) and 212 samples from Pastaza province in the Ama- of the 164 K. pneumoniae isolates showed statistically significant
zon region (65 from humans, 69 from pigs, 78 from chickens). phenotypic resistance to colistin, with these two isolates being
The overall prevalence of E. coli isolates was 86.3% (378/438; identified in pigs (Table 4).
95% CI: 82.1 to 90.5) and the overall prevalence of K. pneumoniae
was 37.4% (164/438; 95% CI: 32.0 to 42.0]. The prevalences of Correlation between colistin-resistant phenotypes
these isolates, stratified by province and by host, are shown in and mcr-1
Table 1 and Table 2, respectively. For E. coli isolates, the prev-
alence was similar in Santo Domingo and in Pastaza among Correlations between phenotypic resistance and the presence
chickens, pigs and humans and ranged between about 85% and of mcr-1 for E. coli and K. pneumoniae are shown in Table 3 and
88%. For K. pneumoniae isolates, however, there was a statisti- Table 4, respectively. For E. coli, mcr-1 was found in all isolates
cally significant higher prevalence in Santo Domingo (42.9%) from humans, pigs and chickens showing phenotypic inter-
compared with Pastaza (31.6%) (P <  0.05) and a significantly mediate resistance (≤ 2 µg/mL) to colistin. The mcr-1 gene was
higher prevalence in humans (72.3%) compared with chickens also found in the two E. coli isolates showing resistance (≥ 4 µg/
(27.9%) and with pigs (15.0%) (P < 0.01). mL) to colistin (Table 3). For K. pneumoniae, mcr-1 was found
in isolates from humans, pigs and chickens with intermediate
Prevalence of mcr-1 in E. coli and K. pneumoniae resistance to colistin. The mcr-1 gene was not found in the two
K. pneumoniae isolates showing resistance to colistin (Table 4).
The overall prevalence of mcr-1 in E. coli isolates was 90.0%
(340/378; 95% CI: 86.1 to 93.8) and the overall prevalence DISCUSSION
in K. pneumoniae isolates was 19.5% (32/164; 95% CI:12.1 to
27.9). The prevalences of mcr-1 in E. coli and K. pneumoniae, To our knowledge, this is the first study in Ecuador, and one
stratified by province and by host, are also shown in Table 1 of the first in Latin America, that has assessed the prevalence of
and Table 2, respectively. For E. coli isolates, there was a sta- fecal carriage of colistin-resistance among E. coli and K. pneumo-
tistically significant higher prevalence in isolates from Santo niae isolates from healthy humans and their backyard animals
Domingo (96.5%, 191/198) compared with Pastaza (82.7%, living in rural areas of two different regions. We found a sur-
149/180) (P <  0.001) and a statistically significant higher prisingly low prevalence of phenotypic resistance to colistin
prevalence in humans (95.6%, 111/116) and chickens (91.8%, (< 1%), but a high prevalence of molecular resistance in E. coli,
123/134) compared with pigs (82.8%, 106/128) (P <  0.0.5). with mcr-1 present in 90% of fecal samples with E. coli.
For K. pneumoniae isolates, the prevalence of mcr-1 was simi- The high prevalence of E. coli in all types of hosts was con-
lar in Santo Domingo and Pastaza, at about 19%, and similar sistent with previous studies, confirming that this bacterium
between humans, chickens and pigs, at about 16–21%. The is one of the most common colonizers of the gastrointestinal
GenBank accession number MW527090, corresponding to the tract in both animals and humans (14). The overall prevalence
phosphoethanolamine–lipid A transferase mcr-1 gene, was of K. pneumoniae in the three types of hosts was lower, at 37%,
obtained from isolate LR37 from a healthy human in Santo although humans had a higher prevalence of the organism
Domingo province. compared with chickens and pigs. K. pneumoniae naturally

