Part I

Regulatory and Manufacturing Requirements

in Compressed Solid Dosage Forms

© 2004 by CRC Press LLC

 1 Bioavailability and Bioequivalence
Studies for Orally Administered
Drug Products
I. INTRODUCTION drug ingredient or active moiety in the test product should
exhibit the same rate and extent of absorption as the ref-
Demonstration of bioequivalence is a critical requirement erence drug product.
for gaining marketing authorization of new drugs and BA and BE studies are required by regulations,
generics. These requirements are specified in the Code of depending on the type of application being submitted.
Federal Regulations, Title 21, Part 320 (21 CFR Part 320) Under § 314.94, BE information is required to ensure
as they apply to dosage forms intended for oral adminis- therapeutic equivalence between a pharmaceutically
tration. Substantial changes to these guidelines have been equivalent test drug product and a reference listed drug.
contemplated. In the discussion that follows, an overview Regulatory requirements for documentation of BA and BE
of the guidelines that will soon become effective is pro- are provided in Part 320, which contains two subparts.
vided. Manufacturers contemplating the development of Subpart A covers general provisions, while Subpart B
new drug products should pay close attention to the contains 18 sections delineating the following general BA
changes described in Sections II.B. and II.C. regarding and BE requirements:
bioavailability (BA) or bioequivalence (BE) studies.
• Requirements for submission of BA and BE
II. BACKGROUND data (§ 320.21)
• Criteria for waiver of an in vivo BA or BE study
A. GENERAL (§ 320.22)
• Basis for demonstrating in vivo BA or BE
Studies to measure BA and establish BE of a product are (§ 320.23)
important elements in support of investigational new drug • Types of evidence to establish BA or BE
applications (INDs), new drug applications (NDAs), (§ 320.24)
abbreviated new drug applications (ANDAs), and their • Guidelines for conducting in vivo BA studies
supplements. As part of INDs and NDAs for orally admin- (§ 320.25)
istered drug products, BA studies focus on determining • Guidelines for design of single-dose BA studies
the process by which a drug is released from the oral (§ 320.26)
dosage form and moves to the site of action. BA data • Guidelines for design of multiple-dose in vivo
provide estimates of the fraction of the drug absorbed, as BA studies (§ 320.27)
well as its subsequent distribution and elimination. BA • Correlations of BA with an acute pharmacolog-
can generally be documented by a systemic exposure pro- ical effect or clinical evidence (§ 320.28)
file obtained by measuring drug or metabolite concentra- • Analytical methods for an in vivo BA study
tion in the systemic circulation over time. The systemic (§ 320.29)
exposure profile determined during clinical trials in the • Inquiries regarding BA and BE requirements
IND period can serve as a benchmark for subsequent BE and review of protocols by the FDA (§ 320.30)
studies. • Applicability of requirements regarding INDs
Studies to establish BE between two products are (§ 320.31)
important when considering certain changes before • Procedures for establishing and amending a BE
approval for a pioneer product in NDA and ANDA sub- requirement (§ 320.32)
missions, and in the presence of certain postapproval • Criteria and evidence to assess actual or poten-
changes in NDAs and ANDAs. In BE studies, an applicant tial BE problems (§ 320.33)
compares the systemic exposure profile of a test drug • Requirements for batch testing and certification
product to that of a reference drug product. For two orally by the FDA (§ 320.34)
administered drug products to be bioequivalent, the active

© 2004 by CRC Press LLC

4 Handbook of Pharmaceutical Manufacturing Formulations: Compressed Solid Products

• Requirements for in vitro batch testing of each before marketing a drug product. The systemic exposure
batch (§ 320.35) profiles of clinical trial material can be used as a bench-
• Requirements for maintenance of records of BE mark for subsequent formulation changes and may be
testing (§ 320.36) useful as a reference for future BE studies.
• Retention of BA samples (§ 320.38)
• Retention of BE samples (§ 320.63) C. BIOEQUIVALENCE

B. BIOAVAILABILITY Bioequivalence is defined in § 320.1 [21 CFR 320.1(e)] as:

Bioavailability is defined in § 320.1 [21 CFR 320.1(a)] as: … the absence of a significant difference in the rate and
extent to which the active ingredient or active moiety in
… the rate and extent to which the active ingredient or pharmaceutical equivalents or pharmaceutical alternatives
active moiety is absorbed from a drug product and becomes available at the site of drug action when admin-
becomes available at the site of action. For drug products istered at the same molar dose under similar conditions
that are not intended to be absorbed into the bloodstream, in an appropriately designed study.
bioavailability may be assessed by measurements
intended to reflect the rate and extent to which the active As noted in the statutory definitions, BE and product
ingredient or active moiety becomes available at the site quality BA focus on the release of a drug substance from
of action. a drug product and subsequent absorption into the sys-
temic circulation. For this reason, similar approaches to
This definition focuses on the processes by which the measuring BA in an NDA should generally be followed
active ingredients or moieties are released from an oral in demonstrating BE for an NDA or an ANDA. Establish-
dosage form and move to the site of action. ing product quality BA is a benchmarking effort, with
From a pharmacokinetic perspective, BA data for a comparisons to an oral solution, an oral suspension, or an
given formulation provide an estimate of the relative frac- intravenous formulation. In contrast, demonstrating BE is
tion of the orally administered dose that is absorbed into usually a more formal comparative test that uses specified
the systemic circulation when compared to the BA data criteria for comparisons and predetermined BE limits for
for a solution, suspension, or intravenous dosage form [21 such criteria.
CFR 320.25(d)(2) and (3)]. In addition, BA studies pro-
vide other useful pharmacokinetic information related to 1. INDs/NDAs
distribution, elimination, effects of nutrients on absorption
of the drug, dose proportionality, linearity in pharmacok- BE documentation may be useful during the IND or NDA
inetics of the active moieties, and, where appropriate, inac- period to establish links between the following:
tive moieties. BA data may also indirectly provide infor-
mation about the properties of a drug substance before • Early and late clinical trial formulations
entry into the systemic circulation, such as permeability • Formulations used in clinical trial and stability
and the influence of presystemic enzymes and transporters studies, if different
(e.g., p-glycoprotein). • Clinical trial formulations and to-be-marketed
BA for orally administered drug products can be doc- drug products
umented by developing a systemic exposure profile • Other comparisons, as appropriate
obtained from measuring the concentration of active ingre-
dients or active moieties and, when appropriate, its active In each comparison, the new formulation or new
metabolites over time in samples collected from the sys- method of manufacture is the test product, and the prior
temic circulation. Systemic exposure patterns reflect formulation or method of manufacture is the reference
release of the drug substance from the drug product and product. The determination to redocument BE during the
a series of possible presystemic and systemic actions on IND period is generally left to the judgment of the sponsor,
the drug substance after its release from the drug product. who may wish to use the principles of relevant guidances
Additional comparative studies should be performed to (see Section II.C.3. and Section II.C.4.) to determine when
understand the relative contribution of these processes to changes in components, composition, or method of man-
the systemic exposure pattern. ufacture suggest that further in vitro and in vivo studies
One regulatory objective is to assess, through appro- should be performed.
priately designed BA studies, the performances of the A test product may fail to meet BE limits because the
formulations used in the clinical trials that provide evi- test product has higher or lower measures of rate and
dence of safety and efficacy [21 CFR 320.25(d)(1)]. The extent of absorption compared with the reference product,
performance of the clinical trial dosage form may be opti- or because the performance of the test or reference product
mized, in the context of demonstrating safety and efficacy, is more variable. In some cases, nondocumentation of BE

