A Source to Outcome Approach for Inhalation Risk ... › groups › ss › RASS › docs › ...• Bob Parr-Dobrzanski, Alex Charlton, Doug Wolf Pacific Northwest National Laboratory

Paul M. Hinderliter, PhDSenior Research ScientistProduct Metabolism ADMESyngenta Crop [email protected] October 11, 2017

A Source to Outcome Approachfor Inhalation Risk AssessmentSOT/RASS Webinar

2

Acknowledgements

● Syngenta- Operator and Consumer Safety

• Sheila Flack, Tharacad Ramanarayanan- Toxicology and Health Sciences

• Bob Parr-Dobrzanski, Alex Charlton, Doug Wolf

● Pacific Northwest National Laboratory- Rick Corley, Andrew Kuprat, Sarah Suffield, Senthil Kabilan

3

Defining the Problem

● Pesticide re-registrations examine the completeness of the toxicological database, including inhalation exposures- requirements often include a subchronic (90-day) inhalation study

(Guideline 870.3465).● “However, there may be “smarter testing”/alternative approaches that

are suitable to inform …inhalation toxicity in lieu of a subchronic inhalation study”

4

Test Chemical

● Respiratory irritant when tested in accordance with guideline standards (MMAD = 1-4 um) – Toxicity Category I.

● 2 week inhalation study was conducted with 2 weeks recovery

Aerosol Concentration MMAD (GSD)Target Gravimetric Analytical(mg chem/L) (mg/L) (mg chem/L) (µm)

0.001 3.82 ± 1.847 0.0011 ± 0.00042 2.13 (3.238) to 3.92 (2.439)

0.003 4.21 ± 2.065 0.0029 ± 0.00078 2.17 (2.780) to 3.47 (2.335)

0.009 4.03 ± 1.045 0.0096 ± 0.00219 2.17 (2.537) to 2.52 (2.471)

0.015 4.07 ± 1.213 0.0143 ± 0.0026 2.21 (2.947) to 3.13 (2.102)Overall Mean1 4.03 2.72 (2.606)

5

Rat in vivo 14 day study endpoint

LOAEC of 0.001 mg/L, based on histopathology observed in the larynx and nasal cavity

Nominal concentration →

Tissue ↓

0.001 mg/L 0.003 mg/L 0.009 mg/L 0.015 mg/L

LarynxSquamous metaplasia 13d

7/10 min3/10 mild

1/10 min6/10 mild3/10 mod

4/10 mild5/10 mod1/10 mark

7/10 mod3/10 mark

LarynxSquamous metaplasia13d + 2d recovery

4/5 min 1/5 min4/5 mild4/5 mild1/5 mod

1/5 mild4/5 mod


3/5 min 4/5 min1/5 mild2/5 min3/5 mild

3/5 min2/5 mild


1/5 min 4/5 min 4/5 min1/5 mild4/5 min1/5 mild

6

Revisiting the problem

● Repeat dose 90-day inhalation study is regulatory requirement- No new or additional information will be forthcoming from such a

study that will improve safety assessment● No additional systemic risk from inhalation exposure

- Contact irritation in respiratory tract● Long history of safe use

7

Exposure Based Risk Assessment

● The inhalation risk for non-volatile pesticides is very different from volatile chemicals

● Many agrochemicals have low or negligible vapor pressure- Inhalation exposure not a standard exposure scenario

● Alternative data generation can provide approaches that are suitable to inform inhalation toxicity in lieu of a acute/subchronic inhalation studies

8

Source to Outcome Approach

● Source- Evaluate the particle size distribution of pesticide

applications. ● Exposure

- Investigate the impact of spray quality on inhalation exposure.

● Dosimetry- Compare deposition in human versus rat airways.

● Outcome- Predict human inhalation toxicity incorporating spray

particle size.

9

Inhalation Exposure

HECRisk

Characterization

Inhalation Exposure

HEC Risk Characterization

Particle Size Distribution of

Inhalable Particles

In vitro TestingC

FD M

odel

ing

Conceptual Model

• Chlorine• Diesel• ultrafine particles• formaldehyde• Glutaraldehyde• Acetaldehyde• vinyl acetate

How far will aerosol droplets go?

What is the exposure needed to get an adequately toxic dose to the various locations in the human respiratory tract to cause a lesion at that site?

