1.

Introduction
2. Current status of regulatory
guidance
3. Interpreting in vitro
equivalence data
4. Interpretation of clinical
bioequivalence studies
5. Integrating guidance with
experience
6. Expert opinion
Review
Establishing bioequivalence for
inhaled drugs; weighing the
evidence
Peter T Daley-Yates

& David A Parkins

GlaxoSmithKline, Stockley Park, Middlesex, UK

Introduction: During drug development and product life-cycle management,
it may be necessary to establish bioequivalence between two pharmaceutical
products. Methodologies to determine bioequivalence are well established
for oral, systemically acting formulations. However, for inhaled drugs, there
is currently no universally adopted methodology, and regulatory guidance
in this area has been subject to debate.
Areas covered: This paper covers the current status of regulatory guidance on
establishing the bioequivalence of topically acting, orally inhaled drugs, the
value and limitations of in vitro and in vivo bioequivalence testing, and the
practical issues associated with various approaches. The reader will gain an
understanding of the issues pertaining to bioequivalence testing of orally
inhaled drugs, and the current status of regulatory approaches to establishing
bioequivalence in different regions.
Expert opinion: Establishing bioequivalence of inhaled drug products involves
a multistep process; however, methodologies for each step have yet to be
fully validated. Our lack of understanding about the relationship between
in vitro, in vivo and clinical data suggests that in most cases, unless there is a
high degree of pharmaceutical equivalence between the test and reference
products, consideration of a combination of preclinical and clinical data may
be preferable to abridged approaches relying on in vitro data alone.
Keywords: bioequivalence, in vitro, inhaled medicines, pharmacodynamics, pharmacokinetics
Expert Opin. Drug Deliv. (2011) 8(10):1297-1308
1. Introduction
During the lifetime of a drug product, a manufacturer may wish to modify their ini-
tial registered formulation and/or inhaler device. For example, this can result from
changes to the manufacturing process, modification of an existing device or switch-
ing to an alternative inhaler device. In addition, new products may also be devel-
oped as alternatives to innovator products. Under these latter circumstances,
where products may not be pharmaceutically equivalent, the robust demonstration
of product equivalence is essential to ensure patients continue to benefit from
products with the same safety and efficacy profile.
In the case of oral, solid dose forms, patients who receive alternative but bioequiva-
lent formulations are unlikely to perceive any difference between the treatments. The
packaging and the colour of the tablet may differ, but the patients interaction with
the product will remain essentially unchanged and will not impact the safe and effective
use of the product. In contrast, for orally inhaled drugs, a patients experience may be
different, particularly in terms of the patients interaction with the inhaler. Even
when an alternative device meets all the required in vitro bioequivalence criteria, the
patient interface with the device and its optimal use could be altered by changes inexter-
nal design features, instructions for use, airflow resistance and something as simple as
the feel in the mouth or throat during inhalation.
10.1517/17425247.2011.592827 2011 Informa UK, Ltd. ISSN 1742-5247 1297
All rights reserved: reproduction in whole or in part not permitted
E
x
p
e
r
t

O
p
i
n
.

D
r
u
g

D
e
l
i
v
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

i
n
f
o
r
m
a
h
e
a
l
t
h
c
a
r
e
.
c
o
m

b
y

N
o
v
a
r
t
i
s

P
h
a
r
m
a

o
n

0
2
/
0
3
/
1
2
F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
The use of pharmacokinetic data obtained in healthy
volunteers is established as the primary means of providing
clinical data that support claims of bioequivalence for system-
ically acting drugs when administered orally or parentally.
The theoretical basis for this is the assumption that the drug
concentration in the systemic circulation is in equilibrium
with the concentration at its site of action. Where this is a
valid assumption, bioequivalence testing can rely on three
steps, comprising: i) qualitative and quantitative sameness
of the active pharmaceutical ingredient and excipients,
ii) in vitro dissolution testing and iii) a human pharmacoki-
netic study (Figure 1). However, this simplistic approach is
not as appropriate for inhaled drugs, as a drugs concentration
in the systemic circulation does not necessarily reflect the
drugs concentration at its (topical) site(s) of action in the
lung. Uncertainty about the relationship between the dose
delivered to the site of action in the lung, topical efficacy
and systemic drug concentrations serves as an obstacle
to relying solely on pharmacokinetic data to assess the
bioequivalence of topically acting orally inhaled drugs. There-
fore, other sources of data must be considered and establish-
ing bioequivalence may take as many as five steps where
data may be required comprising: i) qualitative and quantita-
tive sameness of the active pharmaceutical ingredient and
excipients, ii) device similarity to ensure the product perfor-
mance and the patient device interaction is unchanged,
iii) in vitro device performance testing including emitted
fine particle mass (FPM) dose and particle-size profiling,
iv) in vivo product performance including lung deposition
and systemic pharmacokinetic data and v) confirmation of
equivalent topical efficacy (Figure 1).
The challenge faced in establishing the bioequivalence of
inhaled therapies was first highlighted following the Montreal
Protocol on the use of chlorofluorocarbons (CFCs) as propellants
in metered dose inhalers (MDIs) [1]. It was noted that, despite the
patent for the b
2
-receptor agonist salbutamol (or albuterol as it is
named in some countries) having expired in 1989, the first
generic MDI preparation of salbutamol was not approved by
the US FDA until late 1995 [2]. The main reason for the 6-year
time period between patent expiry and generic approval was the
lack of an acceptable and valid method for establishing in vivo
bioequivalence of the generic salbutamol inhaler. Among the var-
ious methodological approaches, the FDA did not consider the
use of urine and plasma pharmacokinetic data alone [3-5] or
lung deposition data from gamma scintigraphy methods to be
reliable as a means of confirming the relative quantity of drug
delivered to the site of action in the lung. The consequence of
the FDAs concern over industry responding promptly to the
requirements for the Montreal Protocol was the publishing of
their specific guidelines for establishing bioequivalence [6,7]. Reg-
ulatory agencies have since had time to consider the importance
of the issues raised by this problembut have yet to agree a globally
accepted approach on establishing the bioequivalence for inhaled
drugs. Regulatory agencies have consulted with various interested
parties over the type of data they felt would be necessary to deter-
mine the bioequivalence of inhaled drugs. Some guidance has
been provided either in draft or final forms. In 2007, the Euro-
pean Medicines Agency (EMA) released a consultation docu-
ment describing a proposed approach for the determination of
bioequivalence [8]. At the end of the consultation period, a step-
wise method outlining the requirements for clinical documenta-
tion for orally inhaled products was published in which consider-
ation was then given to the use of in vitro data where predefined
criteria for pharmaceutical equivalence have to be met [9]. Other
agencies have placed emphasis on the collection of different types
of data, for example, Health Canada focused more on pharmaco-
dynamic end points for topical efficacy in their draft guidance
published in 2007 [10].
Determination of bioequivalence of inhaled drugs remains
an active area of debate and further discussions were held
at the Product Quality Research Institute Workshop in
2009 [11], at the Respiratory Drug Delivery Conference
2010 [12,13] and at the International Society for Aerosols in
Article highlights.
. Establishing the bioequivalence of inhaled drugs is a
multistep process with little experimental evidence to
validate the use of data from one of the various in vivo
and in vitro methods over those from another.
. This lack of clarity is one factor contributing to the
absence of a consistent approach across regulatory
authorities to establishing bioequivalence.
. In vitro assessments of emitted doses are not universally
predictive of pharmacokinetic and pharmacodynamic
outcomes in clinical studies, pharmacokinetic data do
not always correlate well with in vitro or clinical efficacy
data, and pharmacodynamic data lack the precision of
pharmacokinetic data when used to
determine bioequivalence.
. It is clear that more data and better models are required
to describe the relationship between in vitro and in vivo
observations.
.
A weight-of-evidence approach to new formulations
that can be applied on a case-by-case basis rather than
a rigid regulatory approach may be more acceptable.
This approach places greater responsibility on drug
development scientists to determine the level of
evidence necessary to establish bioequivalence.
.
Our lack of understanding about the relationship
between in vitro, in vivo and clinical data suggests that,
in most cases, unless there is a high degree of
pharmaceutical equivalence between the test and
reference products, consideration of a combination of
preclinical and clinical data may be preferable to
abridged approaches relying on in vitro data alone.
. There are several unresolved issues preventing the
agreement of general guidelines on the best approach
to establishing bioequivalence for all classes of inhaled
drug products. A way forward could involve the
development of product- or drug class-specific
guidelines allowing some of the outstanding issues to
be circumvented.
This box summarises key points contained in the article.
Bioequivalence of inhaled drugs
1298 Expert Opin. Drug Deliv. (2011) 8(10)
E
x
p
e
r
t

O
p
i
n
.

