Processbased toxicity analysis in risk assessment Tjalling Jager

Process-based toxicity analysis in risk assessment Tjalling Jager Bas Kooijman Dept. Theoretical Biology

Contents Ø Ø Ø Dynamic Energy Budget (DEB) theory Current procedures in (eco)tox Introduction to DEBtox Advanced examples The DEB laboratory

Why DEB theory? How do individuals acquire and allocate their resources?

Relation DEB and toxicants ?

Relation DEB and toxicants ?

Relation DEB and toxicants ?

Dynamic Energy Budgets food faeces assimilation reserves somatic maint. maturity maint. structure 1 - maturity offspring

DEB pillars Ø Quantitative theory; “first principles” – time, energy and mass balance Ø Life-cycle of the individual – links levels of organisation: molecule ecosystems Ø Comparison of species – body-size scaling relationships; e. g. metabolic rate Ø Fundamental to biology; many practical applications – (bio)production, (eco)toxicity, climate change …

Chemical-related projects at TB Ø Dutch government (RWS and RIVM) – biaccumulation metals in mussels; biomonitoring – toxicokinetics dioxin in humans Ø Dutch Technology Foundation STW – DEBdeg (bio)degradation of (toxic) compounds – DEBtum tumour induction/growth, analysis tox data – DEBtox ind pop (reprod. modes in nematodes) Ø EU Projects – Model. Key – No. Miracle effects on ecosystems and food chains mixture toxicity More info: http: //www. bio. vu. nl/thb/research/project/

Current procedures in (eco)tox

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

Standard approaches 1. Statistical testing Contr. Response NOEC * LOEC log concentration

What’s wrong with NOEC? Ø No statistically significant effect is not no effect Ø Effect at NOEC regularly 10 -34%, up to >50% Ø Inefficient use of data – only last time point, only lowest doses – for non-parametric tests also values discarded OECD Braunschweig meeting 1996: NOEC is inappropriate and should be phased out!

Standard approaches Response 1. Statistical testing 2. Curve fitting EC 50 log concentration

What’s wrong with ECx? Regression model is purely empirical Ø No estimation of process parameters – not possible to extrapolate to env. conditions Ø Inefficient use of data (last time point only) Ø ECx depends on exposure time

Effects change in time Nonylphenol, survival 1 0. 9 fraction surviving 0. 8 0. 7 0. 6 0. 5 24 hours 0. 4 0. 3 48 hours 0. 2 0. 1 0 0 0. 1 0. 2 0. 3 0. 4 concentration 0. 5 0. 6 0. 7

Why does LC 50 decrease? Toxicokinetics internal concentration – effects are related to internal concentrations – kinetics depend on chemical B chemical A chemical C time

Why does LC 50 decrease? Toxicokinetics internal concentration – effects are related to internal concentrations – kinetics depend on chemical – and species … Daphnia chemical B chemical small fish. A large fish C chemical time

Sub-lethal EC 10 in time does not necessarily decrease 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

Consequences Procedures are inefficient Ø Test protocols yield more data than are used NOEC and LCx/ECx are not representative Ø Change in time, depending on species, body size, chemical and endpoint Standard exposure time leads to systematic error Ø in comparing effects – between chemicals – between species (comparative RA, QSARs …? ) (SSDs … ? ) OECD Braunschweig meeting 1996: Exposure time should be incorporated in data analysis

Introduction to DEBtox

DEBtox OECD Braunschweig meeting 1996: Exposure time should be incorporated in data analysis Mechanistic models should be favoured if they fit the data – Windows software, version 1. 0 in 1996, version 2. 0. 1 in 2004 – Included in draft ISO/OECD guidance on statistical analysis of ecotox data

Why process-based? Understand toxic effects – biology of organism and toxic mechanisms Match experimental set-up – e. g. degradation, pulse exposure Predictions for exposure situation – e. g. populations, food level, varying exposure

DEBtox basics Ø Effect depends on internal concentration – one-compartment model toxicokinetics

DEBtox basics Ø Chemical affects a parameter in DEB – e. g. maintenance rate toxicokinetics target parameter

DEBtox basics Ø Change in target parameter affects endpoint – survival, reproduction, growth toxicokinetics target parameter DEB model

Modes of Action assimilation food faeces assimilation somatic maint. reserves 1 - structure tumour growth costs reproduction costs hazard to embryo maturity maint. hazard (lethal effects) maintenance costs maturity offspring tumour induction endocrine disruption

Windows version Ø User-friendly software, freely downloadable Ø Only for standard tests – – acute survival Daphnia reproduction fish growth algal population growth

