testing equivalence of orally inhaled products: study ... · testing equivalence of orally inhaled...
TRANSCRIPT
Testing Equivalence of Orally Inhaled Products:
Study Designs, Subject Selection, Clinical Study Model
Richard Ahrens
Professor of Pediatrics
University of Iowa
Inhaled Products Commonly Used For Treatment Of Asthma
l Short-acting beta agonists (SABAs; terbutaline, albuterol/salbutamol)n First US generics approved in mid-90s
n Precedent for other drug classes
l Long-acting beta agonist (LABAs; salmeterol, formoterol)
l Corticosteroid (ICS; beclomethasone, triamcinolone, flunisolide, budesonide, fluticasone, ciclesonide)
l Combination productsn Salmeterol/fluticasone (Advair)
n Formoterol/budesonide (Symbicort)
THE PROBLEM
Why Pharmacodynamic Studies
Are Needed.
ORAL ADMINISTRATION FOR SYSTEMIC ACTION
l Rate and extent of absorption into the systemic circulation
n Useful for assessment of BIOEQUIVALENCE
PO
Admin.
GI Tract
1st Pass Metabolism
Site(s) of Action
Systemic Circulation
Site(s) of AEsElimination
TOPICAL DRUG DELIVERY FOR ACTION IN THE LUNGS
l Assessment of drug reaching site of action in the
Lung needed
n PD response reflects BIOPHASE levels
Aerosol Deliver
Device
Dose to Patient
Dose to Oropharynx
Site[s] of Action
(Biophase)
Systemic
Circulation
Site(s) of AEs
Elimination1st Pass
MetabolismGI Tract
Dose to
Lung
Other Pulmonary
Sites
(PK reflects dose to
BIOPHASE?)
Study Design vs. Level of Certainty Test & Ref Deliver Equivalent Amount of Drug to Biophase
Clinical Trial
Parallel Design;
One Dose Test
and Ref
Dose Response
Parallel Design;
Two Doses Test
and Ref
“Dose Axis”
Analysis
Dose Response
Cross-Over;
One Dose Test
and Ref
“Dose Axis”
Analysis
Relative Potency
(Bioassay)
Cross-Over;
One Dose Test
and Ref
“Response Axis”
Analysis
LOW
HIGH
Level of Certainty?
STUDY DESIGN:
1) Active Control Clinical Trial:
� Single dose level of both test and reference.
� Clinically relevant out come reflecting efficacy.
� Compare responses (to equal doses)
LOW LEVEL OF CERTAINTY
OF EQUIVALENT DRUG DELIVERY
Milanowski et al, Respir Med 1999
EFFICACY OF CFC & HFA
BECLOMETHASONE MDI'S
0
100
200
300
400
500
600
0 100 200 300 400 500 600
BDP Dose (mcg/d)
Mo
rnin
g P
ea
k F
low
HFA (Baker Norton)
CFC
Mean and 90% C.I.
Not significantNot significant
N=57 N=56N=54 N=54
Pulmonary deposition of HFA ~2X CFC
Need for Steep Dose-Response
0
10
0 1 2 3
DOSE (Actuations)
RE
SP
ON
SE
Curve B
Curve A
B
A
STUDY DESIGN:
2) Dose Response to Demonstrate Sensitivity:
� Two dose levels of both test and reference.
� Relevant out come reflecting efficacy.
� Parallel Design
� Sensitivity => Significant Difference
Between High and Low Dose Levels
� Compare magnitude of response.
HIGHER LEVEL OF CERTAINTY
OF EQUIVALENT DRUG DELIVERY
Potency of HFA-Beclomethasone (QVAR, 3M) Relative of CFC Beclomethasone (Beclovent, GSK)
12
14
16
18
20
22
24
26
100 1000
Dose (mcg)
Me
an
Ch
an
ge
in
FE
V1
HFA-BDP (QVAR)
CFC-BDP(Beclovent)
400 800200100
1 mcg HFA = 2.6 CFC
(95% C.I. 1.1 – 11.6)
Busse et al, 1998
STUDY DESIGN:
3) Dose Response to Demonstrate Sensitivity:
� Two dose levels of both test and reference.
� Relevant out come reflecting efficacy.