D
r
u
g

D
e
l
i
v
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

i
n
f
o
r
m
a
h
e
a
l
t
h
c
a
r
e
.
c
o
m

b
y

N
o
v
a
r
t
i
s

P
h
a
r
m
a

o
n

0
2
/
0
3
/
1
2
F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
Medicine and International Pharmaceutical Aerosol Consor-
tium on Regulation and Science conference in October
2010 [14]. Although several years of consultation have
occurred, the FDA has yet to publish any general guidance
on demonstrating the bioequivalence of orally inhaled
products. In contrast to other regulatory agencies, the FDA
currently recommends a weight-of-evidence approach [15]
and are also sponsoring studies looking at the potential utility
of surrogate markers of efficacy.
2. Current status of regulatory guidance
Thus far, the only published final guidance has been from the
EMA, Health Canada has only published draft guidance for
the purposes of consultation, the FDA has yet to publish
any general guidelines and, in other regions of the world,
guidelines are also being considered. The EMA guidance has
not only been taken up in Europe but has also been adopted
by other regions such as Australia. After consulting on the
revision of its guidance concerning the development of orally
inhaled drugs in 2007, the EMA adopted its revised guideline
in 2009 [9]. The revised guidance can be considered in con-
junction with the draft EMA regulatory guideline on bio-
equivalence [16]. The EMA advocates a step-wise approach
to the investigation of bioequivalence between test and refer-
ence products: step 1 involves in vitro comparison of formula-
tions (only acceptable if criteria for formulation and device
equivalence have been met); step 2 comprises the comparison
of formulations using lung deposition models (pharma-
cokinetics and scintigraphy); and step 3 involves the use of
pharmacodynamic and clinical efficacy data. Importantly,
the demonstration of bioequivalence at step 1 or step 2 pre-
cludes the need for further comparisons. A schematic of this
approach is shown in Figure 2.
The guideline also addresses the problem of different
requirements depending on the type of inhalation device
under investigation [9]. For dry powder inhalers (DPIs),
consideration of device resistance is required and sufficient
in vitro data must be included in the submitted dossier to
describe the flow deposition characteristics of the products
within the range of clinically relevant pressure drops/flow
limits. For pressurised MDIs, the possible impact of any
new propellant or excipient on clinical efficacy or safety
should be studied and appropriate data should be provided
to support the use of a named spacer device. In vitro character-
isation of the drug product should be conducted prior to any
therapeutic bioequivalence studies for all inhaler types.
Provided that requirements for formulation and device
similarity are met, the EMA guideline allows the applicant
to submit an abridged application that contains only
comparative in vitro data to substantiate a claim of therapeutic
bioequivalence with a reference product. If the in vitro
data satisfy specific criteria, no additional data from clinical
pharmacokinetic or efficacy studies need be provided.
A summary of the criteria from the guideline is presented
in Box 1. The interpretation of this aspect of the guidance
has been the subject of continuing debate [14,17] (see Table 1
and Section 4).
If the in vitro data alone do not support a claim of bio-
equivalence, the next step in terms of the EMA guidelines is
to conduct a pharmacokinetic study. The recommended
design for pharmacokinetic assessment of orally inhaled drugs
by the EMA is a four-period, crossover study where test and
reference drugs are each administered in the presence and
absence of a charcoal block (except where very low oral bio-
availability precludes the need for the assessment of gastroin-
testinal drug absorption). This appears to be a scientifically
sound approach as inhaled administration includes a swal-
lowed fraction that will make some unknown contribution
to systemic pharmacokinetic profiles. However, the method-
ology may need to be specifically validated for each drug
product. Although the EMA guideline states that pharmaco-
kinetic studies should be conducted in the intended patient
population, bioequivalence studies are often conducted in
healthy subjects. The choice of subjects is frequently based
on the greater and less variable systemic exposure data often
API +
Excipients
Dissolution
testing
Human PK
study
API +
Excipients,
propellant,
carrier
Device
design
similarity
Emitted dose
particle size
testing
Patient
device
interface
Human
lung
deposition
Human
PK
study
Human
efficacy and
PD
Figure 1. Comparative pathways for establishing bioequivalence in oral and inhaled medicines.
Daley-Yates & Parkins
Expert Opin. Drug Deliv. (2011) 8(10) 1299
E
x
p
e
r
t

O
p
i
n
.

D
r
u
g

D
e
l
i
v
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

i
n
f
o
r
m
a
h
e
a
l
t
h
c
a
r
e
.
c
o
m

b
y

N
o
v
a
r
t
i
s

P
h
a
r
m
a

o
n

0
2
/
0
3
/
1
2
F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
seen in healthy volunteers compared with patients with
lung disease [18], but having data in patients is invaluable
in understanding the product performance in the target pop-
ulation. A summary of the key regulatory requirements for
pharmacokinetic studies with orally inhaled products are
presented in Box 2. At present, the question remains as to
whether pharmacokinetic data are predictive of topical effi-
cacy either directly or via pharmacodynamic end points, or
more fundamentally, whether the drug concentration mea-
sured in plasma is a marker for topical efficacy at the site of
action in the lung.
Pharmacodynamic assessment of test and reference prod-
ucts is the next step when bioequivalence has not been dem-
onstrated with in vitro or pharmacokinetic data. The EMA
provides guidance on appropriate pharmacodynamic methods
to determine therapeutic bioequivalence. For bronchodilator
and anti-inflammatory compounds, two types of studies
appear to be acceptable to the EMA: studies of bronchodilata-
tion/improved airway function and studies of broncho-
protection. The primary outcome variables are forced
expiratory volume in 1 s (FEV1), such as the measurement
of bronchodilatation over at least 80% of the duration of
Step 1
Are the formulations and
devices similar?
Equivalence
established
Step 2
in vitro similarity?
Equivalence not
established
Step 3
similar lung deposition
profile?
Equivalence not
established
Step 4a
similar
pharmacodynamic
profile?
Step 4b
similar phase III
findings?
No
No
Equivalence not
established
Similar safety?
Similar systemic
profile?
Equivalence
Established
No
Yes
Yes
Yes
N
o
Y
e
s
Figure 2. Schematic of the stepwise approach for establishing bioequivalence advocated by the European Medicines Agency.
Bioequivalence of inhaled drugs
1300 Expert Opin. Drug Deliv. (2011) 8(10)
E
x
p
e
r
t

O
p
i
n
.

