`and Reviewers
`Estimating the Safe Starting Dose in
`Clinical Trials for Therapeutics in
`Adult Healthy Volunteers
`
`DRAFT GUIDANCE
`
`This guidance document is being distributed for comment purposes only.
`
`Comments and suggestions regarding this draft document should be submitted within 60 days of
`publication in the Federal Register of the notice announcing the availability of the draft
`guidance. Submit comments to Dockets Management Branch (HFA-305), Food and Drug
`Administration, 5630 Fishers Lane, rm. 1061, Rockville, MD 20852. All comments should be
`identified with the docket number listed in the notice of availability that publishes in the Federal
`Register.
`
`For questions regarding this draft document contact (CDER) Robert Osterberg, 301-594-5476 or
`(CBER) Martin Green 301-827-5349.
`
`U.S. Department of Health and Human Services
`Food and Drug Administration
`Center for Drug Evaluation and Research (CDER)
`Center for Biologics Evaluation and Research (CBER)
`December 2002
`Pharmacology and Toxicology
`
`\\CDS029\CDERGUID\3814dft.doc
`11/18/02
`
`CFAD Exhibit 1042
`
`
`
`Guidance for Industry
`and Reviewers
`Estimating the Safe Starting Dose in
`Clinical Trials for Therapeutics in
`Adult Healthy Volunteers
`
`Additional copies are available from:
`
`Office of Training and Communications
`Division of Drug Information, HFD-240
`Center for Drug Evaluation and Research
`Food and Drug Administration
`5600 Fishers Lane
`Rockville, MD 20857
`(Tel) 301-827-4573
`http://www.fda.gov/cder/guidance/index.htm
`or
`
`Office of Communication, Training and
`Manufacturers Assistance (HFM-40)
`Center for Biologics Evaluation and Research
`Food and Drug Administration
`1401 Rockville Pike
`Rockville, MD 20852-1448
`(Internet) http://www.fda.gov/cber/guidelines.htm
`Mail: The Voice Information System at 800-835-4709 or 301-827-1800
`
`U.S. Department of Health and Human Services
`Food and Drug Administration
`Center for Drug Evaluation and Research (CDER)
`Center for Biologics Evaluation and Research (CBER)
`December 2002
`Pharmacology and Toxicology
`
`\\CDS029\CDERGUID\3814dft.doc
`11/18/02
`
`
`
`Draft — Not for Implementation
`
`Table of Contents
`
`INTRODUCTION................................................................................................................. 1
`I.
`II. SCOPE ................................................................................................................................... 1
`III. OVERVIEW OF THE ALGORITHM ............................................................................... 3
`IV. STEP 1: NO OBSERVED ADVERSE EFFECT LEVEL (NOAEL)
`DETERMINATION ............................................................................................................ 4
`V. STEP 2: HUMAN EQUIVALENT DOSE (HED) CALCULATION............................... 5
`A. Conversion Based on Body Surface Area ....................................................................................5
`B. Basis for Using Mg/Kg Conversions.............................................................................................7
`C. Other Exceptions to Mg/M2 Scaling Between Species ................................................................8
`VI. STEP 3: MOST APPROPRIATE SPECIES SELECTION............................................ 8
`VII.
`STEP 4: APPLICATION OF SAFETY FACTOR....................................................... 9
`A.
`Increasing the Safety Factor .......................................................................................................10
`B. Decreasing the Safety Factor ......................................................................................................11
`VIII. STEP 5: CONSIDERATION OF THE PHARMACOLOGICALLY ACTIVE
`DOSE (PAD)................................................................................................................................ 11
`IX. SUMMARY ......................................................................................................................... 12
`REFERENCES............................................................................................................................ 13
`APPENDIX A.............................................................................................................................. 15
`APPENDIX B .............................................................................................................................. 17
`APPENDIX C.............................................................................................................................. 22
`APPENDIX D.............................................................................................................................. 23
`APPENDIX E .............................................................................................................................. 25
`GLOSSARY................................................................................................................................. 26
`
`\\CDS029\CDERGUID\3814dft.doc
`11/18/02
`
`
`
`Draft — Not for Implementation
`
`Guidance for Industry and Reviewers1
`
`Estimating the Safe Starting Dose in Clinical Trials
`for Therapeutics in Adult Healthy Volunteers
`
`This draft guidance, when finalized, will represent the Food and Drug Administration's (FDA's) current
`thinking on this topic. It does not create or confer any rights for or on any person and does not operate to
`bind FDA or the public. An alternative approach may be used if such approach satisfies the requirements
`of the applicable statutes and regulations.
