claude beigel, phd. exposure assessment senior scientist research triangle park, usa practical...

22
Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

Upload: carlos-oneill

Post on 27-Mar-2015

215 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

Claude Beigel, PhD.

Exposure Assessment Senior Scientist

Research Triangle Park, USA

Practical session metabolitesPart III: plenary discussion of results

Page 2: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

2

Results Example 1Visual Evaluation of Goodness of fit (Parent + Metabolite1)

ParentMetabolite1

Example 1 data set (SFO)

0 10 20 30 40 50 60 70 80 90 100 110 120

Time (days)

0

10

20

30

40

50

60

70

80

90

100

110

Su

bst

ance

(%

AR

)

Example 1, parent residuals

-20

-15

-10

-5

0

5

10

15

20

0 20 40 60 80 100 120

Time (days)

Res

idu

al (

% A

R)

Example 1, metabolite1 residuals

-3

-2

-1

0

1

2

3

0 20 40 60 80 100 120

Time (days)

Res

idu

al (

% A

R)

Page 3: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

3

Hands-on Example 1Visual Assessment (Parent + Metabolite1)

Graph Assessment / Remarks

ParentOverall fit

Good, initial scattering

Residuals Random distribution

Metabolite1

Overall fit Excellent

Residuals Random distribution

Page 4: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

4

Hands-on Example 1Statistical Indices (Parent + Metabolite1)

2-test Relevant Parameters

Estimated (y/n)Number of Parameters

Minimum 2 Error

Percentage

ParentPini

kP

y

y2 9.2

Metabolite1ffM1

kM1

y

y2 4.9

t-test Estimated Value

Standard Error

Number of Data Points

Number of Estimated

ParametersP-value Conclusion

kP 0.0508 0.0033 38 4 <0.001 Significant

kM1 0.1018 0.0320 38 4 <0.001 Significant

Page 5: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

5

Hands-on Example 1Conclusion and Endpoints (Parent + Metabolite1)

SFO model is considered appropriate for both parent and metabolite

Trigger endpoints for Metabolite 1:

DT50 = 6.8 d and DT90 = 22.6 d

Modeling endpoints:

kP = 0.0508 d-1 (equivalent to half-life of 13.7 d), ffM1= 0.5881 and kM1= 0.1018 d-1 (equivalent to half-life of 6.8 d)

kP_M = 0.0299 d-1 (equivalent to half-life of 23.2 d), kP_S = 0.0209 d-1 (equivalent to half-life of 33.1 d), and kM1= 0.1018 d-1

(equivalent to half-life of 6.8 d)

or

Page 6: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

6

Hands-on Example 1Parent + Metabolite1+ Metabolite2

Initial fit with flow from Metabolite 1 to sink results in formation fraction ffM2 of 0.98 (stepwise fit, parent and M1 parameters fixed) or >1 (simultaneous fit, all parameters free)

The question is: should we remove or keep this flow (does Metabolite 1 degrade exclusively to Metabolite 2, or does it form other metabolites and/or bound residues too)?

Let’s assume that additional information, e.g. a degradation study conducted with Metabolite 1 also suggests 100% formation of Metabolite 2, ffM2 is fixed to 1, i.e. the flow from Metabolite 1 to sink is removed

Parent Metabolite1

Sink

fP_M1

fP_S

Metabolite2

fM1_M2

fM2_S

Parent Metabolite1

Sink

fP_M1

fP_S

Metabolite2

fM1_M2

fM2_S

fM1_S

Page 7: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

7

Results Example 1Visual Evaluation of Goodness of fit (Parent + Met1 +Met2)

ParentMetabolite1Metabolite2

Parent + metabolite1 + metabolite2, SFO-SFO-SFO

0 10 20 30 40 50 60 70 80 90 100 110 120

Time (days)

0

10

20

30

40

50

60

70

80

90

100

110

% A

R

Example 1, parent residuals

-20

-15

-10

-5

0

5

10

15

20

0 20 40 60 80 100 120

Time (days)

Re

sid

ua

l (%

AR

)

Example 1, metabolite1 residuals

-3

-2

-1

0

1

2

3

0 20 40 60 80 100 120

Time (days)

