See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/230841648

Implementation of a Reference-Scaled Average Bioequivalence Approach for

Highly Variable Generic Drug Products by the US Food and Drug
Administration

Article in The AAPS Journal · September 2012

DOI: 10.1208/s12248-012-9406-x · Source: PubMed

CITATIONS READS

94 4,321

12 authors, including:

Barbara Myers Davit Sam H Haidar

Merck & Co. U.S. Food and Drug Administration
44 PUBLICATIONS 1,949 CITATIONS 34 PUBLICATIONS 1,832 CITATIONS

SEE PROFILE SEE PROFILE

Stephanie Choi Christina H Lee

U.S. Food and Drug Administration U.S. Food and Drug Administration
6 PUBLICATIONS 450 CITATIONS 3 PUBLICATIONS 107 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Stemcentrx Hybrid LBA/LC-MS View project

Stability Research View project

All content following this page was uploaded by Sam H Haidar on 05 July 2016.

The user has requested enhancement of the downloaded file.

The AAPS Journal ( # 2012)
DOI: 10.1208/s12248-012-9406-x

Review Article
Theme: Facilitating Oral Product Development and Reducing Regulatory Burden through Novel Approaches to Assess Bioavailability/Bioequivalence
Guest Editors: James Polli, Jack Cook, Barbara Davit, and Paul Dickinson

Implementation of a Reference-Scaled Average Bioequivalence Approach

for Highly Variable Generic Drug Products by the US Food and Drug
Administration

Barbara M. Davit,1,7 Mei-Ling Chen,2 Dale P. Conner,1 Sam H. Haidar,3 Stephanie Kim,4 Christina H. Lee,1
Robert A. Lionberger,4 Fairouz T. Makhlouf,5 Patrick E. Nwakama,1 Devvrat T. Patel,1
Donald J. Schuirmann,6 and Lawrence X. Yu4

Received 27 January 2012; accepted 27 August 2012

Abstract. Highly variable (HV) drugs are defined as those for which within-subject variability (%CV) in
bioequivalence (BE) measures is 30% or greater. Because of this high variability, studies designed to
show whether generic HV drugs are bioequivalent to their corresponding HV reference drugs may need
to enroll large numbers of subjects even when the products have no significant mean differences. To
avoid unnecessary human testing, the US Food and Drug Administration’s Office of Generic Drugs
developed a reference-scaled average bioequivalence (RSABE) approach, whereby the BE acceptance
limits are scaled to the variability of the reference product. For an acceptable RSABE study, an HV
generic drug product must meet the scaled BE limit and a point estimate constraint. The approach has
been implemented successfully. To date, the RSABE approach has supported four full approvals and one
tentative approval of HV generic drug products.
KEY WORDS: bioequivalence; generic drugs; highly variable drugs; reference-scaled average
bioequivalence; US Food and Drug Administration.

INTRODUCTION criteria required by the US Food and Drug Administration

(FDA) for marketing approval of a new generic drug. A
Acceptable bioequivalence (BE) between a generic drug systemically active generic drug is considered to be bioequiva-
product and its corresponding reference product is among the lent to the reference-listed drug if the rate and extent of
absorption of the two products do not show a significant
difference (1). Drug rate and extent of absorption are typically
assessed by conducting in vivo BE studies in human subjects in
1 which generic and reference drug plasma pharmacokinetic (PK)
Office of Generic Drugs, Center for Drug Evaluation and Research,
US Food and Drug Administration, 7520 Standish Place, Metro profiles are characterized and compared. The PK data are used
Park North One, Rockville, Maryland 20855, USA. to obtain peak drug plasma concentration (Cmax) as the BE
2
Office of Pharmaceutical Sciences, Center for Drug Evaluation and measure of rate of absorption and area under the drug plasma
Research, US Food and Drug Administration, 10903 New concentration versus time profile (AUC) as the BE measure of
Hampshire Avenue, White Oak Building 51, Silver Spring, extent of absorption. Most BE studies enroll healthy normal
Maryland 20903, USA. subjects who receive single doses of the generic (test) or
3
Office of Compliance, Center for Drug Evaluation and Research,
reference product via a two-way crossover design. For the most
US Food and Drug Administration, 10903 New Hampshire Avenue,
White Oak Building 51, Silver Spring, Maryland 20903, USA. part, the US FDA asks generic drug applicants to statistically
4
Office of Generic Drugs, Center for Drug Evaluation and Research, compare generic and reference AUC and Cmax values using the
US Food and Drug Administration, 7519 Standish Place, Metro two one-sided tests (TOST) procedure (2). Under the TOST,
Park North Four, Rockville, Maryland 20855, USA. the 90% confidence interval (CI) around the geometric mean
5
Office of Biostatistics, Center for Drug Evaluation and Research, ratio (GMR) of the test and reference values of Cmax and AUC
US Food and Drug Administration, 7520 Standish Place, Metro
is required to fit within BE limits, set from 80% to 125% (3). The
Park North One, Rockville, Maryland 20855, USA.
6
Office of Biostatistics, Center for Drug Evaluation and Research,
width of the 90% CI depends upon the number of subjects in the
US Food and Drug Administration, 10903 New Hampshire Avenue, study and the variability of the BE measure.
White Oak Building 21, Silver Spring, Maryland 20903, USA. Highly variable (HV) drugs are defined as drugs in which
7
To whom correspondence should be addressed. (e-mail: the within-subject variability (defined as the %CV) in one or
barbara.davit@fda.hhs.gov) more of the BE measures is 30% or greater (4,5). A survey of

1550-7416/12/0000-0001/0 # 2012 American Association of Pharmaceutical Scientists

 Davit et al.

