It’s a CEO’s worst nightmare. You’re 11,000 miles away from the office when your cellphone rings late at night. You see that it’s work calling. “Oh, oh,” you exclaim. “It must be bad news. They wouldn’t disturb me otherwise.”
Theo Spierings lived this bad dream as clocks struck midnight in his native Netherlands on Thursday, August 1, 2013. The chief executive officer of Fonterra Cooperative Group Ltd. was back home for a relative’s funeral. He receives the fateful phone call from one of the company’s senior managers at headquarters in Auckland, New Zealand, where it’s 11:00 p.m.
It was bad news. Laboratory tests had come back likely positive for Clostridium botulinum in some Fonterra whey protein concentrate known as WPC80. Thus was sparked a globally publicized multi-million dollar international recall.
“I began my career as a food technologist and although I knew botulism is hardly ever found in dairy, I knew we had to act,” Spierings says. “Food safety must always be our number one priority. So the first thing I did was ask what products had been affected, which customers and which geographies. We then immediately informed our board of directors and the New Zealand Ministry of Primary Industries (MPI), which regulates food safety, and went public with our precautionary recall.”
On Friday, August 2, Fonterra initiated a precautionary recall of 38 tons of its WPC80, advising the eight customers, including three food companies, two beverage companies, and three animal feed producers in Australia, China, Malaysia, New Zealand, Saudi Arabia, Thailand, and Vietnam, that had received the potentially affected ingredient.
Turn Back the Clock
The Fonterra WPC80 recall story actually begins in May 2012. On the last three days of WPC80 manufacture for the 2011/2012 New Zealand dairy season, the Fonterra whey manufacturing facility in Hautapu, New Zealand produced three batches of WPC80 out of a rework process. This manufacture occurred from May 15 to 18, 2012, with the product reprocessed and packed on May 17, 18, and 22.
The reprocessed WPC80 passed the standard, routine testing for compositional and food safety/pathogen requirements for whey powder.
Fonterra first identified a quality issue with this particular WPC80 in March 2013. At that time routine testing conducted at Fonterra’s Darnum site in Australia identified high Clostridia levels in finished product made for a customer.
Microbiology Testing
Immediately after the high Clostridia levels were detected at its Darnum plant, Fonterra began an extensive process of microbiological testing at its research center in Palmerston North, New Zealand, to determine what type of Clostridium it was.
On May 8, 2013, a MALDI-TOF test (Matrix Assisted Laser Desorption/Ionization-Time of Flight), a sophisticated mass spectrometry technique, conducted at the research center indicated that the Clostridia were most likely Clostridium sporogenes, which isn’t a food safety hazard but at high levels can cause food to spoil, but the findings were not definitive. The research center’s further testing was unable to rule out the presence of C. botulinum.
The definitive test for C. botulinum, a mouse bioassay, could only be conducted at two recognized testing facilities in New Zealand.
So on June 26, 2013, Fonterra commissioned AgResearch Limited, a New Zealand government-owned Crown Research Institute, to conduct a mouse bioassay test to confirm whether the identified strain was C. sporogenes or C. botulinum.
On July 31, 2013 Fonterra received AgResearch’s test results, which identified the “potential presence” of C. botulinum in bacteria sourced from product containing the WPC80. The tests showed that this risk was minute, but it was still a risk.
Thus, the midnight call to Spierings while he was in the Netherlands.
Tests in the U.S.
Subsequent to the precautionary recall, MPI commissioned WPC80 testing at two U.S. labs that are accredited for C. botulinum testing, the Centers for Disease Control and Prevention, Atlanta, Ga. and the National Veterinary Services Laboratory, Ames, Iowa.
Tests using a range of technologies were conducted in these two labs and all of the tests definitively confirmed that the recalled WPC80 contained a harmless strain of C. sporogenes and was clear of C. botulinum. As soon as clear and definitive test results were received, the MPI proactively released these results to the public on August 28, 2013, 26 days after the recall was initiated.
Manufacturing Trace Back
The MPI documented its actions in response to the Fonterra WPC80 recall and on August 28, 2013 published a Whey Protein Concentrate Incident Tracing and Verification Report dated August 25, 2013.
The 61-page report includes the following details.
- The dairy material for reworking was reconstituted, chilled, and then processed as usual. Equipment for reconstituting (rehydrating) dairy material had been decommissioned, so temporary (non-standard) equipment was introduced into the plant to facilitate this processing aspect.
- All equipment was subject to cleaning in place (CIP) prior to use and after each day’s manufacture.