4 Rev Panam Salud Publica 47, 2023 | www.paho.org/journal | https://doi.org/10.26633/RPSP.2023.48

Bastidas-Caldes et al. • Colistin resistance in Ecuador Original research

TABLE 2. Prevalence of Klebsiella pneumoniae isolates and the TABLE 3. Association between phenotypic intermediate resist-
mcr-1 gene in fecal samples from humans, pigs and chickens ance and resistance to colistin and presence of the mcr-1 gene
in two rural provinces of Ecuador, 2020 (N = 438) in Escherichia coli isolates from humans and backyard animals
in Ecuador, 2020 (N = 378)a
Variables No. of Prevalence of confirmed Prevalence of mcr-1
samples isolates gene Host Results of broth  mcr-1 gene
n % (95% CI) n % (95% CI) microdilution No. (%) negative No. (%) positive Total
Province       Human I 5 (4.3) 111 (95.7) 116
Pastaza 212 67 31.6 13 19.4 R 0 0
(16.9 to 46.2) (0 to 74.6) Pig I 22 (17.2) 104 (81.3) 128
Santo 226 97 42.9 19 19.6 R 0 2* (1.6)
Domingo (29.9 to 55.8) (0 to 43.1)
Chicken I 11 (8.2) 123 (91.7) 134
Host R 0 0
Human 137 99 72.3 21 21.2 I: intermediate resistance to colistin is determined when growth occurs at ≤ 2 µg/µL of the antibiotic; R: bacteria
(60.3 to 83.6) (0 to 43.9) are considered resistant to colistin if growth occurs at ≥ 4 µg/µL; *: statistically significant difference at P < 0.05
using the χ2 test.
Chicken 154 43 27.9 7 16.3 a
The cut-off points for determining resistance were from the Clinical and Laboratory Standards Institute standard
(10.3 to 45.6) (0 to 51.6) M100, 2022.
Source: Table prepared by the authors with data from their study.
Pig 147 22 15.0 4 18.2
(0 to 34.6) (0 to 67.5)
CI: confidence interval.
Source: Table prepared by the authors with data from their study.
diet, animal husbandry practices and the environment might
play a part.
In K. pneumoniae isolates, we found a lower prevalence of
colonizes the respiratory and gastrointestinal tracts of humans, mcr-1, at less than 20%, with little difference between geograph-
and the bacterium is one of the main etiological pathogens of ical regions and type of host. However, this prevalence in our
infections with clinical relevance (19). The low prevalence of study was higher than those found elsewhere: in Taiwan, 5.5%
K. pneumoniae in chickens and pigs in our study has been found of chickens and 0.4% of pigs were reported to harbor mcr-1 (26,
elsewhere (20) and might be due to their diet, competition in 27), while in China 11% of chickens and 7% of pigs with resis-
the environment or genetic virulence factors responsible for the tant K. pneumoniae isolates harbored the gene (28).
colonization of enteric systems in animals (21, 22). There was no correlation between phenotypic resistance
There was a surprisingly low prevalence of phenotypic resist- and the presence of mcr-1. This finding has been previously
ance to colistin in our study. Our results concur with a recent described by others (29, 30), and it may be explained by the
review that found a global prevalence of colistin resistance of multifactorial aspects of antimicrobial resistance, such as (i)
1.6%, with incidence rates rising during the past 5 years on all the diversity of plasmid types carrying mcr-1 (31), (ii) strong or