Res

idu

al (

% A

R)

Example 1, metabolite2 residuals

-5

-4

-3

-2

-1

0

1

2

3

4

5

0 20 40 60 80 100 120

Time (days)

Res

idu

al (

% A

R)

Page 8: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

8

Hands-on Example 1Visual Assessment (Parent + Met1 +Met2)

Graph Assessment / Remarks

ParentOverall fit

Good, initial scattering

Residuals Random distribution

Metabolite1

Overall fit Excellent

Residuals Random distribution

Metabolite2 Overall fit Excellent

Residuals Random distribution

Page 9: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

9

Hands-on Example 1Statistical Indices (Parent + Met1 +Met2)

2-test Relevant Parameters

Estimated (y/n)Number of Parameters

Minimum 2 Error

Percentage

ParentPini

kP

y

y2 9.2

Metabolite1ffM1

kM1

y

y2 5.0

Metabolite2ffM2

kM2

n (fixed to 1)

Y1 3.9

t-test Estimated Value

Standard Error

Number of Data Points

Number of Estimated

ParametersP-value Conclusion

kP 0.0507 0.0021 56 5 <0.001 Significant

kM1 0.0999 0.0091 56 5 <0.001 Significant

kM2 0.0114 0.0014 56 5 <0.001 Significant

Page 10: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

10

Hands-on Example 1Conclusion and Endpoints (Parent + Met1 +Met2)

SFO model is considered appropriate for parent and both metabolites

Trigger endpoints

Metabolite1 DT50 = 6.9 d and DT90 = 23.1 d

Metabolite2 DT50 = 61.0 d and DT90 = 203 d

Modeling endpoints:

kP = 0.0507 d-1 (equivalent to half-life of 13.7 d), ffM1= 0.5813 and kM1= 0.0999 d-1 (equivalent to half-life of 6.9 d), kM2= 0.0114 d-1

(equivalent to half-life of 61.0 d),

kP_S = 0.0212 d-1 (equivalent to half-life of 32.7 d), kP_M = 0.0295 d-1 (equivalent to half-life of 23.5 d), kM1_M2 = 0.0999 d-1 (equivalent to half-life of 6.9 d), and kM2_S = 0.0114 d-1 (equivalent to half-life of 61.0 d)

or

Page 11: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

11

Results Example 2Visual Evaluation of Goodness of fit (Parent FOMC)

ParentMetabolite

Example data set 2, parent + metabolite, FOMC-SFO

0 10 20 30 40 50 60 70 80 90 100 110 120

Time (days)

0

10

20

30

40

50

60

70

80

90

100

110

% A

R

Example 2 (parent FOMC), parent residuals

-15

-10

-5

0

5

10

15

0 20 40 60 80 100 120

Time (days)

Re

sid

ua

l (%

AR

)

Example 2 (parent FOMC), metabolite residuals

-5

-4

-3

-2

-1

0

1

2

3

4

5

0 20 40 60 80 100 120

Time (days)

Res

idu

al (

% A

R)

Page 12: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

12

Hands-on Example 2, parent FOMCVisual Assessment

Graph Assessment / Remarks

ParentOverall fit Excellent

Residuals Random distribution

Metabolite

Overall fit Excellent

Residuals Random distribution

Page 13: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

13

Hands-on Example 2, parent FOMCStatistical Indices

2-test Relevant Parameters

Estimated (y/n)Number of Parameters

Minimum 2 Error

Percentage

Parent

Pini

P

P

y

y

y

3 5.5

MetaboliteffM

kM

y

y2 4.1

t-test Estimated Value

Standard Error

Number of Data Points

Number of Estimated

ParametersP-value Conclusion

kM 0.0200 0.0019 38 5 <0.001 Significant

Page 14: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

14

Hands-on Example 2Conclusion and Trigger Endpoints (Parent FOMC)

SFO model is considered appropriate for metabolite in combination with FOMC model for parent

Trigger endpoints for Metabolite:

DT50 = 34.7 d and DT90 = 115 d

Endpoints for PEC soil calculations:

