12 EXAMPLES OF FSOs FOR BACTERIA OR TOXINS IN SEAFOOD PRODUCTS (Lone Gram)

12.1 Listeria monocytogenes in RTE seafoods^[9]

Both FAO/WHO (2001) and FDA/FSIS (2001) are currently in the process of carrying out quantitative risk assessments on Listeria monocytogenes in ready-to-eat foods. This section relies heavily on these documents. L. monocytogenes is a ubiquitous bacterium typical of decaying plant material and it is also associated with several animals. L. monocytogenes can cause listeriosis in humans. The main form of listeriosis is a food-borne infection which affects particular risk groups such as immuno-compromised, elderly and neonates. Recently, a milder form of gastro-enteritis affecting otherwise healthy people was reported. Many ready-to-eat (RTE) food products have been linked to listeriosis which typically occurs in sporadic, small outbreaks. L. monocytogenes is halo- and psychrotolerant and capable of multiplying in RTE foods, especially with extended shelf lives. Whilst dairy and meat products seem to be the most common causes of listeriosis, the disease has also been traced to lightly preserved fish products such as smoked mussels or cold-smoked fish (trout).

L. monocytogenes can easily be isolated from RTE food products in low concentrations. Thus between 0 and 80% of samples of cold-smoked fish are positive for the organism. It typically occurs in levels of (10 /gram but is sporadically isolated at higher levels. Inoculated trials have shown that rapid growth may occur in the vacuum-packed chill-stored product. Based on German data on prevalence and levels of L. monocytogenes, Buchanan et al. (1997) developed a dose-response curve for the organism. The study used cold-smoked salmon as the food case. This study, as well as the very thorough studies by FAO/WHO and US FDA conclude that although one cannot define a threshold concentration, i.e. a minimal infectious dose, low levels of the organism ((100 cfu/g) are very unlikely to cause the disease. The WHO/FAO team concluded as part of an expert consultation in May 2001 that if levels of L. monocytogenes were kept below 1000 cfu/g at point of consumption, then 99% of all listeriosis cases would be eliminated.

Due to the widespread occurrence of L. monocytogenes it will be extremely difficult (and expensive) to produce all RTE foods without sporadic occurrence of the organism in low levels. The dose-response relationships (and resulting Risk Estimate) indicates that such low levels constitute a very low risk. In the terminology introduced above, an "appropriate level of protection" / "tolerable level of risk" (ALOP/TLR) could be 100 cfu/g (assuming serving sizes of (100 gram). Following this line of thought, a Food Safety Objective is derived directly from the ALOP and could be 100 L. monocytogenes per gram at point of consumption.

Risk management options

In principle, two interlinked options exist for the management of microbial risks: the implementation of GHP and of HACCP. A HACCP analysis of L. monocytogenes in cold-smoked salmon reveals that with current processing and storage practices, no critical control point exists for the hazard which is growth of L. monocytogenes. The organism survives the processing steps (no listericidal step) and the typical product and storage conditions (vacuum-packed, chill-stored (5°C), NaCl at 3-6% (water phase salt) and pH of approximately 6.2) does not guarantee against growth to hazardous levels. It must be emphasised that CCPs that guarantee that counts do not increase to hazardous levels can be introduced, e.g. by frozen storage or by limiting shelf life. L. monocytogenes is capable of colonizing food processing environments and product contamination typically is caused by contamination during processing rather than by survivors from the raw material. L. monocytogenes may hide in brines, colonise slicers and have its harbouring niches in drains and on floors. Therefore the GHP programme of a food processing plant with L. monocytogenes as an identified hazard, must focus specific actions on eliminating and surveying this bacterium.