`
`I.
`
`INTRODUCTION
`
`This guidance outlines a process (algorithm) and vocabulary for deriving the maximum
`recommended starting dose (MRSD) for "first in human" clinical trials of new molecular entities
`in adult healthy volunteers and recommends a standardized process by which the MRSD can be
`selected. The purpose of this process is to ensure the safety of the human volunteers.
`
`The goals of this guidance are to (1) establish a consistent terminology for discussing the starting
`dose, (2) provide common conversion factors for deriving a human equivalent dose, and (3)
`delineate a strategy for selecting the MRSD for adult healthy volunteers, regardless of the
`projected clinical use. This process is diagrammed with a flow chart that presents the decisions
`and calculations used to generate the MRSD from animal data.
`
`II.
`
`SCOPE
`
`The process identified in this document pertains to determining the MRSD for adult healthy
`subjects when beginning a clinical investigation of any new drug or biological therapeutic that
`has been studied in animals. This document is not pertinent to prophylactic vaccines or
`endogenous proteins (i.e., recombinant clotting factors) used at physiologic concentrations. The
`process outlined in this document does not address dose escalation or maximum allowable doses
`in clinical trials.
`
`
`1 This guidance has been prepared by the Office of New Drugs in the Center for Drug Evaluation and Research
`(CDER) in cooperation with the Center for Biologics Evaluation and Research (CBER) at the Food and Drug
`Administration.
`
`\\CDS029\CDERGUID\3814dft.doc
`11/18/02
`
`1
`
`1
`
`234
`
`5
`
`678 9
`
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`29
`30
`31
`32
`33
`34
`35
`36
`37
`38
`39
`
`
`
`Draft — Not for Implementation
`
`40
`41
`42
`43
`44
`45
`46
`47
`48
`49
`50
`51
`52
`53
`54
`55
`56
`57
`58
`59
`60
`61
`62
`63
`64
`
`Although the process outlined in this document uses observed toxicities, administered doses, and
`an algorithmic approach to calculate the MRSD, an alternative approach could be proposed that
`places primary emphasis on animal pharmacokinetics and modeling rather than dose. In a
`limited number of cases, animal pharmacokinetic data may be useful in determining initial
`clinical doses.2 However, in the majority of new INDs, animal data are not available in
`sufficient detail to construct a scientifically valid, pharmacokinetic model whose aim is to
`accurately project an MRSD.
`
`Toxicity should be avoided at the initial dose. However, doses should be chosen that allow
`reasonably rapid attainment of the phase 1 trial objectives (e.g., assessment of the therapeutic's
`tolerability, pharmacodynamic or pharmacokinetic profile). All of the relevant preclinical data,
`including information on the pharmacologically active dose, the full toxicologic profile of the
`compound, and the pharmacokinetics (absorption, distribution, metabolism, and excretion) of the
`therapeutic, should be considered when determining the MRSD. Starting with doses lower than
`the MRSD is always a possible option and may be particularly appropriate to meet some clinical
`trial objectives.
`
`The remainder of this document will focus on the recommended algorithmic process for starting
`dose extrapolation from animals to humans based on administered doses, since this method will
`likely be useful for the majority of new INDs seeking to investigate new drugs in healthy
`volunteers. Some classes of drugs (e.g., many cytotoxic or biological agents) are commonly
`introduced into initial clinical trials in patient volunteers rather than healthy volunteers.
`Typically, this occurs when a drug is suspected or known to be unavoidably toxic. Although this
`document does not specifically address starting doses in patients, many principles and some
`approaches recommended here may be applicable to designing such trials.
