When the measured “dose”

doesn’t make the poison:

Some thoughts around analytics,

exposure science & toxicology

Ben Smith, Eric Chan & James Chan

A*STAR Innovations in Food & Chemical Safety (IFCS) Programme

“The opinions expressed in this presentation and on the following slides are those of the authors anddo not necessarily represent those of the Innovations in Food and Chemical Safety (IFCS) Programme,

A*STAR Research Entities nor any IFCS collaborators.”

Disclaimer #1My thoughts are my own!

I’m a Toxicologist by training, Risk Assessor by trade….analytical chemist by necessity!

Disclaimer #2Uncertainty/probability vs precision/accuracy

Artiste: Mark Heath (http://www.markheath.com)

Brief Introduction to A*STAR IFCS Programme

Exposure vs Dose

Chemophobia (“Drive to Zero”)

Understanding Exposure & Dose

– some IFCS activities

Regulatory Integration

Tripartite Collaboration

AgendaMore a collection of thoughts…

Dermal Toxicology

Mechanisms of Toxicity

Human organ models

Allergenicity

PBTK & Exposure Modelling

Focus Areas

A*STAR Innovations in Food and Chemical Safety (IFCS) ProgrammeEnhancing consumer safety through sound and sustainable science

Increase Singapore’s economic competiveness1

• Leverage Singapore’s scientific strengths to evolve

regulatory processes and develop relevant safety tools

• Grow Singapore’s branding for high quality and safe

food/consumer products

• Enhance the safety of Singaporean and global

consumers through sound and sustainable science

• Build expertise and grow future talent to stay ahead

of changing regulatory, industry and consumer trends

Protect Singapore’s population2

Food & Chemical Safety

Establish Singapore as the preferred hub in Asia

for companies to innovate & launch new products

Enable a proactive, science-driven and

harmonized regulatory environment

Why IFCS?Creating Growth, Enhancing Lives

Scientists & Regulators Industry

58.2% have

considered using

alternatives or

animal-free methods

57.7% feel that

non-animal

methodologies can

be established in

their field of research

Consumers

“What can I do for the environment?”

Industry needs to balance:

▪ Changing regulatory pressures

▪ Responsible use of natural resources

▪ Recognition social welfare issues

▪ Food/product safety requirements

▪ “Go-to-market” speed & ease“Humans are not 70kg mice”Thomas Hartung,CAAT, John Hopkins University

In Singapore...

Why IFCS?Changing trends across Singapore

“There are 100,000 chemicals in products we use every day and we are missing 90 per cent of the safety information we need”

Thomas Hartung, CAAT, John Hopkins University

A major challenge for society and industry

A global responsibility

Including AsiaTons of tainted melamine milk powder products have been destroyed in China in 2008 (Source: EPA)

Why IFCS?The cost of poor product safety

Etienne Thiry, Belgian Food Safety Agency

“Chemical hazards are largely present in the food chain. Their detection and quantification are of upmost importance. However, a real challenge is their huge diversity. It is essential to assist the risk manager with sound scientific opinions, especially for newly identified contaminants”

▪ 600 Mio people fall ill & 420K die from eating

contaminated food each year

▪ ca. US$ 95 billion in lost productivity & curtailed trade

annually (FAO)

Singapore Food Agency

• > 90% of Singapore’s food is imported

• Increasing novel food ingredient applications

Why IFCS?The complexity of the Singapore Food Story

✓ Are we asking the right questions about chemicals

in our environment (across the exposure to outcome

continuum)?

✓ Are we communicating properly about chemical

risks (what is the role of analytical chemists)?

✓ Are current analytical & toxicological tools being

utilised to best effect in supporting exposure, risk

assessment & safety research?

Session: Advance Food & Water Integrity Through Technology Innovation2019 Food and Water Summit: Fostering sustainability and integrity through collaboration

Analytical Chemists in the Wild West Toxicologists & PBTK Modelers in the Wild West

Exposure & DoseYou say pota(y)to, I say pota(a)to….

And the consumers say…Trace chemical safety and chemophobia

Melamine is toxic but…Trace chemical safety and chemophobia

A statement made by an instrument maker technical expert

at a technical meeting (not Waters!)

“Sensitive methods are requested by

regulators to analyze for highly toxic

contaminants like melamine.”

Melamine is not really “highly” toxic (acute oral LD50 in F344

rats is 3161 mg/kg body weight; NOAEL 8.4–10.9 mg/kg bw/day)

Issue is no one knew about it, sensitive sub-populations,

accumulative exposure…untargeted analysis issue

• US study found BPA concentrations in the urine of 2,157adults and children older than 6 years of age. 92% of thosesampled had BPA in their urine.

