Final Report Summary - PROMISE (Protection of consumers by microbial risk mitigation through combating segregation of expertise)
Executive Summary:
PROMISE has been an EU FP7 funded project running from 1st of January 2012 to 31st of December 2014. The project was coordinated by the University of Veterinary Sciences Vienna, Austria, namely by Prof. Martin Wagner. The consortium comprises of 20 project partners coming from 14 different countries. The primary strategic objective of PROMISE was to improve and increase the integration, collaboration and knowledge transfer between the new member states, old member states (EU15) and candidate countries through a collaborative workplan accompanied by regional training and dissemination actions. The objective was to tackle common food safety threats and hence to protect the European consumers. PROMISE integrates stakeholders like public health authorities and national food safety authorities from the old and new member countries in order to ensure the exploitation of research results into standardization and harmonization efforts and hence to contribute to sustainability of project outcomes. In Europe concerns about food are moving away from issues of ensuring an adequate supply and choice of products towards issues of food safety, animal and plant welfare, labeling and traceability. These changes in emphasis have been shaped by scientific and technological advances within the food chain on the one hand, and health or environmental concerns on the other hand. PROMISE strived for sustainability through involvement of risk communicators, enhanced the knowledge on pathogen transmission and increased risk management capabilities. A main focus of PROMISE was to analyse, assess and interpret the introduction of new variants of pathogens through yet neglected or underresearched routes. PROMISE has met this objective by collecting food samples from air travelers and isolating the most important zoonotic agents. PROMISE studied microtransmission of pathogens on food processing level. We used Listeria monocytogenes (L. monocytogenes) as a model organism and developed a harmonised sampling protocol. PROMISE showed the extent of FPE contamination in 12 European food processing operations (FPO) and developed guidelines how to combat contamination by improved hygiene. PROMISE consortium was working on developing models for better understanding discrepancies in outbreak prevalence data but also in modelling pathogen prevalence data derived from investigations performed on FPO. By intensive discussion, partners were made familiar with culture-dependent (isolation of bugs) and culture independent methods (virus detection, typing, sequencing, phylogenic analysis). Dissemination activity relied both to internal dissemination (within the consortium) and dissemination to stakeholders. Internally, PROMISE was very active in Early Stage Researcher (ESR) exchange and exchange of Senior Researchers (SR). PROMISE bridged the discrepancies in technical abilities by a Technical Training and Workshop Tour (TTWT). All contents developed was made accessible to the consortium through the PROMISE academy that ensured the exploitation and dissemination of research results beyond the funding period. By the end of the project (2014), PROMISE has already published 25 papers mostly in high profile journals and another 16 publications are either in review or in preparation.
Project Context and Objectives:
PROMISE strived for multidimensional networking thus fostering integration
The primary strategic objective of PROMISE was to improve and increase the integration, collaboration and knowledge transfer between the new member states, old member states (EU15) and candidate countries through a collaborative workplan accompanied by regional training and dissemination actions. To reach this goal we have develop a research agenda that was of interest for both the OMS and NMS (see section 1.4 “How did PROMISE meet the specific objectives”?). Furthermore we built up an educational system based on (i) hand-on trainings, (ii) mobility programs and (iii) organisation of fora (annual meetings, regional workshops, stakeholder conferences) enabling to present scientific outcomes to targeted communities.
PROMISE strived for sustainability through involvement of risk communicators
Seven national food safety authorities (FSA) from the old and new member countries were engaged through a special forum: the Sustainability and Policy Board (SPB). The FSAs reflected the data produced, contributed to the elaboration of deliverables, gave key note talks at a variety of meetings and managed to organize 21 regional workshops in total.
PROMISE enhanced the knowledge on pathogen transmission
As described in the specific objectives in more detail, a main focus of PROMISE was to analyse, assess and interpret the introduction of new variants of pathogens through yet neglected or underresearched routes. In WP 1, PROMISE has met this objective by collecting food samples from air travelers and isolating the most important zoonotic agents. Since air traveler arrivals e.g. at airports are a multifold in comparison to the resident population (14 million arrivals to airport Vienna versus 1.8 million residents) there could be a chance to introduce new strain variants from all over the world. The consortium stored these isolates to a database and characterised them by molecular methods. Inspired by the work, the consortium overfulfilled the contract by including agents initially not having been included in the panel of parameters (such as foodborne viruses).
In WP2, PROMISE studied microtransmission of pathogens on food processing level. We used Listeria monocytogenes (L. monocytogenes) as a model organism and developed a harmonised sampling protocol. PROMISE showed the extent of FPE contamination in 12 European food processing operations (FPO) and developed guidelines how to combat contamination by improved hygiene.
PROMISE increases risk management capabilities
Prevalence and typing data are crucial to assess the impact of transmission routes. Modelling is an approach essential to draw conclusions from raw data. In WP3, the PROMISE consortium was working on developing models for better understanding discrepancies in outbreak prevalence data but also in modelling pathogen prevalence data derived from investigations performed on FPO. Since data need to be shown to an audience, PROMISE laid a cornerstone for concepts that might help to better visualize prevalence data.
PROMISE enhanced the harmonization of sampling and control procedures
In terms of all efforts, PROMISE was seeking to use harmonized sampling and detection protocols. By intensive discussion, partners were made familiar with culture-dependent (isolation of bugs) and culture independent methods (virus detection, typing, sequencing, phylogenic analysis). Wherever ISO or CEN methods were existing, PROMISE used those methods. Sampling approaches were built on guidelines or other scientific advice. This approach facilitated the transfer of concepts and methods between countries and some of the participating EU member states conducted systematic sampling of FPEs for L. monocytogenes the first time.
