Critical Orientation in the Jungle of Currently Available Methods and Types of Data for Source Attribution of Foodborne Diseases

Livestock-associated spatial risk factors for human salmonellosis and campylobacteriosis

AIMS

Most human infections with non-typhoid Salmonella (NTS) or Campylobacter are zoonotic in nature and acquired though consumption of contaminated food of mainly animal origin. However, individuals may also acquire salmonellosis or campylobacteriosis through non-foodborne transmission pathways, such as those mediated by the environment. This emphasizes the need to consider both direct and indirect exposure to livestock sources as a possible transmission route for NTS and Campylobacter. Therefore, this study aimed at assessing whether salmonellosis and campylobacteriosis incidence is spatially associated with exposure to livestock (i.e. small ruminants, dairy cows, veal calves, laying hens, broiler chickens and pigs) in the Netherlands for the years 2007-2019 and 2014-2019 respectively.

METHODS AND RESULTS

Risk factors (population-weighted number of animals) and their population attributable fractions were determined using a Poisson regression model with a log-link function fitted using integrated nested Laplace approximation. The analyses were performed for different hexagonal sizes (90, 50, 25 and 10 km2) and accounted for geographical coverage of the diagnostic laboratory catchment areas. Moreover, serological data were used to look into the possible effects of acquired immunity due to repeated exposure to the pathogen through the environment that would potentially hinder the analyses based on the incidence of reported cases. A linear mixed-effects model was then fitted where the postal code areas were included as a random effect. Livestock was not consistently significantly associated with acquiring salmonellosis or campylobacteriosis in the Netherlands.

CONCLUSIONS

Results showed that living in livestock-rich areas in the Netherlands is not a consistently significant, spatially restricted risk factor for acquiring salmonellosis or campylobacteriosis, thereby supporting current knowledge that human infections with Salmonella and Campylobacter are mainly foodborne.

Trends in Salmonella Infantis human illness incidence and chicken carcass prevalence in the United States; 1996-2019

The incidence of human illness due to Salmonella Infantis reported to Foodborne Diseases Active Surveillance Network and the prevalence of Infantis on chicken carcasses reported by the United States Department of Agriculture Food Safety and Inspection Service have increased significantly in the past decade. However, the trends do not appear coincident, as would be expected if the increased prevalence in chicken led to the increase in the incidence of human illness. Salmonella Infantis incidence and prevalence trends are analyzed using penalized B-spline methods for generalized additive regression models. The association between the two time series is analyzed using time-lagged rank-order cross-correlation. Geographic variations in reported incidence and trends are also explored. The increase in human incidence of Salmonella Infantis began circa 2011. The increase in chicken carcass prevalence began circa 2015. A 4-year lag on chicken carcass prevalence maximizes the rank-order cross-correlation with the incidence of illness. While chicken consumption undoubtedly contributes to the incidence of human illness due to Salmonella Infantis, the initial increase in reported illness was likely due to one or more other transmission pathways. Other potential transmission pathways include non-chicken foodborne, waterborne, person-to-person, animal contact, and environmental.

Use of whole genome sequencing for surveillance and control of foodborne diseases: status quo and quo vadis

Improvements in sequencing quality, availability, speed and costs results in an increased presence of genomics in infectious disease applications. Nevertheless, there are still hurdles in regard to the optimal use of WGS for public health purposes. Here, we discuss the current state ("status quo") and future directions ("quo vadis") based on literature regarding the use of genomics in surveillance, hazard characterization and source attribution of foodborne pathogens. The future directions include the application of new techniques, such as machine learning and network approaches that may overcome the current shortcomings. These include the use of fixed genomic distances in cluster delineation, disentangling similarity or lack thereof in source attribution, and difficulties ascertaining function in hazard characterization. Although, the aforementioned methods can relatively easily be applied technically, an overarching challenge is the inference and biological/epidemiological interpretation of these large amounts of high-resolution data. Understanding the context in terms of bacterial isolate and host diversity allows to assess the level of representativeness in regard to sources and isolates in the dataset, which in turn defines the level of certainty associated with defining clusters, sources and risks. This also marks the importance of metadata (clinical, epidemiological, and biological) when using genomics for public health purposes.