Literature to aid in model refinement and verification

11

Adverse Outcome Pathway for squamous cell metaplasiaof the larynx

Necrotic injury to luminal respiratory

epithelial cells

Loss of surface epithelium

Basal stem cell reprogramming

Respiratory epithelium changes to squamous cells

Squamous cell metaplasia

Repeated injury

Conventionally modelledin vitro

Adapted from: Renne R, Brix A, Harkema J, Herbert R, Kittel B, Lewis D, March T, Nagano K, Pino M, Rittinghausen S, Rosenbruch M, Tellier P, Wohrmann T (2009) Proliferative and Nonproliferative Lesions of the Rat and Mouse Respiratory Tract, Tox Path, 37:5D-73S

12

Mucilair in vitro toxicity screen

● MucilAirTM – 3D in vitro cell model of human upper airway epithelium prepared from differentiated primary human cells from a single healthy donor.

● Utility as a prescreen for ‘point of contact’ inhalation toxicity.

13

MucilAir 3D in vitro model

● Measures a variety of membrane and cell damage endpoints as markers of irritation- Trans-epithelial electrical resistance (TEER): measures the integrity of tight

junctions between cells in the membrane - Lactate dehyrogenase (LDH): An enzyme present in most cells released

when cells suffer cytotoxic membrane damage- Resazurin metabolism: reduced to a fluorescent product in viable cells used

as a measure of metabolic competence

14

Design of multi-donor MucilAir study

● Endpoints: TEER, LDH, and resazurin metabolism● Tissues from 5 separate donors● 24 hour topical exposure● 10 concentrations / donor● 6 replicates / concentration / donor

Dose Level

Test Concentration (mg/L)

1 1.9952 5.0123 7.9434 12.595 19.956 31.627 50.128 79.439 125.910 199.5

15

Data from MultiDonor Study (TEER)

TEER BMD values were calculated as being a 10% change from controls

16

Data from MultiDonor study (LDH)

Due to the very low variability in the “no-response” area of the LDH curves a 15% change from controls was used for BMD values

17

Data from MultiDonor study (resazurin)

Low doses induced an apparent increase in viability relative to controls in resazurin measurements

Setting BMD values relative to controls was therefore considered insufficiently conservative and BMD values were defined as a 10% change from the mean 4 lowest concentrations measured

18

MultiDonor MucilAir study results

● Individual donor responses were very similar across the 5 donors used in this study, suggesting low inter-donor variability in sensitivity

● All endpoints (TEER, LHD, resazurin) responded similarly to test material challenge

BMDL (mg/cm2)

TEER LDH Resazurin Mean

Donor 1 0.00430 0.00703 0.00344 0.00470

Donor 2 0.00317 0.00963 0.00224 0.00409

Donor 3 0.00822 0.00878 0.00400 0.00661

Donor 4 0.00937 0.00998 0.00066 0.00395

Donor 5 0.00909 0.00875 0.00674 0.00813

Mean 0.00625 0.00877 0.00267 0.00527

19

In Vitro Endpoint Conclusions

● Necrotic injury to the luminal surface of the respiratory epithelium is a key event in squamous metaplasia of the larynx in treated rats.

● This initial irritant response can be assessed in MucilAir.● TEER, LDH, and resazurin metabolism were measured

following treatment and these results were statistically interrogated to determine BMD and BMDL values.

● BMDL values were in the 30 – 100 mg/L range, with an overall mean mass per unit surface area concentrations of 0.00527 mg/cm2.

20

Pesticide Exposure Data

● Exposure data is commonly collected from agricultural workers using an OSHA Versatile Sampler (OVS) tube.

● Typically, the OVS tube data is only reported as total concentration, without consideration of particle size.- What is the particle size distribution

captured by this device?● Studies of spray particle size undertaken

at Syngenta to compare OVS tube with standard sizing methods.

OVS Tube

21

OVS tube sampling in inhalable particle size range (

22

Particle Sizing vs. Nozzle Type

● Concentration in the inhalable fraction increases from extremely coarse to very fine spray qualities.

● Regardless of nozzle type, systemic exposure is ~40% of total air concentration levels.

