Abstract
The incidence of hepatitis A virus (HAV) infection has been low in developed countries for decades; however, many adults in these countries are susceptible to HAV infection. In recent years, the global trade of food products originating from HAV-endemic countries resulted in HAV outbreaks associated with imported foods in developed countries. This article aims to review the characteristics of selected HAV outbreaks associated with imported food in developed countries during 2012–2018, and discusses improvements in global public health capabilities and new tools for effective detection, control, and prevention of HAV outbreaks.
Background
The incidence of hepatitis A virus (HAV) infection is strongly associated with socioeconomic status (Jacobsen and Koopman, 2005). Globally, with the improvement of hygienic conditions and childhood vaccination, the morbidity of HAV infection has been substantially reduced in both developed and developing countries, but HAV infection rates vary greatly between countries (Jacobsen and Wiersma, 2010; Aggarwal and Goel, 2015; Carrillo-Santisteve et al., 2017). HAV infection is vaccine preventable. Protection against HAV infection is afforded by passive prophylaxis, or vaccination as pre-exposure prophylaxis or as postexposure prophylaxis to an individual within 2 weeks of exposure (Victor et al., 2007). Transmission is through the fecal–oral route by the consumption of contaminated food or water, and through person-to-person contact (Daniels et al., 2009). According to WHO criteria (World Health Organization, 2012), many developed regions including North America, Australia, and European countries have been classified as low or very low endemicity regions (Jacobsen and Wiersma, 2010; Aggarwal and Goel, 2015; Carrillo-Santisteve et al., 2017).
The risk of clinical disease from HAV infection is determined primarily by age of the infected person: infections during early childhood are often entirely asymptomatic, but for adults, the infection may cause severe illness that requires hospitalization and, in rare cases, can be fatal (World Health Organization, 2010; Collier et al., 2015). There has been little circulation of HAV in low endemicity countries, but a large proportion of adults lack the immunity against HAV (Bassal et al., 2017; Yoon et al., 2017). Once the virus is introduced to these countries, either by contaminated food or by person-to-person contact, outbreaks with an extensive spread of HAV may be triggered, and the morbidity and mortality associated with HAV outbreaks in low endemicity countries can be severe. This phenomenon was manifested recently in the United States where >15,000 HAV infection cases have been reported since 2016 (U.S. CDC, 2018), with higher rates of hospitalization (1073 cases, 71%) and death (41 cases, 3%) than usual (Foster et al., 2018). According to the latest publication from the U.S. CDC (Centers for Disease Control and Prevention), the transmission of HAV is believed to be occurring through person-to-person contact. No common food sources were found, despite speculations. Various interventions were deployed, including educational and intensive vaccination campaigns targeting the at-risk population, but HAV outbreak cases are still on the rise in some states (Kentucky Cabinet for Health and Family Services, 2018).
HAV exhibits a high degree of resistance to low pH and temperature (Siegl et al., 1984; Scholz et al., 1989), but can be killed by heating and become inactivated at 90°C for 180 s (Sow et al., 2011). Foodborne HAV outbreaks are often caused by contaminated frozen or dried food that does not involve heating before consumption (Petrignani et al., 2010; Gallot et al., 2011; Collier et al., 2014; U.S. CDC, 2016; Viray et al., 2018). With the industrialization of food production and globalization, many high-risk products (shellfish, fresh or frozen fruits, and vegetables) are produced in HAV-endemic countries and consumed in developed countries (Berger et al., 2010; Strawn et al., 2011), contributing to the HAV outbreaks in food-importing countries (Nordic Outbreak Investigation Team, 2013; Collier et al., 2014; Swinkels et al., 2014; Severi et al., 2015; U.S. CDC, 2016; Viray et al., 2018; Franklin et al., 2019). Foodborne HAV outbreaks may not present as “outbreaks,” as the majority of case reports can be sporadic, and the long incubation period of the infection and low viral counts in food that still can cause infection, make it challenging for epidemiologic and laboratory investigations.
Foodborne outbreaks should be a concern in all countries where there are a considerable number of susceptible adults (Bassal et al., 2017; Yoon et al., 2017). At present, we are not aware of any promising therapeutic options against HAV infection. Foodborne HAV outbreaks may involve a significant number of persons and may lead to high morbidity and necessitate multiple public health interventions, so public health preparedness is critical. This article aims to characterize recent foodborne HAV outbreaks and to propose improvements and modernization of outbreak detection and response.