continents, largely as a result of inappropriate use of the anti- weak plasmid promotors influencing the expression of the genes
biotic in animal husbandry (23). We used standard laboratory (32), and (iii) the fusion plasmid phenomenon that increases the
methodology (broth microdilution) for determining the min- transfer capabilities of mcr genes between hosts (9). Only four
imum inhibitory concentration values for colistin, so we believe bacterial isolates were resistant, and all were found in pigs. The
our results are accurate. Of note, however, when laboratory two resistant E. coli isolates were positive for mcr-1, indicating
susceptibility testing of our samples was conducted in 2020 plasmid-induced resistance, while the two resistant K. pneu-
and 2021, the CLSI break point of ≤ 2 µg/mL of colistin was moniae isolates were negative for mcr-1, presumably indicating
regarded as indicating susceptibility. The 2022 version of the resistance due to chromosomal mutations. The high prevalence
CLSI standard indicates that bacterial growth at that concentra- mcr-1 in E. coli with intermediate resistance, however, is cause
tion should be considered as indicating intermediate resistance, for concern, as this might lead to further full-blown resistance if
which is how we have presented the data. the misuse of colistin continues.
A high prevalence of mcr-1 (90%) was found in the E. coli
isolates, and this aligns with the rapid worldwide spread of Strengths and limitations
this gene since it was first reported in 2016 in China. There are
several variants of the mcr gene, but mcr-1 appears to be the The strengths of this study were the large sample size, which
predominant variant and has spread at a rate that is 95% faster exceeded our estimated sample size; a robust system for col-
than the other 9 variants (mcr-2 to mcr-10) (3, 8). The mcr-1 gene lecting specimens; sound laboratory techniques that followed
is certainly the most common in Latin America and the Carib- international guidelines; and conducting and reporting the
bean, and the country with the highest prevalence is Brazil, with study in line with the STROBE (Strengthening the Reporting
nearly 45% of E. coli isolates testing positive for the gene (24). of Observational Studies in Epidemiology) guidelines (33). A
We found a higher prevalence of mcr-1 in E. coli isolates from further strength was the implementation of the study in two
the coastal region compared with the Amazon region, and in different regions and among rural settlements, together with
humans and chickens compared with pigs. Other studies in the inclusion of humans and two types of backyard animals.
Ecuador have also found differing prevalence rates, depend- However, there were some limitations. It was difficult to
ing on the source. For example, a low prevalence was found collect samples consistently during the study period due to
in dogs, at 25% (11), and irrigation water, at 18% (13), whereas the COVID-19 pandemic, and we had several refusals to par-
in chickens rates of more than 95% were found (12, 25). We do ticipate from households in the settlements. Genetic and
not have a clear explanation for why there are differences, but epidemiological studies are needed to compare these findings

