1
ICH M7 Step 1 Document 1
(15 Aug 2012 Draft, new or revised text in blue) 2
3
Assessment and Control of DNA Reactive (Mutagenic) Impurities in Pharmaceuticals to 4
Limit Potential Carcinogenic Risk 5
6
7
1. INTRODUCTION 8
9
The synthesis of drug substances involves the use of reactive chemicals, solvents, catalysts, and 10
other processing aids. As a result of chemical synthesis or subsequent degradation, low levels of 11
impurities reside in all drug substances and associated drug products. While ICH Q3A(R2): 12
Impurities in New Drug Substances and Q3B(R2): Impurities in New Drug Products provide 13
guidance for the qualification and control for the majority of the impurities, limited guidance is 14
provided for those impurities that are DNA reactive. The purpose of this guidance is to provide a 15
practical framework that can be applied for the identification, categorization, qualification, and 16
control of these mutagenic impurities to limit potential carcinogenic risk. This guidance is 17
intended to complement ICH Q3A(R2), Q3B(R2) (Note 1), and ICH M3(R2): Nonclinical Safety 18
Studies for the Conduct of Human Clinical Trials and Marketing Authorizations for 19
Pharmaceuticals. 20
21
This guidance emphasizes considerations of both safety and quality risk management in 22
establishing levels of mutagenic impurities that pose negligible carcinogenic risk. It outlines 23
recommendations for assessment and control of mutagenic impurities that reside or are 24
reasonably expected to reside in final drug substance or product, taking into consideration the 25
intended conditions of human use. 26
27
28
2. SCOPE OF GUIDELINE 29
30
This document is intended to provide guidance for new drug substances and new drug products 31
during clinical development and applications for marketing. It also applies to new marketing 32
applications and amendments for marketed products, in both cases only where: 33
34
• Changes to the drug substance synthesis result in new impurities or higher specified 35
levels of existing impurities; 36
• Changes in the formulation result in new degradants or higher specified levels of existing 37
degradants; 38
• Changes in indication or dosing regimen are made which significantly affect the 39
acceptable cancer risk level. 40
41
The following types of drug substances are not covered in this guideline: 42
biological/biotechnological, peptide, radiopharmaceutical, fermentation products, herbal 43
products, and crude products of animal or plant origin. Exceptions would be when products such 44
as biologicals and peptides are chemically synthesized or modified (e.g., addition of organic 45
chemical linkers, semi-synthetic products). In such cases an assessment of potential 46
2
mutagenicity is warranted for chemicals likely to exist as impurities/degradants in the drug 47
product. 48
49
This guidance does not apply to drug products intended for advanced cancer indications as 50
outlined in ICH S9. Additionally, a thorough assessment for the presence of potential mutagenic 51
impurities is not warranted if the drug substance itself is genotoxic and therefore expected to be 52
associated with an increased cancer risk. If under these conditions a mutagenic impurity is 53
known to be present in the drug substance/drug product, higher limits than defined in this 54
guideline are usually acceptable. Justification for higher levels should be made on a case by case 55
basis. 56
57
Excipients used in existing marketed products and flavoring agents are excluded from this 58
guideline. Application of this guidance to new excipients and leachables associated with drug 59
product packaging is not intended, but the risk assessment principles of this guidance for limiting 60
potential carcinogenic risk can be used if warranted. 61
62
63
3. GENERAL PRINCIPLES 64
65
The focus of this guidance is on DNA reactive substances that have a potential to directly cause 66
DNA damage when present at low levels leading to mutations and therefore, potentially causing 67
cancer. This type of mutagenic carcinogen is usually detected in a bacterial reverse mutation test 68
(Ames test). Other types of genotoxicants that are Ames negative typically have thresholded 69
mechanisms and usually do not pose carcinogenic risk in humans at the level ordinarily present 70
as impurities (Kirkland et al., 2005). Therefore to limit a possible human cancer risk associated 71
with potentially mutagenic impurities, the Ames test is used to assess the mutagenic 72
potential/effect and the need for controls. Structure-based assessments are useful for predicting 73
Ames outcomes based upon the established knowledge base. There are a variety of approaches 74
to conduct this evaluation including a review of the available literature, expert knowledge, and/or 75
computational toxicology assessment. 76
77
A threshold of toxicological concern (TTC) was developed to define a common dose for any 78
unstudied chemical that will not pose a risk of carcinogenicity or other toxic effects (Munro et 79
al., 1999; Kroes and Kozianowski, 2002). For application of a TTC in the assessment of 80
acceptable limits of mutagenic impurities in drug substances, a value of 1.5 μg/day 81
corresponding to a theoretical 10-5 excess lifetime risk of cancer, can be justified. The methods 82
by which the TTC is based upon are generally considered very conservative since they involved 83
a simple linear extrapolation from the dose giving a 50% tumor incidence (TD50) to a 1 in 106 84
incidence, using TD50 data for the most sensitive species and most sensitive site of tumor 85
induction (several “worst case” assumptions) (Munro et al., 1999). Some structural groups were 86
identified to be of such high potency that intakes even below the TTC would theoretically be 87
associated with a potential for a significant carcinogenic risk (Cheeseman et al., 1999; Kroes et 88
al., 2004). This group of high potency mutagenic carcinogens (“cohort of concern”) comprises 89
aflatoxin-like-, N nitroso-, and azoxy compounds. 90
91
3
During clinical development, it is expected that control strategies and approaches will be less 92
developed in earlier phases where patient populations are smaller and overall development 93
experience is limited. This guidance bases acceptable limits of mutagenic impurities on 94
established risk assessment strategies. Acceptable risk during the early development phase is set 95
at a theoretically calculated level of approximately one additional cancer per million. For later 96
stages in development and marketed products when efficacy has been shown, acceptable 97
increased cancer risk is set at a theoretically calculated level of approximately one in one 98
hundred thousand. These risk levels represent a small theoretical increase in risk when 99
compared to human overall lifetime incidence of developing any type of cancer, which is in the 100
range of 1 in 4 (Reference needed). It is noted that established cancer risk assessments are 101
based on lifetime exposures. Less than lifetime exposures both during development and 102
marketing can have higher acceptable limits of impurities and still maintain comparable risk 103
levels. The use of a numerical cancer risk value (1 in 100,000) and its translation into risk-based 104
doses (TTC) is a highly hypothetical concept that should not be regarded as a realistic indication 105
of the actual risk. The TTC provides an estimate of safe exposures to any mutagenic compound 106
