multisource (generic) pharmaceutical … · working document qas/14.583/rev.1 july 2014 document...
TRANSCRIPT
Working document QAS/14.583/Rev.1
July 2014
Document for comment
1
2
MULTISOURCE (GENERIC) PHARMACEUTICAL 3
PRODUCTS: GUIDELINES ON REGISTRATION 4
REQUIREMENTS TO ESTABLISH 5
INTERCHANGEABILITY. REVISION 6
(JULY 2014) 7
8
REVISED DRAFT FOR COMMENT 9
10
11
© World Health Organization 2014 12
All rights reserved. 13
This draft is intended for a restricted audience only, i.e. the individuals and organizations having received this draft. 14 The draft may not be reviewed, abstracted, quoted, reproduced, transmitted, distributed, translated or adapted, in part or 15 in whole, in any form or by any means outside these individuals and organizations (including the organizations' 16 concerned staff and member organizations) without the permission of the World Health Organization. The draft should 17 not be displayed on any website. 18
Please send any request for permission to: 19
Dr Sabine Kopp, Group Lead, Medicines Quality Assurance, Technologies, Standards and Norms, Department of 20 Essential Medicines and Health Products, World Health Organization, CH-1211 Geneva 27, Switzerland. Fax: (41-22) 21 791 4730; email: [email protected]. 22
The designations employed and the presentation of the material in this draft do not imply the expression of any opinion 23 whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or 24 area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent 25 approximate border lines for which there may not yet be full agreement. 26
The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or 27 recommended by the World Health Organization in preference to others of a similar nature that are not mentioned. 28 Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters. 29
All reasonable precautions have been taken by the World Health Organization to verify the information contained in 30 this draft. However, the printed material is being distributed without warranty of any kind, either expressed or implied. 31 The responsibility for the interpretation and use of the material lies with the reader. In no event shall the World Health 32 Organization be liable for damages arising from its use. 33
This draft does not necessarily represent the decisions or the stated policy of the World Health Organization. 34 35
Should you have any comments on the attached text, please send these to:
Dr Sabine Kopp, Group Lead, Medicines Quality Assurance, Technologies, Standards and Norms, World
Health Organization, 1211 Geneva 27, Switzerland; email: [email protected]; fax: (+41 22) 791 4730
([email protected]) and to Ms Marie Gaspard ([email protected]), by 15 September 2014.
Working documents are sent out electronically and they will also be placed on the Medicines website
for comment. If you do not already receive directly our draft guidelines please let us have your email
address (to [email protected]) and we will add it to our electronic mailing list.
Working document QAS/14.583/Rev.1
page 2
SCHEDULE FOR THE PROPOSED ADOPTION PROCESS OF DOCUMENT QAS/14.583: 36
MULTISOURCE (GENERIC) PHARMACEUTICAL PRODUCTS: GUIDELINES ON 37
REGISTRATION REQUIREMENTS TO ESTABLISH INTERCHANGEABILITY. 38
REVISION 39
40
41
42
Preliminary feedback for chapters to be revised and
discussion during informal consultation with experts and
prequalification assessors
July 2013
Presentation of the outcome of the above consultation to
forty-eighth meeting of the WHO Expert Committee on
Specifications for Pharmaceutical Preparations
October 2013
Revision of guidelines developed by Mr J. Gordon with
input from the Prequalification Team and Professor J.D.
Dressman, WHO Collaborating Centre on
Bioequivalence Testing of Medicines, Frankfurt/Main,
Germany
April 2014
Circulation for comment May–June 2014
Collation of comments 2 July 2014
Discussion during an informal consultation held together
with the Prequalification Assessment Team
6–7 July 2014
Circulation of revised working document July 2014
Collation of comments and review by expert group September 2014
Presentation to forty-ninth meeting of the WHO Expert
Committee on Specifications for Pharmaceutical
Preparations
October 2014
Further follow-up action as required …
Working document QAS/14.583/Rev.1
page 3
BACKGROUND 43
Over the course of time and especially in view of the implementation of the existing 44
guidelines1 the users have indicated that there was a need to review and update certain 45
requirements. The forty-eighth meeting of the World Health Organization (WHO) Expert 46
Committee on Specifications for Pharmaceutical Products discussed this and included the 47
following passage in its report (extract from the forty-eighth report): 48
“WHO published Multisource (generic) pharmaceutical products: Guidelines on 49
registration requirements to establish interchangeability in 2006. In preparation 50
for the revision of the WHO document, the Expert Committee received a report 51
that compared the WHO guidance on interchangeability with that of the European 52
Medicines Agency (EMA) and the United States Food and Drug Administration 53
(US-FDA). Guidance issues proposed for revision were highlighted. It was stated 54
that a first draft of the revision would be available in due course. The Committee 55
was also asked to consider to what extent WHO guidance should extend to non-56
biological complex drugs (NBCDs). The Committee agreed that guidance for this 57
new group of products might be relevant if they were included in the EML and 58
asked the Secretariat to look into it.” 59
In connection with the revision of these guidelines, the following related guidance texts 60
are also under review and update: 61
• Proposal to waive in vivo bioequivalence requirements for WHO Model List of 62
Essential Medicines immediate-release, solid oral dosage forms 63
Annex 8, WHO Technical Report Series 937, 2006; 64
65
1 Multisource (generic) pharmaceutical products: guidelines on registration requirements to establish
interchangeability. In: WHO Expert Committee on Specifications for Pharmaceutical Products. Fortieth
report. World Health Organization, Geneva. WHO Technical Report Series. No. 937, Annex 7, 2006.
Working document QAS/14.583/Rev.1
page 4
• Additional guidance for organizations performing in vivo bioequivalence studies 66
Annex 9, WHO Technical Report Series 937, 2006; 67
68
• Guidance on the selection of comparator pharmaceutical products for equivalence 69
assessment of interchangeable multisource (generic) products 70
Annex 11, WHO Technical Report Series 902, 2002; 71
72
• Guidance on the selection of pharmaceutical products for assessment of 73
interchangeable multisource (generic) products (working document QAS/14.594); 74
75
• List of international comparator products (working document QAS/14.595). 76
77
Working document QAS/14.583/Rev.1
page 5
MULTISOURCE (GENERIC) PHARMACEUTICAL PRODUCTS: GUIDELINES 78
ON REGISTRATION REQUIREMENTS TO ESTABLISH 79
INTERCHANGEABILITY. REVISION 80
Background 81
1. Introduction 82
2. Glossary 83
3. Documentation of equivalence for marketing authorization 84
4. When equivalence studies are not necessary 85
5. When equivalence studies are necessary and types of studies required 86
5.1 In vivo studies 87
5.2 In vitro studies 88
6. In vivo equivalence studies in humans 89
6.1 General considerations 90
6.1.1 Provisions for studies in humans 91
6.1.2 Justification of human bioequivalence studies 92
6.1.3 Selection of investigators 93
6.1.4 Study protocol 94
7. Pharmacokinetic comparative bioavailability (bioequivalence) studies in humans 95
7.1 Design of pharmacokinetic studies 96
7.1.1 Alternative study designs for studies in patients 97
7.1.2 Considerations for active pharmaceutical ingredients with long 98
elimination half-lives 99
7.1.3 Considerations for multiple-dose studies 100
7.1.4 Considerations for modified-release products 101
7.2 Subjects 102
7.2.1 Number of subjects 103
7.2.2 Drop-outs and withdrawals 104
7.2.3 Exclusion of subject data 105
7.2.4 Selection of subjects 106
7.2.5 Monitoring the health of subjects during the study 107
7.2.6 Considerations for genetic phenotyping 108
Working document QAS/14.583/Rev.1
page 6
7.3 Investigational product 109
7.3.1 Multisource pharmaceutical product 110
7.3.2 Choice of comparator product 111
7.4 Study conduct 112
7.4.1 Selection of strength 113
7.4.2 Study standardization 114
7.4.3 Co-administration of food and fluid with the dose 115
7.4.4 Wash-out interval 116
7.4.5 Sampling times 117
7.4.6 Sample fluids and their collection 118
7.4.7 Parameters to be assessed 119
7.4.8 Studies of metabolites 120
7.4.9 Measurement of individual enantiomers 121
7.5 Quantification of active pharmaceutical ingredient 122
7.6 Statistical analysis 123
7.6.1 Two-stage sequential design 124
7.7 Acceptance ranges 125
7.8 Reporting of results 126
7.9 Special considerations 127
7.9.1 Fixed-dose combination products 128
7.9.2 Clinically important variations in bioavailability 129
7.9.3 “Highly variable active pharmaceutical ingredients” 130
8. Pharmacodynamic studies 131
9. Clinical trials 132
10. In vitro equivalence testing 133
10.1 In vitro equivalence testing in the context of the Biopharmaceutics 134
Classification System 135
10.1.1 Biopharmaceutics Classification System 136
10.1.2 Determination of dissolution characteristics of multisource 137
products in consideration of a biowaiver based on the 138
Biopharmaceutics Classification System 139
Working document QAS/14.583/Rev.1
page 7
10.2 Qualification for a biowaiver based on the Biopharmaceutics Classification 140
System 141
10.2.1 Dissolution criteria for biowaivers based on the Biopharmaceutics 142
Classification System according to the properties of active 143
pharmaceutical ingredients 144
10.3 In vitro equivalence testing based on dose-proportionality of formulations 145
10.3.1 Proportional formulations 146
10.3.2 Qualification for biowaivers based on dose-proportionality of 147
formulations 148
10.3.3 Dissolution profile comparison for biowaivers based on dose- 149
proportionality of formulations 150
10.4 In vitro equivalence testing for non-oral dosage forms 151
10. 5 In vitro equivalence testing for scale-up and post-approval changes 152
References 153
Annex. Recommendations for conducting and assessing comparative dissolution profiles 154
155
1. INTRODUCTION 156
157
These guidelines are intended to provide recommendations to regulatory authorities when 158
defining requirements for approval of multisource (generic) pharmaceutical products in 159
their respective countries. The guidance provides appropriate in vivo and in vitro 160
requirements to assure interchangeability of the multisource product without 161
compromising the safety, quality and efficacy of the pharmaceutical product. 162
163
National regulatory authorities (NRA) should ensure that all pharmaceutical products 164
subject to their control conform to acceptable standards of safety, efficacy and quality, 165
and that all premises and practices employed in the manufacture, storage and distribution 166
of these products comply with good manufacturing practice (GMP) standards so as to 167
ensure the continued conformity of the products with these requirements until they are 168
delivered to the end-user. 169
170
Working document QAS/14.583/Rev.1
page 8
All pharmaceutical products, including multisource products, should be used in a country 171
only after approval by the national or regional authority. Regulatory authorities should 172
require the documentation of a multisource pharmaceutical product to meet the following: 173
174
‒ GMP; 175
‒ quality control specifications; 176
‒ pharmaceutical product interchangeability. 177
178
Multisource pharmaceutical products need to conform to the same appropriate standards 179
of quality, efficacy and safety as those required of the innovator’s (comparator) product. 180
In addition, reasonable assurance must be provided that the multisource product is 181
therapeutically equivalent and interchangeable with the comparator product. For some 182
classes of products, including – most evidently – aqueous parenteral solutions, 183
interchangeability is adequately assured by assessment of the composition, 184
implementation of GMP and evidence of conformity with appropriate specifications 185
including relevant pharmacopoeial specifications. For a wide range of pharmaceutical 186
products, the concepts and approaches covered by these guidelines will enable the 187
national regulatory authority to decide whether a given multisource product can be 188
approved. This guidance is generally applicable to orally-administered multisource 189
products, as well as to non-orally-administered pharmaceutical products for which 190
systemic exposure measures are suitable for documenting bioequivalence (e.g. 191
transdermal delivery systems and certain parenteral, rectal and nasal pharmaceutical 192
products). Some information applicable for locally acting products is also provided in this 193
document. For yet other classes of products, including many biologicals such as vaccines, 194
animal sera, products derived from human blood and plasma and products manufactured 195
by biotechnology, as well as non-biological complex products, the concept of 196
interchangeability raises issues that are beyond the scope of this document and these 197
products are consequently excluded from consideration. 198
199
To ensure interchangeability, the multisource product must be therapeutically equivalent 200
to the comparator product. Types of in vivo equivalence studies include comparative 201
Working document QAS/14.583/Rev.1
page 9
pharmacokinetic studies, comparative pharmacodynamic studies and comparative clinical 202
studies. 203
204
Direct demonstration of therapeutic equivalence through a comparative clinical trial is 205
rarely a practical choice, as these trials tend to be insensitive to formulation differences 206
and usually require a very large number of patients. Further, such studies in humans can 207
be financially daunting, are often unnecessary and may be unethical. For these reasons 208
the science of bioequivalence testing has been developed over the last 50 years. 209
According to the tenets of this science, therapeutic equivalence can be assured when the 210
multisource product is both pharmaceutically equivalent/alternative and bioequivalent. 211
212
Assuming that, in the same subject, an essentially similar plasma concentration-time 213
course will result in essentially similar concentrations at the site(s) of action and thus an 214
essentially similar therapeutic outcome, pharmacokinetic data may be used instead of 215
therapeutic results. Further, in selected cases, in vitro comparison of the dissolution 216
profiles of the multisource product with those of the comparator product may be 217
sufficient to provide an indication of equivalence. 218
219
It should be noted that interchangeability includes the equivalence of the dosage form as 220
well as of the indications and instructions for use. Alternative approaches to the 221
principles and practices described in this document may be acceptable provided they are 222
supported by adequate scientific justification. These guidelines should be interpreted and 223
applied without prejudice to obligations incurred through existing international 224
agreement on trade-related aspects of intellectual property rights (1). 225
226
2. GLOSSARY 227
228
Some important terms used in these guidelines are defined below. They may 229
have different meanings in other contexts. 230
231
232
Working document QAS/14.583/Rev.1
page 10
bioavailability 233
The rate and extent to which the active moiety is absorbed from a pharmaceutical dosage 234
form and becomes available at the site(s) of action. Reliable measurements of active 235
pharmaceutical ingredient (API) concentrations at the site(s) of action are usually not 236
possible. The substance in the systemic circulation, however, is considered to be in 237
equilibrium with the substance at the site(s) of action. Bioavailability can be therefore 238
defined as the rate and extent to which the API or active moiety is absorbed from a 239
pharmaceutical dosage form and becomes available in the systemic circulation. Based on 240
pharmacokinetic and clinical considerations it is generally accepted that in the same 241
subject an essentially similar plasma concentration time course will result in an 242
essentially similar concentration time course at the site(s) of action. 243
244
bioequivalence 245
Two pharmaceutical products are bioequivalent if they are pharmaceutically equivalent or 246
pharmaceutical alternatives, and their bioavailabilities, in terms of rate (Cmax and tmax) 247
and extent of absorption (area under the curve (AUC), after administration of the same 248
molar dose under the same conditions, are similar to such a degree that their effects can 249
be expected to be essentially the same. 250
251
biological medicinal product (ECBS) 252
Biological medicinal product is a synonym for biological product or biological described 253
in the WHO Technical Report Series. The definition of a medicinal substance used in 254
treatment, prevention or diagnosis as a“biological” has been variously based on criteria 255
related to its source, its amenability to characterization by physicochemical means alone, 256
the requirement for biological assays, or arbitrary systems of classification applied by 257
regulatory authorities. For the purposes of WHO, including the current document, the list 258
of substances considered to be biologicals is derived from their earlier definition as 259
“substances which cannot be fully characterized by physicochemical means alone, and 260
which therefore require the use of some form of bioassay”. However, developments in 261
the utility and applicability of physicochemical analytical methods, improved control of 262
biological and biotechnology-based production methods, and an increased applicability of 263
Working document QAS/14.583/Rev.1
page 11
chemical synthesis to larger molecules have made it effectively impossible to base a 264
definition of a biological on any single criterion related to methods of analysis, source or 265