generic drug products reviewed at the FDA from 2003 to subjects needed in a crossover design will also increase,
2005 suggested that about 20% of the generic drugs evaluated assuming that all other factors remain constant. Thus, BE
for marketing approval are HV due to their drug substance study sample size is calculated based on a type I error rate of
dispositional characteristics (6). Determining the BE of HV 5% per test, the desired study power, and the best estimates
generic drugs is challenging because the high within-subject of test/reference ratios and within-subject variability. Table I
variability means that large numbers of subjects may be shows how these factors influence the number of subjects
needed to show BE. Figure 1 shows the results of two needed to provide an 80% chance of an acceptable BE study.
hypothetical BE studies with GMRs near 1.0. Product A As shown in Table I, the number of study subjects
meets the BE limits because within-subject variability is needed to show BE increases dramatically for HV drugs. The
relatively low. Product B fails to meet the BE limits because FDA observed in a survey of generic drug product BE studies
of high within-subject variability. Thus, although product B, reviewed from 2003 to 2005 that studies of HV drugs
with a GMR near 1.0, appears to be well-designed to perform generally used more subjects than studies of lower variability
the same as the reference product in vivo, it will be necessary drugs (6).
to increase the number of study subjects—perhaps dramati-
cally—in order for product B to meet the BE limits. SHOULD “ONE-SIZE-FITS-ALL” BE LIMITS APPLY
Several factors influence the sample size needed to meet TO HV DRUGS
the regulatory criteria for acceptable BE. First, each one-
sided test is carried out at the 5% level of significance, which One important observation is that clinical data strongly
corresponds to a 90% CI (7). The 5% level of significance support a conclusion that HV drugs have wide therapeutic
represents the type I error rate (α), which is the probability of indices. Otherwise, there would have been significant safety
incorrectly deeming as bioequivalent two formulations whose issues and lack of efficacy during the pivotal safety and
true (population) GMR fails to meet the BE limits. The efficacy clinical trials required for initial FDA marketing
second factor influencing sample size is study power, defined approval (10). In other words, the reference product, when
as the likelihood or chance of correctly demonstrating BE dosed on different occasions, was safe and efficacious, despite
when it, in fact, exists (8,9). A third factor influencing sample high PK variability. The majority of HV drugs appear to fall
size is the test/reference BE measure ratios. If the true test/ into Biopharmaceutics Classification System class II or IV
reference ratio differs from unity, the overall power to show (low aqueous solubility–high intestinal permeability and low
BE is reduced at any given sample size, resulting in an aqueous solubility–low intestinal permeability, respectively)
increase in the number of study subjects needed. Other (11,12). Dispositional characteristics of HV drugs include
factors influencing sample size include the study design and extensive presystemic metabolism, low bioavailability, high
the expected within-subject variability. For example, a acid lability, and high lipophilicity (6). Consequently, plasma
replicate four-way crossover BE study design, in which each concentrations of these HV drugs are often very low. In such
subject receives the test and reference products twice, situations, it may not be possible to accurately characterize
requires fewer subjects than a two-way crossover BE study PK profiles, with the result that the within-subject variability
design. As within-subject variability increases, the number of of BE measures can exceed 30% (13). As the FDA
discourages unnecessary human testing, these observations
raise questions about whether large numbers of subjects

Table I. The Number of Study Subjects Required to Show BE with

80% Power is a Function of Within-Subject Variability and GMR
(Sample Size Estimations are for the Case σWT =σWR and σD =0)

Sample size Sample size for

for a two-way a four-way
Within- GMR crossover fully replicated
subject %CV (%) study crossover

15 100 10 6
105 12 8
110 20 12
30 100 32 18
105 38 20
110 68 36
Fig. 1. The 80–125% BE limits are represented along the x-axis as 45 100 66 34
two “goal posts.” The BE limits are compared to the hypothetical 105 80 42
90% CIs of the test/reference BE measure GMRs for two drugs, a 110 142 72
drug with normal variability (Drug A) and an HV drug (Drug B). The 60 100 108 56
90% CIs of the two drugs are represented by colored bars. For drug 105 132 66
A (normal variability), the 90% CI (green bar) meets the BE limits. 110 236 118
For drug B (HV), the 90% CI (red bar) fails to meet the acceptance 75 100 156 80
limits. As the width of the CI is influenced by the number of study 105 190 96
subjects, in the hypothetical case of drug B, it is likely that the study 110 340 172
would have met the BE limits if more subjects had been used
Implementation of a Reference-Scaled Average Bioequivalence

should be used in BE studies of drug products for which high & The use of reference scaling should be explored. A
variability does not appear to impact safety and efficacy. It limit on the GMR should be used along with
should be stressed that this concern pertains to high PK reference scaling.
variability due to drug substance dispositional characteristics
and has nothing to do with the formulation performance FDA’s OGD agreed to initiate a series of simulation
assessment that is the key question in BE comparisons. studies to determine whether such an approach was
An additional concern is that the large sample sizes feasible and, if so, how to optimize the approach.
needed for BE studies of HV drugs may serve to deter the Subsequently, the OGD formed an interdisciplinary work-
development of new generic products (14,15). For example, a ing group (hereafter Working Group) to investigate the
generic product line may be abandoned because of a high feasibility of applying a reference-scaled average bioequi-
failure rate obtained in in vivo BE studies or it may be valence (RSABE) approach to generic HV drugs. The
necessary to repeat the in vivo studies several times until the Working Group conducted simulations to investigate
outcome of an acceptable BE study is achieved (16). Not only expanding BE limits as a function of reference product
does this situation lead to unnecessary human testing, but it within-subject variability. Based on the simulation study
can also serve to increase the prices of generic drugs, results, the Working Group members presented a tentative
conflicting with the principle that generic drugs should be proposal for an RSABE method to the Advisory Com-
relatively low-cost because they are not subject to extensive mittee for Pharmaceutical Science and Clinical Pharma-
and unnecessary clinical testing. cology in April of 2006. Speakers from academia and from the
A final concern is that an HV reference product may FDA discussed the advantages and limitations of the following
not be shown to be bioequivalent to itself in a crossover approaches: (a) implementing several different scaling
study using a relatively modest number of subjects (e.g., methods (24–27); (b) recommending a minimum number
18–40) (13). This concern is supported by data from in of subjects for RSABE (25,26); and (c) including a constraint on
vivo studies of brand-name (reference drug product) the GMR of the BE study (25–28). The Advisory Committee’s
formulations of chlorpromazine and verapamil, both of recommendations at the meeting’s conclusion are summarized
which were characterized by high within-subject PK below (29).
variability exceeding 30% (17,18). Such findings raise the & The Advisory Committee was not asked to comment
question of whether the 80–125% BE limits are too stringent on the merits of RSABE as it had previously
for HV drug products. advocated investigating the approach in April 2004.
& The Advisory Committee agreed that these studies
EVOLUTION OF A NEW BE APPROACH FOR HV should enroll a minimum number of subjects. The
GENERIC DRUGS AT THE FDA vote was split as to whether this number should be 24
or 36.
The issues surrounding BE evaluation of HV drugs & The majority of Advisory Committee members either
and proposals for modifying the BE approach for such voted against a GMR constraint or abstained from
products were discussed over many years within the voting on this issue.
pharmaceutical sciences community, in the literature, and
at various national and international venues (17–21). In FDA’s OGD considered the Advisory Committee’s
2004, FDA considered seriously whether an alternative recommendations in finalizing the RSABE approach for HV
BE approach to the “one-size-fits-all” paradigm should be drugs (30).
developed for generic HV drugs. In April of that year,
FDA’s Office of Generic Drugs (OGD) brought the issue FDA’S RECOMMENDATIONS FOR RSABE STUDIES
of optimizing BE study designs for HV drugs before its OF HV DRUGS
Pharmaceutical Science and Clinical Pharmacology Advi-
sory Committee (hereafter Advisory Committee). Speak- Throughout this section, the three basic approaches used
ers from the FDA, academia, and industry speakers for determining if two products are bioequivalent will be
presented the various issues surrounding BE evaluation illustrated using a series of inequalities based on population
of HV drugs to the Advisory Committee (22). Among the parameters.
proposals debated were whether to change the BE limits,
for example, using a GMR constraint of 80–125% or Average Bioequivalence
expanding the 90% CI limits to 70–143% (23) versus
whether to use reference product within-subject PK The usual manner of statistically analyzing BE study data
variability to scale the BE limits (16). Following deliber- is by the average bioequivalence (ABE) approach. The
ations, the Advisory Committee made several recommen- acceptance of BE is stated if the difference between the
dations (22). logarithmic means is between preset regulatory limits, as
shown below:
& There is a need to understand where the variability
originated. Prior knowledge of all biostudies may ðμT
μR Þ2  θ2A
help set more appropriate specifications to make
decisions.
& There is an undue reliance on the use of 80–125% where μT is the population average response of the log-
BE limits for all drug products. As such, a paradigm transformed measure for the test (T) formulation, μR is the
shift for HV drugs is in order. population average response of the log-transformed
 Davit et al.