Rev Panam Salud Publica 47, 2023 | www.paho.org/journal | https://doi.org/10.26633/RPSP.2023.48 5

Original research Bastidas-Caldes et al. • Colistin resistance in Ecuador

TABLE 4. Association between phenotypic intermediate resist- in pigs, chickens and humans (35). Therefore, regulatory policy
ance and resistance to colistin and presence of the mcr-1 gene combined with continual colistin monitoring can work and will
in Klebsiella pneumoniae isolates from humans and backyard help prevent the spread of colistin resistance in Ecuador and
animals in Ecuador, 2020 (N = 164)a elsewhere.

 Host Results of broth mcr-1 gene Authors’ contributions. All authors conceived the original
microdilution No. (%) negative No. (%) positive Total idea, planned the experiments, collected and analyzed the
Human I 78 (78.8) 21 (21.2) 99 data, contributed data or analysis tools, interpreted the results,
R 0 0 and wrote and reviewed the paper. All authors reviewed and
Pig I 16 (72.7) 4 (18.2) 22 approved the final version of the paper.
R 2* (9.1) 0
Chicken I 36 (83.7) 7 (16.3) 43 Acknowledgements. This research study was developed
through the Structured Operational Research and Training
R 0 0
Initiative (SORT IT), a global partnership coordinated by the
I: intermediate resistance to colistin is determined when growth occurs at ≤ 2 µg/µL of the antibiotic; R: bacteria
are considered resistant to colistin if growth occurs at ≥ 4 µg/µL; *: statistically significant difference at P < 0.05 Special Programme for Research and Training in Tropical Dis-
using the χ2 test.
a
The cut-off points for determining resistance were from the Clinical and Laboratory Standards Institute standard eases (TDR), the United Nations Children’s Fund (UNICEF),
M100, 2022.
Source: Table prepared by the authors with data from their study. the United Nations Development Programme (UNDP) and the
World Bank; it is hosted at the World Health Organization. The
specific SORT IT program that led to this study included an
and determine whether there are other strains in the coun- implementation partnership between TDR and the Pan Ameri-
try. Furthermore, studies on the expression of the mcr-1 gene can Health Organization, the WHO country offices of Colombia
are necessary to understand better the relationship between and Ecuador; the Ministry of Health and Social Protection,
molecular and phenotypic resistance because this is dynamic Colombia; the Food and Agriculture Organization, Sierra
and multifactorial for colistin. The clonal distribution between Leone; Sustainable Health Systems, Freetown, Sierra Leone;
humans, chickens and pigs, and the presence of the mcr-1 gene the Tuberculosis Research and Prevention Center, nongovern-
in plasmids were not determined, and these are areas to pursue mental organization, Armenia; the International Union Against
in further research. Finally, we did not collect any information Tuberculosis and Lung Disease, France and India offices
about colistin use in animals in the settlements that we studied. (South-East Asia); Institute of Tropical Medicine, Antwerp, Bel-
gium; Damien Foundation, Belgium; Indian Council of Medical
Conclusions Research–National Institute of Epidemiology; Jawaharlal Insti-
tute of Postgraduate Medical Education and Research; GMERS
The prevalence of colistin resistance in bacterial isolates of E. Medical College, Gotri, India; India Medical College Baroda;
coli and K. pneumoniae was evaluated from fecal samples from Sri Manakula Vinayagar Medical College, India; Public Health
healthy humans and their backyard animals living in rural Ontario, Canada; Universidade Federal de Ciências de Saúde
Ecuador. de Porto Alegre, Brazil; Universidade de Brasilia, Brazil; Uni-
The prevalence of molecular resistance to colistin (identified versidad de Concepción, Chile; Universidad de los Andes,
by the presence of mcr-1) was high, especially in E. coli isolates Colombia; Universidad Pontificia Bolivariana, Colombia; Uni-
(90%). However, the low prevalence of phenotypic resistance versidad Pedagógica y Tecnológica de Colombia; Universidad
associated with these strains was striking (< 1%). In K. pneumo- Central del Ecuador; California State University, Fullerton, Cal-
niae isolates, phenotypic resistance occurred independently of ifornia, USA; and Universidad Autónoma de Yucatán, México.
the presence of mcr-1 (1.2%), with the chromosomal mechanism
of colistin resistance probably being the main factor involved. Conflicts of interest. None declared.
These findings raise important concerns about the potential
spread of drug-resistant pathogens in the community, which Funding. This study was supported by funding from the Uni-
will add to the overall burden of antimicrobial resistance unless versidad de las Américas, Quito, Ecuador (grant number VET.
firm action is taken. MCA.19.03). The SORT IT antimicrobial resistance program is
To counter this, we make two important recommenda- funded by the National Institute for Health and Care Research,
tions. First, our study methodology should serve as a basis for Department of Health and Social Care, United Kingdom, and
ongoing surveillance of colistin resistance in three species of supported by implementing partners. All open access and
hosts, or carriers, and this should be replicated in sentinel sites ethics-related costs were covered by TDR. Funding was also
all over the country. Second, there is an urgent need to regulate provided by Consejería de Economía e Infraestructuras, La
the use of colistin in animal husbandry and to avoid its overuse Junta de Extremadura (grant number IB18047).
in human clinical cases for which there is no clear indication.
In January 2020, Ecuador banned the veterinary use of colis- Disclaimer. Authors hold sole responsibility for the views
tin, and it has since been shown that the sales of colistin as a expressed in the manuscript, which may not necessarily reflect
growth promoter have decreased (34). An encouraging report the opinion or policy of the Revista Panamericana de Salud Públi-
from China showed that banning colistin as an animal growth ca/Pan American Journal of Public Health or those of the Pan
promoter led to substantial decreases in colistin-resistant E. coli American Health Organization.

