The Use of Physiologically Based Pharmacokinetic Analyses — Biopharmaceutics Applications for Oral Drug Product Development, Manufacturing Changes, and Controls : Draft Guidance for Industry

Administrative draft guidance, not enacted legislation. This captured publisher edition preserves the listed document and attachments; later developments are not independently consolidated.

19190003dft.docx
09/28/20
The Use of Physiologically Based
Pharmacokinetic Analyses —
Biopharmaceutics Applications for Oral
Drug Product Development,
Manufacturing Changes, and Controls
Guidance for Industry
DRAFT GUIDANCE
This guidance document is being distributed for comment purposes only.
Comments and suggestions regarding this draft document should be submitted within 60 days of
publication in the Federal Register of the notice announcing the availability of the draft
guidance. Submit electronic comments to https://www.regulations.gov. Submit written
comments to the Dockets Management Staff (HFA-305), Food and Drug Administration, 5630
Fishers Lane, Rm. 1061, Rockville, MD 20852. All comments should be identified with the
docket number listed in the notice of availability that publishes in the Federal Register.
For questions regarding this draft document, contact Paul Seo at 301-796-4874.
U.S. Department of Health and Human Services
Food and Drug Administration
Center for Drug Evaluation and Research (CDER)
October 2020
Pharmaceutical Quality/CMC

The Use of Physiologically Based
Pharmacokinetic Analyses —
Biopharmaceutics Applications for Oral
Drug Product Development,
Manufacturing Changes, and Controls
Guidance for Industry
Additional copies are available from:
Office of Communications, Division of Drug Information
Center for Drug Evaluation and Research
Food and Drug Administration
10001 New Hampshire Ave., Hillandale Bldg., 4th Floor
Silver Spring, MD 20993-0002
Phone: 855-543-3784 or 301-796-3400; Fax: 301-431-6353; Email: druginfo@fda.hhs.gov
https://www.fda.gov/drugs/guidance-compliance-regulatory-information/guidances-drugs
U.S. Department of Health and Human Services
Food and Drug Administration
Center for Drug Evaluation and Research (CDER)
October 2020
Pharmaceutical Quality/CMC

Contains Nonbinding Recommendations
Draft — Not for Implementation
TABLE OF CONTENTS
I. 	INTRODUCTION............................................................................................................. 1
II. 	BACKGROUND ............................................................................................................... 2
III. 	IMPLEMENTATION OF PBPK MODELING FOR BIOPHARMACEUTICS
APPLICATIONS .............................................................................................................. 3
IV. 	DEVELOPMENT AND EVALUATION OF PBPK MODELS FOR
BIOPHARMACEUTICS APPLICATIONS .................................................................. 5
A. 	General Strategy ............................................................................................................................ 5
B. 	General Considerations ................................................................................................................. 7
1. Model Objective(s) ........................................................................................................................... 7
2. Model Development ......................................................................................................................... 7
3. Model Validation and Refinement ................................................................................................... 8
V. 	PBPK MODELING FOR BIOPHARMACEUTICS APPLICATIONS TO SUPPORT
PRODUCT QUALITY ......................................................................................................... 9
A. 	Development of Clinically Relevant Dissolution Specifications (Method and Acceptance
Criteria) .......................................................................................................................................... 9
1. Aid in Biopredictive Dissolution Method Development ................................................................... 9
2. Support Clinically Relevant Dissolution Acceptance Criteria....................................................... 10
B. 	Establishment of Clinically Relevant Drug Product Quality Specifications (Other Than
Dissolution) ................................................................................................................................... 11
C. 	Quality Risk Assessment for Pre- and Postapproval Changes and Risk-Based Biowaivers . 12
GLOSSARY................................................................................................................................. 14
REFERENCES ............................................................................................................................ 16

Contains Nonbinding Recommendations
Draft — Not for Implementation
1
The Use of Physiologically Based Pharmacokinetic Analyses —	1
Biopharmaceutics Applications for Oral Drug Product	2
Development, Manufacturing Changes, and Controls	3
Guidance for Industry1
4
5
6
7
8
This draft guidance, when finalized, will represent the current thinking of the Food and Drug	9
Administration (FDA or Agency) on this topic. It does not establish any rights for any person and is not	10
binding on FDA or the public. You can use an alternative approach if it satisfies the requirements of the	11
applicable statutes and regulations. To discuss an alternative approach, contact the FDA staff responsible	12
for this guidance as listed on the title page.	13
14
15
16
17
I. 	INTRODUCTION	18
19
This guidance provides general recommendations regarding the development, evaluation, and	20
use of physiologically based pharmacokinetic (PBPK) analyses for biopharmaceutics	21
applications employed by sponsors of investigational new drug applications, and applicants for	22
new drug applications, or abbreviated new drug applications, and supplements to these	23
applications,2,3 for oral drug product development, manufacturing changes, and controls. PBPK	24
analyses use models and simulations that combine physiology, population, and drug substance	25
and product characteristics to mechanistically describe the pharmacokinetic (PK) and/or	26
pharmacodynamic behaviors of a drug product.4	
27
28
The application of PBPK modeling in support of drug product development is an evolving field.	29
We note that there are multiple terms used to describe PBPK analyses for biopharmaceutics	30
applications, including PBPK absorption modeling (Zhang et al. 2017), physiologically based	31
absorption modeling (Kesisoglou et al. 2016), and physiologically based biopharmaceutics	32
modeling (PBBM) (Heimbach et al. 2019). This guidance uses the term PBPK analyses (or	33
1 This guidance has been prepared by the Office of Pharmaceutical Quality in the Center for Drug Evaluation and
Research at the Food and Drug Administration.
2 For the purposes of this guidance, sponsor refers to sponsors of investigational new drug applications, applicants
for new drug applications and abbreviated new drug applications, and supplements to those applications.
3 The scientific principles described in this guidance are applicable regardless of whether an original clinical study
demonstrated bioavailability/bioequivalence and are relevant whether or not an application is required.
4 Submission of PBPK analyses to FDA is discussed in the guidance for industry Physiologically Based
Pharmacokinetic Analyses — Format and Content (August 2018). We update guidances periodically. For the most
recent version of a guidance, check the FDA guidance web page at https://www.fda.gov/regulatory-
information/search-fda-guidance-documents.

