Subscriber access provided by UNIV OF NEBRASKA - LINCOLN
Critical Review
Role of the environment in the transmission of antimicrobial resistance to humans: A review Patricia Huijbers, Hetty Blaak, Mart C. M. de Jong, Elisabeth A. M. Graat, Christina M. J. E. Vandenbroucke-Grauls, and Ana Maria De Roda Husman Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b02566 • Publication Date (Web): 10 Sep 2015 Downloaded from http://pubs.acs.org on September 15, 2015
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 44
Environmental Science & Technology
1
Role of the environment in the transmission of antimicrobial resistance to humans: A
2
review
3 4
Patricia M.C. Huijbersa,b, Hetty Blaakb,*, Mart C.M. de Jonga, Elisabeth A.M. Graata,
5
Christina M.J.E. Vandenbroucke-Graulsc, Ana Maria de Roda Husmanb,d
6 7
a
8
(WIAS), Wageningen University, P.O. Box 338, 6700 AH Wageningen, The Netherlands
9
b
Quantitative Veterinary Epidemiology Group, Wageningen Institute of Animal Sciences
Centre for Zoonoses and Environmental Microbiology, Centre for Infectious Disease
10
Control (CIb), National Institute for Public Health and the Environment (RIVM), P.O. Box 1,
11
3720 BA Bilthoven, The Netherlands
12
c
13
P.O. Box 7057, 1007 MB Amsterdam, The Netherlands
14
d
15
University, P.O. Box 80178, 3508 TD Utrecht, The Netherlands
Department of Medical Microbiology and Infection Control, VU University Medical Center,
Faculty of Veterinary Medicine, Institute for Risk Assessment Sciences (IRAS), Utrecht
16
17
* Corresponding author
18
Address: Hetty Blaak, Centre for Zoonoses and Environmental Microbiology, Centre for
19
Infectious Disease Control, National Institute for Public Health and the Environment, P.O.
20
Box 1, 3720 BA Bilthoven, The Netherlands
21
E-mail: hetty.blaak@rivm.nl
22
Telephone: +31 (0)30 2747005
1 ACS Paragon Plus Environment
Environmental Science & Technology
Page 2 of 44
23
Abstract
24
To establish a possible role for the natural environment in the transmission of clinically
25
relevant AMR bacteria to humans, a literature review was conducted to systematically collect
26
and categorize evidence for human exposure to extended-spectrum β-lactamase-producing
27
Enterobacteriaceae, methicillin-resistant Staphylococcus aureus, and vancomycin-resistant
28
Enterococcus spp. in the environment. In total, 239 datasets adhered to inclusion criteria.
29
AMR bacteria were detected at exposure-relevant sites (35/38), including recreational areas,
30
drinking water, ambient air, shellfish, and in fresh produce (8/16). More datasets were
31
available for environmental compartments (139/157), including wildlife, water, soil, and
32
air/dust. Quantitative data from exposure-relevant sites (6/35), and environmental
33
compartments (11/139) were scarce. AMR bacteria were detected in the contamination
34
sources (66/66) wastewater and manure, and molecular data supporting their transmission
35
from wastewater to the environment (1/66) were found. The abundance of AMR bacteria at
36
exposure-relevant sites suggests risk for human exposure. Of publications pertaining to both
37
environmental and human isolates, however, only one compared isolates from samples that
38
had a clear spatial and temporal relationship, and no direct evidence was found for
39
transmission to humans through the environment. To what extent the environment, compared
40
to the clinical and veterinary domains, contributes to human exposure needs to be quantified.
41
AMR bacteria in the environment, including sites relevant for human exposure, originate
42
from contamination sources. Intervention strategies targeted at these sources could therefore
43
limit emission of AMR bacteria to the environment.
