Low dose effect

Alan R Boobis

Imperial College London (a.boobis@imperial.ac.uk)

ILSI Europe 2014

Annual Symposium

20-21 March 2014

Brussels, Belgium

"All substances are poisons; there

is none which is not a poison. The

right DOSE differentiates a poison

from a remedy.“

Areolus Phillipus Theophrastus

Bombastus von Hohenheim

Paracelsus (1493-1541)

Risk assessment/management of genotoxic carcinogens

Management based on assumption of no threshold in

dose-response relationship

ALARA/P (as low as reasonable

achievable/practicable)

Low dose extrapolation, usually linear

Acceptable risk level of 1 in 106 (or 1 in 105)

Margin of exposure (=point of departure/estimated

human exposure)

MOE ≥ 10,000 considered to be of low concern (JECFA,

EFSA)

NB: MOE of 10,000 is equivalent to risk level of 1 in 105

(based on BMDL10 as POD and linear extrapolation)

Cancer ‘mega-studies’ and other relevant studies

‘Mega-rat’ study of nitrosamines (n~4,000) Peto et al (1991a, b)

ED01 study of 2-AAF in mice (n>24,000) Gaylor (1980, 1985)

Study in trout of dibenzo[a,l]pyrene (n>40,000) Bailey et al (2009)

Studies of pre-neoplastic effects of 2-AAF and DEN in rat Williams et al. (1993, 1998, 1999, 2004)

In vivo and in vitro studies of MMS, MNU, EMS and ENU Doak et al. (2007) and Gocke & Muller (2009)

‘Mega-studies’

Plots indicate non-linearity at lower doses

However, there are no experimental data

at acceptable human risk levels (≤ 1 in

105)

The feasibility and resources required to

conduct such studies in vertebrate

species are such that it is not possible to

achieve the necessary power for this

purpose

Epidemiological data for human cancers

Comparison of animal and human data for

IARC class 1 carcinogens

Aflatoxin B1, benzidine, chromium VI and vinyl

chloride

Derive dose resulting in 1 in 105 increase in

cancer incidence from experimental data and

human epidemiology, respectively

The ratio human/animal was ≥ 1 in all cases

Integration of findings

Plot of data from different lines of evidence using same

scale

‘Mega-studies’

Comparison of experimental data with human epidemiology

IARC class 1 carcinogens

IARC class 2a carcinogens (no significant increase in risk in exposed

populations)

Estimates obtained by expert elicitation

Evidence for non-linearity at low exposure levels

Little or no data at acceptable human risk levels (1 in 105)

Uncertainty associated with many of the estimates

The TTC approach

A TTC value is:

A human exposure value for a chemical of unknown

toxicity below which the probability of adverse effects

on human health is considered to be very low

following (oral) exposure for a lifetime

Fre

qu

en

cy

Log hGV (mg/kg/d)

Low probability

hGV will be

below this

value

Murray-Rust et al, 1997

TTC for compounds that are potentially genotoxic

From Kroes et al (2004)

0.15 µg/day

1.5 µg/day

Potency

Specific groups of potent

genotoxic carcinogens

(CoC) excluded

Rela

tive p

robabili

ty d

ensity

-log10 Dose (mg/kg bw per day)

1 in 106 risk

TD50s VSDs

Examples of compounds that should probably not be considered in deriving a TTC for genotoxic carcinogens

Nafenopin Oxazepam

Penobarbital

Chloroform

Retinol acetate

1,4-Dioxane

Re-evaluation of TTC for potential genotoxins

Human relevance of included studies, tumour types and

data points

Study design, maximum dose, route of exposure, dose groups,

number of animals per group, duration, etc

Mode of action (DNA-reactive: yes/no/unknown)

Criteria for DNA reactivity

POD selection for potency estimate

Choice of POD, method of extrapolation

‘Safe dose’ definition and derivation based on current

approaches

Review appropriateness of cohort of concern and update

as necessary

Toxic effect

NOAEL/

BMDL

0 0.1 1 10 100 1000

10 10

RV

UF

Test species Average human Sensitive

human

Derivation of reference values

Dose-association for blood lead levels and IQ

Lanphear et

al. (2005)

Jusko et al.

(2008)

The BMDL01 for

developmental

neurotoxicity = 1.2

μg/dL (B-Pb) (EFSA)

Identification of POD

Andrade et al, 2006a Andrade et al, 2006b

Hypothalamic/preoptic area aromatase

activity in newborn (PND1) male rats,

exposed in utero from GD6-PND1

Sperm morphology in male (PND144) rats

exposed in utero from GD6-PND21

*cf concurrent controls; § cf historical controls

5 6

Mode of action and key events

•External dose

EXPOSURE

•Absorption

KEY EVENT

•Target tissue exposure

KEY EVENT

•Biological perturbation[s]

KEY EVENT [S]

•Pathological change[s]

KEY EVENT[S]

•Adverse health effect

Other factors (e.g. lifestyle, environment, homeostasis)

Host characteristics (e.g.

lifestage, genetics)

Other factors (e.g. lifestyle, environment, homeostasis)

Host characteristics (e.g.

lifestage, genetics)

Other factors (e.g. lifestyle, environment, homeostasis)

Host characteristics (e.g.

lifestage, genetics)

Other factors (e.g. lifestyle, environment, homeostasis)

Host characteristics (e.g.

lifestage, genetics)

Systems-based approach

Sturla et al, 2014

Conclusions

Available evidence tends to support the adequacy

of current risk assessment approaches

The TTC value for potential genotoxins merits re-

evaluation

ILSI Europe Expert Group “Update of the cancer potency

database (CPDB)”

Uncertainties are such that establishing

unequivocally the nature of the dose-response

relationship at human relevant exposures will

require systems-base approaches