if the intake is below the TTC. However, exceeding the TTC is not necessarily associated with 107
an increased cancer risk given the conservative assumptions employed in the derivation of the 108
TTC value. The most likely risk cancer incidence is actually much less than 1 in 100,000 (Kroes 109
et al. 2004) and can even be zero (in case an Ames positive would turn out to be non 110
carcinogenic) but cannot be reasonably estimated in the absence of carcinogenicity data. 111
112
Based on the above considerations, any exposure to an unidentified mutagenic impurity that, in 113
later phases of drug development is identified would not necessarily be associated with an 114
increased cancer risk for patients already exposed to the impurity. If TTC levels were exceeded 115
in such a situation, a risk assessment has to be made in order to decide on a case-by-case basis on 116
any further actions to be taken. 117
118
Where a potential risk has been identified for an impurity, an appropriate control strategy 119
leveraging process understanding and/or analytical controls should be developed to ensure 120
mutagenic impurities are at or below the acceptable cancer risk level. 121
122
There may be cases when a mutagenic impurity is also a metabolite of the drug substance. In 123
such cases, the impurity is considered qualified provided that total exposure as a metabolite is at 124
higher exposures than would be achieved just from the impurity (ICH Q3A/Q3B). 125
126
127
4. CONSIDERATIONS FOR MARKETED PRODUCTS 128
129
While this guidance is not intended to be applied retrospectively to existing marketed products, 130
there are select situations where application to marketed products is warranted. These situations 131
are described below. 132
133
4.1 Post approval changes to the drug substance chemistry, manufacturing, and controls 134
135
Post approval changes involving the drug substance chemistry, manufacturing, and controls 136
(changes to the route of synthesis, reagents, solvents, process conditions etc.) that require an 137
4
amendment to an existing application should be evaluated for potential risk impact associated 138
with mutagenic impurities. Specifically, changes should be evaluated to determine if the change 139
results in any new mutagenic impurities or higher levels of existing mutagenic impurities. 140
Reevaluation of impurities not impacted by the change is not required or expected. For 141
example, when only a portion of the manufacturing process is changed, the assessment of risk 142
from mutagenic impurities should be limited to whether any new mutagenic impurities result 143
from the change, whether any mutagenic impurities formed during the affected step are 144
increased, and whether any known mutagenic impurities from up-stream steps are increased. 145
Regulatory submissions associated with such changes should include a summary of the 146
assessment and if appropriate an updated control strategy. Changes to site of manufacture alone 147
would typically not require a reassessment of mutagenic impurity risk. 148
149
When a new drug substance supplier is proposed, evidence that drug substance produced by this 150
supplier (using same route of synthesis) has been approved for an existing drug product marketed 151
in the assessor’s region is considered to be sufficient evidence of acceptable risk/benefit 152
regarding mutagenic impurities and an assessment per M7 is not required. If this is not the case, 153
then an assessment per M7 is expected. 154
155
4.2 Post approval changes to the drug product 156
157
Changes to drug product (e.g., change in composition, manufacturing process, dosage form) that 158
require an amendment to an existing application should be assessed to determine whether there 159
are any new mutagenic degradants or an increase in existing mutagenic degradants. Regulatory 160
submissions associated with such changes should include a summary of the assessment and if 161
appropriate an updated control strategy. Reevaluation of the drug substance associated with drug 162
products is not required or expected provided there are no changes to the drug substance. 163
164
4.3 Clinical changes to marketed products 165
166
Clinical changes that typically may require a reevaluation of the mutagenic impurity limits 167
include a significant increase in clinical dose, an increase in duration of use (in particular when a 168
mutagenic impurity was controlled above the lifetime TTC for a previous indication that may no 169
longer be appropriate for the longer treatment duration associated with the new indication), or for 170
a change in indication from a serious or life threatening condition where higher limits were 171
justified to an indication for a less serious condition where the existing impurity limits may no 172
longer be appropriate. Clinical changes associated with new routes of administration or 173
expansion into patient populations that include pregnant women and/or pediatrics typically 174
would not require a reevaluation, assuming no changes in daily dose or duration of treatment. 175
176
4.4 Alternative considerations for marketed products 177
178
While this guideline is not intended to be applied retrospectively to existing marketed products, 179
application of this guidance may be warranted if there is specific cause for concern. The 180
existence of impurity structural alerts alone is considered insufficient to trigger follow-up 181
measures and is not a cause for concern. However a specific cause for concern would be new 182
relevant impurity hazard data (classified as Class 1 or 2, Section 6) generated after the overall 183
5
control strategy and specifications for market authorization were established and suggests the 184
potential for increased carcinogenic risk for the product’s intended use conditions. This new 185
relevant impurity hazard data should be derived from high-quality scientific studies consistent 186
with relevant regulatory testing guidelines, with data records or reports readily available. When 187
the applicant becomes aware of this new relevant impurity hazard data, an evaluation should be 188
conducted and if it is concluded by the applicant to affect the acceptable cancer risk/benefit, 189
notification (Section 9) to regulatory authorities with a proposed contemporary control strategy 190
and specification would be warranted. 191
192
193
5. PROCESS AND PRODUCT IMPURITY ASSESSMENT 194
195
As stated in the ICH Q3A/B guidances, actual and potential impurities that are likely to arise 196
during the synthesis, work-up, and storage of a new drug substance and during manufacturing 197
and storage of a new drug product should be summarized. 198
199
The impurity assessment is a two stage process. Firstly, actual impurities that have been 200
identified should be considered for their mutagenic potential. In parallel, an assessment of 201
potential impurities likely to be present in the final drug substance is carried out and a decision 202
taken if further evaluation of their mutagenic potential is required. The steps as applied to 203
synthetic impurities and degradants are described in sections 5.1 and 5.2, respectively. 204
205
5.1 Synthetic Impurities 206
207
Actual impurities include those observed in the drug substance and drug product above the ICH 208
Q3A/Q3B reporting thresholds. Identification of actual impurities is expected when the levels 209
exceed the identification thresholds outlined by ICH Q3A/Q3B. It is acknowledged that some 210
impurities below the ID threshold may also have been identified. 211
212
