Bacteria are by far the most common and best known causes of foodborne illness in the developed world and the food industry is well aware of the risks presented by Salmonella, E. coli O157 and other bacterial pathogens. But there are also a small group of disease-causing viruses that can be transmitted by contaminated food. Noroviruses are the most familiar to many people, but the hepatitis A virus (HAV) is also an important cause of disease worldwide. In 2013 an unexpected spate of foodborne HAV outbreaks in Europe and North America drew attention to the virus as a potential food safety hazard. A closer look at the characteristics of HAV suggests that it would be dangerous to underestimate the threat of the virus in today’s food supply chains.
HAV as a cause of disease
The HAV is an enteric virus and, as the name implies, it causes a disease of the liver in humans and other primates. This is now referred to as hepatitis A, but was previously known as infectious jaundice, or viral hepatitis. There are several other hepatitis viruses, but HAV is the only one known to be transmitted in food, though there is some evidence that the hepatitis E virus can sometimes be foodborne too. Taxonomically, the virus is a member of the genus Hepatovirus, which belongs to the Picornaviridae family, a group of non-enveloped, single stranded RNA viruses. HAV is usually spread by direct person-to-person contact through the faecal-oral route, but many foodborne outbreaks have been recorded around the world, often caused by infected food handlers or contaminated water. The biggest foodborne outbreak of hepatitis A ever recorded occurred in Shanghai in China during 1988. At least 290,000 people were affected and the source of the virus was discovered to be clams harvested from sewage-polluted waters.
The disease hepatitis A is not usually serious and in many cases, especially in children, there may be no apparent symptoms at all. Where there are symptoms, they appear about four weeks after infection and are typically mild, initially consisting of headache, fatigue, a poor appetite, fever, and nausea and vomiting. About seven days later liver disease becomes apparent in the form of jaundice or an elevation in the level of enzymes as the virus multiplies in the cells of the liver itself. Most people recover within about eight weeks in the absence of medical intervention, but a small number may suffer recurring symptoms for several months and in older age groups particularly the disease can occasionally be fatal. Data from the USA indicates that the mortality from HAV infection in the over fifties is almost 2%, whereas in the general population it is about 0.3%. Those infected, including asymptomatic individuals, often excrete large numbers of virus particles in their faeces and this may continue for several months. This prolonged shedding of viruses is the main reason for the high risk of HAV spread by direct contact, or through sewage contaminated water. In fact the main reservoir for HAV is the human intestine.
The incidence of hepatitis A has been falling in many developed countries where clean water supplies are the norm and good hygiene is practised. For instance, the European Centre for Disease Prevention and Control (ECDC) has compiled figures for 21 EU member states (plus Iceland and Norway) for the period from 1995 to 2004. These show that almost 210,000 cases were reported in total, although the number of cases that were foodborne is unknown. There was a peak in the incidence of hepatitis A in the EU in 1996-97 when it reached more than 10 cases per 100,000 people. After that peak the number of cases reported each year began to fall and since 2000 the figure has been less than 4 cases per 100,000. The incidence varies between different European countries and in some, such as Lithuania, Estonia and Slovakia, it remains much higher than the EU average, but the overall trend is downward.
The situation in the developing world is markedly different. In many countries the disease is endemic and many people become infected with HAV during childhood. Since childhood infections are often asymptomatic and confer lifelong immunity from the disease, reported outbreaks of hepatitis A are not common and very many cases almost certainly go unrecognised. This means that a very large reservoir of infection is constantly present in some parts of the world without necessarily causing a serious local public health problem. This dramatic difference in incidence between developed countries and the developing world has the effect of making the citizens of European countries more susceptible to infection. Comparatively few Europeans are now exposed to the disease in childhood and so the great majority have no immunity. Thus travelers to areas where hepatitis A is endemic are at risk of infection through environmental or food contamination, and contaminated food products imported into Europe from high-risk countries are more likely to cause outbreaks among vulnerable consumers.
Not surprisingly therefore, many cases of HAV infection reported in developed countries are now travel-related, but sporadic cases and clusters of infection also occur among people with no history of travelling to countries where hepatitis A is endemic. It is probable that many of these are foodborne, although it can often be very difficult to identify a food vehicle of infection because of the long incubation time before symptoms appear. This dictates that epidemiological investigations are hampered by the interval between consuming a contaminated food and becoming ill. Patients are less likely to remember exactly what they ate a month ago than meals consumed more recently and any remaining food is likely to have been discarded. Nevertheless, plenty of foodborne outbreaks have been identified and documented.
