Precaution in Risk Characterisation of

GM Plants Regarding Food/Feed

Safety and Environmental Impact

Dr. Harry A. Kuiper

RIKILT - Institute of Food Safety,
Wageningen University & Research Centre,
Wageningen
The Netherlands

The Role of Precaution in GMO Policy

Vienna, 18-19 April 2006
 Issues

STRUCTURE RISK ASSESSMENT PROCESS

RISK CHARACTERISATION

UNCERTAINTY ANALYSIS

CONSERVATIVE APPROACH TAKEN IN RISK

ASSESSMENT

2
The Risk Cycle: Components of Risk Analysis,

EU Scientific Steering Committee Report, 2000,

3 2003
 RISK CHARACTERISATION
(EU Scientific Steering Committee)

The quantitative or semi-quantitative estimate

including attendant uncertainties, of the
probability of occurrence and severity of adverse
effects in a given population under defined
conditions based on hazard identification, hazard
characterisation and exposure assessment
(SSC, 2000)

4
 Issues to Look at During Risk
Characterisation
What allowance should be made for the absence of specific
data or for poor quality data?
What allowances should be made for possible unidentified
inter and intra species variations?
What uncertainties are involved in other extrapolations
including route of exposure time (e.g. short-term to long-
term), location (one site to different sites), space (e.g. local to
regional and national and vice versa)?
Can particular at risk groups be identified and how should
they be included in the final risk assessment opinion?
What weighting should be given to simultaneous or co-
exposure to other related agents/media and if so on what
scientific basis?
5
 Uncertainty Analysis is a Key Element
in Risk Assessment
Variability in measurements in data points (systematic
and random errors)
Methodological uncertainties (models, presumptions,
extrapolations valid?)
Significance of test parameters
Assumptions for extrapolation (sensitivity humans-
animals, sensitive groups)
Biological significance of observed changes
Fundamental ignorance: how does the system work?
Numerical/language description

6
 7
The EFSA Journal (2004) 99, 1-93
 International Food Safety Strategies for
Foods Derived from Modern Biotechnology

OECD Group of National Experts on Safety in Biotechnology, 1993,

1994, 1996
OECD Task Force on the Safety of Novel Foods and Feed, 1998-
present
FAO/WHO Expert Consultations, 1991, 1996, 2000, 2001, 2003
CODEX Task Force on Foods Derived from Biotechnology, 1999-
2004
European Commission Directives and Regulations, 1996-present
ENTRANSFOOD, the EU Thematic Network on the Safety
Assessment of Genetically Modified Food Crops, 2000-2003
European Food Safety Authority, Guidance Document GMO Panel

8
 Sequential Steps in the Risk
Assessment of GMOs

HAZARD IDENTIFICATION
characteristics which may cause adverse effects
HAZARD CHARACTERIZATION
potential consequences for man and the environment
EXPOSURE ASSESSMENT
likelihood of occurrence/exposure
TOTAL RISK CHARACTERIZATION
evaluation of risk(s) posed by each identified characteristic

9
 Comparative Safety Assessment
Approach for GMOs

Underlying assumption:
Traditionally cultivated crops have gained a history
of safe use for the environment/consumer/animals

These crops can therefore serve as a baseline for

the environmental and food/feed safety
assessment of GM crops

10
 Comparative Safety Assessment
Approach for GMOs

1. Identification of differences between the GM and non-

GM crop
2. Assessment of the identified differences regarding
environmental/food/feed safety/nutritional impact
Concept of Familiarity
Concept of Substantial Equivalence or Comparative Safety
Assessment
3. No absolute safety assessment in itself

11
 Key Elements for the Assessment of
GMOs
Characterization of donor and host organism
Molecular characterization of the genetic modification
event:
methods
inserted genes
gene expression
Analysis of agronomical and compositional properties
Toxicity/allergenicity/ nutritional testing
Environmental risk assessment
Environmental monitoring/surveillance

12
 Risk-Assessment Strategies for Transgenic Foods

Traceing of differences between the GM food and the

conventional product
Phase 1

Introduced (New) proteins (New) metabolites

genes and toxins

Toxicity / nutritional investigations

Phase 2

Gene Degradation Toxicity

transfer Toxicity
Allergenicity Whole foods
13
 Risk-Analysis Strategies for Transgenic Foods II

Intake studies
Phase 3

Role of the product in the diet

Intake of the product by the consumer
Phase 4

Risk assessment of transgenic foods

14
 EFSA Guidance Document Environmental
Risk Assessment (E.R.A.) Elements

Persistence and invasiveness

Selective advantage or disadvantage
Potential for gene transfer
Interactions between the GM plant and target organisms
Interactions of the GM plant with non-target organisms
Effects on human/animal health
Effects on biogeochemical processes
Impacts of the specific cultivation, management and harvesting
techniques
Potential interactions with the abiotic environment
Monitoring and Surveillance

15
 Tiered Approach to Risk Assessment

Tier 1. Laboratory Experiments > Hazard, Impacts at first

trophic level, direct non targets.