measure for the reference (R) formulation, and θA is equal where σ2WR is the population within-subject variance of the
ÞÞ2
to ln(1.25). reference formulation, θS ¼ ðlnðσ1:25
2 is the BE limit, and σ2W0
As −ln(1.25)=ln(0.8), the BE limits are: W0
is a predetermined constant set by the regulatory agency.
lnð0:8Þ  ðμT
μR Þ  lnð1:25Þ Under this model, the implied limits (which represent
FDA’s desired consumer risk model) on μT − μR are:
 
σWR σWR
Individual Bioequivalence
lnð1:25Þ  μT
μR  lnð1:25Þ
σW0 σW0
For a time, the FDA worked toward implementing an If σWR =σW0, the implied limits are equal to the
individual bioequivalence (IBE) approach for studies submit- standard unscaled BE limits of ±ln(1.25) (0.80 to 1.25). If
ted to New Drug Applications (NDAs) and Abbreviated New σWR >σW0, the implied limits are wider than the standard
Drug Applications (ANDAs, for generic drugs). It was limits. If σWR <σW0, the implied limits are narrower than
argued that requiring drug products to meet an IBE rather the standard limits.
than an ABE criterion would improve formulation switch- The FDA recommends a mixed scaling approach, as the
ability (31,32). The proposed criterion for acceptable IBE Agency has determined that it is acceptable for the implied
included the (a) comparison of test and reference means; (b) limits to be wider than the standard limits only when σWR is
comparison of within-subject variances; (c) assessment of large (as for HV drugs). The mixed scaling model is as shown
subject-by-formulation interaction; and (d) ability to scale the below.
BE limits if within-subject variability following the adminis- T and R are considered BE if:
tration of the reference product exceeded a predetermined
value (15). An additional justification offered for implement- ðμT
μR Þ2 ðlnð1:25ÞÞ2
ing IBE was that the ability to scale BE limits would address  if σWR  σW0
σ2W0 σ2W0
the concerns associated with BE studies of HV drugs (33–35).
Under IBE, the inequality used to determine if two and if:
products are bioequivalent is as follows:

2  ðμT
μR Þ2 ðlnð1:25ÞÞ2
2
ðμT
μR Þ þ σ2D þ σWT
σ2WR  if σWR > σW0
 θI σ2WR σ2W0
σ2WR
FDA sets the value of σW0 at 0.25 (30,40). Under this
where σ2D is the population subject-by-formulation interac- mixed scaling model and with σW0 =0.25, the implied limits on
tion variance components, σ2WT is the population within- μT −μR are as depicted in Fig. 2.
subject variance of the test formulation, σ2WR is the Direct implementation of FDA’s desired consumer risk
population within-subject variance of the reference formula- model is impossible because σWR is a characteristic of the
tion, and θI is the BE limit for IBE. entire population and thus not directly measured in any
From 1999 to 2001, at the FDA’s request, the pharma- particular study. Therefore, FDA has proposed an implemen-
ceutical industry applied the IBE study design and analysis to tation algorithm (described in the “FDA Draft Guidance for
NDAs and ANDAs for modified-release drug products (36).
The IBE was used to evaluate the BE of modified-release
drug products because it was thought that, due to the relative
complexity of modified-release formulations, the likelihood
was greatest of detecting possible subject-by-formulation
interactions with these types of drug products. However,
analysis of these data failed to detect the presence of clinically
significant subject-by-formulation interactions (37). The FDA
concluded that expansion of the IBE approach to all BE
studies submitted for marketing approval was not warranted
and subsequently halted its implementation (38). Notably,
during a discussion of IBE at its November 29, 2001 meeting,
the Advisory Committee for Pharmaceutical Science and
Clinical Pharmacology suggested that the FDA consider
applying reference scaling to ABE studies of HV drugs (39).

Reference-Scaled Average Bioequivalence

Fig. 2. Implied BE limits are plotted as a function of the population
The RSABE for both AUC and Cmax is evaluated as reference product within-subject variability of the BE measure. When
shown below: σW0 ≤0.25, for an acceptable BE study, the 90% CI of the BE
measure test/reference GMRs must fall within the 80–125% limits.
ðμ T
μ R Þ2 When σW0 >0.25, the implied BE limits scale as reference product
 θS within-subject variability increases. The slope of this portion of the
σ2WR curve is determined by the value of σW0
Implementation of a Reference-Scaled Average Bioequivalence