Potential impurities arising from the synthesis of the drug substance could include starting 213
materials, reagents and intermediates, identified impurities in starting materials and 214
intermediates, and reasonably expected reaction by-products based on knowledge of the 215
chemical reactions and conditions involved. Knowledge of the starting material synthesis, in 216
particular the use of mutagenic reagents is an important factor in understanding the potential 217
impurities in the starting materials, especially when there is a reasonable expectation that such 218
impurities may be carried through the synthesis to the drug substance. 219
220
All impurities (actual and potential), where the structures are known, should be evaluated for 221
mutagenic potential as described in Section 6. 222
223
5.2 Degradants 224
225
Actual degradation products are those observed above the ICH Q3A/Q3B reporting threshold 226
during storage of the drug substance and drug product in the proposed long-term storage 227
conditions and primary packaging. Identification of actual degradation products is expected 228
when the levels exceed the identification thresholds outlined by ICH Q3A/Q3B. It is 229
6
acknowledged that some degradation products below the ID threshold may also have been 230
identified. 231
232
Potential degradants in the drug substance and drug product are those that may be reasonably 233
expected to form during storage. Potential degradants include those that form above the 234
ICHQ3A/B identification threshold during accelerated stability studies (e.g. 40oC/75% relative 235
humidity for 6 months) and confirmatory photo-stability studies as described in ICH Q1B, but 236
are yet to be confirmed in the drug substance or drug product in the primary packaging. 237
238
Knowledge of relevant degradation pathways can be used to help guide decisions on the 239
selection of potential degradation products to be evaluated for mutagenicity e.g. from 240
degradation chemistry principles, relevant stress testing studies, and development stability 241
studies. 242
243
Actual and potential degradants likely to be present in the final drug substance or drug product 244
and where the structure is known should be evaluated for mutagenic potential as described in 245
Section 6. 246
247
248
6. HAZARD ASSESSMENT ELEMENTS 249
250
Hazard assessment involves an initial analysis of actual and potential impurities by conducting 251
database and literature searches for carcinogenicity and (Ames) mutagenicity data in order to 252
classify them as Class 1, 2, or 5 according to Table 1. If no data for such a classification is 253
available, an assessment of structure-activity relationships (SAR) that focuses on Ames 254
mutagenicity predictions should be performed (White paper reference to be added). This could 255
lead to a classification into Class 3, 4, or 5. 256
257
Table 1: Impurities Classification with Respect to Mutagenic and Carcinogenic Potential 258
And Resulting Control Actions (according to Muller et al., 2006 with modifications) 259
260
Class
Definition Proposed action for control
(see Section 8)
1
Known mutagenic carcinogens
Control at or below compound-
specific limit (see Section 7)
2
Known mutagens with
unknown carcinogenic potential
(Ames positive*, no rodent carcinogenicity
data)
Control at or below TTC**
3
Alerting structure, unrelated to the
structure of the drug substance;
no mutagenicity data.
Control at or below TTC** or do
Ames test;
If Ames negative = Class 5
If Ames positive = Class 2
4
Alerting structure, same alert in drug
substance which has been tested and is non-
Treat as non-mutagenic impurity
7
mutagenic
5
No structural alerts, or alerting structure
with sufficient data to demonstrate lack of
mutagenicity
Treat as non-mutagenic impurity
261
*Or other relevant positive mutagenicity data indicative of DNA-reactivity related induction of 262
gene mutations (e.g., positive findings in in vivo gene mutation studies) 263
**Or adjusted limit when justified (e.g., shorter treatment duration, see Section 7). 264
265
It is recommended that two different SAR prediction methods with a potential to complement 266
each other be applied such as an expert rule-based and a statistical-based model. SAR systems 267
incorporating these prediction methods should follow the validation principles set forth by the 268
OECD (OECD, 2007). 269
270
The outcome of any computer system-based analysis should be reviewed with use of expert 271
knowledge in order to assess the relevance of any positive or negative prediction and to elucidate 272
underlying reasons in case of conflicting results. 273
274
The absence of structure-based alerts is sufficient to conclude that the impurity is of no concern, 275
and no further genotoxicity testing is required (Class 5 in Table 1). 276
277
To follow up on a structural alert (Class 3 in Table 1), an Ames mutagenicity test can be applied. 278
An appropriately conducted negative Ames test (Note 2) would overrule any structure-based 279
concern, and no further genotoxicity assessments would be required. These impurities (Class 5 280
in Table 1) should be controlled according to ICH Q3A/Q3B. A positive Ames result would 281
warrant further hazard assessment and/or control measures (Class 2 in Table 1). Alternatively 282
adequate control measures in the case of a positive structural alert alone could be applied in place 283
of Ames testing. 284
285
An impurity with a structural alert that is shared with the drug substance (i.e., same alert in the 286
same position and environment in the impurity and the drug substance, as well as similar 287
molecular weight) can be considered as non-mutagenic (Class 4 in Table 1) if the testing of the 288
drug substance in the Ames test was negative. 289
290
Further hazard assessment of an impurity with a positive Ames test result (Class 2 in Table 1) 291
may be needed for instance, when levels of the impurity cannot be controlled at an appropriate 292
TTC limit. In order to understand the relevance of the Ames assay result under in vivo 293
conditions, it is recommended that the impurity is tested in an in vivo gene mutation assay. The 294
selection of other in vivo genotoxicity assays should be scientifically justified based on 295
knowledge of the mechanism of action of the impurity and its organ site of contact (Note 3). In 296
vivo studies should be designed taking into consideration existing guidance and/or expert 297
recommendations (citations to include ICH S2R, OECD, output of relevant workshops, etc.). 298
Negative results in the appropriate in vivo assay, including an adequate margin of safety (i.e., 299
exposure at the highest No Observed Effect Dose (NOED) in the in vivo study compared to 300
estimated clinical exposure to the impurity at the highest clinical dose) may support either setting 301
8
impurity limits in excess of the TTC-based limits or treating the impurity as an ordinary impurity 302
according to ICH Q3A/B. 303
304
305
7. RISK CHARACTERIZATION 306
307
As a result of hazard assessment described in Section 6, each impurity will be assigned to one of 308
the five classes in Table 1. For impurities belonging into Classes 1, 2, and 3 (Class 3 only if 309
presence of a structural alert is not followed up in an Ames test), the principles of risk 310
characterization used to derive acceptable control limits are described in this section. 311
312
7.1 Generic TTC-based limits 313
314
A TTC-based intake of a mutagenic impurity of 1.5 µg per person per day is considered to be 315
