2005N-0147 Sprout Safety Public Meeting
FDA Comment Number : EC5
Submitter : Dr. Keith Warriner Date & Time: 07/20/2005 04:07:33
Organization : University of Guelph
Category : Academia
Issue Areas/Comments
1. What concepts or underlying principles should guide efforts to improve the safety of sprouts?
1. What concepts or underlying principles should guide efforts to improve the safety of sprouts?
Historically, the majority of sprout related outbreaks has been traced to contaminated seed. Therefore, the underlying principle of any guidelines needs to focus on preventing contamination during the seed production stage. This may extend encouraging Good Agricultural Practice on farms and implementing traceability systems. However, given that seed production cannot be closely monitored it is imperative that more effective seed decontamination methods are developed and implemented across the sprout industry. Microbiological screening of spent irrigation water should be implemented but not relied upon to detect contaminated batches of contaminated seed. Indeed, microbiological results should be used to verify the effectiveness of control measures as opposed a test-and-release program.

Implementation of HACCP-based approach should be encouraged to establish universal standards within the sprouting industry. Consideration should also be given to including the end user of sprouted seeds (retail, food service and consummers) to maintain close tempertature control to minimize the risk of pathogen growth.

Consideration to organic growers and identifying risks associated with specific seed types.

2. Which practices primarily contribute to the contamination with harmful pathogens of seeds used for sprouting?
2. Which practices primarily contribute to the contamination with harmful pathogens of seeds used for sprouting?

Evidence to date would suggest that the seed used in sprout production represents the primary source of pathogens. Seeds can potentially become contaminated at each stage of the production. However, relatively few studies have been undertaken to establish which stage of seed production is most critical. Neverthless, our own studies (unpublished data) have shown mung beans can be contaminated with Salmonella when the pathogen is introduced onto the flowers of growing mung bean plants. In parallel studies, when E. coli O157:H7 was inoculated onto the flowers the pathogen was only recovered sporadically on the subsequent seeds. The conclusion derived from the study was that if contaminated irrigation water is applied during the cultivation of plants it is possible that the seed would be contaminated. In addition, the lower persistence of E. coli O157:H7 on growing plants maybe a contributory factor for the low incidence of this pathogen with sprouted seeds.
From work with other plants (lettuce, tomaoto, spinach) is becoming established that human pathogens interact with growing plants to a greater extent than hitherto identified. Indeed, the persistence of pathogens can be enhanced by being encorporated into rhizosphere of growing plants. There has been speculation that human pathogens can become internalized within plant tissue and subsequently distributed to the airial plant tissue.
However, internalization is typically restricted to seedlings as opposed to mature plants and is unlikely to contribute to seed contamination. Of course if seeds contact soil ammended with non-composted manure then it is possible the direct contamintion of the seed will occur.

In several seed producing countries it is common practice for animals to graze of plants following harvesting. In addition, the application of fecally contaminated water onto growing crops does occur in seed producing nations. This may provide a route for enteric human pathogens to be introduced to soil. It should be noted that human pathogens such as Salmonella and E. coli O157:H7 can survive in soil over extended periods. Wild animals can be an additional source of human pathogens.

Crops such as alfalfa are externally pollinated and therefore there is the potential for bees and other insects to indirectly transfer contamination to the flowers of growing plants. Experiments with fruit flies have suggested that this maybe a route by which apples can become contamined when cultivated to adjacent domestic animal farms.

Mung beans are heavily irrigated during the flowering stage and can therefore by susceptible to contamination. In contrast alfalfa is typically sprouted under low moisture conditions that are not condusive to human pathogen survival.

The exclusion of domestic animals and screening of irrigation water are both areas were control can be excercised. However, ensuring complience with overseas farmers on GAP will be unrealistic. In many coutries water resources are scarce and hence the grower is typically limited to alternative sources should fecally contaminated water is detected. Technologies exist to decontaminate water although these would be finacially unfeasible with seeds destined for sprout production.
On-farm HACCP principles have been encouraged in Austrailian mung bean production but not widely applied in alfalfa which is principally used for animal feed.