Example: survival dieldrin time (d) concentration (µg/L) 0 1 2 0. 0 20 20 20 3. 2 20 20 20 5. 6 20 20 19 10 20 20 17 18 20 18 15 32 20 18 9 56 20 17 6 100 20 5 0 3 4 5 6 7 20 20 19 18 19 19 18 18 18 15 14 12 9 8 9 4 4 3 2 2 1 0 0 0

Example: survival dieldrin

Example: survival dieldrin 0 d 1 d 2 d 3 d 4 d 5 d 6 d 7 d NEC Killing rate Elim. rate Blank haz. 5. 2 (2. 7 -6. 9) µg/L 0. 038 L/(µg d) 0. 79 d-1 0. 0084 d-1

Example: survival nonylphenol concentration (mg/L) time 0 h 24 h 48 h 0. 004 20 20 20 0. 032 20 20 20 0. 056 20 20 20 0. 100 20 20 20 0. 180 20 20 16 0. 320 20 13 2 0. 560 20 2 0

Example: survival nonylphenol 0 hrs NEC 0. 14 (0. 094 -0. 17) mg/L Killing rate 0. 66 L/(mg h) Elim. rate 0. 057 h-1 24 hrs 48 hrs

Example: survival nonylphenol NEC LC 50 LC 0

Example: repro cadmium Mode of action NEC Tolerance Max. repro Elim. rate costs for repro 3. 3 e-9 (0 -0. 017) m. M 4. 7 e-9 m. M 14 offspring/d 2. 6 e-9 d-1

Example: repro cadmium EC 0 EC 50

Advantages DEBtox For the standard software Ø Make efficient use of all data points – more accurate parameter estimates – reduce number of test animals … Ø More information obtained – ECx at any time point can be calculated – mode of action; crucial for population response Ø Characterisation of effects – time-independent NEC may replace NOEC and ECx

Advanced examples

DEBtox extensions Simultaneous fits on more data sets – endpoints, chemicals, species … Fit deviating experimental data – degradation, pulse exposure … Extrapolations – time, food level, temperature, (species) … At this moment only available as Mat. Lab scripts

Simultaneous fits Survival and body residues for cadmium (Heugens et al. ) NEC on internal basis: 259 mg/kg dwt (202 -321)

Extrapolation fraction surviving From continuous exposure to a 20 -hour pulse 1 0 mg/L 0. 8 3 mg/L 0. 6 0. 4 4 mg/L 0. 2 0 5 mg/L 10 mg/L 0 20 40 60 time (hours) 80 100

simultaneous fits Survival for 5 OP esters (data De Bruijn & Hermens) Same NEC, elim. rate, killing rate, receptor repair rate Different affinity for receptor

simultaneous fits Reproduction test with cadmium (data Heugens et al. ) body size Mode of action decrease 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) To populations and limiting food 0. 4 0. 3 90% food 0. 2 80% food 0. 1 0 0 0. 05 0. 15 concentration 0. 2

Simultaneous fits Fenvalerate pulse at two food levels (data Pieters et al. ) High food Body length Cumulative offspring Mode of action: assimilation NEC survival: 0. 42 µg/L NEC growth/repro: 0. 051 µg/L Low food Insights • intrinsic sensitivity independent of food • chemical effects fully reversible Fraction surviving

Opportunities 1: Relevant endpoint • ecologically relevant • time independent • integrate endpoints • comparable between chemicals population growth rate NEC impact PEC concentration

Opportunities 1: Relevant endpoint • ecologically relevant • time independent • integrate endpoints • comparable between chemicals population growth rate NEC impact PEC concentration

exposure Opportunities 2: Match exposure scenario survival time

Opportunities 3: Reduce testing needs? Ø Use all of the data points – more data points per parameter – less animals needed Ø Less need to discard ‘poor’ data – disappearance of test compound – change in body weight of test organism – combine low-quality data sets Ø Less need for new tests – better extrapolations from lab data – opportunities for QSAR development …

Relations for alkyl benzenes Fathead minnows, NEC Daphnia pulex, elimination 1 mm juveniles 3 mm adults

The DEB laboratory

Electronic DEB laboratory Freely downloadable from http: //www. bio. vu. nl/thb/deblab/ DEBtox – Windows version 2. 0. 1. – routine applications DEBtool – open source (Octave, Mat. Lab) – full range of DEB research (fundamental+applied) – also advanced DEBtox applications

Finally … In our opinion … – exposure assessment is well ahead of effects assessment – effects assessment will benefit from a processbased approach • more scientific extrapolation • testing needs may be reduced – but … requires major shift in thinking • basic methods are already available • toxicity data are already reported in time

More information These slides are available at: http: //www. bio. vu. nl/thb/users/bas/lectures/ Further reading (paper submitted): http: //www. bio. vu. nl/thb/research/bib/ Jage. Heug 2005. html Further literature: http: //www. bio. vu. nl/thb/research/bib