Risk Assessment for Food Allergens

1
Risk Assessment for Food Allergens
2
Risk Assessment Approaches to Evaluation of Food
Allergen Hazards

• Hazard Identification
• Dose/Response Evaluation
• Exposure Assessment
• Risk Characterization

3
Dose/Response Evaluation

• Trace amounts can elicit allergic reactions
• Severity of response is related directly to dose
• Individuals vary in degree of sensitivity
• How much is too much?

4
Exposure Assessment

• How frequently are food products contaminated
  with potentially hazardous levels of unlabeled
  allergens?
• How frequently do allergic reactions occur?

5
(No Transcript)
6
Sources of CluesRegarding Food Allergen
Thresholds

• Allergen cross contact episodes
• Double-blind, placebo-controlled food challenges
• Immunotherapy trials
• Clinical threshold experiments

7
INTERPRETATION PROBLEMSALLERGEN CROSS CONTACT
EPISODES

• Past analytical methods may not have yielded
  accurate results
• Lack of accurate information on amount of
  allergenic food residues
• Uncertainty about amount of food eaten
• Questions about other sources of allergen
• Individual variability in threshold

8
1st Threshold Conference

• September, 1999
• Hilton Head, South Carolina, USA

9
1st Threshold Conference

• Results
• Taylor et al. 2002. Factors affecting the
  determination of threshold doses for allergenic
  foods how much is too much?
• J. Allergy Clin. Immunol. 109 24-30.

10
1st Threshold Conference

• Results
• Considerable data presented on low dose
  challenges for peanut, egg and cows milk
• More limited data available on low dose
  challenges to other allergenic foods fish,
  mustard, soybean and tree nuts

11
1st Threshold Conference
12
1st Threshold Conference

• Major Conclusions
• Threshold doses do exist for commonly allergenic
  foods
• Thresholds were finite, measurable, and above
  zero
• Premature to reach consensus on threshold doses
  for specific foods

13
1st Threshold Conference

• Secondary Observations/ Opinions
• Reactions occur to hidden/ undeclared allergens
  in foods
• The severity of reactions to undeclared allergens
  increases with the dose of exposure
• Low/ very low dose exposures (LOAELs) result in
  mild, reversible symptoms

14
2nd FARRP Threshold ConferenceMay 20-21,
2002Palm Beach, FL
15
2nd Threshold Conference

• Results
• Taylor et al. 2004. A consensus protocol for the
  determination of the threshold doses for
  allergenic foods how much is too much? Clin.
  Exp. Allergy 34 689-695.

16
3rd FARRP Threshold Conference

• October 4-5, 2004
• Camp De Mar, Mallorca, Spain

17
Why model dose-response?

• It may not be feasible to design studies with
  sufficient power to give the desired degree of
  safety assurance (e.g., 268 patients and no
  reactions are required to say with 95 confidence
  that the reaction rate is less than 1).
• A conventional approach, applying a safety factor
  to a no effect level cannot easily be applied
• many challenge studies do not yield a no-effect
  level
• some challenge studies exclude individuals who
  have suffered a severe response, (are severe
  reactors the most sensitive?)
• No agreement yet on uncertainty factors
• Modelling is an accepted way of defining the
  probability of rare events with potentially
  severe consequences

18
Figure 4 Diagram showing one possible definition
for a allergen threshold using a log-linear model.
19
Classical Risk Assessment

• Typical Uncertainty Factors
• 10x Extrapolation from animals to humans
• 10x Inter-individual variation
• TDI NOAEL 100 (in rats)

20
Risk Assessment for Food Allergens

• Determine the NOAEL for specific allergenic food
  among humans with allergy to that food
• Apply uncertainty factor to obtain TDI

21
Determination of NOAEL

• Challenge large number of allergic individuals
• Identify NOAEL for each patient
• Identify LOAEL for each patient
• Determine variation between individuals in NOAELs
• Standardized protocol leads to consistent
  interpretation of results

22
Uncertainty Factors

• No animal to human extrapolation needed
• Have already selected sensitive subset of human
  population
• Did we include the most sensitive individual?
• Infants vs. adults

23
1st Threshold Conference
24
Relevant Literature Published LOAELs

• Peanut
• May (1976) 3
• Bock et al. (1978) 14
• Atkins et al. (1985) 2
• Pastorello et al. (1988) 2
• Oppenheimer et al. (1992) 11
• Moneret-Vautrin et al. (1995) 2
• Hourihane et al. (1997) 14
• Moneret-Vautrin et al. (1998) 10
• Sicherer et al. (2000) 24
• Wensing et al. (2002) 26
• Taylor et al. (2002) 306
• Morisset et al. (2003) 103
• Grimshaw et al. (2003) 4
• Leung et al. (2003) 84

25
Relevant Literature Published LOAELs

• Egg
• May (1976) 4
• Bock et al. (1978) 10
• Atkins et al. (1985) 1
• Pastorello et al. (1989) 1
• Norgaard Bindslev-Jensen (1992) 7
• Moneret-Vautrin et al. (1995) 8
• Caffarelli et al. (1995) 13
• Sicherer et al. (2000) 267
• Eggesbo et al. (2001) 9
• Taylor et al. (2002) 281
• Osterballe Bindslev-Jensen (2003) 56
• Morisset et al. (2003) 125