associated with a negligible risk (theoretical excess cancer risk of <1 in 100,000 over a lifetime 316
of exposure) and can in general be used for most pharmaceuticals as a default to derive an 317
acceptable control limit. This generic approach would usually be used for mutagenic impurities 318
present in pharmaceuticals for long-term treatment (> 10 years) and where no carcinogenicity 319
data are available (Classes 2 and 3). However a disproportionally high number of members of 320
some structural classes of mutagens, i.e. aflatoxin-like-, N-nitroso-, and azoxy structures, of 321
which some may occur as impurities in pharmaceuticals, display extremely high carcinogenic 322
potency. Acceptable intake limits for these high-potency carcinogens would likely be 323
significantly lower than the TTC levels defined in this document. While the principles of this 324
guideline can be used, a case-by-case approach using e.g. carcinogenicity data from closely 325
related structures, if available, usually needs to be developed to justify limits for pharmaceutical 326
development and marketed products. 327
328
7.2 Limits based on compound-specific risk assessments 329
330
7.2.1 Mutagenic impurities with positive carcinogenicity data (Class 1 in Table 1) 331
332
Compound-specific risk assessments to derive acceptable limits should be applied instead of the 333
TTC-based intakes where sufficient carcinogenicity data exist. For a known mutagenic 334
carcinogen, a compound-specific acceptable limit can be calculated based on carcinogenic 335
potency and linear extrapolation as a default approach. Alternatively, other established risk 336
assessment practices such as those used by international regulatory bodies may be applied either 337
to calculate acceptable limits for daily intakes or to use already existing limits published by 338
regulatory bodies (Note 4). 339
340
Compound-specific calculations for acceptable limits can be applied case-by-case for impurities 341
which are chemically similar to a known carcinogen compound class (class-specific limits) 342
provided that a rationale for chemical similarity and supporting data can be demonstrated (Note 343
5). 344
345
7.2.2 Mutagenic impurities with evidence for a practical threshold 346
347
9
The existence of mechanisms leading to a dose response that is non-linear or has a practical 348
threshold is increasingly recognized, not only for compounds that interact with non-DNA targets 349
but also for DNA-reactive compounds, whose effects may be modulated by, for example, rapid 350
detoxification before coming into contact with DNA, or by effective repair of induced damage. 351
The regulatory approach to such compounds can be based on the identification of a critical no-352
observed-genotoxic effect level (NOGEL) and use of uncertainty factors when data are available 353
(Note 6). 354
355
The acceptable limit derived from compound-specific risk assessments can be adjusted for 356
shorter term use in the same proportions as the staged TTC approach (see Section 7.3.1 and 357
7.3.2). 358
359
7.3 Adjusted limits in relation to less-than-lifetime (LTL) exposure 360
361
The TTC-based limit of 1.5 µg/day is considered to be protective for a lifetime of daily exposure. 362
To address LTL exposures to mutagenic impurities in pharmaceuticals, an approach is applied in 363
which the acceptable cumulative lifetime dose (1.5 µg x 25,550 days) is uniformly distributed 364
over the total number of exposure days during LTL exposure (Felter et al., 2011). This would 365
allow higher daily intake of mutagenic impurities than would be the case for lifetime exposure 366
and still maintain comparable risk levels for daily and non-daily treatment regimens. 367
368
7.3.1 Clinical development 369
370
Using this concept, adjusted daily intake levels of mutagenic impurities are recommended for 371
limited treatment periods during clinical development of up to 1 month, 1 to 12 months and more 372
than one year up to completion of Phase III clinical trials (Table 2). These adjusted intake levels 373
maintain a 10-6 risk level in early clinical development when benefit has not yet been established 374
and a 10-5 risk level for later stages in development (Note 7). 375
376
Table 2: Acceptable daily intakes for LTL exposure: Clinical development 377
378
Duration
of
treatment
< 1 month
1 - 12
months
>1 year
Daily
intake
[µg/day]
120
20
10
379
Clinical development trials up to 14 days 380
381
An alternative approach to the strict use of a LTL control limit for any mutagenic impurity could 382
be applied for Phase I clinical trials of up to 14 days. Only impurities that are known mutagenic 383
carcinogens (Class 1) and known mutagens of unknown carcinogenic potential (Class 2), as well 384
as impurities in the cohort of concern chemical class, should be controlled to acceptable limits as 385
described in this section. All other impurities would be treated as ordinary impurities. This 386
10
includes impurities which contain structural alerts, which alone would not trigger action for an 387
assessment for this limited Phase I duration. 388
389
7.3.2 Post-marketing products 390
391
Standard risk assessments of known carcinogens operate under the assumption that cancer risk 392
increases as a function of cumulative dose. Thus, cancer risk of a continuous low dose over a 393
lifetime would be equivalent to the cancer risk associated with an identical cumulative exposure 394
averaged over a shorter duration or lifetime average daily dose. This assumption has been 395
advocated by other regulatory agencies (USEPA 2005) and proposed elsewhere (Felter et al. 396
2011). 397
398
For marketed product treatments with cumulative exposures of less than 10 years (continuous or 399
total of intermittent treatments), the acceptable daily intake can be adjusted to at most 10 µg. 400
With a 10-5 cancer risk level, the 10 µg/day limit incorporates extra safety factors of 120 at 1 401
month) to the calculated acceptable daily intake without extra safety factors at 10 years. 402
403
7.4 Control limits for multiple mutagenic impurities 404
405
The TTC value should be applied to each individual impurity. When there are more than three 406
mutagenic impurities specified on the drug substance specification, total mutagenic impurities 407
should be limited as described below: 408
409
• For the lifetime TTC for marketed products, total mutagenic impurities should be 410
controlled to <5 ug/day. 411
• For the less than lifetime TTC for marketed products, total mutagenic impurities should 412
be controlled to <30 ug/day 413
• For clinical development, total mutagenic impurities should be controlled to <3 times the 414
individual limit as described in Table 2. 415
416
Only impurities that are specified on the drug substance specification contribute to the 417
calculation for total. Mutagenic impurities which are controlled through an Option 2, 3 or 4 418
approach (see Section 8) and therefore not specified in the drug substance would not be included 419
in this total. In addition, degradation products would not be included in this total but controlled 420
separately. The above approach is supported by a detailed analysis of the effect of combining 421
multiple impurities that are in similar or different chemical classes and by the conservative 422
assumptions incorporated into the TTC, and the low likelihood of synergistic carcinogenic 423
effects at very low mutagenic impurity levels (Bercu et al., 2008). 424
425