is ≥30% and, as such,
is determined

 by using the conversion
formula of s2 ¼ ln CV 2 þ 1 .
FDA determined that there are advantages to choosing
an sWR cutoff somewhat larger than σW0. The Working Group
conducted a series of simulations of BE study results
(1,000,000 per condition) to compare the effects of applying
different values to σW0. The study results showed that using a
σW0 of 0.25 (1) increased the study power compared to ABE,
without causing relatively large numbers of studies to pass
when the GMR≥1.2 and (2) resulted in a lower inflation of
type I error compared to a σW0 of 0.294 (42,43).
The FDA recommends a secondary point estimate con-
straint criterion of 0.8 to 1.25 on the GMR (30,40). As it is
possible that, using RSABE, two products can be shown to be
bioequivalent but have an estimated GMR outside of the 0.8 to
1.25 range, it is thought that use of the secondary point estimate
constraint will improve the confidence of clinicians and patients
Fig. 3. Decision tree showing the process by which FDA’s OGD (15,28). Several simulation studies investigating the relationship
decide whether it is suitable to use RSABE or unscaled ABE. The between FDA’s recommended RSABE approach and study
first condition to be met is that the study protocol must state a priori power have shown that, when within-subject variability exceeds
that RSABE will be the method of statistical analysis 50–60%, the point estimate constraint becomes the dominant
regulatory criterion rather than the scaling (43–45). It has also
Industry, Bioequivalence Recommendations for Progesterone been shown that applying the point estimate constraint can
Oral Capsules”) and discussed in more detail later in this increase the sample size needed to show BE for drugs with
paper. In the FDA-recommended implementation algorithm within-subject variability >50–60% (45). Thus, using a point
for mixed scaling studies, the observed within-subject vari- estimate constraint results in a situation where the benefits of
ability of the reference sWR (determined in the BE study) is using RSABE are reduced when applied to drugs with very high
compared to a cutoff value of 0.294, above which reference within-subject variability.
scaling is used. This implementation reduces the type I error However, it has also been argued that, without a point
(defined relative to FDA’s desired consumer risk model) estimate constraint in effect, the permissiveness of the RSABE
when the within-subject variability is near σW0. approach can become excessively high (42). Another argument
favoring incorporating a point estimate constraint is that it has a
long history of successful application by regulatory agencies as a
FDA RECOMMENDATIONS FOR DESIGNING RSABE BE study acceptance criterion (15). Using a point estimate
STUDIES constraint on the GMR as a criterion for study outcome
acceptance is used by the Therapeutic Products Directorate of
The reference product is administered twice in order to Health Canada for Cmax in all BE studies (46) and, until 2003, by
determine its within-subject variability (30). As such, the BE the FDA for both AUC and Cmax in BE studies that were
study can use either a partial replicate (three-way crossover: conducted under fed conditions.
RTR, RRT, or TRR) or full replicate (four-way crossover:
RTRT or TRTR) design, but should enroll a minimum of 24
subjects (40,41). The FDA recommends an sWR cutoff value APPLICATION OF RSABE TO REGULATORY
of 0.294, at or above which reference scaling is permitted and REVIEW OF ANDAS FOR GENERIC HV DRUGS
below which the unscaled limits of 0.8 to 1.25 are applied
(30,40). The selection of 0.294 as the variation at which use of Using the RSABE approach recommended by the FDA,
reference scaling of the limits is consistent with the general two products are bioequivalent when the 95% upper confidence
2
understanding that drugs may be considered HV if the within- bound for ðμTσ
2 μR Þ is ≤θS or, equivalently, a 95% upper
WR
subject coefficient of variation (%CV) observed in the study confidence bound for ðμT
μR Þ2
θS σ2WR is ≤0. In

Table II. Summary BE Statistics for Two HV Generic Drugs, Analyzed Using RSABE or Unscaled ABE, Depending upon the Value of sWR

Summary of statistical analysis-scaled data: least-square geometric means, point estimates and 90% confidence intervals
Parameter T/R ratio Lower 90%CL Upper 90%CL s2WR sWR Criteria Method Outcome
bound used
Drug A: fed bioequivalence study, N=43
AUC0−t 1.17 93.91 132.91 0.5723 0.7565 −0.2892 Scaled/PE Pass
AUC∞ 1.11 94.64 124.91 0.3452 0.5875 −0.1767 Scaled/PE Pass
Cmax 1.19 100.47 133.26 0.3768 0.6138 −0.1684 Scaled/PE Pass
Drug B: fasting bioequivalence study, N=36
AUC0−t 0.99 89.82 109.53 0.07425 0.2725 −0.03914 Unscaled Pass
AUC∞ 1.02 92.57 111.77 0.06561 0.2561 −0.03126 Unscaled Pass
Cmax 0.98 85.70 110.54 0.1069 0.3270 −0.05294 Scaled Pass
 Davit et al.

Case Studies

Two examples, provided in Table II, show how FDA

applies RSABE to studies of HV generic drugs. Both cases
are of approved generic drugs, for which the BE studies used
three-way partial replicate study designs. As FDA scientific
reviewers conduct their own independent calculations in
performing BE analysis, the data in Table II are from FDA’s
Fig. 4. Outcome of the review of 64 RSABE studies reviewed at the calculations, performed using SAS® Version 9.2.
OGD. All but two of the studies met the BE acceptance criteria. For As shown in Table II, in the case of drug A, the first step
36 of these studies, there were no other deficiencies related to the BE of the analysis determined that values of sWR exceeded the
portion of the ANDA submission. For 26 studies, other miscellaneous
regulatory cutoff of 0.294 for AUC0−t, AUC∞, and Cmax.
deficiencies unrelated to RSABE analysis (most often related to the
Thus, the reviewer proceeded to use RSABE for all three BE
bioanalytical method used) were present in the BE portion of the
submission. One study did not meet the point estimate constraint and measures. Table II shows that the calculated 95% upper
thus the study did not meet the RSABE acceptance criteria. Another confidence bound values for all three parameters were <0.
study could not be analyzed using RSABE because the sWR for Cmax Therefore, the product met the first BE acceptance criterion.
was <0.294. However, when the TOST was applied, the 90% CI for In addition, the point estimates for all three BE measures are
Cmax did not fall within the 80–125 limits; thus, this study did not meet within 0.8 and 1.25. Thus, in this study, drug A met the two
the ABE acceptance criteria RSABE acceptance criteria and was deemed bioequivalent to
its corresponding reference.
addition, the GMR of the two products should fall between 0.8 The analysis of drug B, shown in Table II, was conducted
in a different manner because it was not suitable to use
and 1.25.
RSABE for all BE measures. In the first part of the analysis,
Steps for Conducting an RSABE Analysis sWR for Cmax exceeded 0.294 and values of sWR for AUC0−t
and AUC∞ were <0.294. Thus, ABE with the TOST
It should be noted that we do not know the values of the procedure was used for AUC, for which the 90% CIs met
above population parameters and can never know their the 80–125% limits. RSABE analysis was only applied to
values. What we can and do in actuality is calculate CIs Cmax, which met both acceptance criteria, as the 95% upper
around the BE metrics, log(AUC) or log(Cmax). confidence bound was <0 and the test/reference GMR point
FDA posted a Guidance for Industry providing step- estimate was within 0.8 and 1.25. The case of drug B
by-step instructions on how to statistically analyze BE illustrates the important points that (a) both ABE and
study data using RSABE (40). The intention to use RSABE analysis can be used for data from the same study
RSABE for an HV drug should be stated a priori in the depending on BE measure variability and (b) there is no
study protocol. The first step in the analysis is to determine penalty if the applicant uses a three-way or four-way study
sWR, the within-subject standard deviation (SD) of the design with the intent of using RSABE but sWR fails to make
reference product estimated from the study, for the BE the 0.294 cutoff.
measures AUC and Cmax. If sWR <0.294, then the TOST
procedure should be used to determine ABE for the BE Statistics on BE Study Designs in Recent ANDA Submissions
measure. If sWR ≥0.294, then RSABE can be used for the of Generic HV Drugs
BE measure. Figure 3 illustrates the decision-making
process that the FDA uses when determining whether it Since 2007, the FDA has evaluated 46 ANDAs containing
is suitable to use RSABE for a generic drug. 64 studies in which RSABE was applied. Figure 4 displays the
Once it has been decided that RSABE can be applied to statistics on these studies. Sixty-two of these studies met the
a BE measure, the next step is to determine the 95% upper RSABE acceptance criteria, although some studies were consid-
confidence bound for ðμT
μR Þ2
θS σ2WR based on Howe’s ered “incomplete” for other reasons, many due to questions
Approximation I (47). The test and reference product are relating to the bioanalytical method (48). Only two studies were
concluded to be bioequivalent if: unacceptable because they did not meet the acceptance criteria.
In the first instance, as one BE measure, sWR, was <0.294, the
(a) the 95% upper confidence bound for ðμT
μR Þ2
RSABE analysis could not be applied. Thus, it was necessary to
θS σ2WR is ≤0 AND apply ABE and the TOST procedure, but the study did not meet
(b) the test/reference GMR in the study falls within the acceptance limits because the upper confidence limit
[0.8, 1.25]. exceeded 125%. In the second instance, the study data met the