6 Rev Panam Salud Publica 47, 2023 | www.paho.org/journal | https://doi.org/10.26633/RPSP.2023.48

Bastidas-Caldes et al. • Colistin resistance in Ecuador Original research

REFERENCES

1. Hee Park S. Management of multi-drug resistant organisms in antimicrobial resistance: a univariate and multivariable analysis.
healthcare settings. J Korean Med Assoc. 2018;61(1):26–35. Lancet Planet Health. 2018;2(9):e398–405.
2. Bastidas-Caldes C, Romero-Alvarez D, Valdez-Vélez V, Morales RD, 20. Franklin-Alming FV, Kaspersen H, Hetland MAK, Bakksjø RJ,
Montalvo-Hernández A, Gomes-Dias C, et al. Extended-spectrum Nesse LL, Leangapichart T, et al. Exploring Klebsiella pneumoniae in
beta-lactamases producing Escherichia coli in South America: a sys- healthy poultry reveals high genetic diversity, good biofilm-forming
tematic review with a One Health perspective. Infect Drug Resist. abilities and higher prevalence in turkeys than broilers. Front
2022;15:5759–79. Microbiol. 2021;12:725414.
3. Elbediwi M, Li Y, Paudyal N, Pan H, Li X, Xie S, et al. Global bur- 21. Young T, Bray A, Nagpal R, Caudell D, Yadav H, Zafar M. Animal
den of colistin-resistant bacteria: mobilized colistin resistance genes model to study Klebsiella pneumoniae gastrointestinal colonization
study (1980–2018). Microorganisms. 2019;7(10):1–18. and host-to-host transmission. Infect Immun. 2020;88(11):e00071-20.
4. WHO Advisory Group on Integrated Surveillance of Antimicrobial 22. Cao L, Li X, Xu Y, Shen J. Prevalence and molecular characteris-
Resistance. Critically important antimicrobials for human medicine: tics of mcr-1 colistin resistance in Escherichia coli: isolates of clinical
Ranking of medically important antimicrobials for risk management infection from a Chinese university hospital. Infect Drug Resist.
of antimicrobial resistance due to non-human use, sixth revision. 2018;11:1597–603.
Geneva: World Health Organization; 2019. https://www.who.int/ 23. Binsker U, Käsbohrer A, Hammerl JA. Global colistin use: a review
publications/i/item/9789241515528 of the emergence of resistant Enterobacterales and the impact on
5. Poirel L, Jayol A, Nordmann P. Polymyxins: antibacterial activity, their genetic basis. FEMS Microbiol Rev. 2022;46(1):fuab049.
susceptibility testing, and resistance mechanisms encoded by plas- 24. Fernandes MR, Moura Q, Sartori L, Silva KC, Cunha MP, Esposito
mids or chromosomes. Clin Microbiol Rev. 2017;30(2):557–96. F, et al. Silent dissemination of colistin-resistant Escherichia coli in
6. Skov RL, Monnet DL. Plasmid-mediated colistin resistance (mcr- South America could contribute to the global spread of the mcr-1
1 gene): three months later, the story unfolds. Euro Surveill. gene. Euro Surveill. 2016;21(17):pii=30214.
2016;21(9):30155. 25. Hmede Z, Kassem II. The colistin resistance gene mcr-1 is preva-
7. Liu YY, Wang Y, Walsh TR, Yi LX, Zhang R, Spencer J, et al. Emer- lent in commensal Escherichia coli isolated from preharvest poultry
gence of plasmid-mediated colistin resistance mechanism MCR-1 in in Lebanon. Antimicrob Agents Chemother. 2018;62(11):e01304-18.
animals and human beings in China: a microbiological and molecu- 26. Yang YQ, Li YX, Lei CW, Zhang AY, Wang HN. Novel plasmid-
lar biological study. Lancet Infect Dis. 2016;16(2):161–8. mediated colistin resistance gene mcr-7.1 in Klebsiella pneumoniae. J
8. Bastidas-Caldes C, de Waard JH, Salgado MS, Villacís MJ, Coral- Antimicrob Chemother. 2018;73(7):1791–5.
Almeida M, Yamamoto Y, et al. Worldwide prevalence of mcr- 27. Lai CC, Lin YT, Lin YT, Lu MC, Shi ZY, Chen YS, et al. Clinical
mediated colistin-resistance Escherichia coli in isolates of clinical characteristics of patients with bacteraemia due to the emergence
samples, healthy humans, and livestock—a systematic review and of mcr-1-harbouring Enterobacteriaceae in humans and pigs in Tai-
meta-analysis. Pathogens. 2022;11(6):659. wan. Int J Antimicrob Agents. 2018 Nov 1;52(5):651–7.
9. Hoa HTT, Higashi A, Yamaguchi T, Kawahara R, Calvopina M, 28. Wang X, Wang Y, Zhou Y, Li J, Yin W, Wang S, et al. Emergence of