44 45
Keywords: ESBL, MRSA, VRE, soil, water, air, wildlife, wastewater, manure
46 47
2 ACS Paragon Plus Environment
Page 3 of 44
Environmental Science & Technology
48
Introduction
49
The occurrence and spread of antimicrobial resistant (AMR) bacteria are pressing public
50
health problems worldwide. Recently, in its first global report on surveillance of
51
antimicrobial resistance, the World Health Organization (WHO) reported very high rates of
52
resistance in bacteria (e.g. Escherichia coli, Klebsiella pneumoniae, Staphylococcus aureus)
53
that cause common healthcare-associated and community-acquired infections in people in all
54
WHO regions.1 Infections caused by AMR bacteria are associated with excess mortality,
55
prolongation of hospital stay, and increased costs.2 In order to formulate effective
56
intervention strategies to combat intractable infections caused by AMR bacteria, it is
57
important to discern which fraction of the total disease burden and costs are attributable to
58
different sources, including animal reservoirs and vehicles such as foods, or the
59
environment.3
60
Enteric bacteria are introduced into the environment with human and animal feces,
61
and people may be exposed to these bacteria through e.g. recreation in contaminated surface
62
water, consumption of contaminated drinking water, fresh produce or (shell)fish, and
63
inhalation of bioaerosols. Previous studies indicate that the risk of contracting Salmonella
64
spp. or Campylobacter spp. in recreational waters is higher than or equal to the risk of
65
contracting the organisms through chicken consumption.4,5,6 In the USA, 9% of all outbreaks
66
with the pathogenic E. coli O157 are waterborne.7 Based on global WHO risk assessments, it
67
was estimated that there are over 120 million cases annually of gastrointestinal disease from
68
exposure to coastal waters via recreation or by eating raw or lightly cooked shellfish.8
69
Multiple studies have described outbreaks of infections with Enterobacteriaceae associated
70
with consumption of fresh produce.9 Examples include the 2011 outbreak of E. coli O104:H4
71
associated with sprouts in Europe and Northern America,10 and the 2006 outbreak of E. coli
72
O157:H7 associated with spinach in the USA.11
3 ACS Paragon Plus Environment
Environmental Science & Technology
Page 4 of 44
73
The natural environment has been identified as a pathway by which transmission of
74
AMR bacteria to humans might occur.12 The extent to which this occurs remains unknown,
75
however. It is important to establish the relative role of the environment in the transmission
76
of AMR bacteria to humans, compared with the spread of AMR bacteria through contact with
77
animal carriers, consumption of food of animal origin, (international) travel, and their spread
78
in healthcare and community settings. The aim of the current study was to establish a possible
79
role for the natural environment, defined as soil, water, air/dust, and wildlife in the
80
transmission of clinically relevant AMR bacteria to humans by systematically reviewing the
81
peer-reviewed literature. For this purpose, three AMR bacteria were selected: extended-
82
spectrum β-lactamase (ESBL)-producing Enterobacteriaceae (ESBL-Ent), methicillin-
83
resistant S. aureus (MRSA), and vancomycin-resistant Enterococcus spp. (VRE).
84
Additionally, the contribution of fecal contamination sources (human wastewater and animal
85
manure) to the burden of the AMR bacteria in the environment was explored.
86
87
Methods
88 89
Search strategy
90
Two databases (Medline and Scopus) were searched on April 10th, 2014 to identify
91
publications describing one or more of three AMR bacteria in relation to the environment.
92
The bacterial species were selected to represent Gram-negative and Gram-positive species,
93
and species of fecal and non-fecal origin. Moreover, each of the bacterial species colonizes
94
both animals and humans, has a clinically relevant type of resistance, and is sufficiently
95
prevalent in humans and/or animals to anticipate its presence in the environment.
96
Accordingly, ESBL-Ent (i.e. E. coli, Enterobacter spp., K. pneumoniae), MRSA, and VRE
97
were selected, and carbapenemase-producing Enterobacteriaceae were not selected.
4 ACS Paragon Plus Environment
Page 5 of 44
Environmental Science & Technology
98
For the purpose of the present review, environment was defined as natural
99
environment (or ‘outdoor’ environment). Consequently, indoor environments such as
100
hospitals or livestock housing were excluded. Four environmental compartments were
101
distinguished: soil, water, air/dust and wildlife (Figure 1). Wildlife was categorized as an
102
environmental compartment because these animals are not treated with antibiotics, and
103
carriage of AMR bacteria is most likely explained by uptake from the natural environment
104
during foraging and drinking. Wildlife can therefore be considered as an extension of the
105
specified environmental compartments, as well as a vehicle for multiplication and spread of
106
these bacteria. Within the environmental compartments, sites relevant for human exposure
107
were defined and included in the search strategy: recreational areas (i.e. beach sand and
108
recreational water), irrigation water, drinking water, urban water (e.g. fountains), shellfish
109
and ambient air. Fresh produce i.e. food of plant origin such as vegetables and fruits, may be
110
contaminated with AMR bacteria during growth in contaminated soil and/or irrigation with
111
contaminated water. Consequently, this was also included as a ‘site’ relevant for human
112
exposure. Finally, the main sources of fecal contamination of the environment, wastewater
113
and manure, were included in the search strategy. For the purpose of the current study,
114
manure was defined as animal feces that is intentionally (e.g. for use as fertilizer) or
115
unintentionally (e.g. droppings of free-range animals) introduced into the natural
116
environment, or that is stored, presumably for use as fertilizer. Consequently, only
117
measurements of AMR bacteria in animal feces introduced in natural environments (e.g. not
118
fecal swabs or droppings sampled in stables) were included.