� Cross-Over Design
� Sensitivity => Significant Difference
Between High and Low Dose Levels
� Compare responses (to equal doses)
Greater Statistical Power
STUDY DESIGN:
4) BIOASSAY: Estimation of Relative Potency:
� Two dose levels of both test and reference.
� Relevant out come reflecting efficacy.
� Cross-Over Design
� Sensitivity => Significant Difference
Between High and Low Dose Levels
� Compare DOSE needed (for RESPONSE)
HIGHEST LEVEL OF CERTAINTY
OVERALL STUDY DESIGN:2 by 2 Bioassay
l VALIDITY TESTS
n Highly Significant
DRC
n Parallelism Contrast
n Preparations
Contrast
DOSE
RE
SP
ON
SE
STANDARD
TEST
Potency of Norton MDI Relative to Ventolin
Bioequivalence Criteria: 0.67 – 1.50
1 4
32
16
8
4
DOSE OF ALBUTEROL
(Actuations)
PC
20F
EV
1
Norton
MDI
Ventolin
MDI
1 puff Norton = 0.95 puff Ventolin
(90% C.I. 0.69 - 1.40)
Subject Selection & Clinical Study Model:
To Maximize The Steepness
of the Dose-Response:
� Subjects who require higher doses
to achieve response
� Choose outcome measure
with sufficiently steep dose-response
� Choose best clinical circumstance under which to SEE this dose-response
FAQ
If it’s so difficult to show a dose-response relationship, doesn’t that mean that dose doesn’t really matter?
Need for Steep Dose-Response
0
10
0 1 2 3
DOSE (Actuations)
RE
SP
ON
SE
Curve B
Curve A
B
A
Mean FEV1 Response to Albuterol:Daytime versus Nocturnal Awakening
1
1.5
2
2.5
3
3.5
0 20 40 60 80 100
TIME (Min)
FE
V1
(L
)
DAY
NIGHT
0
1 2 4 8 16
MDI ACTUATIONS
Median Dose to Achieve >80% Personal Best FEV1
0
2
4
6
8
10
12
14
16
18
20
22
0.2
DO
SE
(A
ctu
ati
on
s)
DAY NIGHT
TREATMENT PERIOD
SUMMARY
l Demonstration of “Sensitivity”
n => Significant Dose-Response
n Essential
l Bioassay Study with Estimation of
Relative Potency Yields the Highest
Level of Certainty of Bioequivalence.
SABALABA
ICS
STRATEGY DEPENDS ON DRUG CLASS
BIOEQUIVALENCE OF SABAs
Inhaled Beta Agonists:A Relatively Mature Field
To Ensure a Sufficiently Steep Dose-Response is Present:
• Protection against bronchoprovocation (methacholine, histamine)
• Bronchodilation under CAREFULLY controlled clinical circumstances
BIOEQUIVALENCE OF LABAs
LONG-ACTINGBETA AGONISTS
l With minor modifications, methods for
BE of inhaled albuterol should apply
LABA BE Studies Must Account For Differences Between LABAs & Albuterol
Onset
of Effect
15 – 30 Mininutes
1 – 4 Hours
Duration
of Action
2 – 6 Hours
> 12 Hours
Tolerance with Chronic Admin.
Minimal Profound(bronchprotection)
ALBUTEROL LABA
BIOEQUIVALENCE OF ICS
Less Mature Field
This concept has not been successfully applied to inhaled corticosteroids
“… it is difficult to draw firm conclusions about the comparative efficacy of different inhaled corticosteroids.”
“ …partially explained by differences between the designs of studies, the flat dose-response relationship for inhaled corticosteroids, the differences between inhalers, and the lack of control over important confounding factors in many studies.”
“Additional well-designed comparisons, perhaps examining different outcome parameters, are required before any definite conclusionscan be made.”
Barnes et al. AM J RESPIR CRIT CARE MED 157;S11, 1998
FAQ
If it’s so difficult to show a dose-response relationship, doesn’t that mean that dose doesn’t really matter?
An Asthma Clinician’s Paradox
l Clinical studies with hundreds of
patients show ICS dose-responses that
are flat to non-existent.
l Asthma clinicians think they see ICS
dose-response every day in individual
patients.
Which is illusion, which is reality?