`
`
`2 If the parent drug is measured in the plasma at multiple times and fits the range of toxic dose for two or more
`animal species, it may be possible to develop a pharmacokinetic model predicting human doses and concentrations
`and draw inferences about human safe plasma levels in the absence of prior human data. While quantitative
`modeling for this purpose may be straightforward, the following points suggest this approach may present a number
`of difficulties when evaluating estimates of a safe starting dose. Generally, at the time of IND initiation, there are a
`number of unknowns regarding animal toxicity and comparability of human and animal pharmacokinetics and
`metabolism: (1) human bioavailability and metabolism may differ significantly from that of animals; (2)
`mechanisms of toxicity may not be known (i.e., toxic accumulation in a peripheral compartment; and/or (3) toxicity
`may be due to an unidentified metabolite, not parent drug. Thus, to rely on pharmacokinetic models (based on
`parent drug in plasma) to gauge starting doses would require multiple untested assumptions. Modeling may be used
`with greatest validity to estimate human starting doses in special cases where few underlying assumptions would be
`necessary. Such cases are exemplified by large molecular weight proteins (like humanized monoclonal antibodies),
`which are intravenously administered, are removed from circulation by endocytosis rather than metabolizism, have
`immediate and detectable effects on blood cells, and have a volume of distribution limited to the plasma volume.
`Here, allometric, pharmacokinetic, and pharmacodynamic models have been useful in identifying the human mg/kg
`dose that would be predicted to correlate with safe drug plasma levels in nonhuman primates. Even in these cases,
`uncertainties (such as differences between human and chimpanzee receptor sensitivity or density) have been shown
`to affect human pharmacologic or toxicologic outcomes, and the use of safety factors as described in this document
`is still warranted.
`
`\\CDS029\CDERGUID\3814dft.doc
`11/18/02
`
`2
`
`
`
`65
`66
`67
`68
`69
`70
`71
`72
`73
`74
`75
`76
`77
`78
`79
`80
`81
`82
`83
`84
`85
`86
`87
`88
`89
`90
`91
`92
`93
`94
`95
`96
`97
`98
`99
`100
`101
`102
`103
`104
`105
`106
`
`Draft — Not for Implementation
`
`III.
`
`OVERVIEW OF THE ALGORITHM
`
`The process for selecting the MRSD is presented in Figure 1 and described in this section. The
`major elements C the determination of the no observed adverse effect levels (NOAELs) in the
`tested species, conversion of NOAELs to human equivalent dose (HED), selection of the most
`appropriate species, and application of a safety factor C are all discussed in greater detail in
`subsequent sections. Situations are also discussed in which the algorithm should be modified.
`The algorithm is intended to be used for systemically administered therapeutics. Topical,
`intranasal, intra-tissue, and compartmental administration routes and depot formulations may
`have additional considerations, but similar principles should apply.
`
`The process of calculating the MRSD should begin after the toxicity data have been analyzed.
`Although only the NOAEL should be used directly in the algorithm for calculating a MRSD,
`other data (exposure/toxicity relationships, pharmacologic data, or prior clinical experience with
`related drugs) can affect the choice of most appropriate species, scaling, and safety factors.
`
`The NOAEL for each species tested should be identified, then each should be converted to the
`human equivalent dose (HED) using appropriate scaling factors. For most systemically
`administered therapeutics, this conversion should be based on the normalization of doses to body
`surface area. Although body surface area conversion is the usual way to approximate equivalent
`exposure if no further information is available, in some cases, extrapolating doses based on other
`parameters may be more appropriate. This decision should be based on the data available for the
`individual case. The body surface area normalization and the extrapolation of the animal dose to
`human dose should be done in one step by dividing the NOAEL in each of the animal species
`studied by the appropriate body surface area conversion factor (BSA-CF). This is a unitless
`number that converts mg/kg dose for each animal species to the mg/kg dose in humans, which is
`equivalent to the animal’s NOAEL on a mg/m2 basis. The resulting figure is called a human
`equivalent dose (HED). The species that generates the lowest HED is called the most sensitive
`species.
`
`When information indicates that a particular species is most relevant for assessing human risk
`(and deemed the most appropriate species), the HED for that species should be used in
`subsequent calculations, regardless of whether this species was the most sensitive. This case is
`common for biologic therapies, many of which have high selectivity for binding to human target
`proteins, and limited reactivity in species commonly used for toxicity testing. In such cases, in
`vitro binding and activity studies should be done to select appropriate, relevant species before
`toxicity studies are designed (please refer to the ICH3 guidance for industry S6 Preclinical Safety
`Evaluation of Biotechnology-Derived Pharmaceuticals for more details). Additionally, a species
`might be considered an inappropriate toxicity model for a given drug if a dose-limiting toxicity
`in that species was concluded to be of limited value for human risk assessment (based on
`historical comparisons of toxicities in species to those in humans across a therapeutic class). In
`
`3 International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for
`Human Use (ICH).