Contains Nonbinding Recommendations
Draft — Not for Implementation
2
modeling) for biopharmaceutics applications to emphasize the focus on drug product quality	34
attributes and a mechanistic understanding of their interaction with physiology to affect in vivo	35
drug performance.	36
37
This guidance applies only to orally administered, systemically active drug products. It does not	38
apply to locally acting drug products, including orally delivered gastrointestinal (GI) drug	39
products that reach the site of action before entering systemic circulation. The use of PBPK	40
analyses for biopharmaceutics applications for locally acting drug products will be considered on	41
a case-by-case basis and via communication with FDA.	42
43
In general, FDA’s guidance documents do not establish legally enforceable responsibilities.	44
Instead, guidances describe the Agency’s current thinking on a topic and should be viewed only	45
as recommendations, unless specific regulatory or statutory requirements are cited. The use of	46
the word should in Agency guidances means that something is suggested or recommended, but	47
not required.	48
49
50
II. 	BACKGROUND	51
52
Several guidances for industry advocate the use of biopharmaceutics tools,5 such as in vitro	53
dissolution, and in vivo bioavailability (BA)/bioequivalence (BE) studies, along with modeling	54
approaches to support drug product quality. In addition, quality by design (QbD) principles	55
recognize that drug product quality cannot be tested into drug products; quality should be built	56
into drug products by design.6 In this regard, QbD enables an in-depth understanding of the	57
relationship among critical quality attributes (CQAs), critical material attributes (CMAs), critical	58
process parameters (CPPs), and predefined clinical performance metrics (e.g., systemic exposure	59
such as C max and area under the curve (AUC)). Data describing this relationship are essential for	60
establishing an in vitro-in vivo link. Establishing an in vitro-in vivo link supports clinically	61
relevant drug product specifications.	62
63
5 See the guidances for industry Dissolution Testing of Immediate Release Solid Oral Dosage Forms (August 1997),
Extended Release Oral Dosage Forms: Development, Evaluation, and Application of In Vitro/In Vivo Correlations
(September 1997), and Immediate Release Solid Oral Dosage Forms Scale-Up and Postapproval Changes:
Chemistry, Manufacturing, and Controls, In Vitro Dissolution Testing, and In Vivo Bioequivalence Documentation
(November 1995).
6 See the ICH guidance for industry Q8(R2) Pharmaceutical Development (November 2009).

Contains Nonbinding Recommendations
Draft — Not for Implementation
3
Although the pharmaceutical industry has in some cases been successful in developing in vitro/in	64
vivo correlations (IVIVCs) to support biowaiver7 requests in lieu of in vivo BE studies for major	65
manufacturing changes (Nguyen et al. 2017), development of an adequate IVIVC for regulatory	66
submission remains challenging (Suarez-Sharp et al. 2016). FDA recognizes this challenge and	67
encourages the development and use of new tools and approaches for linking pharmaceutical	68
quality to clinical performance. Advances in modeling and simulation have enabled the	69
integration of factors such as the physicochemical properties of the active pharmaceutical	70
ingredient (API), dissolution data, and the physiology of the GI tract into the development of	71
PBPK models. As such, PBPK modeling has become a promising tool in predicting systemic	72
drug exposure (Kostewicz et al. 2014a) and has been used for dose selection, food effect	73
assessment, and drug interaction potential evaluation (Wagner et al. 2015a; Wagner et al. 2016;	74
Huang et al. 2013; Wagner et al. 2015b).	75
76
77
III. 	IMPLEMENTATION OF PBPK MODELING FOR BIOPHARMACEUTICS	78
APPLICATIONS	79
80
The application of PBPK modeling could be expanded to pharmaceutical drug product	81
development, manufacturing changes, and controls. One feasible approach is to combine an in	82
vitro drug product test (e.g., biopredictive dissolution) with PBPK models where in vitro	83
dissolution data provide input to predict absorption (Heimbach et al. 2019). As such, dissolution	84
testing is a key modeling input, because it probes both the extent and rate of in vivo drug product	85
release.	86
87
The purpose of PBPK analyses for biopharmaceutics applications is to combine dissolution	88
modeling/biopredictive dissolution or other in vitro testing inputs with PBPK modeling strategies	89
to quantitatively describe (or characterize) the potential interactions of formulation variants with	90
the body and their effect on drug exposure. This modeling approach should include relevant	91
mechanisms pertaining to the absorption process, such as GI tract local metabolism (if	92
applicable) and drug transport, and incorporate drug product quality properties to predict	93
systemic drug exposure.	94
95
7 In addition to waiver of an in vivo BE requirement under 21 CFR 320.22, there are certain circumstances in which
BE can be evaluated using in vitro approaches under 21 CFR 320.24(b)(6). The scientific principles described in this
guidance regarding waiver of an in vivo requirement also apply to consideration of in vitro data under that
regulation. In such circumstances, an in vivo data requirement is not waived, but rather, FDA has determined that in
vitro data is the most accurate, sensitive, and reproducible approach for establishing BE, as required under 21 CFR
320.24(a). Nonetheless, for ease of the reader, this guidance refers to either the decision to waive an in vivo BE
requirement under 21 CFR 320.22 or the decision to accept in vitro BE data in accordance with 21 CFR 320.24(a) as
a biowaiver.

