[PPT] - Literature Model for FEV1 in COPD Trials Separating the Dynamic PowerPoint Presentation

SLIDE 1

Literature Model for FEV1 in COPD Trials

Separating the Dynamic Components of Placebo Effect, Disease Progression and Interacting Drug Effects

Global Pharmacometrics

Jakob Ribbing, Christine Falcoz, Itzela Correa and Steven W Martin 21st PAGE meeting, Venice (Italy) 6th June 2012

SLIDE 2

Acknowledgements

Lutz Harnisch
Margherita Bennetts
Tracy Higgins
Thomas Kerbusch
John C Lukas

2

SLIDE 3

The FEV1 biomarker in COPD

Chronic obstructive pulmonary disease

(COPD)

– Third leading cause of death in US and projected to increase world-wide due to projected to increase world wide due to smoking

Forced expiratory volume in one second

(FEV1)

– Important endpoint for diagnosis and the primary biomarker for dose selection in ph2b – Ph3 need to show reduced exacerbations

3

SLIDE 4

Example of how literature analysis aids internal development at Pfizer

Internal Data Literature Data FEV1 Model prediction of Drugs X/Y Dose- Response in FEV1 Model prediction of FEV1-treatment effects across published compounds FEV1 comparison to competitor drugs (MDI)

4

Exacerbation rate (ER) Model prediction of FEV1-ER relation across published compounds Predicted efficacy in ER across published compounds Model prediction of Drugs X/Y Dose- Response in ER ER comparison to competitor drugs (MDI)

SLIDE 5

Drug Class of interest in COPD

Long-acting bronchodilators (inhaled)

– Long-acting β2 agonists (LABA) L ti ti h li i (LAAC) – Long-acting anticholinergics (LAAC)

Anti-inflammatory therapy

– Inhaled corticosteroids (ICS) – PDE4-inhibitors (PDE4i)

5

SLIDE 6

Literature FEV1 in COPD

Change from baseline trough FEV1 (up to wk 26) by treatment class

ICS ICS LABA ICS LABA LAAC

0.0 0.1 0.2

LAAC

FB (L)

Absolute FEV1 is analysed, but displayed as

6

0.0 0.1 0.2 5 10 15 20 25

LABA LABA LAAC

5 10 15 20 25

PBO PDE4i

Time (weeks) Trough FEV1 CF

displayed as change from baseline, due to dominating variability in baseline!

SLIDE 7

Linear disease progression Disease modifying effects?

0.2
0.1

0.0

Pauwels (1999) Ref=637 Trial ID=43 N=1277 Post-BD

1 1 1 1 1 1 1 1 1 1 1 1 1

0.1

0.0 0.1

Tashkin (2008) NCT00144339 Ref=661 Trial ID=75 N=5993 Pre-BD

1 1 1 1 1 1 1 1 1 1

UPLIFT study

7

50 100 150

0.3

PBO Budesonide 800 UG/DAY 1

Time (weeks) CFB FEV1 (L)

50 100 150 200

0.2
PBO

Tiotropium 18 UG/DAY 1

Time (weeks) CFB FEV1 (L)

50 100 150

0.2
0.1

0.0 0.1

Celli (2008) NCT00268216, SCO30003 Ref=663 Trial ID=81 N=6112 Post-BD

PBO 1 1 1 1 1 1 1 Fluticasone/Salmeterol 1000/100 UG/DAY 1 2 2 2 2 2 2 2 Fluticasone 1000 UG/DAY 2 3 3 3 3 3 3 3 Salmeterol 100 UG/DAY 3

Time (weeks) CFB FEV1 (L)

50 100 150 200

0.3
0.2
0.1

0.0

UNK (2000) Ref=741 Trial ID=60 N=1116 Pre-BD

PBO 1 1 1 1 1 1 1 1 1 Triamcinolone 1200 UG/DAY 1

Time (weeks) CFB FEV1 (L)

48 / LIKE COPD 110621 l / f 1 19 / 04J 2012

y TORCH study Week CFB in FEV1(L)

SLIDE 8

Illustration of ISV in placebo effect at typical disease progression

1.22

Short study duration Long study duration

Pre-SABA/SAAC

Placebo

95% Conf. Int.