© 2004 by CRC Press LLC

Bioavailability and Bioequivalence Studies for Orally Administered Drug Products 5

may arise because of inadequate numbers of subjects in and modified-release drug products approved as NDAs
the study relative to the magnitude of intrasubject vari- or ANDAs in the presence of specified postapproval
ability and not because of the high or low relative BA of changes is provided in the FDA guidance for industry
the test product. Adequate design and execution of a BE titled SUPAC-IR: Immediate Release Solid Oral Dosage
study will facilitate an understanding of the causes of Forms: Scale-Up and Post-Approval Changes: Chemistry,
nondocumentation of BE. Manufacturing, and Controls, In Vitro Dissolution Testing,
Where the test product generates plasma levels sub- and In Vivo Bioequivalence Documentation (November
stantially above those of the reference product, the regu- 1995) and SUPAC-MR: Modified Release Solid Oral Dos-
latory concern is not therapeutic failure but the adequacy age Forms: Scale-Up and Post-Approval Changes: Chem-
of the safety database from the test product. Where the istry, Manufacturing, and Controls, In Vitro Dissolution
test product has levels that are substantially below those Testing, and In Vivo Bioequivalence Documentation (Sep-
of the reference product, the regulatory concern becomes tember 1997). In the presence of certain major changes in
therapeutic efficacy. When the variability of the test prod- components, composition, or methods of manufacture
uct rises, the regulatory concern relates to safety and effi- after approval, in vivo BE should be redemonstrated. For
cacy because it may suggest that the test product does not approved NDAs, the drug product after the change should
perform as well as the reference product, and the test be compared with the drug product before the change. For
product may be too variable to be clinically useful. approved ANDAs, the drug product after the change
Proper mapping of individual dose–response or con- should be compared with the reference listed drug. Under
centration–response curves is useful in situations where § 506A(c)(2)(B) of the Federal Food, Drug, and Cosmetic
the drug product has plasma levels that are either higher Act (the Act) [21 USC 356a(c)(2)(B)], postapproval
or lower than the reference product and are outside usual changes requiring completion of studies in accordance
BE limits. In the absence of individual data, population with Part 320 must be submitted in a supplement and
dose–response or concentration–response data acquired approved by the FDA before distributing a drug product
over a range of doses, including doses above the recom- made with the change.
mended therapeutic doses, may be sufficient to demon-
strate that the increase in plasma levels would not be
accompanied by additional risk. Similarly, population
III. METHODS TO DOCUMENT BA AND BE
dose– or concentration–response relationships observed As noted in § 320.24, several in vivo and in vitro methods
over a lower range of doses, including doses below the can be used to measure product quality BA and establish
recommended therapeutic doses, may be able to demon- BE. In descending order of preference, these include phar-
strate that reduced levels of the test product compared macokinetic, pharmacodynamic, clinical, and in vitro
with the reference product are associated with adequate studies. These general approaches are discussed in the
efficacy. In either event, the burden is on the sponsor to following sections of this guidance. Product quality BA
demonstrate the adequacy of the clinical trial dose– or and BE frequently rely on pharmacokinetic measures,
concentration–response data to provide evidence of ther- such as AUC and Cmax, that are reflective of systemic
apeutic equivalence. In the absence of this evidence, a exposure.
failure to document BE may suggest that the product
should be reformulated, the method of manufacture for A. PHARMACOKINETIC STUDIES
the test product should be changed, or the BE study should
be repeated. 1. General Considerations

2. ANDAs The statutory definitions of BA and BE, expressed in terms

of rate and extent of absorption of the active ingredient or
BE studies are a critical component of ANDA submis- moiety to the site of action, emphasize the use of phar-
sions. The purpose of these studies is to demonstrate BE macokinetic measures in an accessible biological matrix,
between a pharmaceutically equivalent generic drug prod- such as blood, plasma, or serum, to indicate release of the
uct and the corresponding reference listed drug [21 CFR drug substance from the drug product into the systemic
314.94 (a)(7)]. Together with the determination of phar- circulation. This approach rests on an understanding that
maceutical equivalence, establishing BE allows a regula- measuring the active moiety or ingredient at the site of
tory conclusion of therapeutic equivalence to be drawn. action is generally not possible, and furthermore, that
some relationship exists between the efficacy and safety
3. Postapproval Changes and concentration of the active moiety and its important
Information on the types of in vitro dissolution and in vivo metabolite or metabolites in the systemic circulation. To
BE studies that should be conducted for immediate-release measure product quality BA and establish BE, reliance on
pharmacokinetic measurements may be viewed as a