Seed mixing should be avoided to enhance traceability. Direct sourcing of seeds from reliable growers should be encouraged. The key problem, however, is that very few seed producers are willing to meet on-farm HACCP since seeds destined for sprouting represent only a small part of their buisness.

Alfalfa is grown in areas that have high density of dairy cattle. Hence direct or indirect contamination from bovine fecal material is a strong possibility.

For alfalfa seed, and other perennial crops, the season begins when the forage that has grown during the winter or early
3. Which practices primarily contribute to the contamination with harmful pathogens of sprouts?
3. Which practices primarily contribute to the contamination with harmful pathogens of sprouts?
The contribution of seedborne vs environment to contamination of sprouts remains an unexplored area. Given that the majority of sprout related outbreaks have been traced to contaminated seed batches it could be concluded that seedborne contamination remains the most significant source of pathogens. This is supported by studies that have illustrated that the microflora of sprouts is independent of being sprouted under commercial or asceptic conditions. Neverthless, with all plants the release of nutrients by germinating seeds is thought to attact and support the growth of microflora from the surrounding environment. In recent research we have illustrated that the prevalence of oppertunitic contaminats (generic Escherichia coli, thermotolerant coliforms and mesophilic Aeromonas) on sprouting mung beans changes over time (8 weeks in the present study) that would suggest that environmental sources could potentially provide a source of contamination. Although non-pathogenic the recovery of traditional fecal indicators (i.e. E. coli and thermotolerant coliforms) may indicate that enteric human pathogens may also follow a similar temporal trend. Although sprout related outbreaks associated with Listeria monocytogenes have been very rare it is possible that the pathogen can become endemic within sprouting facilities. Therefore, although there is strong evidence to confirm seeds as the main source of contamination the potential
for environmental sources should not be dismissed without further research.

Seed decontamination still remains the most effective method of controlling human pathogens in sprout production. Although many methods have been evaluated none have proven totally effective. However, through studies initiated back in 1999 we have developed and evaluated a seed decontamination method based on using a sanitizer subsequently referred to as SDH. By using mung beans as a model system, it was demonstrated that SDH could be used to inactivate a five strain cocktail of either Escherichia coli O157:H7 or Salmonella introduced onto beans at a loading of 103-104 cfu/g. Salmonella was more tolerant to SDH compared to E. coli O157:H7 with sanitizer levels of >150ppm and 100 ppm respectively, being required to ensure pathogen free sprouts. The decontamination efficacy was also found to be dependent on treatment time and the seed to sanitizer ratio (>1:4). With the optimized treatment no pathogens were detected in sprouts or spent irrigation water derived from sprouting seed beds. Mung bean batch size (25-500g) or time of sampling (48-96h into the sprouting process) had no effect on the efficacy of the SDH seed decontamination treatment.

Naturally contaminated seeds derived from inoculated mung bean plants could be effectively decontaminated using the optimized SDH based protocol. Interestingly Meliagradis (originally isolated from alfalfa) was the only Salmonella serovar recovered from naturally contaminated seed and dominated the microflora when artificially introduced along with 4 other strains.

SDH treatment did not exhibit phytotoxic effects as germination and sprout yields were not significantly (P>0.05) different compared to non- treated controls. Although human pathogens could be effectively eliminated form beans the total viable count on subsequent sprouts was not significantly different to non-treated controls. The diversity of microbial populations (determined through 16S rRNA-DGGE analysis) associated with bean sprouts were not significantly affected by SDH treatment. However, in sprouts derived from SDH treated seed it was noted that Klebsiella and Herbasprillium (both common plant endophytes) were absent. It is unclear how SDH appears to selectively inactivate microbial populations associated with sprouted seeds.

When a further range of seed types were evaluated it was found that alfalfa, clover, cress, flax and soybean could be decontaminated using SDH.
4. Although FDA's current recommendations address practices by all parties, efforts to promote adoption of effective preventive controls have focused largely on sprouting facilities.
4. Although FDA's current recommendations address practices by all parties, efforts to promote adoption of effective preventive controls have focused largely on sprouting facilities.
Yes, the guidelines should be made more specific and avoid vauge references to sanitation procedures, see decontamination and microbiological screening. Specifically, provide a list of seed decontamination methods.
4. Do the preventive controls recommended in FDA's sprout guidances need to be explained?
4. Do the preventive controls recommended in FDA's sprout guidances need to be explained?
Any on-farm safty strategies must have a strong scientific foundation for seed producers to realize they have a responsibility for supplying contamination-free seed.