Contains Nonbinding Recommendations
Draft — Not for Implementation
4
With these mechanistic elements defined, PBPK modeling for biopharmaceutics applications	96
could predict the effect of variations from the CMAs, CPPs, and CQAs on drug exposure toward	97
the establishment of a safe space via either IVIVCs or in vivo-in vitro relationships combined	98
with virtual BE. A safe space (Abend 2018) is defined by the boundaries demarcated by in vitro	99
specifications (i.e., dissolution or, when applicable, other relevant drug product quality	100
attributes), within which drug product variants are anticipated to be bioequivalent to one another.	101
Less optimally, but still possible (e.g., for modified-release (MR) formulations with appropriate	102
additional supporting data), safe space represents specifications within which drug product	103
variants are anticipated to be bioequivalent to the pivotal clinical batch(es).8 Building a safe	104
space may also reduce the need for in vivo data to support regulatory assessment.9 Although safe	105
spaces can be used for new and generic drug products, building a safe space for a generic drug	106
product necessitates the identification of a range of virtual dissolution profiles within which the	107
proposed drug products are found to be bioequivalent to one another and to the reference or	108
target drug product (e.g., via virtual BE analysis). Also, the range of virtual dissolution profiles	109
should contain the target (i.e., biobatch or pivotal clinical batch) dissolution profile.	110
111
The implementation of PBPK analyses for biopharmaceutics applications to support drug	112
product quality should consider a risk-based approach (e.g., Kuemmel et al. 2020) and	113
contemplate several factors such as: (1) whether in vivo dissolution (as opposed to permeability)	114
is the rate-limiting step toward drug absorption; (2) the in vitro and in vivo data collected to	115
develop, verify, and validate10 the proposed model; and (3) the complexity of the drug product	116
formulation. For example, the use of PBPK analyses for biopharmaceutics applications to	117
support major manufacturing changes for immediate-release (IR) drug products containing high	118
solubility APIs generally is not warranted.11	
119
120
121
8 If a clinical investigation (i.e., any experiment other than a BA study in which a drug is administered or dispensed
to, or used on, human subjects) is necessary to demonstrate the safety or effectiveness of a proposed drug product,
generally this type of study goes beyond the scope of information that may be relied upon as necessary for approval
in an abbreviated new drug application (see 21 CFR 314.108(a) and the guidance for industry Determining Whether
to Submit an ANDA or a 505(b)(2) Application (May 2019)). However, for ease of the reader, use of the term
clinical in this guidance may refer to clinical investigations conducted to support the demonstration of safety or
effectiveness in a drug product submitted in a new drug application, or to in vivo studies submitted to support a
demonstration of BE or other requirements under section 505(j) of the Federal Food, Drug, and Cosmetic Act and
FDA’s implementing regulations.
9 See for example 21 CFR 320.25(a) (“[t]he basic principle in an in vivo bioavailability study is that no unnecessary
human research should be done”).
10 In the most general terms, verification refers to an assessment of model components, for example by examining
computer codes and equations, to evaluate whether they accurately implement model assumptions; and validation
refers to an assessment of the model performance in comparison with observed in vivo data.
11 See the guidance for industry Dissolution Testing and Acceptance Criteria for Immediate-Release Solid Oral
Dosage Form Drug Products Containing High Solubility Drug Substances (August 2018).

Contains Nonbinding Recommendations
Draft — Not for Implementation
5
IV. 	DEVELOPMENT AND EVALUATION OF PBPK MODELS FOR	122
BIOPHARMACEUTICS APPLICATIONS	123
124
A. 	General Strategy	125
126
The general recommended process of developing (Zhang et al. 2011), evaluating, and applying a	127
PBPK model for biopharmaceutics applications for an oral dosage form is presented in Figure 1	128
and described in this section and section V., PBPK Modeling for Biopharmaceutics Applications	129
to Support Product Quality.	130
131
A complete study report of the modeling and simulation work using a PBPK model for	132
biopharmaceutics applications should be submitted to FDA for evaluation and included in the	133
electronic common technical document Module 5.3.1.3.12	
134
135
12 When such a model is used for other purposes, other modules may be more appropriate. For additional
information, see the guidance for industry Providing Regulatory Submissions in Electronic Format — Certain
Human Pharmaceutical Product Applications and Related Submissions Using the eCTD Specifications (February
2020).

Contains Nonbinding Recommendations
Draft — Not for Implementation
6
Figure 1. Recommended Workflow Describing Development and Evaluation of a PBPK	136
Model for Biopharmaceutics Applications	137
138
139
The data needed for model development include, but are not limited to:	140
141
• 	Drug data comprised of drug substance physicochemical properties; formulation	142
attributes; the drug product release mechanism; the absorption, distribution, metabolism,	143
and excretion properties of the drug product; as well as other relevant clinical data (e.g.,	144
BA/BE or other PK data)	145
146
• 	System data (i.e., anatomical structure and physiological parameters) for the GI tract and	147
other organs and/or tissues, if applicable	148
Model Development
(Section IV.B.2)
Model Validation
(Section IV.B.3)	
Model Refinement
(IV.B.3)
Model Application
(Section V)
Identification of Model Objective(s)
(Section IV.B.1)
Structure 	Assumptions	Parameters
Acceptable?
Data collection/integration
Model Verification
No
Yes