PROMISE contributes to risk mitigation through tailor-made dissemination strategies
Dissemination activity relied both to internal dissemination (within the consortium) and dissemination to stakeholders. Internally, PROMISE was very active in Early Stage Researcher (ESR) exchange and exchange of Senior Researchers (SR). Whereas the ESR exchange fostered the integration of young phDs into the consortium and facilitated the exchange of expertise, the SR exchange helped in managing the project, writing up papers, deliverables and guidelines. Furthermore PROMISE bridged the discrepancies in technical abilities by a Technical Training and Workshop Tour (TTWT). All contents developed was made accessible to the consortium through the PROMISE academy that ensured the exploitation and dissemination of research results beyond the funding period.
Externally, dissemination of results to stakeholders was fueled by an enormous publication activity already within the timespan of the project. By the end of the project (2014), PROMISE has already published 25 papers mostly in high profile journals and another 16 publications are either in review or in preparation. Most important to us was the information of the local stakeholders in the EU member states. To reach this ambitious goal, PROMISE set up a new TRAIN the TRAINER/TRAINEE concept endorsed by a massive involvement of “endusers” such food enterprises and public health authorities. PROMISE organized a JOINT CONFERENCE with a relevant COST funded expert network termed BacFoodNet inspired by the consideration that BacFoodNet people are working on issues related to PROMISE research. This conference was an outstanding success through the participation of EFSA and WHO experts, and broadcasted through national radio and TV stations. The food business operations (FBO) and the trans-european science community were informed at stakeholder meetings. To foster communication of results at national level, an impressive number of regional workshops reached more than 1100 experts in all member states participating to PROMISE.
Project Results:
(
i) Identification of yet unknown risks for transmission of pathogens through exogenous routes of transmission
In WP 1, the consortium tested food originating from more than 50 countries worldwide that has been carried along by travellers in their luggage and collected more than 500 isolates that are of interest for further characterization by molecular research. The involved partners collected 2641 animal food samples confiscated at airports (Frankfurt, Vienna, Bilbao), ports, black and food markets selling predominantly homegrown food. The pathogens investigated were Campylobacter spp., Listeria monocytogenes, Salmonella spp. and potentially hazardous E. coli. Some partners were looking for staphylococci and multi-drug resistant E. coli. The amount of food carried with global travellers was surprising (2.7 – 4.2 kg; 2,4% of all passengers). Generally the prevalence of Salmonella and verotoxin-producing E. coli (VTEC) was in a range that could have been expected from data revealed from analysis of domestic food samples. Regarding L. monocytogenes and in comparison to the recently published EU baseline study, the prevalence was quite high in some food items (eg in fish samples collected from a Romanian grey market). Campylobacter was found very rarely what is explainable by the fact that >80% of food items were highly processed and of reduced free water activity. Multi Locus Sequence Typing revealed yet unknown Salmonella and VTEC variants. Analysis of virulence determinants in Salmonella, VTEC and L. monocytogenes revealed new virulotypes in Salmonella, whereas all but one VTEC isolate were found to be non-EHEC strains. In L. monocytogenes, analysis resulted in a rather high abundance of ST121, St155, ST9 strains signified by a low virulence profile what was explainable by mutation in some virulence markers (hemolysin gene and internalin A). Abundance of antibiotic resistance markers was studied in E. coli and staphylococci. It was surprising that almost all VTEC stains had very few AB resistance genes whereas allmost all non-VTEC showed up to 12 AB resistance genes. Methicillin resistence in staphylococci was rare and associated with strains isolated from food arriving from Bolivia. Finally, one partner was looking for abundance of foodborne viruses and found a high percentage of Hepatitis E virus but no indication of food contamination with norovirus and Hepatitis A virus. Conclusively, PROMISE provided the first data on microbiological hazards arriving into EU-28 via travellers at airports and by cross-boarder traffic. The percentage of passengers carrying food in their luggage was considerable. Since most food was confiscated from passagers that are part of the national immigrant scene we concluded that the food is not only consumed in homes but might get vendored in ethnic cuisine. New genetic variants of zoonotic agents were described in Salmonella and VTEC, however no evidence was found that these isolates might be a matter of increased concern. Due to the limited amount of food usually transferred, we concluded that outbreaks encompassing multiple individuals are not expectable.
(ii) Identification of yet unknown risks for transmission of pathogens through indigenous routes of transmission
The macro-transmission aspect is completed by research on another neglected source of transmission of food-borne pathogens, namely in-house cross-contamination. For this research direction, L. monocytogenes was chosen as a model organism. From the very beginning, it was the ambition of PROMISE to collaborate with food processing operations (FBO) rather than to work lab-scale with in vitro models. Six PROMISE partners had to find FBO collaborators that were accepting that their food processing environment (FPE) is going to be sampled by a very thorough approach (with respect to number of samples taken). The basic idea was to demonstrate whether “presumed negative” FPEs stay negative (even when a lot of samples are taken by exerts) and to which extent the contamination pattern fluctuates in “positive” FPEs (through e.g. application of intensified cleaning and disinfection). A side effect was to train the partners that had never done FPE sampling for L. monocytogenes before. FBOs were also asked to appraise their contamination status through thorough check up of their self-control data. In four comprehensive sampling events, 12 FPE in six European countries were tested by using a harmonized sampling procedure. By the end of the reporting period the PROMISE consortium collected and investigated 2987 samples (1234 from presumed uncontaminated premises and 1754 from known contaminated premises). We compared the culture dependent species detection with data as achieved by culture independent species detection technology (real time PCR) and found PCR (if performed on secondary enrichments) equivalent to culture. These data recommend public domain real time PCR to be used as a rapid L. monocytogenes detection technology in FBOs.