Performance standards

The performance standard (PS) is the level of the hazard (here L. monocytogenes) that the processor must meet. In several RTE products, L. monocytogenes will not grow during storage and the PS for instance at the end of processing can then equal the FSO. However, if growth of the organism is possible/likely during storage and distribution, the FSO must be translated to PS depending on the amount of growth expected between sampling and consumption. It has been demonstrated that in naturally contaminated cold-smoked salmon stored at 5°C, approximately 1 log increase occurs during a 3 week storage period (Jørgensen and Huss, 1998). Thus, using a shelf life limit of 3 weeks or shorter at chill temperatures, a PS of 10 Listeria per gram off processing line will allow the FSO to be met. Most processors will set a PS of <10 Listeria per gram to built in safety margins.

Process and Product Criteria

Process or product criteria are levels of e.g. a heat treatment or a salting concentration that ensures that the hazard is under control. The preservation and safety of cold-smoked salmon depends on use of appropriate raw materials and combinations of salt and low temperature after processing. Since no listericidal step is included in the processing and neither of the food preservation parameters will control growth of L. monocytogenes, process or product criteria cannot be identified.

Microbiological criteria

It is possible, when appropriate, to develop microbiological criteria using an FSO of 100/g or PS of lower values. Such criteria may be used as acceptance criteria in situations where the pre-history of the product is not known, such as at port-of-entry or at certain retail outlets. Clearly, such criteria should only be used where other acceptance criteria, such as product criteria, cannot be developed. Also, the product should be epidemiologically linked to the hazard - or a hazard analysis should indicate reason for concern (van Schothorst, 1996). This is the case with cold-smoked salmon since listeriosis has been linked to cold-smoked trout (Swedish outbreak with 9 infected people and 2 fatalities), and several inoculated food trials have demonstrated growth in the product (Table 12.1).

Table 12.1 Questions evaluating the use of MC for Listeria monocytogenes in cold-smoked salmon (based on CAC, 2001)

As with other microbiological criteria, careful consideration must be given to the choice of sampling plans and the degree of assurance it provides. Currently spreadsheet systems are available that allows one to determine the performance of a particular sampling plan (http://www.foodscience.afisc.csiro.au/icmsf/samplingplans. htm). For instance, if a sampling plan with 20 samples are used and c= 0 and m=100, then there is a 95% (or higher) probability of rejecting lots if the mean concentration of Listeria in the lot is (15 cfu/g. It therefore follows that even with 20 samples, the probability of accepting a lot which actually contains L. monocytogenes increases rapidly if the mean concentrations drops below 15 cfu/g.

Similarly, a sampling plan with 10 samples and c=0 and m=100 has a 95% (or higher) probability of rejecting the lot if the mean concentration is (30 cfu/g. If a sampling plan uses only 5 samples and c=0 and m=100, then there is a 95% probability of rejecting the lot if the mean concentration is (80 cfu/g. These figures emphasise the well-known fact, that low levels of pathogens are difficult to control using product sampling and testing.

If products are inspected just before consumption or the products do not support growth, the MC can equal the FSO. Depending on the assurance required from the sampling, i.e. the probability of only accepting acceptable lots, the number of samples is decided upon. If growth is supported, a PS and a MC of "not detectable" in 25 g may be opted for.

12.2 Staphyloccocal enterotoxin in cooked crustaceans

Staphylococcus aureus is, as described in Chapter 5, a mesophilic, Gram-positive bacterium associated with warm-blooded animals. It is a common member of the skin and nasal microflora of humans. Many strains of S. aureus may produce enterotoxins which upon ingestion causes a sudden reaction in terms of cramps, abdominal pain and vomiting. Several different enterotoxins may be produced and they have, based on antigenic properties been divided into sero-types A to J. Enterotoxin A is assumed to be the most commonly involved in food-poisoning outbreaks, however, recently type C has become prevalent (Jablonski and Bohack, 1997).

S. aureus is commonly detected from foods, either raw foods from warm-blooded animals or foods that have been manually handled. S. aureus can be detected sporadically on raw fish but is clearly more typical of seafood products that have been heat treated and manually handled, such as crustacean products (Table 12.2).

Table 12.2 Prevalence of S. aureus in seafood commodities (modified from Jablonsky and Bohach, 1997).