Legend

FEV1active = Baseline + Effplacebo + Progressiondisease

Graph illustrates gradual placebo

8

Week FEV1 (L)

1.18 1.19 1.20 1.21 1 2 3 4 5 6

Year

1.05 1.10 1.15 1.20 1 2 3 4

gradual placebo response, but mixture component allowed immediate placebo response as well

SLIDE 9

Placebo, Salmeterol (LABA) and Fluticasone (ICS)

1.40

Short study duration

1.4

Long study duration

Pre-SABA/SAAC

Placebo Salmeterol 50 mcg BID Fluticasone 250 mcg BID Legend

FEV1active = Baseline + Effplacebo + Progressiondisease + Effdrug

9

Week FEV1 (L)

1.20 1.25 1.30 1.35 1 2 3 4 5 6

Year

1.1 1.2 1.3 1 2 3 4

SLIDE 10

Salmeterol interaction with FEV1 post-SABA/SAAC

1.40

Short study duration

1.4

Long study duration

Pre-SABA/SAAC

Placebo Salmeterol 50 mcg BID Fluticasone 250 mcg BID

Post-SABA/SAAC

Legend

FEV1active = Baseline + Effplacebo + Progressiondisease + Effdrug

10

Week FEV1 (L)

1.20 1.25 1.30 1.35 1 2 3 4 5 6

Year

1.1 1.2 1.3 1 2 3 4

SLIDE 11

Covariates

Pre-specified

– baseline parameter, based on inclusion criteria:

Min/max disease severity:mild/moderate/severe/v. severe
Restricted medical history limits more severe patients
min/max #exacerbations in previous year (no, any#, ≥1,

≥2)

– Disease prog. proportional to (ipred) baselineuntreated

Selected at p<0.001:

– Baseline decline with Age – Baseline < 1.2 L: Linear decline in anti-inflammatory efficacy

11

SLIDE 12

η-diagnostics: Base model Baseline affects drug response

typical efficacy

1.2 1.4

ETA on AI efficacy

1.0 1.5 2.0

ETA on BD efficacy

12

Untreated study baseline (L) Ratio: ipred efficacy over

0.4 0.6 0.8 1.0 1.0 1.5 2.0

SLIDE 13

η-diagnostics: Base model Baseline affects drug response

typical efficacy

1.2 1.4

fluticasone trials budesonide trials roflumilast trials 599 648 637 165 663 668 684 714 577 308 261 ETA on AI efficacy

1.0 1.5 2.0

salmeterol trials formoterol trials tiotropium trials indacaterol trials 693 353 573 490 ETA on BD efficacy

13

Untreated study baseline (L) Ratio: ipred efficacy over

0.4 0.6 0.8 1.0 1.0 1.5 2.0

785 649 649 648 573 750 679 657 647 609 695 664 668 779 786 708 706 666 164 251 270 288 328 509 609 621 695 664 663 668 779684 708 706 714 577 308 709 353 605 639 610261 573 653 659 660 673 690 776 778 807 679 657 678 664 647 656 1 536 536 765 251 607 658767 686 683 288 455 488 492 493 494 621 653 659 660 662 673 690 695 762 676 661 664 668 779 687 647 656 667 694 807 798 790

SLIDE 14

ETA diagnostics: Final model

typical efficacy

1 05 1.10 1.15

fluticasone trials budesonide trials roflumilast trials 649 648 663 261 ETA on AI efficacy

1.0 1.5 2.0

salmeterol trials formoterol trials tiotropium trials indacaterol trials 693 684 714 353 639 573 653 807 536 765 490 494 653 676 807 ETA on BD efficacy