In total 440 samples were positive for L. monocytogenes (54 from presumed uncontaminated premises and 386 from contaminated premises). All FBOs proved to be positive to some extent, either those that were known to be contaminated and those that denied contamination. Meat FPEs were higher contaminated than dairy FPEs. On strains isolated from three FBOs that proved to be highly contaminated (Ireland, Romania, Austria) we conducted intensified typing (to determine whether the isolated clones were persistent), (ii) performed growth experiments in foods under processing, (iii) studied biofilm formation and (iv) tested for resistance against sanitizers. We further matched the sites of positive findings in FPEs with spatial organisation and personnel movement patterns. This approach allowed us to describe three contamination scenarios: widely disseminated contamination, hotspot contamination, directed contamination (close to entrance). Since the contamination scenario “directed contamination” implied that hygiene measures were insufficient at people entering the FPE we performed a study describing the functionality of hygiene barriers. We found that most hygiene barriers lacked to decontaminate the working boots and in one case we could isolate bugs resistant to sanitizers (hypochlorite) from such a hygiene barrier (what would boost cross contamination dramatically). Sequencing was performed on the hypochlorite resistant isolates, however no marker genes explaining the phenotype was identified yet. Isolate characterization by molecular typing allowed the consortium to study the initial colonization pattern of a newly built plant in Spain and to elucidate the spreading pattern of L. monocytogenes during reconstruction of already existing plant in Austria. These data increased our understanding concerning risks for cross-contamination in cases where production routines are changed or interrupted. In another highly contaminated FBO, we compared the occurrence of L. monocytogenes in FPE and matched the data with occurrence of L. monocytogenes in the food under processing. By collaboration with WP3 the PROMISE consortium developed a model that enables to predict the impact of occurrence of L. monocytogenes in FPE on the safety of the food being processed. Such models are needed to manage decision-making with regard to “big data” that are going to be delivered during application of self-control programs.
Further molecular studies on contamination patterns in FPE revealed drains as a hotspot for cross contamination. Conclusively we studied the microbiome of L. monocytogenes contaminated drains and compared these data with drains that were L. monocytogenes negative. We found a rather weak match of the microbiomes residing in drain water and drain biofilms. We could describe non-listerial species that obviously co-occur with L. monocytogenes in drains. By sequencing persistent L. monocytogenes of ST121, we were lucky to identify a novel marker that could explain why this particular MLST type is so prevalent in FPEs. Taking all together, PROMISE was highly successful in shedding a light on the persistent lifestyle of L. monocytogenes in FPEs. We developed management strategies that might help to combat L. monocytogenes in FPEs more efficiently. We are convinced that monitoring of critical environmental points (as a completion of HACCPs concepts that focus on food under processing) could avoid cross-continuation into the food chain. A further leap forward would be to tackle the assumption underlying Listeria legislation that all L. monocytogenes isolates should be considered as equally virulent. As shown by us and other authors, highly prevalent L. monocytogenes ST121 and ST 155 are very low in virulence (due to a truncated internalin A), however, the legal consequence remains the same! Therefore a scientifically unnecessary damage happens to FBOs whenever the same legal consequences are applied to the presence ofhighly and low virulent strains. We propose that a ST type dependent risk assessment concept should be developed and further discussed. All these conclusions were checked and balanced in discussions with the Food Safety Authorities and a guideline was developed that should help FBOs to successfully combat L. monocytogenes contamination of FPEs.
(iii) Development of models to analyse the impact of food chain parameters on pathogen prevalence
The transmission studies were completed by a small modeling work package that researches novel ways of understanding the boundaries of prevalence data and helps to draw the right conclusions from transmission studies.
Details regarding the occurrence of pathogens in food are always difficult to ascertain and equally difficult to interpret. PROMISE addressed these issues in relation to food borne illness in the EU by considering three questions; what are the appropriate data sources to quantify food borne pathogens in the EU? How well does the data supply represent the real incidence of infectious disease in EU MSs? And can the reported incidence be used to infer particular patterns in the occurrence of disease?
In Europe the EFSA and the ECDC collect, collate and publish data about food borne illness and zoonoses within the community each year. National data is submitted annually by each EU MS and covers a variety of issues from reported outbreaks of food borne illness to the prevalence of zoonotic agents in flocks of laying hens. PROMISE reviewed the published data and built a small data base of supplementary data sources, contributed by project partners, but also confirmed that the EFSA annual summaries, and the national submissions which support them, represent the dominant, and harmonizing, EU food safety information supply.
The reporting process, for food borne illness cases or outbreaks, is highly variable and difficult to quantify. Following on from previous work performed by the MedVetNet group PROMISE examined factors that drive variability in reporting of pathogen specific food borne illness within distinct EU MSs. A mathematical model, supported by well-defined national data supplies, can be used to quantify the rate, per case, at which infectious intestinal disease enters national records and, inversely, can support the transformation of national reports to give a picture of actual incidence rates. Within the reporting process some key features can be identified, such as the way in which bloody diarrhoea is reported by cases and by practitioners, and these form the basis for guiding improved reporting and increased harmonization. This development provides a clear picture of information supplies that are necessary to establish increased confidence in reporting efficiency and, therefore, that are necessary for inference and interventions based on actual incidence.
The developing EU data supply has increasing continuity; currently extending over nine years and 27 MSs. This continuity increasingly prompts comparisons between the incidence rate values for a particular MS at different times (trends) and between values for different MSs at the same time (spatial variation) but the heterogeneity of the data collection makes clear expression of these patterns difficult. Notwithstanding with each annual release of data there is an increasing requirement, by many stakeholder groups, for improved interpretation and visualization of the patterns. In PROMISE a variety of time series models were compared, and tested on EFSA data, and this prompted a view that analysis based on discrete (annual) changes, that is more commonly associated with bio-surveillance, may have advantages in comparison with the traditional (regression and moving average) continuous interpretation. Geographical variation of rate information is currently encoded, by EFSA and ECDC, in coloured maps but a review of this process, in PROMISE, revealed a wide variety of modern visualization techniques such as the Spatio-Temporal epidemiology modelling, that harness GIS and multimedia methods, which could add value.
(iv) Mitigating the impact of zoonotic disease in New Member Countries, Candidate Countries and Old Member Countries by improving and strengthening integration in food safety research
Since integration, collaboration and knowledge transfer between the new and old member states of the European is a core objective of PROMISE, work packages four to six are designed as training (WP4), risk communication (WP5) and dissemination packages (WP6). A series of training sessions and mobility programs, both for senior and junior researchers, was organized in WP 4. Since achieving sensitive data, we have created a WP5 essentially working on best practice for risk communication with national food safety authorities. WP6 aimed at disseminating appropriate PROMISE results to regulatory bodies (results from WP1), stakeholders such as food business operators (results from WP2) and risk assessors (EFSA, results from WP3). To achieve this important milestone, a PROMISE academy was founded that makes PROMISE content available through e-learning.