119
Combinations of key search terms for AMR bacteria and defined environmental
120
compartments, contamination sources and exposure-relevant sites, were used to interrogate
121
the online databases (Table 1). Specifically, the titles, abstracts and key words of publications
122
included in the online databases were screened for these key search terms.
5 ACS Paragon Plus Environment
Environmental Science & Technology
Page 6 of 44
123 124
Selection criteria
125
Two reviewers independently screened titles and abstracts, resulting from the automated
126
database search, for exclusion criteria. Publications were excluded if they were in a language
127
other than English, Dutch or German, contained non-original research (e.g. reviews), or did
128
not investigate the specified compartments or AMR bacteria (Figure 2). Publications solely
129
describing resistance genes (i.e. genes encoding ESBL or resistance to methicillin or
130
vancomycin), independently of bacterial hosts, were excluded. In case of conflict of opinion
131
about inclusion of an article based on title or abstract, this was discussed by both reviewers
132
until consensus was reached. Full texts were retrieved for included publications, and for
133
publications where the decision for inclusion or exclusion could not be based on the contents
134
of title and abstract, or abstracts were not available. Full-texts were further assessed for
135
compliance with the inclusion and exclusion criteria (Figure 2A).
136 137
Data extraction
138
Study characteristics and data from included publications were extracted and exported to a
139
database using a custom-made form. Full details of the extraction fields are available in Table
140
S1 (Supporting Information). In case multiple publications described the same or overlapping
141
sets of samples or isolates, these publications were considered as one data source i.e. one
142
publication. Publications were categorized based on the type of investigated compartment
143
(contamination sources, environmental compartments, exposure-relevant sites) and the type
144
of AMR bacteria. Individual publications could be included within more than one category.
145
To ensure the quality of the dataset, the method used to isolate AMR bacteria was
146
assessed. Firstly, it was determined whether selective culture methods were used to detect the
147
AMR bacteria. Selective culture was defined as culture (pre-enrichment or direct plating) in
6 ACS Paragon Plus Environment
Page 7 of 44
Environmental Science & Technology
148
the presence of relevant antibiotics: 3rd generation cephalosporin-supplemented medium or
149
commercial screening medium for the isolation of ESBL-Ent; methicillin-, oxacillin- or
150
cefoxitin-supplemented medium or commercial screening medium for MRSA; and
151
vancomycin-supplemented medium for VRE. The rationale for this assessment is the lack of
152
sensitivity of non-selective culturing methods for the detection of AMR bacteria, and the
153
resulting unreliability for drawing conclusions based on the lack of detection when non-
154
selective culturing was used. Secondly, it was determined whether the antibiotic resistance
155
was confirmed using phenotypic and/or molecular methods. Phenotypic confirmation tests
156
deemed valid included: growth of isolates on media containing concentrations equal to or
157
above clinical breakpoints, established Minimal Inhibitory Concentrations (MICs) equal or
158
above clinical breakpoints or, in the case of disc diffusion assays, inhibition zones smaller
159
than clinical breakpoints as described in the Clinical and Laboratory Standards Institute
160
(CLSI) guidelines.13 VanB-carrying VRE variants with MICs below the current CLSI break-
161
point concentration (32 µg/ml) have been described,14 therefore growth at concentrations >16
162
µg/ml was also considered a valid phenotypic confirmation. Polymerase chain reactions
163
(PCRs) detecting mecA, and vanA or vanB genes were considered valid molecular tests for
164
MRSA and VRE, respectively. For ESBL-Ent, PCRs detecting CTX-M-genes, PCRs
165
detecting TEM-, SHV- and OXA-genes in combination with sequencing to establish
166
subtypes, or PCRs detecting other ESBL-genes were considered valid molecular confirmation
167
tests. Sequencing of TEM-, SHV-, and OXA-genes is required since non-ESBL-encoding
168
alleles exist of these genes. Studies in which AMR bacteria were investigated, but not
169
detected using only non-selective culture methods, or where resistance of isolates was not
170
confirmed appropriately, were excluded from analysis.