14

Untreated study baseline (L) Ratio: ipred efficacy over

0.85 0.90 0.95 1.00 1.05 1.0 1.5 2.0

599 785 649 648 573 637 750 679 657 647 165 609 695 664 668 779 684 786 708 706 714 666 577 308 164 261 251 270 288 328 509 609 621 695 664 663 668 779 708 706 714 577 308 709 605 610 261 659 660 673 690 776 778 807 679 657 678 664 647 656 1 536 765 251 607 658 767 686 683 288 455 488 492 493 621 659 660 662 673 690 695 762 676 661 664 668 779 687 647 656 667 694 807 798 790

SLIDE 15

Efficacy in moderate COPD

formoterol 9 mcg BID indacaterol 75 mcg QD SX 50mcg BID and Spiriva 18 mcg QD SXFP 50/250 mcg BID

86 ml 149 ml 162 ml 154 ml

Pre-SABA/SAAC

LABA + ICS same efficacy as the sum

f the two mono

components

15

Placebo-corrected FEV1 (L)

roflumilast 500 mcg QD budesonide 400/320 mcg BID Fluticasone 250 mcg BID Spiriva 18 mcg QD Respimat 5 mcg QD salmeterol 50 mcg BID

0.05 0.10 0.15

36 ml 61 ml 83 ml 92 ml 128 ml 116 ml

SLIDE 16

Efficacy in moderate COPD

formoterol 9 mcg BID indacaterol 75 mcg QD SX 50mcg BID and Spiriva 18 mcg QD SXFP 50/250 mcg BID

86 ml 149 ml 162 ml 154 ml

Pre-SABA/SAAC

LABA+LAAC interaction estimated as a relative reduction in LABA effect

16

Placebo-corrected FEV1 (L)

roflumilast 500 mcg QD budesonide 400/320 mcg BID Fluticasone 250 mcg BID Spiriva 18 mcg QD Respimat 5 mcg QD salmeterol 50 mcg BID

0.05 0.10 0.15

36 ml 61 ml 83 ml 92 ml 128 ml 116 ml

SLIDE 17

Efficacy in moderate COPD

95% CI excluding treatment heterogeneity

Indacaterol (QD) has superior trough efficacy in comparison to the two BID LABAs.

formoterol 9 mcg BID indacaterol 75 mcg QD SX 50mcg BID and Spiriva 18 mcg QD SXFP 50/250 mcg BID

86 ml 149 ml 162 ml 154 ml

95% CI: 131-167 ml 95% CI: 135-195 ml 95% CI: 136-171 ml 95% CI: 68-106 ml

Pre-SABA/SAAC

17

Several large studies indicating poor efficacy for budesonide. Due to protocol/study conduct

r compound?

Placebo-corrected FEV1 (L)

roflumilast 500 mcg QD budesonide 400/320 mcg BID Fluticasone 250 mcg BID Spiriva 18 mcg QD Respimat 5 mcg QD salmeterol 50 mcg BID

0.05 0.10 0.15

36 ml 61 ml 83 ml 92 ml 128 ml 116 ml

95% CI: 117-139 ml 95% CI: 101-131 ml 95% CI: 22-51 ml 95% CI: 48-72 ml 95% CI: 71-97 ml 95% CI: 80-106 ml

SLIDE 18

Efficacy in moderate COPD

95% CI including treatment heterogeneity (ISV Emax)

formoterol 9 mcg BID indacaterol 75 mcg QD SX 50mcg BID and Spiriva 18 mcg QD SXFP 50/250 mcg BID

86 ml 149 ml 162 ml 154 ml

95% CI: 57-123 ml 95% CI: 100-196 ml 95% CI: 112-224 m 95% CI: 115-187 ml

Pre-SABA/SAAC

18

Placebo-corrected FEV1 (L)

roflumilast 500 mcg QD budesonide 400/320 mcg BID Fluticasone 250 mcg BID Spiriva 18 mcg QD Respimat 5 mcg QD salmeterol 50 mcg BID

0.05 0.10 0.15 0.20

36 ml 61 ml 83 ml 92 ml 128 ml 116 ml

95% CI: 20-57 ml 95% CI: 38-77 ml 95% CI: 57-113 ml 95% CI: 65-127 ml 95% CI: 87-167 ml 95% CI: 78-154 ml