Control measure could include:
Madatory seed testing
Irrigation water testing
Restricting access of domestic and wild animals to seed destined for sprout production.
5. Is a regulation likely to be an effective means of achieving the goal of minimizing foodborne illness associated with the consumption of sprouts?
5. Is a regulation likely to be an effective means of achieving the goal of minimizing foodborne illness associated with the consumption of sprouts?
Regulation will not improve the safety of sprouts until an effective seed decontamination method is implemented and/or representative microbiological screening method developed.
6. How can progress toward the overarching goal (to minimize foodborne illness associated with sprout consumption) be effectively measured?
6. How can progress toward the overarching goal (to minimize foodborne illness associated with sprout consumption) be effectively measured?
Monitoring results of spent irrigation water testing
Screen seeds for pathogens at the primary production level
Surveillance of outbreaks linked to sprouts.

7. There is broad variation within the seed and sprout industry, including variations in size of establishments, types of sees and sprouts produced, practices used in production
7. There is broad variation within the seed and sprout industry, including variations in size of establishments, types of sees and sprouts produced, practices used in production
There is an increasing number of different seed types used within the sprout industry. It follows that the characteristics of sprouts and production methods also differ, For example, mung beans are sprouted in bins as opposed to alfalfa that is cultivated on trays. Hence, the distribution of contamination within the sprouting seed bed would be more heterogenous for bean sprouts (mung beans) than for alfalfa (sprouted as a monolayer). This was indeed found to be the case in a study we recently performed (abstract below).

Abstract

Abstract
The reliability of spent irrigation water testing to assess the microbiological status of sprouting mung bean beds has been investigated. In
commercial trials, the distribution of opportunistic contaminants within 32 bean sprout beds (25 kg mung beans/ bin) was assessed 48 h post germination. The prevalence of generic E. coli, Thermotolerant coliforms and Aeromonas in sprouts (n = 288) was 5%, 11% and 39% respectively, while it was 57%, 70% and 79% in the corresponding spent irrigation water samples (n=96). Contamination was heterogeneously distributed within the seedbed. In laboratory trials, beans inoculated with a 5-strain cocktail of either Salmonella or Escherichia coli O157:H7 were introduced (1g/500 g non-inoculated seed) at defined locations (top, middle or base) and sprouted over a 48 h period. With inoculated seed introduced at the base or top of the seedbed the pathogens were typically restricted to these sites and resulted in 44% of the spent irrigation water samples returning false-negative results. Introducing inoculated beans into the middle or at the pre-soak stage enhanced the distribution of both pathogens within the subsequent sprout bed and resulted in comparable levels recovered in spent irrigation water. The study demonstrated that even though screening single spent irrigation water is more reliable than testing sprouts directly it does not provide an accurate assessment on the microbiological status of sprouting mung bean beds. Such limitations may be addressed by ensuring bean batches are mixed prior to use and by taking spent irrigation water samples from multiple sites at the latter stages of the sprouting process.

The rate of sprout development should also be considered. Bean spouts germinate within 24h of the sprouting process, however, broccoli can take up to 3 days (buck wheat even longer). Therefore, it can be envisaged that testing spent irrigation water after 48h for broccoli seed would be inappropriate.

The relative risk of seeds being contaminated by enteric pathogens should also be considered. For example, alfalfa is at a greater risk of being contaminated with pathogens compared to seeds such as broccoli or wheat. Therefore, it is unlikely that all seed types0 require to be decontaminated prior to the sprouting process.

8. Are there existing food safety systems or standards (such as international standards) that FDA should consider as part of the agency's efforts to minimize foodborne illness associated with the consumption of sprouts?
8. Are there existing food safety systems or standards (such as international standards) that FDA should consider as part of the agency's efforts to minimize foodborne illness associated with the consumption of sprouts?
There are guidelines being prepared by the Food Standards Agency (United Kingdom) but these are predomonantly based on those published by the FDA.