InhalableThoracicRespirable

Medium

ExtremelyCoarse

systemic exposure

0% 20% 40% 60% 80% 100%

Very Fine

ExtremelyCoarse

23

Selection of Human Relevant Particle Size Distribution

● Mixing/loading (pouring into water) or atomization

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

Mixing/Loading Flat fan, XR8003VS Hollow cone, TXVS-3

Con

c. (m

g/m

3 )

Medium Very Fine

Total concentration by Activity

0% 20% 40% 60% 80% 100%

1.7 µm 4 µm 13 µm

Mixing/Loading

Flat fan, XR8003VS(Medium)

Hollow cone, TXVS-3(Very Fine)

Proportion within Thoracic Fraction

● Conclusion: Air concentration varies with activity and spray quality, but particle size distributions are similar. Use “thoracic” fraction (13 µm) for mixing/loading as a health protective number.

24

What particle size distribution to use for CFD model?Selection of Human Relevant Particle Size Distribution

5% Respirable40% Thoracic 60% Extrathoracic

(D50=10 µm)

% of AirConcentration during spraying

(D50=4 µm)

Conclusion: Identified particle size distribution of 35 µm (GSD=1.5) for applicator; 13 µm (GSD = 1.5) for mixer/loader.

The ISO / ACGIH / ECEN sampling conventions for the inhalable, thoracic and respirable aerosol fractions can be used to derive Inhalable PSD

25

Animal versus Human Exposures

● Particle size distributions between rodent inhalation toxicity studies and spray applications are extremely different

● Sample rodent study found 2.33 (3.996) µm [MMAD (GSD)]

Droplet Size Cutoff (µm)

Fraction of Total Mass BelowCutoff (%)

2.5 0.77

10 1.03

30 5.5

100 13.8

150 22.4

200 37.3

Droplet Size Cutoff (µm)

Fraction of Total Mass BelowCutoff (%)

0.37 16.5

0.90 22.8

1.6 28.4

3.7 40.4

5.6 70.6

8.0 94.1

Fine/Medium reference nozzle Rat Inhalation Study

MMAD= Mass Median Aerodynamic DiameterGSD= Geometric Standard Deviation

26

● Multiple-Path Particle Dosimetry Model (MPPD v 2.11)- Version 3.04 is now available

● Computational model for estimating human and rat airway particle dosimetry.

● Calculates the deposition and clearance of monodisperse and polydisperse aerosols in the respiratory tracts of rats and human.

● Within each airway, deposition is calculated using theoretically derived efficiencies for deposition by diffusion, sedimentation, and impaction within the airway or airway bifurcation.

● Filtration of aerosols by the nose and mouth is determined using empirical efficiency functions.

● http://www.ara.com/products/mppd.htm

http://www.ara.com/products/mppd.htm

27

Comparative Deposition (Human vs. Rat)

Human

Rat

28

Computational Fluid Dynamics Modeling

● The outputs of the MPPD model, however, are not precise enough to define exact locations in the respiratory tract (e.g. the U-shaped cartilage in the rat larynx).

● More advanced models have previously been developed. These models use computational fluid dynamics (CFD) to simulate the deposition of a large number of individual particles.

29

Computational Fluid Dynamics Modeling

● Simulates the fluid dynamics of the airflow in the respiratory tract.● Surfaces are meshes of polygons, each of which can be monitored.● Commonly used in simulation of smaller particulates like bacteria

spores and diesel particulates.

30

CFD Simulation of Small Particles

Small particle deposition in the rabbit (left) and human (right)

31

Mass Deposited at End of Inhalation

Rat CFD/Aerosol Simulation

32

(65.64%)

(6.91%)

(0.44%)(0.03%)

(0.25%)

(0.01%)

(0.02%)

(0.03%)(0.04%)

(0.08%)

(0.03%)(0.11%)

(0.11%)

(0.05%)

74% Deposited in Nose, Pharynx, Larynx0.5% Deposited in Remaining Airways11% Escape the 3D Lung14.5% Remain suspended in airways at end of

inhalation (would be part of next breath).