7.5 Exceptions and flexibility in approaches 426
427
• Higher limits may be justified when human exposure to the impurity will be much greater 428
from other sources e.g., food. 429
• Case-by-case exceptions to the use of the appropriate TTC can be justified in cases of severe 430
disease, reduced life expectancy, late onset but chronic disease, or with limited therapeutic 431
alternatives. 432
11
• A rationale to deviate from the above risk scenarios can be made when it is not practical to 433
achieve the recommended limit. 434
435
The above risk approaches are applicable to all routes of administration and no corrections to 436
acceptable limits are generally warranted. Exceptions to consider may include situations where 437
data justifies route-specific concerns that need to be evaluated case-by-case. These approaches 438
are also applicable to all patient populations based upon the conservative nature of the risk 439
approaches being applied. 440
441
442
8. CONTROL 443
444
A control strategy is a planned set of controls, derived from current product and process 445
understanding that assures process performance and product quality (ICH Q10). A control 446
strategy can include, but is not limited to, the following: 447
448
• Controls on material attributes (including raw materials, starting materials, intermediates, 449
reagents, primary packaging materials) 450
• Controls implicit in the design of the manufacturing process 451
• In-process controls (including in-process tests and process parameters) 452
• Controls on drug substance and drug product (e.g., release testing) 453
454
When an impurity has been characterized as mutagenic, it is important to develop a control 455
strategy that assures that the level of this impurity in the drug substance and/or drug product is 456
below the acceptable limit. A thorough knowledge of the chemistry associated with the drug 457
substance manufacturing process, the drug product manufacturing process, along with an 458
understanding of the overall stability of the drug substance and drug product is fundamental to 459
developing the appropriate controls. Developing a strategy to mitigate mutagenic impurities in 460
the drug product is consistent with risk management processes identified in ICH Q9. A control 461
strategy that is based on product and process understanding and utilisation of risk management 462
principles will lead to a combination of process design and control and appropriate product 463
testing, which can also provide an opportunity to shift controls upstream and minimize the need 464
for end-product testing. 465
466
8.1 Control of process related impurities 467
468
There are 4 potential approaches to development of a control strategy for drug substance: 469
470
Option 1 471
472
• Specified impurity on the drug substance specification: Include a test for the impurity in the 473
drug substance specification with an acceptance criterion at or below the acceptable limit 474
using an appropriate analytical procedure. 475
476
Option 2 477
478
12
• Non specified impurity in the drug substance: Include a test for the impurity in the 479
specification for a raw material, starting material or intermediate, or as an in-process control, 480
with an acceptance criterion at or below the acceptable limit using an appropriate analytical 481
procedure. 482
483
Option 3 484
485
• Non specified impurity in the drug substance: Include a test for the impurity in the 486
specification for a raw material, starting material or intermediate, or as an in-process control, 487
with an acceptance criterion above the acceptable limit using an appropriate analytical 488
procedure coupled with demonstrated understanding of fate and purge and associated process 489
controls that assure the level in the drug substance is below the acceptable limit without the 490
need for any additional testing. 491
492
Option 4 493
494
Non specified impurity in the drug substance: Understanding of process parameters and impact 495
on residual impurity levels (including fate and purge knowledge) with sufficient confidence that 496
the level of the impurity in the drug substance will be below the acceptable limit such that no 497
analytical testing is needed for this impurity. 498
499
8.2 Discussion of control approaches 500
501
A control strategy that relies on process controls in lieu of analytical testing (Option 4) can be 502
appropriate if scientific understanding of the chemistry and process parameters that impact levels 503
of mutagenic impurities is understood and the risk of an impurity residing in the final drug 504
substance or drug product above the acceptable limit is determined to be negligible. Elements of 505
a scientific risk assessment/chemistry rationale should include an assessment of various factors 506
that influence the fate and purge of an impurity including chemical reactivity, solubility, 507
volatility, ionizability and any physical process steps designed to remove impurities. This option 508
is especially useful for those impurities that are inherently unstable (e.g. thionyl chloride that 509
reacts rapidly and completely with water) or for those impurities introduced early in the synthesis 510
and are effectively purged. 511
512
For Option 4 approaches where justification based on scientific principles alone is not considered 513
sufficient, as well as for Option 3 approaches, analytical data to support the control approach is 514
expected. In these cases it is important to demonstrate that the fate/purge argument for the 515
impurity is shown to be robust and will consistently assure a negligible probability of an 516
impurity residing in the final drug substance above the acceptable limit. Where the purge factor 517
is based on developmental data, it is important to address the expected scale-dependence or 518
independence. In the case that the small scale model used in the development stage is considered 519
to not represent the commercial scale, confirmation of suitable control in pilot scale and/or initial 520
commercial batches is necessary. The need for data from pilot/commercial batches is influenced 521
by the magnitude of the purge factor calculated from laboratory or pilot scale data, point of entry 522
of the impurity, and knowledge of downstream process purge points. 523
524
13
If Options 3 and 4 cannot be justified, then a test for the impurity on the specification for a raw 525
material, starting material or intermediate, or as an in-process control (Option 2) or drug 526
substance (Option 1) at the acceptable limit should be included. A test for this impurity in the 527
drug substance specification is not needed when the impurity is controlled by an Option 2 528
approach. 529
530
The application of ‘as low as reasonably practicable’ (ALARP) is not necessary if the level of the 531
mutagenic impurity is below acceptable limits. Similarly, it is not necessary to demonstrate that 532
alternate routes of synthesis have been explored. 533
534
In cases where control efforts cannot reduce the level of the mutagenic impurity to below the 535
recommended limits (Section 7) and levels are as low as reasonably practical, higher limits may 536
be justified based on a risk/benefit analysis. 537
538
8.3 Control of degradants 539
540
For a potential degradant that has been characterized as mutagenic, it is important to understand 541
if the degradation pathway is relevant to the drug substance and drug product manufacturing 542
processes and/or their proposed packaging and storage conditions. A well-designed accelerated 543