method of production. Nevertheless, many biologicals are produced using in vitro culture 266
systems. 267
268 In small print: Developers of such medicinal products that do not fit the definition of biological medicinal 269
products provided in this document should consult the relevant NRAs for product classification and the 270
licensing application pathway. 271
272
Biopharmaceutics Classification System (BCS) 273
The BCS is a scientific framework for classifying active pharmaceutical ingredients (API) 274
based upon their aqueous solubility and intestinal permeability. When combined with the 275
dissolution of the pharmaceutical product and the critical examination of the excipients of 276
the pharmaceutical product, the BCS takes into account the major factors that govern the 277
rate and extent of API absorption (exposure) from immediate-release oral solid dosage 278
forms: excipient composition, dissolution, solubility, and intestinal permeability. 279
280
biowaiver 281
The term biowaiver is applied to a regulatory pharmaceutical product approval process 282
when the dossier (application) is approved based on evidence of equivalence other than 283
through in vivo equivalence testing. 284
285
comparator product 286
The comparator product is a pharmaceutical product with which the multisource product 287
is intended to be interchangeable in clinical practice. The comparator product will 288
normally be the innovator product for which efficacy, safety and quality have been 289
established. If the innovator product is no longer marketed in the jurisdiction, the 290
selection principle as described in “Guidance on the selection of comparator 291
pharmaceutical products for equivalence assessment of interchangeable multisource 292
(generic) products” should be used to identify a suitable alternative comparator product. 293
294
295
Working document QAS/14.583/Rev.1
page 12
dosage form 296
The form of the completed pharmaceutical product, e.g. tablet, capsule, elixir or 297
suppository. 298
299
equivalence requirements 300
In vivo and/or in vitro testing requirements for approval of a multisource pharmaceutical 301
product for a marketing authorization. 302
303
equivalence test 304
A test that determines the equivalence between the multisource product and the 305
comparator product using in vivo and/or in vitro approaches. 306
307
fixed-dose combination (FDC) 308
A combination of two or more active pharmaceutical ingredients in a fixed ratio of doses. 309
This term is used generically to mean a particular combination of active pharmaceutical 310
ingredients irrespective of the formulation or brand. It may be administered as single-311
entity products given concurrently or as a finished pharmaceutical product. 312
313
fixed-dose combination finished pharmaceutical product (FDC-FPP) 314
A finished pharmaceutical product that contains two or more active pharmaceutical 315
ingredients. 316
317
generic product 318
See multisource pharmaceutical products. 319
320
innovator pharmaceutical product 321
Generally, the innovator pharmaceutical product is that which was first authorized for 322
marketing, on the basis of complete documentation of quality, safety and efficacy. 323
324
interchangeable pharmaceutical product 325
Working document QAS/14.583/Rev.1
page 13
An interchangeable pharmaceutical product is one which is therapeutically equivalent to 326
a comparator product and can be interchanged with the comparator in clinical practice. 327
328
in vitro equivalence dissolution test 329
An in vitro equivalence test is a dissolution test that includes comparison of the 330
dissolution profile between the multisource product and the comparator product, typically 331
in at least three media: pH 1.2, pH 4.5 and pH 6.8 buffer solutions. 332
333
in vitro quality control dissolution test 334
A dissolution test procedure identified in the pharmacopoeia for routine quality control of 335
product batches, generally a one-time point dissolution test for immediate-release 336
products and a three- or more time points dissolution test for modified-release products. 337
338
multisource pharmaceutical products 339
Pharmaceutically equivalent or pharmaceutically alternative products that may or may 340
not be therapeutically equivalent. Multisource pharmaceutical products that are 341
therapeutically equivalent are interchangeable. 342
343
non-biological (Oxford) 344
Not involving or derived from biology or living organisms. 345
346
pharmaceutical alternatives 347
Products are pharmaceutical alternative(s) if they contain the same active pharmaceutical 348
moiety(s) but differ in dosage form (e.g. tablets versus capsules), strength, and/or 349
chemical form (e.g. different salts, different esters). Pharmaceutical alternatives deliver 350
the same active moiety by the same route of administration but are otherwise not 351
pharmaceutically equivalent. They may or may not be bioequivalent or therapeutically 352
equivalent to the comparator product. 353
354
355
356
Working document QAS/14.583/Rev.1
page 14
pharmaceutical equivalence 357
Products are pharmaceutical equivalents if they contain the same molar amount of the 358
same active pharmaceutical ingredient(s) (API) in the same dosage form, if they meet 359
comparable standards, and if they are intended to be administered by the same route. 360
Pharmaceutical equivalence does not necessarily imply therapeutic equivalence, as 361
differences in the API solid state properties, the excipients and/or the manufacturing 362
process and some other variables can lead to differences in product performance. 363
364
quantitatively similar amounts (concentrations) of excipients 365
The relative amount of excipient present in two solid oral FPPs is considered to be 366
quantitatively similar if the differences in amount fall within the limits described in the 367
following table. 368
369
Excipient type Percentage difference (w/w) out of
total product (core) weight
Filler 5.0%
Disintegrant
Starch
Other
3.0%
1.0%
Binder 0.5%
Lubricant
Ca or Mg stearate
Other
0.25%
1.0%
Glidant
Talc
Other
1.0%
0.1%
370
If an excipient serves multiple functions (e.g. microcrystalline cellulose as a filler and as 371
a disintegrant), then the most conservative recommended range should be applied (e.g. ± 372
1.0% for microcrystalline cellulose should be applied in this example). 373
374
Working document QAS/14.583/Rev.1
page 15
The relative concentration of an excipient present in two aqueous solution FPPs is 375
considered to be similar if the difference is ≤ 10%. 376
377
therapeutic equivalence 378
Two pharmaceutical products are considered to be therapeutically equivalent if they are 379
pharmaceutically equivalent or pharmaceutical alternatives and after administration in the 380
same molar dose, their effects, with respect to both efficacy and safety, are essentially the 381
same when administered to patients by the same route under the conditions specified in 382
the labelling. This can be demonstrated by appropriate equivalence studies, such as 383
pharmacokinetic, pharmacodynamic, clinical or in vitro studies. 384
385
3. DOCUMENTATION OF EQUIVALENCE FOR MARKETING AUTHORIZATION 386
387
Multisource pharmaceutical products must be shown, either directly or indirectly, to be 388
therapeutically equivalent to the comparator product if they are to be considered 389
interchangeable. Suitable test methods to assess equivalence are: 390
391
‒ comparative pharmacokinetic studies in humans, in which the active 392
pharmaceutical ingredient (API) and/or its metabolite(s) are measured as a 393
function of time in an accessible biological fluid such as blood, plasma, serum or 394
urine to obtain pharmacokinetic measures, such as AUC and Cmax that are 395
reflective of the systemic exposure; 396
‒ comparative pharmacodynamic studies in humans; 397
‒ comparative clinical trials; 398
‒ comparative in vitro tests. 399
400
The applicability of each of these four methods is discussed below. Detailed information 401
is provided on conducting an assessment of equivalence studies using pharmacokinetic 402
measurements and in vitro methods, which are currently the methods most often used to 403
document equivalence for most orally-administered pharmaceutical products for systemic 404
exposure. 405
Working document QAS/14.583/Rev.1
page 16
406
Acceptance of any test procedure in the documentation of equivalence between two 407
pharmaceutical products by a NRA depends on many factors, including the 408
characteristics of the API and the pharmaceutical product. Where an API produces 409
measurable concentrations in an accessible biological fluid such as plasma, comparative 410
pharmacokinetic studies can be performed. This type of study is considered to be the gold 411
standard in equivalence testing; however, where appropriate, in vitro testing, e.g. BCS-412
based biowaivers for immediate-release pharmaceutical products can also assure 413
equivalence between the multisource product and the comparator product (see sections 5 414
and 10). Where an API does not produce measurable concentrations in an accessible 415
biological fluid and a BCS-based biowaiver is not an option, comparative 416
pharmacodynamics studies may be an alternative method for documenting equivalence. 417
Further, in certain cases when it is not possible to assess equivalence through other 418
methods, comparative clinical trials may be considered appropriate. 419
420
The criteria that indicate when equivalence studies are necessary are discussed in the 421
following two sections of the guideline. 422
423
4. WHEN EQUIVALENCE STUDIES ARE NOT NECESSARY 424
425
The following types of multisource pharmaceutical product are considered to be 426
equivalent without the need for further documentation: 427
428
(a) when the pharmaceutical product is to be administered parenterally (e.g. 429
intravenously, subcutaneously or intramuscularly) as an aqueous solution 430
containing the same API in the same molar concentration as the comparator product 431
and the same or similar excipients in comparable concentrations as in the 432
comparator product. Certain excipients (e.g. buffer, preservative and antioxidant) 433
may be different provided it can be shown that the change(s) in these excipients 434
would not affect the safety and/or efficacy of the pharmaceutical product. The same 435
principles are applicable for parenteral oily solutions, but in this case the use of the 436
Working document QAS/14.583/Rev.1
page 17
same oily vehicle is essential. Similarly, for micellar solutions, solutions containing 437
complexing agents or solutions containing co-solvents of the same qualitative and 438
quantitative composition of the functional excipients, is necessary to waive 439
equivalence studies and the change of other excipients should be critically reviewed; 440
441
(b) when pharmaceutically-equivalent products are solutions for oral use (e.g. syrups, 442
elixirs and tinctures), contain the API in the same molar concentration as the 443
comparator product, and contain the same excipients in similar concentrations. 444
445
(c) when pharmaceutically-equivalent products are in the form of powders for 446
reconstitution as an aqueous solution and the resultant solution meets either 447
criterion (a) or criterion (b) above; 448
449
(d) when pharmaceutically-equivalent products are gases; 450
451
(e) when pharmaceutically-equivalent products are otic or ophthalmic products 452
prepared as aqueous solutions and contain the same API(s) in the same molar 453
concentration and the same excipients in similar concentrations. Certain excipients 454
(e.g. preservative, buffer, substance to adjust tonicity or thickening agent) may be 455
different provided their use is not expected to affect bioavailability, safety and/or 456
efficacy of the product; 457
458
(f) when pharmaceutically-equivalent products are topical products prepared as 459
aqueous solutions and contain the same API(s) in the same molar concentration and 460
the same excipients in similar concentrations (note that this waiver does not apply 461
to other topical dosage forms like gels, emulsions or suspensions, but might be 462
applicable to oily solutions if the vehicle composition is sufficiently similar); 463
464
(g) when pharmaceutically-equivalent products are aqueous solutions for nebulization 465
or nasal drops, intended to be administered with essentially the same device, 466
contain the same API(s) in the same concentration and contain the same excipients 467
Working document QAS/14.583/Rev.1
page 18
in similar concentrations (note that this waiver does not apply to other dosage forms 468
like suspensions for nebulization, nasal drops where the API is in suspension, nasal 469
sprays in solution or suspension, dry powder inhalers, or pressurized metered dose 470
inhalers in solution or suspensions). The pharmaceutical product may include 471
different excipients provided their use is not expected to affect bioavailability, 472
safety and/or efficacy of the product. 473
474
For situations (b), (c), (e), (f) and (g) above, it is incumbent upon the applicant to 475
demonstrate that the excipients in the pharmaceutically-equivalent product are the same 476
and in concentrations similar to those in the comparator product or, where applicable (i.e. 477
(e) and (g)), that their use is not expected to affect the bioavailability, safety and/or 478
efficacy of the product. In the event that the applicant cannot provide this information 479
and the NRA does not have access to the relevant data, it is incumbent upon the applicant 480
to perform appropriate studies to demonstrate that differences in excipients or devices do 481
not affect product performance. 482
483
5. WHEN IN VIVO EQUIVALENCE STUDIES ARE NECESSARY AND 484
TYPES OF STUDY REQUIRED 485
486
Except for the cases discussed in section 4, these guidelines recommend that 487
documentation of equivalence with the comparator product be required by registration 488
authorities for a multisource pharmaceutical product. Studies must be carried out using 489
the product intended for marketing (see also section 7.3). 490
491
5.1 In vivo studies 492
493
For certain APIs and dosage forms in vivo documentation of equivalence, through either 494
a pharmacokinetic comparative bioavailability (bioequivalence) study, a comparative 495
pharmacodynamic study or a comparative clinical trial, is regarded as especially 496
important. In vivo documentation of equivalence is necessary when there is a risk that 497
Working document QAS/14.583/Rev.1
page 19
possible differences in bioavailability may result in therapeutic inequivalence (2). 498
Examples are listed below. 499
500
(a) Oral, immediate-release pharmaceutical products with systemic action, except for 501
the conditions outlined in section 10. 502
503
(b) Non-oral, non-parenteral pharmaceutical products designed to act systemically 504
(such as transdermal patches, suppositories, nicotine chewing gum, testosterone gel 505
and skin-inserted contraceptives). 506
507
(c) Modified-release pharmaceutical products designed to act systemically, except for 508
the conditions outlined in section 10. 509
510
(d) Fixed-dose combination (FDC)products with systemic action, where at least one of 511
the APIs requires an in vivo study (3). 512
513
(e) Non-solution pharmaceutical products, which are for non-systemic use (e.g. for oral, 514
nasal, ocular, dermal, rectal or vaginal application) and are intended to act without 515
systemic absorption. In these cases, the equivalence is established through, e.g. 516
comparative clinical or pharmacodynamic studies, local availability studies and/or 517
in vitro studies. In certain cases, measurement of the concentration of the API may 518
still be required for safety reasons, i.e. in order to assess unintended systemic 519
absorption. 520
521
5.2 In vitro studies 522
523
For certain APIs and dosage forms, in vitro documentation of equivalence may be 524
appropriate. In vitro approaches for systemically-acting oral products are discussed in 525
section 10. 526
527
6. IN VIVO EQUIVALENCE STUDIES IN HUMANS 528
Working document QAS/14.583/Rev.1
page 20
529
6.1 General considerations 530
531
6.1.1 Provisions for studies in humans 532
533
Pharmacokinetic, pharmacodynamic and clinical studies are clinical trials and should 534
therefore be carried out in accordance with the provision and prerequisites for a clinical 535
trial, as outlined in the World Health Organization (WHO) guidelines for good clinical 536
practice (GCP) for trials on pharmaceutical products (4). Additional guidance for 537
organizations performing in vivo equivalence studies is available from WHO (5). 538
539
All research involving human subjects should be conducted in accordance with the 540
ethical principles contained in the most recent version of the Declaration of Helsinki, 541
including respect for persons, beneficence (“maximize benefits and minimize harms and 542
wrongs”) and non-maleficence (“do no harm”). 543
544
As defined by the International Ethical Guidelines for Biomedical Research Involving 545
Human Subjects issued by the Council for International Organizations of Medical 546
Sciences (CIOMS), or laws and regulations of the country in which the research is 547
conducted, whichever represents the greater protection for subjects. 548
549
6.1.2 Justification of human bioequivalence studies 550
551
Most pharmacokinetic and pharmacodynamic equivalence studies are non-therapeutic 552
studies in which no direct clinical benefit accrues to the subject. 553
554
It is important for anyone preparing a trial of a medicinal product in humans that the 555
specific aims, problems and risks or benefits of the proposed human study be thoroughly 556
considered and that the chosen design be scientifically sound and ethically justified. It is 557
assumed that people involved in the planning of a study are familiar with 558
pharmacokinetic theories underlying bioavailability and bioequivalence studies. The 559
Working document QAS/14.583/Rev.1
page 21
overall design of the bioequivalence study should be based on the knowledge of the 560
pharmacokinetics, pharmacodynamics and therapeutics of the API. Information about 561