Methods
We reviewed HAV foodborne outbreaks from January 1, 2012 to December 31, 2018. Eligible studies in English were searched in PubMed, Web of Science, and proMED. The search strategy was to combine terms on hepatitis A and outbreak, and the search was adapted for each database. The search string used in PubMed was as follows: “Hepatitis A virus”[Mesh] OR “hav”[Title] OR “hepatitis a virus”[Title/Abstract] OR “hepatitis a viruses”[Title/Abstract] OR “infectious hepatitis a”[Title/Abstract] “hepatitis a virus hav”[Title/Abstract] OR “hepatitis a virus infection”[Title/Abstract] AND outbreak*[Title].
Only outbreaks in developed countries were included. We defined developed countries according to the gross domestic product by the International Monetary Fund (IMF, 2008) and the United States, Japan, Germany, France, Italy, the United Kingdom, and Canada constituted the group of developed countries. The 15 members of the European Economic Area (EEA) and the 4 newly industrialized Asian economies are also included as developed countries. We excluded outbreaks that occurred before 2012; HAV outbreaks not caused by contaminated food were excluded; review articles, letter and news, and nonhuman studies on HAV were also excluded.
We extracted data on characteristics of foodborne outbreaks from the articles including country, size, duration, and implicated food and genotype of outbreak strains. We also extracted the demographic information and clinical characteristics of the outbreak-related patients (age, hospitalization rate, and death rate). Meanwhile, details about the detection, testing, and intervention measures were also reviewed.
Results
In total, nine outbreaks were included in the final review (Fig. 1). The characteristics of foodborne outbreaks and demographic and clinical characteristics of outbreak-related patients are shown in the table 1. All nine HAV outbreaks were associated with imported food, and frozen food served as the main vehicle for transmission (Nordic Outbreak Investigation Team, 2013; Collier et al., 2014; Swinkels et al., 2014; Severi et al., 2015; The Institute of Environmental Science and Research Ltd., 2016; U.S. CDC, 2016; Enkirch et al., 2018; Viray et al., 2018; Franklin et al., 2019). All the food-exporting countries except for China and Poland are considered high or intermediate endemicity countries for HAV (Jacobsen and Wiersma, 2010). The presence of HAV were confirmed in four outbreaks (Severi et al., 2015; U.S. CDC, 2016; Viray et al., 2018; Franklin et al., 2019), and the methods of food testing were available in two outbreak reports: one was based on ISO/ST specification 15216-1 (Franklin et al., 2019), and the other was conducted according to FDA protocol (Viray et al., 2018).

Summary of literature search.
Foodborne Hepatitis A Outbreak in Developed Countries, 2012–2018
The median age of all patients involved were ranging from 24 to 47 years.
Europe economic area countries are consisted of Austria, Bulgaria, Denmark, England, Finland, France, Germany, Ireland, Italy, the Netherlands, Norway, Poland, and Sweden in this article.
NA, not available; U.S. CDC, The United States Centers for Disease Control and Prevention.
Only the outbreak in Hawaii was local, and all the others were multi-state/county or even multi-country outbreaks, and unknown links between cases occurring at different times and in distant locations were revealed by molecular methods. In general, patient specimens from various points in time were sent to laboratory for RNA sequence analysis. Sequence analysis was carried out to characterize the outbreak strain and to perform the sequence alignment with strain sequences from other HAV outbreaks in the national or international sequence database. Subsequently, the outbreak-related cases were defined as a person with HAV infection and with a sequence identical to the HAV outbreak strain. In Australia (Franklin et al., 2019) and Sweden (Enkirch et al., 2018), samples from HAV infection patients were sequenced using the HAV network protocol (460 nucleotides in the VP1-2B region of the HAV genome). The HAV sequencing protocols in European countries were not harmonized before 2015; hence, HAV genomic fragments of different lengths were characterized (Severi et al., 2015). During the Canada outbreak, 373 nucleotides in the VP1-2A region were analyzed (Swinkels et al., 2014). For the United States, comparison of the sequencing results was generally based on 315 nucleotide segments of the VP1-2B junction (Collier et al., 2014; Viray et al., 2018).
The affected population size and time course varied greatly between the nine outbreaks: the most important outbreak associated with frozen berries involved 1589 cases across 13 EEA countries (Severi et al., 2015), with the most cases occurring in Italy (1438 cases). The duration of this EEA outbreak was comparatively longer than those in the United States (Collier et al., 2014) and Canada (Swinkels et al., 2014) (Table 1).