Bastidas-Caldés A, et al. Fusion plasmid carrying the colistin resis- a novel mobile colistin resistance gene, mcr-8, in NDM-producing
tance gene mcr of Escherichia coli isolated from healthy residents. J Klebsiella pneumoniae. Emerg Microbes Infect. 2018;7(1):122.
Glob Antimicrob Resis. 2022;30:152–4. 29. Pillonetto M, Mazzetti A, Becker GN, Siebra CA, Arend LNVS, Barth
10. Ortega-Paredes D, Barba P, Zurita J. Colistin-resistant Escherichia AL. Low level of polymyxin resistance among nonclonal mcr-1–
coli clinical isolate harbouring the mcr-1 gene in Ecuador. Epidemiol positive Escherichia coli from human sources in Brazil. Diagn Micro-
Infect. 2016;144(14):2967–70. biol Infect Dis. 2019;93(2):140–2.
11. Albán MV, Núñez EJ, Zurita J, Villacís JE, Tamayo R, Sevillano G, et al. 30. Lentz SA, de Lima-Morales D, Cuppertino VM, Nunes LDS, da
Canines with different pathologies as carriers of diverse lineages of Motta AS, Zavascki AP, et al. Escherichia coli harbouring mcr-1 gene
Escherichia coli harbouring mcr-1 and clinically relevant β-lactamases isolated from poultry not exposed to polymyxins in Brazil. Euro
in central Ecuador. J Glob Antimicrob Resist. 2020;22:182–3. Surveill. 2016;21(26):pii=30267.
12. Yamamoto Y, Calvopina M, Izurieta R, Villacres I, Kawahara R, 31. Wu R, Yi LX, Yu LF, Wang J, Liu Y, Chen X, et al. Fitness advantage
Sasaki M, et al. Colistin-resistant Escherichia coli with mcr genes in of mcr-1-bearing IncI2 and IncX4 plasmids in vitro. Front Microbiol.
the livestock of rural small-scale farms in Ecuador. BMC Res Notes. 2018;9:331.
2019;12(1):121. 32. Lu S, Li D, Wang L, Bi Y, Wang M, Yang F. Promoter variations asso-
13. Montero L, Irazabal J, Cardenas P, Graham JP, Trueba G. Extended- ciated with expression of mcr-1 gene and level of colistin resistance.
spectrum beta-lactamase producing-Escherichia coli isolated Int J Antimicrob Agents. 2021;58(2):106371.
from irrigation waters and produce in Ecuador. Front Microbiol. 33. Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC,
2021;12:709418. Vandenbrouckef JP. The Strengthening the Reporting of Observa-
14. Bastidas-Caldes C, Ochoa J, Guerrero-Latorre L, Moyota-Tello tional Studies in Epidemiology (STROBE) statement: guidelines
C, Tapia W, Rey-Pérez JM, et al. Removal of extended-spectrum for reporting observational studies. Bull World Health Organ.
beta-lactamase-producing Escherichia coli, ST98, in water for human 2007;85(11):867–72.
consumption by black ceramic water filters in low-income Ecuador- 34. Butzin-Dozier Z, Waters WF, Baca M, Vinueza RL, Saraiva-Garcia
ian highlands. Int J Environ Res Public Health. 2022;19(8):4736. C, Graham J. Assessing upstream determinants of antibiotic use
15. Haziq M, Momin FA, Bean DC, Hendriksen RS, Haenni M, Phee in small-scale food animal production through a simulated client
LM, et al. CHROMagar COL-APSE: a selective bacterial culture method. Antibiotics (Basel). 2021;10(1):2.
medium for the isolation and differentiation of colistin-resistant 35. Wang Y, Xu C, Zhang R, Chen Y, Shen Y, Hu F, et al. Changes in
Gram-negative pathogens. J Med Microbiol. 2017;66:1554–61. colistin resistance and mcr-1 abundance in Escherichia coli of animal
16. Suenaga E, Nakamura H. Evaluation of three methods for effective and human origins following the ban of colistin-positive additives
extraction of DNA from human hair. J Chromatogr B Analyt Technol in China: an epidemiological comparative study. Lancet Infect Dis.
Biomed Life Sci. 2005;820(1):137–41. 2020;20(10):1161–71.
17. Yamaguchi T, Kawahara R, Harada K, Teruya S, Nakayama T,
Motooka D, et al. The presence of colistin resistance gene mcr-1 and
-3 in ESBL producing Escherichia coli isolated from food in Ho Chi
Minh City, Vietnam. FEMS Microbiol Lett. 2018;365(11):fny100.
18. QIAGEN. QIAGEN multiplex PCR handbook. Hilden (Germany):
QIAGEN; 2010.
19. Collignon P, Beggs JJ, Walsh TR, Gandra S, Laxminarayan R. Manuscript submitted 26 October 2022. Revised version accepted for publication
Anthropological and socioeconomic factors contributing to global on 29 October 2022.