(v) Dissemination of research results and scientific support to policy making
Potential Impact:
PROMISE, as a highly active network of 12 academic partners of multiple disciplines including 6 national food safety authorities has exploited all intellectual and investigative resources to fulfill all subjects laid down in the contract. We have performed the program without any need for deviation due to contingency. This is remarkable since most work was performed in a sensitive setting (work at border stations, work at food business operations) requiring legislative support (permissions) and a burden of pre-laboratory preparation.
PROMISE performed a hazard characterization on isolates shuttling around at either a global scale or at food premises.
(1) We have been the first that have intensively looked into zoonotic microorganisms associated with illegal food imports through travelers. The impact is fully given since (private or business) traveling is expected to launch to yet unseen numbers in the years coming (The German Ministry for Transportation reported an increase in air passengers from 167.000.000 in 2010 to 225.000.000 in 2025)http://web.ard.de/themenwoche_2011/?p=2905). PROMISE is one of the first projects that have systematically investigated the nature of zoonotic strains shuttled by passengers. Due to the inspiring collaboration PROMISE has overfulfilled the number of food items tested (2500 increased to 2621) and organized research on pathogens initially not part of the contract (viruses).
(2) We have provided data both on virulence profiles and on transmission of antibiotics resistance genes and phenotypes coming into EU-28 through this channel. Our data will stimulate discussions on whether more caution should be shared with neglected or under-researched gates of transmission, either through increased research endeavours, preventive action (increasing control effort) or intensified traveler information. Some of our data were used by EFSA for their recently published risk assessment on Ebola virus transmission through food items.
The micro-transmission study aimed at shedding a light on the importance of Article 5 of the current EU wide legislation (Dir 2073/2005) that stipulates environmental sampling as a means to combat L. monocytogenes transmission. Food business operators (FBOs) manufacturing ready-to-eat foods are obliged to test processing areas and equipment for the absence/presence of L. monocytogenes as part of their sampling scheme. Supporting guidelines for sampling of food processing areas and equipment for the detection of L. monocytogenes have been already published however a systemic approach or vigilance concept is currently not existing.
(3) PROMISE has shown the dimension of the problem and it must be concluded that almost every animal food production is endangered by transmission of L. monoyctogenes from the processing environment to the FPE. A concept looking to eradicate the pathogen from FPEs is unrealistic (promoted as the “Seek and destroy” concept). An essential impact of PROMISE is that some countries started to look into this issue for the first time. It is all about knowing (problem evidence) and to increase the willingness to react (response). If food business operators learn to control L. monocytogenes in their processing environment then the risk for cross contamination into food under processing will be reduced and the margin of safety will be elevated. Most problems occur in FBO through “silenced” contamination scenarios (contamination scenarios that are not getting to the public). FBO have to recall food lots, suffer from increased costs for investigations. Developing procedures to control the pathogen in the environment might help to reduce such costs.
(4) We have exchanged protocols, and shared the conceptual ideas. This is best documented by the long list of publications that were set up by researchers from OMS an NMS.
(5) Beyond scientific excellence a further impact lies in the conceptual work towards such a vigilance concept thus increasing the at-line response capabilities of European food producers. The concept has been published in a book edited by core members of the PROMSIE consortium (Kieran Jordan; Ed Fox; Martin Wagner).
The modelling WP will try to show what data sources (outbreak, prevalence), as reported through PROMISE or at (super)national level mean in terms of interpretability and reliability. It will provide a framework for improved appreciation of fluctuations in outbreak and case numbers annually reported by EFSA or ECDC
(6) The impact lies in a better comparability of prevalence data. Prevalence estimates are subject to the phenomenon called “magic of numbers” and very often (mis)used for political statements (“member state X must be good in public health cause it reports half the prevalence of disease in comparison to member state Y”). PROMISE has tried to shed a light behind the curtain and tried to develop a model allow to quantify factors that contribute to over- or under reporting.
The PROMISE research efforts were embedded into activities allowing us to level out discrepancies among participating partners both with respect to research capabilities, surveillance strategies and risk communication procedures (WP 4 and WP6).
(7) PROMISE has increased the access to state-of-the-art analysis through exchange of personal, knowledge and technical skills. Experts organized Technical Training and Workshops for phD students and junior PostDocs basically educating the methods that were successfully used for WP1 – WP3. The mobility program supported 14 highly ambitious young students to learn about the methods in a hand-on style. Some collaboration has led to a long lasting collaboration and future career tracks (eg Luminita Cirolacu, Romania, now working in Belgium; Jasna Kovac SLO now going to work in the USA)
(8) Subconsortia were successful in novel grant applications (e.g. Turkey/Spain/Austria in the Leonardo program Safe Meat).
The exchange with Food Safety Authorities was a core activity in PROMISE (WP5). The collaboration proved to be based on confidence, and the researchers learnt a lot regarding risk management and communication.
(9) The impact is given in the fact that all scientific information was put into the appropriate channels for further reception and dissemination. It is the legislative bodies that have to decide on action, countermeasures and intensified communication (with stakeholder bodies; food processing operations). The food safety authorities were part in the project but also partners for dissemination. More than 20 workshops, many of them done in NMS in the mother tongue of the organizers, had a huge impact on the spread of knowledge and critical discussion of results. We are very proud that these approaches have stimulated discussions on risk management tremendously.
Taking all together the PROMISE impacts are multiple and highly satisfying the expectations of the partners. Almost all PhDs have finished their programs and are currently doing their final exams. Attracting young individuals for food science is one of the most important goals of PROMISE and all (wishes for) personal careers have been documented on the PROMISE Academy by video.