© 2004 by CRC Press LLC

6 Handbook of Pharmaceutical Manufacturing Formulations: Compressed Solid Products

bioassay that assesses release of the drug substance from The recommended method for analysis of nonreplicate
the drug product into the systemic circulation. A typical or replicate studies to establish BE is average bioequiva-
study is conducted as a crossover study. In this type of lence, as discussed in Section IV. General recommenda-
study, clearance, volume of distribution, and absorption, tions for nonreplicate study designs are provided in the
as determined by physiological variables (e.g., gastric Appendix. Recommendations for replicate study designs
emptying, motility, pH), are assumed to have less interoc- can be found in the Guidance for Industry Statistical
casion variability compared with the variability arising Approaches to Establishing Bioequivalence (January 2001)
from formulation performance. Therefore, differences (http://www.fda.gov/cder/guidance/3616fnl.htm).
between two products because of formulation factors can
be determined. 5. Study Population

2. Pilot Study Unless otherwise indicated by a specific guidance, sub-

jects recruited for in vivo BE studies should be 18 years
If the sponsor chooses, a pilot study among a small num- of age or older and capable of giving informed consent.
ber of subjects can be carried out before proceeding with This guidance recommends that in vivo BE studies be
a full BE study. The study can be used to validate analyt- conducted in individuals who represent the general pop-
ical methodology, assess variability, optimize sample col- ulation, taking into account age, sex, and race. If the drug
lection time intervals, and provide other information. For product is intended for use in both sexes, the sponsor
example, for conventional immediate-release products, should attempt to include similar proportions of males and
careful timing of initial samples may avoid a subsequent females in the study. If the drug product is to be used
finding in a full-scale study that the first sample collection predominantly in the elderly, the sponsor should attempt
occurs after the plasma concentration peak. For modified- to include as many subjects who are 60 years old, or older,
release products, a pilot study can help determine the as possible. The total number of subjects in the study
sampling schedule with which to assess lag time and dose should provide adequate power for BE demonstration, but
dumping. A pilot study that documents BE may be appro- it is not expected that there will be sufficient power to
priate, provided its design and execution are suitable, and draw conclusions for each subgroup. Statistical analysis
a sufficient number of subjects (e.g., 12) completed the of subgroups is not recommended. Restrictions on admis-
study. sion into the study should generally be based solely on
safety considerations. In some instances, it may be useful
3. Pivotal Bioequivalence Studies to admit into BE studies patients for whom a drug product
is intended. In this situation for the duration of the BE
General recommendations for a standard BE study based study, sponsors and applicants should attempt to enter
on pharmacokinetic measurements are provided in the patients who have a disease process that is stable. In accor-
Appendix to this chapter. dance with § 320.31, for some products that will be sub-
mitted in ANDAs, an IND may be required for BE studies
4. Study Designs to ensure patient safety.
Nonreplicate study designs are recommended for BE stud-
ies of immediate-release and modified-release dosage 6. Single-Dose/Multiple-Dose Studies
forms. However, sponsors and applicants have the option Instances where multiple-dose studies may be useful are
of using replicate designs for BE studies for these drug defined under § 320.27(a)(3). The new guidelines gener-
products. Replicate study designs offer several scientific ally recommend single-dose pharmacokinetic studies for
advantages compared with nonreplicate designs. The both immediate- and modified-release drug products to
advantages of replicate study designs are that they: demonstrate BE because they are generally more sensitive
in assessing the release of the drug substance from the
• Allow for comparisons of within-subject vari- drug product into the systemic circulation (see Section V).
ances for the test and reference products If a multiple-dose study design is important, appropriate
• Indicate whether a test product exhibits higher dosage administration and sampling should be carried out
or lower within-subject variability in the bio- to document the attainment of a steady state.
availability measures when compared with the
reference product 7. Bioanalytical Methodology
• Provide more information about the intrinsic
factors underlying formulation performance Bioanalytical methods for BA and BE studies should be
• Reduce the number of subjects needed in the accurate, precise, selective, sensitive, and reproducible. A
BE study separate FDA guidance titled Bioanalytical Method

© 2004 by CRC Press LLC

Bioavailability and Bioequivalence Studies for Orally Administered Drug Products 7