Table III. Number of Failed BE Studies Reviewed at FDA’s OGD Since the “All Bioequivalence Studies” Rule Became Effective in July 2009

BE Studies ANDAs

Description Number Percent of total Number Percent of total

Within-subject variability of AUC and Cmax ≥30% 92 45 45 37

Within-subject variability of AUC and Cmax <30% 113 55 76 63
Totals 205 100 121 100
Implementation of a Reference-Scaled Average Bioequivalence

Table IV. Changes Made to the Study Design or Formulation to all BE studies conducted on the same drug product formu-
Achieve Successful Pivotal BE Studies of HV Drugs after the Initial lation as the one intended for marketing (50,51), it is hoped
BE Studies Failed to Meet the Acceptance Criteria that such submissions will provide valuable information about
the development of HV drug formulations.
No. of To test the hypotheses that a high percentage of BE
Change made ANDAs Percent
studies fail because they are of (a) HV drugs and (b) do not
Increase in sample size in 41 91 enroll enough subjects, we surveyed all failed BE studies of
two-way crossover BE study generic drugs reviewed from the time that the Final Rule
Changed to three-way study design and 3 7 went into effect until the time of this writing. As shown in
RSABE Table III, 205 failed BE studies were reviewed during this
Reformulated 1 2 time period. Of these, 92 of these failed BE studies were of
Total 45 100 HV drugs, representing 45% of the total. This appears to be a
disproportionately high percentage of failed BE studies, as
HV drugs are thought to represent about 20% of drug
criteria for applying the RSABE analysis, and the 90% upper products submitted to FDA in ANDAs.
confidence bound for all three BE measures was <0, but the In addition, we surveyed the reasons why these 92 BE
GMRs of all BE measures exceeded 1.25. Interestingly, the studies submitted in 45 ANDAs failed. The results confirm
applicant conducted a second successful study on this product that a high percentage of studies of HV drugs appear to fail
which met the BE limits using RSABE analysis without the BE acceptance criteria due to underpowering. As shown
reformulating or increasing the number of subjects. in Table IV, for all but one of 45 different ANDAs, generic
As of this writing, FDA has fully approved four HV products that initially failed one or more two-way crossover
generic drug products for which RSABE was used. In studies with TOST analysis met the BE limits in subsequent
addition, one HV generic drug product for which RSABE studies when power was increased, either by enrolling more
was used was tentatively approved (because a patent related subjects or by changing to an RSABE design.
to the reference product has not yet expired).
One of the major reasons offered to justify moving away
from the “one-size-fits-all” BE limits for HV drugs was RSABE APPROACHES FOR HV DRUGS
concern that BE studies would be unnecessarily repeated IN OTHER JURISDICTIONS
because of underpowering. Until the “All Bioequivalence
Studies” Final Rule became effective in 2009, it was not In August 2010, the European Medicines Agency
possible for the FDA to assess the prevalence of failed and (EMA) issued a new Guideline on the Investigation of
repeated BE studies on HV drugs because generic applicants Bioequivalence, recommending a scaling approach for BE
were not required to submit failed BE studies (49). As a assessment of HV drugs (15,42,52,53). EMA recommends
result, FDA’s assessment of the scope of BE problems in that BE limits scale with variability only for Cmax and only up
regulatory submissions of HV drugs was possibly biased pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
to a maximum within-subject
2
ffi
variability value of 50% (where
because the Agency did not see the results of failed BE %CV is defined as 100 esWR
1 ), after which they remain
studies. As applicants are now required to submit to ANDAs constant at the static limits of 69.84% to 143.19%. In BE

Fig. 5. This timeline summarizes the evolution of the RSABE approach at the FDA. The FDA first
proposed reference scaling as a component of IBE analysis, although the FDA subsequently
decided not to continue implementing IBE. Three years later, in 2004, the FDA began developing
RSABE using the scaling approach from IBE. The first ANDA using RSABE was submitted in
2007. The first tentative approval of an HV drug for which RSABE was successfully applied was in
2009. The first full approval of an HV drug supported by acceptable RSABE was in 2011. From
2004 to 2010, FDA scientists published several peer-reviewed scientific articles on RSABE and
made a number of presentations at national and international venues. The FDA has also posted a
Draft Guidance for Industry with a step-by-step description of how to perform RSABE
 Davit et al.

studies, Cmax is usually more variable than AUC (15,54,55). 30% is substantially reduced by the evaluation approach used
Either a three-way or four-way crossover design, in which the by BE review scientists in FDA’s OGD. Although the
reference product is administered twice, can be used in the RSABE analysis is triggered when sWR ≥0.294 in a three-
BE study, in order to determine the reference product within- way or four-way replicated study, review scientists are
subject variability for scaling. The applicant should demon- encouraged to evaluate the appropriateness of reference
strate that the within-subject variability of the reference scaling in the context of other available information on the
product Cmax is >30%. The extent of the widening of the PK variability of a given drug. The three-level review process
acceptance limits is defined based upon the within-subject for all BE submissions in the OGD (reviewer, team leader,
variability seen in the BE study using scaled average BE and division director) also encourages critical evaluation of
according to [U, L]=exp [±k×sWR], where U is the upper data and consistent policy application.
limit of the acceptance range, L is the lower limit of the
acceptance range, k is the regulatory constant set to 0.760, FUTURE DIRECTIONS
and sWR is the within-subject SD of the log-transformed
values of Cmax of the reference product (15,42,52). Using Lack of harmonization in setting the regulatory constant
these values, the regulatory limits range from 80% to 125% and cutoff for RSABE could potentially be resolved by using
when the within-subject %CV is 30% to 69.84–143.19% when the approach for all generic products, as was formerly
the within-subject variability is 50%. EMA also recommends proposed for IBE. Such an approach could also be more
a point estimate constraint on the Cmax GMR if scaling is favorable to ensuring drug product switchability. The FDA
used. recently proposed to adapt RSABE for BE studies of narrow
The South African Medicines Control Council suggests therapeutic index (NTI) drugs (58). Since NTI drugs are
that the RSABE may be used for AUC and Cmax if specified a characterized by low within-subject variability, this would
priori in the protocol and justified by sound scientific principles; effectively result in narrower BE limits.
applicants also have the option of applying static BE limits with
a wider acceptance interval for HV Cmax (56,57). CONCLUSIONS