manufacturing procedures and data from tests performed on the product batch to be used 562
in the study should establish that the product under investigation is of suitable quality. 563
564
6.1.3 Selection of investigators 565
566
The investigator(s) should have the appropriate expertise, qualifications and competence 567
to undertake the proposed study. Prior to the trial the investigator(s) and the sponsor 568
should draw up an agreement on the protocol, monitoring, auditing, standard operating 569
procedures (SOP) and the allocation of trial-related responsibilities. The identity and 570
duties of the individuals responsible for the study and safety of the subjects participating 571
in the study must be specified. The logistics and premises of the trial site should comply 572
with requirements for the safe and efficient conduct of the trial. 573
574
6.1.4 Study protocol 575
576
A bioequivalence study should be carried out in accordance with a protocol agreed upon 577
and signed by the investigator and the sponsor. The protocol and its attachments and/or 578
appendices should state the aim of the study and the procedures to be used, the reasons 579
for proposing the study to be undertaken in humans, the nature and degree of any known 580
risks, assessment methodology, criteria for acceptance of bioequivalence, the groups 581
from which it is proposed that trial subjects be selected and the means for ensuring that 582
they are adequately informed before they give their consent. The investigator is 583
responsible for ensuring that the protocol is strictly followed. Any change(s) required 584
must be agreed on and signed by the investigator and sponsor and appended as 585
amendments, except when necessary to eliminate an apparent immediate hazard or 586
danger to a trial subject. 587
588
The protocol and attachments/appendices should be scientifically and ethically appraised 589
by one or, if required by local laws and regulations, more review bodies, e.g. institutional 590
Working document QAS/14.583/Rev.1
page 22
review board, peer review committee, ethics committee, NRA, constituted appropriately 591
for these purposes and independent of the investigator(s) and sponsor. 592
593
The signed and dated study protocol should be approved by the NRA before commencing 594
the study, if required by national and regional laws and regulations. The study report 595
forms the integral part of the registration dossier of the multisource product in order to 596
obtain the marketing authorization for the multisource product. 597
598
7. PHARMACOKINETIC COMPARATIVE BIOAVAILABILITY 599
(BIOEQUIVALENCE) STUDIES IN HUMANS 600
601
7.1 Design of pharmacokinetic studies 602
603
Bioequivalence studies are designed to compare the in vivo performance of a multisource 604
product with that of a comparator product. Such studies on products designed to deliver 605
the API for systemic exposure serve two purposes: 606
607
• as a surrogate for clinical evidence of the safety and efficacy of the multisource 608
product; 609
• as an in vivo measure of pharmaceutical quality. 610
611
The design of the study should maximize the sensitivity to detect difference between 612
products, minimize the variability that is not caused by formulation effects and eliminate 613
bias as far as possible. Test conditions should reduce variability within and between 614
subjects. In general, for a bioequivalence study involving a multisource product and a 615
comparator product, a randomized, two-period, two-sequence, single-dose, cross-over 616
study conducted with healthy volunteers is the preferred study design. In this design each 617
subject is given the multisource and the comparator product in randomized order. An 618
adequate wash-out period should follow the administration of each product. 619
620
Working document QAS/14.583/Rev.1
page 23
It should be noted however that under certain circumstances an alternative, well-621
established and statistically-appropriate study design may be more appropriate. 622
623
7.1.1 Alternative study designs for studies in patients 624
625
For APIs that are very potent or too toxic to administer in the highest strength to healthy 626
volunteers (e.g. because of the potential for serious adverse events, or the trial 627
necessitates a high dose), it is recommended that the study be conducted using the API at 628
a lower strength in healthy volunteers. For APIs that show unacceptable pharmacological 629
effects in healthy volunteers, even at lower strengths, a study conducted in patients may 630
be required. Depending on the dosing posology, this may be a multiple dose, steady state 631
study. As above, such studies should employ a cross-over design if possible; however, a 632
parallel group-design study in patients may be required in some situations.The use of 633
such an alternative study design should be fully justified by the sponsor and should 634
include patients whose disease process is stable for the duration of the bioequivalence 635
study if possible. 636
637
7.1.2 Considerations for active pharmaceutical ingredients with long elimination half- 638
lives 639
640
A single-dose, cross-over bioequivalence study of an orally-administered product with a 641
long elimination half-life is preferred, provided an adequate wash-out period between 642
administrations of the treatments is possible. The interval between study days should be 643
long enough to permit elimination of essentially all of the previous dose from the body. 644
Ideally the interval should not be less than five terminal elimination half-lives of the 645
active compound or metabolite, if the latter is measured. If the cross-over study is 646
problematic due to a very long elimination half-life, a bioequivalence study with a 647
parallel design may be more appropriate. A parallel design may also be necessary when 648
comparing some depot formulations. 649
650
Working document QAS/14.583/Rev.1
page 24
For both cross-over and parallel-design studies of oral products, sample collection time 651
should be adequate to ensure completion of GI transit (approximately 2–3 days) of the 652
pharmaceutical product and absorption of the API. Blood sampling should be conducted 653
for up to 72 hours following administration but sampling beyond this time is not 654
generally necessary. 655
656
The number of subjects should be derived from statistical calculations but generally more 657
subjects are needed for a parallel study design than for a cross-over study design. 658
659
7.1.3 Considerations for multiple-dose studies 660
661
In certain situations multiple-dose studies may be considered appropriate. 662
Multiple-dose studies in patients are most useful in cases where the API being studied is 663
considered to be too potent and/or too toxic to be administered to healthy volunteers, 664
even in single doses (see also 7.1.1). In this case, a multiple-dose, cross-over study in 665
patients may be performed without interrupting therapy. 666
667
The dosage regimen used in multiple-dose studies should follow the usual dosage 668
recommendations. 669
670
Other situations in which multiple-dose studies may be appropriate are as follows: 671
672
‒ cases where the analytical sensitivity is too low to adequately characterize the 673
pharmacokinetic profile after a single dose; 674
‒ extended-release dosage forms with a tendency to accumulate (in addition to 675
single-dose studies). 676
677
In steady-state studies the wash-out of the last dose of the previous treatment can overlap 678
with the approach to steady state of the second treatment, provided the approach period is 679
sufficiently long (at least five times the terminal half-life). Appropriate dosage 680
Working document QAS/14.583/Rev.1
page 25
administration and sampling should be carried out to document for the attainment of a 681
steady state. 682
683
7.1.4 Considerations for modified-release products 684
685
Modified-release products include extended-release products and delayed-release 686
products. Extended-release products are variously known as controlled-release, 687
prolonged-release and sustained-release products. 688
689
Due to the more complex nature of modified-release products relative to immediate-690
release products, additional data is required to ensure the bioequivalence of two 691
modified-release products. Factors such as the co-administration of food, which 692
influences API bioavailability and also, in certain cases, bioequivalence, must be taken 693
into consideration. The presence of food can affect product performance both by 694
influencing the release of the API from the formulation and by causing physiological 695
changes in the GI tract. In this regard a significant concern with regard to modified-696
release products is the possibility that food may trigger a sudden and abrupt release of the 697
API leading to “dose dumping”. This would most likely be manifested as a premature and 698
abrupt rise in the plasma concentration time profile. Therefore, bioequivalence studies 699
conducted under both fasted and fed conditions are required for orally-administered, 700
modified-release pharmaceutical products. 701
702
Unless single-dose studies are not possible for reasons such as those discussed above in 703
section 7.1.1 single-dose, cross-over bioequivalence studies conducted under both fasted 704
and fed conditions comparing the highest strength of the multisource product and the 705
comparator product must be performed to demonstrate bioequivalence. Single-dose 706
studies are preferred to multiple-dose studies as single-dose studies are considered to 707
provide more sensitive measurement of the release of API from the pharmaceutical 708
product into the systemic circulation. In addition to single-dose studies, multiple-dose 709
studies may be considered for extended-release dosage forms with a tendency to 710
Working document QAS/14.583/Rev.1
page 26
accumulate, e.g. after a single dose of the highest strength the AUC for the dosing 711
interval covers <90% of AUC extrapolated to infinity. 712
713
The comparator product in these studies should be a pharmaceutically-equivalent, 714
modified-release product. The bioequivalence criteria for modified-release products are 715
essentially the same as for conventional-release dosage forms except acceptance criteria 716
should also be applied to Cmin (Ctau) in case of multiple dose studies. As release 717
mechanisms of pharmaceutical products become more complex, e.g. products with an 718
immediate-release and modified-release component, additional parameters such as partial 719
AUC measures may be necessary to ensure the bioequivalence of two products. 720
721
The fed-state bioequivalence study should be conducted after the administration of an 722
appropriate standardized meal at a specified time (usually not more than 30 minutes) 723
before taking the pharmaceutical product. A meal that will promote the greatest change in 724
GI tract conditions relative to the fasted state should be employed. Refer to section 7.4.3 725
for more recommendations for the content of the meal. The composition of the meal 726
should take local diet and customs into consideration. The composition and caloric 727
breakdown of the test meal should be provided in the study protocol and report. 728
729
7.2 Subjects 730
731
7.2.1 Number of subjects 732
733
The number of subjects required for a bioequivalence study is determined by: 734
735
‒ the error variance (coefficient of variation) associated with the primary 736
parameters to be studied, as estimated from a pilot experiment, from previous 737
studies or from published data; 738
‒ the significance level desired (5%); 739
‒ the statistical power desired; 740
Working document QAS/14.583/Rev.1
page 27
‒ the mean deviation from the comparator product compatible with bioequivalence 741
and with safety and efficacy; 742
‒ the need for the 90% confidence interval around the geometric mean ratio to be 743
within bioequivalence limits, normally 80–125%, for log-transformed data. 744
745
The number of subjects to be recruited for the study should be estimated by considering 746
the standards that must be met using an appropriate method (see, for example, Julious 747
2004 (6)). In addition, an extra number of subjects should be recruited, dosed and their 748
samples analysed based on the expected rate of drop-outs and/or withdrawals, which 749
depends on the safety and tolerability profile of the API. The number of subjects 750
recruited should always be justified by the sample-size calculation provided in the study 751
protocol. A minimum of 12 subjects is required. 752
753
In some situations reliable information concerning the expected variability in the 754
parameters to be estimated may not be available. In such situations, a two-stage 755
sequential study design can be employed as an alternative to conducting a pilot study. 756
Refer to section 7.6.1 for more information. 757
758
7.2.2 Drop-outs and withdrawals 759
760
Sponsors should select a sufficient number of study subjects to allow for possible drop-761
outs or withdrawals. Because replacement of subjects during the study could complicate 762
the statistical model and analysis, drop-outs generally should not be replaced. Reasons for 763
withdrawal (e.g. adverse reaction or personal reasons) must be reported. If a subject is 764
withdrawn due to an adverse event after receiving at least one dose of the study 765
medication, the subject’s plasma/serum concentration data should be provided. 766
767
The concentration-time profiles of subjects who exhibit pre-dose concentrations higher 768
than 5% of the corresponding Cmax should be excluded from the statistical analysis. The 769
concentration-time profiles of subjects who exhibit pre-dose concentrations equal to or 770
Working document QAS/14.583/Rev.1
page 28
less than 5% of the corresponding Cmax should be included in the statistical analysis 771
without correction. 772
773
7.2.3 Exclusion of subject data 774
775
Extreme values can have a significant impact on bioequivalence study data because of the 776
relatively small number of subjects typically involved; however, it is rarely acceptable to 777
exclude data. Potential reasons for excluding subject data and the procedure to be 778
followed should be included in the study protocol. Exclusion of data for statistical or 779
pharmacokinetic reasons alone is not acceptable. Retesting of subjects is not 780
recommended. 781
782
7.2.4 Selection of subjects 783
784
Bioequivalence studies should generally be performed with healthy volunteers. Clear 785
criteria for inclusion and exclusion should be stated a priori in the study protocol. If the 786
pharmaceutical product is intended for use in both sexes, the sponsor should include both 787
males and females in the study. The potential risk to women will need to be considered 788
on an individual basis and, if necessary, they should be warned of any possible dangers to 789
the foetus if they should become pregnant. The investigators should ensure that female 790
volunteers are not pregnant or likely to become pregnant during the study. Confirmation 791
should be obtained by urine tests just before administration of the first and last doses of 792
the product under study. 793
794
Generally subjects should be between the ages of 18 and 55 years, and their weight 795
should be within the normal range with a Body Mass Index (BMI) between 18 and 30 796
kg/m2. The subjects should have no history of alcohol or drug-abuse problems and should 797
preferably be non-smokers. 798
799
The volunteers should be screened for their suitability using standard laboratory tests, a 800
medical history and a physical examination. If necessary, special medical investigations 801
Working document QAS/14.583/Rev.1
page 29
may be carried out before and during studies depending on the pharmacology of the 802
individual API being investigated, e.g. an electrocardiogram if the API has a cardiac 803
effect. The ability of the volunteers to understand and comply with the study protocol has 804
to be assessed. Subjects who are being or have previously been treated for any GI 805
problems, or convulsive, depressive or hepatic disorders, and in whom there is a risk of a 806
recurrence during the study period, should be excluded. 807
808
If a parallel design-study is planned standardization of the two groups of subjects is 809
important in order to minimize variation not attributable to the investigational products 810
(see 7.2.6.). 811
812
If the aim of the bioequivalence study is to address specific questions (e.g. 813
bioequivalence in a special population) the selection criteria should be adjusted 814
accordingly. 815
816
7.2.5 Monitoring the health of subjects during the study 817
818
In keeping with GCP (4), the health of volunteers should be monitored during the study 819
so that onset of side-effects, toxicity or any intercurrent disease may be recorded and 820
appropriate measures taken. The incidence, severity and duration of any adverse reactions 821
and side-effects observed during the study must be reported. The probability that an 822
adverse effect is API-induced is to be judged by the investigator. 823
824
Health-monitoring before, during and after the study must be carried out under the 825
supervision of a qualified medical practitioner licensed in the jurisdiction in which the 826
study is conducted. 827
828
7.2.6 Considerations for genetic phenotyping 829
830
Phenotyping for metabolizing activity can be of importance for studies with high-831
clearance APIs that are metabolized by enzymes that are subject to genetic polymorphism, 832
Working document QAS/14.583/Rev.1
page 30
e.g. propranolol. In such cases slow metabolizers will have a higher bioavailability of the 833
active parent moiety while the bioavailability of possible active metabolites will be lower. 834
Phenotyping of subjects can be considered for studies of APIs that show phenotype-835
linked metabolism and for which a parallel group design is to be used, because it allows 836
fast and slow metabolizers to be evenly distributed between the two groups of subjects. 837
838
Phenotyping could also be important for safety reasons, determination of sampling times 839