Of the 9 outbreaks, the pomegranate aril-associated outbreak in the United States had 8 secondary cases that occurred in households or through other contact with the patients, representing 4% of the outbreak cases; of the 908 cases with available information, the number of secondary cases in the European outbreak was 39, also accounting for 4% of the cases (Table 1). Australia had the highest rate of secondary transmission (10%) (Franklin et al., 2019). Of the outbreaks with information on genotypes, five were genotype IB and one was genotype IA (Table 1).
With regard to controlling measures, all countries that faced outbreaks recalled the contaminated food except Australia, but the health authority of Australia did request that the relevant exporters provide labeling on their products indicating that these foods must be heated before consumption (Enkirch et al., 2018). Vaccinations were provided in six outbreaks, included outbreaks in Sweden, Hawaii, the EEA, the United States (2013, 2016), and Canada. Many of them were offered through the local public health authorities and communicated through news releases to media channels. In the EEA, postexposure prophylaxis was provided to people who had been in contact with cases. The authorities of Hawaii and the United States (2013) took a step further by contacting individuals who had consumed the contaminated food directly and suggesting that they monitor themselves for symptoms of hepatitis and get vaccinated if indicated.
Discussion
Of the nine foodborne outbreaks we reviewed, seven were caused by contaminated food products originating from regions with high or intermediate endemicity for HAV. Frozen berries and seafood that have not been heated before eating may pose a greater risk (Collier et al., 2014; U.S. CDC, 2016; Viray et al., 2018; Franklin et al., 2019). The foodborne outbreaks led to hospitalization in significant proportions of cases and resulted in secondary transmissions to contacts. Moreover, food contaminated by HAV may result in outbreaks that involved as many as more than 1000 cases (Table 1). Multiple public health interventions were often needed, including providing immunoglobulins and vaccinations to contacts or individuals at high risk, and these interventions can be costly. As more foods are imported from endemic regions through the global supply chain, local and international health authorities, the food industry, and consumers should all be aware of HAV contamination.
Improvements in the infrastructure for molecular characterization of HAV isolates can improve the detection of HAV outbreaks. HAV infections often occur sporadically over weeks and even months because of the long incubation period up to 45 days (Lemon, 1985), making it difficult to establish an association between food and illness. Molecular characterization can detect the unnoticed link between cases occurring at different times and in different places (Amon et al., 2005; Verhoef et al., 2011; Collier et al., 2014; Severi et al., 2015; U.S. CDC, 2016; Enkirch et al., 2018; Franklin et al., 2019), and patients with identical sequence patterns and similar epidemiologic characteristics usually suggest a common-source exposure, which makes it possible to trace to the origin of contaminated food (Nainan et al., 2006). But the lack of molecular databases hampers strain comparison when countries use different protocols for HAV sequencing (Severi et al., 2015). Standardization and sharing of sequences with public health agencies, like the PulseNet, Foodborne Diseases Active Surveillance Network (FoodNet), and the Epidemic Intelligence Information System for Food- and Water-borne Diseases (EPIS-FWD), and real-time comparison and analysis of genome sequencing may speed foodborne outbreak detection (Food and Drug Administration, 2019).
The EPIS-FWD was supported by epidemiologists and microbiologists from all European Union (EEU)/EEA countries. In our review, three multi-country foodborne HAV outbreaks were reported through EPIS-FWD. In these outbreaks, shortly after Sweden, Denmark and Italy first reported HAV infections and shared urgent enquiries in EPIS-FWD, other Nordic and EEU countries reported locally acquired and travel-related cases associated with the same outbreak strain. In the end, because of the collaboration among public health authorities of the affected countries, imported frozen berries were identified as the vehicle of the abovementioned outbreaks. Prompt analyses of surveillance information and timely reports triggered a rapid and coordinated response among affected countries and European CDC (Nordic Outbreak Investigation Team, 2013; Severi et al., 2015; Enkirch et al., 2018). A standard sequencing protocol followed by a broad sharing of sequencing data across borders had assisted countries in early detection and control.
