Transcript FOOD ALLERGEN TESTING

FOOD ALLERGEN TESTING
Peter Cressey
ESR
Allergen Seminar
Waipuna Hotel and Conference Centre
21 March 2007
Specialist Science Solutions
Manaaki Tangata Taiao Hoki
protecting people and their environment through science
Presentation Content
• What are we testing for?
• What techniques are available?
• What are the strengths and
weaknesses of various methods?
• Common pitfalls and problems in
allergen testing
• When to use testing
© ESR 2007
What are we testing for?
• True food allergies are an immunological
response to specific food proteins
• Test methods can target proteins or the genetic
material that produces the proteins (DNA)
• The test method needs to be as specific as
possible for the food material of concern
• The test method need to be able to detect the
material of concern in as wide as possible a
range of circumstances (e.g. processed informs
of the food including heat treated)
© ESR 2007
What Techniques are Available?
• Protein-based methods (usually antibody based)
that determine proteins characteristic of the
allergenic source material (but not necessarily
the allergenic proteins)
• DNA-based methods that determine a
characteristic sequence of base pairs (not
necessarily related to the allergenic proteins)
Most allergen testing utilises protein-based
methods (Enzyme-Linked ImmunoSorbent Assay
– ELISA) detecting a specific sequence of amino
acids (epitope)
© ESR 2007
Rapid Antibody-Based Methods
• So called ‘dipstick’ or ‘lateral flow’
methods
• Extracted proteins are captured on
coloured particles with antibodies
attached
• Particles are allowed to migrate through a
support medium, containing a zone of
antibodies
• A positive result is visualised as a
coloured line on a dipstick or test strip
© ESR 2007
© ESR 2007
Rapid Allergen Methods
•
•
•
•
Fast (<10 minutes)
Sensitive (detection limits ~5 ppm)
Qualitative
Suitable for screening of foods and
environmental swabs
• Currently available for peanut,
gliadin (gluten), milk
© ESR 2007
Sandwich ELISA
© ESR 2007
Sandwich ELISA
• Extracts added to reaction well
coated with bound antibodies
• Second set of antibodies are then
added with enzyme linked to the
antibody
• Coloured chemical added, the
enzyme converts the chemical to a
different form with a different colour
© ESR 2007
© ESR 2007
Sandwich ELISA
• Test methods are exacting and
usually take 1-2 hours to complete
• Sensitive (sub-ppm to ~5 ppm,
depending on kit)
• Quantitative*
• Available for most major allergenic
source materials, except fish
© ESR 2007
Pitfalls with ELISA
• Detection level too high (some kits
designed for adulteration detection –
percentage level, not ppm)
• Kit not validated for substrate or state
of allergen (heat-treated, non-heattreated)
• Not following kit insert instructions
• Hydrolysed or fermented proteins
• Non-homogeneity in test material
• Too low detection limit – what do you
do with the result?
© ESR 2007
Limit of Detection too High
• Until recently one of the few soy kits
available had a detection limit of
5000 ppm (0.5%) soy protein
• Soy allergy sufferers have been
reproted to react to as little as 88 mg
soy protein or 18 g of a food
containing 5000 ppm of soy protein
• This kit was designed for grain
adulteration not allergen detection
© ESR 2007
Kit not Validated for Substrate or
State of Allergen
• With some kits cooked egg protein only showed
5-20% of the response of raw egg protein.
However, most egg allergic individuals are
allergic to both raw and cooked egg
• Almond protein was recovered from a biscuit
matrix quantitatively, however recoveries from a
fruit pulp were less than 20% of expected values
• Tropomyosin, the target protein for crustacean
kits varies significantly in content between prawn
species and between prawns an dotehr crustacea
© ESR 2007
Not Following Kit Insert
• Elisa kit methods are highly empirical and
changing times, temperatures and concentration
may not have the expected results
• A US regulatory agency caused an unjustified
Class I recall through a decision to multiply and
analytical result by a factor of 25
• We have no local examples of this, BECAUSE WE
DON’T DO IT
© ESR 2007
Hydrolysed or Fermented Proteins
• The sequence of amino acids utilised by
the test may be different to the sequence
causing the allergic reaction
• Hydrolysis and fermentation break up
protein structure and may destroy either
the test or allergic sequence
• Consequently, the food may test positive
while it is not allergenic (probably okay) or
it may test negative when it is still
allergenic (not okay)
© ESR 2007
Non-Homogeneity of Test Sample
• The test is only as good as the
sample tested!
• Allergen contamination of a food has
the potential to be distributed in a
non-uniform manner (‘hot spots’)
• Sampling needs to take this into
account
© ESR 2007
Limit of Detection Too Low
• Some kits are extremely sensitive (e.g. limit of
detection 0.5 ppm) and will elicit positive test
results at even lower concentration (e.g. 0.1 ppm)
• The most sensitive individiuals reported in the
literature react to about 0.1 mg of protein from the
allergenic source
• At a detection level of 0.1 ppm, the most sensitive
allergy sufferers would need to consume 1 kg of
the offending food to experience an adverse
reaction
• However, positive test results cannot be ignored,
no matter how low
© ESR 2007
Qualitative or Quantiative?
• Many kits can adequately detect the qualitative
presence of allergenic source material
• Quantitative measurement requires a target
protein that is:
- Always present at about the same
concentration
- Unaffected by food processing
- Able to be extracted quantitatively from a
range of foods
• Many kits appear to be quantitative when, in fact,
they are only qualitative
© ESR 2007
Test kit validation – what is ESR
doing?
• Sensitivity – what is the actual limit of detection?
• Specificity – does the kit cross react with other
than the target proteins?
• Precision – how repeatable are results?
• Accuracy – does the kit result represent the
actual amount of material present?
• Confounders – does processing affect the
performance of the kit (e.g. heating of
ingredients)?
© ESR 2007
Test kit validation – typical results
(Tepnel peanut kit)
• Sensitivity – LOD approximately 1.5 ppm of
peanut
• Specificity – no detectable cross reaction with
range of foods (soy, almond, cashew, lentils,
sesame, cornmeal, linseed)
• Precision – CoV of approximately 20%
• Accuracy – average spike recovery 106%
• Heat affect – 90% recovery after heating for 2
hours at 100 C
• Validated for biscuit matrix, other matrices would
need to be validated on a case by case basis
© ESR 2007
Performance of Currently
Available Test Kits
• While kits all appear similar, there seems to be
considerable variation in the degree of
development and validation carried out
• Peanut kits from major manufacturers are
generally the best and have been through AOAC
method assessment process
• Some kits are perfectly satisfactory for screening,
but not really suitable for quantitative testing
• Some kits (e.g. crustacea) would only be of
limited value for screening, unless the test
situation was well defined (i.e. known allergen
source in known food)
© ESR 2007
Conclusions
• A range of antibody-based assays are
available for most common food allergens
• Assays require no or minimal laboratory
resources (unless quantitation is required)
• Limitations of assay need to be
acknowledged
• Assays should be validated for particular
situations
© ESR 2007
When Should Allergen Testing be
used?
• Investigation of consumer
complaints (critical incidents)
• Validation of sanitation methods
• Incoming raw product validation
• (Finished product testing?)
© ESR 2007