Validation (May 2001) is available to assist sponsors in • Area under the plasma/serum/blood concentra-
validating bioanalytical methods (http://www.fda.gov/cder/ tion–time curve from time 0 to time infinity
guidance/4252fnl.htm). (AUC0-°), where AUC0-° = AUC0-t + Ct/lz, Ct is
the last measurable drug concentration, and lz
8. Pharmacokinetic Measures of Systemic is the terminal or elimination rate constant cal-
Exposure culated according to an appropriate method; the
terminal half-life (t1/2) of the drug should also
Direct (e.g., rate constant, rate profile) and indirect (e.g.,
be reported.
Cmax, Tmax, mean absorption time, mean residence time,
Cmax normalized to area under the curve [AUC]) pharma-
For steady-state studies, the measurement of total
cokinetic measures are limited in their abilities to assess
exposure should be the area under the plasma, serum, or
rate of absorption. This guideline, therefore, recommends
blood concentration–time curve from time 0 to time t over
a change in focus from these direct or indirect measures
a dosing interval at steady state (AUC0-t ), where t is the
of absorption rate to measures of systemic exposure. The
Cmax and AUC values can continue to be used as measures length of the dosing interval.
for product quality BA and BE, but more in terms of their
capacity to assess exposure than their capacity to reflect B. PHARMACODYNAMIC STUDIES
the rate and extent of absorption. Reliance on systemic Pharmacodynamic studies are not recommended for orally
exposure measures should reflect comparable rates and administered drug products when the drug is absorbed into
extents of absorption, which, in turn, should achieve the the systemic circulation, and a pharmacokinetic approach
underlying statutory and regulatory objective of ensuring can be used to assess systemic exposure and establish BE.
comparable therapeutic effects. Exposure measures are However, in those instances where a pharmacokinetic
defined relative to early, peak, and total portions of the approach is not possible, suitably validated pharmacody-
plasma, serum, or blood concentration–time profile, as namic methods can be used to demonstrate BE.
described in Section III.A.8.a. through Section III.A.8.c.
a. Early Exposure C. COMPARATIVE CLINICAL STUDIES
For orally administered immediate-release drug products, Where no other means are available, well-controlled clin-
BE may generally be demonstrated by measurements of ical trials in humans may be useful to provide supportive
peak and total exposure. An early exposure measure may
evidence of BA or BE. However, the use of comparative
be informative on the basis of appropriate clinical efficacy
clinical trials as an approach to demonstrate BE is gener-
and safety trials or pharmacokinetic and pharmacody-
ally considered insensitive and should be avoided when
namic studies that call for better control of drug absorption
possible (21 CFR 320.24). The use of BE studies with
into the systemic circulation (e.g., to ensure rapid onset
clinical trial end points may be appropriate to demonstrate
of an analgesic effect or to avoid an excessive hypotensive
BE for orally administered drug products, when measure-
action of an antihypertensive). In this setting, the guidance
ment of the active ingredients or active moieties in an
recommends use of partial AUC as an early exposure
accessible biological fluid (pharmacokinetic approach) or
measure. The partial area should be truncated at the pop-
pharmacodynamic approach is infeasible.
ulation median of Tmax values for the reference formula-
tion. At least two quantifiable samples should be collected
before the expected peak time to allow adequate estima- D. IN VITRO STUDIES
tion of the partial area. Under certain circumstances, product quality BA and BE
b. Peak Exposure can be documented using in vitro approaches [21 CFR
320.24(b)(5) and 21 CFR 320.22(d)(3)]. For highly solu-
Peak exposure should be assessed by measuring the peak
drug concentration (Cmax) obtained directly from the data ble, highly permeable, rapidly dissolving, orally adminis-
without interpolation. tered drug products, documentation of BE using an in vitro
approach (dissolution studies) is appropriate based on the
c. Total Exposure biopharmaceutics classification system. This approach
For single-dose studies, the measurement of total exposure may also be suitable under some circumstances in assess-
should be as follows: ing BE during the IND period, for NDA and ANDA sub-
missions, and in the presence of certain postapproval
• Area under the plasma/serum/blood concentra- changes to approved NDAs and ANDAs. In addition, in
tion–time curve from time 0 to time t (AUC0-t), vitro approaches to document BE for nonbioproblem
where t is the last time point with measurable drugs approved before 1962 remain acceptable (21 CFR
concentration for individual formulation 320.33).

© 2004 by CRC Press LLC

8 Handbook of Pharmaceutical Manufacturing Formulations: Compressed Solid Products

Dissolution testing is also used to assess batch-to- This guidance recommends that dissolution data from
batch quality, where the dissolution tests, with defined three batches, for NDAs and ANDAs, be used to set dis-
procedures and acceptance criteria, are used to allow batch solution specifications for modified-release dosage forms,
release. Dissolution testing is also used to provide process including extended-release dosage forms.
control and quality assurance and assess whether further
BE studies relative to minor postapproval changes should
be conducted, where dissolution can function as a signal
IV. COMPARISON OF BA MEASURES IN
of bioinequivalence. Dissolution characterization in vitro BE STUDIES
is encouraged for all product formulations investigated An equivalence approach was and continues to be recom-
(including prototype formulations), particularly if in vivo mended for BE comparisons. The recommended approach
absorption characteristics are defined for the different relies on:
product formulations. Such efforts may enable the estab-
lishment of an in vitro–in vivo correlation. When an in 1. Criterion to allow the comparison
vitro–in vivo correlation or association is available [21 2. Confidence interval for the criterion
CFR 320.24(b)(1)(ii)], the in vitro test can serve not only 3. BE limit; log transformation of exposure mea-
as a quality control specification for the manufacturing sures before statistical analysis is recommended
process but also as an indicator of how the product will
perform in vivo. The following guidances provide recom- BE studies are performed as single-dose, crossover
mendations on the development of dissolution methodol- studies. To compare measures in these studies, data were
ogy, the setting of specifications, and the regulatory appli- analyzed using an average BE criterion. This guidance
cations of dissolution testing: Dissolution Testing of recommends continued use of an average BE criterion to
Immediate Release Solid Oral Dosage Forms (August compare BA measures for replicate and nonreplicate BE
1997) and Extended Release Oral Dosage Forms: Devel- studies of immediate- and modified-release products.
opment, Evaluation, and Application of In Vitro/In Vivo
Correlations (September 1997).
The following information should generally be V. DOCUMENTATION OF BA AND BE
included in the dissolution method development report for
An in vivo study is generally recommended for all solid
solid oral dosage forms.
oral dosage forms approved after 1962 and for bioproblem
For an NDA, the following should be included:
drug products approved before 1962. Waiver of in vivo
studies for different strengths of a drug product may be
• The pH solubility profile of the drug substance
granted under § 320.22(d)(2), when (1) the drug product
• Dissolution profiles generated at different agi-
is in the same dosage form, but in a different strength; (2)
tation speeds (e.g., 100 to 150 rpm for U.S.
this different strength is proportionally similar in its active
Pharmacopoeia [USP] Apparatus I [basket], and
and inactive ingredients to the strength of the product for
50 to 100 rpm for USP Apparatus II [paddle])
which the same manufacturer conducted an acceptable in
• Dissolution profiles generated on all strengths
vivo study; and (3) the new strength meets an appropriate
in at least three dissolution media (pH 1.2, 4.5,
in vitro dissolution test. This guidance defines proportion-
and 6.8 buffer); water can be used as an addi-
ally similar in the following ways:
tional medium; if the drug being considered is
poorly soluble, appropriate concentrations of
• All active and inactive ingredients are in the
surfactants should be used
same proportion between different strengths
(e.g., a tablet of 50-mg strength has all the
The agitation speed and medium that provide the best
inactive ingredients, exactly half that of a tablet
discriminating ability, taking into account all the available
of 100 mg strength, and twice that of a tablet
in vitro and in vivo data, will be selected.
of 25 mg strength).
For ANDAs, the following should be included:
• Active and inactive ingredients are not in
exactly the same proportion between different
• USP method
strengths, as stated previously, but the ratios of
• If a USP method is not available, the FDA
inactive ingredients to total weight of the dos-
method for the reference listed drug should be
age form are within the limits defined by the
used.
SUPAC-IR and SUPAC-MR guidances (up to
• If USP or FDA methods are not available, the
Level II).
dissolution method development report
• For high-potency drug substances, where the
described previously should be submitted.
amount of the active drug substance in the