D
r
u
g

D
e
l
i
v
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

i
n
f
o
r
m
a
h
e
a
l
t
h
c
a
r
e
.
c
o
m

b
y

N
o
v
a
r
t
i
s

P
h
a
r
m
a

o
n

0
2
/
0
3
/
1
2
F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
action after a single inhalation (FEV1 AUC), change in
FEV1 at appropriate time points or the provocative con-
centration (or dose) that produces a 20% fall in FEV1
(PC20 FEV1). High variability in pharmacodynamic end
points reported within the literature questions the reliability
and repeatability of measures of orally inhaled drugs in a
clinical study population through this type of study. At
least two doses on the steep part of the dose--response curve
need to be studied. The demonstration of equivalence
in anti-inflammatory efficacy is difficult as it requires the
demonstration of a dose--response relationship where a flat
dose--response curve is often seen.
Health Canada released proposals on the assessment of bio-
equivalence for inhaled corticosteroids [10], but these are less
detailed than the EMA guideline. In the Health Canada guide-
line, in addition to comparative in vitro product testing, second
entry medicines must be assessed for bioequivalence in terms of
clinical efficacy criteria. The recommended co-primary end
points are sputum eosinophil count and FEV1. The draft
guideline also states that sponsors should characterise systemic
exposure profiles in terms of area under the plasma concentra-
tion--time curve (AUC) and maximum plasma concentration
(C
max
). In contrast to the EMA and Health Canada, applica-
tions to the FDA are viewed on a case-by-case weight of evi-
dence basis (Box 3). Although specific guidelines have not
been published, it is assumed that comparisons between test
and reference will need to provide data on all aspects of product
performance. Hence, although the emphasis on in vivo and
in vitro may differ in each case, the principles underlying the
interpretation of the data and hurdles for approval are likely
to be similar for the FDA and EMA.
3. Interpreting in vitro equivalence data
Characterising the in vitro particle-size distribution of inhaled
therapies using cascade impaction has become a standard
quality control approach to describing the aerodynamic
particle-size distribution (APSD) performance of these prod-
ucts. This technique relies on separating the emitted aerosol
into a series of size ranges based on their aerodynamic diame-
ter. In the case of the Andersen Cascade Impactor operated at
28 l/min, the larger (> 9 m) particles are trapped in the
induction port with the subsequent stages covering the aero-
dynamic particle-size range from 9 (stage 0) to 0.4 m (stage
7). The relevant size range for lung deposition is typically
expected to be in the range of 1 -- 5 m and, for MDIs, group-
ing of data for stages 3 -- 5 is frequently described as the FPM.
There is an implied assumption that the particle-size profile,
particularly the FPM of an inhaled therapy, is predictive of
the in vivo lung deposition profile. Although this is not
universally true, larger particles tend to be deposited in the
mouth, throat and upper airways and smaller particles are
likely to achieve greater lung penetration.
The ability to correlate clinical lung deposition and pene-
tration with in vitro physical characteristics of an inhaled
drug product is compelling as it would greatly simplify prod-
uct evaluation. However, this assumption is more supported
by theoretical than by empirical data. There is also a lack of
data to support the definition of acceptable thresholds for
demonstrating similarity between particle-size profiles. In
this regard, the current EMA guidance to a large extent leaves
the applicant to justify the testing of a product and the level
of acceptance.
Many of the deficiencies encountered when attempts to
correlate in vitro results with lung deposition data are inherent
to the currently available methodologies. For example, differ-
ences in relative humidity can mean that the delivered
particle-size distribution for a product can differ due to
hygroscopicity or evaporation effects.
Cascade impactors generally use a metal induction port
that does not replicate deposition that is observed in the
oropharyngeal region of the respiratory tract. Furthermore,
Box 1. Regulatory criteria leading to the acceptance of in vitro data alone as proof of bioequivalence to a
reference medicinal product (European Medicines Agency 2009).
The product contains the same active substance (i.e., same salt, ester, hydrate or solvate, etc.)
The pharmaceutical dosage form is identical (e.g., pressurised MDI, nonpressurised MDI, DPI, etc.)
The active substance is in the solid state (powder, suspension): any differences in crystalline structure and/or polymorphic form
should not influence the dissolution characteristics, the performance of the product or the aerosol particle behaviour
Any qualitative and/or quantitative differences in excipients should not influence the performance of the product (e.g., delivered
dose uniformity, etc.), aerosol particle behaviour (e.g., hygroscopic effect, plume dynamic and geometry) and/or be likely to
affect the inhalation behaviour of the patient (e.g., particle-size distribution affecting mouth/throat feel or cold Freon effect)
Any qualitative and/or quantitative differences in excipients should not change the safety profile of the product
The inhaled volume through the device to enable a sufficient amount of active substance into the lungs should be similar
(within 15%)
Handling of the inhalation devices for the test and the reference products in order to release the required amount of the active
substance should be similar
The inhalation device has the same resistance to airflow (within 15%)
The target delivered dose should be similar (within 15%)
DPI: Dry powder inhaler; MDI: Metered dose inhaler.
Daley-Yates & Parkins
Expert Opin. Drug Deliv. (2011) 8(10) 1301
E
x
p
e
r
t

O
p
i
n
.

D
r
u
g

D
e
l
i
v
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

i
n
f
o
r
m
a
h
e
a
l
t
h
c
a
r
e
.
c
o
m

b
y

N
o
v
a
r
t
i
s

P
h
a
r
m
a

o
n

0
2
/
0
3
/
1
2
F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
T
a
b
l
e
1
.
R
e
l
a
t
i
o
n
s
h
i
p
s
b
e
t
w
e
e
n
p
h
a
r
m
a
c
o
k
i
n
e
t
i
c
d
a
t
a
a
n
d
i
n
v
i
t
r
o
p
a
r
t
i
c
l
e
-
s
i
z
e
p
r
o
f
i
l
e
s
a
n
d
c
l
i
n
i
c
a
l
e
f
f
i
c
a
c
y
f
i
n
d
i
n
g
s
.
E
x
a
m
p
l
e
I
n
v
i
t
r
o
P
h
a
r
m
a
c
o
k
i
n
e
t
i
c
s
E
f
f
i
c
a
c
y
D
e
s
i
g
n
B
u
d
e
s
o
n
i
d
e
T
u
r
b
u
h
a
l
e
r

v
s
B
u
d
e
s
o
n
i
d
e
F
l
e
x
h
a
l
e
r

1
,
2
G
o
o
d
m
a
t
c
h
E
q
u
i
v
a
l
e
n
t
(
n
o
c
h
a
r
c
o
a
l
b
l
o
c
k
)
N
o
t
e
q
u
i
v
a
l
e
n
t
O
p
e
n
-
l
a
b
e
l
,
r
a
n
d
o
m
i
s
e
d
,
a
c
t
i
v
e
-
c
o
n
t
r
o
l
,
c
r
o
s
s
o
v
e
r
,
s
i
n
g
l
e
d
o
s
e
;
n
=
3
7
D
o
u
b
l
e
-
b
l
i
n
d
,
r
a
n
d
o
m
i
s
e
d
,
p
l
a
c
e
b
o
-
c
o
n
t
r
o
l
l
e
d
,
p
a
r
a
l
l
e
l
g
r
o
u
p
,
1
2
-
w
e
e
k
;
n
=
6
2
1
D
o
u
b
l
e
-
b
l
i
n
d
,
r
a
n
d
o
m
i
s
e
d
,
p
l
a
c
e
b
o
-
c
o
n
t
r
o
l
l
e
d
,
p
a
r
a
l
l
e
l
g
r
o
u
p
,
1
2
-
w
e
e
k
;
n
=
5
1
6
B
u
d
e
s
o
n
i
d
e
T
u
r
b
u
h
a
l
e
r

v
s
B
u
d
e
s
o
n
i
d
e
E
a
s
y
h
a
l
e
r

3
,
4
S
m
a
l
l
d
i
f
f
e
r
e
n
c
e
s
E
q
u
i
v
a
l
e
n
t
(
c
h
a
r
c
o
a
l
b
l
o
c
k
)
E
q
u
i
v
a
l
e
n
t
O
p
e
n
-
l
a
b
e
l
,
r
a
n
d
o
m
i
s
e
d
,
a
c
t
i
v
e
-
c
o
n
t
r
o
l
,
c
r
o
s
s
o
v
e
r
,
s
i
n
g
l
e
d
o
s
e
;
n
=
3
3
D
o
u
b
l
e
-
b
l
i
n
d
,
r
a
n
d
o
m
i
s
e
d
,
a
c
t
i
v
e
c
o
n
t
r
o
l
,
p
a
r
a
l
l
e
l
g
r
o
u
p
,
1
2
-
w
e
e
k
;
n
=
1
6
1
B
D
P
I
n
n
o
v
a
t
o
r
C
F
C
M
D
I
v
s
B
D
P
G
e
n
e
r
i
c
C
F
C
M
D
I
5
G
o
o
d
m
a
t
c
h
E
q
u
i
v
a
l
e
n
t
(
n
o
c
h
a
r
c
o
a
l
b
l
o
c
k
)
E
q
u
i
v
a
l
e
n
t
D
o
u
b
l
e
-
b
l
i
n
d
,
r
a
n
d
o
m
i
s
e
d
,
c
r
o
s
s
o
v
e
r
,
s
i
n
g
l
e
d
o
s
e
;
n
=
5
1
F
P
/
S
a
l
m
e
t
e
r
o
l
I
n
n
o
v
a
t
o
r
p
M
D
I
v
s
F
P
/
S
a
l
m
e
t
e
r
o
l
G
e
n
e
r
i
c
p
M
D
I
6
S
i
m
i
l
a
r
E
q
u
i
v
a
l
e
n
t
f
o
r
F
P
;
n
o
t
e
q
u
i
v
a
l
e
n
t
f
o
r
s
a
l
m
e
t
e
r
o
l
N
D
O
p
e
n
,
r
a
n
d
o
m
i
s
e
d
,
c
r
o
s
s
o
v
e
r
,
s
i
n
g
l
e
d
o
s
e
(
c
h
a
r
c
o
a
l
b
l
o
c
k
)
;
n
=
3
1
D
o
u
b
l
e
-
b
l
i
n
d
,
r
a
n
d
o
m
i
s
e
d
,
c
r
o
s
s
o
v
e
r
,
s
i
n
g
l
e
d
o
s
e
;
n
=
3
1
F
P
D
i
s
k
u
s
v
s
F
P
D
i
s
k
h
a
l
e
r