`
`\\CDS029\CDERGUID\3814dft.doc
`11/18/02
`
`3
`
`
`
`107
`108
`109
`110
`111
`112
`113
`114
`115
`116
`117
`118
`119
`120
`121
`122
`123
`124
`125
`126
`127
`128
`129
`130
`131
`132
`133
`134
`135
`136
`137
`138
`139
`140
`141
`142
`143
`144
`145
`146
`147
`148
`149
`150
`151
`
`Draft — Not for Implementation
`
`this case, data from that species should not be used to derive the HED. Without any additional
`information to guide the choice of the most appropriate species for assessing human risk, the
`most sensitive species is designated the most appropriate, because using the lowest HED would
`generate the most conservative starting dose.
`
`A safety factor should then be applied to the HED to increase assurance that the first dose in
`humans will not cause adverse effects. The use of the safety factor should be based on the
`possibility that humans may be more sensitive to the toxic effects of a therapeutic agent than
`predicted by the animal models, that bioavailability may vary across species, and that the models
`tested do not evaluate all possible human toxicities. For example, ocular disturbances or pain
`(such as severe headaches) in humans can be significant dose-limiting toxicities that may go
`undetected in animal studies.
`
`In general, a safety factor of 10 is recommended. The MRSD should be obtained by dividing the
`HED by the safety factor. Safety concerns or design shortcomings noted in animal studies may
`increase the safety factor, and thus reduce the MRSD further. Alternatively, information about
`the pharmacologic class (well-characterized classes of therapeutics with extensive human clinical
`and preclinical experience) may allay concerns and form the basis of reducing the magnitude of
`the default safety factor and increasing the MRSD. Although a dose lower than the MRSD can
`be used as the actual starting dose, the process described here will derive the maximum
`recommended starting dose. This algorithm generates a MRSD in units of mg/kg, a common
`method of dosing used in phase 1 trials, but the equations and conversion factors provided in this
`document (Table one, second column) can be used to generate final dosing units in the mg/m2
`form if desired.
`
`As previously stated, for purposes of initial clinical trials in adult healthy volunteers, the HED
`should ordinarily be calculated from the animal NOAEL. If the HED is based on an alternative
`index of effect, such as the pharmacologically active dose (PAD), this exception should be
`prominently stipulated in descriptions of starting dose calculations.
`
`The remainder of this document provides a description of the individual steps in the
`recommended process and the reasoning behind each step. The method is supported by a general
`review and analysis by CDER and CBER examining the results from a number of therapeutics
`entered into development.
`
`STEP 1: NO OBSERVED ADVERSE EFFECT LEVEL (NOAEL)
`IV.
`DETERMINATION
`
`The first step in determining the MRSD is to review and evaluate the available animal data so
`that a NOAEL can be determined for each study. Several differing definitions of NOAEL exist,
`but for selecting a starting dose, the following is used here: the highest dose level that does not
`produce a significant increase in adverse effects. In this context, adverse effects that are
`statistically significant and adverse effects that may be clinically significant (even if they are not
`statistically significant) should be considered in the determination of the NOAEL. The NOAEL
`
`\\CDS029\CDERGUID\3814dft.doc
`11/18/02
`
`4
`
`
`
`152
`153
`154
`155
`156
`157
`158
`159
`160
`161
`162
`163
`164
`165
`166
`167
`168
`169
`170
`171
`172
`173
`174
`175
`176
`177
`178
`179
`180
`181
`182
`183
`184
`185
`186
`187
`188
`189
`190
`191
`192
`193
`194
`195
`196
`
`Draft — Not for Implementation
`
`is a generally accepted benchmark for safety when derived from appropriate animal studies and
`can serve as the starting point for determining a reasonably safe starting dose of a new
`therapeutic in healthy (or asymptomatic) human volunteers.
`
`The NOAEL is not the same as the no observed effect level (NOEL), which refers to any effect,
`not just adverse ones, although in some cases the two might be identical. The definition of the
`NOAEL, in contrast to that of the NOEL, reflects the view that some effects observed in the
`animal may be acceptable pharmacodynamic actions of the therapeutic and may not raise a safety
`concern. The NOAEL should not be confused with lowest observed adverse effect level
`(LOAEL) or maximum tolerated dose (MTD). Both of the latter concepts are based on findings
`of adverse effects and are not generally used as benchmarks for establishing safe starting doses
`in adult healthy volunteers. The term level refers to dose or dosage, generally expressed as
`mg/kg or mg/kg/day.