© 2004 by CRC Press LLC

Bioavailability and Bioequivalence Studies for Orally Administered Drug Products 9

dosage form is relatively low, the total weight active and inactive ingredients to the reference listed drug,
of the dosage form remains nearly the same for an in vivo BE demonstration of one or more lower
all strengths (within ± 10% of the total weight strengths can be waived to the reference listed drug based
of the strength on which a biostudy was per- on dissolution tests and an in vivo study on the highest
formed), the same inactive ingredients are used strength.
for all strengths, and the change in any strength For an NDA, biowaivers of a higher strength will be
is obtained by altering the amount of the active determined to be appropriate based on clinical safety and
ingredients and one or more of the inactive efficacy studies, including data on the dose and the desir-
ingredients. The changes in the inactive ingre- ability of the higher strength; linear elimination kinetics
dients are within the limits defined by the over the therapeutic dose range; the higher strength being
SUPAC-IR and SUPAC-MR guidances (up to proportionally similar to the lower strength; and the same
Level II). dissolution procedures being used for both strengths and
similar dissolution results obtained. A dissolution profile
Exceptions to these definitions may be possible, if should be generated for all strengths.
adequate justification is provided. If an appropriate dissolution method was established
(see Section III.D.), and the dissolution results indicate
A. SOLUTIONS that the dissolution characteristics of the product are not
dependent on the product strength, then dissolution pro-
For oral solutions, elixirs, syrups, tinctures, or other sol-
files in one medium are usually sufficient to support waiv-
ubilized forms, in vivo BA or BE can be waived [21 CFR
ers of in vivo testing. Otherwise, dissolution data in three
320.22(b)(3)(i)]. Generally, in vivo BE studies are waived
media (pH 1.2, 4.5, and 6.8) are recommended.
for solutions on the assumption that release of the drug
The f2 test should be used to compare profiles from
substance from the drug product is self-evident, and that
the different strengths of the product. An f2 value ≥ 50
the solutions do not contain any excipient that significantly
indicates a sufficiently similar dissolution profile, such
affects drug absorption [21 CFR 320.22(b)(3)(iii)]. How-
that further in vivo studies are not necessary. For an f2
ever, certain excipients, such as sorbitol or mannitol, can
value < 50, further discussions with review staff from the
reduce the bioavailability of drugs with low intestinal per-
U.S. FDA Center for Drug Evaluation and Research
meability in amounts sometimes used in oral liquid dosage
(CDER) may help to determine whether an in vivo study
forms.
is necessary [21 CFR 320.22(d)(2)(ii)]. The f2 approach
B. SUSPENSIONS is not suitable for rapidly dissolving drug products (e.g.,
≥ 85% dissolved in 15 min or less).
BA and BE for a suspension should generally be estab- For an ANDA, conducting an in vivo study on a
lished just as they are for immediate-release solid oral strength that is not the highest may be appropriate for
dosage forms, and both in vivo and in vitro studies are reasons of safety, subject to approval by the Division of
recommended. Bioequivalence, Office of Generic Drugs, and provided
that the following conditions are met:
C. IMMEDIATE-RELEASE PRODUCTS: CAPSULES AND
TABLETS • Linear elimination kinetics were shown over the
entire therapeutic dose range.
1. General Recommendations • Higher strengths of the test and reference prod-
For product quality, BA and BE studies, where the focus is ucts are proportionally similar to their corre-
on release of the drug substance from the drug product into sponding lower strengths.
the systemic circulation, a single-dose, fasting study should • Comparative dissolution testing on the higher
be performed. The in vivo BE studies should be accompanied strength of the test and reference products was
by in vitro dissolution profiles on all strengths of each prod- submitted and found to be appropriate.
uct. For ANDAs, the BE study should be conducted between
the test product and reference listed drug using the strength b. NDAs and ANDAs: Postapproval
specified in Approved Drug Products with Therapeutic Information on the types of in vitro dissolution and in vivo
Equivalence Evaluations (Orange Book). BE studies for immediate-release drug products approved
as either NDAs or ANDAs in the presence of specified
2. Waivers of In Vivo BE Studies (Biowaivers) postapproval changes are provided in an FDA guidance
for industry titled SUPAC-IR: Immediate Release Solid
a. INDs, NDAs, and ANDAs: Preapproval Oral Dosage Forms: Scale-Up and Post-Approval
When the drug product is in the same dosage form but in Changes: Chemistry, Manufacturing, and Controls, In
a different strength and is proportionally similar in its Vitro Dissolution Testing, and In Vivo Bioequivalence

© 2004 by CRC Press LLC

10 Handbook of Pharmaceutical Manufacturing Formulations: Compressed Solid Products

Documentation (November 1995). For postapproval • The drug product meets the controlled-release
changes, the in vitro comparison should be made between claims made for it
the prechange and postchange products. In instances • The BA profile established for the drug product
where dissolution profile comparisons are recommended, rules out the occurrence of any dose dumping
an f2 test should be used. An f2 value of ≥ 50 suggests a • The drug product’s steady-state performance is
sufficiently similar dissolution profile, and no further in equivalent to a currently marketed noncon-
vivo studies are needed. When in vivo BE studies are trolled release or controlled-release drug prod-
recommended, the comparison should be made for NDAs uct that contains the same active drug ingredient
between the prechange and postchange products and for or therapeutic moiety and is subject to an
ANDAs between the postchange and reference listed drug approved full NDA
products. • The drug product’s formulation provides con-
sistent pharmacokinetic performance between
D. MODIFIED-RELEASE PRODUCTS individual dosage units