Recent foodborne outbreaks
Outbreaks of hepatitis A are often associated with filter-feeding shellfish, like oysters and mussels, taken from sewage-contaminated waters. Viruses in the water can accumulate within the bodies of the shellfish and cause infection if cooking is inadequate. However, there is another category of food that can also be contaminated with HAV and cause outbreaks. Fresh produce grown and harvested in regions with a high incidence of hepatitis A has been implicated regularly. Outbreaks linked to imported strawberries, blueberries, raspberries, lettuce and fresh orange juice have been reported and in 2003 green onions grown in Mexico were responsible for an outbreak affecting more than 600 people in the USA, three of whom died as a consequence of illness.
More recently semi-dried tomatoes were implicated in outbreaks in Australia (2009) in the Netherlands (2010) and the UK (2011) and genetic analysis of the viruses involved suggested the possibility of an international outbreak from a single source. Investigation of another outbreak in France in 2010 revealed that the food involved was a batch of frozen semi-dried tomatoes imported from Turkey and defrosted before being placed in oil and used in sandwiches. While it has not been established that these outbreaks were all linked, and the original sources of the contamination have never been identified, these events hinted at the possibility for large-scale international hepatitis A outbreaks caused by imported foods.
That possibility seems more like a probability now after a series of outbreaks were reported in Europe and North America in 2013. In the USA, an outbreak across ten states had affected at least 159 people by mid-August. An epidemiological investigation established a link with a blend of frozen berries and seeds marketed as “Townsend Farms Organic Anti-oxidant Blend” and sold through two retail chains. Further investigation suggested that the most likely vehicle for the HAV was a shipment of pomegranate seeds supplied by a Turkish company. The virus involved in the outbreak was characterised as genotype 1B, which is rare in North America, but common in North Africa and the Middle East. The same genotype was identified in a 2012 outbreak in Canada linked to a frozen berry blend containing pomegranate seeds from Egypt.
Meanwhile, another significant outbreak was under investigation in Scandinavian countries. By August 106 cases of infection by HAV genotype 1B had been reported in Denmark, Finland, Norway and Sweden since October 2012, all among people with no history of travel outside western Europe. Epidemiological data pointed to frozen berries as the culprit, but further investigations by Danish public health experts identified frozen strawberries imported from Morocco and Egypt as the most likely vehicle.
Simultaneously, a much larger multinational outbreak was in progress in Europe. According to the European Food Safety Authority (EFSA), at least 1440 people in 12 countries were infected by HAV between January 2013 and September 2014. Of these, 331 cases in Denmark, Finland, France, Germany, Ireland, Norway, the Netherlands, Poland, Sweden and the UK were confirmed by genotyping as being caused by the same strain of genotype 1A HAV. A large number of potentially related cases were also reported in Italy. Once again, an epidemiological link with frozen berries was established and HAV was detected in frozen mixed berries and mixed berry cakes/pastries in Italy, France and Norway. Evidence supporting the link was also found in Ireland, the Netherlands and Sweden. Eventually, 43 different lots of frozen mixed berries were traced back to source and the most common ingredients were found to be Bulgarian blackberries and Polish redcurrants, but no single point source of contamination was ever identified. The investigators concluded that the virus probably entered the supply chain either during primary production or at the freezing stage.
Very recently, in February 2015, another outbreak of HAV infection occurred in Australia and affected at least 14 people in New South Wales, Queensland, Victoria and Western Australia. Once more, frozen berries were identified as the most likely vehicle of infection, resulting in a recall of four frozen mixed berry and raspberry products marketed by company based in Victoria. The implicated products all contained raspberries and were packed in China, but the exact source of the contamination has not been identified and it is quite possible that berries from other countries could be involved.
While there is no evidence that all these recent outbreaks are linked – two different HAV genotypes were identified and the outbreak strains differed significantly – it is very unusual to see several at the same time. Researchers from the European Centre for Disease Prevention and Control who studied the outbreaks have recommended that more research is needed into the risks presented by frozen berries, along with better surveillance and increased laboratory testing for HAV contamination in food.