7
-
-
9
S
m
a
l
l
d
i
f
f
e
r
e
n
c
e
s
N
o
t
e
q
u
i
v
a
l
e
n
t
i
n
h
e
a
l
t
h
y
.
C
l
o
s
e
m
a
t
c
h
i
n
a
s
t
h
m
a
t
i
c
s
(
n
o
c
h
a
r
c
o
a
l
b
l
o
c
k
)
E
q
u
i
v
a
l
e
n
t
D
o
u
b
l
e
-
b
l
i
n
d
,
r
a
n
d
o
m
i
s
e
d
,
p
l
a
c
e
b
o
-
c
o
n
t
r
o
l
l
e
d
,
c
r
o
s
s
o
v
e
r
,
4
-
d
a
y
d
o
s
i
n
g
;
n
=
2
1
D
o
u
b
l
e
-
b
l
i
n
d
,
r
a
n
d
o
m
i
s
e
d
,
a
c
t
i
v
e
-
c
o
n
t
r
o
l
,
c
r
o
s
s
o
v
e
r
,
s
i
n
g
l
e
d
o
s
e
;
n
=
1
2
D
o
u
b
l
e
-
b
l
i
n
d
,
r
a
n
d
o
m
i
s
e
d
,
a
c
t
i
v
e
-
c
o
n
t
r
o
l
,
c
r
o
s
s
o
v
e
r
,
s
i
n
g
l
e
d
o
s
e
;
n
=
1
2
O
p
e
n
-
l
a
b
e
l
,
n
o
n
r
a
n
d
o
m
i
s
e
d
,
u
n
c
o
n
t
r
o
l
l
e
d
,
7
.
5
d
a
y
;
n
=
2
4
D
o
u
b
l
e
-
b
l
i
n
d
,
r
a
n
d
o
m
i
s
e
d
,
a
c
t
i
v
e
-
c
o
n
t
r
o
l
,
c
r
o
s
s
o
v
e
r
,
6
-
w
e
e
k
;
n
=
2
3
2
D
o
u
b
l
e
-
b
l
i
n
d
,
r
a
n
d
o
m
i
s
e
d
,
a
c
t
i
v
e
-
c
o
n
t
r
o
l
,
c
r
o
s
s
o
v
e
r
,
1
2
-
w
e
e
k
;
n
=
2
1
2
F
P
H
F
A
M
D
I
v
s
F
P
C
F
C
M
D
I
1
0
G
o
o
d
m
a
t
c
h
N
o
t
e
q
u
i
v
a
l
e
n
t
a
t
a
l
l
d
o
s
e
s
(
n
o
c
h
a
r
c
o
a
l
b
l
o
c
k
)
E
q
u
i
v
a
l
e
n
t
O
p
e
n
-
l
a
b
e
l
,
r
a
n
d
o
m
i
s
e
d
,
a
c
t
i
v
e
-
c
o
n
t
r
o
l
,
c
r
o
s
s
o
v
e
r
,
s
i
n
g
l
e
d
o
s
e
;
n
=
2
3
S
a
l
m
/
F
P
D
i
s
k
u
s

v
s
S
a
l
m
/
F
P
R
P
I
D

1
1
G
o
o
d
m
a
t
c
h
N
o
t
e
q
u
i
v
a
l
e
n
t
(
n
o
c
h
a
r
c
o
a
l
b
l
o
c
k
)
E
q
u
i
v
a
l
e
n
t
D
o
u
b
l
e
-
b
l
i
n
d
,
r
a
n
d
o
m
i
s
e
d
,
a
c
t
i
v
e
c
o
n
t
r
o
l
,
c
r
o
s
s
o
v
e
r
,
1
4
d
a
y
;
n
=
2
2
B
e
c
l
o
m
e
t
a
s
o
n
e
D
P
C
F
C
M
D
I
v
s
B
e
c
l
o
m
e
t
a
s
o
n
e
D
P
H
F
A
M
D
I
5
B
i
g
d
i
f
f
e
r
e
n
c
e
s
N
o
t
e
q
u
i
v
a
l
e
n
t
(
n
o
c
h
a
r
c
o
a
l
b
l
o
c
k
)
N
o
t
e
q
u
i
v
a
l
e
n
t
D
o
u
b
l
e
-
b
l
i
n
d
,
r
a
n
d
o
m
i
s
e
d
,
c
r
o
s
s
o
v
e
r
,
s
i
n
g
l
e
d
o
s
e
;
n
=
5
1
F
l
u
n
i
s
o
l
i
d
e
C
F
C
M
D
I
v
s
F
l
u
n
i
s
o
l
i
d
e
H
F
A
M
D
I
1
2
B
i
g
d
i
f
f
e
r
e
n
c
e
s
N
o
t
e
q
u
i
v
a
l
e
n
t
(
n
o
c
h
a
r
c
o
a
l
b
l
o
c
k
)
N
o
t
e
q
u
i
v
a
l
e
n
t
O
p
e
n
-
l
a
b
e
l
,
r
a
n
d
o
m
i
s
e
d
,
a
c
t
i
v
e
-
c
o
n
t
r
o
l
,
p
a
r
a
l
l
e
l
g
r
o
u
p
,
1
3
.
5
d
a
y
s
;
n
=
3
1
1
P
e
r
s
s
o
n
2
0
0
8
;
2
K
e
r
w
i
n
2
0
0
8
;
3
L
a

h
e
l
m
a

2
0
0
4
;
4
T
u
k
i
a
i
n
e
n
2
0
0
2
;
5
D
a
l
e
y
-
Y
a
t
e
s
1
9
9
9
;
6
C
l
e
a
r
i
e
2
0
1
0
;
7
F
a
l
c
o
z
2
0
0
0
;
8
M
a
c
k
i
e
2
0
0
0
;
9
K
u
n
k
a
2
0
0
0
;
1
0
D
a
l
e
y
-
Y
a
t
e
s
2
0
0
9
;
1
1
N
o
l
t
i
n
g
a
2
0
0
1
.
Bioequivalence of inhaled drugs
1302 Expert Opin. Drug Deliv. (2011) 8(10)
E
x
p
e
r
t

O
p
i
n
.

