Upload
claribel-palmer
View
217
Download
0
Tags:
Embed Size (px)
Citation preview
WP2: Cumulative dietary exposure and hazard assessment
Bernadette Ossendorp en Polly Boon
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 2
Contents
• What is probabilistic modeling
• Build upon previous work done for and by EFSA
• What need to be done in WP2- Exposure
- Effect modeling
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 3
consumptiondatabase
residuedatabase
0.0
0.2
0.4
0.0 0.7 1.4 2.2 2.9 3.6-0.7-1.4-2.2-2.9-3.6
99, 99.9, and/or 99.99 percentile
RPF index
compound
Probabilistic modeling cumulative exposure
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 4
• Performed by RIKILT (RIVM), NFA, NIPH, CRD, AFSSA, INRAN, EVIRA
• Is probabilistic assessment possible and can it be used at the international level addressing both acute and chronic toxicity.
- actual exposure using monitoring results
- exposure using MRLs or STMRs
- acute toxicity corresponds with short-term exposure
- chronic toxicity corresponds with long-term exposure
EFSA contract probabilistic assessment
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 5
NLIT
SE
0.0
0.2
0.4
0.0 0.7 1.4 2.2 2.9 3.6-0.7-1.4-2.2-2.9-3.6
UK
MCRA-software
E-platform of databases (food and residue)
CZ
FR
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 6
Short-term acute intake in different countries
CZ FR IT NL SEUK
90
95
9999.9
0
1
2
3
4
5
6
7
8
9
10
Exp
osur
e (u
g/kg
bw
/d)
Country
CZ FR IT NL SEUK
90
95
9999.9
0
1
2
3
4
5
6
7
8
9
10
Expo
sure
(ug/
kg b
w/d
)
Country
Residue per country All residues pooled
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 7
Models and bimodal distributions
• No real cumulative model available
• Two alternative approaches were used
• Serious limitations
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 8
Other limitation in models for MRL setting
• BBN do not handle bimodal distributions properly
• Observed Individual Mean model as alternative
• Further model development will be done in WP5
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 9
Alternative model Observed Individual Mean
• OIM is conservative which is allowed in Risk Assessment
• Real cumulative models are not used
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 10
Main conclusions in EFSA opinion, RIKILT report
• Common Assessment Groups should be clear and the same over Europe
• Lot of work to solve compatibility issues and to organize data
• More understanding and model development regarding cumulative probabilistic modelling (EFSA works on opinion on probabilistic modeling)
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 11
Is MoE or level of protection adequate?
Is Margin of Exposure (or safety) acceptable
Hazard assessment(tiers)
Exposure assessment (tiers)
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 12
Tasks WP2
• Description of the work to be performed in WP2, including
- Deliverables
- Milestones
• Tasks are to be performed by all WP2 partners including AFSSA, unless stated otherwise
• Case studies
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 13
Task 2.1 Input data organisation• Food consumption and pesticide monitoring data for use in case studies:
- Organophosphates, conazoles, synthetic pyrethroids
• RPFs (NFA)
• Processing, variability factors and unit weights (CRD)
• All data will be linked in such to allow for cumulative assessments by M12 (ms. 2.2)
• Paper on availability and quality input data by M12 (del. 2.1)
NL ITSE
0.0
0.2
0.4
0.0 0.7 1.4 2.2 2.9 3.6-0.7-1.4-2.2-2.9-3.6
UK
MCRA-software
CZ
FR
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 14
Task 2.2 Harmonising input data
• Link between foods consumed and products (RACs) analysed- Guidance document by M9 (del. 2.2) + training (RIVM)
- Conversion own food consumption data using national recipes by M15
• Pooling of food consumption data: vulnerable groups- Development of a methodology to merge different food surveys into a common database
(RIVM)
- Harmonisation food coding and compound coding
- M24
- Paper by M36, including uncertainties (del 2.5)
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 15
Task 2.3 Benchmark dose modeling, RPFs
• Summarise toxicological studies relevant for cumulative effects (NFA/UMIL +
EPA member advisory board) by M12 (ms 2.1)
• BMD modelling by M18 (NFA): needed for Task 2.4.
• What has been done:
- SAFE FOODS
- EFSA opinion
-3.0 -2.5 -2.0 -1.5 -1.0
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
log10-dose.mgkgbw
Tm
PA
H.C
i.m
OTA
E4: y = a*[c-(c-1)exp(-bx)]Proast23.8 v ar- 0.1513 a- 0.9628 b- 25.63 c 0.05854 loglik -24.68 conv : 1 sf .x : 1 dty pe : 10 selected : dur.mth 3 106 CES 0.05 BMD 0.002129
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 16
Task 2.4 Cumulative exposure and risk modelling (1)• Guidance document on (RIVM)
- Usage of the models developed in WP5
- Performance of cumulative exposure calculations 4 scenario’s• Acute and chronic exposure, actual (monitoring results)
• Acute and chronic exposure, MRL setting
• Use of model (exposure) by M18- Case studies
- Chlorpyrifos: results are input for quantitative link with biomarker measurements WP4 (ms 2.3)
- Vulnerable groups
- Scientific paper by M18 (del 2.3)
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 17
Task 2.4 Cumulative exposure and risk modelling (2)• After integration exposure with BMD modelling
- Estimating cumulative margin of exposure (MoE) using integrated model
- Uncertainty analyses
- User and reference guideline integrated model + uncertainty analysis by M30 (ms 5.5)
• Scientific paper by M36 (del. 2.4)
MoE
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 18
Task 2.5. Identification uncertainties
• Identification uncertainties expected to affect cumulative exposure and effect + qualitative importance (ms 5.1)
- M1-6
- Based on an approach developed in WP5
• Providing input (data + expert knowledge) for WP5 to develop models to assess most important uncertainties quantitatively
• Reviewing approaches developed in WP5
• Inclusion of uncertainties in cumulative assessments (Task 2.4)
Kick-off meeting ACROPOLIS, 7-8 June 2010, Utrecht, The Netherlands 19
Work to be done in coming months
• Discuss quality of data to be used
• Discuss data organisation based on previous work
• Both for exposure as for BMD modelling
Thanks
Bernadette Ossendorp en Polly Boon