stability study (e.g., 40 oC/75% relative humidity, 6 months) in the proposed packaging, with 544
appropriate analytical procedures is recommended to determine the relevance of the potential 545
degradation product. Alternatively, a well designed kinetically equivalent shorter term thermal 546
study in the proposed commercial package may be used to determine the relevance of the 547
degradation pathway prior to initiating longer term stability studies. This type of study would be 548
especially useful to understand the relevance of those potential degradants that are based on 549
knowledge of potential degradation pathways but not yet observed in the product. 550
551
Based on the result of these accelerated studies, if it is anticipated that the degradant will form at 552
levels approaching the acceptable limit under the proposed packaging and storage conditions, 553
then efforts to control formation of the degradant is expected. The extent of degradation can 554
often be lowered through formulation development and/or packaging designed to protect from 555
moisture, light, or oxygen. Monitoring for the drug substance or drug product degradant in long 556
term primary stability studies at the proposed storage conditions (in the proposed commercial 557
pack) will generally be expected in these cases. The determination of the need for a specification 558
for the mutagenic degradant will generally depend on the results from these stability studies. 559
560
If it is anticipated that formulation development and packaging design options are unable to 561
control mutagenic degradant levels to less than the recommended limit (Section 7) and levels are 562
as low as reasonably practicable, higher limits can be justified based on a risk/benefit analysis. 563
564
8.4 Lifecycle management 565
566
The quality system elements and management responsibilities described in ICH Q10 are intended 567
to encourage the use of science-based and risk-based approaches at each lifecycle stage, thereby 568
promoting continual improvement across the entire product lifecycle. Product and process 569
14
knowledge should be managed from development through the commercial life of the product up 570
to and including product discontinuation. 571
572
The development and improvement of a drug substance or drug product manufacturing process 573
usually continues over its lifecycle. Manufacturing process performance, including the 574
effectiveness of the control strategy, should be periodically evaluated. Knowledge gained from 575
commercial manufacturing can be used to further improve process understanding and process 576
performance and to adjust the control strategy to ensure appropriate quality. 577
578
Any proposed change to the manufacturing process should be evaluated for the impact on the 579
quality of drug substance and drug product. This evaluation should be based on scientific 580
understanding of the manufacturing process and should determine appropriate testing to analyze 581
the impact of the proposed change. Throughout the lifecycle of the product, it will be important 582
to reassess if “risk-driven” testing is needed when intended or unintended changes outside of the 583
regulatory process description occur in the process. This is especially important when there is no 584
routine monitoring at the acceptable limit (Option 3 or Option 4 control approaches). The 585
appropriate testing to analyze the impact of the proposed change could include, but is not limited 586
to, an assessment of current and potential new impurities and an assessment of the test 587
procedures’ abilities to detect any new impurities. This testing should be performed at an 588
appropriate point in the manufacturing process. 589
590
The use of statistical process monitoring can enhance assurance of continued suitability and 591
capability of processes to provide adequate control on the impurity. 592
593
All changes should be subject to internal change management processes as part of the quality 594
system (ICH Q10). Changes to information filed and approved in a dossier should be reported to 595
regulatory authorities in accordance with regional regulations and guidelines. 596
597
8.5 Considerations for clinical development 598
599
It is recognized that product and process knowledge increases over the course of development 600
and therefore it is expected that data to support control strategies in the clinical development trial 601
phases will be less than at the marketing registration phase. A risk-based approach based on 602
process chemistry fundamentals is encouraged to prioritize analytical efforts on those impurities 603
with the highest likelihood of being present in the drug substance or drug product, typically those 604
impurities that are introduced late in the synthesis. In a situation where analytical data may be 605
needed to support a control approach for a marketing application, that data may not be available 606
in early clinical development but the approach would be acceptable in early development. It is 607
also recognized that commercial formulation design occurs later in clinical development and 608
therefore efforts associated with drug product degradants will be limited in the earlier phases. 609
610
611
9. DOCUMENTATION 612
613
Information relevant to the application of this guidance should be provided at the following 614
stages: 615
15
616
Clinical Development Trial Applications 617
618
• It is expected that the number of structures assessed for mutagenicity, and the collection of 619
analytical data will both increase throughout the clinical development period. 620
• For phase I clinical trials of 14 days or less, a brief high level summary of efforts to mitigate 621
risks of mutagenic impurities should be provided with a focus on Class 1 and 2 impurities as 622
outlined in Section 7. A short description of the origin of the impurity (e.g. “reagent used in 623
step 2”, “final intermediate”, “reaction by-product from step 1”, etc.) should be provided. 624
For other clinical development trials including phase I studies of longer than 14 days, any 625
Class 3 actual and potential impurities should be described along with the corresponding 626
control strategy. When supported by a risk assessment, chemistry arguments may be 627
appropriate instead of analytical data for potential impurities that present a low risk as 628
described in Section 8.4. Where appropriate, a statement of future plans for the control 629
strategy may be of value. 630
631
Common Technical Document (Marketing Application) 632
633
• For all actual and potential process related impurities and degradants, the mutagenic impurity 634
classification and rationale for this classification should be provided. 635
o This would include a description of the outcome of the in silico systems used, any 636
assays conducted, and how supporting literature or databases are used to arrive at the 637
overall conclusion. 638
o When Ames assays were performed on an impurity, the study report should be 639
provided. 640
• Justification for the proposed specification limits and the approach to control should be 641
provided. 642
• A summary of the control strategy in the marketing application should be provided (e.g., Q11 643
example 5b). This information could include the safety threshold, the location and sensitivity 644
of relevant routine monitoring (if applicable), a summary of current knowledge of the purge 645