33


● Surface concentration: 1.83x10-7 mg/cm2/breath● Breathing rate: 85 breaths/minute or 30600 breaths/6hr● Surface concentration: 5.60x10-3 mg/cm2/6hr

● MucilAir BMDL 5.27x10-3 mg/cm2/6hr- This is a human in vitro endpoint but response should be similar

34

(0%)

(98.4%)(1.3%)

(0.0002%)(0.0009%)

(0.0007%)

Human CFD/Aerosol Simulation(50 µm MMAD, 4 mg/L)

% of Inhaled Mass Deposited by Tissue Type

99.7% Deposited in Nose, Pharynx, Larynx0.0007% Deposited in Remaining Airways0% Escape the 3D Lung0.3% Remain suspended in airways at end

of inhalation

35

(98.4%)(1.3%)

(0%)

(0.0002%)

(0.0009%)

Human CFD/Aerosol Simulation(50 µm MMAD, 4 mg/L)

36

Deposition Patterns in Humans

Side

Ventral

larynx

larynx

37

Polydispersity

38

Deposition from the CFD model – Human Simulation

● Deposition at 75th percentile

Aerosol Diameter (µm)

Deposition (mg/cm2)Vestibule Respiratory Olfactory Pharynx Larynx Trachea

1 4.29E-06 3.04E-06 5.22E-06 1.71E-06 2.16E-06 7.34E-073 3.39E-06 2.43E-06 4.55E-06 1.26E-06 2.48E-06 7.69E-075 5.73E-06 2.86E-06 1.26E-05 1.48E-06 3.08E-06 6.36E-0710 1.62E-04 4.50E-06 1.77E-06 5.41E-06 1.40E-05 1.30E-0615 2.91E-04 2.90E-06 9.73E-07 3.49E-06 8.46E-06 1.39E-0620 2.76E-04 2.27E-06 6.50E-07 1.85E-06 2.67E-06 5.57E-0730 1.50E-04 1.89E-06 - 5.61E-07 1.03E-06 2.13E-07

39

Reconstruction of Polydisperse Distribution● From specific particle sizes

Aerosol Diameter (µm)

Percentage of DistributionMMAD=35GSD=1.5

MMAD=25GSD=1.5

MMAD=13GSD=1.5

1 0.00% 0.00% 0.00%3 0.00% 0.00% 0.14%5 0.00% 0.09% 6.52%10 0.48% 3.83% 37.50%15 3.66% 14.35% 31.89%20 14.80% 29.72% 18.05%30 81.06% 52.01% 5.91%

40

Calculation of Human Exposure Concentration - Applicators

● Aerosol: 35 (1.5) µm polydisperse● Breathing rate: 16.4 breaths/minute or 7872 breaths/8hr● Dilution Factor 0.438 lb/lb diluted formulation● In vitro BMDL 5.27x10-3 mg/cm2

● The exposure concentration is the air concentration necessary to match the in vivo 8 hour deposition with the in vitro BMDL

Respiratory Olfactory Pharynx Larynx TracheaDeposition

(mg/cm2/breath) 2.00E-06 1.41E-07 8.82E-07 1.61E-06 3.12E-07

Deposition(mg/cm2/8hr) 6.89E-04 4.85E-05 3.04E-04 5.54E-04 1.08E-04

Exposure Concentration(mg/L) 0.34 4.8 0.76 0.42 2.1

41

Calculation of Human Exposure Concentration - Mixer/Loaders







42

Calculation of Human Exposure Concentration - Bystander







43

Comparison of Simulated and in vitro tissue concentrations

Tissue Median Conc (mg/cm2) Fold greater than minimum tested MucilAir dose

Vestibule 8.76 x 10-5 0.55

Respiratory 4.40 x 10-8 1099

Olfactory 0 NA

Pharynx 1.35 x 10-8 3583

Larynx 1.11 x 10-8 4357

Mucilair NOEC > 1000 x MOE vs modelled median conc. in all tissues beyond anterior nasal cavity.

44

Conceptual Model

Inhalation Exposure

HEC

Risk Characterization

Particle Size Distribution of

Inhalable Particles

In vitro Testing

CFD

Mod

elin

g

Inhalation Evaluation

Systemic Exposure Assessment

MPPD

Portal of Entry

Toxicity Likely?

Adequate In vivo study available?