Development and validation of a random forest algorithm for source attribution of animal and human Salmonella Typhimurium and monophasic variants of S. Typhimurium isolates in England and Wales utilising whole genome sequencing data

Source attribution has traditionally involved combining epidemiological data with different pathogen characterisation methods, including 7-gene multi locus sequence typing (MLST) or serotyping, however, these approaches have limited resolution. In contrast, whole genome sequencing data provide an overview of the whole genome that can be used by attribution algorithms. Here, we applied a random forest (RF) algorithm to predict the primary sources of human clinical Salmonella Typhimurium (S. Typhimurium) and monophasic variants (monophasic S. Typhimurium) isolates. To this end, we utilised single nucleotide polymorphism diversity in the core genome MLST alleles obtained from 1,061 laboratory-confirmed human and animal S. Typhimurium and monophasic S. Typhimurium isolates as inputs into a RF model. The algorithm was used for supervised learning to classify 399 animal S. Typhimurium and monophasic S. Typhimurium isolates into one of eight distinct primary source classes comprising common livestock and pet animal species: cattle, pigs, sheep, other mammals (pets: mostly dogs and horses), broilers, layers, turkeys, and game birds (pheasants, quail, and pigeons). When applied to the training set animal isolates, model accuracy was 0.929 and kappa 0.905, whereas for the test set animal isolates, for which the primary source class information was withheld from the model, the accuracy was 0.779 and kappa 0.700. Subsequently, the model was applied to assign 662 human clinical cases to the eight primary source classes. In the dataset, 60/399 (15.0%) of the animal and 141/662 (21.3%) of the human isolates were associated with a known outbreak of S. Typhimurium definitive type (DT) 104. All but two of the 141 DT104 outbreak linked human isolates were correctly attributed by the model to the primary source classes identified as the origin of the DT104 outbreak. A model that was run without the clonal DT104 animal isolates produced largely congruent outputs (training set accuracy 0.989 and kappa 0.985; test set accuracy 0.781 and kappa 0.663). Overall, our results show that RF offers considerable promise as a suitable methodology for epidemiological tracking and source attribution for foodborne pathogens.

A Review on Microbiological Source Attribution Methods of Human Salmonellosis: From Subtyping to Whole-Genome Sequencing

Salmonella is one of the main causes of human foodborne illness. It is endemic worldwide, with different animals and animal-based food products as reservoirs and vehicles of infection. Identifying animal reservoirs and potential transmission pathways of Salmonella is essential for prevention and control. There are many approaches for source attribution, each using different statistical models and data streams. Some aim to identify the animal reservoir, while others aim to determine the point at which exposure occurred. With the advance of whole-genome sequencing (WGS) technologies, new source attribution models will greatly benefit from the discriminating power gained with WGS. This review discusses some key source attribution methods and their mathematical and statistical tools. We also highlight recent studies utilizing WGS for source attribution and discuss open questions and challenges in developing new WGS methods. We aim to provide a better understanding of the current state of these methodologies with application to Salmonella and other foodborne pathogens that are common sources of illness in the poultry and human sectors.

Using SNP addresses for Salmonella Typhimurium DT104 in routine veterinary outbreak detection

SNP addresses are a pathogen typing method based on whole-genome sequences (WGSs), assigning groups at seven different levels of genetic similarity. Public health surveillance uses it for several gastro-intestinal infections; this work trialled its use in veterinary surveillance for salmonella outbreak detection. Comparisons were made between temporal and spatio-temporal cluster detection models that either defined cases by their SNP address or by phage type, using historical data sets. Clusters of SNP incidents were effectively detected by both methods, but spatio-temporal models consistently detected these clusters earlier than the corresponding temporal models. Unlike phage type, SNP addresses appeared spatially and temporally limited, which facilitated the differentiation of novel, stable, or expanding clusters in spatio-temporal models. Furthermore, these models flagged spatio-temporal clusters containing only two to three cases at first detection, compared with a median of seven cases in phage-type models. The large number of SNP addresses will require automated methods to implement these detection models routinely. Further work is required to explore how temporal changes and different host species may impact the sensitivity and specificity of cluster detection. In conclusion, given validation with more sequencing data, SNP addresses are likely to be a valuable addition to early warning systems in veterinary surveillance.