factor, and identification of factors providing control (e.g., process steps, solubility in wash 646
solutions, etc.). This becomes especially important when routine measurement at the safety 647
threshold is not part of the control approach (e.g., Options 3 and 4). 648
649
650
10. NOTES 651
652
Note 1 653
In both, ICH Q3A (Attachment 3, Decision Tree for Identification and Qualification) and ICH 654
Q3B (Attachment 3, Decision Tree for Identification and Qualification of a Degradation 655
Product), it is proposed to consider the conduct of genotoxicity studies (point mutation, 656
chromosomal aberration) and general toxicity studies (one species, usually 14 to 90 days) if 657
untested impurities are present above the qualification threshold. The ICH M7 guidance 658
recommendations provide a state-of-the-art approach for assessing the genotoxic potential of 659
impurities and ensure that such impurities are controlled to safe levels so that below or above the 660
16
qualification threshold no further qualification for genotoxic potential is required. This includes 661
the initial use of (Q)SAR tools to predict Ames mutagenicity. 662
663
Note 2 664
To assess the mutagenic potential of impurities, a single Ames test can be carried out with a fully 665
adequate protocol according to ICH S2R and OECD 471 guidelines. The assays are expected to 666
be performed in compliance with GLP; however, it is noted that the test article may not be 667
prepared or analyzed in compliance with GLP regulations. Lack of full GLP compliance does 668
not necessarily mean that the data cannot be used to support clinical trials and marketing 669
authorizations. Such deviations should be described in the study report. In some cases, the 670
selection of Ames tester strains may be limited to those proven to be sensitive to an alert. For 671
degradants where it is not feasible to isolate or synthesize or when compound quantity is limited, 672
it may not be possible to achieve the highest test concentrations recommended for an ICH-673
compliant Ames assay according to the current testing guidelines. In this case miniaturized 674
bacterial mutagenicity assays with proven high concordance to the ICH-compliant Ames assay 675
(Reference) may be used to enable testing at high concentrations. In any way, confidence in 676
detection of mutagens requires testing concentrations at levels ≥250 µg/plate (Kenyon et al. 677
2007). 678
679
Note 3 680
Tests to investigate the in vivo relevance of in vitro mutagens (Ames-positives) 681
682
In vivo test
Mechanistic data to justify choice of test
as fit-for-purpose
Transgenic mutation assays
(target organs liver, blood)
• No further justification needed
pig-A assay
• For directly acting mutagens (Ames positive without S9)
or for indirectly acting mutagens (requiring metabolic
activation) with justification about sufficient exposure to
metabolite(s)
Micronucleus test
blood or bone marrow
• For directly acting mutagens (Ames positive without S9)
or for indirectly acting mutagens (requiring metabolic
activation) with justification about sufficient exposure to
metabolite(s)
and
• Known to be clastogenic (in vitro data or class effect)
Rat liver UDS test
• Ames positive with S9 only
• Responsible liver metabolite known
o to be generated in test species used
o to induce bulky adducts
Comet assay
(target organ liver, blood; can
be combined with
micronucleus)
• Justification needed (chemical class specific mode of
action to form alkaline labile sites or single-strand breaks
as preceding DNA damage that can potentially lead to
mutations
Others • With convincing scientific justification
683
17
Note 4 684
Example of linear extrapolation from the TD50 685
686
It is possible to calculate a compound-specific limit based on rodent carcinogenicity potency data 687
such as TD50 values (doses giving a 50% tumor incidence equivalent to a cancer risk probability 688
level of 1:2). Linear extrapolation to a probability of 1 in 100,000 (i.e., the accepted lifetime risk 689
level used) is achieved by simply dividing the TD50 by 50,000. This procedure is similar to that 690
employed for derivation of the TTC. 691
692
Calculation example: Ethylene oxide 693
694
TD50 values for ethylene oxide according to the Cancer Potency Data Base (Reference) are 21.3 695
mg/kg body weight/day (rat) and 63.7 mg/kg body weight/day (mouse). For the calculation of an 696
acceptable control limit, the lower (i.e., more conservative) value of the rat is used. 697
To derive a dose to cause tumors in 1 in 100,000 animals, divide by 50,000: 698
21.3 mg/kg ÷ 50,000 = 0.42 µg/kg 699
To derive a total human daily dose: 700
0.42 µg/kg/day x 60 kg body weight = 25.2 µg/person/day 701
702
Hence, a daily life-long intake of 25 µg ethylene oxide would correspond to a theoretical cancer 703
risk of 10-5 and therefore be an acceptable intake when present as an impurity in a drug 704
substance. 705
706
Alternative methods and published regulatory limits for cancer risk assessment 707
708
As an alternative of using the most conservative TD50 value from rodent carcinogenicity studies 709
irrespective of its relevance to humans, an in–depth toxicological expert assessment of the 710
available carcinogenicity data can be done in order to initially identify the findings (species, 711
organ etc) with highest relevance to human risk assessment as a basis for deriving a reference 712
point for linear extrapolation. Also, in order to better take into account directly the shape of the 713
dose-response curve, a benchmark dose such as a benchmark dose lower confidence limit 10% 714
(BMDL10, an estimate of the lowest dose which is 95% certain to cause no more than a 10% 715
cancer incidence in rodents) may be used instead of TD50 values as a numerical index for 716
carcinogenic potency. Linear extrapolation to a probability of 1 in 100,000 (i.e., the accepted 717
lifetime risk level used) is then achieved by simply dividing the BMDL10 by 10,000. 718
719
Compound-specific acceptable limits can also be derived from published recommended limits 720
from internationally recognized bodies such as WHO or others [References] using the 721
appropriate 10-5 lifetime risk level. In general, a regulatory limit that is applied should be based 722
on the most current and scientifically supported data and/or methodology. 723
724
Note 5 725
A compound-specific calculation of acceptable limits for mutagenic impurities may be applied 726
for mutagenic impurities (without carcinogenicity data) which are structurally similar to a 727
chemically-defined class of known carcinogen. For example, factors that are associated with the 728
carcinogenic potency of alkyl halides have been identified (Brigo and Muller, 2011) and can be 729
18
used to modify the safe acceptable daily intake of monofunctional alkyl halides, a group of alkyl 730
halides commonly used in drug synthesis. Compared to multifunctional alkyl halides the 731
monofunctional compounds are much less potent carcinogens with TD50 values ranging from 732
2.96 to 1810 mg/kg/day (n=16) (Brigo and Muller, 2011). A TD50 value of 12.5 mg/kg/day can 733
thus be used as a still very conservative class-specific potency reference point for calculation of 734
acceptable intakes for monofunctional alkyl halides. This potency level is ten fold lower than the 735
TD50 of 1.25 mg/kg/day corresponding to the default lifetime TTC of 1.5 µg/day and therefore 736
justifies lifetime and less-than-lifetime daily intakes for monofunctional alkyl halides ten times 737
the default ones. 738
739
Note 6 740