Contains Nonbinding Recommendations
Draft — Not for Implementation
7
149
• 	Study design data (e.g., the tested drug product or formulation information, dosing	150
regimen, and study subject population) of the in vivo studies selected for model	151
development and validation.	152
153
B. 	General Considerations	154
155
The following subsections provide general considerations for model development and evaluation	156
in supporting pharmaceutical quality.	157
158
1. 	Model Objective(s)	159
160
The specific drug product quality issue(s) or question(s) to be addressed by PBPK modeling for	161
biopharmaceutics applications should be clearly described in the study report. Sponsors should	162
provide an analysis of how the specified quality issue(s) or question(s) affect the PK	163
performance of the drug product, the rationale for conducting the modeling and simulation, as	164
well as any strategies undertaken to mitigate the risk of the change to PK performance. The	165
analysis should also incorporate a description of the level of confidence in the modeling outcome	166
based on additional data available to support the verification and/or validation of the model, and	167
other factors, such as the model application, the therapeutic indication, and the therapeutic	168
window of the drug. FDA will evaluate on a case-by-case basis the adequacy of the model for the	169
intended purpose and data sufficiency for model verification and/or validation.	170
171
2. 	Model Development	172
173
Model development should consist of the following three general elements.	174
175
a. 	Model structure	176
177
The model structure should provide a mechanistic framework of drug oral absorption by	178
representing the in vivo drug absorption process and accounting for the relevant product quality	179
attribute(s) that affect drug dissolution and absorption. The construction of an absorption model	180
should consider the model objective(s), as well as multiple factors affecting drug dissolution and	181
absorption and their interactions. These factors include but are not limited to: physicochemical	182
properties of the drug substance; formulation/process characteristics; drug release mechanism; in	183
vivo drug dissolution process; supersaturation and precipitation processes; location and duration	184
of absorption; drug permeation and transport pathway; and the effect of GI tract physiology on	185
absorption. Finally, model construction should consider the supporting data and knowledge	186
available to justify the model structure. Sponsors should also document the approaches taken to	187
integrate quality attributes, such as dissolution, and other factors into the model.	188
189
Because the focus of the model is on in vivo dissolution and absorption, it is appropriate to	190
combine a mechanistic absorption model with a simplified disposition model (e.g., a classic	191
compartmental PK model or a reduced PBPK model that lumps tissue/organ compartments) for	192
the prediction of systemic exposure following absorption. Such simplification is recommended if	193
it does not compromise the ability to adequately describe processes governing the drug BA.	194

Contains Nonbinding Recommendations
Draft — Not for Implementation
8
When drug disposition involves complexity (e.g., nonlinear PK, saturation of clearance	195
pathway), we recommend an alternative approach, such as incorporating enzyme kinetics in the	196
disposition model. Any modification to the initial model structure should include sufficient	197
justification (e.g., the addition of structural elements may be supported by comprehensive	198
sensitivity analyses and/or appropriate proof that clearly demonstrates the significance of the	199
metrics of interest).	200
201
b. 	Model assumptions	202
203
The assumptions that underly the model structure and parameters should be clearly presented	204
(e.g., the assumptions made upon drug disintegration, dissolution, precipitation, degradation,	205
transport, first-pass effect, distribution, and clearance). The assumptions should be scientifically	206
justified with supportive information and data, when available. The effect of these assumptions	207
on model structure and/or parameter(s) should be described.	208
209
c. 	Model parameters	210
211
The approach taken to incorporate drug product quality attributes into the model and the	212
selection of parameters and parameter values as model inputs should be clearly presented and	213
scientifically justified. Selection and evaluation of CMAs (such as drug substance	214
physicochemical properties and excipient(s) level), CPPs (such as compression force), and CQAs	215
(such as hardness, disintegration, and in vitro dissolution) as model inputs should consider	216
whether these attributes and parameters can affect drug in vivo dissolution and absorption.	217
218
3. 	Model Validation and Refinement	219
220
The predictive performance of a model should be validated for its intended purpose. Depending	221
on the clinical risk and the intended purpose, the amount and type of data needed for model	222
validation may vary. Independent datasets not used in model development are recommended to	223
evaluate the predictive performance of the model. In general, for addressing pharmaceutical	224
development and quality issues, the adequacy of the model to predict the effect of model inputs	225
on the PK performance of the studied drug product should be demonstrated by establishing a	226
clear rank-order relationship between in vitro testing (e.g., in vitro release/dissolution) and in	227
vivo PK study results.	228
229
To increase confidence in the model, we strongly recommend that sponsors demonstrate the	230
model’s predictive performance based on PK data from batches exhibiting unacceptable BA, in	231
addition to those that exhibited acceptable BA (compared to a target and/or reference product).	232
In this context, BA would be considered unacceptable when, based on BE criteria, the 90 percent	233
confidence interval of the test-to-reference geometric mean ratio of C max and AUC fall outside	234
the range of 80 to 125 percent. Model validation acceptance criteria should be established a	235
priori and the criteria should be appropriate for the specified application. For instance, the	236
acceptance criteria for a mechanistic IVIVC model to support biowaiver should comply with the	237
criteria provided in the guidance for industry Extended Release Oral Dosage Forms:	238
Development, Evaluation, and Application of In Vitro/In Vivo Correlations.	239
240

