Fonterra’s botulism bungle
07 Nov, 2013
In August 2013, dairy company Fonterra confirmed that milk whey protein from the company’s Hautapu manufacturing facility had become contaminated with the harmful botulism-causing bacteria Clostridium botulinum.
This prompted the Ministry for Primary Industries (MPI) to issue a warning not to use certain brands of infant formula, fearing that it contained the bacteria.
The reaction from retailers and consumers throughout the world was swift, with some shops removing all New Zealand milk products from their shelves and temporary trade bans being imposed on New Zealand milk exports. The reaction seemed extreme and the economic impact on our country in the millions but, as microbiologist Dr Siouxsie Wiles from the School of Medical Sciences at University of Auckland explained at the time, the ‘toxin A’ produced by C. botulinum is one of the most toxic known substances. “One kilogram of it would be enough to kill the entire human population!”
C. botulinum can cause botulism, a sometimes fatal illness, characterised by flaccid paralysis, blurred vision, drooping eyelids, nausea, vomiting, diarrhoea and/or constipation and cramps. However, the disease is rare. John Brooks, Professor of Food Microbiology at AUT, reported that there have only been a couple of cases in New Zealand in the last 35 years. It was unthinkable that it could be in milk formula destined to be consumed by the most vulnerable sector of the population.
Contaminant incorrectly identified
Except that it turns out that it wasn’t C. botulinum at all. The contaminant had been misidentified.
By the end of August 2013, independent testing by MPI (contaminated samples were sent to four overseas accredited laboratories) on Fonterra’s whey protein concentrate had found no trace of C. botulinum. Testing did find the closely related bacteria Clostridium sporogenes, which could cause food spoilage but posed no risk to human health.
On releasing the results, MPI’s acting Director-General Scott Gallacher said, “A failure of hygiene during processing remains a concern for customers incorporating whey protein concentrate into their products. However, the concern primarily relates to quality and the potential for spoilage when used in foods that support growth of Clostridium sporogenes from spores.”
Politicians were outraged at the bungle and demanded answers about how this trade catastrophe and false positive could have happened. Enquiries are on-going, especially around the long timelines involved between suspecting a contaminant (the suspect whey was made in May 2012) and releasing the (mis)information. Fonterra has already said the test that incorrectly detected C. botulinum was done by AgResearch. However, in several media reports, AgResearch says it urged Fonterra to seek further testing.
Two bacteria almost identical
Dr Wiles explained to the Science Media Centre how the mix-up could have happened, saying that the two bacteria are almost identical except for the presence of the genes for botulinum toxin.
“There are different ways of identifying contaminating organisms, the cheapest of which is to isolate the microbe by growing on selective agar and then performing biochemical tests to determine species. Species can also be identified genetically by looking for particular DNA sequences – like a barcode. On these kinds of tests, C. botulinum and C. sporogenes would be identical.
“Next comes the test to determine if the strains have the capacity to produce toxins – again this can be done by looking for the presence of the specific genes. If positive for toxin genes, the strains are tested to see if they actually produce toxin and if that toxin is active using a mouse bioassay. The mouse bioassay could take some weeks to carry out but the genetic identification of the presence or absence of the toxin genes can be done in hours to days.
Public alerted before confirmation
“It sounds to me like Fonterra alerted their customers to the presence of C. botulinumin their product before confirming whether the contaminating strain could produce toxin and was therefore really C. botulinum. From a public relations point of view, this was the right decision as the backlash against Fonterra would have been fierce if they had waited for confirmation of toxin production, had it proven to be positive.
“Fonterra have a lot to learn about how they could have handled the situation better. Just publicising the fact they didn’t know if the strain they had identified had the ability to produce toxin or not, rather than staying silent, would have shown they were being proactive in ensuring the safety of their product.
Bacteria – friends or foes?
“As an aside, it looks like researchers are trying to use C. sporogenes to help kill tumours. I’ve just read something from a couple of years ago where scientists in Europe have engineered the bug with an enzyme that activates a drug to kill cancers. Apparently, they may be moving into clinical trials. Just shows how context is everything when thinking of whether bacteria are friends or foes!”
Somewhat prophetically, while expressing concern that “it’s a serious lapse in process control and obviously should not have occurred” if the dirty pipe scenario is true, Professor Brooks also expressed surprise at the initial identification, saying back in early August 2013 that it is not usual to test dairy products for the presence of Clostridium botulinum. “When bacteria occur in a product at very low level and very infrequently, testing is ineffective in assuring safety and the cost is prohibitive. An Australian specification for whey protein concentrate does not mention Clostridia.”
Because of the rarity of the botulism-causing bacteria in New Zealand, we have no laboratories accredited to test for it. Accreditation means that the labs have been examined by a regulatory organisation and found to have the appropriate test equipment and scientist competency and that the particular test is carried out routinely with some results independently checked. For this reason, MPI sent the test samples overseas to accredited laboratories to be tested and appropriately identified – the results were returned within 3 and half weeks.
- 07 November 2013