Some published data give reliable experimental evidence for (practical) thresholds in the dose 741
response for Ames positive mutagens. This includes examples of thresholds in error-free repair 742
capacity of the mutagenic DNA-ethylating agent ethyl methanesulfonate (EMS) (Mueller et al., 743
2009) or similarly for methylating agents (Wirtz, 2010). Thresholds involving metabolic 744
detoxification processes also appear to exist for vinyl chloride (Bolt, 2005) and 1, 3-butadiene 745
(Koc, 1999). Further, a threshold for oxidative DNA damage associated with the buildup of 746
hemosiderin has been shown for p-chloroaniline hydrochloride (Gehlhaus, 2011). Aside of 747
mechanistic considerations supporting an experimentally observed threshold, it is important that 748
a proper statistical analysis supports this assumption as well (Lutz, 2009). 749
750
Note 7 751
Establishing less-than-lifetime limits for mutagenic impurities in pharmaceuticals has precedent 752
in the establishment of the staged TTC limits for clinical development (Mueller et al., 2006). 753
The calculation of less-than-lifetime acceptable daily intakes (ADI) is predicated on the principle 754
of Haber’s rule, a fundamental concept in toxicology where concentration (C) x time (T) = a 755
constant (k). Therefore, the carcinogenic effect is based on both dose and duration of exposure. 756
757
19
1
ye
ar
1
m
on
th1
da
y
5
ye
ar
s
1
1
10
100
1000
10000
1 10 100 1000
Number of treatment days
D
os
e[
µg
/p
er
so
n/
da
y]
g
iv
en
on
t
re
at
m
en
td
ay
s 38250 µg
1270 µg
100 µg
10 µg
1
da
y
1
m
on
th
1
ye
ar
10
y
ea
rs
70
y
ea
rs
1,5 µg
120 µg
20 µg
10 µg
1,5 µg
SF: 60‐5x
SF: 10‐1x
SF: 300‐10x
SF: 7‐1x
30 365 3650 25500
Calculated dose corresp. to 10‐5 cancer risk
Proposed acceptable dose
SF: “Safety Factor” (difference (max./min.) between
calculated and proposed doses
1
ye
ar
1
m
on
th1
da
y
5
ye
ar
s
1
1
10
100
1000
10000
1 10 100 1000
Number of treatment days
D
os
e[
µg
/p
er
so
n/
da
y]
g
iv
en
on
t
re
at
m
en
td
ay
s 38250 µg
1270 µg
100 µg
10 µg
1
da
y
1
m
on
th
1
ye
ar
10
y
ea
rs
70
y
ea
rs
1,5 µg
120 µg
20 µg
10 µg
1,5 µg
SF: 60‐5x
SF: 10‐1x
SF: 300‐10x
SF: 7‐1x
30 365 3650 25500
Calculated dose corresp. to 10‐5 cancer risk
Proposed acceptable dose
SF: “Safety Factor” (difference (max./min.) between
calculated and proposed doses
758
759
Figure 1: Illustration of calculated daily dose of a mutagenic impurity 760
corresponding to a theoretical 1:100,000 cancer risk as a function of duration of 761
treatment in comparison to the acceptable daily intake levels as recommended in 762
Section 7.3. 763
764
The solid line in Figure 1 represents the linear relationship between the amount of daily intake of 765
a mutagenic impurity corresponding to a 10-5 cancer risk and the number of treatment days. The 766
calculation is based on the TTC level as applied in this guidance for life-long treatment i.e., 1.5 767
µg per person per day using the formula: 768
769
Less-than-lifetime adjusted ADI = 1.5 µg x (365 days x 70 years lifetime = 25,550) 770
Total number of treatment days 771
772
The calculated daily intake levels would thus be 1.5 µg for treatment duration of 70 years, 10 µg 773
for 10 years, 100 µg for 1 year, 1270 µg for 1 month and 38250 µg as a single dose, all resulting 774
in the same cumulative dose and therefore theoretically in the same cancer risk (1:100,000). 775
776
The dashed step-shaped curve represents the actual daily intake levels adjusted to less-than-777
lifetime exposure as recommended in Section 7 of this guidance for products in clinical 778
development. These proposed levels are in general significantly lower than the calculated values 779
thus providing safety factors (SF) that increases with shorter treatment durations. 780
781
20
The proposed accepted daily intakes are also in compliance with a 10-6 cancer risk level if 782
treatment durations are not longer than 6 months* and are therefore applicable in early clinical 783
trials with volunteers/patients where benefit has not yet been established. In this case the safety 784
factors as shown in the upper graph would be reduced by a factor of 10. 785
786
*At 6 months the calculated dose at a 10-6 risk level would be 20 µg which is identical to the recommended accepted 787
dose i.e. there is no extra safety factor; at longer duration the theoretical 10-6 risk level would be exceeded. 788
789
790
11. GLOSSARY 791
792
793
12. REFERENCES 794
795
Bercu, J.P., Hoffman, W.P., Lee, C., Ness, D.K. (2008). Quantitative assessment of cumulative 796
carcinogenic risk for multiple genotoxic impurities in a new drug substance. Regul. Toxicol. 797
Pharmacol. 51, 270-277. 798
799
Brigo, A. and Müller, L. (2011) Development of the Threshold of Toxicological Concern 800
Concept and its Relationship to Duration of Exposure, in Genotoxic Impurities (ed A. Teasdale), 801
John Wiley & Sons, Inc., Hoboken, NJ, USA. doi: 10.1002/9780470929377.ch2 802
803
Cheeseman M.A., Machuga E.J., Bailey A.B. (1999). A tiered approach to threshold of 804
regulation. Food Chem Toxicol 37, 387-412. 805
806
Dobo K.L., Greene N., Cyr M.O., Caron S., Ku W.W. (2006). The application of structure-based 807
assessment to support safety and chemistry diligence to manage genotoxic impurities in active 808
pharmaceutical ingredients during drug development, Reg Tox Pharm 44, 282-293. 809
810
Felter, S. P., Conolly, R. B., Bercu, J. P., Bolger, P. M., Boobis, A. R, Bos, P. M. J., Carthew, 811
P., Doerrer, N. G, Goodman, J. I., Harrouk, W. A, Kirkland, D. J., Lau, S. S., Llewellyn, G. C., 812
Preston, R. J., Schoeny, R., Schnatter, A. R., Tritscher, A., van Velsen, F., Williams, G. M. 813
(2011). A proposed framework for assessing risk from less-than-lifetime exposures to 814
carcinogens. Critical reviews in toxicology 41, 507-44. 815
816
Kenyon, M.O., J.R. Cheung, K.L. Dobo, W.W. Ku (2007). An evaluation of the sensitivity of the 817
Ames assay to discern low-level mutagenic impurities. Regul Toxicol Pharmacol 48, 75-86. 818
819
Kirkland, D., Aardema, M., Henderson, L., Muller, L. (2005). Evaluation of the ability of a 820
battery of three in vitro genotoxicity tests to discriminate rodent carcinogens and non-821
carcinogens I. Sensitivity, specificity and relative predictivity. Mutat. Res. 584 (1–2), 1–256. 822
823
Kroes R., Kozianowski G. (2002). Threshold of toxicological concern (TTC) in food safety 824
assessment. Toxicol Letters 127, 43-46. 825
826
Kroes R., Renwick A.G., Cheeseman M., Kleiner J., Mangelsdorf I., Piersma A., Schilter B., 827
Schlatter J., van Schothorst F., Vos J.G., Würtzen G. (2004). Structure-based threshold of 828
21
toxicological concern (TTC): guidance for application to substances present at low levels in the 829
diet. Food Chem Toxicol 42, 65-83. 830
831
Munro I.C., Kennepohl E., Kroes R. (1999). A procedure for the safety evaluation of flavouring 832
substances. Food Chem Toxicol 37, 207-232. 833
834
Müller L., Mauthe R.J., Riley C.M., Andino M.M., De Antonis D., Beels C., DeGeorge J., De 835
Knaep A.G.M., Ellison D., Fagerland J.A., Frank R., Fritschel B., Galloway S., Harpur E., 836
Humfrey C.D.N., Jacks A.S.J., Jagota N., Mackinnon J., Mohan G., Ness D.K., O’Donovan 837
M.R., Smith M.D., Vudathala G., Yotti L. (2006). A rationale for determining, testing, and 838
controlling specific impurities in pharmaceuticals that possess potential for genotoxicity, Reg 839
Tox Pharm 44, 198-211. 840
841
842
ATTACHMENTS 843
844
845
APPENDICES 846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
04.08.2014
Datei
PD