D
r
u
g

D
e
l
i
v
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

i
n
f
o
r
m
a
h
e
a
l
t
h
c
a
r
e
.
c
o
m

b
y

N
o
v
a
r
t
i
s

P
h
a
r
m
a

o
n

0
2
/
0
3
/
1
2
F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
in vivo the oropharyngeal cross section can vary with device
type, device resistance, tongue position and patient effort [19].
Attempts to design induction ports that better reflect
and/or mimic physiological reality have included changing
the dimensions and angles in the port, using sticky surfaces
to reflect those present in vivo and using casts from
cadavers [20-22]. Magnetic resonance imaging data have also
been used to determine the retention effect of the oropharyn-
geal airspace during the administration of drug aerosols as
well as to generate mouth--throat models that can be used
with patient relevant flow profiles to evaluate emitted
APSD [23,24].
Deposition within the cascade impactor occurs through
impaction that differs from the in vivo situation where depo-
sition occurs through impaction, sedimentation and diffusion,
the extent of which is dependent on particle size and the inha-
lation manoeuvre of the patient. The cascade impaction test is
conducted at a fixed flow rate so as to allow the aerosol to be
characterised into predefined size ranges, this differs from the
in vivo situation where the inhalation profile is dynamic with
varying flow and acceleration rate.
In an MDI, drug delivery is ballistic in nature being driven
by the propellant present in the formulation, this means that
the patient has to carefully coordinate their inhalation
manoeuvre with actuation of the device; any differences
between MDIs that may impact this coordination would
not be detected by the cascade impaction test. In the case of
the DPI, aerosolisation is driven by the patients inhalation
manoeuvre. This means that, when comparing DPIs, the
resistance of the device itself must be considered as few devices
have the same airflow resistance [25], thus affecting the flow
rates achieved by patients.
As DPI performance is often a function of flow rate, which
can differ from device to device [26], it is necessary to compare
in vitro product performance across a range of flow rates.
Several different approaches have been adopted to address
these issues. Flow rate dependency can be studied using a
range of fixed flow rates such as 30, 60 and 90 l/min; how-
ever, this approach results in different pressure drops across
the device for devices of differing resistivity [27]. An alternative
approach is to compare devices using the same pressure drop
[typically 4 kPa (European Pharmacopedia)], but this means
that the devices operate at different flow rates. Therefore,
the stage cutoffs for the two test conditions will differ. Conse-
quently, attributes such as fine particle dose (< 5 m) can only
be derived by interpolation, and comparison of full cascade
profiles is not straightforward. Using the same pressure drop
with different flow rates may still not allow the comparison
of devices from the perspective of comparing devices with
the same patient inspiratory effort [28]. Cascade impactors
are limited in the range of flow rates at which they can operate
to ensure that stage cutoffs are in the appropriate size
range [29]. This means that the flow rates at which they are
operated may not be the same as those actually achieved by
some patient populations using the device in clinical practice.
In addition to a fixed flow rate, cascade impactors also use
fixed high acceleration rates and airflow profiles that do not
Box 2. Key regulatory requirements for pharmacokinetic studies of orally inhaled drugs in the
European Medicines Agency guideline.
The demonstration of bioequivalence for orally inhaled drugs requires that standard criteria be fulfilled, that is, 90% confidence
intervals for the log-transformed test/reference C
max
and AUC
(0--t)
ratios should lie within 80 -- 125% Tighter limits for AUC and
possibly C
max
may be appropriate for drugs with a narrow therapeutic index. Widened limits for C
max
may be acceptable for
highly variable products
Bioequivalence should be confirmed for partial AUC as a measure of early exposure where a rapid onset of effect is important
Both pulmonary deposition and total systemic exposure should be assessed, unless drug absorption via the oral route is very low
such that pulmonary and systemic bioavailabilities are essentially the same
Total systemic exposure may be acceptable as a surrogate of systemic safety
Dose selection should be based on pharmacokinetic linearity/nonlinearity
Both urinary or plasma pharmacokinetic studies are acceptable in adults, whereas in children only the latter are advocated.
Where urinary pharmacokinetic studies are undertaken, plasma C
max
should be estimated, if feasible, alongside the urinary data
Pharmacokinetic data for parent compounds/pro-drugs should be presented alongside that of active metabolites, assuming
pharmacokinetics of the former are linear and plasma concentrations easily measurable, as C
max
for parent compounds is more
sensitive to detect differences between products
For pressurised metered dose inhalers, pharmacokinetic data comparing test and reference products in conjunction with spacers
should be provided unless comparative in vitro spacer data satisfy stringent criteria for bioequivalence
For dry powder inhalers, the relevance of differences in intrinsic device resistance should be considered with respect to children
AUC
(0--t)
: Area under the plasma concentration--time curve from time zero to infinity; C
max
: Maximum plasma concentration.
Box 3. Key components of the aggregate
weight of evidence approach described by the
FDA [15].
Similarity of formulation
Similarity of device design
Comparative in vitro tests
Comparative systematic exposure studies
Pharmacodynamic or clinical end point studies
Daley-Yates & Parkins
Expert Opin. Drug Deliv. (2011) 8(10) 1303
E
x
p
e
r
t

O
p
i
n
.

D
r
u
g

D
e
l
i
v
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

i
n
f
o
r
m
a
h
e
a
l
t
h
c
a
r
e
.
c
o
m

b
y

N
o
v
a
r
t
i
s

P
h
a
r
m
a

o
n

0
2
/
0
3
/
1
2
F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
reflect patient profiles [30]. This is relevant when comparing
DPIs, as the performance of a device may be influenced
by the acceleration rate used in testing [31]. In contrast,
others have concluded that constant flow rates could be used
providing that the rates selected reflected those expected
in vivo [32].
The differences discussed might explain some discrepancies
between clinical observations and in vitro data. For example, a
comparison of two DPIs raised concerns about the predict-
ability of in vitro FPM data [33]: the reservoir powder inhala-
tion device (RPID) and Diskus

(GlaxoSmithKline, Ware,
UK) inhalers administered the same lactose blend of salme-
terol and fluticasone propionate. The performances of the
inhalers were comparable over a range of flow rates. In vitro
FPM data were similar for the two devices, but pharmacoki-
netic data showed that the devices could not be considered
bioequivalent in terms of fluticasone propionate AUC
[RPID:Diskus estimated ratio: 2.00; 90% confidence interval
(CI): 1.56 -- 2.55] and salmeterol C
max
(ratio: 1.92; 90% CI:
1.64 -- 2.25). However, data showed the two inhalers to be
bioequivalent in terms of mean morning peak expiratory
flow in a clinical efficacy study. There was clearly a lack of
association both between in vitro FPM and in vivo pharmaco-
kinetic data and between the pharmacokinetic data and clini-
cal efficacy data. The key point to takeaway from this
observation is that the difference in pharmacokinetic profile
between the two inhalers could not be predicted from the
in vitro data alone.
4. Interpretation of clinical bioequivalence
studies
There are relatively few well-documented examples of inhaled
product bioequivalence assessments that contain data from
in vitro profiling, pharmacokinetics, lung deposition and clinical
efficacy findings. Some of the best documented examples were
summarised recently [34]. These studies included comparisons
of generic and innovator products and studies assessing the effect
of changes made to propellant or inhaler type. In most examples,
in vitro, pharmacokinetic and clinical efficacy data were not in
agreement. In general, in vitro data did not predict correctly
the findings of pharmacokinetic and/or clinical efficacy studies
(Table 1), and the pharmacodynamic data were less discriminat-
ing than pharmacokinetic assessments. Of the examples shown
in Table 1, three are studies where bioequivalence was concluded
from pharmacokinetic data. However, it would appear that,
when other data are considered more closely, these conclusions
may be unreliable. For example, a comparison of budesonide
administered using either a Flexhaler (AstraZeneca,
Soldertalje, Sweden) (budesonide 180 g) or a Turbuhaler

(AstraZeneca, Soldertalje, Sweden) (budesonide 200 g) fea-

tures two products designed to have similar in vitro performance.
Pharmacokinetic analysis demonstrated bioequivalence within
the confidence limits of 0.8 -- 1.25 for AUC (treatment
comparison: 0.96; 90% CI: 0.91 -- 1.02) and C
max
(treatment
comparison: 1.00; 90%CI: 0.92 -- 1.09) [35]. A subsequent clin-
ical efficacy study, although not designed to test equivalence,
showed that lung function responses were lower for the Flexhaler
formulation than Turbuhaler [36]. In the second example, bude-
sonide administered by Easyhaler and Turbuhaler showed bio-
equivalence in terms of AUC (Easyhaler:Turbuhaler ratio:
1.02; 90% CI: 0.96 -- 1.09) and C
max
(ratio: 0.94; 90% CI:
0.86 -- 1.03) [34]. However, the in vitro performance of these
devices is likely to be different [37]. As charcoal block methodol-
ogy was used in the study, only lung deposition was assessed,
with no assessment of total systemic exposure. Another study
presented total systemic exposure data, but with insufficient
data to allow an assessment of bioequivalence in terms of sys-
temic safety as only limited sampling was conducted around
C
max
and the Easyhaler showed greater variability than the Tur-
buhaler [38]. Therefore, it would be unwise to conclude true
bioequivalence without additional and robust systemic pharma-
cokinetic data. A third comparison investigated the administra-
tion of beclometasone dipropionate using innovator and
generic MDIs. Reviewing the pharmacokinetic data obtained
at the 250-g strength product, bioequivalence was concluded
for pharmacokinetic, in vitro and clinical studies. However,
when data obtained at different dose levels were reviewed, clear
indications of nonbioequivalence emerged: differences were
observed in the FPM and pharmacokinetics data at other doses
and, therefore, it was concluded that bioequivalence could not
be assumed for the high and low doses [39].
Pharmacokinetic data in various studies appear to sug-
gest only partial bioequivalence when comparing inhaled
products [40-42]. In a study that compared the pharmacokinetics
of fluticasone propionate administered via the Diskus or
Diskhaler