Designed to survive
The HAV has a number of characteristics that render it uniquely well adapted for foodborne transmission. Like all other viruses, it is unable to multiply outside the cells of its host, and so cannot grow in foods or in water. However, the number of virus particles needed to cause infection is thought to be as few as 10-100, so that foods do not need to be heavily contaminated to cause disease. Unlike many other viruses, HAV seems able to survive in quite harsh environments for long periods. For example, virus particles present in human faeces are able to remain infective for several months in the soil or in contaminated water. HAV is also unusually resistant to chlorine and other disinfectants and can survive exposure to acid environments.
Importantly from a food industry point of view HAV is more resistant to heat than other enteroviruses and has been shown to survive a temperature of 70oC for up to 10 minutes. Crucially, this means that many standard cooking processes designed to kill Listeria and bacterial pathogens (e.g. 70oC for at least 2 minutes) are insufficient to inactivate HAV. In fact the World Health Organisation recommends that shellfish from HAV-contaminated waters should be heated to 90oC for 4 minutes or steamed for 90 seconds. Equally important is the ability of the virus to survive at low temperatures for long periods. It has been demonstrated that HAV can remain infective even after storage at -20oC for up to two years. This helps to explain some of the recent hepatitis A outbreaks associated with frozen berries and their long duration.
Until quite recently it has also been quite difficult to detect HAV in food samples. Traditional cell culture-based methods do not work well for detecting HAV, partly because the virus does not grow easily in cell culture, but also because it does not usually cause visible changes in animal cells when freshly isolated from the environment. The advent of molecular biology-based detection methods has changed this and there are now published international standard methods for quantitative (ISO/TS 15216-1:2013) and qualitative (ISO/TS 15216-2:2013) determination of HAV and norovirus in food. These methods are based on amplifying and detecting target sequences within viral RNA by real-time RT-PCR. Commercial products using the same technology are also now available. An example is the Hepatitis A Virus Detection Kit developed and manufactured by CEERAM SAS and distributed by Life Technologies. This is designed to be suitable for detection of HAV in environmental and food samples and is claimed to be sensitive enough to detect from 1-10 copies, depending on the quality of RNA purification.
The recent hepatitis A outbreaks in Europe and North America highlight the risk posed by HAV in the food chain and the difficulty of eliminating it from contaminated foods. For instance, during the outbreak in Nordic countries consumers were advised to boil imported frozen berries before consumption to ensure safety. Since the outbreak was linked to frozen strawberries, it was unlikely that the advice could be followed without reducing the fruit’s appeal considerably. In fact the current consensus is that the decontamination of most fresh foods containing HAV is not practical unless they are intended for consumption only after thorough cooking.
The best strategy for controlling HAV in the food supply chain is to take measures to prevent the initial contamination. Where contamination might occur in a food processing or preparation environment as a result of contact with infected food handlers, standard personal hygiene best practice should be sufficient to achieve this. It is also important to monitor shellfish harvesting waters for sewage contamination and to close fisheries if pollution occurs. But for food production in areas where hepatitis A infection is endemic, the situation is more complicated. For example, water used to irrigate and process crops must come from unpolluted or disinfected sources and it is important to keep young children – who may be carrying an asymptomatic infection – away from fresh produce growing and harvesting operations as much as possible. Measures as basic as providing proper toilet and hand-washing facilities for workers in the fields and in packing facilities can also make a significant contribution to reducing the risk of contamination. In short, what is needed is the application of good agricultural practice (GAP) and improved hygiene in primary production, along with a good standard of hygiene and sanitation at the packing and processing stage.
The recent outbreaks should act as a warning to all food businesses sourcing fresh produce from regions where hepatitis A is endemic. The HAV can be widespread in the growing and harvesting environment and is well equipped to survive, even in frozen foods. Those businesses need to be aware of the risk and be careful how they source fresh produce, ensuring that they know its exact origin so that they can confirm that the growers have the proper control measures in place. It may also be possible to monitor the supply chain for HAV periodically by testing now that rapid and reliable detection methods are available. Supplier assurance is becoming ever more difficult as supply chains become longer and more complex and the demand for out-of-season fruit and vegetables all year round increases, but marketing imported produce that is not completely traceable could have disastrous consequences, both for the business and for its customers.