Modified-release products include delayed-release prod- As noted in 21 CFR 320.25(f)(2), “the reference mate-
ucts and extended-controlled-release products. rial(s) for such a bioavailability study shall be chosen to
As defined in the USP, delayed-release drug products permit an appropriate scientific evaluation of the con-
are dosage forms that release the drugs at a time later than trolled release claims made for the drug product,” such as
immediately after administration (i.e., these drug products the following:
exhibit a lag time in quantifiable plasma concentrations).
Typically, coatings (e.g., enteric coatings) are intended to • Solution or suspension of the active drug ingre-
delay the release of medication until the dosage form dient or therapeutic moiety
passed through the acidic medium of the stomach. The in • Currently marketed noncontrolled-release drug
vivo tests for delayed-release drug products are similar to product containing the same active drug ingre-
those for extended-release drug products. The in vitro dient or therapeutic moiety and administered
dissolution tests for these products should document that according to the dosage recommendations in
they are stable under acidic conditions and that they the labeling
release the drug only in a neutral medium (e.g., pH 6.8). • Currently marketed controlled-release drug
Extended-release drug products are dosage forms that product subject to an approved full NDA con-
allow a reduction in dosing frequency as compared with taining the same active drug ingredient or ther-
when the drug is present in an immediate-release dosage apeutic moiety and administered according to
form. These drug products can also be developed to reduce the dosage recommendations in the labeling
fluctuations in plasma concentrations. Extended-release
products can be capsules, tablets, granules, pellets, and This guidance recommends that the following BA
suspensions. If any part of a drug product includes an studies be conducted for an extended-release drug product
extended-release component, the following recommenda- submitted as an NDA:
tions apply.
• Single-dose, fasting study on all strengths of
1. NDAs: BA and BE Studies tablets and capsules and highest strength of
An NDA can be submitted for a previously unapproved beaded capsules
new molecular entity or for a new salt, new ester, prodrug, • Single-dose, food-effect study on the highest
or other noncovalent derivative of a previously approved strength
new molecular entity, formulated as a modified-release • Steady-state study on the highest strength
drug product. The first modified-release drug product for
a previously approved immediate-release drug product BE studies are recommended when substantial
should be submitted as an NDA. Subsequent modified- changes in the components or composition or method of
release products that are pharmaceutically equivalent and manufacture for an extended-release drug product occur
bioequivalent to the listed drug product should be submit- between the to-be-marketed NDA dosage form and the
ted as ANDAs. BA requirements for the NDA of an clinical trial material.
extended-release product are listed in 21 CFR 320.25(f).
The purpose of an in vivo BA study for which a controlled- 2. ANDAs: BE Studies
release claim is made is to determine if all the following For modified-release products submitted as ANDAs, the
conditions are met: following studies are recommended: a single-dose, non-
replicate, fasting study comparing the highest strength of

© 2004 by CRC Press LLC

Bioavailability and Bioequivalence Studies for Orally Administered Drug Products 11

the test and reference listed drug product, unless the drug an f2 test should be used. An f2 value of ≥ 50 suggests a
or drug product is highly variable, in which case a replicate similar dissolution profile. A failure to demonstrate similar
design study is recommended; and a food-effect, nonrep- dissolution profiles may indicate that an in vivo BE study
licate study comparing the highest strength of the test and should be performed. When in vivo BE studies are con-
reference product (see Section VI.A.). Because single- ducted, the comparison should be made for NDAs between
dose studies are considered more sensitive in addressing the prechange and postchange products and for ANDAs
the primary question of BE (i.e., release of the drug sub- between the postchange product and the reference listed
stance from the drug product into the systemic circula- drug.
tion), multiple-dose studies are generally not recom-
mended, even in instances where nonlinear kinetics are E. MISCELLANEOUS DOSAGE FORMS
present.
Rapidly dissolving drug products, such as buccal and sub-
3. Waivers of In Vivo BE Studies (Biowaivers): lingual dosage forms, should be tested for in vitro disso-
lution and in vivo BA and BE. Chewable tablets should
NDAs and ANDAs
also be evaluated for in vivo BA and BE. Chewable tablets
a. Beaded Capsules — Lower Strength (as a whole) should be subject to in vitro dissolution
For modified-release beaded capsules, where the strength because they might be swallowed by a patient without
differs only in the number of beads containing the active being properly chewed. In general, in vitro dissolution test
moiety, a single-dose, fasting BE study should be carried conditions for chewable tablets should be the same as for
out only on the highest strength, with waiver of in vivo nonchewable tablets of the same active ingredient or moi-
studies for lower strengths based on dissolution profiles. ety. Infrequently, different test conditions or acceptance
A dissolution profile should be generated for each strength criteria may be indicated for chewable and nonchewable
using the recommended dissolution method. The f2 test tablets, but these differences, if they exist, should be
should be used to compare profiles from the different resolved with the appropriate review division.
strengths of the product. An f2 value of ≥ 50 can be used
to confirm that further in vivo studies are not needed. VI. SPECIAL TOPICS
b. Tablets — Lower Strength
For modified-release tablets, when the drug product is in A. FOOD-EFFECT STUDIES
the same dosage form but in a different strength, is pro- Coadministration of food with oral drug products may
portionally similar in its active and inactive ingredients, influence drug BA and BE. Food-effect BA studies focus
and has the same drug release mechanism, an in vivo BE on the effects of food on the release of the drug substance
determination of one or more lower strengths can be from the drug product as well as the absorption of the
waived based on dissolution profile comparisons, with an drug substance. BE studies with food focus on demon-
in vivo study only on the highest strength. The drug prod- strating comparable BA between test and reference prod-
ucts should exhibit similar dissolution profiles between ucts when coadministered with meals. Usually, a single-
the highest strength and the lower strengths, based on the dose, two-period, two-treatment, two-sequence crossover
f2 test in at least three dissolution media (e.g., pH 1.2, 4.5, study is recommended for food-effect BA and BE studies.
and 6.8). The dissolution profile should be generated on
the test and reference products of all strengths. B. MOIETIES TO BE MEASURED
4. Postapproval Changes 1. Parent Drug vs. Metabolites