and wash-out periods in cross-over design studies. 840
841
842
7.3 Investigational product 843
844
7.3.1 Multisource pharmaceutical product 845
846
The multisource pharmaceutical product used in the bioequivalence studies for 847
registration purposes should be identical to the projected commercial pharmaceutical 848
product. Therefore, not only the composition and quality characteristics (including 849
stability), but also the manufacturing methods (including equipment and procedures) 850
should be the same as those to be used in the future routine production runs. Test 851
products must be manufactured under GMP regulations. Batch-control results, lot number, 852
manufacturing date and, if possible, expiry date for the multisource product should be 853
stated. 854
855
Samples should ideally be taken from batches of industrial scale. When this is not 856
feasible pilot or small-scale production batches may be used, provided that they are not 857
smaller than 10% of expected full production batches, or 100 000 units, whichever is 858
larger, and are produced with the same formulation and similar equipment and process as 859
that planned for commercial production batches. A biobatch of less than 100 000 units 860
may be accepted on provision that this is the proposed production batch size, with the 861
understanding that future scale up for production batches will not be accepted. 862
863
Working document QAS/14.583/Rev.1
page 31
It is recommended that potency and in vitro dissolution characteristics of the multisource 864
and the comparator pharmaceutical products be ascertained prior to performance of an 865
equivalence study. Content of the API(s) of the comparator product should be close to the 866
label claim and the difference between two products being compared should not be more 867
than ± 5%. If, because of the lack of availability of different batches of the comparator 868
product, it is not possible to study batches with potencies within ± 5%, potency correction 869
may be required on the statistical results from the bioequivalence study. 870
871
7.3.2 Choice of comparator product 872
873
The innovator pharmaceutical product is usually the most logical comparator product for 874
a multisource pharmaceutical product because its quality, safety and efficacy should have 875
been well assessed and documented in premarketing studies and postmarketing 876
monitoring schemes. Preferably this will mean employing the innovator product available 877
on the market when studying multisource products for national and regional approval. 878
There will be situations, however, where this is not feasible. Detailed guidance for the 879
selection of comparator products for use in national and regional applications is provided 880
in the comparator guidance (7). 881
882
7.4 Study conduct 883
884
7.4.1 Selection of strength 885
886
In bioequivalence studies the molar equivalent dose of multisource and comparator 887
product must be used. 888
889
For a series of strengths that can be considered proportionally formulated (see section 890
10.3), the strength with the greatest sensitivity for bioequivalence assessment should be 891
administered as a single unit. This will usually be the highest marketed strength. A higher 892
dose, i.e. more than one dosage unit, may be employed when analytical difficulties exist. 893
In this case the total single dose should not exceed the maximal daily dose of the dosage 894
Working document QAS/14.583/Rev.1
page 32
regimen. In certain cases a study performed with a lower strength can be considered 895
acceptable if this lower strength is chosen for reasons of safety or if the API is highly 896
soluble and its pharmacokinetics are linear over the therapeutic range. 897
898
7.4.1.1 Non-linear pharmacokinetics 899
900
When the API in a series of strengths, that are considered proportionally formulated, 901
exhibits non-linear pharmacokinetics over the range of strengths, special consideration is 902
necessary when selecting the strength for study. 903
904
For APIs exhibiting non-linear pharmacokinetics within the range of strengths resulting 905
in greater than proportional increases in AUC with increasing dose, the comparative 906
bioavailability study should be conducted on at least the highest marketed strength. 907
908
For APIs with non-linear pharmacokinetics within the range of strengths due to saturable 909
absorption and resulting in less than proportional increases in AUC with increasing dose, 910
the bioequivalence study should be conducted on at least the lowest strength (or a 911
strength in the linear range). 912
913
For APIs with non-linear pharmacokinetics within the range of strengths due to limited 914
solubility of the API and resulting in less than proportional increases in AUC with 915
increasing dose, bioequivalence studies should be conducted on at least the lowest 916
strength (or a strength in the linear range) and the highest strength. 917
918
7.4.2 Study standardization 919
920
Standardization of study conditions is important to minimize the magnitude of variability 921
other than in the pharmaceutical products. Standardization between study periods is 922
critical to a successful study. Standardization should cover exercise, diet, fluid intake, 923
posture, as well as the restriction of the intake of alcohol, caffeine, certain fruit juices and 924
concomitant medicines for a specified time period before and during the study. 925
Working document QAS/14.583/Rev.1
page 33
926
Volunteers should not take any other medicine, alcoholic beverages or over-the-counter 927
(OTC) medicines and supplements for an appropriate interval before or during the study. 928
In the event of emergency the use of any non-study medicine must be reported (dose and 929
time of administration). 930
931
Physical activity and posture should be standardized as far as possible to limit their 932
effects on GI blood flow and motility. The same pattern of posture and activity should be 933
maintained for each day of the study. The time of day at which the study product is to be 934
administered should be specified. 935
936
7.4.3 Co-administration of food and fluid with the dose 937
938
Finished pharmaceutical products (FPPs) are usually given after an overnight fast of at 939
least 10 hours and participants are allowed free access to water. On the morning of the 940
study no water is allowed during the hour prior to FPP administration. The dose should 941
be taken with a standard volume of water (usually 150–250 ml). Two hours after FPP 942
administration water is again permitted ad libitum. A standard meal is usually provided 943
four hours after FPP administration. All meals should be standardized and the 944
composition stated in the study protocol and report. 945
946
There are situations when the investigational products should be administered following 947
consumption of a meal (under fed conditions). These situations are described below. 948
949
7.4.3.1 Immediate-release formulations 950
951
Fasted-state studies are generally preferred. However, when the product is known to 952
cause GI disturbances if given to subjects in the fasted state, or if the labelling of the 953
comparator product restricts administration to subjects in the fed state, then the fed-state 954
study becomes the preferred approach. 955
956
Working document QAS/14.583/Rev.1
page 34
For products with specific formulation characteristics (e.g. micro-emulsions, solid 957
dispersions), bioequivalence studies performed under both fasted and fed conditions are 958
required, unless the product taken is only in a fasted or fed state. 959
960
Typically a meal meeting the composition recommendations identified below in section 961
7.4.3.2 should be employed in fed-state studies. The exact composition of the meal may 962
depend on local diet and customs as determined by the NRA. For studies conducted with 963
immediate-release products, there may be situations where it is appropriate to employ a 964
pre-dose meal with a different caloric/fat content. 965
966
The test meal should be consumed beginning 30 minutes prior to administration of the FPP. 967
968
7.4.3.2 Modified-release formulations 969
970
In addition to a study conducted under fasted conditions, food-effect studies are 971
necessary for all multisource modified-release formulations to ensure that the interaction 972
between the varying conditions in the GI tract and the product formulations does not 973
differentially impact the performance of the multisource and comparator products. The 974
presence of food can affect product performance both by influencing the release of the 975
API from the formulation and by causing physiological changes in the GI tract. A 976
significant concern with regard to modified-release products is the possibility that food 977
may trigger a sudden and abrupt release of the API leading to “dose dumping”. 978
979
In these cases the objective is to select a meal that will challenge the robustness of the 980
new multisource formulation to prandial effects on bioavailability. To achieve this a meal 981
that will provide a maximal perturbation to the GI tract relative to the fasted state should 982
be employed, e.g. a high-fat (approximately 50% of the total caloric content of the meal), 983
high-calorie (approximately 800 to 1000 kilocalories) test meal has been recommended 984
(FDA, EMA guidelines referenced). The meal selected should take into account local 985
customs and diet. The caloric breakdown of the test meal should be provided in the study 986
report. 987
Working document QAS/14.583/Rev.1
page 35
988
The test meal should be consumed within a 30-minute interval prior to administration of 989
the FPP. 990
991
7.4.4 Wash-out interval 992
993
The interval (wash-out period) between doses of each formulation should be long enough 994
to permit the elimination of essentially all of the previous dose from the body. The wash-995
out period should be the same for all subjects and should normally be more than five 996
times the median terminal half-life of the API. Consideration should be given to 997
extending this period in some situations, e.g. if active metabolites with longer half-lives 998
are produced or if the elimination rate of the API has high variability between subjects. In 999
this second case a longer wash-out period should be considered to allow for the slower 1000
elimination in subjects with lower elimination rates. Just prior to administration of the 1001
treatment during the second study period, blood samples are collected and assayed to 1002
determine the concentration of the API or metabolites. The minimum wash-out period 1003
should be at least seven days unless a shorter period is justified by a short half-life. The 1004
adequacy of the wash-out period can be estimated from the pre-dose concentrations of the 1005
API in the second study period which should be less than 5% of the observed Cmax. 1006
1007
7.4.5 Sampling times 1008
1009
Blood samples should be taken at a frequency sufficient for assessing Cmax, AUC and 1010
other parameters. Sampling points should include a pre-dose sample, at least 1–2 points 1011
before Cmax, 2 points around Cmax and 3–4 points during the elimination phase. 1012
Consequently at least seven sampling points will be necessary for estimation of the 1013
required pharmacokinetic parameters. 1014
1015
For most APIs the number of samples necessary will be higher to compensate for 1016
between-subject differences in absorption and elimination rate and thus enable accurate 1017
determination of the maximum concentration of the API in the blood (Cmax) and 1018
Working document QAS/14.583/Rev.1
page 36
terminal elimination rate constant in all subjects. Generally sampling should continue for 1019
long enough to ensure that 80% of the AUC (0→infinity) can be accrued, but it is not 1020
necessary to sample for more than 72 hours. The exact duration of sample collection 1021
depends on the nature of the API and the input function from the administered dosage 1022
form. 1023
1024
7.4.6 Sample fluids and their collection 1025
1026
Under normal circumstances blood should be the biological fluid sampled to measure the 1027
concentrations of the API. In most cases the API or its metabolites are measured in serum 1028
or plasma. If it is not possible to measure the API in blood/plasma/serum, the API is 1029
excreted unchanged in the urine and there is a proportional relationship between plasma 1030
and urine concentrations; urine can be sampled for the purpose of estimating exposure. 1031
The volume of each sample must be measured at the study centre, where possible 1032
immediately after collection and included in the report. The number of samples should be 1033
sufficient to allow the estimation of pharmacokinetic parameters. However, in most cases 1034
the exclusive use of urine excretion data should be avoided as this does not allow 1035
estimation of the tmax and the maximum concentration. Blood/plasma/serum/urine 1036
samples should be processed and stored under conditions that have been shown not to 1037
cause degradation of the analytes. These conditions should be included in the analytical 1038
validation report (see section 7.5). 1039
1040
The sample collection methodology must be specified in the study protocol. 1041
1042
7.4.7 Parameters to be assessed 1043
1044
In bioavailability studies the shape and area under the plasma concentration versus time 1045
curves are mostly used to assess rate (Cmax, tmax) and extent (AUC) of exposure. 1046
Sampling points or periods should be chosen such that the concentration versus time 1047
profile is sufficiently defined to allow calculation of relevant parameters. For single-dose 1048
studies, the following parameters should be measured or calculated: 1049
Working document QAS/14.583/Rev.1
page 37
1050
• area under the plasma/serum/blood concentration–time curve from time zero to 1051
time t (AUC0–t), where t is the last sampling time point with a measurable 1052
concentration of the API in the individual formulation tested. The method of 1053
calculating AUC values should be specified. Non-compartmental methods should 1054
be used for pharmacokinetic calculations in bioequivalence studies; 1055
• Cmax is the maximum or peak concentration observed representing peak exposure 1056
of API (or metabolite) in plasma, serum or whole blood. 1057
1058
Usually AUC0–t and Cmax are considered to be the most relevant parameters for 1059
assessment of bioequivalence. In addition it is recommended that the following 1060
parameters be estimated: 1061
1062
• area under the plasma/serum/blood concentration–time curve from time zero to 1063
time infinity (AUC0-∞) representing total exposure, where AUC0-∞ = AUC0–t + 1064
Clast/Ke; Clast is the last measurable analyte concentration and Ke is the terminal 1065
or elimination rate constant calculated according to an appropriate method; 1066
• tmax is the time after administration of the FPP at which Cmax is observed. 1067
1068
For additional information the elimination parameters can be calculated: 1069
1070
• T1/2 is the plasma (serum, whole blood) half-life. 1071
1072
For steady-state studies conducted with modified-release products, the following 1073
parameters should be calculated: 1074
1075
• AUCτ is AUC over one dosing interval (τ) at steady state; 1076
• Cmax; 1077
• Cmin (Ctau) is concentration at the end of a dosing interval; 1078
• peak trough fluctuation is percentage difference between Cmax and Cmin. 1079
1080
Working document QAS/14.583/Rev.1
page 38
As release mechanisms of pharmaceutical products become more complex, e.g. products 1081
with an immediate-release and a modified-release component, additional parameters such 1082
as partial AUC measures may be necessary to ensure the bioequivalence of two products. 1083
1084
When urine samples are used cumulative urinary recovery (Ae) and maximum urinary 1085
excretion rate are employed instead of AUC and Cmax. 1086
1087
7.4.8 Studies of metabolites 1088
1089
Generally evaluation of bioequivalence will be based on the measured concentrations of 1090
the active parent moiety released from the dosage form rather than the metabolite. The 1091
concentration–time profile of the active parent moiety is more sensitive to changes in 1092
formulation performance than a metabolite, which is more reflective of metabolite 1093
formation, distribution and elimination. 1094
1095
It is important to state a priori in the study protocol which chemical entities (pro-drug, 1096
active moiety or metabolite) will be analysed in the samples. 1097
1098
In rare cases it may be necessary to measure concentrations of a primary active 1099
metabolite rather than those of the active parent moiety if concentrations of the active 1100
parent moiety are too low to allow reliable analytical measurement in blood, plasma or 1101
serum for an adequate length of time, or when the parent compound is unstable in the 1102
biological matrix. 1103
1104
It is important to note that measurement of one analyte, API or metabolite, carries the risk 1105
of making a type-I error (the consumer risk) to remain at the 5% level. However, if more 1106
than one of several analytes is selected retrospectively as the bioequivalence determinant, 1107
then both the consumer and producer risks change (8). The analyte whose data will be 1108
used to assess bioequivalence cannot be changed retrospectively. 1109
1110
Working document QAS/14.583/Rev.1
page 39
When measuring active metabolites, wash-out period and sampling times may need to be 1111
adjusted to enable adequate characterization of the pharmacokinetic profile of the 1112
metabolite. 1113
1114
7.4.9 Measurement of individual enantiomers 1115
1116
A non-stereoselective assay is acceptable for most bioequivalence studies. A 1117
stereospecific assay measuring the individual enantiomers should be employed when the 1118
enantiomers exhibit different pharmacokinetic properties, different pharmacodynamic 1119
properties and the exposure of the enantiomers, as estimated by their AUC ratio or Cmax 1120
ratio, changes when there is a change in the rate of absorption. 1121
1122
7.5 Quantification of active pharmaceutical ingredient 1123