For the detection of HAV in food, laboratory methods are improving and should be optimized for testing of high-risk products. Collecting contaminated food promptly is challenging because of the long incubation period of HAV. In the meanwhile, viral loads in food are typically lower than in clinical samples, and more sensitive methods are needed (Sanchez et al., 2007). At present, the ISO method 15216-1:2017 is recommended to detect virus in food by the European Committee for Standardizations and the International Organization for Standardization (ISO, 2017). Chou and Williams-Hill (2018) and Fraisse et al. (2017) claimed their methods may provide an improvement of virus detection in food with low viral copies, but these methods need to be further evaluated.
High-quality epidemiologic investigations are also important. Canada and the United States halted a potentially worsening outbreak by recalling suspected contaminated food promptly in the absence of identifying HAV from food because the epidemiological evidence was considered sufficient to trigger a product recall (Collier et al., 2014; Swinkels et al., 2014). Conversely, in the case of the European outbreak with inconclusive epidemiology or laboratory evidence, the authorities could not pin down the exact contaminated food or promptly remove suspected product(s) from the market, leading to continued transmission over a period of 1 year in at least five different countries (Severi et al., 2015); high-quality epidemiologic investigations are essential for both detection and control of outbreaks, including coordination of the trace-back and trace-forward activities to curb HAV outbreaks.
Modern investigation of and response to HAV outbreaks associated with imported food requires analyses of various data sources. Data on complex supply chains, global HAV epidemiology and genotype distribution, membership data at the retail level, and food frequency data of the general population and subpopulations were critical for narrowing down the contaminated food product, for fast product recall, and targeted vaccination in Canada and the United States (Collier et al., 2014; Swinkels et al., 2014; Viray et al., 2018). The ability to follow the global movement of a food product and its constituents through the stages of production, processing, and distribution is challenging, but modern technology has enabled all supply-chain partners to store relevant data within their databases and provide the data to trusted partners or government authorities on request (Zhang and Bhatt, 2014). Complex data could be managed using new blockchain technology, which has no central database and provides several advantages including scalability and robustness (Bhatt and Gooch, 2017).
Notifying people at high risk for HAV infection and delivering postexposure prophylaxis to them are essential for fast control of outbreaks. The outbreak associated with imported pomegranate arils in the United States was halted because the food retailer used automated phone calls to notify >250,000 customers who bought contaminated food according to their membership information (Collier et al., 2014). Finally, HAV infection can be prevented through pre-exposure prophylaxis (Fiore et al., 2006), and many countries, including the United States and China (Wasley et al., 2005; Cui et al., 2009) have introduced HAV vaccines into their national childhood vaccination programs. Providing pre-exposure and postexposure prophylaxis in the broad community by vaccination is crucial for halting the spread of outbreaks, but several U.S. cities were constrained by vaccine supply (Furlow, 2017). Vaccine stockpile and distribution, and implementation research carried out in advance on gaps in knowledge about mass vaccination should be part of the preparedness to respond to remerging HAV outbreaks.
In addition, halting contaminated food into low endemic countries, enhancing prevention in the food-producing countries through vaccination, and risk-based food safety inspections targeting high-risk products should be considered. In the United States, the FDA placed contaminated food products on import alerts that were critical in controlling the spread of HAV during the outbreak (Collier et al., 2014). In Australia, local authorities have engaged in testing commodities from food manufacturers that were linked to the HAV outbreak (Franklin et al., 2019).
Our study also has limitations: first, we addressed the HAV issue by using data from developed countries, but some developing countries, such as China, have become low endemicity countries in recent times. As more countries go through epidemic transitions, greater awareness is needed among all relevant stakeholders regarding the risk of HAV outbreaks associated with imported food. Second, we only obtained data from articles that were published, so the number of HAV foodborne outbreaks worldwide may be underestimated.
Conclusion
In today's globally intertwined world, HAV infection poses a threat to all countries with large numbers of susceptible adults. The HAV outbreaks associated with imported food in developed countries require new approaches and tools for early detection and control. For public health preparedness within and beyond the country borders, surveillance data on HAV infection should be monitored and viral sequencing data should be collected and shared. Good examples from the EEA and other jurisdictions have shown the utility of these kinds of platforms. In response to known outbreaks, public notifications, product recalls, and mass vaccination campaigns are important strategies. Improvements along the spectrum of awareness by the food industry to enlist big data technology are needed for protection and effective control of HAV infection.
Footnotes
Author Disclosure Statement
No competing financial interests exist.
Funding Information
This study was funded by National Science and Technology Major Project of China (Grant No. 2018ZX10201002).