Animal sources of antimicrobial-resistant bacterial infections in humans: a systematic review

Bacterial antimicrobial resistance (AMR) is among the leading global health challenges of the century. Animals and their products are known contributors to the human AMR burden, but the extent of this contribution is not clear. This systematic literature review aimed to identify studies investigating the direct impact of animal sources, defined as livestock, aquaculture, pets, and animal-based food, on human AMR. We searched four scientific databases and identified 31 relevant publications, including 12 risk assessments, 16 source attribution studies, and three other studies. Most studies were published between 2012 and 2022, and most came from Europe and North America, but we also identified five articles from South and South-East Asia. The studies differed in their methodologies, conceptual approaches (bottom-up, top-down, and complex), definitions of the AMR hazard and outcome, the number and type of sources they addressed, and the outcome measures they reported. The most frequently addressed animal source was chicken, followed by cattle and pigs. Most studies investigated bacteria-resistance combinations. Overall, studies on the direct contribution of animal sources of AMR are rare but increasing. More recent publications tailor their methodologies increasingly towards the AMR hazard as a whole, providing grounds for future research to build on.

Comparison of Source Attribution Methodologies for Human Campylobacteriosis

Campylobacter spp. are the most common cause of bacterial gastrointestinal infection in humans both in Denmark and worldwide. Studies have found microbial subtyping to be a powerful tool for source attribution, but comparisons of different methodologies are limited. In this study, we compare three source attribution approaches (Machine Learning, Network Analysis, and Bayesian modeling) using three types of whole genome sequences (WGS) data inputs (cgMLST, 5-Mers and 7-Mers). We predicted and compared the sources of human campylobacteriosis cases in Denmark. Using 7mer as an input feature provided the best model performance. The network analysis algorithm had a CSC value of 78.99% and an F1-score value of 67%, while the machine-learning algorithm showed the highest accuracy (98%). The models attributed between 965 and all of the 1224 human cases to a source (network applying 5mer and machine learning applying 7mer, respectively). Chicken from Denmark was the primary source of human campylobacteriosis with an average percentage probability of attribution of 45.8% to 65.4%, representing Bayesian with 7mer and machine learning with cgMLST, respectively. Our results indicate that the different source attribution methodologies based on WGS have great potential for the surveillance and source tracking of Campylobacter. The results of such models may support decision makers to prioritize and target interventions.

Genomic insights into zoonotic transmission and antimicrobial resistance in Campylobacter jejuni from farm to fork: a one health perspective

BACKGROUND

Campylobacteriosis represents a global public health threat with various socio-economic impacts. Among different Campylobacter species, Campylobacter jejuni (C. jejuni) is considered to be the foremost Campylobacter species responsible for most of gastrointestinal-related infections. Although these species are reported to primarily inhabit birds, its high genetic and phenotypic diversity allowed their adaptation to other animal reservoirs and to the environment that may impact on human infection.

MAIN BODY

A stringent and consistent surveillance program based on high resolution subtyping is crucial. Recently, different epidemiological investigations have implemented high-throughput sequencing technologies and analytical pipelines for higher resolution subtyping, accurate source attribution, and detection of antimicrobial resistance determinants among these species. In this review, we aim to present a comprehensive overview on the epidemiology, clinical presentation, antibiotic resistance, and transmission dynamics of Campylobacter, with specific focus on C. jejuni. This review also summarizes recent attempts of applying whole-genome sequencing (WGS) coupled with bioinformatic algorithms to identify and provide deeper insights into evolutionary and epidemiological dynamics of C. jejuni precisely along the farm-to-fork continuum.

CONCLUSION

WGS is a valuable addition to traditional surveillance methods for Campylobacter. It enables accurate typing of this pathogen and allows tracking of its transmission sources. It is also advantageous for in silico characterization of antibiotic resistance and virulence determinants, and hence implementation of control measures for containment of infection.