Tier 2. Growth Room/Glasshouse > Interactions, 2nd trophic

level, indirect non targets.

Tier 3. Field Experiments > Exposure at range of trophic levels,

indirect/agronomic effects

Tier 4. Monitoring > long term/large scale impacts

16
 Comparative Comprehensive Risk
Assessment

All evidence from the molecular, agronomical,

compositional, toxicological/nutritional and
environmental impact characterization should be
taken into account for final conclusions on risk
assessment

Science evolves continuously and therefore there is

permanent need for further method development

17
Uncertainty Analysis in GM Food/Feed
Safety Assessment

Completeness and quality of the molecular, compositional

and toxicological/nutritional dataset
Adequacy of performance of experiments, data collection
and processing and statistical analysis
Possibilities of unintentional insertion and expression of
gene sequences
Possible occurrence of open reading frames (ORFs)
Possible occurrence of unintended alterations in the
composition (new proteins/metabolites)
Safety/nutritional impact of newly formed products

18
Uncertainty Analysis in GM Food/Feed
Safety Assessment

Uncertainties in assumptions in toxicological, allergenicity or

nutritional analysis
Strength data provided on the allergenic potential of newly
expressed proteins/GM crop
Extrapolation of data from animals to humans
Delayed onset of effects (short term versus chronic exposure)
Need for specific toxicological data (reproductive and
developmental toxicity)
Risks associated with variations in susceptibility in populations
Evaluation of intake estimations
Necessity for post-market monitoring?

19
 Animal Feeding Trials with Whole GM
Foods/Feed

Foods are extremely complex matrices with many

biologically active compounds, that may cause adverse
reactions
Very few conventional foods have been subject to
toxicological studies, yet they are accepted as being
safe
Very little known about long term effects of any food
Wide genetic variability among consumers
Changing diets over time.

20
 When Animal Feeding Trials With GM
Foods/Feed?

Profound changes in the composition of the GM plant

Indications for potential unexpected effects (molecular
characterization, agronomic, compositional analysis)
90-days study in rodents recommended
Reassurance study
Protocols (OECD) for low molecular weight chemicals
testing should be adapted for testing whole foods.

EFSA Guidance Document

Codex Alimentarius Guidelines
21
 Difficulties with Animal Feeding Trials
with GM Foods/Feed

Natural bulkiness of food,

Effects on satiety
Need to maintain nutritional balance
Limit of dietary administration (5%) in order to
prevent dietary imbalance
Matrix effects
Semi-synthetic diets can be prepared with inclusion
levels as high as 60% or more.

22
 Predictivity of Short-term Rodent
Studies for Long term Effects?

US NTP: Studies of industrial and agro chemicals indicated that

for 70% of the compounds toxicological findings in the 2 year
rodent test were also seen in or predicted by the 3 month
subchronic test (British Toxicology Society, 1994)
Review of data sources including monographs of JECFA,
covering 613 substances, indicated that in many cases, the
lowest and most conservative NOEL for a substance came from
a subchronic study (Munro et al.,1996)
Similar observations in dog studies (Box and Spielman, 2005)

23
 Uncertainty Analysis in the Environmental
Risk Assessment (E.R.A.) of GMOs

Soundness of predictions on stability of introduced

and expressed traits under representative
environmental conditions
Extrapolation of data from environmental laboratory
studies to complex ecosystems
Predictability of potential adverse environmental
effects in the long term

24
 Assumptions and Uncertainty Analysis in
the E.R.A. of GMOs

ERA starts with data on maximum or high dose tests in tier 1 in

order to catch potentially all possible effects.
Identified effects in tier 2 will trigger more detailed studies in higher
tier 3 and 4 testing.
Experience from chemical pesticide testing has shown, that most
harmful effects in a range of 50% above or below normal in a
simplified system will not lead to significant biological effects in a
more complicated environment.
The (GM-) Organism specific uncertainty concerning self-replication
and gene flow to wild relatives is addressed using
prediction tools developed (e.g. introduced biocontrol organism)
simulation and modelling of worst-cases (e.g. field experiments
with out crossed GM wild plants)

25
 Assumptions and Uncertainty Analysis in
the E.R.A. of GMOs

Most crop plants are annuals, some biennials and few

perennials. They are not generally climactic species in natural
ecosystems.
Most crops have been developed to grow in managed
environments ( i.e. with fertilizer supply and little competition to
weeds): the likelihood that one or two transgenes will have
dramatic effects is very low.
Crossable wild plants are highly evolved and adapted to certain
niches. They are variable with a constant flux in populations and
genotypes (e.g. in Genus Beta and Brassica).
Introduction of a novel gene is unlikely to have a similar effect
on all populations in all niches.