Rev Panam Salud Publica 47, 2023 | www.paho.org/journal | https://doi.org/10.26633/RPSP.2023.48 7

Original research Bastidas-Caldes et al. • Colistin resistance in Ecuador

Resistencia a la colistina de las bacterias Escherichia coli y Klebsiella

pneumoniae en humanos y animales de granja en Ecuador
RESUMEN Objetivo. La colistina es un antibiótico de último recurso para tratar infecciones graves por bacterias gramneg-
ativas. Sin embargo, su uso indebido, especialmente para estimular el crecimiento animal, ha contribuido con
el aumento de la resistencia a los antimicrobianos, mediada principalmente por la transferencia de plásmidos
del gen mcr-1. En este estudio se evaluó la prevalencia de la resistencia fenotípica y molecular a la colistina
de las bacterias Escherichia coli y Klebsiella pneumoniae en humanos sanos, sus pollos y cerdos en Ecuador.
Métodos. Se recolectaron muestras fecales de humanos, así como de sus pollos y cerdos, en dos zonas
rurales de la región costera y la región amazónica entre abril y agosto del 2020. Se aislaron las bacterias
gramnegativas y se identificaron empleando técnicas convencionales. Se determinó la resistencia fenotípica
mediante la técnica de microdilución en caldo y se detectó el gen mcr-1 con la técnica convencional de reac-
ción en cadena de la polimerasa.
Resultados. Se obtuvo un total de 438 muestras fecales de 137 humanos, 147 cerdos y 154 pollos. La preva-
lencia de E. coli en las cepas aisladas fue del 86,3% (378/438) y la de K. pneumoniae, del 37,4% (164/438). En
general, se detectó el gen mcr-1 en el 90% (340/378) de las cepas aisladas de E. coli y la mayor prevalencia
encontrada fue en cepas aisladas de la región costera (96,5%, 191/198), humanos (95,6%, 111/116) y pollos
(91,8%, 123/134); en el caso de K. pneumoniae, el gen se encontró en el 19,5% (32/164) de las cepas, con
una distribución equitativa entre regiones y hospedadores. Únicamente cuatro cepas aisladas, dos de E. coli
y dos de K. pneumoniae, mostraron resistencia fenotípica: el gen mcr-1 estaba presente en ambas cepas de
E. coli y ausente en las cepas de K. pneumoniae.
Conclusiones. Si bien hubo una baja prevalencia de resistencia fenotípica a la colistina, la alta prevalencia
del gen mcr-1 en E. coli es preocupante. Es necesario hacer cumplir la prohibición del uso de colistina en la
cría de animales en Ecuador, así como realizar un seguimiento continuo de la situación.

