Aflatoxin Risk Assessment Red Book Model Exercise

1
Aflatoxin Risk Assessment Red Book Model
Exercise

• Charles Yoe, Ph.D.
• College of Notre Dame of Maryland

2
Risk Analysis

Risk Management
Risk Assessment
Risk Communication
3
Risk Assessment

• What can go wrong?
• How can it happen?
• How likely is it?
• What is the magnitude of the effect?

4
What are the steps?

• CODEX
• Hazard identification
• Hazard characterization
• Exposure assessment
• Risk characterization

• NAS
• Hazard identification
• Dose-response assessment
• Exposure assessment
• Risk characterization

5
Risk Assessment CODEX

• Hazard Identification
• The identification of known or potential health
  effects associated with a particular agent.
• Hazard Characterization
• The qualitative and/or quantitative evaluation of
  the nature of the adverse effects associated with
  biological, chemical, and physical agents which
  may be present in food. Dose-response
  assessments should be performed if the data are
  available.

6
Risk Assessment CODEX

• Exposure Assessment
• The qualitative and/or quantitative evaluation of
  the degree of intake likely to occur.
• Risk Characterization
• Integration of hazard identification, hazard
  characterization and exposure assessment into an
  estimation of the adverse effects likely to occur
  in a given population, including attendant
  uncertainties.

7
Risk Assessment NAS

• Hazard Identification
• Determine if exposure to an agent causes an
  increased incidence of an adverse health effect.
• Dose-Response Assessment
• Characterize the relationship between exposure
  (at different levels or doses) and the incidence
  of the adverse health effect.

8
Risk Assessment NAS

• Exposure Assessment
• Measure or estimate the intensity, frequency, and
  duration of actual or hypothetical exposures of
  humans to the identified agent
• Risk Characterization
• estimate the probability of specific harm to an
  exposed individual or population based on
  information from dose-response and exposure
  assessments.

9
Turkey X Disease

• 1960 1000s turkey poults died in England
• Major investigation
• Turkeys poisoned by agent in peanut meal
  component of their feed
• Agent found in peanuts contaminated with certain
  mold
• Mold, Aspergillus flavus, not responsible for
  poisoning

10
Turkey X Disease

• 1965 MIT team solved mystery of turkey X
• Aflatoxin discovered

11
Mycotoxins

• No awareness of mold-related disease before 1960s
• Imported peanut meal killed 1000s of turkeys in
  England 1960s
• The mold Aspergillus flavus produced toxins that
  fluoresced under analysis
• aflatoxin blue (AFB)
• aflatoxin green (AFG)
• Over 100 mycotoxins identified since aflatoxin

12
Molds and Mycotoxins

• Considerable worldwide significance
• Public health
• Agriculture
• Economics
• Aflatoxin cost 20M to US peanut crop 1989
• Foods that are ground present particular problems

13
What do we know about aflatoxin?
14
Aflatoxin

• Mixture of 4 closely related chemicals
• Two emit blue fluorescence B1 B2
• Two emit green fluorescence G1 G2
• Research showed them regularly
• peanuts some peanut products
• corn
• nuts
• Fed to animals can show up in derived food
  products

15
Aflatoxin

• Experimental studies showed
• potent liver poison
• malignant tumors in rats, ferrets, guinea pigs,
  mice, monkeys, sheep, ducks, trout
• Results reported 1961-1976
• Low level but not infrequent contaminant of some
  human foods

16
Some Questions About Aflatoxin

• What is to be done?
• Are aflatoxins a threat to public health?
• How many cancers can be attributed to them?
• Why is there no clear link to human cancers?
• If a menace, how can we control it?
• How much of our resources is this worth?

17
Aflatoxicosis

• Poisoning from mold-produced metabolites
• Affects all tested species and humans
• Occurs when food supplies are limited and people
  ate moldy grains
• Flabby heart, edema, abdominal pain, liver
  necrosis, palpable liver
• Chronic ingestion--liver tumors

18
FDA and Aflatoxin

• Decided limits were in order, based on what could
  be detected
• 1968 gt30 ppb in peanut products unfit
• Lowered to 20 ppb soon after
• No completely safe level can be established for
  cancer causing chemicals
• Does this mean as science gets better food
  becomes less safe?