1124
For the measurement of concentrations of the active compound and/or metabolites in 1125
biological matrices, such as serum, plasma, blood, urine and saliva, the applied 1126
bioanalytical method should be well-characterized, fully validated and documented to a 1127
satisfactory standard in order to yield reliable results. 1128
1129
The validation of bioanalytical methods and the analysis of subject samples for clinical 1130
trials in humans should be performed following the principles of good clinical practices 1131
(GCP). 1132
1133
State of the art principles and procedures for bioanalytical method validation and analysis 1134
of study samples should be employed. 1135
1136
The main characteristics of a bioanalytical method that are essential to ensure the 1137
acceptability of the performance and the reliability of analytical results are: selectivity; 1138
lower limit of quantification; the response function and calibration range (calibration 1139
curve performance); accuracy; precision; matrix effects; stability of the analyte(s) in the 1140
biological matrix; stability of the analyte(s) and of the internal standard in the stock and 1141
Working document QAS/14.583/Rev.1
page 40
working solutions, and in extracts under the entire period of storage and processing 1142
conditions. 1143
1144
In general: 1145
1146
• the analytical method should be able to differentiate the analyte(s) of interest, and 1147
if employed, the internal standard (IS) from endogenous components in the matrix 1148
or other components in the sample; 1149
• the lower limit of quantification (LLOQ), being the lowest concentration of 1150
analyte in a sample, should be estimated to prove that the analyte at this 1151
concentration can be quantified reliably, with an acceptable accuracy and 1152
precision; 1153
1154
• the response of the instrument with regard to the concentration of analyte should 1155
be known and should be evaluated over a specified concentration range. The 1156
calibration curve should be prepared in the same matrix as the matrix of the 1157
intended subject samples by spiking the blank matrix with known concentrations 1158
of the analyte. A calibration curve should consist of a blank sample, a zero sample 1159
and 6–8 non-zero samples covering the expected range; 1160
• within-run and between-run accuracy and precision should be assessed on 1161
samples spiked with known amounts of the analyte, the quality control (QC) 1162
samples, at a minimum of three different concentrations; 1163
• matrix effects should be investigated when using mass spectrometric methods; 1164
• stability of the analyte in the stock solution and in the matrix should be proven 1165
covering every step taken during sample preparation and sample analysis, as well 1166
as the storage conditions used; 1167
• when more than one analyte is present in subject samples, it is recommended to 1168
demonstrate the stability of the analytes in the matrix in the presence of the other 1169
analytes; 1170
Working document QAS/14.583/Rev.1
page 41
• in case changes are made to an analytical method that has already been validated, 1171
a full validation may not be necessary, depending on the nature of the changes 1172
implemented. A partial validation may be acceptable; 1173
• a cross-validation is needed in cases where data are obtained from different 1174
methods within and across studies or when data are obtained within a study from 1175
different laboratories, applying the same method; 1176
• analysis of subject samples should be carried out after validation of the analytical 1177
method. Before the start of the analysis of the subject samples the performance of 1178
the bioanalytical method should have been verified; 1179
• reasons for reanalysis, reinjection and reintegration of subject samples should be 1180
predefined in the protocol, study plan or SOP. Re-injection of a full analytical run 1181
or of individual calibration standard samples or QC samples, simply because the 1182
calibration or QCs failed, without any identified analytical cause, is considered 1183
not acceptable For bioequivalence studies, reanalysis, reinjection or reintegration 1184
of subject samples for reasons related to pharmacokinetic fit is normally not 1185
acceptable, as this may affect and bias the outcome of such a study; 1186
• it is recommended to evaluate accuracy of incurred samples by reanalysis of 1187
subject samples in separate runs at different days; 1188
• it is recommended that all the samples from one subject (all periods) should be 1189
analysed in the same analytical run, if possible. 1190
1191
Validation procedures, methodology and acceptance criteria should be specified in the 1192
analytical protocol and/or the SOP. All experiments used to support claims or draw 1193
conclusions about the validity of the method should be described in a report (method 1194
validation report). 1195
1196
The results of subject sample determination should be given in the analytical report 1197
together with calibration and quality control sample results, repeat analyses, reinjections 1198
and reintegrations (if any) and a representative number of sample chromatograms. 1199
1200
1201
Working document QAS/14.583/Rev.1
page 42
7.6 Statistical analysis 1202
1203
The primary concern in bioequivalence assessment is to limit the risk of a false 1204
declaration of equivalence. Statistical analysis of the bioequivalence trial should 1205
demonstrate that a clinically significant difference in bioavailability between the 1206
multisource product and the comparator product is unlikely. The statistical procedures 1207
should be specified in the protocol before the data collection starts. 1208
1209
The statistical method for testing bioequivalence is based on the determination of the 90% 1210
confidence interval around the ratio of the log-transformed population means 1211
(multisource/comparator) for the pharmacokinetic parameters under consideration and by 1212
carrying out two one-sided tests at the 5% level of significance (9). To establish 1213
bioequivalence, the calculated confidence interval should fall within a preset 1214
bioequivalence limit. The procedures should lead to a decision scheme which is 1215
symmetrical with respect to the formulations being compared (i.e. leading to the same 1216
decision whether the multisource formulation is compared to the comparator product or 1217
the comparator product to the multisource formulation). 1218
1219
All concentration-dependent pharmacokinetic parameters (e.g. AUC and Cmax) should be 1220
log-transformed using either common logarithms to the base 10 or natural logarithms. 1221
The choice of common or natural logs should be consistent and should be stated in the 1222
study report. 1223
1224
Logarithmically transformed, concentration-dependent pharmacokinetic parameters 1225
should be analysed using analysis of variance (ANOVA). Normally the ANOVA model 1226
should include formulation, period, sequence and subject factors. 1227
1228
Parametric methods, i.e. those based on normal distribution theory, are recommended for 1229
the analysis of log-transformed bioequivalence measures. 1230
1231
Working document QAS/14.583/Rev.1
page 43
The general approach is to construct a 90% confidence interval for the quantity µT−µR 1232
and to reach a conclusion of pharmacokinetic equivalence if this confidence interval is 1233
within the stated limits. The nature of parametric confidence intervals means that this is 1234
equivalent to carrying out two one-sided tests of the hypothesis at the 5% level of 1235
significance (9, 10). The antilogs of the confidence limits obtained constitute the 90% 1236
confidence interval for the ratio of the geometric means between the multisource and 1237
comparator products. 1238
1239
The same procedure should be used for analysing parameters from steady-state trials or 1240
cumulative urinary recovery, if required. 1241
1242
For tmax descriptive statistics should be given. In those cases where tmax is considered 1243
clinically relevant, median and range of tmax should be compared between test and 1244
comparator to exclude numerical differences with clinical importance. A formal 1245
statistical comparison is rarely necessary. Generally the sample size is not calculated to 1246
have enough statistical power for tmax. However, if tmax is to be subjected to a statistical 1247
analysis this should be based on non-parametric methods and should be applied to 1248
untransformed data. A sufficient number of samples around predicted maximal 1249
concentrations should have been taken to improve the accuracy of the tmax estimate. For 1250
parameters describing the elimination phase (T1/2), only descriptive statistics should be 1251
given. 1252
1253
See section 7.2.3 for information on the handling of extreme data. Exclusion of data for 1254
statistical or pharmacokinetic reasons alone is not acceptable. 1255
1256
7.6.1 Two-stage sequential design 1257
1258
In some situations reliable information concerning the expected variability in the 1259
parameters to be estimated may not be available. In such situations a two-stage sequential 1260
study design can be employed such that an accurate estimate of the variability can be 1261
determined in the first stage of the study. The number of subjects employed in the first 1262
Working document QAS/14.583/Rev.1
page 44
stage is generally based on the most likely intra-subject variance estimate with some 1263
added subjects to protect against drop-outs. The analysis undertaken at the end of the first 1264
stage is treated as an interim analysis. If bioequivalence is proven at this point the study 1265
can be terminated. If bioequivalence is not proven at the end of the first stage, the second 1266
stage is conducted employing an appropriate number of additional subjects as determined, 1267
based on the variance estimates and point estimate calculated from the stage 1 data. At 1268
the end of the second stage the results from both groups combined are used in the final 1269
analysis. In order to employ a two-stage design, adjustments must be made to protect the 1270
overall Type I error rate and maintain it at 5%. In order to do this both the interim and 1271
final analyses must be conducted at adjusted levels of significance, with the confidence 1272
intervals calculated using the adjusted values. 1273
1274
It is recommended that the same alpha for both stages be employed which gives an alpha 1275
of 0.0294 for this case (11), however, the amount of alpha to be spent at the time of the 1276
interim analysis can be set at the study designer’s discretion. For example, the first stage 1277
may be planned as an analysis where no alpha is spent in the interim analysis since the 1278
objective of the interim analysis is to obtain information on the point estimate difference 1279
and variability and where all the alpha is spent in the final analysis with the conventional 1280
90% confidence interval. In this case no test against the acceptance criteria is made 1281
during the interim analysis and bioequivalence cannot be proven at that point. The 1282
proposed statistical plan must be clearly defined in the study protocol, including the 1283
adjusted significance level that is to be employed during each analysis. 1284
1285
A factor for stage should be included in the ANOVA model for the final analysis of the 1286
combined data from the two stages. 1287
1288
This approach can be employed in both cross-over and parallel study designs. 1289
1290
1291
1292
1293
Working document QAS/14.583/Rev.1
page 45
7.7 Acceptance ranges 1294
1295
Area under the curve-ratio 1296
The 90% confidence interval for this measure of relative bioavailability should lie within 1297
a bioequivalence range of 80.00–125.00%. If the API is determined to possess a narrow 1298
therapeutic index (NTI) the bioequivalence acceptance range should be restricted 90.00–1299
111.11%. 1300
1301
Cmax-ratio 1302
For maximal concentration data the acceptance limit of 80.00–125.00% should be applied 1303
to the 90% confidence interval for the mean Cmax-ratio. However, this measure of relative 1304
bioavailability is inherently more variable than, for example, the AUC-ratio, and in 1305
certain cases this variability can make proving bioequivalence challenging. Refer to 1306
section 7.9.3 for information on an approach for proving bioequivalence when the Cmax 1307
parameter intra-subject variability is high. If the API is determined to possess a narrow 1308
therapeutic index (NTI) the bioequivalence acceptance range may need to be restricted to 1309
90.00–111.11%, if appropriate. 1310
1311
tmax-difference 1312
Statistical evaluation of tmax makes sense only if there is a clinically relevant claim for 1313
rapid onset of action or concerns about adverse effects. In such a case comparison of the 1314
median and range data for each product should be undertaken. 1315
1316
For other pharmacokinetic parameters the same considerations as outlined above apply. 1317
1318
7.8 Reporting of results 1319
1320
The report of a bioequivalence study should give the complete documentation of its 1321
protocol, conduct and evaluation complying with GCP rules (4). The relevant ICH 1322
guideline (12) can be used in the preparation of the study report. The responsible 1323
investigator(s) should sign the respective sections of the report. Names and affiliations of 1324
Working document QAS/14.583/Rev.1
page 46
the responsible investigator(s), site of the study and period of its execution should be 1325
stated. 1326
1327
The names and batch numbers of the pharmaceutical products used in the study as well as 1328
the composition(s) of the tests product(s) should be given. Results of in vitro dissolution 1329
tests conducted in pH 1.2, 4.5 and 6.8 media and the QC media, if different, should be 1330
provided. In addition the applicant should submit a signed statement confirming that the 1331
test product is identical to the pharmaceutical product that is submitted for registration. 1332
1333
The bioanalytical validation report should be attached. The bioanalytical report should 1334
include the information recommended in the SRA guidance chosen as a guide for the 1335
bioanalytical portion of a study (see section 7.5). 1336
1337
All results should be presented clearly. All concentrations measured in each subject and 1338
the sampling time should be tabulated for each formulation. Tabulated results showing 1339
API concentration analyses according to analytical run (including runs excluded from 1340
further calculations, including all calibration standards and quality control samples from 1341
the respective run) should also be presented. The tabulated results should present the date 1342
of run, subject, study period, product administered (multisource or comparator) and time 1343
elapsed between FPP administration and blood sampling in a clear format. The procedure 1344
for calculating the parameters used (e.g. AUC) from the raw data should be stated. Any 1345
deletion of data should be documented and justified. 1346
1347
Individual blood concentration/time curves should be plotted on a linear/linear and 1348
log/linear scale. All individual data and results should be given, including information on 1349
those subjects who dropped out. The drop-outs and/or withdrawn subjects should be 1350
reported and accounted for. All adverse events that occurred during the study should be 1351
reported along with the study physician’s classification of the events. Further, any 1352
treatments employed to address adverse events should be reported. 1353
1354
Working document QAS/14.583/Rev.1
page 47
Results of all measured and calculated pharmacokinetic parameters should be tabulated 1355
for each subject–formulation combination together with descriptive statistics. The 1356
statistical report should be sufficiently detailed to enable the statistical analyses to be 1357
repeated if necessary. If the statistical methods applied deviate from those specified in the 1358
study protocol the reasons for the deviations should be stated. 1359
1360
7.9 Special considerations 1361
1362
7.9.1 Fixed-dose combination products 1363
1364
If the bioequivalence of FDC products is assessed by in vivo studies the study design 1365
should follow the same general principles as described in previous sections. The 1366
multisource FDC product should be compared with the pharmaceutically-equivalent 1367
comparator FDC product. In certain cases (e.g. when no comparator FDC product is 1368
available on the market) separate products administered in free combination can be used 1369
as a comparator (3). Sampling times should be chosen to enable the pharmacokinetic 1370
parameters of all APIs to be adequately assessed. The bioanalytical method should be 1371
validated with respect to all analytes measured in the presence of the other analytes. 1372
Statistical analyses should be performed with pharmacokinetic data collected on all active 1373
ingredients; the 90% confidence intervals of test/comparator ratio of all active ingredients 1374
should be within acceptance limits. 1375
1376
7.9.2 Clinically important variations in bioavailability 1377
1378
Innovators should make all efforts to provide formulations with good bioavailability 1379
characteristics. If a better formulation is developed over time by the innovator this should 1380
then serve as the comparator product. A new formulation with a bioavailability outside 1381
the acceptance range for an existing pharmaceutical product is not interchangeable by 1382
definition. 1383
1384
1385
Working document QAS/14.583/Rev.1
page 48
7.9.3 “Highly variable active pharmaceutical ingredients” 1386
1387
A “highly variable API” has been defined as an API with a within-subject variability of > 1388
30% in terms of the ANOVA-CV (13). Proving the bioequivalence of FPPs containing 1389
“highly variable APIs” can be problematic because the higher the ANOVA-CV, the 1390
wider the 90% confidence interval. Thus large numbers of subjects must be enrolled in 1391
studies involving “highly variable APIs” to achieve adequate statistical power. 1392
1393
Although there is variability in how regulatory authorities deal with the issue of highly 1394