List of Websites:
http://www.promise-net.eu
http://www.promise-academy.eu
Prof. Martin Wagner
martin.wagner@vet
PROMISE has been an EU FP7 funded project running from 1st of January 2012 to 31st of December 2014. The project was coordinated by the University of Veterinary Sciences Vienna, Austria, namely by Prof. Martin Wagner. The consortium comprises of 20 project partners coming from 14 different countries. The primary strategic objective of PROMISE was to improve and increase the integration, collaboration and knowledge transfer between the new member states, old member states (EU15) and candidate countries through a collaborative workplan accompanied by regional training and dissemination actions. The objective was to tackle common food safety threats and hence to protect the European consumers. PROMISE integrates stakeholders like public health authorities and national food safety authorities from the old and new member countries in order to ensure the exploitation of research results into standardization and harmonization efforts and hence to contribute to sustainability of project outcomes. In Europe concerns about food are moving away from issues of ensuring an adequate supply and choice of products towards issues of food safety, animal and plant welfare, labeling and traceability. These changes in emphasis have been shaped by scientific and technological advances within the food chain on the one hand, and health or environmental concerns on the other hand. PROMISE strived for sustainability through involvement of risk communicators, enhanced the knowledge on pathogen transmission and increased risk management capabilities. A main focus of PROMISE was to analyse, assess and interpret the introduction of new variants of pathogens through yet neglected or underresearched routes. PROMISE has met this objective by collecting food samples from air travelers and isolating the most important zoonotic agents. PROMISE studied microtransmission of pathogens on food processing level. We used Listeria monocytogenes (L. monocytogenes) as a model organism and developed a harmonised sampling protocol. PROMISE showed the extent of FPE contamination in 12 European food processing operations (FPO) and developed guidelines how to combat contamination by improved hygiene. PROMISE consortium was working on developing models for better understanding discrepancies in outbreak prevalence data but also in modelling pathogen prevalence data derived from investigations performed on FPO. By intensive discussion, partners were made familiar with culture-dependent (isolation of bugs) and culture independent methods (virus detection, typing, sequencing, phylogenic analysis). Dissemination activity relied both to internal dissemination (within the consortium) and dissemination to stakeholders. Internally, PROMISE was very active in Early Stage Researcher (ESR) exchange and exchange of Senior Researchers (SR). PROMISE bridged the discrepancies in technical abilities by a Technical Training and Workshop Tour (TTWT). All contents developed was made accessible to the consortium through the PROMISE academy that ensured the exploitation and dissemination of research results beyond the funding period. By the end of the project (2014), PROMISE has already published 25 papers mostly in high profile journals and another 16 publications are either in review or in preparation.
Project Context and Objectives:
PROMISE strived for multidimensional networking thus fostering integration
The primary strategic objective of PROMISE was to improve and increase the integration, collaboration and knowledge transfer between the new member states, old member states (EU15) and candidate countries through a collaborative workplan accompanied by regional training and dissemination actions. To reach this goal we have develop a research agenda that was of interest for both the OMS and NMS (see section 1.4 “How did PROMISE meet the specific objectives”?). Furthermore we built up an educational system based on (i) hand-on trainings, (ii) mobility programs and (iii) organisation of fora (annual meetings, regional workshops, stakeholder conferences) enabling to present scientific outcomes to targeted communities.
PROMISE strived for sustainability through involvement of risk communicators
Seven national food safety authorities (FSA) from the old and new member countries were engaged through a special forum: the Sustainability and Policy Board (SPB). The FSAs reflected the data produced, contributed to the elaboration of deliverables, gave key note talks at a variety of meetings and managed to organize 21 regional workshops in total.
PROMISE enhanced the knowledge on pathogen transmission
As described in the specific objectives in more detail, a main focus of PROMISE was to analyse, assess and interpret the introduction of new variants of pathogens through yet neglected or underresearched routes. In WP 1, PROMISE has met this objective by collecting food samples from air travelers and isolating the most important zoonotic agents. Since air traveler arrivals e.g. at airports are a multifold in comparison to the resident population (14 million arrivals to airport Vienna versus 1.8 million residents) there could be a chance to introduce new strain variants from all over the world. The consortium stored these isolates to a database and characterised them by molecular methods. Inspired by the work, the consortium overfulfilled the contract by including agents initially not having been included in the panel of parameters (such as foodborne viruses).
In WP2, PROMISE studied microtransmission of pathogens on food processing level. We used Listeria monocytogenes (L. monocytogenes) as a model organism and developed a harmonised sampling protocol. PROMISE showed the extent of FPE contamination in 12 European food processing operations (FPO) and developed guidelines how to combat contamination by improved hygiene.
PROMISE increases risk management capabilities
Prevalence and typing data are crucial to assess the impact of transmission routes. Modelling is an approach essential to draw conclusions from raw data. In WP3, the PROMISE consortium was working on developing models for better understanding discrepancies in outbreak prevalence data but also in modelling pathogen prevalence data derived from investigations performed on FPO. Since data need to be shown to an audience, PROMISE laid a cornerstone for concepts that might help to better visualize prevalence data.
PROMISE enhanced the harmonization of sampling and control procedures
In terms of all efforts, PROMISE was seeking to use harmonized sampling and detection protocols. By intensive discussion, partners were made familiar with culture-dependent (isolation of bugs) and culture independent methods (virus detection, typing, sequencing, phylogenic analysis). Wherever ISO or CEN methods were existing, PROMISE used those methods. Sampling approaches were built on guidelines or other scientific advice. This approach facilitated the transfer of concepts and methods between countries and some of the participating EU member states conducted systematic sampling of FPEs for L. monocytogenes the first time.
PROMISE contributes to risk mitigation through tailor-made dissemination strategies
Dissemination activity relied both to internal dissemination (within the consortium) and dissemination to stakeholders. Internally, PROMISE was very active in Early Stage Researcher (ESR) exchange and exchange of Senior Researchers (SR). Whereas the ESR exchange fostered the integration of young phDs into the consortium and facilitated the exchange of expertise, the SR exchange helped in managing the project, writing up papers, deliverables and guidelines. Furthermore PROMISE bridged the discrepancies in technical abilities by a Technical Training and Workshop Tour (TTWT). All contents developed was made accessible to the consortium through the PROMISE academy that ensured the exploitation and dissemination of research results beyond the funding period.