Contains Nonbinding Recommendations
Draft — Not for Implementation
9
To demonstrate model predictive performance, sponsors should provide graphical and numerical	241
comparisons of the predicted and observed in vivo drug concentrations (e.g., in plasma) versus	242
time profiles as well as PK parameter estimates (e.g., C max, Tmax, and AUC) and statistical	243
analysis of those estimates (e.g., confidence intervals). Any significant deviation of the model	244
prediction from clinical PK observation (e.g., failure to meet pre-defined model acceptance	245
criteria) will be subject to evaluation by FDA.	246
247
When model refinement or optimization is necessary, we recommend uncertainty analyses on	248
model structure and parameters. Such analyses can be performed by reevaluation of model	249
assumptions and/or parameter sensitivity analysis. The model structure and/or parameters that	250
are modified should be clearly presented and scientifically justified. When a model parameter is	251
optimized, sponsors should provide, in addition to the scientific justification(s) and rationale, the	252
selected initial values and range of parameters, the estimation method and optimization	253
algorithm, and the in vitro and in vivo data used for optimization.	254
255
256
V. 	PBPK MODELING FOR BIOPHARMACEUTICS APPLICATIONS TO	257
SUPPORT PRODUCT QUALITY	258
259
The in vivo prediction capability provided by PBPK modeling for biopharmaceutics applications	260
allows for a wide array of uses in the pharmaceutical industry, including formulation	261
development, biopredictive dissolution method development, clinically relevant product	262
specifications setting, quality risk assessment, and drug product life cycle management. The	263
implementation of PBPK modeling for biopharmaceutics applications may reduce the number of	264
in vivo BA/BE studies (e.g., due to formulation and/or manufacturing process changes)	265
conducted during the initial approval process, as well as support product scale-up and	266
postapproval changes (SUPAC). The major regulatory uses of PBPK models for	267
biopharmaceutics applications with respect to supporting product quality are presented below.	268
For cases not discussed in this guidance, sponsors are encouraged to contact FDA.	269
270
A. 	Development of Clinically Relevant Dissolution Specifications (Method and	271
Acceptance Criteria)	272
273
1. 	Aid in Biopredictive Dissolution Method Development	274
275
Although progress has been made with the use of biorelevant media and appropriate testing	276
conditions to create physiologically based dissolution methods (Kostewicz et al. 2014b), the use	277
of in vitro dissolution data to quantitatively predict drug absorption is challenging. A	278
biopredictive dissolution method can be used to generate a dissolution profile. This profile can	279
be used to predict systemic exposure after oral administration of solid dosage forms (Suarez-	280
Sharp et al. 2018). The dissolution data could be effectively used as a surrogate to assess the	281
clinically relevant effect of drug product variants, thereby streamlining drug product	282
development.	283
284
We encourage development of biopredictive dissolution methods at the early stage of a drug	285
product development program, especially for drug products with dissolution as the rate-limiting	286

Contains Nonbinding Recommendations
Draft — Not for Implementation
10
step for absorption, such as MR drug products and IR drug products containing poorly soluble	287
APIs.	288
289
With the mechanistic platform to delineate the complex mechanisms underlying drug absorption,	290
PBPK modeling for biopharmaceutics applications can provide an estimation of a drug product’s	291
in vivo dissolution profile, based on the simulation of in vivo process of drug absorption in the	292
GI tract. Although the estimated in vivo dissolution profile may be used as a reference,	293
understanding of the physicochemical properties of the API and drug product quality attributes,	294
as well as their potential effect on in vivo dissolution, is critical for the development of a	295
biopredictive/clinically relevant dissolution method. The critical physicochemical properties of	296
the API and drug product quality attributes include, but are not limited to: (1) its solubility in	297
aqueous media within physiological pH range (e.g., 1 to 6.8) and/or biorelevant media (e.g.,	298
simulated gastric and intestinal fluid mimicking fasted or fed conditions); (2) saturation or	299
supersaturation and precipitation properties; (3) mechanism of release; and (4) the in vitro	300
dissolution characteristics in media at different pH within the physiological range.	301
302
By exploring dissolution methodologies (e.g., medium, apparatus, and hydrodynamics), an in	303
vitro dissolution method can be developed with the intention to predict the in vivo dissolution	304
profile of a drug product. The use of biorelevant dissolution methodology is encouraged as a	305
starting point in the development of a biopredictive dissolution method.	306
307
To evaluate whether a dissolution method is biopredictive, sponsors should incorporate	308
dissolution profiles generated by such method into the PBPK model and the predicted systemic	309
exposure should be comparable (±10 percent) to the observed in vivo PK data. To evaluate the	310
method, we recommend that sponsors use observed in vivo PK data of formulations with	311
different release rates.	312
313
Implementation of a biopredictive method is encouraged for establishing a quality control (QC)	314
dissolution test. If a biopredictive method is determined inappropriate to be employed for routine	315
use (e.g., due to method complexity), an alternative dissolution method can be selected as the	316
primary method for the QC dissolution test while the biopredictive dissolution method can be	317
retained as an alternate testing approach for aiding in quality assessment when needed. For	318
instance, dissolution data from a biopredictive method can supersede the primary QC dissolution	319
testing results to support chemistry, manufacturing, and controls (CMC) changes in terms of	320
maintaining desired in vivo performance.	321
322
2. 	Support Clinically Relevant Dissolution Acceptance Criteria	323
324
The term clinically relevant dissolution acceptance criteria is defined as a metric that can	325
identify and reject drug product batches that are not bioequivalent to the pivotal clinical drug	326
product (Abend et al. 2018). It often refers to the acceptance criteria set for a dissolution method	327
to minimize the possibility of releasing batches that would have clinical performance differences.	328
A clinically relevant dissolution acceptance criterion can be wider than that set based on the	329