Figure 5 provides a timeline showing how RSABE

CONTROVERSIES evolved at the FDA, from inception to application to ANDAs
for generic HV drugs. FDA implemented RSABE for generic
One of the more controversial aspects of the RSABE HV wide therapeutic index drugs to ease regulatory burden.
approach is the selection of the value of σW0, the variability at In the RSABE approach recommended by the FDA, the
which the BE limits can potentially begin to scale. The FDA within-subject variability of the reference product BE mea-
sets σW0 at 0.25, but also proposes an implementation sure is determined in a three-way or four-way replicated
algorithm that does not permit BE limits to scale until the study which should use at least 24 subjects. The implied BE
observed within-subject variability reaches the cutoff value of limits scale to reference within-subject variability once
0.294 for sWR. Endrenyi and Tothfalusi (44) argue that the σWR =σW0 =0.25 or greater. However, to preserve an accept-
FDA’s approach to RSABE is discontinuous and conducted able (<5%) type I error rate, applicants cannot apply
simulations claiming that, when σW0 ranges from 0.246 to reference scaling to calculate BE limits until the reference
0.294, consumer risk is substantially larger than 5% and the product within-subject SD in the BE study is at least 0.294.
regulatory uncertainty about a decision on acceptance or Acceptance criteria are that the 95% upper confidence bound
rejection is enhanced. This claim is based on a different for the GMR test/reference BE measure ratio is ≤0 and the
definition of consumer risk than FDA uses. Endrenyi and point estimate of the BE measure GMR is within 0.8 to 1.25.
Tothfalusi define consumer risk relative to BE limits of 80– The approach has been applied successfully to ANDAs, with
125% (44). FDA’s position is that a true population variance 64 RSABE studies submitted and reviewed since 2007. To
of σWR ≥0.25 justifies using BE limits that are wider than 80– date, FDA has granted four full marketing approvals and one
125%. The use of sWR ≥0.294 as part of FDA’s implementa- tentative approval for several diverse HV generic drug
tion method affects the producer risk—there will be a higher products based on the RSABE analysis. The RSABE is not
probability of failing a product that is acceptable according to a “one-size-fits-all” approach, but rather a system that self-
FDA’s consumer risk model. adjusts to the characteristics of the reference. As a result,
Results of simulations conducted by members of the HV RSABE could be used to potentially set appropriately strict
Drug Working Group support the position that using a cutoff BE limits for all products.
value of 0.294 for sWR maintains an acceptable type I error
rate relative to FDA’s desired consumer risk model. This can
be illustrated by using a value of 0.29399 for σWR. In this case, ACKNOWLEDGMENTS
the acceptable limits on μT −μR are ±ln(1.25)×0.29399/0.25=
±0.262408. Results of simulated three-way partial replicate The authors thank Hoainhon Nguyen Caramenico,
studies of 36 subjects result in 46,434 conclusions of BE out of Parthapratim Chandaroy, Svetlana Cherstniakova, Li
1,000,000 simulated studies or an estimated type I error rate Gong, Heather Lim, Anitha Palamakula, and Xiaofai
of 0.046434 relative to FDA’s desired consumer risk model. It Wang for contributing data to the examples described
is important to carefully note the distinction between σWR throughout the manuscript. The opinions stated in this article
and sWR in the analysis of the scaling method. represent those of the authors and do not necessarily
Finally, likelihood of regulatory uncertainty in evaluating represent the official position of the US Food and Drug
RSABE study results when within-subject variability is near Administration.
Implementation of a Reference-Scaled Average Bioequivalence