26
Relevant Literature Published LOAELs

• Milk
• May (1976) 1
• Bock et al. (1978) 10
• Bernstein et al. (1982) 5
• Hill et al. (1988) 68
• Pastorello et al. (1989) 7
• Baehler et al. (1995) 10
• Norgaard Bindslev-Jensen (1992) 4
• Moneret-Vautrin et al. (1995) 5
• Sicherer et al. (2000) 117
• Taylor et al. (2002) 299
• Morisset et al. (2003) 59
• Meglio et al. (2004) 13

27
Uncertainties Regarding Establishment of
Threshold Doses

• Adults vs. children (mg vs. mg/kg)
• Nature of challenge material
• Allergen content of challenge material

28
Nature of Challenge Material
29
Uncertainties About Challenge Materials

• Studies should be compared by using protein
  content
• If the protein content of the challenge material
  was not determined experimentally or cannot be
  determined with reliable factors, then the study
  should be rejected from consideration in
  establishment of thresholds
• Well-characterized challenge material NFDM,
  dried egg white, soy flour
• Thresholds should be established in terms that
  can be related to analytical methods (mg food)

30
Uncertainties Regarding Establishment of
Threshold Doses

• Blinding of challenges (single-blind vs.
  double-blind)
• Oral vs. labial challenges
• Choice of dosages for challenges

31
Uncertainties Regarding Establishment of
Threshold Doses

• LOAELs vs. NOAELs
• Uncertainty in using LOAEL to establish threshold
  dose
• Patient selection and exclusion of severely
  affected patients
• Variability in individual threshold doses

32
NOAELs vs. LOAELs

• Diagnostic challenges report only LOAELs
• NOAELs may not be recorded
• In many cases (how many?), the patient has
  responded to the lowest dose administered
• How far above the NOAEL is the LOAEL?
• If using LOAEL, how big should the UF be?

33
NOAELs vs. LOAELsSize of UF

• UF 1 if LOAEL based upon subjective symptoms
• UF 2 if LOAEL based on mild, objective symptoms
  at first dose and very low doses (0.1 20 mg)
  were given
• UF ? If LOAEL based on objective symptoms, at
  first dose and higher doses (400 mg) were given
• Use NOAEL in cases where objective symptoms
  occurred at doses above the initial dose

34
Uncertainties Regarding Establishment of
Threshold Doses

• LOAELs vs. NOAELs
• Uncertainty in using LOAEL to establish threshold
  dose
• Patient selection and exclusion of severely
  affected patients
• Variability in individual threshold doses

35
Selection of Subjects

• Diagnostic evaluations (DBPCFC) may be
  representative of the whole population of
  allergic individuals
• Referral clinics?
• Clinical threshold studies may be skewed toward
  the more highly sensitive

36
Selection of Subjects

• Sicherer et al. 2000. Dose-response in
  double-blind, placebo-controlled oral food
  challenge in children with atopic dermatitis. J.
  Allergy Clin. Immunol. 105 582-586.
• DBPCFC of 53 soy-allergic children
• 28 reacted at first dose (400 or 500 mg)
• 53 reacted at intermediate doses
• 19 reacted at final dose (2.0 or 2.5 g) or on
  open challenge
• Soy protein or soy flour?

37
Minimal Eliciting Dose Peanut (n103)

• Patients with suggestive symptoms (OAS shock)
• SBPCFC or DBPCFC, 5 active doses
• 20 min interval
• Dose range 5 700mg/10 5000mg
• Roasted crushed peanuts in mashed potatoes
• Severe symptoms abdominal pain (3.3), asthma
  (20), drop in BP (3)
• Age ED (cum)
• Children lt 15 mg - gt 7110 mg
• Adults lt 15 mg - gt 7110 mg
• (one patient 5 mg)

Morisset et al. CEA 2003331046-51.
38
Severely Affected Individuals

• Have they been excluded from challenge trials?
• Do they have lower minimal eliciting doses?
• Do they experience severe reactions at very low
  doses?
• Have they simply made big mistakes in their
  avoidance diets?

39
(No Transcript)
40
Minimal Eliciting Dose Peanut (n26)

• Adult patients with suggestive history (OAS
  shock)
• Double blind challenge, 7 active doses, 7 placebo
• Randomly interspersed, 15 30 min interval
• 85 defatted roasted peanut flour in mashed
  potatoes
• 26/26 had symptoms, 5/26 objective
• Symptoms OAS (26), lipswelling (3), nausea (2),
  vomiting (2)
• Clinical score mild, moderate, severe
• Age (mean) ED obj (µg) ED obj (cum) (µg)
  
• 25 10 - 30 14.43 44

Wensing et al. JACI 2002110915-20.
41
(No Transcript)
42
Uncertainties Regarding Establishment of
Threshold Doses

• LOAELs vs. NOAELs
• Uncertainty in using LOAEL to establish threshold
  dose
• Patient selection and exclusion of severely
  affected patients
• Variability in individual threshold doses