variable APIs, the most rigorous of the current approaches involve the scaling of 1395
bioequivalence acceptance criteria based on the within-subject standard deviation 1396
observed in the relevant parameters for the comparator product (14-16). Of the two most 1397
common assessment parameters Cmax is the parameter that is subject to the highest 1398
variability and hence, the parameter for which a modified approach is most needed. 1399
1400
For highly variable FPPs it is recommended that a three-way partial replicate (where the 1401
comparator product is administered twice) or four-way fully replicated cross-over 1402
bioequivalence study be conducted and reference-scaled average bioequivalence be 1403
employed to widen the acceptance interval for the Cmax parameter, if the within-subject 1404
variability for Cmax following replicate administrations of the comparator product 1405
is >30%. If this is the case, the acceptance criteria for Cmax can be widened to a maximum 1406
of 69.84 – 143.19%. The applicant should justify that the calculated intra-subject 1407
variability is a reliable estimate and that it is not the result of outliers. 1408
1409
The extent of the widening of the acceptance interval for Cmax is defined based upon the 1410
within-subject variability seen in the bioequivalence study using scaled-average-1411
bioequivalence according to [U, L] = exp [±k·sWR], where U is the upper limit of the 1412
acceptance range, L is the lower limit of the acceptance range, k is the regulatory 1413
constant set to 0.760 and sWR is the within-subject standard deviation of the log-1414
transformed values of Cmax of the reference product. The table below gives examples of 1415
Working document QAS/14.583/Rev.1
page 49
how different levels of variability lead to different acceptance limits using this 1416
methodology. 1417
1418
Within-subject CV (%) Lower limit Upper limit
30 80.00 125.00
35 77.23 129.48
40 74.62 134.02
45 72.15 138.59
≥50 69.84 143.19
1419
���%� = ���� − 1
1420
The geometric mean ratio (GMR) for Cmax should lie within the conventional acceptance 1421
range 80.00-125.00%. 1422
1423
The standard bioequivalence acceptance criterion for AUC should be maintained without 1424
scaling. If the within-subject variability for Cmax, following replicate administration of the 1425
comparator, is found to <30%, standard bioequivalence acceptance criteria should be 1426
applied to both AUC and Cmax without scaling. 1427
1428
The approach to be employed should be clearly defined prospectively in the study 1429
protocol. The regulatory authority of the country to which the study data will be 1430
submitted should be consulted prior to study conduct to confirm that the proposed 1431
approach is acceptable for that jurisdiction. 1432
1433
8. PHARMACODYNAMIC STUDIES 1434
1435
Studies in healthy volunteers or patients using pharmacodynamic measurements may be 1436
used for establishing equivalence between two pharmaceutical products when the 1437
pharmacokinetic approach is not feasible. Pharmacodynamic equivalence studies may 1438
become necessary if quantitative analysis of the API and/or metabolite(s) in plasma or 1439
urine cannot be made with sufficient accuracy and sensitivity; however, this is extremely 1440
unlikely given current technology. Furthermore, pharmacodynamic equivalence studies in 1441
humans are required if measurements of API concentrations cannot be used as surrogate 1442
Working document QAS/14.583/Rev.1
page 50
end-points for the demonstration of efficacy and safety of the particular pharmaceutical 1443
product such as pharmaceutical products designed to act locally. However, local 1444
availability studies based on pharmacokinetic studies alone or in combination with in 1445
vitro dissolution studies, are being considered as surrogate end-points for the 1446
demonstration of equivalent biopharmaceutical quality and release at the site of action for 1447
some products acting locally. In addition, bioequivalence studies are also required in 1448
order to demonstrate equivalent systemic exposure for systemic safety purposes. 1449
1450
Pharmacodynamic studies are not recommended for orally-administered, pharmaceutical 1451
products for systemic action when the API is absorbed into the systemic circulation and a 1452
pharmacokinetic approach can be used to assess systemic exposure and establish 1453
bioequivalence. This is because the sensitivity to detect differences between products in 1454
their biopharmaceutical quality, release and absorption is lower with pharmacodynamic 1455
or clinical end-points. As the dose–response curve for pharmacodynamics or clinical end-1456
points is usually flatter than the relationship between dose and PK parameters, it is 1457
essential to ensure the internal validity of the study by showing assay sensitivity, i.e. the 1458
ability to distinguish the response obtained by adjacent doses (two-fold or even four-fold 1459
difference in dose).It is essential to perform the comparison at the dose level where the 1460
dose response is steepest, which may require a previous pilot study for its identification. 1461
Furthermore, variability in pharmacodynamic measures is usually greater than that in 1462
pharmacokinetic measures. In addition, pharmacodynamic measures are often subject to 1463
significant placebo effects, which add to the variability and complicate experimental 1464
design. The result is often that huge numbers of patients would have to be enrolled in 1465
pharmacodynamic studies to achieve adequate statistical power. 1466
1467
If pharmacodynamic studies are to be used they must be performed as rigorously as 1468
bioequivalence studies and the principles of GCP must be followed (4). 1469
The following requirements must be recognized when planning, conducting and assessing 1470
the results of a study intended to demonstrate equivalence by measuring 1471
pharmacodynamic responses: 1472
1473
• the response measured should be a pharmacological or therapeutic 1426 effect which 1474
Working document QAS/14.583/Rev.1
page 51
is relevant to the claims of efficacy and/or safety; 1475
• the methodology must be validated for precision, accuracy, reproducibility and 1476
specificity; 1477
• neither the test product nor the comparator product should produce a maximal 1478
response in the course of the study, since it may be impossible to detect differences 1479
between formulations given in doses which give maximum or near maximum effects. 1480
Investigation of dose–response relationships may be a necessary part of the design; 1481
• the response should be measured quantitatively, preferably under double-blind 1482
conditions, and be recordable by an instrument that produces and records the results of 1483
repeated measurements to provide a record of the pharmacodynamic events, which are 1484
substitutes for measurements of plasma concentrations. Where such measurements are 1485
not possible, recordings on visual analogue scales may be used. Where the data are 1486
limited to qualitative (categorized) measurements appropriate special statistical 1487
analysis will be required; 1488
• participants should be screened prior to the study to exclude non-responders. The 1489
criteria by which responders are distinguished from non-responders must be stated in 1490
the protocol; 1491
• in instances where an important placebo effect can occur, comparison between 1492
pharmaceutical products can only be made by a priori consideration of the potential 1493
placebo effect in the study design. This may be achieved by adding a third phase with 1494
placebo treatment in the design of the study; 1495
• the underlying pathology and natural history of the condition 1455 must be considered 1496
in the study design. There should be knowledge of the reproducibility of baseline 1497
conditions; 1498
• a cross-over design can be used. Where this is not appropriate a parallel group study 1499
design should be chosen. 1500
1501
The selection basis for the multisource and comparator products should be the same as 1502
described in section 7.3. 1503
1504
1505
Working document QAS/14.583/Rev.1
page 52
In studies in which continuous variables can be recorded, the time-course of the intensity 1506
of the action can be described in the same way as in a study in which plasma 1507
concentrations are measured and parameters can be derived which describe the area under 1508
the effect–time curve, the maximum response and the time at which the maximum 1509
response occurred. 1510
1511
The comparison between the test and the comparator product can be performed in two 1512
different ways: 1513
1514
(a) dose-scale analysis or relative potency, which is defined as the ratio of the 1515
potency of the test product to that of the reference product. It is a way of 1516
summarising the relationship between the dose-response curves of the test and 1517
comparator product; 1518
1519
(b) response-scale analysis, which consist of demonstration of equivalence (for at 1520
least two dose levels) on the pharmacodynamic end-point. 1521
1522
For either approach to be acceptable a minimum requirement is that the study has assay 1523
sensitivity. For a study to have assay sensitivity at least two non-zero levels need to be 1524
studied and one dose level needs to be shown to be superior to the other. Therefore, it is 1525
recommended that unless otherwise justified more than one dose of both the test and 1526
reference products are studied. However, it is essential that doses on the steep part of the 1527
dose-response curve are studied. If a dose too low on the dose response curve is chosen 1528
then demonstrating equivalence between two products is not convincing, as this dose 1529
could be sub-therapeutic. Equally if a dose at the top of the dose response curve is 1530
included similar effects will be seen for doses much higher than that studied and hence 1531
demonstrating equivalence at this dose level would also not be convincing. 1532
1533
The results using both approaches should be provided. In both cases the observed 1534
confidence intervals comparing test and reference products should lie within the chosen 1535
equivalence margins to provide convincing evidence of equivalence. Like in 1536
Working document QAS/14.583/Rev.1
page 53
bioequivalence studies, 90% confidence intervals should be calculated for relative 1537
potency, whereas 95% confidence intervals should be calculated for the response-scale 1538
analysis. It should be noted that the acceptance range as applied for bioequivalence 1539
assessment may not be appropriate. For both approaches the chosen equivalence ranges 1540
should be pre-specified and appropriately justified in the protocol. 1541
1542
9. CLINICAL TRIALS 1543
1544
In some instances (see example (e) in section 5.1, “In vivo studies”) plasma concentration 1545
time–profile data may be not suitable for assessing equivalence between two 1546
formulations. Although in some cases pharmacodynamic equivalence studies can be an 1547
appropriate tool for establishing equivalence, in others this type of study cannot be 1548
performed because of a lack of meaningful pharmacodynamic parameters that can be 1549
measured; a comparative clinical trial then has to be performed to demonstrate 1550
equivalence between two formulations. However, it is preferable to assess equivalence by 1551
performing a pharmacokinetic equivalence study rather than a clinical trial that is less 1552
sensitive and would require a huge number of subjects to achieve adequate statistical 1553
power. For example, it has been calculated that 8600 patients would be required to give 1554
adequate statistical power to detect a 20% improvement in response to the study API 1555
compared with placebo (18). Similarly it was calculated that 2600 myocardial infarct 1556
patients would be required to show a 16% reduction in risk. A comparison of two 1557
formulations of the same API based on such end-points would require even greater 1558
numbers of subjects (19). 1559
1560
If a clinical equivalence study is considered as being undertaken to prove equivalence, 1561
the same statistical principles apply as for the bioequivalence studies, although a 95% 1562
confidence interval might be necessary for pharmacodynamic and clinical end-points in 1563
contrast to the 90% confidence level employed conventionally for pharmacokinetic 1564
studies. The number of patients to be included in the study will depend on the variability 1565
of the target parameters and the acceptance range and is usually much higher than the 1566
number of subjects needed in bioequivalence studies. 1567
Working document QAS/14.583/Rev.1
page 54
1568
The methodology for establishing equivalence between pharmaceutical products by 1569
means of a clinical trial in patients with a therapeutic end-point has not yet evolved as 1570
extensively as for bioequivalence studies. However, some important items that need to be 1571
defined in the protocol can be identified as follows: 1572
1573
• the target parameters that usually represent relevant clinical end-points from 1574
which the onset, if applicable and relevant, and intensity of the response are to be 1575
derived; 1576
• the size of the acceptance range has to be defined case by case, taking into 1577
consideration the specific clinical conditions. These include, among others, the 1578
natural course of the disease, the efficacy of available treatments and the chosen 1579
target parameter. In contrast to bioequivalence studies (where a conventional 1580
acceptance range is applied) the size of the acceptance range in clinical trials 1581
should be set individually according to the therapeutic class and indication(s); 1582
• the presently used statistical method is the confidence interval approach; 1583
• the confidence intervals can be derived from either parametric or nonparametric 1584
methods; 1585
• where appropriate a placebo leg should be included in the design; 1586
• in some cases it is relevant to include safety end-points in the final comparative 1587
assessments. 1588
1589
The selection basis for the multisource and comparator products should be the same as 1590
described in section 7.3. 1591
1592
10. IN VITRO EQUIVALENCE TESTING 1593
1594
Over the past three decades dissolution testing has evolved into a powerful tool for 1595
characterizing the quality of oral pharmaceutical products. The dissolution test, at first 1596
exclusively a quality control test, is now emerging as a surrogate equivalence test for 1597
certain categories of orally-administered, pharmaceutical products. For these products 1598
Working document QAS/14.583/Rev.1
page 55
(typically solid oral dosage forms containing APIs with suitable properties) similarity in 1599
in vitro dissolution profile, in addition to excipient comparisons and a risk benefit 1600
analysis, can be used to document equivalence of a multisource product with a 1601
comparator product. 1602
1603
It should be noted that, although the dissolution tests recommended in The International 1604
Pharmacopoeia (Ph.Int.) (20) for quality control have been designed to be compatible 1605
with the biowaiver dissolution tests, they do not fulfill all the requirements for evaluating 1606
equivalence of multisource products with comparator products. Dissolution tests for 1607
quality control purposes do not generally correspond to the test conditions required for 1608
evaluating equivalence of multisource products and should not be applied for this purpose. 1609
1610
10.1 In vitro equivalence testing in the context of the Biopharmaceutics 1611
Classification System 1612
1613
10.1.1 Biopharmaceutics Classification System 1614
1615
The Biopharmaceutics Classification System (BCS) is based on aqueous solubility and 1616
intestinal permeability of the API. It classifies the API into one of four classes: 1617
1618
— Class 1: high solubility, high permeability 1619
— Class 2: low solubility, high permeability 1620
— Class 3: high solubility, low permeability 1621
— Class 4: low solubility, low permeability. 1622
1623
Combining the dissolution results and a critical examination of the excipients of the 1624
pharmaceutical product with these two properties of the API takes the four major factors 1625
that govern the rate and extent of API absorption from immediate-release, solid dosage 1626
forms into account (21). On the basis of their dissolution properties, immediate-release 1627
dosage forms can be categorized as having “very rapid”, “rapid”, or “not rapid” 1628
dissolution characteristics. 1629
Working document QAS/14.583/Rev.1
page 56
1630
On the basis of solubility and permeability of the API, excipient nature, excipient content 1631
and dissolution characteristics of the dosage form, the BCS approach provides an 1632
opportunity to waive in vivo bioequivalence testing for certain categories of immediate-1633
release FPPs. Oral FPPS containing an API possessing a narrow therapeutic index are not 1634
eligible for a so-called “biowaiver” based on the BCS approach.. 1635
1636
10.1.1.1 High solubility 1637
1638
An API is considered highly soluble when the highest dosage strength or highest single 1639
therapeutic dose as determined by the relevant regulatory authority, typically defined by 1640
the labelling for the innovator product, is soluble in 250 ml or less of aqueous media over 1641
the pH range of 1.2–6.8. The pH-solubility profile of the API should be determined at 37 1642
± 1 °C in aqueous media. A minimum of three replicate determinations of solubility at 1643
each pH condition is recommended. 1644
1645
[Note from the Secretariat: 1646
1647
Major discussion took place on whether or not to use highest dosage strength or highest 1648
single therapeutic dose, or both. A number of regulatory authorities favour the highest 1649
single therapeutic dose approach; however, some favour the highest dosage strength. In 1650
view of this discussion during the informal consultation there was a proposal to have 1651
both options in the text. Your feedback is sought. The approach used will have 1652
repercussions on the biowaiver guidance document, which is based on the essential 1653
medicines list.] 1654
1655
10.1.1.2 High permeability 1656
1657