Polyphyly in widespread Salmonella enterica serovars and using genomic proximity to choose the best reference genome for bioinformatics analyses

Salmonella is the most common cause of gastroenteritis in the world. Over the past 5 years, whole-genome analysis has led to the high-resolution characterization of clinical and foodborne Salmonella responsible for typhoid fever, foodborne illness or contamination of the agro-food chain. Whole-genome analyses are simplified by the availability of high-quality, complete genomes for mapping analysis and for calculating the pairwise distance between genomes, but unfortunately some difficulties may still remain. For some serovars, the complete genome is not available, or some serovars are polyphyletic and knowing the serovar alone is not sufficient for choosing the most appropriate reference genome. For these serovars, it is essential to identify the genetically closest complete genome to be able to carry out precise genome analyses. In this study, we explored the genomic proximity of 650 genomes of the 58 Salmonella enterica subsp. enterica serovars most frequently isolated in humans and from the food chain in the United States (US) and in Europe (EU), with a special focus on France. For each serovar, to take into account their genomic diversity, we included all the multilocus sequence type (MLST) profiles represented in EnteroBase with 10 or more genomes (on 19 July 2021). A phylogenetic analysis using both core- and pan-genome approaches was carried out to identify the genomic proximity of all the Salmonella studied and 20 polyphyletic serovars that have not yet been described in the literature. This study determined the genetic proximity between all 58 serovars studied and revealed polyphyletic serovars, their genomic lineages and MLST profiles. Finally, we enhanced the open-access databases with 73 new genomes and produced a list of high-quality complete reference genomes for 48 S. enterica subsp. enterica serovars among the most isolated in the US, EU, and France.

A statistical modelling approach for source attribution meta-analysis of sporadic infection with foodborne pathogens

Numerous source attribution studies for foodborne pathogens based on epidemiological and microbiological methods are available. These studies provide empirical data for modelling frameworks that synthetize the quantitative evidence at our disposal and reduce reliance on expert elicitations. Here, we develop a statistical model within a Bayesian estimation framework to integrate attribution estimates from expert elicitations with estimates from microbial subtyping and case-control studies for sporadic infections with four major bacterial zoonotic pathogens in the Netherlands (Campylobacter, Salmonella, Shiga toxin-producing E. coli [STEC] O157 and Listeria). For each pathogen, we pooled the published fractions of human cases attributable to each animal reservoir from the microbial subtyping studies, accounting for the uncertainty arising from the different typing methods, attribution models, and year(s) of data collection. We then combined the population attributable fractions (PAFs) from the case-control studies according to five transmission pathways (domestic food, environment, direct animal contact, human-human transmission and travel) and 11 groups within the foodborne pathway (beef/lamb, pork, poultry meat, eggs, dairy, fish/shellfish, fruit/vegetables, beverages, grains, composite foods and food handlers/vermin). The attribution estimates were biologically plausible, allowing the human cases to be attributed in several ways according to reservoirs, transmission pathways and food groups. All pathogens were predominantly foodborne, with Campylobacter being mostly attributable to the chicken reservoir, Salmonella to pigs (albeit closely followed by layers), and Listeria and STEC O157 to cattle. Food-wise, the attributions reflected those at the reservoir level in terms of ranking. We provided a modelling solution to reach consensus attribution estimates reflecting the empirical evidence in the literature that is particularly useful for policy-making and is extensible to other pathogens and domains.

Salmonella enterica subsp. enterica Welikade: guideline for phylogenetic analysis of serovars rarely involved in foodborne outbreaks

BACKGROUND

Salmonella spp. is a major foodborne pathogen with a wide variety of serovars associated with human cases and food sources. Nevertheless, in Europe a panel of ten serovars is responsible for up to 80% of confirmed human cases. Clustering studies by single nucleotide polymorphism (SNP) core-genome phylogenetic analysis of outbreaks due to these major serovars are simplified by the availability of many complete genomes in the free access databases. This is not the case for outbreaks due to less common serovars, such as Welikade, for which no reference genomes are available. In this study, we propose a method to solve this problem. We propose to perform a core genome MLST (cgMLST) analysis based on hierarchical clustering using the free-access EnteroBase to select the most suitable genome to use as a reference for SNP phylogenetic analysis. In this study, we applied this protocol to a retrospective analysis of a Salmonella enterica serovar Welikade (S. Welikade) foodborne outbreak that occurred in France in 2016. Finally, we compared the cgMLST and SNP analyses. SNP phylogenetic reconstruction was carried out considering the effect of recombination events identified by the ClonalFrameML tool. The accessory genome was also explored by phage content and virulome analyses.