The enterotoxins are part of a larger family of toxins produced by S. aureus (and Streptococcus pyogenes), the pyrogenic toxin family that can act as so-called super-antigens. These toxins can provoke a very strong response from the host immune defence system.

Disease symptoms may occur with ng levels of enterotoxin per gram, however, in most disease outbreaks, an estimated 1 to 5 mg has been ingested. The dose depends on the food matrix and the consumer - thus several children became ill after eating chocolate containing 100-200 ng enterotoxin. The enterotoxins are small molecules that are not degraded by gut proteases and that are relatively heat-stable and require prolonged boiling to inactivate.

Enterotoxins are produced in extremely low amounts during exponential growth but production increases markedly in the late exponential phase and stationary phase. It therefore follows that marked growth of S. aureus has to take place before toxic levels of enterotoxin are formed. Thus, it is the growth and toxin-production of the organism that is the hazard - not its mere presence. Stewart et al. (2002) when assaying 94 samples detected toxin in all samples with a positive OD-reading (i.e. cfu (10⁷ /ml) and not in any samples with a negative OD-reading. In a range in inoculated foods, Notermans and Otterdijk (1985) found that enterotoxin was not detected in any sample with less than 10⁷ cfu/g. Many samples with higher S. aureus counts were positive (>0.1 mg enterotoxin/100g) but also several samples with counts of 10⁹-10¹⁰ were negative. Hence, it can be concluded that (10⁶ cfu/g food are required to produce toxin at hazardous levels (Adams and Moss, 2000).

S. aureus is a poor competitor with respect to other microorganisms and outbreaks have mostly been associated with cooked foods, that have been manually handled and temperature abused. Thus cooked, hand-peeled crustaceans which may be temperature abused are high-risk products.

Quantitative risk assessments using mathematical representations of all steps from farm-to-fork have not been conducted on staphyloccocal enterotoxins. However, based on evaluations of dose-response, an FSO of 1 mg per gram (of cheese) has been suggested as an example of an FSO (van Schothorst, 1998). However, as indicated above lower levels have been causing disease when ingested in a protective (fatty) food matrix and when consumed by children. Therefore an FSO of 50 ng may be a safer option.

Risk management options

The presence of S. aureus in cooked seafoods is clearly a result of cross-contamination from people handling the product. Therefore the GHP procedures must specify the hygienic level during handling of cooked foods. Obviously people with sores or infected scratches must not handle such foods. Also, sneezing and coughing which will spread S. aureus from the nasal reservoir must be avoided. Gloves and mouth protection ware may, if used appropriately, minimise the spread. However, control of the hazard (growth and toxin production) is, in principle straight forward. Clearly, all storage/processing conditions that prevent growth (see tables in Chapter 5) will control the hazard. Proper cooling is essential to prevent growth. Also, staphylococcal enterotoxins are formed under a more limited range of conditions compared with growth (ICMSF, 1996). Whilst S. aureus may grow down to 7°C, toxin is not formed below 10°C and only very limited amounts are produced between 10 and 20°C. Growth occurs down to a_w values of 0.83 but toxin production stops at 0.87 (Table 12.3).

Performance Standards

Although the FSO is based on the agent, the toxin, it is not likely that producers of foods in which S. aureus growth is a risk will measure toxin on a regular basis. Therefore most criteria and standards "translate" the toxin levels to levels of S. aureus. As mentioned, significant growth is required for toxin to be produced.

Table 12.3 Limits for growth and enterotoxin production (from ICMSF, 1996)

1. NaCl % (water phase salt) calculated based on water activity values. Note that some studies report 20% as maximum for growth and 10-15% as maximum for toxin production.

Therefore the FSO of 50 ng toxin/gram could theoretically be translated to 10⁵-10⁶ cfu/g. This requires that the number is reached by growing, toxin producing bacteria and not a result of massive recontamination. However, the producer is unlikely to set e.g. 10⁵ S. aureus per gram but will target a much lower level to incorporate extra safety. In most foods, a PS of 100 S. aureus per gram will ensure that the FSO is met - assuming that the appropriate controls are in place.