19
FDA and Aflatoxin

• Meeting 20 ppb not too great a burden on peanut
  butter industry
• discolored peanuts could be eliminated by sorting
  machines
• required substantial new quality control measures
• Did this make scientific sense?
• If aflatoxin can be detected it is unacceptable
  if it cannot it is acceptable

20
Yes

• Potent cancer causing agent in animals
• Do not wait for human data to control it
• Animal tests are reliable indicators of human
  risk
• Risky at any level of intake
• Eliminate human exposure or reduce it to lowest
  possible level

21
No

• Animal cancers occur at levels well above FDA
  limit
• Provide some safety to humans but 20 ppb is too
  low
• Policy of no safe level is not supported by
  science
• Animals not proven reliable indicators of human
  risk
• Carcinogenic potency highly variable among
  species
• No evidence of cancer in humans

22
FDA and Aflatoxin

• Easy to detect 5ppb in some labs
• 1 ppb almost routine in some labs
• FDA did not call for these lower limits
• Large fraction of peanut butter would fail 1 ppb
  standard
• Economic impact of 1ppb could be very large

23
Detection

• Analytical chemists can now measure levels
  toxicologists are unable to evaluate for
  biological significance
• 1 ppm is like a second in 11.6 days
• 1 ppb is a second in 32 years
• 1 ppt is a second in 3,169 years

24
ppb

• Weight of contaminant divided by weight of food
• In kg of peanut butter, 20 ppb is 20 micrograms

25
Aflatoxin Occurrence 1989
26
A Few More Points

• Corn responsible for most human exposure
• Peanuts and peanut butter in US
• Drought and other damage encourage mold
• Heat not enough to destroy mycotoxin
• Processing not effective in destroying mycotoxins
• Preventing formation is crucial

27
Aflatoxin and Peanuts

• Average concentration in peanuts and peanut
  butter is 2 ppb
• FDA defect action level (DAL) to seize peanuts is
  20 ppb
• In practice anything over 15 ppb is rejected
• Average daily intake estimate is 0.005 ppb from
  peanuts

28
Science and Economics

• Just how certain is our science on matters like
  this?
• Size of economic consequence should not influence
  scientific thinking, but it influences scientists
  and policy makers when there are scientific
  uncertainties

29
Aflatoxin Management Options

• Constant testing
• more in drought years
• Seize contaminated crops
• Destroy contaminated crop residues
• Agricultural techniques
• forced air drying of crops
• controlled storage conditions
• Minimize exposure to moldy foods

30
Lets look at a CODEX/NAS risk assessment
31
Hazard Identification

• Evolving understanding
• Turkey X
• JECFA 1987
• JECFA 1997

32
JECFA 1987

• Evaluated at 31st meeting of JECFA 1987
• Considered potential human carcinogen
• Insufficient information to set tolerable intake
  level
• Urge reduction to lowest practicable level

33
JECFA 1997

• One of most potent mutagenic and carcinogenic
  substances known
• Liver cancer in most species
• Some evidence humans are at lower risk than other
  species
• Epidemiological studies show no detectable
  independent risk
• Ongoing studies--Shanghai, Thailand, Qidong

34
JECFA 1997

• Hepatitis B virus may increase liver cancer risk
• Estimated 50 to 100 of liver cancers are
  associated with Hepatitis B

35
What is the hazard?
36
Hazard Identification

• The Committee considered that the weight of
  scientific evidence, which includes
  epidemiological data, laboratory animal studies
  in vivo and in vitro metabolism studies, supports
  a conclusion that aflatoxins should be treated as
  carcinogenic food contaminants, the intake of
  which should be reduced to levels as low as
  reasonably achievable
• Source JECFA 1997

37
Hazard Characterization

• We will use a simple dose-response analysis
• This makes the two models, CODEX and NAS
  essentially equivalent

38
Aflatoxin Toxicity

• B1 (AFB1) most common, most studied, most toxic
• Toxicity varies by species
• LD50 .5 mg/kg for duckling
• LD50 60 mg/kg for mouse
• Binds to nucleic acids in some species
• Difficult to assess for humans
• Death usually from liver damage

39
Dose-Response Analysis

• Limitations of available aflatoxin data
• Confounded by concurrent Hepatitis B
• Reliability precision of aflatoxin exposure in
  study population are unknown
• Shape of dose-response relationship unknown

40
Sources of Information

• Animal bioassays
• Human feeding trials
• Epidemiological data
• Cell lines (tissue cultures)
• Animal studies most common for cancers

41
Animal Studies

• Relatively high dose to relatively few animals
• Absence of data in low dose region
• Which mathematical model best approximates
  dose-response in low dose region
• Fit data that exists
• Linear extrapolation to zero from fitted curve or
  95 confidence interval

42
Dose Response Linear Interpolation
Upper Confidence Limit
Actual Data
Excess Tumor Rate
Alternative Extrapolations
Estimated Dose Response
Linear Extrapolation
Dosage
Experimental Range
43
Low Dose Response