`
`Initial IND submissions for first in human studies by definition lack human data or formal
`allometric comparison of pharmacokinetics. Measurements of systemic levels or exposure (i.e.,
`AUC or Cmax) cannot be employed for setting a safe starting dose in humans, and it is critical to
`rely on dose and observed toxic response data from adequate and well-conducted toxicology
`studies. However, there are cases where data on bioavailability, metabolite profile, and plasma
`drug levels associated with toxicity may influence the choice of the NOAEL. One such case
`would be when saturation of drug absorption occurs at a dose that produces no toxicity. In this
`case, the lowest saturating dose, not the highest (non-toxic) dose, should be used for calculating
`the HED.
`
`There are essentially three types of findings in nonclinical toxicology studies that can be used to
`determine the NOAEL: (1) overt toxicity (e.g., clinical signs, macro- and microscopic lesions);
`(2) surrogate markers of toxicity (e.g., serum liver enzyme levels); and (3) exaggerated
`pharmacodynamic effects. Although the nature and extent of adverse effects can vary greatly
`with different types of therapeutics and it is anticipated that in many instances experts will
`disagree on the characterization of effects as being adverse or not, the use of NOAEL as a
`benchmark for dose-setting in healthy volunteers should be acceptable to all responsible
`investigators. As a general rule, an adverse effect observed in nonclinical toxicology studies
`used to define a NOAEL for the purpose of dose-setting should be based on an effect that would
`be unacceptable if produced by the initial dose of a therapeutic in a phase 1 clinical trial
`conducted in adult healthy volunteers.
`
`V.
`
`STEP 2: HUMAN EQUIVALENT DOSE (HED) CALCULATION
`
`A.
`
`Conversion Based on Body Surface Area
`
`After the NOAELs in the relevant animal studies have been determined, they are converted to
`human equivalent doses (HEDs). A decision should be made regarding the most appropriate
`method for extrapolating the animal dose to the equivalent human dose. Toxic endpoints for
`therapeutics administered systemically to animals, such as the MTD or NOAEL, are usually
`
`\\CDS029\CDERGUID\3814dft.doc
`11/18/02
`
`5
`
`
`
`197
`198
`199
`200
`201
`202
`203
`204
`205
`206
`207
`208
`209
`210
`211
`212
`213
`214
`215
`216
`217
`218
`219
`220
`221
`222
`223
`224
`225
`226
`227
`228
`229
`230
`231
`232
`233
`
`Draft — Not for Implementation
`
`assumed to scale well between species when doses are normalized to body surface area (i.e.,
`mg/m2). The basis for this assumption lies primarily with the work of Freireich et al. (1996) and
`Schein et al. (1970). These investigators reported that, for antineoplastic drugs, doses lethal to
`10 percent of rodents (LD10s) and MTDs in non-rodents both correlated with the human MTD
`when the doses were normalized to the same administration schedule and expressed as mg/m2.
`Despite the subsequent analyses showing that the MTDs for this set of drugs scale best between
`species when doses are normalized to W0.75 rather than W0.67 (inherent in body surface area
`normalization), normalization to body surface area has remained a widespread practice for
`estimating an HED based on an animal dose.
`
`An analysis of the impact of the allometric exponent on the conversion of an animal dose to the
`HED was conducted (see Appendix A). Based on this analysis and on the fact that correcting for
`body surface area increases clinical trial safety by resulting in a more conservative starting dose
`estimate, it was concluded that the approach of converting NOAEL doses to an HED based on
`body surface area correction factors (i.e., W0.67) should be maintained for selecting starting doses
`for initial studies in adult healthy volunteers. Nonetheless, use of a different dose normalization
`approach, such as directly equating the human dose to the NOAEL in mg/kg, may be appropriate
`in some circumstances. Deviations from the surface area approach should be justified. The basis
`for justifying direct mg/kg conversion and examples in which other normalization methods are
`appropriate are described in the following subsection.