Palabras clave Colistina; Escherichia coli; Klebsiella pneumoniae; humanos; animales; resistencia a medicamentos; genes
MDR; investigación operativa; Ecuador.

Resistência à colistina em Escherichia coli e Klebsiella pneumoniae em

humanos e animais de quintal no Equador
RESUMO Objetivo. A colistina é um antibiótico de último recurso para o tratamento de infecções graves por bac-
térias Gram-negativas. Entretanto, o uso indevido da colistina, principalmente como promotor de crescimento
animal, tem contribuído para o aumento da resistência a antimicrobianos, principalmente por transferência
horizontal do gene mcr-1 mediada por plasmídeos. Este estudo avaliou a prevalência de resistência fenotípica
e molecular à colistina em Escherichia coli e Klebsiella pneumoniae no Equador em humanos hígidos e em
galinhas e porcos por eles criados.
Métodos. Entre abril e agosto de 2020, foram coletadas amostras de fezes de habitantes de duas regiões
litorâneas e amazônicas do Equador e de galinhas e porcos por eles criados. Bactérias Gram-negativas
foram isoladas e identificadas por meio de técnicas convencionais. A resistência fenotípica foi determinada
pela técnica de microdiluição em caldo, e o gene mcr-1 foi detectado por reação em cadeia da polimerase
convencional.
Resultados. Foram obtidas 438 amostras fecais de 137 humanos, 147 suínos e 154 galinhas. A prevalência
de isolados de E. coli foi de 86,3% (378/438), e de K. pneumoniae, 37,4% (164/438). Em geral, o gene mcr-1
foi encontrado em 90% (340/378) dos isolados de E. coli, com maiores prevalências encontradas em isola-
dos de regiões litorâneas (96,5%, 191/198), humanos (95,6%, 111/116) e galinhas (91,8%, 123/134); para K.
pneumoniae, o gene foi encontrado em 19,5% (32/164) dos isolados, com igual distribuição entre regiões e
hospedeiros. Somente quatro isolados, dois de E. coli e dois de K. pneumoniae, demonstraram resistência
fenotípica: o gene mcr-1 estava presente em ambas as cepas de E. coli, mas ausente nas de K. pneumoniae.
Conclusões. Apesar da baixa prevalência de resistência fenotípica à colistina, a alta prevalência do gene
mcr-1 em E. coli é preocupante. É preciso fiscalizar a proibição ao uso agropecuário de colistina no Equador
e implementar o monitoramento contínuo da situação.

Palavras-chave Colistina; Escherichia coli; Klebsiella pneumoniae; humanos; animais; resistência a medicamentos; genes
MDR; Equador.

8 Rev Panam Salud Publica 47, 2023 | www.paho.org/journal | https://doi.org/10.26633/RPSP.2023.48