An API is considered highly permeable when the extent of absorption in humans is 85% 1658
or more based on a mass balance determination or in comparison with an intravenous 1659
comparator dose. Ideally the mass balance study /i.v. comparison would be conducted at 1660
Working document QAS/14.583/Rev.1
page 57
the same dose as that used for the solubility classification. If this is not possible, dose 1661
linearity of pharmacokinetics should be used to justify the use of other doses. 1662
1663
An acceptable alternative test method for permeability determination of the 1664
API could be: 1665
1666
(i) in vivo intestinal perfusion in humans. 1667
1668
When this method is used for permeation studies, suitability of the methodology should 1669
be demonstrated, including determination of permeability relative to that of a reference 1670
compound whose fraction of dose absorbed has been documented to be at least 85%, as 1671
well as use of a negative control. 1672
1673
Supportive data can be provided by the following additional test methods: 1674
1675
(ii) in vivo or in situ intestinal perfusion using animal models; 1676
1677
(iii) in vitro permeation across a monolayer of cultured epithelial cells (e.g. Caco-2) 1678
using a method validated using APIs with known permeabilities, although data from 1679
neither method (ii) nor (iii) would be considered acceptable on a stand-alone basis. In 1680
these experiments high permeability is assessed with respect to the high permeability of a 1681
series of reference compounds with documented permeabilities and fraction absorbed 1682
values, including some for which fraction of dose absorbed is at least 85% (22). 1683
1684
1685
10.1.2 Determination of dissolution characteristics of multisource products in 1686
consideration of a biowaiver based on the Biopharmaceutics Classification System 1687
1688
For exemption from an in vivo bioequivalence study, an immediate-release, multisource 1689
product should exhibit very rapid or rapid in vitro dissolution characteristics (see below), 1690
Working document QAS/14.583/Rev.1
page 58
depending on the BCS properties of the API. In vitro data should also demonstrate the 1691
similarity of dissolution profiles between the test and comparator products. 1692
1693
10.1.2.1 Very rapidly dissolving 1694
1695
A multisource product is considered to be very rapidly dissolving when no less than 85% 1696
of the labelled amount of the API dissolves in 15 minutes at 37 ± 1oC using a paddle 1697
apparatus at 75 rpm or a basket apparatus at 100 rpm in a volume of 900 mL or less in 1698
each of the following media: 1699
1700
‒ pH 1.2 HCl solution or buffer; 1701
‒ a pH 4.5 acetate buffer; 1702
‒ a pH 6.8 phosphate buffer. 1703
1704
Pharmacopoeial buffers (e.g. Ph.Int.) are recommended for use at these three pH values. 1705
Surfactants should not be used in the dissolution media. Enzymes (pepsin at pH 1.2 and 1706
pancreatin at pH 6.8) may be used if the pharmaceutical product contains gelatin (e.g. 1707
capsules or caplets), due to the possibility of cross linking. 1708
1709
(See also section 10.2, dissolution profile comparison.) 1710
1711
10.1.2.2 Rapidly dissolving 1712
1713
A multisource product is considered to be rapidly dissolving when no less than 85% of 1714
the labelled amount of the API dissolves in 30 minutes at 37 ± 1oC using a paddle 1715
apparatus at 75 rpm or a basket apparatus at 100 rpm in a volume of 900 mL or less in 1716
each of the following media: 1717
1718
‒ pH 1.2 HCl solution or buffer; 1719
‒ pH 4.5 acetate buffer; 1720
‒ pH 6.8 phosphate buffer. 1721
Working document QAS/14.583/Rev.1
page 59
1722
Surfactants should not be employed. 1723
1724
10.2 Qualification for a biowaiver based on the Biopharmaceutics Classification 1725
System 1726
1727
A biowaiver based on the BCS considers: 1728
1729
(a) the solubility and permeability of the API (see section 10.1); 1730
1731
(b) the similarity of the dissolution profiles of the multisource and comparator 1732
products in pH 1.2, 4.5 and 6.8 media (see below); 1733
1734
(c) the excipients used in the formulation (see below); 1735
1736
(d) the risks of an incorrect biowaiver decision in terms of the therapeutic index of, 1737
and clinical indications for, the API (see section 5.1 for cases where an in vivo study 1738
would be required to demonstrate bioequivalence). Only when there is an acceptable 1739
benefit–risk balance in terms of public health and risk to the individual patient should 1740
bioequivalence testing be waived and the in vitro methods described in this section 1741
applied as a test of product equivalence. 1742
1743
Risk reduction and assessment of excipients 1744
1745
The risk of reaching an incorrect decision that the multisource product is equivalent to the 1746
comparator product can be reduced by correct classification of the API and by following 1747
the recommendations for dissolution testing and comparison of the dissolution profiles. In 1748
all cases it should be further demonstrated that the excipients included in the formulation 1749
of the multisource product are well established for use in products containing that API 1750
and that the excipients used will not lead to differences between the comparator and 1751
multisource product with respect to processes affecting absorption (e.g. by effects on 1752
Working document QAS/14.583/Rev.1
page 60
GI motility or interactions with transport processes), or which might lead to interactions 1753
that alter the pharmacokinetics of the API. 1754
1755
In all cases well-established excipients in usual amounts should be employed in 1756
multisource products. Excipients that might affect the bioavailability of the API, e.g. 1757
mannitol, sorbitol or surfactants, should be identified and an assessment of their impact 1758
provided. These critical excipients should not differ qualitatively and must be 1759
quantitatively similar between the test product and comparator product. 1760
1761
For biowaivers for products containing Class 1 APIs, there is some flexibility in the 1762
excipients employed with the exception of critical excipients as discussed above. It is 1763
recommended that the excipients employed be present in the comparator product or be 1764
present in other products which contain the same API as the multisource product and 1765
which have marketing authorizations in ICH-associated countries. 1766
1767
For biowaivers for products containing Class 3 APIs, all excipients in the proposed 1768
product formulation should be qualitatively the same and quantitatively similar to that of 1769
the comparator product, as defined by the WHO quality limits on allowable quantitative 1770
changes in excipients for a variation (23). 1771
1772
As a general rule the closer the composition of the multisource product to that of the 1773
comparator product with regard to excipients, the lower the risk of an inappropriate 1774
decision on equivalence using a biowaiver based on the BCS. 1775
1776
Sub- and supra-bioavailable products 1777
1778
A further consideration is the potential risk to public health and to the individual patient, 1779
should an inappropriate decision with respect to bioequivalence be reached. Essentially 1780
there are two possible negative outcomes. 1781
1782
Working document QAS/14.583/Rev.1
page 61
The first arises when the multisource product is sub-bioavailable. In this case substitution 1783
of the comparator with the multisource product could lead to reduced therapeutic efficacy. 1784
APIs which must reach a certain concentration to be effective (e.g. antibiotics) are most 1785
susceptible to problems of sub-bioavailability. 1786
1787
The second negative outcome arises when the multisource product is supra-bioavailable. 1788
In this case substitution of the comparator with the multisource product could lead to 1789
toxicity. APIs which exhibit toxic effects at concentrations close to the therapeutic range 1790
are most susceptible to problems of supra-bioavailability. For these reasons therapeutic 1791
index is an important consideration in determining whether the biowaiver based on BCS 1792
can be applied or not. 1793
1794
Dissolution profile comparison 1795
1796
Approval of multisource formulations using comparative in vitro dissolution studies 1797
should be based on the generation of comparative dissolution profiles rather than a 1798
single-point dissolution test. For details refer to the Annex. 1799
1800
10.2.1 Dissolution criteria for biowaivers based on the Biopharmaceutics 1801
Classification System according to the properties of active pharmaceutical ingredients 1802
1803
The major application of BCS is to provide criteria for biowaiver of multisource products. 1804
It is recommended that products containing the following BCS classes of APIs be eligible 1805
for a biowaiver: 1806
1807
• BCS Class 1 APIs, if the multisource and comparator product are very rapidly 1808
dissolving or similarly rapidly dissolving; 1809
• BCS Class 3 APIs, if the multisource and comparator product are very rapidly 1810
dissolving. 1811
1812
Working document QAS/14.583/Rev.1
page 62
In summary, biowaivers for solid oral dosage forms based on BCS can be considered 1813
under the following conditions: 1814
1815
1. Dosage forms of APIs that are highly soluble, highly permeable (BCS Class 1) 1816
with acceptable excipient content and favourable risk-benefit analysis, and which are 1817
rapidly dissolving, are eligible for a biowaiver based on the BCS provided: 1818
1819
(i) the dosage form is rapidly dissolving (as defined in section 10.1.2.2) and the 1820
dissolution profile of the multisource product is similar to that of the 1821
comparator product at pH 1.2, pH 4.5 and pH 6.8 buffer using the paddle 1822
method at 75 rpm or the basket method at 100 rpm (as described in section 1823
10.2) and meets the criteria of dissolution profile similarity, f2 ≥50 (or 1824
equivalent statistical criterion); 1825
(ii) if both the comparator and the multisource dosage forms are very rapidly 1826
dissolving (as defined in section 10.1.2.1) the two products are deemed 1827
equivalent and a profile comparison is not necessary. 1828
1829
2. Dosage forms of APIs that are highly soluble and have low permeability (BCS 1830
Class 3) are eligible for biowaivers provided all the criteria (a–d) listed in section 10.2 are 1831
met and the risk–benefit is additionally addressed in terms of extent, site and mechanism 1832
of absorption. 1833
1834
In general the risks of reaching an inappropriate biowaiver decision need to be more 1835
critically evaluated when the extent of absorption is lower (especially if fabs < 50%); 1836
therefore it is essential that the excipients in the proposed product formulation be 1837
scrutinized carefully. In order to minimize the risk of an inappropriate decision, 1838
excipients in the proposed product formulation should be qualitatively the same and 1839
quantitatively similar to that of the comparator. 1840
1841
If it is deemed that the risk of reaching an inappropriate biowaiver decision and its 1842
associated risks to public health and for individual patients is acceptable, the multisource 1843
Working document QAS/14.583/Rev.1
page 63
product is eligible for a biowaiver based on BCS when both the comparator and the 1844
multisource dosage forms are very rapidly dissolving (85% dissolution in 15 minutes as 1845
described in section 10.1.2.1). 1846
1847
10.3 In vitro equivalence testing based on dose-proportionality of formulations 1848
1849
Under certain conditions approval of different strengths of a multisource product can be 1850
considered on the basis of dissolution profiles if the formulations have proportionally 1851
similar compositions. 1852
1853
10.3.1 Proportional formulations 1854
1855
For the purpose of this guidance proportional formulations can be defined in two ways, 1856
based on the strength of dosage forms. 1857
1858
(i) All active and inactive ingredients are exactly in the same proportions in the 1859
different strengths (e.g. a tablet of 50 mg strength has all the active and inactive 1860
ingredients exactly half that of a tablet of 100 mg strength and twice that of a 1861
tablet of 25 mg strength). 1862
1863
(ii) For an FPP, where the amount of the API in the dosage form is relatively low (up 1864
to 10 mg per dosage unit or not more than 5% of the weight of the dosage form), 1865
the total weight of the dosage form remains similar for all strengths. 1866
1867
A waiver is considered: 1868
1869
• if the amounts of the different excipients or capsule contents are the same for the 1870
concerned strengths and only the amount of the API has changed; 1871
• if the amount of filler is changed to account for the change in amount of API: the 1872
amounts of other core excipients or capsule content should be the same for the 1873
concerned strengths. 1874
Working document QAS/14.583/Rev.1
page 64
1875
10.3.2 Qualification for biowaivers based on dose-proportionality of formulations 1876
1877
10.3.2.1 Immediate-release tablets 1878
1879
A biowaiver based on dose-proportionality of formulations for a series of strengths of a 1880
multisource product, when the pharmaceutical products are manufactured with the same 1881
manufacturing process may be granted when: 1882
1883
(i) an in vivo equivalence study has been performed on at least one of the strengths 1884
of the formulation (see section 7.4.1 Selection of strength, usually the highest 1885
strength, unless a lower strength is chosen for reasons of safety or the API is 1886
highly soluble and displays linear pharmacokinetics); 1887
1888
(ii) all strengths are proportionally similar in formulation to that of the strength 1889
studied; 1890
1891
(iii) the dissolution profiles for the different strengths are similar at pH 1.2, 4.5, 6.8 1892
and the QC media, unless justified by the absence of sink conditions. In case the 1893
different strengths of the test product do not show similar dissolution profiles due 1894
to the absence of sink conditions in any of the above media, this should be 1895
substantiated by showing similar dissolution profiles when testing the same dose 1896
per vessel (e.g. two tablets of 5 mg vs one tablet of 10 mg) or by showing the 1897
same behaviour in the comparator product. 1898
1899
As for the BCS-based biowaiver, if both strengths release 85% or more of the label 1900
amount of the API in 15 minutes, using all three dissolution media as recommended in 1901
section 10.2, the profile comparison with an f2 test is unnecessary. 1902
1903
Working document QAS/14.583/Rev.1
page 65
In the case where an immediate-release dosage form with several strengths deviates from 1904
proportionality, a bracketing approach is possible, so that only two strengths representing 1905
the extremes need to be studied in vivo. 1906
1907
10.3.2.2 Delayed-release tablets and capsules 1908
1909
For delayed-release tablets, for a series of strengths of a multisource product where the 1910
strengths are proportionally similar in formulation to that of the strength studied in an in 1911
vivo equivalence study, a lower strength can be granted a biowaiver if it exhibits similar 1912
dissolution profiles, f2 ≥ 50, in the recommended test condition for delayed-release 1913
product, e.g. dissolution test in acid medium (pH 1.2) for 2 hours followed by dissolution 1914
in pH 6.8. When evaluating proportionality in composition, it is recommended to 1915
consider the proportionality of gastro-resistant coating with respect to the surface area 1916
(not to core weight) to have the same gastro-resistance (mg/cm2). 1917
1918
For delayed-release capsules, where different strengths have been achieved solely by 1919
means of adjusting the number of beads containing the API, similarity in the dissolution 1920
profile of the new (lower) strength to that of the approved strength (f2 > 50) under the 1921
test conditions recommended for delayed-release products (see above) is sufficient for a 1922
biowaiver. 1923
1924
10.3.2.3 Extended-release tablets and capsules 1925
1926
(a) For extended-release tablets, when there is a series of strengths of a multisource 1927
product that are proportionally similar in their active and inactive ingredients and 1928
have the same API-release mechanism, in vivo bioequivalence studies should be 1929
conducted with the lowest and highest strengths; however, an intermediate 1930
strength can be granted a biowaiver if it exhibits similar dissolution profiles, f2 ≥ 1931
50, in three different pH buffers (between pH 1.2 and 7.5) and the QC media by 1932
the recommended test method. 1933
1934
Working document QAS/14.583/Rev.1
page 66
(b) For extended release tablets, with an osmotic pump release mechanism the 1935
dissolution profile comparison (f2 ≥ 50) under one recommended test condition is 1936
sufficient for a biowaiver based on dose-proportionality of formulation. 1937
1938
(c) For extended-release, beaded capsules, where different strengths have been 1939
achieved solely by means of adjusting the number of beads containing the API, 1940
dissolution profile comparison (f2 ≥ 50) under one recommended test condition is 1941
sufficient for a biowaiver based on dose-proportionality of formulation. 1942
1943
10.3.3 Dissolution profile comparison for biowaivers based on dose-proportionality of 1944
formulations 1945
1946
As for biowaivers based on the BCS a model independent mathematical approach (e.g. f2 1947
test) can be used for comparing the dissolution profiles of two products. The dissolution 1948
profile of the two products (reference strength2 and additional strength) should be 1949
measured under the same test conditions. 1950
1951
The dissolution sampling times for both reference strength and additional strength 1952
profiles should be the same, for example: 1953
1954
‒ for immediate-release products 5, 10, 15, 20, 30, 45 and 60 minutes; 1955
‒ for 12-hour extended-release products 1, 2, 4, 6 and 8 hours; 1956
‒ for 24-hour extended-release products 1, 2, 4, 6, 8 and 16 hours. 1957
1958
For the application of the f2 value, please refer to the Annex. 1959
1960
1961
1962
1963
2 The reference strength is the strength of the FPP that was compared to the comparator product in an in
vivo equivalence study.