Externally, dissemination of results to stakeholders was fueled by an enormous publication activity already within the timespan of the project. By the end of the project (2014), PROMISE has already published 25 papers mostly in high profile journals and another 16 publications are either in review or in preparation. Most important to us was the information of the local stakeholders in the EU member states. To reach this ambitious goal, PROMISE set up a new TRAIN the TRAINER/TRAINEE concept endorsed by a massive involvement of “endusers” such food enterprises and public health authorities. PROMISE organized a JOINT CONFERENCE with a relevant COST funded expert network termed BacFoodNet inspired by the consideration that BacFoodNet people are working on issues related to PROMISE research. This conference was an outstanding success through the participation of EFSA and WHO experts, and broadcasted through national radio and TV stations. The food business operations (FBO) and the trans-european science community were informed at stakeholder meetings. To foster communication of results at national level, an impressive number of regional workshops reached more than 1100 experts in all member states participating to PROMISE.
Project Results:
(
i) Identification of yet unknown risks for transmission of pathogens through exogenous routes of transmission
In WP 1, the consortium tested food originating from more than 50 countries worldwide that has been carried along by travellers in their luggage and collected more than 500 isolates that are of interest for further characterization by molecular research. The involved partners collected 2641 animal food samples confiscated at airports (Frankfurt, Vienna, Bilbao), ports, black and food markets selling predominantly homegrown food. The pathogens investigated were Campylobacter spp., Listeria monocytogenes, Salmonella spp. and potentially hazardous E. coli. Some partners were looking for staphylococci and multi-drug resistant E. coli. The amount of food carried with global travellers was surprising (2.7 – 4.2 kg; 2,4% of all passengers). Generally the prevalence of Salmonella and verotoxin-producing E. coli (VTEC) was in a range that could have been expected from data revealed from analysis of domestic food samples. Regarding L. monocytogenes and in comparison to the recently published EU baseline study, the prevalence was quite high in some food items (eg in fish samples collected from a Romanian grey market). Campylobacter was found very rarely what is explainable by the fact that >80% of food items were highly processed and of reduced free water activity. Multi Locus Sequence Typing revealed yet unknown Salmonella and VTEC variants. Analysis of virulence determinants in Salmonella, VTEC and L. monocytogenes revealed new virulotypes in Salmonella, whereas all but one VTEC isolate were found to be non-EHEC strains. In L. monocytogenes, analysis resulted in a rather high abundance of ST121, St155, ST9 strains signified by a low virulence profile what was explainable by mutation in some virulence markers (hemolysin gene and internalin A). Abundance of antibiotic resistance markers was studied in E. coli and staphylococci. It was surprising that almost all VTEC stains had very few AB resistance genes whereas allmost all non-VTEC showed up to 12 AB resistance genes. Methicillin resistence in staphylococci was rare and associated with strains isolated from food arriving from Bolivia. Finally, one partner was looking for abundance of foodborne viruses and found a high percentage of Hepatitis E virus but no indication of food contamination with norovirus and Hepatitis A virus. Conclusively, PROMISE provided the first data on microbiological hazards arriving into EU-28 via travellers at airports and by cross-boarder traffic. The percentage of passengers carrying food in their luggage was considerable. Since most food was confiscated from passagers that are part of the national immigrant scene we concluded that the food is not only consumed in homes but might get vendored in ethnic cuisine. New genetic variants of zoonotic agents were described in Salmonella and VTEC, however no evidence was found that these isolates might be a matter of increased concern. Due to the limited amount of food usually transferred, we concluded that outbreaks encompassing multiple individuals are not expectable.
(ii) Identification of yet unknown risks for transmission of pathogens through indigenous routes of transmission
The macro-transmission aspect is completed by research on another neglected source of transmission of food-borne pathogens, namely in-house cross-contamination. For this research direction, L. monocytogenes was chosen as a model organism. From the very beginning, it was the ambition of PROMISE to collaborate with food processing operations (FBO) rather than to work lab-scale with in vitro models. Six PROMISE partners had to find FBO collaborators that were accepting that their food processing environment (FPE) is going to be sampled by a very thorough approach (with respect to number of samples taken). The basic idea was to demonstrate whether “presumed negative” FPEs stay negative (even when a lot of samples are taken by exerts) and to which extent the contamination pattern fluctuates in “positive” FPEs (through e.g. application of intensified cleaning and disinfection). A side effect was to train the partners that had never done FPE sampling for L. monocytogenes before. FBOs were also asked to appraise their contamination status through thorough check up of their self-control data. In four comprehensive sampling events, 12 FPE in six European countries were tested by using a harmonized sampling procedure. By the end of the reporting period the PROMISE consortium collected and investigated 2987 samples (1234 from presumed uncontaminated premises and 1754 from known contaminated premises). We compared the culture dependent species detection with data as achieved by culture independent species detection technology (real time PCR) and found PCR (if performed on secondary enrichments) equivalent to culture. These data recommend public domain real time PCR to be used as a rapid L. monocytogenes detection technology in FBOs.