- Immediately prior to and following manufacture of the WPC80, the plant also manufactured a protein called hydrolysate. While there was common processing equipment used for both WPC80 and hydrolysate, the temporary equipment associated with the contaminated WPC80 was not used for the hydrolysate. This helped to identify the likely source of contamination.
- The rework was completed using manufacturing equipment common to the hydrolysate and fresh WPC80 manufacturing processes, with the exception of food grade flexi-hoses and the microfiltration line.
- The equipment operator’s investigation concluded that the cause was either within the product lines used to deliver the reconstituted product to the routine process or within a microfiltration line.
- The data suggested that either the reconstituted product line or the microfiltration line contained a protein deposit/biofilm. This is likely to have harbored Clostridia spores and may not have received the full effect of the CIP. The majority of the evidence pointed to the reconstituted product line being the cause of the problem.
- So it was determined that Clostridia contamination occurred during this reprocessing. To complete the reprocessing, an item of nonstandard equipment was used, namely a transfer pipe.
- Although standard cleaning procedures were followed, including those for the transfer pipe, before and during reprocessing, subsequent analysis of the plant by the MPI shows that the source of Clostridia was likely to have been this transfer pipe. The pipe has since been decommissioned.
Fonterra Reviews
On August 12, the Fonterra board of directors established the WPC80 Inquiry Committee, and charged them to oversee an independent review of the circumstances that led to the WPC80 recall and the subsequent chain of events. “The review is due to conclude at the end of October 2013,” says Maury Leyland, Fonterra’s group director of strategy.
Soon after initiating the precautionary recall, Spierings commissioned an inhouse Operational Review to enable the cooperative to understand the cause of the contamination and evaluate how the business handled the incident and precautionary recall. The Operational Review team was also tasked with making recommendations on how Fonterra could prevent an event like this from happening again.
Led by Leyland, the Operational Review team presented its findings and recommendations to Spierings on August 30.
The Operational Review of Fonterra’s WPC80 concluded that no single occurrence led to the precautionary recall and the events that ensued, and that no single issue was the sole contributor.
“The WPC80 incident was the result of a number of separate and unrelated events occurring in an unforeseen sequence,” Leyland says. “One of those events was the decision to reprocess the original WPC80 and not downgrade the product, in combination with the use of an item of non-standard equipment, which was the cause of the contamination.”
While the root cause of the contamination was an item of nonstandard equipment used in reprocessing the WPC80, a computer systems upgrade, as well as lapses in escalation and communication to higher levels of management also played a part in contributing to the event, Leyland points out.
“Actions have been taken to immediately address these issues,” she says. “Specifically, in relation to the reprocessing of the WPC80 and use of nonstandard equipment, we have clarified our reprocessing procedures with our teams and these have also been embedded into our HACCP plans.
“In New Zealand each year Fonterra produces 2.3 million tons of high quality dairy products from 17 billion liters of milk,” Leyland says. “To achieve this we have world-class manufacturing facilities, quality systems, and robust testing regimes in place. This event stress-tested all of them. Overall our systems worked well, while some aspects showed room for further improvement.”
To that end, the WPC80 event definitely fostered learnings, Leyland relates, and as a result, several areas have been identified where the cooperative needs to improve. “We are now well underway in implementing these improvements and we’re confident that the associated action plan will make Fonterra stronger,” Leyland emphasizes.
Government Reviews
The MPI and the New Zealand government are conducting WPC80 recall reviews.
The Government Inquiry into the whey protein concentrate contamination incident has three parts that have been divided into two tranches. Part A is an inquiry into how the potentially contaminated whey protein concentrate entered New Zealand and international markets, and how this was subsequently addressed. Parts B and C involve inquiries into regulatory and best practice requirements, and comparison of Part B’s findings to similar matters in other comparable jurisdictions, respectively.
The government has said that the inquiry will prepare an interim report on Parts B and C within three months of the September 12, 2013 publication of notice in the New Zealand Gazette, the official newspaper of the Government of New Zealand. The final full report will be produced after the MPI compliance investigation and any subsequent court proceedings have reached a stage where they cannot be prejudiced by the inquiry’s processes.
All recalled Fonterra WPC80 was accounted for or contained by the time that MPI released its Tracing and Verification Report on August 28. That same day, MPI acting director-general Scott Gallacher issued a statement confirming that, based on testing and additional information, the identified batches of WPC80 and all recalled infant formula products that contained this WPC80 were not contaminated with C. botulinum and posed no risks to consumers of contracting botulism.
To date, there have been no confirmed reports of illness as a result of any person or animal consuming this recalled WPC80 or any products that contained it. Moreover, there never was actually a botulism risk associated with the WPC80 to begin with.