Contains Nonbinding Recommendations
Draft — Not for Implementation
11
average dissolution data of pivotal clinical batches (e.g., beyond plus or minus 10 percent	330
variation range for an extended-release (ER) drug product).13	
331
332
PBPK modeling for biopharmaceutics applications links in vitro dissolution to PK performance	333
and hence supports the establishment of clinically relevant dissolution acceptance criteria. PK	334
predictions from PBPK models for biopharmaceutics applications (e.g., virtual BE studies) based	335
on in vitro dissolution profile(s) representing the desired limits and/or range of dissolution rate	336
can be used to justify the clinical relevance of proposed acceptance criteria. The approach should	337
consider comparing PK predictions based on in vitro dissolution profile(s) representing the	338
desired dissolution limits and PK predictions based on dissolution profile of pivotal clinical	339
batches (as a reference). Sponsors should consider the following when conducting virtual BE	340
studies: (1) the estimated intra- and intersubject variability for PK parameters (such as C max and	341
AUC) should be representative of the observed intra- and intersubject variability; (2) the number	342
of subjects for virtual BE trials should be justified and comparable to in vivo BE studies; and (3)	343
the number of virtual BE trials used to estimate the probability of concluding BE should be	344
justified.	345
346
If the outcome of the analysis meets BE acceptance criteria, the proposed dissolution acceptance	347
criterion could support clinical relevance. For ER drug products, clinically relevant dissolution	348
acceptance criteria preferably should be set such that all lots/batches that have dissolution	349
profiles within the upper and lower limits of the specification are bioequivalent to one another.	350
Less optimally, lots/batches exhibiting dissolution profiles at the upper and lower dissolution	351
limits should be bioequivalent to the clinical/BA lots/batches or to an appropriate reference	352
standard, but not necessarily to one another.14 For generic drug products, the predicted PK	353
performance corresponding to the upper and lower limits of dissolution should support that	354
product variants are bioequivalent to each other and to the reference listed drug.	355
356
Parameter sensitivity analysis (PSA) also can be performed to evaluate the effect of the	357
dissolution rate change on systemic exposure using validated PBPK models for	358
biopharmaceutics applications, in support of the clinical relevance of proposed dissolution	359
acceptance criteria.	360
361
B. 	Establishment of Clinically Relevant Drug Product Quality Specifications	362
(Other Than Dissolution)	363
364
QbD is a systematic approach for pharmaceutical development and manufacturing to enhance	365
drug product quality with more consideration of the drug product’s intended use by the patients;	366
nevertheless, it is often challenging to establish clinically relevant specifications for drug	367
substances, excipients, in-process materials, and finished drug products. Current quality testing	368
or control is largely based on in vitro testing/performance (including in vitro dissolution) of	369
clinical, development, and registration batches. Although clinical data alone are often insufficient	370
to inform appropriate drug product specifications, the overall clinical pharmacology and	371
13 See the guidance for industry Extended Release Oral Dosage Forms: Development, Evaluation, and Application
of In Vitro/In Vivo Correlations.
14 See the guidance for industry Dissolution Testing of Immediate Release Solid Oral Dosage Forms.

Contains Nonbinding Recommendations
Draft — Not for Implementation
12
biopharmaceutics information gathered during drug product development can be leveraged for	372
the development of PBPK models for biopharmaceutics applications. These models may help in	373
establishing a desired in vitro-in vivo link, a key element in building clinical relevance for drug	374
product quality attributes.	375
376
Provided that the quality attributes and process parameters are incorporated either directly or	377
indirectly in the model, the effect of these attributes and parameters on in vivo dissolution and	378
absorption can be assessed. The quality attributes can include: (1) drug substance quality	379
attributes (e.g., particle size distribution, physical form, polymorphic form); (2) excipient quality	380
attributes (e.g., type and/or level of release rate controlling excipient); (3) in-process quality	381
attributes (e.g., granule particle size, coating weight gain); and (4) finished drug product	382
attributes (e.g., disintegration). Manufacturing process parameters include, but are not limited to,	383
coating parameters and compression force. A biopredictive dissolution profile can be used to	384
assess the in vivo effect of the quality attributes and process parameters that cannot be directly	385
input into the model.	386
387
Similar to setting clinically relevant dissolution acceptance criteria, clinically relevant drug	388
product specifications for quality attributes other than dissolution can be established based on	389
modeling predictions to ensure BE of batches within the specification limits to the pivotal	390
clinical/BA batches (see section V.A.2., Support Clinically Relevant Dissolution Acceptance	391
Criteria) or to the reference listed drug for generic drugs.	392
393
C. 	Quality Risk Assessment for Pre- and Postapproval Changes and Risk-Based	394
Biowaivers	395
396
Sponsors can employ PBPK analyses for biopharmaceutics applications as an advanced tool for	397
quality risk assessment and management in pharmaceutical development and drug product life	398
cycle. Specifically, enhanced understanding can be provided by the modeling approach on how	399
quality attributes affect clinical performance, thereby aiding in risk assessment as part of	400
formulation and process development and the establishment of the control strategy, as well as	401
supporting postapproval changes.	402
403
The use of PBPK analyses for biopharmaceutics applications at pre- and postapproval stages can	404
include:	405
406
• 	Preapproval Stage	407
408
‒ 	Establishing clinically relevant manufacturing design space and control strategy to	409
mitigate quality risks in support of patient-centric drug product development	410
411
‒ 	Bridging clinical batches to the to-be-marketed commercial product accounting for	412
the CMC changes such as formulation, manufacturing process, and manufacturing	413
site changes made during pharmaceutical development	414
415