In total 440 samples were positive for L. monocytogenes (54 from presumed uncontaminated premises and 386 from contaminated premises). All FBOs proved to be positive to some extent, either those that were known to be contaminated and those that denied contamination. Meat FPEs were higher contaminated than dairy FPEs. On strains isolated from three FBOs that proved to be highly contaminated (Ireland, Romania, Austria) we conducted intensified typing (to determine whether the isolated clones were persistent), (ii) performed growth experiments in foods under processing, (iii) studied biofilm formation and (iv) tested for resistance against sanitizers. We further matched the sites of positive findings in FPEs with spatial organisation and personnel movement patterns. This approach allowed us to describe three contamination scenarios: widely disseminated contamination, hotspot contamination, directed contamination (close to entrance). Since the contamination scenario “directed contamination” implied that hygiene measures were insufficient at people entering the FPE we performed a study describing the functionality of hygiene barriers. We found that most hygiene barriers lacked to decontaminate the working boots and in one case we could isolate bugs resistant to sanitizers (hypochlorite) from such a hygiene barrier (what would boost cross contamination dramatically). Sequencing was performed on the hypochlorite resistant isolates, however no marker genes explaining the phenotype was identified yet. Isolate characterization by molecular typing allowed the consortium to study the initial colonization pattern of a newly built plant in Spain and to elucidate the spreading pattern of L. monocytogenes during reconstruction of already existing plant in Austria. These data increased our understanding concerning risks for cross-contamination in cases where production routines are changed or interrupted. In another highly contaminated FBO, we compared the occurrence of L. monocytogenes in FPE and matched the data with occurrence of L. monocytogenes in the food under processing. By collaboration with WP3 the PROMISE consortium developed a model that enables to predict the impact of occurrence of L. monocytogenes in FPE on the safety of the food being processed. Such models are needed to manage decision-making with regard to “big data” that are going to be delivered during application of self-control programs.
Further molecular studies on contamination patterns in FPE revealed drains as a hotspot for cross contamination. Conclusively we studied the microbiome of L. monocytogenes contaminated drains and compared these data with drains that were L. monocytogenes negative. We found a rather weak match of the microbiomes residing in drain water and drain biofilms. We could describe non-listerial species that obviously co-occur with L. monocytogenes in drains. By sequencing persistent L. monocytogenes of ST121, we were lucky to identify a novel marker that could explain why this particular MLST type is so prevalent in FPEs. Taking all together, PROMISE was highly successful in shedding a light on the persistent lifestyle of L. monocytogenes in FPEs. We developed management strategies that might help to combat L. monocytogenes in FPEs more efficiently. We are convinced that monitoring of critical environmental points (as a completion of HACCPs concepts that focus on food under processing) could avoid cross-continuation into the food chain. A further leap forward would be to tackle the assumption underlying Listeria legislation that all L. monocytogenes isolates should be considered as equally virulent. As shown by us and other authors, highly prevalent L. monocytogenes ST121 and ST 155 are very low in virulence (due to a truncated internalin A), however, the legal consequence remains the same! Therefore a scientifically unnecessary damage happens to FBOs whenever the same legal consequences are applied to the presence ofhighly and low virulent strains. We propose that a ST type dependent risk assessment concept should be developed and further discussed. All these conclusions were checked and balanced in discussions with the Food Safety Authorities and a guideline was developed that should help FBOs to successfully combat L. monocytogenes contamination of FPEs.
(iii) Development of models to analyse the impact of food chain parameters on pathogen prevalence
The transmission studies were completed by a small modeling work package that researches novel ways of understanding the boundaries of prevalence data and helps to draw the right conclusions from transmission studies.
Details regarding the occurrence of pathogens in food are always difficult to ascertain and equally difficult to interpret. PROMISE addressed these issues in relation to food borne illness in the EU by considering three questions; what are the appropriate data sources to quantify food borne pathogens in the EU? How well does the data supply represent the real incidence of infectious disease in EU MSs? And can the reported incidence be used to infer particular patterns in the occurrence of disease?
In Europe the EFSA and the ECDC collect, collate and publish data about food borne illness and zoonoses within the community each year. National data is submitted annually by each EU MS and covers a variety of issues from reported outbreaks of food borne illness to the prevalence of zoonotic agents in flocks of laying hens. PROMISE reviewed the published data and built a small data base of supplementary data sources, contributed by project partners, but also confirmed that the EFSA annual summaries, and the national submissions which support them, represent the dominant, and harmonizing, EU food safety information supply.
The reporting process, for food borne illness cases or outbreaks, is highly variable and difficult to quantify. Following on from previous work performed by the MedVetNet group PROMISE examined factors that drive variability in reporting of pathogen specific food borne illness within distinct EU MSs. A mathematical model, supported by well-defined national data supplies, can be used to quantify the rate, per case, at which infectious intestinal disease enters national records and, inversely, can support the transformation of national reports to give a picture of actual incidence rates. Within the reporting process some key features can be identified, such as the way in which bloody diarrhoea is reported by cases and by practitioners, and these form the basis for guiding improved reporting and increased harmonization. This development provides a clear picture of information supplies that are necessary to establish increased confidence in reporting efficiency and, therefore, that are necessary for inference and interventions based on actual incidence.
The developing EU data supply has increasing continuity; currently extending over nine years and 27 MSs. This continuity increasingly prompts comparisons between the incidence rate values for a particular MS at different times (trends) and between values for different MSs at the same time (spatial variation) but the heterogeneity of the data collection makes clear expression of these patterns difficult. Notwithstanding with each annual release of data there is an increasing requirement, by many stakeholder groups, for improved interpretation and visualization of the patterns. In PROMISE a variety of time series models were compared, and tested on EFSA data, and this prompted a view that analysis based on discrete (annual) changes, that is more commonly associated with bio-surveillance, may have advantages in comparison with the traditional (regression and moving average) continuous interpretation. Geographical variation of rate information is currently encoded, by EFSA and ECDC, in coloured maps but a review of this process, in PROMISE, revealed a wide variety of modern visualization techniques such as the Spatio-Temporal epidemiology modelling, that harness GIS and multimedia methods, which could add value.
(iv) Mitigating the impact of zoonotic disease in New Member Countries, Candidate Countries and Old Member Countries by improving and strengthening integration in food safety research
Since integration, collaboration and knowledge transfer between the new and old member states of the European is a core objective of PROMISE, work packages four to six are designed as training (WP4), risk communication (WP5) and dissemination packages (WP6). A series of training sessions and mobility programs, both for senior and junior researchers, was organized in WP 4. Since achieving sensitive data, we have created a WP5 essentially working on best practice for risk communication with national food safety authorities. WP6 aimed at disseminating appropriate PROMISE results to regulatory bodies (results from WP1), stakeholders such as food business operators (results from WP2) and risk assessors (EFSA, results from WP3). To achieve this important milestone, a PROMISE academy was founded that makes PROMISE content available through e-learning.