Information on the types of in vitro dissolution and in vivo The moieties to be measured in biological fluids collected
BE studies for extended-release drug products approved in BA and BE studies are either the active drug ingredient
as either NDAs or ANDAs in the presence of specified or its active moiety in the administered dosage form (par-
postapproval changes are provided in an FDA guidance ent drug) and, when appropriate, its active metabolites [21
for industry titled SUPAC-MR: Modified Release Solid CFR 320.24(b)(1)(i)]. This guidance recommends the fol-
Oral Dosage Forms: Scale-Up and Post-Approval lowing approaches for BA and BE studies.
Changes: Chemistry, Manufacturing, and Controls, In For BA studies (see Section II.B.), determination of
Vitro Dissolution Testing, and In Vivo Bioequivalence moieties to be measured in biological fluids should take
Documentation (September 1997). For postapproval into account concentration and activity. Concentration
changes, the in vitro comparison should be made between refers to the relative quantity of the parent drug or one or
the prechange and postchange products. In instances more metabolites in a given volume of an accessible bio-
where dissolution profile comparisons are recommended, logical fluid, such as blood or plasma. Activity refers to

© 2004 by CRC Press LLC

12 Handbook of Pharmaceutical Manufacturing Formulations: Compressed Solid Products

the relative contribution of the parent drug and its metab- • Nonlinear absorption is present (as expressed
olite in the biological fluids to the clinical safety and by a change in the enantiomer concentration
efficacy of the drug. For BA studies, the parent drug and ratio with a change in the input rate of the drug)
its major active metabolite should be measured, if analyt- for at least one of the enantiomers.
ically feasible.
For BE studies, measurement of only the parent drug In such cases, BE criteria should be applied to the
released from the dosage form, rather than the metabolite, enantiomers separately.
is generally recommended. The rationale for this recom-
mendation is that the concentration–time profile of the 3. Drug Products with Complex Mixtures as the
parent drug is more sensitive to changes in formulation Active Ingredients
performance than a metabolite, which is more reflective
of metabolite formation, distribution, and elimination. The Certain drug products may contain complex drug sub-
following are exceptions to this general approach: stances (i.e., active moieties or active ingredients that are
mixtures of multiple synthetic and natural source compo-
• Measurement of a metabolite may be preferred nents). Some or all the components of these complex drug
when parent drug levels are too low to allow substances cannot be characterized with regard to chem-
reliable analytical measurement in blood, ical structure or biological activity. Quantification of all
plasma, or serum for an adequate length of time. active or potentially active components in pharmacoki-
The metabolite data obtained from these studies netic studies to document BA and BE is neither necessary
should be subject to a confidence interval nor desirable. Rather, BA and BE studies should be based
approach for BE demonstration. If there is clin- on a small number of markers of rate and extent of absorp-
ical concern related to efficacy or safety for the tion. Although necessarily a case-by-case determination,
parent drug, sponsors and applicants should criteria for marker selection include the amount of the
contact the appropriate review division to deter- moiety in the dosage form, plasma or blood levels of the
mine whether the parent drug should be mea- moiety, and biological activity of the moiety relative to
sured and analyzed statistically. other moieties in the complex mixture. Where pharmaco-
• Metabolite may be formed as a result of gut kinetic approaches are not feasible to assess the rate and
wall or other presystemic metabolism. If the extent of absorption of a drug substance from a drug
metabolite contributes meaningfully to safety product, in vitro approaches may be preferred. Pharmaco-
and efficacy, the metabolite and the parent drug dynamic or clinical approaches may be called for if no
should be measured. When the relative activity quantifiable moieties are available for in vivo pharmaco-
of the metabolite is low and does not contribute kinetic or in vitro studies.
meaningfully to safety and efficacy, it does not
need to be measured. The parent drug measured C. LONG HALF-LIFE DRUGS
in these BE studies should be analyzed using a
In a BA or pharmacokinetic study involving an oral prod-
confidence interval approach. The metabolite
uct with a long half-life drug, adequate characterization
data can be used to provide supportive evidence
of the half-life calls for blood sampling over a long period
of comparable therapeutic outcome.
of time. For a BE determination of an oral product with
a long half-life drug, a nonreplicate, single-dose, crossover
2. Enantiomers vs. Racemates
study can be conducted, provided an adequate washout
For BA studies, the measurement of individual enanti- period is used. If the crossover study is problematic, a BE
omers may be important. For BE studies, this guidance study with a parallel design can be used. For either a
recommends measurement of the racemate using an crossover or parallel study, sample collection time should
achiral assay. Measurement of individual enantiomers in be adequate to ensure completion of gastrointestinal tran-
BE studies is recommended only when all the following sit (approximately 2 to 3 days) of the drug product and
conditions are met: absorption of the drug substance. The Cmax, and a suitably
truncated AUC, can be used to characterize peak and total
• Enantiomers exhibit different pharmacody- drug exposure, respectively. For drugs that demonstrate
namic characteristics. low intrasubject variability in distribution and clearance,
• Enantiomers exhibit different pharmacokinetic an AUC truncated at 72 h (AUC0-72 h) can be used in place
characteristics. of AUC0-t or AUC0-° . For drugs demonstrating high
• Primary efficacy and safety activity reside with intrasubject variability in distribution and clearance, AUC
the minor enantiomer. truncation warrants caution. In such cases, sponsors and
applicants should consult the appropriate review staff.

© 2004 by CRC Press LLC

Bioavailability and Bioequivalence Studies for Orally Administered Drug Products 13