• Laboratory tests of the umbilical cord blood of 10 newbornsfound an average of 200 chemicals that can cause cancer,brain damage, birth defects and other health ailments,according to a study sponsored by the EnvironmentalWorking Group.

• A National Geographic writer chronicled his efforts to learnwhat chemicals were in his body. He was told that a fireretardant, PBDE, was found in his blood at levels 10 timesthe average found in a small study of US residents.

Biomonitoring reported in the press…Trace chemical safety and chemophobia

• All created fear/news stories, all commented on the sensitivity

of the analytical methods, none demonstrated that the

levels of chemicals detected posed any risk of harm.

Peter Sandman’s Formulae: (Perceived) Risk = 𝐻𝑎𝑧𝑎𝑟𝑑 (𝑂𝑢𝑡𝑟𝑎𝑔𝑒)

Safety / Risk

Assessments

Emotional

Response

David Galbraith, Cardno ChemRisk (2005)

Presentation to the International Society of Regulatory

Toxicology & Pharmacology

“Our vastly improved abilities to detect

have often outstripped our abilities to

detect meaning.”

Problems with data interpretationPutting exposure back into risk assessment

Communicating with granting bodies is a different matter!Using chemophobia for good!

So what are we doing to better understand exposure & dose?The IFCS road map: Looking across the exposure to outcome continuum

Inhalation

Ingestion

Dermal

So what are we doing to better understand exposure & dose?Current & developing areas under IFCS

Intake and Consumption HabitsConsumption surveys

Local exposure assessment case studies

Total Diet Study

(Bottom-up) Physiologically-based

ToxicokineticsTranslate external exposure to

internal tissue dosage

Singapore Branded FoodProducts

DatabaseData focus on Nutrition & Food Safety

(ingredients, packaging material, source)

BiomonitoringTargeted/untargeted analysis, body burden;

Link to exposure models

Model DevelopmentAggregate exposure for multi-use ingredients

Probabilistic Chemical Model®

Frequency Amount Substance Concentration

40g

Diary Day 1

500g

200g

90g

0.6%

0.02%

1.5%

0.007%

Material Concentration

20%

1%

5%

40%

Concentration data

Bornyl acetate

Food Chemical Intake & Probability of Addition/PresenceRefining the amount on the plate!

Daily

Average

Daily

Average

Daily

Average

Probabilistic Chemical Model®

PA= volume flavor material sold per food category/Total volume sold all flavors

X

X

X XX

1

0.51

10.2

0.2 0.5

1

0.80.7

XXX

XXXX X

0.5

Food Chemical Intake & Probability of Addition/PresenceRefining the amount on the plate!

Results

Benzaldehyde

Lower Higher

MeanMedian Consumer Exposure95th Percentile

Consumer Exposure

Population Exposure

Results

Benzaldehyde

Aggregate Exposure (fragrances)Multiple sources, concentrations, routes & even application sites

Raw Food AmountBioavailability

CorrectionChemical

Concentration

Actual Intake

*

Recipe Fraction Correction

*

Edible Portion Correction

*

Raw Product Correction

Correction Factor

Assay Concentration

*

Processing Correction

*

Limit of Reporting

Correction

Presence Probability

Chemical Intake =

× ×

∑ What you eat What’s in the food

How much you absorb× ×

Tissue

Dose

Correction Factor

×

How much reaches the

target×

Physiological Based Toxicokinetic ModellingExternal Exposure to Internal Dose

Systems data:

Human physiology

Metabolism Fraction

unbound

Transporters

Blood/plasma

ratio

PermeabilityLog P, Log D,

pKa

Chemical data:

In vitro ADME assaysPredicted

Toxicokinetics

+ =

Concept: Assemble multiple pieces of in vitro

data pertaining to ADME processes to predict

toxicokinetic profiles

Case study on 3 statins. Between 8-15

sets of in vitro data were used to build

PBTK models for each compound

Bottom-up PBTK ModellingAdopting a systems biology approach

KidneyIntestine Liver

J Pharm Sci. 2016 Feb;105(2):443-459

Acta Pharm Sin B. 2016 Sep; 6(5): 363–373

http://transportal.compbio.ucsf.edu/organs/small-intestine/

Metabolism/

ExcretionAbsorption Excretion

Statins as a first model for transportersUnderstanding the role of active transport in “chemical” uptake

Quantitative proteomics to measure transporter

and metabolic enzyme abundance – scaling factors

Transporter Cultured hepatocyte Liver tissue Whole liver

pmol/min/106

hepatocytes

Inter-cell type

difference

Down-regulation

in culturepmol/min/mg

total cell proteinMembrane vs

Whole cell

pmol/min/g liver pmol/min

Liver weight

Statins as a first model for transportersScaling factors for in vitro-to-in vivo extrapolation for transporters