`
`Although normalization to body surface area is an appropriate method for extrapolating doses
`between species, consistent factors for converting doses from mg/kg to mg/m2 have not always
`been used. Given that body surface area normalization provides a reasonable approach for
`estimating an HED, the factors used for converting doses from each species should be
`standardized. Since surface area varies with W0.67, the conversion factors are therefore
`dependent on the weight of the animals in the studies. However, analyses conducted to address
`the effect of body weight on the actual BSA-CF (body surface area - conversion factor)
`demonstrated that a standard factor provides a reasonable estimate of the HED over a broad
`range of human and animal weights (see Appendix B). The conversion factors and divisors
`shown in Table 1, below, are therefore recommended as the standard values to be used for
`interspecies dose conversions for NOAELs in CDER and CBER. These factors may also be
`applied when comparing safety margins for other toxicity endpoints (e.g., reproductive toxicity
`and carcinogenicity) when other data for comparison, (i.e., AUCs) are unavailable or are
`otherwise inappropriate for comparison.
`
`Table 1: Conversion of Animal Doses to Human Equivalent Doses
`(HED) Based on Body Surface Area
`To convert animal
`To convert animal dose in mg/kg
`to HEDa in mg/kg, either:
`dose in mg/kg to
`dose in mg/m²,
`Divide
`Multiply
`multiply by km
`animal dose by:
`Animal dose by:
`below:
`
`Species
`
`\\CDS029\CDERGUID\3814dft.doc
`11/18/02
`
`6
`
`
`
`Draft — Not for Implementation
`
`37
`25
`3
`5
`6
`7
`8
`12
`20
`
`---
`---
`12.3
`7.4
`6.2
`5.3
`4.6
`3.1
`1.8
`
`---
`---
`0.08
`0.13
`0.16
`0.19
`0.22
`0.32
`0.54
`
`Human
`Child (20 kg)b
`Mouse
`Hamster
`Rat
`Ferret
`Guinea pig
`Rabbit
`Dog
`Primates:
`Monkeysc
`0.32
`3.1
`12
`0.16
`6.2
`6
`Marmoset
`0.19
`5.3
`7
`Squirrel monkey
`0.54
`1.8
`20
`Baboon
`0.73
`1.4
`27
`Micro-pig
`0.95
`1.1
`35
`Mini-pig
`a Assumes 60 kg human. For species not listed or for weights outside the standard ranges, human
`equivalent dose can be calculated from the formula:
`HED = animal dose in mg/kg x (animal weight in kg/human weight in kg)0.33.
`b This km is provided for reference only since healthy children will rarely be volunteers for phase 1 trials.
`c For example, cynomolgus, rhesus, stumptail.
`
`234
`235
`236
`237
`238
`239
`240
`241
`242
`243
`244
`245
`246
`247
`248
`249
`250
`251
`252
`253
`254
`255
`256
`257
`258
`259
`260
`
`B.
`
`Basis for Using Mg/Kg Conversions
`
`The factors in Table 1 for scaling animal NOAEL to HEDs are based on the assumption that
`doses scale 1:1 between species when normalized to body surface area. However, there are
`occasions for which scaling based on body weight (i.e., setting the HED (mg/kg) = NOAEL
`(mg/kg)) may be more appropriate. To consider mg/kg scaling for a therapeutic, the available
`data should show that the NOAEL occurs at a similar mg/kg dose across species. The factors
`below should be satisfied before extrapolating to the HED on a mg/kg basis rather than using the
`mg/m2 approach. Note that mg/kg scaling will give a 12-, 6-, and 2- fold higher HED than the
`default mg/m2 approach for mice, rats, and dogs, respectively. If these factors cannot be met, the
`mg/m2 scaling approach for determining the HED should be followed as it will lead to a safer
`MRSD.
`
`1.
`
`2.
`
`NOAELs occur at a similar mg/kg dose across test species (for the studies with a
`given dosing regimen relevant to the proposed initial clinical trial).
`
`If only two NOAELs from toxicology studies in separate species are available,
`one of the following criteria should also be true:
`
`• The therapeutic is administered orally and the dose is limited by local
`toxicities. Gastrointestinal (GI) compartment weight scales by W0.94 . GI
`
`\\CDS029\CDERGUID\3814dft.doc
`11/18/02
`
`7
`
`
`
`261
`262
`263
`264
`265
`266
`267
`268
`269
`270
`271
`272
`273
`274
`275
`276
`277
`278
`279
`280
`281
`282
`283
`284
`285
`286
`287
`288
`289
`290
`291
`292
`293
`294
`295
`296
`297
`298
`299
`300
`301
`302
`303
`304
`305
`
`Draft — Not for Implementation
`
`volume determines the concentration of the therapeutic in the GI tract. It is
`thus reasonable that the toxicity of the therapeutic would scale by mg/kg
`(W1.0).