D. FIRST POINT CMAX help in understanding the degree of systemic exposure that
occurs following administration of a drug product
The first point of a concentration–time curve in a BE study intended for local action in the gastrointestinal tract.
based on blood and plasma measurements is sometimes
the highest point, which raises a question about the mea- F. NARROW THERAPEUTIC RANGE DRUGS
surement of true Cmax because of insufficient early sam-
pling times. A carefully conducted pilot study may avoid The guidance defines narrow therapeutic range drug prod-
this problem. Making collections at an early time point, ucts as those containing certain drug substances that are
between 5 and 15 min after dosing, followed by making subject to therapeutic drug concentration or pharmacody-
additional sample collections (e.g., two to five) in the first namic monitoring, and where product labeling indicates
hour after dosing may be sufficient for assessing early a narrow therapeutic range designation. Examples include
peak concentrations. If this sampling approach is fol- digoxin, lithium, phenytoin, theophylline, and warfarin.
lowed, data sets should be considered adequate, even when Because not all drugs subject to therapeutic drug concen-
the highest observed concentration occurs at the first time tration or pharmacodynamic monitoring are narrow ther-
point. apeutic range drugs, sponsors and applicants should con-
tact the appropriate review division at CDER to determine
E. ORALLY ADMINISTERED DRUGS INTENDED FOR whether a drug should or should not be considered to have
LOCAL ACTION a narrow therapeutic range.
The guidance recommends that sponsors consider
Documentation of product quality BA for NDAs, where additional testing and controls to ensure the quality of
the drug substance produces its effects by local action in drug products containing narrow therapeutic range drugs.
the gastrointestinal tract, can be achieved using clinical The approach is designed to provide increased assurance
efficacy and safety studies or suitably designed and vali- of interchangeability for drug products containing speci-
dated in vitro studies. Similarly, documentation of BE for fied narrow therapeutic range drugs. It is not designed to
ANDAs and for NDAs, as well as for ANDAs in the influence the practice of medicine or pharmacy.
presence of certain postapproval changes, can be achieved Unless otherwise indicated by a specific guidance, this
by using BE studies with clinical efficacy and safety end guidance recommends that the traditional BE limit of 80
points or suitably designed and validated in vitro studies, to 125% for nonnarrow therapeutic range drugs remain
if the latter studies are reflective of important clinical unchanged for the bioavailability measures (AUC and
effects or are more sensitive to changes in product perfor- Cmax) of narrow therapeutic range drugs.
mance compared with a clinical study. To ensure compa-
rable safety, additional studies with and without food may

© 2004 by CRC Press LLC

© 2004 by CRC Press LLC
Appendix 1A — General Pharmacokinetic
Study Design and Data Handling
For replicate and nonreplicate in vivo pharmacokinetic BE drug or metabolites are measured in serum or
studies, the following general approaches are recom- plasma. However, in certain cases, whole blood
mended, recognizing that the elements may be adjusted may be more appropriate for analysis. Blood
for certain drug substances and drug products. samples should be drawn at appropriate times
to describe the absorption, distribution, and
elimination phases of the drug. For most drugs,
STUDY CONDUCT 12 to 18 samples, including a predose sample,
should be collected per subject per dose. This
• The test or reference products should be admin- sampling should continue for at least three or
istered with about 8 oz (240 ml) of water to an more terminal half-lives of the drug. The exact
appropriate number of subjects under fasting timing for sample collection depends on the
conditions, unless the study is a food-effect BA nature of the drug and the input from the admin-
and BE study. istered dosage form. The sample collection
• Generally, the highest marketed strength should should be spaced in such a way that the maxi-
be administered as a single unit. If warranted mum concentration of the drug in the blood
for analytical reasons, multiple units of the (Cmax) and terminal elimination rate constant
highest strength can be administered, providing (lz) can be estimated accurately. At least three
the total single-dose remains within the labeled to four samples should be obtained during the
dose range. terminal log-linear phase in order to obtain an
• An adequate washout period (e.g., more than 5 accurate estimate of lz from linear regression.
half-lives of the moieties to be measured) The actual clock time when samples are drawn
should separate each treatment. as well as the elapsed time related to drug
• The lot numbers of both test and reference listed administration should be recorded.
products and the expiration date for the refer-
ence product should be stated. The drug content
of the test product should not differ from that SUBJECTS WITH PREDOSE PLASMA
of the reference listed product by more than CONCENTRATIONS
5%. The sponsor should include a statement of
• If the predose concentration is less than or equal
the composition of the test product and, if pos-
to 5% of the Cmax value in that subject, the
sible, a side-by-side comparison of the compo-
subject’s data, without any adjustments, can be
sitions of test and reference listed products. In
included in all pharmacokinetic measurements
accordance with 21 CFR 320.38, samples of the
and calculations. If the predose value is greater
test and reference listed product must be
than 5% of Cmax, the subject should be dropped
retained for 5 years.
from all BE study evaluations.
• Before and during each study phase, subjects
should be allowed water, as desired, except for
1 h before and after drug administration; be DATA DELETION DUE TO VOMITING
provided standard meals no less than 4 h after
drug administration; and abstain from alcohol • Data from subjects who experience emesis dur-
for 24 h before each study period and until after ing the course of a BE study for immediate-
the last sample from each period is collected. release products should be deleted from statis-
tical analysis if vomiting occurs at or before
two times median Tmax. In the case of modified-
SAMPLE COLLECTION AND SAMPLING TIMES release products, the data from subjects who
• Under normal circumstances, blood, rather than experience emesis any time during the labeled
urine or tissue, should be used. In most cases, dosing interval should be deleted.

15

© 2004 by CRC Press LLC

16 Handbook of Pharmaceutical Manufacturing Formulations: Compressed Solid Products

PHARMACOKINETIC INFORMATION BIOEQUIVALENCE DEMONSTRATION

RECOMMENDED FOR SUBMISSION MEASURES

• Plasma concentrations and time points • Logarithmic transformation should be provided

• Subject, period, sequence, treatment for measures used for BE demonstration.
• AUC0-t, AUC0-°, Cmax, Tmax, kz, and t1/2
• Intersubject, intrasubject, and total variability,
if available
CONFIDENCE INTERVAL VALUES
• Cmin (concentration at the end of a dosing inter-
val), Cav (average concentration during a dosing • Confidence interval (CI) values should not be
interval), degree of fluctuation [(C m a x - rounded off; therefore, to pass a CI limit of 80
Cmin)/Cav], and swing [(Cmax - Cmin)/Cmin], if to 125, the value should be at least 80.00 and
steady-state studies are employed not more than 125.00.
• Partial AUC, requested only as discussed in
Section III.A.9.a. of this chapter

STATISTICAL INFORMATION FOR AUC0-T,

AUC0-°, AND CMAX:

• Geometric mean
• Arithmetic mean
• Ratio of means
• Confidence intervals

© 2004 by CRC Press LLC