(GlaxoSmithKline, Ware, UK), data from patients

with asthma differed markedly from that obtained from healthy
subjects. In vitro analysis showed similarities between the Diskus
and Diskhaler for FPM (17.0 and 16.9%, respectively),
although a difference was observed for particles < 2 m
(2.0 and 3.4%, respectively) [43]. Data from healthy subjects
demonstrated higher fluticasone propionate systemic exposures
following Diskus inhaler administration and in vitro FPM dif-
fered between the two devices [44]. In patients with mild-to-mod-
erate asthma, systemic exposure to fluticasone propionate was
~50% that in healthy subjects with little difference between
the two devices within population groups [40]. These findings
emphasised the need to carry out studies in the target patient
population. In another example, the delivery profiles were com-
pared for two different propellants, CFCand hydrofluoroalkane
(HFA), following the administration of fluticasone propionate
via two MDIs [41]. The inhalers had similar in vitro FPM. At
the 250-g dose, systemic exposure was similar for the two for-
mulations in terms of AUC (HFA:CFC ratio: 0.88; 90% CI:
0.75 -- 1.05) and C
max
(ratio: 1.01; 90% CI: 0.86 -- 1.18).
However, at the 125-g dose, there was a marked difference
between the formulations, with systemic exposure approxi-
mately one-third lower for the HFA formulation than CFC in
terms of AUC (ratio: 0.67; 90% CI: 0.57 -- 0.79) and C
max
Bioequivalence of inhaled drugs
1304 Expert Opin. Drug Deliv. (2011) 8(10)
E
x
p
e
r
t

O
p
i
n
.

D
r
u
g

D
e
l
i
v
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

i
n
f
o
r
m
a
h
e
a
l
t
h
c
a
r
e
.
c
o
m

b
y

N
o
v
a
r
t
i
s

P
h
a
r
m
a

o
n

0
2
/
0
3
/
1
2
F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
(ratio: 0.63; 90%CI: 0.75 -- 1.05). It was clear that the pharma-
cokinetic profile for the 125-g dose could not be extrapolated
from the 250-g dose, and therefore, the two formulations
should not be considered bioequivalent at all doses. Another
study showed similarity of an innovator and generic fluticasone
propionate/salmeterol combination product in terms of overall
fine particle dose (particles < 4.7 m) for both the fluticasone
propionate and salmeterol components [42]. On pharmacoki-
netic assessment, the salmeterol moiety of the generic product
produced significantly greater systemic exposure than the
innovator product, although bioequivalent systemic exposure
was observed for fluticasone propionate. No major differences
were observed in pharmacodynamic end points. As the authors
noted, the results of this study highlighted the potential pitfalls
of extrapolating in vitro data for fine particle dose to a clinical
setting for in vivo pharmacokinetic and pharmacodynamic out-
comes. Taken together, the findings of these three comparisons
highlight the difficulty in cases where in vitro data alone are
used to assess the bioequivalence of two inhaled drug products.
In three of the examples presented in Table 1 [33,39,45],
pharmacokinetic data did not establish bioequivalence in
terms of AUC and C
max
. In comparisons of MDIs delivering
drugs with either CFC or HFA propellants, the in vitro FPM
profiles, median mass aerodynamic diameter and systemic
availability were sufficiently different to offer a low expecta-
tion that in vivo bioequivalence would be found, and this
was confirmed [39,45]. An example in which an innovator
MDI administering beclometasone dipropionate propelled
by CFC was compared with a generic inhaler propelled
by HFA indicated that pharmacokinetic data may be more
discriminating than pharmacodynamic data [39]. In this
instance, both in vitro and pharmacokinetic assessments
showed differences between the two inhalers in terms of
the emitted FPM dose and systemic exposure to the active
metabolite. Serum cortisol AUC data showed little differ-
ence between the inhalers, so inappropriate conclusions of
bioequivalence would have been made for the lower dose
inhalers if the bioequivalence assessment had been based
on serum cortisol AUC alone.
As stated earlier, the role of pharmacokinetic data in estab-
lishing the bioequivalence of topically acting inhaled drugs
relies on an assumption that the systemic concentrations of
an orally inhaled topically acting drug reflect those at the
site of action. The validity of this assumption was assessed
indirectly in a study originally designed to investigate whether
the therapeutic benefits of inhaled fluticasone propionate are
mediated through topical effects in the lungs rather than via
systemic effects. The study compared the effect on morning
FEV1 of inhaled (100 and 500 g twice daily) and orally
(20 mg once daily) administered fluticasone propionate and
placebo after 6 weeks in 274 patients with asthma [46]. High
oral doses of fluticasone propionate achieved steady-state
plasma concentrations approximately twofold greater than
those following inhaled fluticasone propionate. However,
although both inhaled doses improved morning FEV1, oral
dosing showed no significant improvement. This finding pro-
vides further compelling evidence that the systemic plasma
concentrations and the concentrations at the site of action
in the lung are not in equilibrium, and consequently, it
must be considered that pharmacokinetic data are not a valid
surrogate of topical efficacy. However, it is important to
emphasise that pharmacokinetic data remain a valid surrogate
of systemic safety.
Looking to the future, imaging technologies have been pro-
posed as a means to provide further information on lung
deposition of inhaled drugs. Three methods are available: pla-
nar gamma scintigraphy, currently the most frequently used
technique, single photon emission computed tomography
and positron emission tomography (PET). The first two tech-
niques involve passive radiolabelling of the aerosol particles or
droplets and PET requires direct labelling of the drug mole-
cule. At the RDD 2010 conference, a workshop session dis-
cussed this area in particular. The output describing these
discussions has recently been published [47]. There are two
main problems to overcome in adopting these techniques,
first, the process of labelling the drug particles has the poten-
tial to alter the drug product such that it is no longer repre-
sentative of a commercially manufactured batch of the test
or reference product. Second, the drug particles may only
be surfaced-labelled or the marker is too labile and hence
does not give a true picture of lung deposition of the
unadulterated product.
5. Integrating guidance with experience
Rigorous standards are needed to ensure that alternative
bioequivalent products are truly interchangeable without
compromising patient safety or efficacy. Guidance is begin-
ning to emerge on what might be regarded as an acceptable
level of proof. As these guidelines are new, or not finalised,
it is likely that they will go through a process of evolution as
more data become available. The step-wise approach pro-
posed by the EMA places the burden on the regulatory
authorities to develop a robust and appropriate framework
by which sponsors can determine bioequivalence unequivo-
cally. In Canada, focus has been placed on surrogate marker
data for efficacy, hence the burden falls on the Sponsor iden-
tifying and conducting appropriate clinical studies. In the
weight-of-evidence approach suggested by the FDA, the bur-
den also falls on the Sponsor to deliver a well-conceived and
executed bioequivalence programme that provides data to
support claims of bioequivalence.
Clinical experience, including those studies presented in
this review, shows that the determination of bioequivalence
of topically acting, orally inhaled drugs is complex particularly
where significant differences in formulation and/or device are
present. In these cases, the evidence to support the reliance on
a single source of data such as in vitro or pharmacokinetic data
was lacking. In such examples, a complete package of in vivo
and in vitro data is likely to be required to make a decision
Daley-Yates & Parkins
Expert Opin. Drug Deliv. (2011) 8(10) 1305
E
x
p
e
r
t

O
p
i
n
.