No

Yes

Portal of Entry Exposure AssessmentCFD + in vivo

Portal of Entry Exposure Assessment

CFD + in vitro

No

Yes

Inhalation Assessment Decision Tree

Residential Bystander Risk Assessment

● Max Average Exposure Range = 0.002 – 4.86 µg/m3

● Toxicity (HEC) Range (mg/L converted to mg/m3) = 112 – 1,590 mg/m3

LOC = 10

LOC = 100

Operator Risk Assessment

● Exposure Ranges:- Applicators = 0.11 – 313 µg/m3; Mixer/Loader = 0.33 – 259 µg/m3

● Toxicity (HEC) Ranges (mg/L converted to mg/m3):- Applicator = 335 – 4,760 mg/m3; Mixer/Loader = 77 – 616 mg/m3

LOC = 10

LOC = 100

48

Summary

● Human in vitro (MucilAir human airway epithelium) BMDL = 0.00527 mg/cm2

● Exposure particle size distribution for applicators/bystanders (MMAD = 35 µm, GSD 1.5) and mixer/loaders (MMAD = 13 µm, GSD 1.5)

● Exposure model derived HECs (0.077 – 4.8 mg/L) ● Compared HEC to human exposure values for risk

assessment

49

Conclusions

● This Source to Outcome Analysis Addresses:- Inhalation Study Requirement- Point of Departure- Uncertainty Factors

• Database• LOAEL to NOAEL• Interspecies

50

Questions and Feedback

Thank you

51

Excess Slides

52

Abstract

The Source to Outcome Approach provides a framework for improving and modernizing human health risk assessments where uncertainty is reduced through integrating hazard and exposure characterization. For non-volatile pesticides, using human-relevant particle size distributions from agricultural spray applications and mixing/loading scenarios with refined inhalation dosimetry models, such as computational fluid dynamic (CFD) models, more accurately estimates a human equivalent concentration (HEC) for human health risk assessments. Inhalation exposure data used by USEPA in human health risk assessments for operators, residential handlers and bystanders is typically measured with personal air monitors and environmental sampling without consideration of particle size distributions. Side-by-side air sampling with a Respicon particle sampler characterized the size distribution of aerosols captured on personal air monitors (i.e., OVS tubes) during spraying to derive a particle size distribution for respiratory dosimetry modelling. CFD models further link this exposure information with target site-specific dosimetry for the human by calculating surface concentrations of deposited particle formulations in discrete regions of the respiratory tract. CFD models are especially useful to provide a more accurate reflection of the deposition necessary to initiate the cascade of events resulting from irritant mediated response in the upper respiratory tract. The dosimetry modelled for the human can be compared to the benchmark dose level derived from human in vitro models (e.g., MucilAir) to calculate an HEC for risk assessments. The Source to Outcome Approach has been used to refine the human health risk assessment of pesticides by integrating exposure characterization, site-specific respiratory dosimetry (CFD model) and a human in vitro (MucilAir) model to derive an HEC. This approach can also be tailored for other aerosols where refinement of inhalation risk is warranted.

53

Available in vitro models for assessment of damage to respiratory epithelial cells

Easy to use and to maintain?

Able to model cell-cell interactions in response to toxicants?

Representative of in vivo tissue organisation?

Able to simulate the mechanical action of the respiratory system?

Suitable for long term tests?

Assessed for applicability of results to in vivo inhalation toxicity?

Immortalised cell monoculture (eg. A549 as reported in Mazzarella et al., 2007)

Yes No No No No No

Primary cell monoculture (Ghio et al., 2013)

No No No No No No

2D/3D cellular co-culture (eg. Rothen-Rutishauser et al., 2005)

No Yes Yes No No No

Lung-on-a-Chip (Huh et al., 2010) Yes

Yes (Limited

cell types)No Yes No No

MucilAir Yes Yes Yes No PotentiallyYes

(Current Work)

A Source to Outcome Approach�for Inhalation Risk Assessment�SOT/RASS WebinarAcknowledgementsDefining the ProblemTest ChemicalRat in vivo 14 day study endpointRevisiting the problemExposure Based Risk AssessmentSource to Outcome Approach Conceptual ModelLiterature to aid in model refinement and verificationAdverse Outcome Pathway for squamous cell metaplasia�of the larynxMucilair in vitro toxicity screenMucilAir 3D in vitro model Design of multi-donor MucilAir studyData from MultiDonor Study (TEER)Data from MultiDonor study (LDH)Data from MultiDonor study (resazurin)MultiDonor MucilAir study resultsIn Vitro Endpoint ConclusionsPesticide Exposure DataOVS tube sampling in inhalable particle size range (

Documents

A Source to Outcome Approach for Inhalation Risk ... › groups › ss › RASS › docs › ...• Bob Parr-Dobrzanski, Alex Charlton, Doug Wolf Pacific Northwest National Laboratory