Process and Product Criteria

In cooked crustaceans, the most important criteria ensuring control of the hazard is keeping the temperature low (< 10°C). The products do not per se include other preservation parameters that can be relied upon for growth control. However, if the cooked crustaceans are used for brined foods, the combination of low pH, salt and preservation compounds such as sorbate or benzoate can guarantee that growth does not occur.

Microbiological Criteria

The EU (EC, 1991) has set a microbiological standard for S. aureus in cooked crustaceans with a 5 sample sampling plan and c = 2, m= 100 cfu/g and M=1000 cfu/g. Such standards are widely used at port-of-entry where there is no knowledge of the GHP or HACCP programmes of the producer.

References

Adams, M.R. and M.O. Moss 2000. Food Microbiology. 2^nd ed. Royal Society of Chemistry, Cambridge, UK.

Buchanan R.L., W.G. Damert, R.C. Whiting & M. van Schothorst 1997. Use of epidemiologic and food survey data to estimate a purposefully conservative dose-response relationship for Listeria monocytogenes levels and incidence of listeriosis. Journal of Food Protection 60, 918-922.

CAC (Codex Alimentarius Commission) 2001. Report on the thirty-fourth session of the Codex Committee on food hygiene. Alinorm 03/13. Food and Agriculture Organization / World Health Organization, Rome, Italy.

EC (European Commission) 1991. Council Directive 91/493/EEC of 22 July laying down the health conditions for the production and placing on the market of fishery products. Official Journal of the European Communities. No. L268. pp.15-34.

FAO/WHO (Food and Agriculture Organization/World Health Organization) 2001. Joint FAO/WHO Expert Consultation on Risk Assessment of Microbiological Hazards in Foods. Risk characterization of Salmonella spp. in eggs and broilers and Listeria monocytogenes in ready-to-eat foods. FAO Food and Nutrition Paper No. 72.

FDA/FSIS (US Food and Drug Administration/Food Safety Inspection Service) 2001. Draft Assessment of the Relative Risk to Public Health from Foodborne Listeria monocytogenes among Selected Categories of Ready-to-Eat Foods. FDA Center for Food Safety and Applied Nutrition.

ICMSF (International Commission on Microbiological Specifications for Foods) 1996. Microorganisms in Foods 5. Characteristics of Microbial Pathogens. Blackie Academic & Professionals.

ICMSF (International Commission on Microbiological Specifications for Foods) 2002. Microorganisms in Foods 7. Microbiological Testing in Food Safety Management. Aspen Publishers Inc.

Jablonsky, L.M. and G.A. Bohach 1997. Staphylococcus aureus. In Doyle, M.P., L.R. Beauchat and T.J. Montville (eds.) Food Microbiology. Fundamentals and Frontiers. ASM Press, Washington, USA. pp.353-375.

Jørgensen, L.V. and H.H. Huss 1998 Prevalence and growth of Listeria monocytogenes in Danish Seafood. International Journal of Food Microbiology 42, 127-131.

Notermans, S. and R.L.M. van Otterdijk 1985. Production of enterotoxin A by Staphylococcus aureus in food. International Journal of Food Microbology 2, 145-149.

Stewart, C.M., M.B. Cole, J.D. Legan, L. Slade, M.H. Vandeven and D.W. Schaffner 2002. Staphylococcus aureus growth boundaries: Moving towards mechanistic predictive models based on solute-spectific effects. Applied and Environmental Microbiology 68, 1864-1871.

van Schothorst, M. 1996. Sampling plans for Listeria monocytogenes. Food Control 7, 203-208.

van Schothorst, M. (International Commission on Microbiological Specifications for Foods) 1998. Principles for the establishment of microbiological food safety objectives and related control measures. Food Control 9, 379-384.