P = 0.9425, P = 4.436, ffM= 0.8018 and kM= 0.0200 d-1

Page 15: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

15

Results Example 2Visual Evaluation of Goodness of fit (Parent DFOP)

ParentMetabolite

Example data set 2, parent + metabolite, DFOP-SFO

0 10 20 30 40 50 60 70 80 90 100 110 120

Time (days)

0

10

20

30

40

50

60

70

80

90

100

110

% A

R

Example 2 (parent DFOP), parent residuals

-15

-10

-5

0

5

10

15

0 20 40 60 80 100 120

Time (days)

Re

sid

ual

(%

AR

)

Example 2 (parent DFOP), metabolite residuals

-10

-8

-6

-4

-2

0

2

4

6

8

10

0 20 40 60 80 100 120

Time (days)

Res

idu

al (

% A

R)

Page 16: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

16

Hands-on Example 2, Parent DFOPVisual Assessment

Graph Assessment / Remarks

ParentOverall fit

Excellent up to DT90, slight overestimation

afterward

ResidualsRandom distribution

up to DT90

Metabolite

Overall fit Excellent

Residuals Random distribution

Page 17: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

17

Hands-on Example 2, Parent DFOPStatistical Indices

2-test Relevant Parameters

Estimated (y/n)Number of Parameters

Minimum 2 Error

Percentage

Parent

Pini

g

k1

k2

y

y

y

y

4 6.5

MetaboliteffM

kM

y

y2 3.6

t-test Estimated Value

Standard Error

Number of Data Points

Number of Estimated

ParametersP-value Conclusion

k1 0.3227 0.0613 38 6 <0.001 Significant

k2 0.0340 0.0064 38 6 <0.001 Significant

kM 0.0216 0.0021 38 6 <0.001 Significant

Page 18: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

18

Hands-on Example 2Conclusion and Modeling Endpoints (Parent DFOP)

SFO model is considered appropriate for metabolite in combination with DFOP model for parent

Modeling endpoints (higher Tier approach based on parent DFOP):

g = 0.5509, k1 = 0.3227 d-1 (equivalent to half-life of 2.15 d), k2 = 0.0340 d-1 (equivalent to half-life of 20.4 d), ffM= 0.8332 and kM= 0.0216 d-1 (equivalent to half-life of 32.0 d)

Page 19: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

19

Results Example 2Visual Evaluation of Goodness of Fit (Metabolite Decline)

Metabolite

Example data set 2, metabolite decline

0 10 20 30 40 50 60 70 80 90 100 110 120

Time (days)

0

10

20

30

40

50

60

70

80

90

100

110

% A

R

Example 2, metabolite decline

-10

-8

-6

-4

-2

0

2

4

6

8

10

0 20 40 60 80 100 120

Time (days)

Res

idu

al (

% A

R)

Page 20: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

20

Hands-on Example 2, Metabolite DeclineVisual Assessment

Graph Assessment / Remarks

Metabolite decline

Overall fitGood, slight

underestimation at last time points

Residuals No distinct pattern

Page 21: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

21

Hands-on Example 2, Metabolite DeclineStatistical Indices

2-test Relevant Parameters

Estimated (y/n)Number of Parameters

Minimum 2 Error

Percentage

MetaboliteMmax

kM

y

y2 5.7

t-test Estimated Value

Standard Error

Number of Data Points

Number of Estimated

ParametersP-value Conclusion

kM 0.0216 0.0021 12 2 <0.001 Significant

Page 22: Claude Beigel, PhD. Exposure Assessment Senior Scientist Research Triangle Park, USA Practical session metabolites Part III: plenary discussion of results

22

Hands-on Example 2Conclusion and Endpoints (Metabolite Decline)

SFO model is considered appropriate for metabolite decline

Metabolite decline rate may be used as worst-case estimate for trigger endpoints

Trigger endpoints: DT50 = 49.7 d and DT90 = 165 d (compared to DT50 = 34.7 d and DT90 = 115 d from actual degradation rate)

Decline rate may also be used as modeling endpoint for metabolite, if calculated from maximum observed

Modeling endpoint: kM= 0.0139 d-1 (equivalent to half-life of 49.7 d)