MPI is undertaking a compliance investigation into the WPC80 incident that will determine whether regulatory requirements under New Zealand’s Food Act 1981 and Animal Products Act 1999 were met by all parties involved, or whether any parties may have committed any breaches or offences.
“This investigation is still underway,” Gallacher says. “As such, it’s too early to provide a timeframe for when the investigation will be completed, or when a final report of findings will be available.”
In-House Recommendations
Based on comprehensive findings, and to achieve the intended outcomes, the Fonterra Operational Review recommended 20 actions which cover four key elements of the cooperative’s business, namely people, products, systems, and response. (See sidebar below to view Fonterra’s recommendations.)
“The Operational Review concluded that Fonterra’s quality and care systems in our manufacturing and testing are robust, and underpin our reputation for leadership in the global dairy trade,” Leyland says “Transparency of information, internally and externally–despite the lapses in information sharing and escalation noted in the review’s findings–and commitment to public safety, were reflected throughout this event. Notwithstanding the shortcomings identified by the review, our staff, on most occasions, acted conscientiously on new information, and generally sought to do the right thing.
“Although the initial likely identification of C. botulinum proved not to be a food safety risk, the significant impact of the recall stress-tested a great many systems across our company’s business,” she says. “Now we are using the lessons and improvements from this event to enable us to take a leadership position in product traceability and food safety and quality in the global food and dairy industries.”
Spierings is quick to concur. “As a global leader in supplying dairy nutrition, I am confident that our action plan will make us even stronger,” he says. “Fonterra is committed to providing high quality dairy products to our customers and to people around the world, and to putting food safety above everything else we do.”
Leake is a food safety consultant based in Wilmington, N.C. Reach her at [email protected].
Botulism and Babies: It’s Not About Formula
The likelihood of getting infant botulism from infant formula is extremely rare, says Eric Johnson, ScD, a professor at the University of Wisconsin Department of Bacteriology. A respected scientist whose 30 years of research focuses on Clostridium botulinum and its toxins, Dr. Johnson has been providing expertise to Fonterra Cooperative Group Ltd. relative to its whey protein concentrate recall.
“Infant botulism almost always comes from other sources, such as vegetables or dust,” Dr. Johnson emphasizes. “There is only one reported case of infant botulism linked to the consumption of infant formula in medical literature. In this case the result was inconclusive, with experts noting that the illness might well have been caused by sources other than the infant formula.”
C. botulinum is a spore-forming bacterium and the spores are highly resistant forms of the organism that are prevalent in nearly every region of the world. To grow, these bacteria require an oxygen free environment that is warm and moist. C. botulinum bacteria are commonly found in soil, dust, and some marine environments, so most people are exposed to the spores likely on a daily basis.
Like plant seeds, spores can lie dormant for years. Spores are not threatening until they encounter an adequate environment for growth. The spores that germinate produce the deadly botulinum toxin, which is the most poisonous substance known. (The affected Fonterra WPC 80 was potentially contaminated with spores of C. botulinum, not with toxin.)
Since C. botulinum is present in the environment, many foods do become contaminated with the spores. “Most of these foods are vegetables and associated foods that contact the soil,” Dr. Johnson points out. “The number of spores in infant formula is either non-existent or extraordinarily few and so the likelihood of getting infant botulism from infant formula is extremely rare.
“Indeed, it is possible that the spores are there but they’re at extremely low concentrations because milk has, it’s been estimated, only one spore per liter when they are there,” he continues. “The number of spores in a food correlates with the likelihood of coming down with the illness.”
The number of spores required to cause infant botulism is not known, but based on honey studies it’s thought to be as low as 10 spores on ingestion up to much more—100 spores or more. “So in infant formula, the number of spores present is generally much lower than the dose required to lead to disease,” Dr. Johnson says.
There are likely less than 150 cases of hospitalized infant botulism globally each year, but about 70 percent of them occur in the U.S., Dr. Johnson says. “That’s probably because of the high incidence of the type of C. botulinum spores that cause infant botulism and the rigorous surveillance practices in place in this country.”
Botulism Testing
The diagnosis for botulism is detecting the toxin. Since it’s the toxin, not the organism, that is solely responsible for the disease, detection of the toxin is the only definitive proof. With the recalled Fonterra WPC80, no toxin was detected from clostridial isolates in the product when grown and tested.
The standard procedure for the detection of botulinum toxin is the mouse lethality assay. The test is based on an intraperitoneal injection into laboratory mice of sample diluted in phosphate buffer. If the sample contains toxin, the mice develop typical signs of botulism, including fuzzy hair, muscle weakness, and respiratory failure that manifests as a wasp-like narrowed waist.