26
 Application of Worst Case Scenarios in the
E.R.A. of GMOs

Application of worst case scenarios, e.g. cultivating Bt maize

(MON810) three years continuously at the same field addressing Bt
protein accumulation and long-term effects on non-target organisms:
no effect detectable e.g. for the most Cry1Ab sensitive
lepidoperan species known (Plutella xylostella) feeding on weed
leaves inside the Bt maize plots, (to be published)
experimental system sensitive to demonstrate chemical
insecticide effects
adequate extrapolation to the maize agro-ecosystems in the EU
Environmental biosafety research is an evolving discipline
(databases now host >1900 records)

27
 Monitoring is an Essential Element of the
E.R.A. of Transgenic Plants

6-year-field evaluation of Bt maize in Spain (Event 176,Cry1Ab)

(Matilde Eizaguirre et al., Transgenic Research (2006) 15:112
Scope of the study: the potential of the target corn borers Sesamia
nonagrioides and Ostrinia nubilalis to evolve resistance to Bt and
effects on non-target organisms.
Resistance to Bt in target corn borers in Spain has not developed as
of 2005, after 7 years of growing Bt maize
Separate fields of non-Bt maize as refuge fields may be more
advisable than mixing Bt and non-Bt seed (dispersal capacity of the
corn borer larvae)
Transgenic maize did not have a negative impact on non-target
pests: more aphids and leafhoppers but similar numbers of
cutworms and wireworms in Bt versus non-Bt fields;
No difference in the numbers of the most relevant predators in fields
containing transgenic or no transgenic maize

28
 Self Task Activities of the GMO Panel

Biosafety of antibiotic resistance marker genes

Post-market environmental monitoring of GM crops
The use of animal feeding trials for the safety evaluation of whole GM
foods/feed
Update the approaches for allergenicity assessment of GMOs
Strategies for statistical analysis in comparative analysis and animal
studies
Assessment of GM plants used as production platform for non-
food/feed products
Assessment of GM plants with enhanced nutritional properties (not
started)

29
When Should the Precautionary Principle be
Invoked?

Where scientific information is insufficient,

inconclusive, or uncertain, or where there are
indications that the possible effects on the
environment, or human or plant health may be
potentially dangerous and inconsistent with the
chosen level of protection, the precautionary
approach may be invoked (Commission, 2000)

30
 Risk Characterisation of GM Plants
(Chapter IV EFSA Guidance Document)

Application of the precautionary approach is distinct

from the normal conservative approach taken during
risk assessment when safety or extrapolation factors
are applied

Application of the precautionary approach is the

responsibility of the risk manager and not of the risk
assessor (COM 2000)

31
When Should the Precautionary Principle be
Invoked?

Identify potential risks in an early phase

Describe risks, as quantitative as possible:
Type and severity,
Reversibility/irreversibility
Acute/long- term effects on human/animal health and the
environment
Who will be affected
Options for risk handling and control
Present comprehensive risk assessments, taking all
available information into account

32
 Precaution in Risk Analysis:
Task for Risk Assessors, Risk Managers and
other Stakeholders

The precautionary principle is an essential approach

(attitude) throughout the whole Risk Analysis
Process, and should highlight the unknowns
(uncertainty analysis)
Scientific experts should always apply caution in
their assessment of scientific data (quality and
uncertainty)
Risk managers (and other stakeholders) should
consider what is an acceptable level of risk for the
society

33
 Role of the Scientific Expert
In the Risk Analysis Process

Interact with Risk

Managers
Risk Assessor Possible
consequences Interact with
Characterise
Options Social
Risk problems
Choices Economic
Complexity
Experts
Uncertainty Interact with Stakeholders
Producers
Consumers

34
Promoting Food Safety through a New Integrated
Risk Analysis Approach for Foods

K ASSESSO
- RIS RS
ERS -
D RE
EE G
SYSTEMIC RIS

BR
NT

UL
E KS

NT
M EV

AT
S

LA
ES

O
AL
S
- P

RY
AS

UA
SK

TIO

INST
ODUCERS

PECIFIC RI

N RI
INTEGRATED

ITUTIONS -
RISK ANALYSIS

SK MANA
APPROACH FOR
FOODS
S S
PR

G
E
OD

E
C I

ME
LI

E
FO

FS
T
P

OP
NT

A
E TI O
-

-
N R S
ISK ASSES
S

C
OD
ER EX
UM
AL NS
IM E O
NTARIUS - C

SAFE FOODS (www.safefoods.nl)

35
 Conclusions

The risk assessment of GM plants regarding food/feed

safety and their environmental impact follows an
integrated approach consisting of a case-by-case
detailed analysis of key issues representing potential
risks for humans, animals and/or the environment

Risk characterisation, i.e. a (semi)-quantitative estimate

of associated uncertainties, and of the probability of
occurrence and severity of adverse effects in a given
population and/or the environment is an essential part
of the risk assessment procedure of GM plants and
derived food/feed products
36