Working document QAS/14.583/Rev.1
page 67
10.4 In vitro equivalence testing for non-oral dosage forms 1964
1965
In the case of intravenous micellar solutions with the same qualitative and quantitative 1966
composition of the surfactant, but significant changes in other excipients, an in vitro 1967
comparison might avoid the need for in vivo studies if a similar micellar system and API 1968
release from the micelle after dilution of the FPP or API administration into the blood 1969
system is ensured (26). 1970
1971
Locally-applied, locally-acting products in the form of aqueous suspensions containing 1972
the same API(s) in the same molar concentration and essentially the same excipients in 1973
comparable concentrations might be waived from the demonstration of equivalence by 1974
means of local availability, pharmacodynamic or clinical studies if in vitro 1975
characterization is able to ensure a similar crystallographic structure and particle size 1976
distribution as well as any other in vitro test specific for each dosage form, e.g. 1977
dissolution. The methodological details for the techniques mentioned below are not 1978
covered in this guideline. Additional information regarding these techniques should be 1979
sought from guidelines of stringent regulatory authorities or state-of-the-art literature. 1980
1981
(a) Suspensions for nebulization with the same qualitative and quantitative 1982
composition as the comparator product might be waived from in vivo studies if the 1983
particles in the suspensions are shown to have the same crystallographic structure, 1984
and particle size distribution as those from the comparator product, as well as 1985
comparability in any other appropriate in vitro test, e.g. dissolution. In addition, 1986
the nebulized droplets should exhibit a similar aerodynamic particle size 1987
distribution to that of the comparator product. 1988
1989
(b) Suspensions for nebulization with the different qualitative and quantitative 1990
composition might be waived if, in addition to the requirements defined above 1991
under (a), the difference in excipient composition does not alter the nebulizer 1992
efficiency (e.g. by the presence or absence of a different surfactant/preservative) 1993
and the aerodynamic particle size distribution (e.g. altering product hygroscopicity 1994
Working document QAS/14.583/Rev.1
page 68
by the presence of a different amount of salt as isotonic agent). To this end the 1995
appropriate state-of-the-art in vitro test should be conducted to ensure product 1996
equivalence. Any excipient difference should be critically reviewed because 1997
certain excipients that are considered irrelevant in other dosage forms (e.g. 1998
preservative, substance to adjust tonicity or thickening agent) may affect safety 1999
and/or efficacy of the product. 2000
2001
(c) Nasal drops where the API is in suspension with the same qualitative and 2002
quantitative composition as the comparator product might be waived from in vivo 2003
studies if the particles in suspension are shown to have the same crystallographic 2004
structure and similar particle size distribution to that of the comparator product, as 2005
well as comparability in any other appropriate in vitro test, e.g. dissolution. 2006
2007
(d) Nasal drops where the API is in suspension with qualitative or quantitative 2008
differences in excipient composition with respect to the comparator product might 2009
be waived from in vivo studies if, in addition to the requirements defined above 2010
under (c), the difference in excipient composition does not affect efficacy and 2011
safety (e.g. a different preservative may affect the safety profile due to a greater 2012
irritation and a different viscosity or thixotropy may affect the residence time in 2013
the site of action). Therefore any excipient difference should be critically 2014
reviewed. 2015
2016
(e) Nasal sprays in solution with the same qualitative and quantitative composition in 2017
excipients can be waived based on a battery of in vitro tests as defined by SRAs 2018
(17, 29). 2019
2020
(f) Nasal sprays in solution with qualitative and quantitative differences in the 2021
excipient composition might be waived if, in addition to showing similarity in the 2022
battery of in vitro tests referenced under (e), excipients differences are critically 2023
reviewed as described above under (d). 2024
2025
Working document QAS/14.583/Rev.1
page 69
(g) Nasal sprays in suspension with the same qualitative and quantitative composition 2026
in excipients might be waived if, in addition to the battery of in vitro tests 2027
referenced above under (e), the particles in suspension are shown to have the same 2028
crystallographic structure and similar particle size distribution, as well as 2029
comparability in any other appropriate in vitro test, e.g. dissolution. 2030
2031
(h) Nasal sprays in suspension with qualitative and quantitative differences in 2032
excipient composition might be waived if, in addition to the battery of in vitro 2033
tests referenced above under (e) and (g), excipients differences are critically 2034
reviewed as described above under (d). 2035
2036
(i) In case of pressurized metered dose inhalers in solution or suspension, in vivo 2037
studies might be waived if similarity is shown in a battery of in vitro tests as 2038
described in specific guidelines of stringent regulatory authorities (27). 2039
2040
(j) When pharmaceutically-equivalent products are topical prepared gels where the 2041
API is in solution, contain the same API(s) in the same molar concentration and 2042
essentially the same excipients in comparable concentrations, equivalence can be 2043
demonstrated by means of in vitro membrane diffusion studies (28). 2044
2045
(k) Otic and ophthalmic suspensions with the same qualitative and quantitative 2046
composition in excipients might be waived if the particles in suspension are 2047
shown to have the same crystallographic structure and similar particle size 2048
distribution, as well as comparability in any other appropriate in vitro test, e.g. 2049
dissolution. 2050
2051
(l) Locally-acting products acting in the GI tract containing highly soluble APIs (as 2052
defined by the BCS) in immediate release dosage forms might be waived from in 2053
vivo equivalence studies based on the same dissolution requirements applied for 2054
the BCS based biowaiver. 2055
2056
Working document QAS/14.583/Rev.1
page 70
10.5 In vitro equivalence testing for scale-up and post-approval changes 2057
2058
Although these guidelines comment primarily on registration requirements for 2059
multisource pharmaceutical products, it should be noted that under certain conditions, 2060
following permissible formulation or manufacturing changes after FPP approval, in vitro 2061
dissolution testing may also be suitable to confirm similarity of product quality and 2062
performance characteristics. More information on when dissolution testing may be used 2063
to support product variations is provided in WHO guidances on variations in 2064
pharmaceutical products. 2065
2066
[Note from the Secretariat: references will be updated when the document is finalized.] 2067
2068
REFERENCES 2069
2070
1. Marrakesh Agreement Establishing the World Trade Organization. Marrakesh, 2071
General Agreement on Tariffs and Trade, 1994, Annex 1 C, Article 39. 2072
2073
2. HHS/FDA Guidance for industry: bioavailability and bioequivalence studies 2074
for orally administered medicine products – general considerations. Rockville, MD, 2075
Department of Health and Human Services, US Food and Drug Administration. 2076
2003 (http://www.fda.gov/cder/guidance/index.htm). 2077
2078
3. Guidelines for registration of fixed-dose combination medicinal products. In: 2079
WHO Expert Committee on Specifications for Pharmaceutical Preparations. 2080
Thirty-ninth report. Geneva, World Health Organization, 2005 (WHO Technical 2081
Report Series, No. 929), Annex 5:94–142. 2082
2083
4. Guidelines for good clinical practice for trials on pharmaceutical products. 2084
In: The use of essential drugs. Sixth report of the WHO Expert Committee. 2085
Geneva, World Health Organization, 1995 (WHO Technical Report 2086
Series, No. 850):97–137. 2087
Working document QAS/14.583/Rev.1
page 71
2088
5. Guidelines for organizations performing in vivo bioequivalence studies. In: 2089
WHO Expert Committee on Specifications for Pharmaceutical Preparations. 2090
Fortieth report. Geneva, World Health Organization (WHO Technical Report 2091
Series, No. 937), 2006, Annex 9. 2092
2093
6. Julious SA. Sample sizes for clinical trials with normal data. Statistics in Medicine, 2094
2004, 23(12): 1921-1986. 2095
2096
7. Revision/update of the guidance on the selection of comparator pharmaceutical 2097
products for equivalence assessment of interchangeable multisource (generic) 2098
products. (In preparation for WHO Technical Report Series.) 2099
2100
8. Midha KK, Rawson MJ, Hubbard JW. Commentary: The Role of Metabolites 2101
in Bioequivalence. Pharmaceutical Research, 2004, 21:1331–1344. 2102
2103
9. Schuirmann DJ. A comparison of the two one-sided tests procedure and the 2104
power approach for assessing the equivalence of average bioavailability. 2105
Journal of Pharmacokinetics and Biopharmaceutics, 1987, 15:657–680. 2106
2107
10. Westlake WJ. Bioavailability and bioequivalence of pharmaceutical formulations. 2108
In: Peace KE, ed. Biopharmaceutical statistics for drug development. 2109
New York, Marcel Dekker, Inc., 1988: 329–352. 2110
2111
11. Pocock SJ. Group sequential methods in the design and analysis of clinical trials. 2112
Biometrika, 1977, 64(2): 191-199. 2113
2114
12. ICH E3, Structure and content of clinical study reports. Geneva, International 2115
Conference on Harmonisation (ICH) Secretariat: IFPMA, 1995. 2116
2117
2118
Working document QAS/14.583/Rev.1
page 72
13. Blume HH, Midha KK. Bio-International 92, Conference on bioavailability, 2119
bioequivalence and pharmacokinetic studies. Journal of Pharmaceutical 2120
Sciences, 1993, 82:1186–1189. 2121
2122
14. Tothfalusi L, et al. Evaluation of bioequivalence of highly variable drugs. 2123
Pharmaceutical Research, 2001, 18:728–733. 2124
2125
15 Tothfalusi L, Endrenyi L, Midha KK. Scaling of wider bioequivalence limits for 2126
highly variable drugs and for the special case of Cmax. International 2127
Journal of Clinical Pharmacology and Therapeutics, 2003, 41:217–225. 2128
2129
16. Tothfalusi L, Endrenyi L. Limits for scaled average bioequivalence of highly 2130
variable drugs and drug products. Pharmaceutical Research, 2003, 20:382– 2131
2132
17. Guideline on the investigation of bioequivalence, London, Committee for 2133
Medicinal Products for Human Use (CHMP), The European Medicines Agency, 2134
2010 (http://www.ema.europa.eu). 2135
2136
18. Yusif et al. Journal of the American Medical Association, 1988, 260: 2259–2263. 2137
2138
19. The Studies of Left Ventricular Dysfunction (SOLVD) Investigators. Effect of 2139
enalapril on survival in patients with reduced left ventricular ejection fractions and 2140
congestive heart failure. New England Journal of Medicine, 1991, 325:293–302. 2141
2142
20. The International Pharmacopoeia. Geneva, World Health Organization 2143
(www.who.int/medicines/publications/pharmacopoeia/). 2144
2145
21. Amidon GL, et al. A theoretical basis for a biopharmaceutic drug classification: 2146
The correlation of in vitro drug product dissolution and in vivo bioavailability. 2147
Pharmaceutical Research, 1995, 12:413–420. 2148
2149
Working document QAS/14.583/Rev.1
page 73
22. Yu LX, et al. Biopharmaceutics Classification System: The scientific basis for 2150
biowaiver extensions. Pharmaceutical Research, 2002, 19:921–925. 2151
2152
23. WHO guidelines on variations to a prequalified product. In: WHO Expert 2153
Committee on Specifications for Pharmaceutical Preparations. Forty-seventh 2154
report. Geneva, World Health Organization, 2013 (WHO Technical Report Series, 2155
No. 981): 154. 2156
2157
24. Moore JW, Flanner HH. Mathematical comparison of curves with an emphasis 2158
on in vitro dissolution profiles. Pharmaceutical Technology, 1996, 20:64–74. 2159
2160
25. Shah VP, et al. In vitro dissolution profile comparison – statistics and analysis 2161
of the similarity factor, f2. Pharmaceutical Research, 1998, 15:889–896. 2162
2163
26. HHS/FDA Draft Guidance for Industry, Bioavailability and Bioequivalence Studies 2164
for Nasal Aerosols and Nasal Sprays for Local Action U.S. Department of Health 2165
and Human Services, Food and Drug Administration, Center for Drug Evaluation 2166
and Research (CDER) April 2003. 2167
2168
27. Guideline on the requirements for clinical documentation for orally inhaled 2169
products (OIP) including the requirements for demonstration of therapeutic 2170
equivalence between two inhaled products for use in the treatment of Asthma and 2171
Chronic Obstructive Pulmonary Disease (COPD). (CPMP/EWP/4151/00 rev 1) 2172
London, Committee for Medicinal Products for Human Use (CHMP), The 2173
European Medicines Agency 2009. (http://www.ema.europa.eu). 2174
2175
28. HHS/FDA Guidance for Industry, Nonsterile Semisolid Dosage Forms 2176
Scale-Up and Postapproval Changes: Chemistry, Manufacturing, and Controls; 2177
In Vitro Release Testing and In Vivo Bioequivalence Documentation. U.S. 2178
Department of Health and Human Services, Food and Drug Administration, Center 2179
for Drug Evaluation and Research (CDER) May 1997. 2180
Working document QAS/14.583/Rev.1
page 74
2181
29. European Medicines Agency – Guideline on product specific guidance on 2182
demonstration of bioequivalence. 2183
2184
30. ICHE6. Good Clinical Practice: Consolidated Guidance. 2185
2186
31. WHO: Good clinical laboratory practice (GCLP). 2187
2188
32. Reflection paper on guidance for laboratories that perform the analysis or 2189
evaluation of clinical trial samples; EMA/INS/GCP/532137/2010. 2190
2191
33. Fares HM, Zats JL. January 1995. Measurement of drug release from topical gels 2192
using two types of apparatus. Pharmaceutical Technology, 52-58. 2193
2194
34. WHO general guidance on variations to multisource pharmaceutical products 2195
(unpublished working document QAS/14.575). 2196
2197
35. Guidance on variations to a prequalified product dossier. Annex 6, WHO 2198
Technical Report Series; No. 943, 2007 2199
2200
2201
Working document QAS/14.583/Rev.1
page 75
ANNEX 2202
RECOMMENDATIONS FOR CONDUCTING AND ASSESSING 2203
COMPARATIVE DISSOLUTION PROFILES 2204
2205
2206
The dissolution measurements of the two FPPs (e.g. test and comparator or two different 2207
strengths) should be made under the same test conditions. A minimum of three time-2208
points (zero excluded) should be included, the time-points for both reference (comparator) 2209
and test product being the same. The sampling intervals should be short for a 2210
scientifically sound comparison of the profiles (e.g. 5, 10, 15, 20, 30, 45 (60, 90, 120) 2211
minutes). The 15-minute time-point is critical to determine whether a product is very 2212
rapidly dissolving and to determine whether f2 must be calculated. For extended release 2213
FPPs, the time-points should be set to cover the entire duration of expected release, e.g. 1, 2214
2, 3, 5 and 8 hours for a 12-hour release and additional test intervals for longer duration 2215
of release. 2216
2217
Studies should be performed in at least three media covering the physiological range, 2218
including pH 1.2 hydrochloric acid, pH 4.5 buffer and pH 6.8 buffer. International 2219
Pharmacopoeia buffers are recommended; other pharmacopoeial buffers with the same 2220
pH and buffer capacity are also accepted. Water may be considered as an additional 2221
medium, especially when the API is unstable in the buffered media to the extent that the 2222
data are unusable. 2223
2224
If both the test and reference (comparator) products show more than 85% dissolution in 2225
15 minutes, the profiles are considered similar (no calculations required). Otherwise: 2226
2227
• similarity of the resulting comparative dissolution profiles should be calculated 2228
using the following equation that defines a similarity factor (f2): 2229
; 2230
Working document QAS/14.583/Rev.1
page 76
• where Rt and Tt are the mean per cent API dissolved in reference (comparator) 2231
and test product, respectively, at each time-point. An f2 value between 50 and 100 2232
suggests that the two dissolution profiles are similar; 2233
2234
• a maximum of one time-point should be considered after 85% dissolution of the 2235
reference (comparator) product has been reached; 2236
2237
• in the case where 85% dissolution cannot be reached due to poor solubility of the 2238
API, the dissolution should be conducted until an asymptote (plateau) has been 2239
reached; 2240
2241
• at least 12 units should be used for determination of each profile. Mean 2242
dissolution values can be used to estimate the similarity factor, f2. To use mean 2243
data, the percentage coefficient of variation at time points up to 10 minutes should 2244
be not more than 20% and at other time-points should be not more than 10%; 2245
2246
• when delayed-release products (e.g. enteric coated) are being compared, the 2247
recommended conditions are acid medium (pH 1.2) for 2 hours and buffer pH 6.8 2248
medium; 2249
2250
• when comparing extended-release beaded capsules, where different strengths 2251
have been achieved solely by means of adjusting the number of beads containing 2252
the API, one condition (normally the release condition) will suffice; 2253
2254
• surfactants should be avoided in comparative dissolution testing. 2255
2256
A statement that the API is not soluble in any of the media is not sufficient and profiles in 2257
the absence of surfactant should be provided. The rationale for the choice and 2258
concentration of surfactant should be provided. The concentration of the surfactant 2259
should be such that the discriminatory power of the test will not be compromised. 2260
*** 2261