BiologyBased Modelling Tjalling Jager Bas Kooijman Dept Theoretical
Biology-Based Modelling Tjalling Jager Bas Kooijman Dept. Theoretical Biology
Contents Ø Ø Current procedures in (eco)tox Biology-based methods A glimpse into the future? Workshop announcement
Risk assessment EXPOSURE EFFECTS “RISK” Available data Assessment factor Three LC 50 s 1000 One NOEC 100 Two NOECs 50 Three NOECs 10
Exposure assessment Lab. experiments Process parameters at env. conditions Integrated model for system PEC Extrapolation to e. g. : • different compartment sizes • different temperature • time-varying emissions
Effects assessment Lab. experiments PNEC
Analysing test data 1. Statistical testing Contr. Response NOEC * log concentration
Analysing test data Response 1. Statistical testing 2. Regression EC 50 log concentration
Limitations Practical problems as summary statistic – NOEC has received serious criticism – ECx and NOEC change with exposure time
EC 10 changes in time carbendazim pentachlorobenzene 2. 5 140 120 2 100 survival 1. 5 80 body length 60 1 40 0. 5 0 0 5 10 time (days) 15 cumul. reproduction 20 0 0 2 4 6 8 10 12 time (days) 14 16 Alda Alvarez et al. , 2006 (ES&T)
Limitations Practical problems as summary statistic – NOEC has received serious criticism – ECx and NOEC change with exposure time Not all data are used – only data at last time point – test protocols prescribe more time points – in some cases, more endpoints available
Survival dichlorobutene 20 survival 15 10 4 days 1 -3 days 5 0 0. 01 0. 1 concentration (m. M/L) 1 Data from Geiger et al. , 1985
Reproduction phenol 120 total eggs/female 100 21 days 80 60 9 -20 days Further available - survival data - possibly body weight 40 20 0 0. 01 0. 1 1 10 concentration (mg/L) Data from OECD ring test
Limitations Practical problems as summary statistic – NOEC has received serious criticism – ECx and NOEC change with exposure time Not all data are used – only data at last time point – test protocols prescribe more time points – in some cases, more endpoints available No biological mechanism – selected curve for ECx is descriptive – limits analysis of non-standard data (e. g. degradation) – limits extrapolation to other scenarios …
Effects assessment Lab. experiments Regression process orparameters? statistical test Extrapolation to e. g. : • different body sizes? • different temperature? • time-varying exposure? • populations? integrated Assessment factor model? Available data PNEC Assessment factor Three LC 50 s 1000 One NOEC 100 Two NOECs 50 Three NOECs 10
Biology-based modelling
Biology based methods OECD recommendations (1996) – NOEC is inappropriate and should be phased out – incorporate exposure time in data analysis – favour mechanistic models if they fit the data ISO and OECD guidances (2006) – statistical analysis of ecotoxicity data – biology-based methods included next to standard methods
Biology based methods exposure concentration calculated by fate models toxicokinetics internal concentration e. g. one-comp. model or PBPK toxicodynamics e. g. DEB model effects in time
Dynamic Energy Budgets (DEB) assimilation reproduction DEB provides simple rules for resource acquisition and growth allocation maintenance
DEB parameter Effect of toxicants in DEB NEC blank value internal concentration Potential targets • probability to die • maintenance costs • assimilation • costs for growth • costs for reproduction • death of embryos
DEBtox Ø Currently only method with general applicability – combines toxicokinetics and -dynamics – analysis of all regular ecotox endpoints – data from standard OECD protocols Ø Large body of literature over last 10+ years Ø Windows software – version 1. 0 in 1996, version 2. 0. 1 in 2004 – freely downloadable
Windows version Only for standard tests – – acute survival Daphnia reproduction fish growth algal population growth
Example: Hexachlorobutadiene Fitting 4 parameters for entire data set No-effect concentration: NEC = 0. 13 (0. 091 -0. 16) Data from Geiger et al. , 1985
Example: Hexachlorobutadiene external concentration (mol/L) iso-effect lines LC 0 LC 10 LC 25 LC 50 NEC 1 0. 8 0. 6 0. 4 0. 2 0 0 20 40 60 time (hours) 80 100
Example: Phenol Fitting 4 parameters for entire data set Mo. A: ‘costs for growth’ NEC = 0. 82 (0. 52 -0. 91) Data from OECD ring test
Example: Phenol iso-effect lines external concentration (mg/L) 1. 8 EC 0 EC 10 EC 25 EC 50 NEC 1. 6 1. 4 1. 2 1 0. 8 0. 6 0. 4 0. 2 0 0 5 10 time (days) 15 20
Advantages Using standard test data … Ø Make efficient use of all data points in time – more accurate parameter estimates – reduce number of test animals Ø Extract more information from same data – kinetic information on toxicity (e. g. , ECx at any time point) – physiological mode of action (for population response) Ø Yield a more robust summary statistic – time-independent NEC may replace NOEC and ECx
Advantages Furthermore … Ø Fit non-standard data – loss of compound from test system – limited tests (single dose, no control, one indiv. per dose) – no need to discard existing test data Ø Simultaneous fit on multiple data sets – more endpoints (life-cycle toxicity) – different compounds: mixture toxicity Ø Educated extrapolations – intermittent exposure, food limitation, population etc. – between compounds (QSARs), between species …
A glimpse into the future?
Characterisation of impacts? EXPOSURE EFFECTS PEC/PNEC EXPOSURE EFFECTS “IMPACTS”
Example Ø Daphnia magna exposed to Cd Ø Test based on OECD repro test – 15 days – survival and offspring daily – body weight at end of test Data from Heugens et al. , 2006 (ET&C) Analysis in Jager et al. , 2006 (Ecotox)
Simultaneous fits body size Mode of action Cd decreases assimilation 120 1 100 0. 8 80 0. 6 60 0. 4 40 survival reproduction 0. 2 0 20 0 2 4 6 8 10 12 14 16 0 0 5 10 15
Extrapolations population growth rate (1/day) NEC 0. 4 impact 0. 3 • ecologically relevant • time independent • integrates endpoints 0. 2 0. 1 PEC 0 0 0. 05 0. 15 concentration 0. 2
Extrapolations population growth rate (1/day) NEC 0. 4 0. 3 impact • ecologically relevant • time independent • integrates endpoints • compare chemicals 0. 2 0. 1 PEC 0 0 0. 05 0. 15 concentration 0. 2
population growth rate (1/day) Extrapolations 0. 4 0. 3 90% food 0. 2 80% food 0. 1 0 0 0. 05 0. 15 concentration 0. 2
emission exposure Matching release scenario time survival time
Announcing a workshop Ø There is potential for improvement of RA with biologybased methods Ø Biology-based methods are available and described by ISO and OECD Workshop organised by ECB – introduce these methods to regulatory community – discuss potential for implementation – planned for 2 days in week of 4 -8 June 2007 http: //www. bio. vu. nl/thb/deb
- Slides: 36