Typical ICS Study DesignParallel Treatment Groups 1-12 Months of Treatment
(a) high variability (S)(b) shallow sloped DRC (B)(c) carry-over (prevents cross-over)
Baseline
PFTs, Sx, etc.
Improvement in PFTs, Sx,
PEFR
ICS
Statistical Power Associated with a Bioassay Study
Related to:
l Variability of responses (S)
l Steepness of dose-response slope (B)
l S + B do not function independently, but in concert as the ratio of S/B
l The smaller S/B, the more powerful the study
Relationship Between S/B and Computed Sample Size
0
1000
2000
3000
4000
5000
6000
7000
0 0.5 1 1.5 2 2.5 3
S/B Ratio
Co
mp
ute
d S
am
ple
Siz
e
2 by 2 Study Design
Typical Study??
Busse Study (HFA-BDP, QVAR)
Asthma Stability Following Prednisone “Burst”
1. 100 and 800 mcg/d HFA-BDP2. Cross-over
3. Multiple variables examined (58)4. Estimate S/B for each5. The lower S/B, the better
Prednisone “burst”
Baseline PFTs, Sx,
etc.
ICS Deterioration in PFTs, Sx,
etc. ??
Relationship Between S/B and Computed Sample Size
0
50
100
150
200
0 0.5 1 1.5
S/B Ratio
Co
mp
ute
d S
am
ple
Siz
e
2 by 2 Study Design
AM FEF 25-75
AM PEFR
PM FEV1
AM FEV1
PC20 FEV1 (Methacholine)
Relationship Between Crossover and Parallel Study Designs
0
500
1000
1500
2000
2500
0 0.5 1 1.5 2 2.5
S/B Ratio
Co
mp
ute
d S
am
ple
Siz
e Parallel Design
AM FEF25-75
Busse Study (HFA-BDP, QVAR)
Crossover Design
PC20 FEV1
AM PEFR
RUN-IN PERIOD
� 4 weeks LOW-DOSE ICS
(CFC-BDP 100 mcg/d)
� 2 weeks high-dose ICS
(Budesonide DPI 1600 mcg/d)
� ENTRANCE CRITERIA:
>7% predicted increase between HIGH and LOW-DOSE ICS
Asthma Stability Following High-Dose Corticosteroid “Burst”
QVAR versus Flovent Diskus
2.55
2.6
2.65
2.7
2.75
2.8
2.85
2.9
2.95
0 100 200 300 400
Dose (mcg/day)
FE
V1
(L
ite
rs)
HFA-BDP
FP-Diskus
p-value
Regress. 0.01
Parallel 0.16
1 mcg QVAR = 0.84 mcg FP
(90% CI : 0.35 – 1.74)
Asthma Stability Following High-Dose Corticosteroid “Burst”
RUN-IN PERIOD
Approx. 20% of subjects
met this entrance criteria
Lessons LearnedFrom ICS Experience To Date
l “Traditional” model insensitive to dose.
l Dose DOES matter to at least some patients.
l Therefore, need a model that CAN detect
differences in delivered dose.
Lessons LearnedFrom ICS Experience To Date
The solution? -- Control carry-over
-- Do cross-over study
-- Sensitive outcome measure
(at least one available)
-- Select subjects most likely to show a dose-response
l Number of subjects will depend on
n Design, outcome, subjects selected
n Width for BE interval (e.g. 67-150% vs 80-125%)
The Solution?Other Outcomes?
� Exhaled nitric oxide
� Adenosine challenge
� Sputum Eosinophilia
Lessons LearnedFrom ICS Experience To Date
l STUDY MODEL MUST:
n Use a clinically relevant outcome (reflect delivery to site of clinical action)
n Sufficiently steep dose-response
=> Sufficient statistical power.
APPLYING THESE METHODS TO COMBINATION DPIs
FACTORS ADDING TO COMPLEXITY OF EVALUATING COMBINATION DPIs
l Establishing PHARMACEUTICAL EQUIVALENCE is likely to be more difficult than:n Single agent inhalers
n DPIs vs. MDIs
l Need to assess SYSTEMIC EXPOSURE for both components (r.e. potential for adverse effects)n PK evaluations
l Assessment of IN VIVO BE for LABA and CS n Different outcomes and study designs
n => 2 separate BE studies
THANK YOU FOR YOUR ATTENTION!!
??? QUESTIONS ???