(v) Dissemination of research results and scientific support to policy making
Potential Impact:
PROMISE, as a highly active network of 12 academic partners of multiple disciplines including 6 national food safety authorities has exploited all intellectual and investigative resources to fulfill all subjects laid down in the contract. We have performed the program without any need for deviation due to contingency. This is remarkable since most work was performed in a sensitive setting (work at border stations, work at food business operations) requiring legislative support (permissions) and a burden of pre-laboratory preparation.
PROMISE performed a hazard characterization on isolates shuttling around at either a global scale or at food premises.
(1) We have been the first that have intensively looked into zoonotic microorganisms associated with illegal food imports through travelers. The impact is fully given since (private or business) traveling is expected to launch to yet unseen numbers in the years coming (The German Ministry for Transportation reported an increase in air passengers from 167.000.000 in 2010 to 225.000.000 in 2025)http://web.ard.de/themenwoche_2011/?p=2905). PROMISE is one of the first projects that have systematically investigated the nature of zoonotic strains shuttled by passengers. Due to the inspiring collaboration PROMISE has overfulfilled the number of food items tested (2500 increased to 2621) and organized research on pathogens initially not part of the contract (viruses).
(2) We have provided data both on virulence profiles and on transmission of antibiotics resistance genes and phenotypes coming into EU-28 through this channel. Our data will stimulate discussions on whether more caution should be shared with neglected or under-researched gates of transmission, either through increased research endeavours, preventive action (increasing control effort) or intensified traveler information. Some of our data were used by EFSA for their recently published risk assessment on Ebola virus transmission through food items.
The micro-transmission study aimed at shedding a light on the importance of Article 5 of the current EU wide legislation (Dir 2073/2005) that stipulates environmental sampling as a means to combat L. monocytogenes transmission. Food business operators (FBOs) manufacturing ready-to-eat foods are obliged to test processing areas and equipment for the absence/presence of L. monocytogenes as part of their sampling scheme. Supporting guidelines for sampling of food processing areas and equipment for the detection of L. monocytogenes have been already published however a systemic approach or vigilance concept is currently not existing.
(3) PROMISE has shown the dimension of the problem and it must be concluded that almost every animal food production is endangered by transmission of L. monoyctogenes from the processing environment to the FPE. A concept looking to eradicate the pathogen from FPEs is unrealistic (promoted as the “Seek and destroy” concept). An essential impact of PROMISE is that some countries started to look into this issue for the first time. It is all about knowing (problem evidence) and to increase the willingness to react (response). If food business operators learn to control L. monocytogenes in their processing environment then the risk for cross contamination into food under processing will be reduced and the margin of safety will be elevated. Most problems occur in FBO through “silenced” contamination scenarios (contamination scenarios that are not getting to the public). FBO have to recall food lots, suffer from increased costs for investigations. Developing procedures to control the pathogen in the environment might help to reduce such costs.
(4) We have exchanged protocols, and shared the conceptual ideas. This is best documented by the long list of publications that were set up by researchers from OMS an NMS.
(5) Beyond scientific excellence a further impact lies in the conceptual work towards such a vigilance concept thus increasing the at-line response capabilities of European food producers. The concept has been published in a book edited by core members of the PROMSIE consortium (Kieran Jordan; Ed Fox; Martin Wagner).
The modelling WP will try to show what data sources (outbreak, prevalence), as reported through PROMISE or at (super)national level mean in terms of interpretability and reliability. It will provide a framework for improved appreciation of fluctuations in outbreak and case numbers annually reported by EFSA or ECDC
(6) The impact lies in a better comparability of prevalence data. Prevalence estimates are subject to the phenomenon called “magic of numbers” and very often (mis)used for political statements (“member state X must be good in public health cause it reports half the prevalence of disease in comparison to member state Y”). PROMISE has tried to shed a light behind the curtain and tried to develop a model allow to quantify factors that contribute to over- or under reporting.
The PROMISE research efforts were embedded into activities allowing us to level out discrepancies among participating partners both with respect to research capabilities, surveillance strategies and risk communication procedures (WP 4 and WP6).
(7) PROMISE has increased the access to state-of-the-art analysis through exchange of personal, knowledge and technical skills. Experts organized Technical Training and Workshops for phD students and junior PostDocs basically educating the methods that were successfully used for WP1 – WP3. The mobility program supported 14 highly ambitious young students to learn about the methods in a hand-on style. Some collaboration has led to a long lasting collaboration and future career tracks (eg Luminita Cirolacu, Romania, now working in Belgium; Jasna Kovac SLO now going to work in the USA)
(8) Subconsortia were successful in novel grant applications (e.g. Turkey/Spain/Austria in the Leonardo program Safe Meat).
The exchange with Food Safety Authorities was a core activity in PROMISE (WP5). The collaboration proved to be based on confidence, and the researchers learnt a lot regarding risk management and communication.
(9) The impact is given in the fact that all scientific information was put into the appropriate channels for further reception and dissemination. It is the legislative bodies that have to decide on action, countermeasures and intensified communication (with stakeholder bodies; food processing operations). The food safety authorities were part in the project but also partners for dissemination. More than 20 workshops, many of them done in NMS in the mother tongue of the organizers, had a huge impact on the spread of knowledge and critical discussion of results. We are very proud that these approaches have stimulated discussions on risk management tremendously.
Taking all together the PROMISE impacts are multiple and highly satisfying the expectations of the partners. Almost all PhDs have finished their programs and are currently doing their final exams. Attracting young individuals for food science is one of the most important goals of PROMISE and all (wishes for) personal careers have been documented on the PROMISE Academy by video.
List of Websites:
http://www.promise-net.eu
http://www.promise-academy.eu
Prof. Martin Wagner
martin.wagner@vet