RESULTS

Our findings revealed high clustering concordance using cgMLST and SNP analyses. Nevertheless, SNP analysis allowed for better assessment of the genetic distance among strains. The results revealed epidemic clones of S. Welikade circulating within the poultry and dairy sectors in France, responsible for sporadic and non-sporadic human cases between 2012 and 2019.

CONCLUSIONS

This study increases knowledge on this poorly described serovar and enriches public genome databases with 42 genomes from human and non-human S. Welikade strains, including the isolate collected in 1956 in Sri Lanka, which gave the name to this serovar. This is the first genomic analysis of an outbreak due to S. Welikade described to date.

Clinical Metagenomics Is Increasingly Accurate and Affordable to Detect Enteric Bacterial Pathogens in Stool

Stool culture is the gold standard method to diagnose enteric bacterial infections; however, many clinical laboratories are transitioning to syndromic multiplex PCR panels. PCR is rapid, accurate, and affordable, yet does not yield subtyping information critical for foodborne disease surveillance. A metagenomics-based stool testing approach could simultaneously provide diagnostic and public health information. Here, we evaluated shotgun metagenomics to assess the detection of common enteric bacterial pathogens in stool. We sequenced 304 stool specimens from 285 patients alongside routine diagnostic testing for Salmonella spp., Campylobacter spp., Shigella spp., and shiga-toxin producing Escherichia coli. Five analytical approaches were assessed for pathogen detection: microbiome profiling, Kraken2, MetaPhlAn, SRST2, and KAT-SECT. Among analysis tools and databases compared, KAT-SECT analysis provided the best sensitivity and specificity for all pathogens tested compared to culture (91.2% and 96.2%, respectively). Where metagenomics detected a pathogen in culture-negative specimens, standard PCR was positive 85% of the time. The cost of metagenomics is approaching the current combined cost of PCR, reflex culture, and whole genome sequencing for pathogen detection and subtyping. As cost, speed, and analytics for single-approach metagenomics improve, it may be more routinely applied in clinical and public health laboratories.

Antimicrobial-resistant bacteria in international travelers

PURPOSE OF REVIEW

Antimicrobial resistance (AMR) in bacteria poses a major risk to global public health, with many factors contributing to the observed increase in AMR. International travel is one recognized contributor. The purpose of this review is to summarize current knowledge regarding the acquisition, carriage and spread of AMR bacteria by international travelers.

RECENT FINDINGS

Recent studies have highlighted that travel is an important risk factor for the acquisition of AMR bacteria, with approximately 30% of studied travelers returning with an acquired AMR bacterium. Epidemiological studies have shown there are three major risk factors for acquisition: travel destination, antimicrobial usage and travelers' diarrhea (TD). Analyses have begun to illustrate the AMR genes that are acquired and spread by travelers, risk factors for acquisition and carriage of AMR bacteria, and local transmission of imported AMR organisms.

SUMMARY

International travel is a contributor to the acquisition and dissemination of AMR organisms globally. Efforts to reduce the burden of AMR organisms should include a focus on international travelers. Routine genomic surveillance would further elucidate the role of international travel in the global spread of AMR bacteria.

Sources and transmission routes of campylobacteriosis: A combined analysis of genome and exposure data

OBJECTIVES

To determine the contributions of several animal and environmental sources of human campylobacteriosis and identify source-specific risk factors.

METHODS

1417 Campylobacter jejuni/coli isolates from the Netherlands in 2017-2019 were whole-genome sequenced, including isolates from human cases (n = 280), chickens/turkeys (n = 238), laying hens (n = 56), cattle (n = 158), veal calves (n = 49), sheep/goats (n = 111), pigs (n = 110), dogs/cats (n = 100), wild birds (n = 62), and surface water (n = 253). Questionnaire-based exposure data was collected. Source attribution was performed using core-genome multilocus sequence typing. Risk factors were determined on the attribution estimates.

RESULTS

Cases were mostly attributed to chickens/turkeys (48.2%), dogs/cats (18.0%), cattle (12.1%), and surface water (8.5%). Of the associations identified, never consuming chicken, as well as frequent chicken consumption, and rarely washing hands after touching raw meat, were risk factors for chicken/turkey-attributable infections. Consuming unpasteurized milk or barbecued beef increased the risk for cattle-attributable infections. Risk factors for infections attributable to environmental sources were open water swimming, contact with dog faeces, and consuming non-chicken/turkey avian meat like game birds.