`
`• The toxicity in humans (for a particular class) is dependent on an exposure
`parameter that is highly correlated across species with dose on a mg/kg basis.
`For example, complement activation by systemically administered antisense
`oligonucleotides in humans is believed to be dependent upon Cmax (Geary et
`al., 1997). For some antisense drugs, the Cmax correlates across nonclinical
`species with mg/kg dose and in such instances mg/kg scaling would be
`justified.
`
`• Other pharmacologic and toxicologic endpoints also scale between species by
`mg/kg for the therapeutic. Examples of such endpoints include the MTD,
`lowest lethal dose, and the pharmacologically active dose.
`
`Other Exceptions to Mg/M2 Scaling Between Species
`
`Therapeutics administered by alternative routes (e.g., topical, intranasal,
`subcutaneous, intramuscular) for which the dose is limited by local toxicities.
`Such therapeutics should be normalized to concentration (mg/area of application,
`for instance) or amount of drug (mg) at the application site.
`Therapeutics administered into anatomical compartments that have little
`subsequent distribution outside of the compartment. Examples are intrathecal,
`intravesical, intraocular, intrapleural, and intraperitoneal administration. Such
`therapeutics should be normalized between species according to the
`compartmental volumes and concentrations of the therapeutic.
`Biological products administered intravascularly with Mr > 100,000 daltons. Such
`therapeutics should be normalized to mg/kg.
`
`C.
`
`1.
`
`2.
`
`3.
`
`VI.
`
`STEP 3: MOST APPROPRIATE SPECIES SELECTION
`
`After the HEDs have been determined from the NOAELs from all toxicology studies relevant to
`the proposed human trial, the next step is to pick one HED for subsequent derivation of the
`MRSD. This HED should be chosen from the most appropriate species. In the absence of data
`on species relevance, a default position is that the most appropriate species for deriving the
`MRSD for a trial in adult healthy volunteers is the most sensitive species (i.e., the species in
`which the lowest HED can be identified).
`
`Factors that could influence the choice of the most appropriate species rather than the default to
`the most sensitive species include: (1) differences in the absorption, distribution, metabolism and
`elimination (ADME) of the therapeutic between the species; (2) class experience that may
`indicate a particular model is predictive of human toxicity; or (3) limited biological cross-species
`pharmacologic reactivity of the therapeutic. This latter point is especially important for
`
`\\CDS029\CDERGUID\3814dft.doc
`11/18/02
`
`8
`
`
`
`Draft — Not for Implementation
`
`biological therapeutics as many are human proteins that bind to human or non-human primate
`targets (see ICH guidance S6).
`
`When determining the MRSD for the first dose of a new therapeutic in humans, absorption,
`distribution, and elimination parameters will not be known for humans. Comparative
`metabolism data, however, might be available based on in vitro studies. These data are
`particularly relevant when there are marked differences in both the in vivo metabolite profiles
`and HEDs in animals. Class experience implies that previous studies have demonstrated that a
`particular animal model is more appropriate for the assessment of safety for a particular class of
`therapeutics. For example, in the nonclinical safety assessment of the phosphorothioate
`antisense drugs, the monkey is considered the most appropriate species because monkeys
`experience the same dose limiting toxicity as humans, (i.e., complement activation), whereas
`rodents do not. For this class of therapeutics, the MRSD would usually be based on the HED for
`the NOAEL in monkeys regardless of whether it was lower than that in rodents, unless unique
`dose limiting toxicities were observed with the new antisense compound in the rodent species.
`Similarities of biochemistry and physiology between the species and humans that are relevant to
`the limiting toxicities of the therapeutic should also be considered under class experience. If a
`species is the most sensitive but has differences in physiology compared to humans that sensitize
`it to the therapeutic, it may not be the most appropriate species for selecting the MRSD.
`
`VII.
`
`STEP 4: APPLICATION OF SAFETY FACTOR
`
`Once the HED of the NOAEL in the most appropriate species has been determined, a safety
`factor is then applied in order to provide a margin of safety for protection of human subjects
`receiving the initial clinical dose. This safet