SLIDE 19

Efficacy with untreated baseline 1 L

95% CI including treatment heterogeneity

formoterol 9 mcg BID indacaterol 75 mcg QD SX 50mcg BID and Spiriva 18 mcg QD SXFP 50/250 mcg BID

82 ml 142 ml 156 ml 127 ml

95% CI: 53-119 ml 95% CI: 92-191 ml 95% CI: 108-212 m 95% CI: 92-161 ml

Pre-SABA/SAAC

19

Placebo-corrected FEV1 (L)

roflumilast 500 mcg QD budesonide 400/320 mcg BID Fluticasone 250 mcg BID Spiriva 18 mcg QD Respimat 5 mcg QD salmeterol 50 mcg BID

0.05 0.10 0.15 0.20

23 ml 39 ml 54 ml 88 ml 122 ml 110 ml

95% CI: 13-38 ml 95% CI: 23-54 ml 95% CI: 35-76 ml 95% CI: 60-121 ml 95% CI: 81-159 ml 95% CI: 74-148 ml

SLIDE 20

Predicted efficacy: Morning trough Pre- vs. Post-SABA/SAAC

LABA and LAAC efficacy when FEV1 is measured post- SABA/SAAC

Spiriva Respimat 5 mcg QD salmeterol 50 mcg BID formoterol 9 mcg BID indacaterol 75 mcg QD

48 ml 83 ml 52 ml 65 ml

Post-SABA/SAAC

In TORCH (ref 650) reports a disappointing 40

20

LABA and LAAC efficacy when FEV1 is measured pre- SABA/SAAC Placebo-corrected FEV1 (L)

Spiriva 18 mcg QD Respimat 5 mcg QD salmeterol 50 mcg BID formoterol 9 mcg BID indacaterol 75 mcg QD

0.05 0.10 0.15

86 ml 149 ml 92 ml 128 ml 116 ml

Pre-SABA/SAAC

Spiriva 18 mcg QD

72 ml

reports a disappointing 40 ml efficacy for salmeterol. Low efficacy due to measuring post SABA

SLIDE 21

Model Predicts LABA/LAAC Interaction

LABA with LAAC background and vice versa

indacaterol 75 mcg QD

95 ml

60% background

Spiriva 18 mcg QD Respimat 5 mcg QD salmeterol 50 mcg BID formoterol 9 mcg BID indacaterol 75 mcg QD

34 ml 59 ml 36 ml 72 ml 60 ml

100% background

UPLIFT (661) reports 87-103 ml efficacy for Spiriva. Canadian study

21

Placebo-corrected FEV1 (L)

Spiriva 18 mcg QD Respimat 5 mcg QD salmeterol 50 mcg BID formoterol 9 mcg BID indacaterol 75 mcg QD

0.05 0.10 0.15

86 ml 149 ml 92 ml 128 ml 116 ml

No DD interaction

Spiriva 18 mcg QD Respimat 5 mcg QD salmeterol 50 mcg BID formoterol 9 mcg BID 75 mcg QD

55 ml 59 ml 95 ml 83 ml

Canadian study (686) reports 80- 120 ml. 60% and 54% LABA background, respectively.

SLIDE 22

Discussion

Placebo effect has no typical direction

– Still important to account for: individual studies have marked placebo effect

Dropout due to lack of efficacy

p y

– Masks any disease progression – Reduce signs of disease modifying effects

Low study baseline
lower efficacy of anti-inflammatory drugs
Weak trend towards lower efficacy in

bronchodilators (LABA or LAAC)

Mixed message in literature on individual data

22

SLIDE 23

Conclusion

DD-interactions and covariates account for

the sometimes disappointing results on FEV1 in late-stage trials

TORCH UPLIFT & rolumilast exacerbation trials – TORCH, UPLIFT & rolumilast exacerbation trials – Except LABA/LAAC interaction (UPLIFT) these do not necessarily translate into lower effect on exacerbations!

Model provides efficacy bench mark for

published compounds, accounting for treatment heterogeneity (ISV Emax)

23