CONCLUSIONS

Poultry and cattle are the main livestock sources of campylobacteriosis, while pets and surface water are important non-livestock sources. Foodborne transmission is only partially consistent with the attributions, as frequency and alternative pathways of exposure are significant.

Molecular typing tools for identifying and characterizing lactic acid bacteria: a review

Identification and classification of beneficial microbes is of the highest significance in food science and related industries. Conventional phenotypic approaches pose many challenges, and they may misidentify a target, limiting their use. Genotyping tools show comparatively better prospects, and they are widely used for distinguishing microorganisms. The techniques already employed in genotyping of lactic acid bacteria (LAB) are slightly different from one another, and each tool has its own advantages and disadvantages. This review paper compiles the comprehensive details of several fingerprinting tools that have been used for identifying and characterizing LAB at the species, sub-species, and strain levels. Notably, most of these approaches are based on restriction digestion, amplification using polymerase chain reaction, and sequencing. Nowadays, DNA sequencing technologies have made considerable progress in terms of cost, throughput, and methodology. A research journey to develop improved versions of generally applicable and economically viable tools for fingerprinting analysis is ongoing globally.

AB_SA: Accessory genes-Based Source Attribution - tracing the source of Salmonella enterica Typhimurium environmental strains

The partitioning of pathogenic strains isolated in environmental or human cases to their sources is challenging. The pathogens usually colonize multiple animal hosts, including livestock, which contaminate the food-production chain and the environment (e.g. soil and water), posing an additional public-health burden and major challenges in the identification of the source. Genomic data opens up new opportunities for the development of statistical models aiming to indicate the likely source of pathogen contamination. Here, we propose a computationally fast and efficient multinomial logistic regression source-attribution classifier to predict the animal source of bacterial isolates based on 'source-enriched' loci extracted from the accessory-genome profiles of a pangenomic dataset. Depending on the accuracy of the model's self-attribution step, the modeller selects the number of candidate accessory genes that best fit the model for calculating the likelihood of (source) category membership. The Accessory genes-Based Source Attribution (AB_SA) method was applied to a dataset of strains of Salmonella enterica Typhimurium and its monophasic variant (S. enterica 1,4,[5],12:i:-). The model was trained on 69 strains with known animal-source categories (i.e. poultry, ruminant and pig). The AB_SA method helped to identify 8 genes as predictors among the 2802 accessory genes. The self-attribution accuracy was 80 %. The AB_SA model was then able to classify 25 of the 29 S. enterica Typhimurium and S. enterica 1,4,[5],12:i:- isolates collected from the environment (considered to be of unknown source) into a specific category (i.e. animal source), with more than 85 % of probability. The AB_SA method herein described provides a user-friendly and valuable tool for performing source-attribution studies in only a few steps. AB_SA is written in R and freely available at https://github.com/lguillier/AB_SA.

Source Attribution Study of Sporadic Salmonella Derby Cases in France

Salmonella enterica subsp. enterica serovar Derby is one of the most frequent causes of gastroenteritis in humans. In Europe, this pathogen is one of the top five most commonly reported serovars in human cases. In France, S. Derby has been among the ten most frequently isolated serovars in humans since the year 2000. The main animal hosts of this serovar are pigs and poultry, and white meat is the main source of human contamination. We have previously shown that this serovar is polyphyletic and that three distinct genetic lineages of S. Derby cohabit in France. Two of them are associated with pork and one with poultry. In this study, we conducted a source attribution study based on single nucleotide polymorphism analysis of a large collection of 440 S. Derby human and non-human isolates collected in 2014-2015, to determine the contribution of each lineage to human contamination. In France, the two lineages associated with pork strains, and corresponding to the multilocus sequence typing (MLST) profiles ST39-ST40 and ST682 were responsible for 94% of human contaminations. Interestingly, the ST40 profile is responsible for the majority of human cases (71%). An analysis of epidemiologic data and the structure of the pork sector in France allowed us to explain the spread and the sporadic pattern of human cases that occurred in the studied period.