REFERENCES 19. Blume H, McGilvery I, Midha K. Bio-International 94 confer-

ence on bioavailability, bioequivalence and pharmacokinetic
studies. Eur J Pharm Sci. 1995;3:113–24.
1. Federal Food, Drug, and Cosmetic Act, Chapter V, Subchapter 20. Blume HH. “One-size-fits-all” in bioavailability and bioequiva-
A, Drugs and Devices Section 355(j)(8)(B)(i). lence testing. Int J Clin Pharmacol Ther. 2009;47:419–20.
2. Schuirmann DJ. A comparison of the two one-sided tests 21. Yu LX. Bioequivalence of highly variable drugs: issues and
procedure and the power approach for assessing the equivalence challenges. In: FDA Advisory Committee for Pharmaceutical
of average bioavailability. J Pharmacokinet Biopharm. Sciences and Clinical Pharmacology Meeting Transcript. Apr 14
1987;15:657–80. 2004. US Food and Drug Administration Dockets. http://
3. Westlake WJ. Bioequivalence testing: a need to rethink. www.fda.gov/ohrms/dockets/ac/04/transcripts/4034T2.htm.
Biometrics. 1981;37:589–94. Accessed 23 Jan 2012.
4. Blume HH, Midha KK. Bio-International 92, conference on 22. Executive Secretary, Advisory Committee for Pharmaceutical
bioavailability, bioequivalence, and pharmacokinetic studies. J Science and Clinical Pharmacology. Summary meeting minutes.
Pharm Sci. 1993;82:1186–9. In: FDA Advisory Committees meeting materials. US Food and
5. Shah VP, Yacobi A, Barr WH, Benet LZ, Breimer D, et al. Drug Administration Dockets. Apr 14 2004. http://www.fda.gov/
Evaluation of orally administered highly variable drugs and drug ohrms/dockets/ac/04/minutes/4034M1.htm. Accessed 23 Jan
formulations. Pharm Res. 1996;13:1590–4. 2012.
6. Davit BM, Conner DP, Fabian-Fritsch B, Haidar SH, Jiang X, et 23. Haidar SH. FDA perspective. In: FDA Advisory Committee for
al. Highly variable drugs: observations from bioequivalence data Pharmaceutical Sciences and Clinical Pharmacology meeting
submitted to the FDA for new generic drug applications. transcript . US Food and Drug Administration Dockets. Apr 14
AAPSJ. 2008;10:148–56. 2004. http://www.fda.gov/ohrms/dockets/ac/04/transcripts/
7. FDA Guidance for Industry, statistical approaches to establish- 4034T2.htm. Accessed 23 Jan 2012.
ing bioequivalence. US Department of Health and Human 24. Midha K. Bioequivalence of highly variable drugs. In: FDA
Services Food and Drug Administration Center for Drug Advisory Committee for Pharmaceutical Sciences and Clinical
Evaluation and Research. 2001. http://www.fda.gov/downloads/ Pharmacology meeting transcript. US Food and Drug
Drugs/GuidanceComplianceRegulatoryInformation/Guidances/ Administration Dockets. Apr 6 2006. http://www.fda.gov/ohrms/
UCM070244.pdf. Accessed 23 Jan 2012. dockets/ac/06/transcripts/2006-4241t2-01.pdf. Accessed 23 Jan
8. Patterson SD, Zariffa NMD, Montague TH, Howland K. Non- 2012.
traditional study designs to demonstrate average bioequivalence 25. Haidar SH. Evaluation of a scaling approach for highly variable
for highly variable drug products. Eur J Clin Pharmacol. drugs. In: FDA Advisory Committee for Pharmaceutical
2001;57:663–70. Sciences and Clinical Pharmacology meeting transcript. US
9. Phillips KF. Power of the two one-sided tests procedure in Food and Drug Administration Dockets. Apr 6 2006. Accessed
bioequivalence. J Pharmacokinet Biopharm. 1990;18:137–44. 23 Jan 2012.
10. Benet L. Why highly variable drugs are safer. In: FDA Advisory 26. Davit BM. FDA’s proposal. In: FDA Advisory Committee for
Committee for Pharmaceutical Sciences and Clinical Pharmaceutical Sciences and Clinical Pharmacology meeting
Pharmacology meeting transcript. US Food and Drug transcript. US Food and Drug Administration Dockets. Apr 6
Administration Dockets. Apr 14 2004. http://www.fda.gov/ 2006. http://www.fda.gov/ohrms/dockets/ac/06/transcripts/2006-
ohrms/dockets/ac/04/transcripts/4034T2.htm. Accessed 23 Jan 4241t2-01.pdf. Accessed 23 Jan 2012.
2012. 27. Endrenyi L. Some issues on the determination of bioequivalence
11. Cook JA, Davit BM, Polli JE. Impact of Biopharmaceutics for highly variable drugs. In: FDA Advisory Committee for
Classification System-based biowaivers. Mol Pharmaceutics. Pharmaceutical Sciences and Clinical Pharmacology meeting
2010;7:1539–44. transcript. US Food and Drug Administration Dockets. Apr 6
12. Amidon GL. Sources of variability: physicochemical and gastro- 2006. http://www.fda.gov/ohrms/dockets/ac/06/transcripts/2006-
intestinal. In: FDA Advisory Committee for Pharmaceutical 4241t2-01.pdf. Accessed 23 Jan 2012.
Sciences and Clinical Pharmacology meeting transcript. US Food 28. Benet L. Therapeutic considerations of highly variable drugs. In:
and Drug Administration Dockets. Apr 14 2004. http://www.fda. FDA Advisory Committee for Pharmaceutical Sciences and
gov/ohrms/dockets/ac/04/transcripts/4034T2.htm. Accessed 23 Clinical Pharmacology meeting transcript. US Food and Drug
Jan 2012. Administration Dockets. Apr 6 2006. http://www.fda.gov/ohrms/
13. Conner DP. Bioequivalence methods for highly variable drugs and dockets/ac/06/transcripts/2006-4241t2-01.pdf. Accessed 23 Jan
drug products. In: FDA Advisory Committee for Pharmaceutical 2012.
Sciences and Clinical Pharmacology meeting transcript. US Food and 29. Executive Secretary, Advisory Committee for Pharmaceutical
Drug Administration Dockets. Aug 5 2009. http://www.fda.gov/down Science and Clinical Pharmacology. Summary meeting minutes.
loads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/ In: FDA Advisory Committees meeting materials. US Food and
AdvisoryCommitteeforPharmaceuticalScienceandClinicalPharmaco Drug Administration Dockets. Apr 6 2006. http://www.fda.gov/
logy/UCM179891.pdf. Accessed 23 Jan 2012. ohrms/dockets/ac/06/minutes/2006-4241m2.pdf. Accessed 23 Jan
14. Diliberti CE. Why bioequivalence of highly variable drugs is an 2012.
issue. In: FDA Advisory Committee for Pharmaceutical Sciences 30. Haidar SH, Davit BM, Chen M-L, Conner D, Lee LM, et al.
and Clinical Pharmacology meeting transcript. US Food and Bioequivalence approaches for highly variable drugs and drug
Drug Administration Dockets. Apr 14 2004. http://www.fda.gov/ products. Pharm Res. 2008;25:237–41.
ohrms/dockets/ac/04/transcripts/4034T2.htm. Accessed 23 Jan 31. Hauck WH, Hyslop T, Chen M-L, Patnaik R, Williams RL.
2012. Subject-by-formulation interaction in bioequivalence: conceptual
15. Tothfalusi L, Endrenyi L, Arieta AG. Evaluation of bioequiva- and statistical terms. Pharm Res. 2000;17:375–80.
lence for highly variable drugs with scaled average bioequiva- 32. Patnaik RN, Lesko LJ, Chen M-L, Williams RL. Individual
lence. Clin Pharmacokinet. 2009;48:725–43. bioequivalence: new concepts in the statistical assessment of
16. Endrenyi L. Bioequivalence methods for highly variable drugs. bioequivalence metrics. Clin Pharmacokinet. 1997;33:1–6.
In: FDA Advisory Committee for Pharmaceutical Sciences and 33. Chen M-L. Background and concepts of individual bioequivalence.
Clinical Pharmacology meeting transcript. US Food and Drug In: FDA Advisory Committee for Pharmaceutical Sciences and
Administration Dockets. Apr 14 2004 http://www.fda.gov/ohrms/ Clinical Pharmacology meeting transcript .US Food and Drug
dockets/ac/04/transcripts/4034T2.htm. Accessed 23 Jan 2012. Administration Dockets. Nov 29 2001. http://www.fda.gov/ohrms/
17. Midha KK, Rawson MJ, Hubbard JW. The bioequivalence of dockets/ac/01/transcripts/3804t2_03_Afternoon_Session.pdf.
highly variable drugs and drug products. Int J Clin Pharmacol Accessed 23 Jan 2012.
Ther. 2005;43:485–98. 34. Patnaik R. Results from replicate design studies in ANDAs. In:
18. Tsang YC, Pop R, Gordon P, Hems J, Spino M. High variability FDA Advisory Committee for Pharmaceutical Sciences and
in drug pharmacokinetic complicates determination of bioequi- Clinical Pharmacology meeting transcript. US Food and Drug
valence: experience with verapamil. Pharm Res. 1996;13:846–50. Administration Dockets. Nov 29 2001. http://www.fda.gov/ohrms/
 Davit et al.

dockets/ac/01/transcripts/3804t2_03_Afternoon_Session.pdf. formulations used for systemic effects. Therapeutic Products