C. botulinum is tested by its route of intoxication, which involves absorption of botulinum toxin into the blood and then trafficking to the nerve, binding to a receptor, internalization into the nerve and cleavage of a protein substrate.
“To my knowledge, there are no foods that are routinely tested for the spores of C. botulinum,” Dr. Johnson says. “Such testing requires considerable time and there would be so many non-existent or negative results that it would not be practical.
“That takes quite a long period, so the only test that will follow all those steps of intoxication currently is the mouse bioassay, which is very expensive, uses a number of animals, but is a very sensitive assay,” Dr. Johnson relates. “There are other assays being developed, including one in my lab in which neuronal cells are used to test for the toxin, but right now mouse assay is the only definitive test.”
There’s an alternative technique called Sulphite Reducing Clostridia (SRC) testing that is used in dairy manufacturing to ensure products are low in Clostridia such as Clostridium sporogenes, which isn’t a food safety hazard but at high levels can cause food to spoil. SRC testing is general in nature and tests for a number of different Clostridia species. SRC is not intended specifically for C. botulinum, which is almost unheard of in dairy products. Moreover, of the seven serotypes of C. botulinum, only three can cause botulism.
Even in the absence of evidence that the sulfite-reducing organisms that were detected in the WPC80 were C. botulinum, Fonterra embraced a proactive approach and complimented the SRC testing with a mouse bioassay where several mice were injected with C. botulinum and one of injected mice did have a fatality.
“But antibodies were not available at the time, so that test could have detected many lethal substances in food or in the culture medium in which the isolates were grown,” says Dr. Johnson in review of the Fonterra WPC80 testing regime.
In the end it was proven that the recalled WPC80 was not contaminated with C. Botulinum and no consumers, babies included, were ever at risk of botulism poisoning as a result of consuming this product.
“The development of rapid new tests for inexpensive screening for spores could be beneficial to the food industry,” Dr. Johnson adds. “In the future, it will be beneficial to replace the complex mouse bioassay with a complete assay, such as neuronal cells, that is more sensitive than the mouse bioassay and avoids use of animals for testing.”
—L.L.L.
Life After the WPC80 Recall
To prevent a similar food safety scare from happening again, the Fonterra Operational Review identified 20 recommendations.
People – A Focus on Food Safety
1. Create a group director of food safety and quality reporting directly to the CEO.
2. Strengthen the remit and scope of Fonterra’s Food Integrity Council, chaired by the group director food safety and quality.
3. Incorporate food safety expectations in all new employment agreements.
4. Improve the management of customer complaints and export non-conformances with respect to escalation, root cause analysis, response, and continuous improvement opportunities.
5. Amend people’s performance objectives across the organization to reinforce the emphasis on food quality and safety as a mandatory objective.
6. Develop a comprehensive suite of people initiatives to lift the focus on food safety and quality.
Products – Quality Control and Testing Regimes During Manufacturing
7. Reset of Hautapu site. (“This encompassed further checking and servicing of all equipment and processes to ensure it continues to meet the highest possible international standards,” Maury Leyland explains.)
8. Establish interim Sulphite Reducing Clostridia testing program.
9. Conduct a specialized audit of Fonterra’s most sensitive plants that make ingredients or products consumed by infants and very young children.
10. Identify and audit any nonstandard or temporary manufacturing setups in Fonterra’s New Zealand plants.
11. Introduce more robust systems to authorize and ensure compliance of any nonstandard processing (including rework) and testing.
12. Conduct specialized quality audits for all Fonterra sites, including offshore businesses, third-party manufacturing, and joint ventures.
13. Conduct a full review of product standards with clear focus on when product is placed on hold and authorizations around release.
14. Explore options to more proactively align with customers and regulators on standards.
Systems – Traceability
15. Define, document, and communicate traceability protocols for all areas of the business.
16. Conduct an immediate review of significant system changes planned for product traceability and customer complaint management to ensure adequate change and risk management.
17. Conduct a wider review and overhaul of Fonterra-wide product traceability capability, including product outside of Fonterra’s control.
18. Institute a quality hotline for Fonterra staff and contractors to encourage early escalation.
Response – Transparency
19. Ensure Fonterra has a well-rehearsed crisis management capability.
20. Review and clarify communication protocols with respect to third parties (including customers, industry bodies, and government) in a crisis.
—L.L.L.
ACCESS THE FULL VERSION OF THIS ARTICLE
To view this article and gain unlimited access to premium content on the FQ&S website, register for your FREE account. Build your profile and create a personalized experience today! Sign up is easy!
GET STARTED
Already have an account? LOGIN