D
r
u
g

D
e
l
i
v
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

i
n
f
o
r
m
a
h
e
a
l
t
h
c
a
r
e
.
c
o
m

b
y

N
o
v
a
r
t
i
s

P
h
a
r
m
a

o
n

0
2
/
0
3
/
1
2
F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
on bioequivalence. For some comparisons, pharmacokinetic
data have shown bioequivalence of two products, whereas
other data have shown differences in terms of in vitro data
or a lack of bioequivalence for clinical efficacy.
Overall, the published data highlight how contrasting find-
ings between in vitro, pharmacokinetic and clinical efficacy
studies imply that reliance solely on in vitro data may not pro-
vide a complete profile (including the justification of pharma-
ceutical equivalence) that permits the observer to conclude
that sufficient bioequivalence between inhaled products has
been established. The most widely used in vitro particle-
size testing methodologies do not always predict in vivo
performance despite its widespread and almost universal
adoption for in vitro testing. The balance of evidence favour-
ing this approach is theoretical rather than empirical. In the
future, imaging technologies may assist in determining lung
deposition patterns; however, two issues need to be resolved:
first, whether there is sufficient methodological agreement
between laboratories to allow direct comparison of imaging
data and, second, acceptance by regulators that passive label-
ling, which changes the drug formulation, does not represent
adulteration of the test sample.
Finally, it is clear that although pharmacokinetic data do
not always correlate well with in vitro, efficacy or safety
data, they are an important component in establishing
the bioequivalence of topically acting inhaled products. As
systemic pharmacokinetic data do not necessarily reflect
drug concentrations at the site of action in the lung, their
primary role should be to establish bioequivalence in terms
of the systemic safety profile. In this role, pharmacokinetic
data are likely to be superior to pharmacodynamic end points.
6. Expert opinion
Establishing the bioequivalence of inhaled drugs is a multistep
process with little experimental evidence to validate the use
of data from one of the various in vivo and in vitro methods
over those from another. This lack of clarity is one factor
contributing to the absence of a consistent approach across
regulatory authorities to establishing bioequivalence. In vitro
assessments of emitted doses are not universally predictive of
pharmacokinetic and pharmacodynamic outcomes in clinical
studies, pharmacokinetic data do not always correlate well
with in vitro or clinical efficacy data, and pharmacodynamic
data lack the precision of pharmacokinetic data when used
to determine bioequivalence. It is clear that more data and
better models are required to describe the relationship
between in vitro and in vivo observations. A weight-of-evi-
dence approach to new formulations that can be applied on
a case-by-case basis rather than a rigid regulatory approach
may be more acceptable. This approach places greater respon-
sibility on drug development scientists to determine the level
of evidence necessary to establish bioequivalence. Unless there
is a high degree of pharmaceutical equivalence between the
test and reference products, as might arise from minor
changes to an existing product, confidence in the predictabil-
ity of in vitro data is likely to be low. Our lack of understand-
ing about the relationship between in vitro, in vivo and
clinical data suggests that, in most cases, the consideration
of a combination of preclinical and clinical data may be pref-
erable to abridged approaches relying on in vitro data alone.
There are several unresolved issues preventing the agreement
of general guidelines on the best approach to establishing bio-
equivalence for all classes of inhaled drug products. A way for-
ward could involve the development of product- or drug class-
specific guidelines allowing some of the outstanding issues to
be circumvented.
Acknowledgements
The authors recognise the role of professional medical writers
in the preparation of this manuscript: T Hardman and J Cook
from Niche Science and Technology.
Declaration of interest
Both authors are employees of GlaxoSmithKline, who
develop and manufacture inhaled drug products.
Bioequivalence of inhaled drugs
1306 Expert Opin. Drug Deliv. (2011) 8(10)
E
x
p
e
r
t

O
p
i
n
.

D
r
u
g

D
e
l
i
v
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

i
n
f
o
r
m
a
h
e
a
l
t
h
c
a
r
e
.
c
o
m

b
y

N
o
v
a
r
t
i
s

P
h
a
r
m
a

o
n

0
2
/
0
3
/
1
2
F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
Bibliography
Papers of special note have been highlighted as
either of interest () or of considerable interest
() to readers.
1. The Montreal Protocol. The Montreal
protocol on substances that deplete the
ozone layer. Final act (Nairobi: UNEP
1987. Fed Reg 1994;59:56276-98
2. Stewart BA, Ahrens RC, Carrier S, et al.
Demonstration of in vivo Bioequivalence
of generic albuterol metered dose inhaler
to Ventolin. Chest 2000;117(3):714-21
3. Hindle M, Newton DA. Chrystyn H.
investigation of an optimal inhaler
technique with the use of urinary
salbutamol excretion as a measure of
relative bioavailability to the lung.
Thorax 1993;48:607-10
4. Hindle M, Chrystyn H. Relative
bioavailability of salbutamol to the lung
following inhalation using metered dose
inhalation method and spacer devices.
Thorax 1994;49:549-53
5. Hindle LM, Newton DA, Chystyn H.
Dry powder in-halers are bioequivalent
to metered dose inhalers: a study using a
new urinary albuterol (salbutamol) assay
technique. Chest 1995;107:629-33
6. Food and Drug Administration. Interim
guidance for documentation of In vivo
bioequivalence of albuterol aerosols
(metered-dose inhalers. Rockville, MD:
Division of bioequivalence. Office of
generic drugs, food and drug
administration; January 27, 1994;
Publication No. 1-27
7. Chystyn H. Standard for bioequivalence
of inhaled products. Clin Pharmacokinet
1994;26:1-6
8. European Medicines Agency. Committee
for Medicinal Products for Human Use.
(2007). Guideline on the requirements
for clinical documentation for orally
inhaled products (OIP) including the
requirements for demonstration of
therapeutic equivalence between two
inhaled products for use in the treatment
of asthma and chronic obstructive
pulmonary disease (COPD. Draft.
CPMP/EWP/4151/00 Rev 1
9. European Medicines Agency. Committee
for Medicinal Products for Human Use.
Guideline on the requirements for
clinical documentation for orally inhaled
products (OIP) including the
requirements for demonstration of
therapeutic equivalence between two
inhaled products for use in the treatment
of asthma and chronic obstructive
pulmonary disease (COPD) in adults and
for use in the treatment of asthma in
children and adolescents.. Doc Ref.
CPMP/EWP/4151/00 Rev.1. 22 January
2009 Available from: http://www.ema.
europa.eu/docs/en_GB/document_
library/Scientific_guideline/2009/09/
WC500003504.pdf
[Accessed 30 August 2010]
.. Key Regulatory guideline
10. Healthy Canada. Release of Draft
Guidance Document: Submission
Requirements for Subsequent Market
Entry Inhaled Corticosteroid Products
for Use in the Treatment of Asthma.
2 August 2007. Available from: http://
www.hc-sc.gc.ca/dhp-mps/alt_formats/
hpfb-dgpsa/pdf/prodpharma/
inhal_corticost-eng.pdf
[Accessed 30 August 2010]
.. Key Regulatory guideline
11. Adams WP, Ahrens RC, Chen M-L,
et al. Demonstrating bioequivalence of
locally acting orallyinhaled drug
products (OIPs): workshop summary
report. J Aerosol Med Pulm Drug Deliv
2010;23:1-29
.. Provides a detailed exploration of
the issues
12. Respiratory Drug Delivery. ILISBN:
Book 1, 1-933722-41-X: In: Dalby RN,
Byron PR, Peart J, et al. editors.
FloridaDavis Healthcare International
Publishing; Orlando, River Grove:
2010; vii
13. Respiratory Drug Delivery Online.
Available from: http://www.
rddonline.com/index.php
[Accessed 19 January 2011]
14. International Society for Aerosols in
Medicine/International Pharmaceutical
Aerosol Consortium on Regulation and
Science Workshop in Frankfurt,
12--13 Oct 2010. Available from:
http://isam-org.activaero.de/component/
content/article/13/38-ipac
[Accessed 19 January 2010]
15. Lee SL, Adams WP, Li BV, et al.
In vitro considerations to support
bioequivalence of locally acting drugs in
dry powder inhalers for lung diseases.
AAPS J 2009;11:414-23
.
Describes the FDA perspective on
the subject
16. Committee for Medicinal Products for
Human Use. European medicines
agency. 2008). Guideline on the
investigation of bioequivalence, Draft.
CPMP/EWP/QWP/1401/98 Rev 1
17. Product Quality Research Institute
workshop on demonstrating
bioequivalence of locally acting
orally inhaled drug products. Available
from: http://www.pqri.org/workshops/
imagespdfs/Bioequivalence2009_
Final_Agenda.pdf
[Accessed 19 January 2011]
18. Dissanayake S. Application of the EU
Guidelines for pharmacokinetic studies of
locally acting orally inhaled drug
products. In: Dalby RN, Byron PR,
Peart J, Suman JD, Farr SJ, Young PM.
editors. Respiratory Drug Delivery.
ISBN: Book 1, 1-933722-41-X: David
Healthcare International Publishing;
Orlando Florida, River Grove IL;
2010. p. 293-304
. Provides an insight into the
application of the EU guidance
19. Ehtezazi T, Horsfield MA, Barry PW,
et al. Dynamic change of the upper
airway during inhalation via aerosol
delivery devices. J Aerosol Med
2004;17:325-34
20. Dolovich M, Rhem R. Impact of
oropharyngeal deposition on inhaled
dose. J Aerosol Med 1998;11:S112-15
21. Newhouse MT, Dobranowski JA,
Chambers C, et al. Development and
preliminary validation studies with an
ex-vivo oropharyngeal model. In:
Byron PR, Dalby RN, Farr SJ, editors.
Proceedings of Respiratory Drug Delivery
VI. Interpharm Press, Inc., Buffalo
Grove, IL.1998, p. 389-91
22. Cheng Y-S, Zhou Y, Chen BT. Particle
deposition in a cast of human oral
airways. Aerosol Sci Tech
1999;31:286-300
23. Burnell PKP, Asking L, Borgstrom L,
et al. Studies of the human
oropharyngeal airspaces using magnetic
resonance imaging IV -- the
oropharyngeal retention effect for four
inhalation delivery systems.
J Aerosol Med 2007;20:269-81
24. Olsson B, Berg E, Svennsson M.
Comparing aerosol size distributions that
penetrate mouth-throat models under
realistic inhalation conditions.
Daley-Yates & Parkins
Expert Opin. Drug Deliv. (2011) 8(10) 1307
E
x
p
e
r
t

O
p
i
n
.