Accessed 23 Jan 2012. Directorate, Canada Minister of Health, Ottawa. 1997. http://
35. Benet L. Individual bioequivalence: have the opinions of the www.hc-sc.gc.ca/dhp-mps/alt_formats/hpfb-dgpsa/pdf/prod
scientific community changed? In: FDA Advisory Committee for pharma/bio-a-eng.pdf. Accessed 23 Jan 2012.
Pharmaceutical Sciences and Clinical Pharmacology meeting 47. Howe WG. Approximate confidence limits on the mean of X+Y
transcript. US Food and Drug Administration Dockets. Nov 29 where X and Y are two tabled independent variables. J Amer
2001. http://www.fda.gov/ohrms/dockets/ac/01/transcripts/ Statist Assoc. 1974;69:789–94.
3804t2_03_Afternoon_Session.pdf. Accessed 23 Jan 2012. 48. Liu Q, Davit BM, Cherstniakova SA, Dandamudi S, Walters JF,
36. Executive Secretary, Advisory Committee for Pharmaceutical et al. Common deficiencies with bioequivalence submissions in
Science and Clinical Pharmacology. Briefing document, bioequi- abbreviated new drug applications assessed by the FDA.
valence criteria research program. In: FDA Advisory Committee AAPSJ. 2011. doi:10.1208/s12248-011-9312-7.
Meeting Materials. US Food and Drug Administration Dockets. 49. US Food and Drug Administration, Department of Health and
Nov. 29 2001. http://www.fda.gov/ohrms/dockets/ac/01/briefing/ Human Services. Final rule, 21 CFR Parts 314 and 320, Docket
3804b2_08_Bioequiv%20Criteria.pdf. Accessed 23 Jan 2012. No. FDA-2003-N-0209. Fed Regist 2009; 74:2849–62.
37. Chen M-L. Results from replicate design studies in NDAs and FDA 50. FDA Guidance for Industry, submission of summary bioequiva-
database. In: FDA Advisory Committee for Pharmaceutical Sciences lence data for abbreviated new drug application. US Department
and Clinical Pharmacology meeting transcript . US Food and Drug of Health and Human Services Food and Drug
Administration Dockets. Nov 29 2001. http://www.fda.gov/ohrms/ Administration Center for Drug Evaluation and Research,
dockets/ac/01/transcripts/3804t2_03_Afternoon_Session.pdf. Silver Spring. 2011. http://www.fda.gov/downloads/Drugs/
Accessed 23 Jan 2012. GuidanceComplianceRegulatoryInformation/Guidances/
38. FDA Guidance for Industry, bioavailability and bioequivalence UCM134846.pdf. Accessed 23 Jan 2012.
studies for orally administered drug products: general consid- 51. US Food and Drug Administration, Department of Health and
erations. US Department of Health and Human Services Food Human Services. Bioequivalence. In: Title 21 Code of Federal
and Drug Administration Center for Drug Evaluation and Regulations, Food and Drugs. Office of the Federal Register
Research. 2003. http://www.fda.gov/downloads/Drugs/ National Archives and Records Administration. Washington;
GuidanceComplianceRegulatoryInformation/Guidances/ 2011. p. 132–3.
UCM070124.pdf. Accessed 23 Jan 2012. 52. European Medicines Agency. Guideline on the investigation of
39. Members of the FDA Advisory Committee for Pharmaceutical bioequivalence. 2010. http://www.emea.europa.eu/docs/en_GB/
Science and Clinical Pharmacology. Discussion of individual document_library/Scientific_guideline/2010/01/
bioequivalence issues. In: FDA Advisory Committee for WC500070039.pdf. Accessed 23 Jan 2012.
Pharmaceutical Sciences and Clinical Pharmacology meeting 53. Morais JA, Lobato Mdo R. The new European Medicines
transcript. US Food and Drug Administration Dockets. Nov 29 Agency Guideline on the investigation of bioequivalence. Basic
2001. http://www.fda.gov/ohrms/dockets/ac/01/transcripts/ Clin Pharmacol Toxicol. 2010;106:251–60.
3804t2_03_Afternoon_Session.pdf. Accessed 23 Jan 2012. 54. Davit B. Bioequivalence of highly variable drugs case studies. In:
40. FDA Draft Guidance for Industry, bioequivalence recommenda- FDA Advisory Committee for Pharmaceutical Sciences and
tions for progesterone oral capsules. US Department of Health and Clinical Pharmacology Meeting Transcript. US Food and Drug
Human Services Food and Drug Administration Center for Drug Administration Dockets. Apr 16 2004. http://www.fda.gov/ohrms/
Evaluation and Research, Silver Spring. 2011. http://www.fda.gov/ dockets/ac/04/transcripts/4034T2.htm. Accessed 23 Jan 2012.
downloads/Drugs/GuidanceComplianceRegulatoryInformation/ 55. Davit BM, Nwakama PE, Buehler GJ, Conner DP, Haidar SH, et
Guidances/UCM209294.pdf. Accessed 23 Jan 2012. al. Comparing generic and innovator drugs: a review of 12 years
41. Davit BM, Conner DP. The United States of America. In: Kanfer of bioequivalence data from the United States Food and Drug
I, Shargel L, editors. Generic drug product development: Administration. Ann Pharmacother. 2009;43:1583–97.
international regulatory requirements for bioequivalence. New 56. Walker RB, Kanfer I, Skinner MF. Bioequivalence assessment of
York: Informa Healthcare; 2010. p. 254–81. generic products: an innovative South African approach. Clin
42. Karalis V, Sylmillides M, Macheras P. Bioequivalence of highly Res Regul Affair. 2006;23:11–20.
variable drugs: a comparison of the newly-proposed regulatory 57. Medicines Control Council. Human medicines guideline,
approaches by FDA and EMA. Pharm Res. 2011. doi:10.1007/ biostudies. MCC Guidelines and Forms. South African
s11095-011-0651-y. Registrar of Medicines. Capetown. 2007. http://www.
43. Haidar SH, Makhlouf F, Schuirmann DJ, Hyslop T, Davit B, mccza.com/dynamism/default_dynamic.asp?grpID=
et al. Evaluation of a scaling approach for the bioequivalence of 30&doc=dynamic_generated_page.asp&categID=177&
highly variable drugs. AAPSJ. 2008;10:450–4. groupID=30. Accessed 23 Jan 2012.
44. Endrenyi L, Tothfalusi L. Regulatory conditions for the deter- 58. Davit B. FDA Proposal for bioequivalence of generic narrow
mination of bioequivalence of highly variable drugs. J Pharm therapeutic index drug products. In: FDA Advisory Committee
Pharmaceut Sci. 2009;12:138–49. for Pharmaceutical Sciences and Clinical Pharmacology
45. Tothfalusi L, Endrenyi L. Sample size for designing bioequiva- meeting transcript. US Food and Drug Administration
lence studies for highly variable drugs. J Pharm Pharmaceut Sci. Dockets. Jul 26 2011. http://www.fda.gov/downloads/
2011;15:73–84. AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/
46. Health Canada Guidance for Industry, conduct and analysis of AdvisoryCommitteeforPharmaceuticalScienceandClinicalPhar
bioavailability and bioequivalence studies—part A: oral dosage macology/UCM272112.pdf. Accessed 23 Jan 2012.

View publication stats