Contains Nonbinding Recommendations
Draft — Not for Implementation
13
• 	Post-approval Stage	416
417
‒ 	Conducting risk assessment/risk classification as per SUPAC and/or the draft ICH	418
guidances for industry Q12 Technical and Regulatory Considerations for	419
Pharmaceutical Product Lifecycle Management: Core Guideline (November 2017)	420
and Q12 Technical and Regulatory Considerations for Pharmaceutical Product	421
Lifecycle Management: Annex (November 2017)15 on postapproval CMC changes	422
such as formulation, manufacturing process, and manufacturing site changes	423
424
‒ 	Supporting biowaivers for postapproval changes	425
426
Risk assessment can be performed using the same approach as illustrated in section V.A. and B.	427
in setting clinically relevant drug product specifications. In this regard, model prediction(s) or	428
PSA results may be used to support high-impact CMC changes that may otherwise need an in	429
vivo BE study per the guidances for industry Immediate Release Solid Oral Dosage Forms	430
Scale-Up and Postapproval Changes: Chemistry, Manufacturing, and Controls, In Vitro	431
Dissolution Testing, and In Vivo Bioequivalence Documentation and SUPAC-MR: Modified	432
Release Solid Oral Dosage Forms Scale-Up and Postapproval Changes: Chemistry,	433
Manufacturing, and Controls; In Vitro Dissolution Testing and In Vivo Bioequivalence	434
Documentation (September 1997). FDA may grant a biowaiver request supported by PBPK	435
modeling for biopharmaceutics applications after evaluation of the outcome of the risk	436
assessment, the level of impact of the proposed change, and the totality of the provided	437
information.	438
439
15 When final, these guidances will represent the FDA’s current thinking on these topics.

Contains Nonbinding Recommendations
Draft — Not for Implementation
14
GLOSSARY	440
441
Biopharmaceutics: The study of the physical and chemical properties of a drug, its dosage form,	442
and formulation, as related to the onset, duration, and intensity of drug action.	443
444
Biopredictive dissolution method: A set of testing conditions for which in vitro dissolution	445
profiles are capable of predicting PK profiles. These are typically based on classical or	446
mechanistic IVIVC.	447
448
Biorelevant dissolution method: A set of testing conditions (e.g., media and hydrodynamics)	449
for monitoring in vitro dissolution designed to closely mimic a relevant biological fluid and a	450
physiological environment.	451
452
Clinically relevant dissolution specification: A specification that takes into consideration the	453
clinical effect of variations in dissolution ensuring a consistent safety and efficacy profile.	454
455
Critical process parameter (CPP): A process parameter whose variability has an effect on a	456
CQA and therefore should be monitored or controlled to ensure the process produces the desired	457
quality.16	
458
459
Critical quality attribute (CQA): A physical, chemical, biological, or microbiological property	460
or characteristic that should be within an appropriate limit, range, or distribution to ensure the	461
desired drug product quality.17 CQAs are generally associated with the drug substance,	462
excipients, intermediates (in-process materials), and drug product.	463
464
Design space: The multidimensional combination and interaction of input variables (e.g.,	465
material attributes) and process parameters that have been demonstrated to ensure quality.	466
Working within the design space is not considered to be a change. Movement out of the design	467
space is considered to be a change and would normally initiate a regulatory postapproval change	468
process. Design space is proposed by the sponsor and is subject to regulatory assessment and	469
approval.18	
470
471
In vitro/in vivo correlation (IVIVC): A predictive mathematical model describing the	472
relationship between an in vitro property of an ER dosage form (usually the rate or extent of drug	473
dissolution or release) and a relevant in vivo response (e.g., plasma drug concentration or amount	474
of drug absorbed).19	
475
476
In vitro-in vivo relationship: A qualitative rank-order relationship between a relevant in vivo	477
response and in vitro release profiles.	478
16 See ICH Q8(R2).
17 Ibid.
18 Ibid.
19 See the guidance for industry Extended Release Oral Dosage Forms: Development, Evaluation, and Application
of In Vitro/In Vivo Correlations.

Contains Nonbinding Recommendations
Draft — Not for Implementation
15
479
Parameter sensitivity analysis (PSA): A series of analyses targeting the same estimand, with	480
differing assumptions to explore the robustness of inferences from the main estimator to	481
deviations from its underlying modeling assumptions and limitations in the data.20	
482
483
Physiologically based pharmacokinetic (PBPK) analysis: An analysis using models and	484
simulations that combine physiology, population, and drug characteristics to mechanistically	485
describe the PK and/or pharmacodynamic behaviors of a drug product.21	
486
487
Risk assessment: A systematic process of organizing information to support a risk decision to be	488
made within a risk management process. It consists of the identification of hazards and the	489
analysis and evaluation of risks associated with exposure to those hazards.22	
490
491
Safe space: Boundaries defined by in vitro specifications, such as dissolution or other relevant	492
drug product quality attributes, within which drug product variants are anticipated to be	493
bioequivalent to one another.	494
20 See the draft ICH guidance for industry E9(R1) Statistical Principles for Clinical Trials: Addendum: Estimands
and Sensitivity Analysis in Clinical Trials (June 2017). When final, this guidance will represent the FDA’s current
thinking on this topic.
21 See the guidance for industry Physiologically Based Pharmacokinetic Analyses — Format and Content.
22 See the ICH guidance for industry Q9 Quality Risk Management (June 2006).