Chem. Commun., 2016, 52, 8918-8934

(& modellers)

Building such models does requires chemical specific methods! Chemistry is the central science…

Parameter(units)

Clinical data

No scaling (Fold-difference)

With scaling (Fold-difference)

Cmax (ng/mL)37.10

177.96(4.80)

47.56(1.28)

AUC (ng/mL.h)

164.00 1695.49 (10.34)236.31(1.44)

CL(L/h)

48.904.72

(0.097)33.85(0.69)

IV 8 mg,

4 h

infusion

Rosuvastatin The importance of understanding physiological processes

PET imaging of [11C] labelled Rosuvastatin in humans

Clin Pharmacol Ther. 2019 May 18 (epub ahead of print)

Rosuvastatin liver concentrations

Interplay between

liver uptake,

metabolism,

efflux, excretion

Rosuvastatin Validation of predicted tissue concentrations

Ki = 1.11 ng/mLIC50 = 2.60 ng/mL

Rosuvastatin Inhibition of HMG-CoA Reductase

Ki = 1.11 ng/mL

Duration of action

Simulated Rosuvastatin Plasma and

Liver Concentrations

Time dimension – cellular recovery? Irreversible cellular injury?

Rosuvastatin The time makes the poison!

• Poly- and perfluoroalkyl substances (most widely studied arePFOA & PFOS)

• Found ubiquitously in humans (99% of blood samples)

• Exposure routes include consumer products, food, contaminatedwater sources

• Metabolically inert

• Biologically persistent with long half-lives (between 3.5-7.5 years)– why?

• Different routes of elimination (renal, biliary)

• Highly dependent on transporters

• Some SAR can be observed (chain length)

• PBPK models not sufficiently mechanistic, extrapolate from animaldata

• Vastly different animal and human disposition

PFOS, PFAS, PFOAAdvancing the Pace of Chemical Risk Assessment

In vitro to in vivo extrapolation (IVIVE) & reverse dosimetryHumans are not 70kg mice… but equally we are not 70kg culture plates!

Predict organ concentrations of chemicals to guide interpretation of in vitro toxicodynamic data

Reverse dosimetry

Scaling factor 1: microsomal protein per gram of liverIn vitro

CLuint

CLuint per g liver

Scaling factor 2: liver weight CLuint per

liver

WSM: liver blood flow, fraction unbound

CLH: Hepatic clearance

1. Break an in vivo process down (mechanistically) to a smaller, measurable unit

2. Design an in vitro assay to quantitatively measure that process

3. Scale the measurement (mechanistically) back to an in vivo process

Inter-individual variability Special populations Disease populations

Intra-species differences

“When CODEX started, toxicity and exposure data was primarily from the EU and US. This is not, especially for exposure, necessarily representative for Asia and other parts of the world. Although EU and US diets have been shown to be “similar”, Asian diets are very different and how Asians metabolize chemicals may be very different….we are now (in the panels & working groups) starting to see a lot more data coming from Asia and will keep seeing an increase in country specific data”

Dr. Yukiko Yamada (Japan); presented at CIFSQ, 7-8 Nov., Shanghai, China

Reducing testing, Reducing UncertaintyExtrapolation between species and sub-populations

Dede et al. Environ Toxicol Pharmacol. 2018 Jan; 57:104-114

PBTK as a complimentary tool for biomonitoringTiming makes the poison

Internal TTC Approach

Regulatory Toxicology and Pharmacology 103 (2019) 63–72

Regulatory Toxicology and Pharmacology 103 (2019) 63-72

The ultimate goalLinking PBTK to Toxicodynamic data & risk assessment approaches

Blaauboer et al. Considering new methodologies in strategies for safety assessment of foods and food

ingredients. FCT 91 (2016) 19-35

Integration into regulatory frameworksPartnering with the SFA on 21st century safety

Collaboration is the KeyA*STAR IFCS are keen to collaborate on case studies (across models) in this space

Future Ready Food Safety HubTripartite collaboration for Food Safety Leadership

Regulatory Readiness | Science & Innovation | Talent Development

✓ Building local food safety capabilities to support growing innovation in food production & manufacturing

✓ Developing the science for new food safety standards for industry and the region

Research is required to address gaps

in food safety capabilities and

technologies:

(i) Ascertain food safety of novel foods

and new food processing solutions

(ii) Develop intelligent supply chains

(iii) Understand consumer behaviour

towards food

(iv) Develop talent & training

Possible industry collaborators

But the question is how much are we actually exposed to? This is truly the ‘Golden’ age of exposure modelling supported by analytical chemistry!

Please check out the A*STAR IFCS Programme @ www.a-star.edu.sg/ifcs

Thanks – Any Questions?