171
For each publication, it was assessed whether AMR bacteria were enumerated, and
172
whether environmental isolates were compared with isolates from fecal contamination
7 ACS Paragon Plus Environment
Environmental Science & Technology
Page 8 of 44
173
sources or with human isolates for phenotypic or genetic relatedness. From the perspective of
174
the current review, i.e. study of documentation on transmission from the environment to
175
humans, and emission from contamination sources to the environment, only comparisons
176
between isolates from samples with a spatiotemporal relationship were considered relevant.
177
Tests frequently used for establishing relatedness of isolates included: pulsed-field gel
178
electrophoresis (PFGE), multilocus sequence typing (MLST), antibiotic resistance profiling,
179
Staphylococcus protein A gene (spa)-typing (specifically for MRSA), and PhenePlate (PhP)
180
biochemical fingerprinting (specifically for VRE). Additional tests included those identifying
181
resistance genes and their genetic environment e.g. identification of ESBL-genotype and
182
plasmid characterization for ESBL-Ent, and staphylococcal cassette chromosome SCCmec-
183
typing for MRSA.
184
185
Results
186
General
187
The selection process yielded 241 publications that met the inclusion criteria (Figure 2A), and
188
from these publications data were extracted. Fifteen publications described the same or
189
overlapping samples or isolates, and these were combined to represent seven publications.
190
Some publications described multiple AMR bacteria, environmental compartments,
191
contamination sources, and/or exposure-relevant sites. For further study, data from included
192
publications were divided into datasets, with each dataset describing one type of AMR
193
bacteria (i.e. ESBL-Ent, MRSA, VRE) in combination with one type of environmental
194
compartment (i.e. soil, water, air/dust, wildlife), contamination source (i.e. wastewater,
195
manure) or fresh produce. The 241 included publications resulted in 312 datasets (Figure 2B).
196
After applying the defined quality criteria, 64 datasets were excluded. For 9 datasets, the
197
results could not be interpreted due to lack of specification in which of the studied
8 ACS Paragon Plus Environment
Page 9 of 44
Environmental Science & Technology
198
compartment(s) AMR bacteria were found. These had not been not excluded earlier in the
199
selection process because the publications also contained other, valid datasets. This left 239
200
datasets for further study. Details on publications describing these datasets are available in
201
Table S2 (Supporting Information). The majority of datasets described wildlife (n=71),
202
wastewater (mostly municipal or urban sewage) (n=60), and water (n=56), followed by soil
203
(n=25), fresh produce (n=16), manure (n=6), and air/dust (n=5). ESBL-Ent (n=102), and
204
VRE (n=88) were described more often compared to MRSA (n=49).
205
All continents were represented, however more than half of the datasets described
206
isolates collected in Europe (n=139), followed by North America (n=39), Asia (n=38), Africa
207
(n=9), South America (n=8), Oceania (n=3), and the Antarctic (n=2). One dataset contained
208
isolates that were collected from both Europe and Asia.
209 210
Exposure-relevant sites and fresh produce
211
Of 157 datasets concerning environmental compartments (Table 2), 24% (38/157) included
212
sites that are relevant for human exposure (Table 3). Of these, almost two thirds were about
213
recreational water (n=15), or beach sand (n=12). The remaining datasets were about drinking
214
water (n=5), ambient air (n=3), shellfish (n=2), and irrigation water (n=1). No datasets
215
pertaining to urban water (e.g. fountains) were identified. Furthermore, 16 datasets concerned
216
fresh produce such as vegetables, fruits, sprouts and herbs (Table 3). Overall, AMR bacteria
217
were detected in 92% (35/38) of datasets about exposure-relevant sites, and 50% (8/16) of
218
datasets about fresh produce.
219
Only six datasets concerning exposure-relevant sites included quantitative data on the
220
AMR bacteria (Table 4). Five datasets concerned concentrations in recreational areas,
221
ranging from 100.1-103 CFU/100ml.15-18 One dataset provided information on concentrations
222
in blue mussels, which was