Contains Nonbinding Recommendations
Draft — Not for Implementation
16
REFERENCES	495
496
Abend A, Heimbach T, Cohen M, Kesisoglou F, Pepin X, and Suarez-Sharp S, 2018, Dissolution	497
and Translational Modeling Strategies Enabling Patient-Centric Drug Product Development: The	498
M-CERSI Workshop Summary Report, AAPS J, 20:60.	499
500
Babiskin A and Zhang X, 2015, Application of Physiologically Based Absorption Modeling for	501
Amphetamine Salts Drug Products in Generic Drug Evaluation, J Pharm Sci, Sep;104(9):3170–	502
3182.	503
504
Heimbach T, Suarez-Sharp S, Kakhi M, Holmstock N, Olivares-Morales A, Pepin X, Sjögren E,	505
Tsakalozou E, Seo P, Li M, Zhang X, Lin HP, Montague T, Mitra A, Morris D, Patel N, and	506
Kesisoglou F, 2019, Dissolution and Translational Modeling Strategies Toward Establishing an	507
In Vitro-In Vivo Link — A Workshop Summary Report, AAPS J, Feb 11;21(2):29.	508
509
Huang SM, Abernethy DR, Wang Y, Zhao P, and Zineh I, 2013, The Utility of Modeling and	510
Simulation in Drug Development and Regulatory Review, J Pharm Sci, 102(9):2912–2923.	511
512
Kesisoglou F, Chung J, Van Asperen J, and Heimbach T, 2016, Physiologically Based	513
Absorption Modeling to Impact Biopharmaceutics and Formulation Strategies in Drug	514
Development-Industry Case Studies, J Pharm Sci, Sep;105(9):2723–2734.	515
516
Kostewicz ES, Aarons L, Bergstrand M, Bolger MB, Galetin A, Hatley O, Jamei M, Lloyd R,	517
Pepin X, Rostami-Hodjegan A, Sjögren E, Tannergren C, Turner DB, Wagner C, Weitschies W,	518
and Dressman J, 2014a, PBPK Models for the Prediction of In Vivo Performance of Oral Dosage	519
Forms, Eur. J. Pharm. Sci., 57:300–321.	520
521
Kostewicz ES, Abrahamsson B, Brewster M, Brouwers J, Butler J, Carlert S, Dickinson PA,	522
Dressman J, Holm R, Klein S, Mann J, McAllister M, Minekus M, Muenster U, Müllertz A,	523
Verwei M, Vertzoni M, Weitschies W, and Augustijns P, 2014b, In Vitro Models for the	524
Prediction of In Vivo Performance of Oral Dosage Forms, Eur J Pharm Sci, Jun 16;57:342–366.	525
526
Kuemmel C, Yang Y, Zhang X, Florian J, Zhu H, Tegenge M, Huang SM, Wang Y, Morrison T,	527
and Zineh I, 2020, Consideration of a Credibility Assessment Framework in Model-Informed	528
Drug Development: Potential Application to Physiologically-Based Pharmacokinetic Modeling	529
and Simulation, CPT Pharmacometrics Syst Pharmacol, Jan;9(1):21–28.	530
531
Nguyen MA, Flanagan T, Brewster M, Kesisoglou F, Beato S, Biewenga J, Crison J, Holm R, Li	532
R, Mannaert E, McAllister M, Mueller-Zsigmondy M, Muenster U, Ojala K, Page S, Parr A,	533
Rossenu S, Timmins P, Van Peer A, Vermeulen A, and Langguth P, 2017, A Survey on	534
IVIVC/IVIVR Development in the Pharmaceutical Industry – Past Experience and Current	535
Perspectives, Eur J Pharm Sci, May 1;102:1–13.	536
537
Suarez-Sharp S, Cohen M, Kesisoglou F, Abend A, Marroum P, Delvadia P, Kotzagiorgis E, Li	538
M, Nordmark A, Bandi N, Sjögren E, Babiskin A, Heimbach T, Kijima S, Mandula H, Raines K,	539

Contains Nonbinding Recommendations
Draft — Not for Implementation
17
Seo P, and Zhang X, 2018, Applications of Clinically Relevant Dissolution Testing: Workshop	540
Summary Report, AAPS J, Aug 27;20(6):93.	541
542
Suarez-Sharp S, Li M, Duan J, Shah H, and Seo P, 2016, Regulatory Experience With In Vivo In	543
Vitro Correlations (IVIVC) in New Drug Applications, AAPS J, Nov;18(6):1379–1390.	544
545
Wagner C, Pan Y, Hsu V, Grillo JA, Zhang L, Reynolds KS, Sinha V, and Zhao P, 2015a,	546
Predicting the Effect of Cytochrome P450 Inhibitors on Substrate Drugs: Analysis of	547
Physiologically Based Pharmacokinetic Modeling Submissions to the US Food and Drug	548
Administration, Clin Pharmacokinet, 54(1):117–127.	549
550
Wagner C, Pan Y, Hsu V, Sinha V, and Zhao P, 2016, Predicting the Effect of CYP3A Inducers	551
on the Pharmacokinetics of Substrate Drugs Using Physiologically Based Pharmacokinetic	552
(PBPK) Modeling: An Analysis of PBPK Submissions to the US FDA, Clin Pharmacokinet,	553
55(4):475–483.	554
555
Wagner C, Zhao P, Pan Y, Hsu V, Grillo J, Huang SM, and Sinha V, 2015b, Application of	556
Physiologically Based Pharmacokinetic (PBPK) Modeling to Support Dose Selection: Report of	557
an FDA Public Workshop on PBPK, CPT Pharmacometrics Syst Pharmacol, 4(4):226–230.	558
559
Zhang X, Duan J, Kesisoglou F, Novakovic J, Amidon GL, Jamei M, Lukacova V, Eissing T,	560
Tsakalozou E, Zhao L, and Lionberger R, 2017, Mechanistic Oral Absorption Modeling and	561
Simulation for Formulation Development and Bioequivalence Evaluation: Report of an FDA	562
Public Workshop, CPT Pharmacometrics Syst. Pharmacol, 6:492–495.	563
564
Zhang X, Lionberger RA, Davit BM, and Yu LX, 2011, Utility of Physiologically Based	565
Absorption Modeling in Implementing Quality by Design in Drug Development, AAPS J,	566
Mar;13(1):59–71.	567
568