D
r
u
g

D
e
l
i
v
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

i
n
f
o
r
m
a
h
e
a
l
t
h
c
a
r
e
.
c
o
m

b
y

N
o
v
a
r
t
i
s

P
h
a
r
m
a

o
n

0
2
/
0
3
/
1
2
F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
In: Dalby RN, Byron PR, Peart J, et al.
editors. Respiratory Drug Delivery.
David Healthcare International
Publishing; Orlando Florida, River Grove
IL; 2010. p. 225-33
..
Provides an overview of the issues
relating to the role of Pk studies
25. Raid AAM, Tarsin WY, Assi KH, et al.
Can all patients with COPD use the
correct inhalation flow with all inhalers
and does training help? Resp Med
2007;101:2395-401
26. Tarsin W, Assi KH, Chrystyn H.
In-vitro inhalation flow rate dependent
dosage emission and inconsistency from a
combination of budesonide and
efometrol in a dry powder inhaler.
J Aerosol Med 2004;17:25-32
27. Malton A, Sumby BS, Smith IJ.
A comparison of in vitro drug delivery
from two multi-dose powder inhalation
devices. Eur J Clin Res 1995;7:1-17
28. Prime D. Determining flow and flow
acceleration conditions for in-vitro
testing of dry powder inhalers. Abstracts
from international society of aerosols in
medicine, 15th congress; Perth;
2005, p. P68
29. Mitchell J, Newman S, Chan H-K.
In vitro and in vivo aspects of cascade
impactor tests and inhaler performance:
a review. AAPS PharmSciTech
2007;8:Article 110
30. Bisgaard H, Klug B, Sumby BS, et al.
Fine particle mass from the Diskus
inhaler and Turbuhaler inhaler in
children with asthma. Eur Respir J
1998;11:1111-15
31. de Boer AH, Bolhuis GK, Gjaltema D,
et al. Inhalation characteristics and their
effects on in vitro drug delivery from dry
powder inhalers - Part 3: the effect of
flow increase rate (FIR) on the in vitro
drug release from the Pulmicort 200.
Turbuhaler 1997;11:67-77
32. Finlay WH, Gehmlich MG. Inertial
sizing of aerosol inhaled from two dry
powder inhalers with realistic breath
patterns versus constant flow rates.
Int J Pharm 2000;210:83-95
33. Daley-Yates PT, Parkins DA,
Thomas MJ, et al. Pharmacokinetic,
pharmacodynamic, efficacy, and safety
data from two randomized, double-blind
studies in patients with asthma and an
in vitro study comparing two dry powder
inhalers delivering a combination of
salmeterol 50 mug and fluticasone
propionate 250mug: implications for
establishing bioequivalence of inhaled
products. Clin Ther 2009;31:370-85
34. Daley-Yates P. The clinical utility of
pharmacokinetics: theoretical
consideration across drug classes. In:
Dalby RN, Byron PR, Peart J,
Suman JD, Farr SJ, Young PM. editors.
Respiratory Drug Delivery. ISBN: Book
1, 1-933722-41-X: David Healthcare
International Publishing; Orlando
Florida, River Grove IL: 2010. p. 273-84
.
Provides a review on inhaled
bioequivalance studies
35. Persson G, Ankerst J, Gillen M, et al.
Relative systemic availability of
budesonide in patients with asthma after
inhalation from two dry powder inhalers.
Curr Med Res Opin 2008;24:1511-17
36. Kerwin EM, Pearlman DS, de Guia T,
et al. Evaluation of efficacy, safety, and
pharmacokinetics of budesonide delivered
via two dry powder inhalers. Curr Med
Res Opin 2008;4:1497-510
37. Lahelma S, Kirjavainen M, Kela M, et al.
Equivalent lung deposition of budesonide
in vivo: a comparison of dry powder
inhalers using a pharmacokinetic method.
Br J Clin Pharmacol 2004;59:167-73
38. Tukiainen H; on behalf of the Finnish
Study Group. Safety, tolerability and
acceptability of two dry powder inhalers
in the administration of budesonide in
steroid-treated asthmatic patients.
Respir Med 2002;96:221-9
39. Daley-Yates PT, Bagen S, Tournant J,
et al. Beclomethasone dipropionate
chlorofluorocarbon and
hydrofluoroalkane metered dose inhalers:
relationship between systemic exposure,
dose, fine particle mass and particle size
in healthy volunteers. Eur Resp J
1999;14:1358
40. Falcoz C, Horton J, Mackie AE, et al.
Pharmacokinetics of inhaled fluticasone
propionate delivered via the Diskhaler
and Diskus delivery devices in patients
with mild to moderate asthma.
Clin Pharmacokinet
2000;39(Suppl 1):31-7
41. Kunka R, Andrews S, Pimazzoni M,
et al. Dose proportionality of fluticasone
propionate from hydrofluoroalkane
inhalers and comparability with
chlorofluorocarbon inhalers. Resp Med
2000;94(Suppl B):10-16
42. Clearie KL, Williamson PA,
Vaidyanathan S, et al. Systemic
bioavailability of hydrofluoroalkane
(HFA) formulations of fluticasone/
salmeterol in healthy volunteers via
pMDI alone and spacer. Br J
Clin Pharmacol 2010;69:637-44
43. Johnson M. Fluticasone propionate:
pharmacokinetic and pharmacodynamic
implications of different aerosol delivery
systems. In: Byron PR, Dalby RN,
Farr SJ. editors. Respiratory Drug
Delivery VI. Interpharm Press, Inc;
Buffalo Grove, IL: 1998. p. 61-70
44. Mackie AE, McDowall JE, Falcoz C,
et al. Pharmacokinetics of fluticasone
propionate inhaled via the Diskhaler and
Diskus powder devices in healthy
subjects. Clin Pharmacokinet
2000;39(Suppl 1):23-30
45. Noltinga A, Sistab S, Abramowitza W.
Flunisolide HFA vs Flunisolide CFC:.
Pharmacokinetic comparison in healthy
volunteers. Biopharm Drug Dispos
2001;22:373-82
46. Lawrence M, Wolfe J, Webb DR, et al.
Efficacy of inhaled fluticasone propionate
in asthma results from topical and not
from systemic activity. Am J Respir Crit
Care Med 1997;156:744-51
47. OConnor D, Adams WP, Chen M-L,
et al. Role of pharmacokinetics in
establishing bioequivalence for orally
inhaled drug products: workshop
summary report. J Aerosol Med Pulm
Drug Deliv 2011;24:1-17
Affiliation
Peter T Daley-Yates
1
& David A Parkins
2

Author for correspondence

1
GlaxoSmithKline, Respiratory Medicines
Discovery and Development,
Stockley Park,
Middlesex UB11 1BT, UK
Tel: +44 020 8587 5281;
Fax: +44 020 8587 5400;
E-mail: [email protected]
2
GlaxoSmithKline,
Inhaled Product and Device Technology,
Park Road, Ware Hertfordshire, UK
Bioequivalence of inhaled drugs
1308 Expert Opin. Drug Deliv. (2011) 8(10)
E
x
p
e
r
t

O
p
i
n
.

D
r
u
g

D
e
l
i
v
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

i
n
f
o
r
m
a
h
e
a
l
t
h
c
a
r
e
.
c
o
m

b
y

N
o
v
a
r
t
i
s

P
h
a
r
m
a

o
n

0
2
/
0
3
/
1
2
F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.