• Threshold/No threshold assumption is
  significant
• Many mathematical models possible
• Determines potency estimate
• Does not rely on safety factors

44
Dose-Response

• Potential biases in potency
• Only studies with association were used
• Historical levels ignored in favor of current
  levels of intake
• Hepatitis B prevalence systematically
  underestimated in early studies
• Non-primary liver cancers may have been included
• Interpolation method

45
Dose-Response

• Population risks
• Vary from population to population
• Geographically
• Culturally--diet
• Susceptibility--base health

46
Dose Response Factors

• Diet also affects toxicity
• Human response variable
• males and children more susceptible
• Hepatitis B increases cancer risk

47
Potency Values

• HbsAg
• 0.3 cancers/year per 100,000 population per ng
  aflatoxin/kg bw per day
• Uncertainty range 0.05 to 0.5

• HBsAg-
• 0.01 cancers/year per 100,000 population per ng
  aflatoxin/kg bw per day
• Uncertainty range 0.002 to 0.03

48
Exposure Assessment

• Estimating frequency and intensity of exposure to
  agent
• Magnitude, duration, schedule and route of
  exposure
• Size, nature and class of exposed population
• Detailing associated uncertainties

49
Aflatoxin Exposure Assessment

• Contamination levels data appear biased
• Studies focus on commodity lots thought
  contaminated
• Contamination levels must be used with caution
  for patterns of importance not exact
  contamination estimates

50
CDF Aflatoxin in US Maize
Contamination (µg/kg)
Cumulative density
51
Hypothetical Standards

• Assume 20 µg/kg rejection level
• 4 maize crop rejected
• mean aflatoxin level of 0.91 µg/kg
• Assume 10 µg/kg rejection level
• 6.2 maize crop rejected
• mean aflatoxin level of 0.58 µg/kg
• Standards remove most highly contaminated ,
  reducing average

52
Risk Characterization

• Combine dose-response and exposure assessments
• Describe risk in meaningful and useful fashion

53
Cancer Incidence

• Combine
• Aflatoxin potency estimates (risk per unit dose)
• Dose response
• Estimates of aflatoxin intake (dose per person)
• Exposure
• What are the uncertainties in these analyses?

54
Lets do a simple risk assessment.
55
Sample Data and Assumptions

• Assume
• Low contamination of food
• Small prevalence of hepatitis B (1 carriers)
• Potency .3 with HBSAG
• Potency .01 with HBSAG-
• European diet intake 19 ng/person per day
• Adult human weighs 60 kg
• Population of 30 million

56
Aflatoxin Risk Assessment

• 1) Calculate estimated population potency
• a) What is potency for HBSAG?
• b) What is of HBSAG?
• c) What is potency for HBSAG-?
• d) What is of HBSAG-?
• e) ab cd population potency

57
Aflatoxin Risk Assessment

• 2) Calculate intake per kg bw
• a) What is intake per person?
• b) What is weight of a person?
• c) a/b
• 3) Calculate increased cancer rate due to
  aflatoxin
• a) 1e x 2c

58
Aflatoxin Risk Assessment

• 4) Calculate increased number of cancers
• a) What is cancer rate?
• b) What is population
• c) ab
• 5) Repeat calculations for uncertain range of
  potency

59
Calculate Cancers per Year

• 0.01 x 99 0.03 x 1 0.013 cancers/year per
  100,000 population per ng aflatoxin/kg bw per day
• Range in cancer deaths is 0.002 to 0.035
• 19 ng/person per day ? 60 kg bw per person
  .317 ng/kg bw per day
• .317 ng/kg bw per day x 0.013 cancers/year per
  100,000 population per ng aflatoxin/kg bw per day
  0.0041 cancers/year per 100,000 people

60
Calculate Cancers per Year

• 30,000,000 people x 0.0041 cancers/year per
  100,000 people 1.23 cancers/year

61
Risk Assessment Model

• Which steps were hazard identification?
• Which steps were dose-response assessment?
• Which steps were exposure assessment?
• Which steps were risk characterization?

62
Model Comparison

• Food safety vs... CODEX/NAS Model
• Did food safety have four steps?
• What were major differences?
• What were similarities?
• Which was easier? Why?
• Which do you prefer? Why?

63
Lets see how we can address the uncertainty in a
model like this.
64
(No Transcript)
65
(No Transcript)
66
Whats Next?

• Once the risk has been assessed
• Risk management decides what to do about it
• Risk Communication
• The risk is described to others
• Management options are explained

67
The End