Overview – Internal review

• Presentation title: Introduction to toxicology Assessment

• Track title: Contamination / toxicology
• Speaker: Andrew Bartholomaeus
• Date / Time: Tuesday
• Time allotted: 11:00 -12:30
• Dot point overview:
• A basic introduction to the toxicology relevant to establishing a PDE or OEL
• Introduction to the basic vocabulary and concepts
• General overview of the types of studies and toxicological endpoints that must be
considered
• Where to get the data to support the toxicology assessment
• Establishing a point of departure (NOAELS

Slide 1 © PharmOut 2015

 RISK IDENTIFICATION FOR
MANUFACTURE IN SHARED
FACILITIES

Introduction to Toxicology
Assessment

PROF. ANDREW BARTHOLOMAEUS

SCHOOL OF PHARMACY, UNIVERSITY OF CANBERRA
THERAPEUTIC RESEARCH CENTRE,
SCHOOL OF MEDICINE, UNIVERSITY OF QUEENSLAND
CEO, BARTCROFTS SCIENTIFIC SERVICES PTY LTD

EMAIL; [email protected]

BartCrofts Scientific Services Pty Ltd; [email protected]

 Basic Concepts
3

risk = f (hazard*exposure)

BartCrofts Scientific Services Pty Ltd; [email protected]

 Key Terms
4

 NOAEL - No Observed Adverse Effect Level

 The highest dose in a study at which no adverse effects of treatment are observed

 LOAEL – Lowest Observed Adverse Effect Level

 The lowest dose in a study at which adverse effects were observed

 ADI, TDI - Acceptable or Tolerable Daily Intake (oral) – food & food contaminants
 The amount of a substance that can be taken orally every day for a lifetime without appreciable toxicological effect

 PDE / ADE -Permitted (or Acceptable) daily Exposure

 A dose that is unlikely to cause an adverse effect if an individual is exposed every day for a lifetime
 Route specific but not limited to oral intake
 Equivalent to Acceptable Daily Exposure
 OELs Occupational Exposure Limits
 TLV, PEL – Threshold limit value, Permissible Exposure limit
 OEB – Occupational Exposure Band
 LD50: lethal dose for 50% of animals treated
 Historical Controls
 A collation of data from control groups of the same strain, sex, source and age of animal constructed over a period
of time
 Provides normative data for the range of standard experimental parameters

BartCrofts Scientific Services Pty Ltd; [email protected]

 Defining toxicity
5

 Everything is toxic at some level of exposure by some route

 Oxygen, Nitrogen and water are all toxic
 The Objective of Toxicological Risk Assessment is Identification of the
boundary (zone) between safety and toxicity
 How toxic an agent is will depend on:
 Physical and chemical properties of the toxicant
 Route of administration
 Dose, duration and frequency of exposure
 How system metabolises and eliminates the toxic agent (ADME)
 Susceptibility of subject to toxic agent
 Local versus systemic toxicity

 DOSE METRICS are defined by

 How much is given/taken (total dose)
 Per kg of body mass (dose per unit of body weight)
 Over what period (the duration of the exposure eg iv infusion over 1 minute or 1 hour)
 Frequency (constant, hourly, daily, weekly etc)
 Duration – for a week, month, year, lifetime

BartCrofts Scientific Services Pty Ltd; [email protected]

 Drug Toxicity
6

 Drugs may exhibit toxicity due to

 Excessive pharmacological action (primary pharmacology)
 Clinically predictable/manageable
 Occurs at the upper dose range for the individual patient
 Secondary pharmacological effects due to lack of specificity for
the primary target (receptor)
 Generally clinically manageable
 Generally occurs across a defined dose range
 Off target toxicity unrelated to the primary pharmacological
action
 May be idiosyncratic and unpredictable
 Often not readily managed clinically
 May occur in some patients at low or sub therapeutic doses

BartCrofts Scientific Services Pty Ltd; [email protected]

 Guidelines Applicable to APIs
7

 Preclinical studies on APIs generally need to be

conducted in accordance with specific guidelines on
study design, the studies required by specific regulators
and on GLP certification
 OECD
 Industrial chemicals, pesticides, some drug studies
 ICH
 Human pharmaceuticals
 VICH
 Veterinary pharmaceuticals
 EMA (European Medicines Agency
 FDA (US Food and Drug Administration)
 TGA (Therapeutic Goods Administration) – mostly adopts from
EU and USA
OECD – Guidelines
8

–http://www.oecd-ilibrary.org/environment/oecd-guidelines-for-the-testing-of-chemicals-section-4-health-effects_20745788
ICH - Pharmaceuticals
9
 Data Quality: Good Laboratory Practice (GLP)
10

 Codes of GLP ensure that the processes of design, conduct and

documentation of nonclinical studies are of a very high standard
 But do not of themselves make a study good or bad
 GLP-compliance not required for pharmacodynamic or
pharmacokinetic studies, but routinely expected for toxicity studies
 Published studies not conducted by the innovator will not usually be GLP
 In toxicological assessment, more weight is generally given to GLP-
compliant than non-compliant studies
 But common sense and judgement need to be applied here
 Ideally all pivotal safety studies should be under GLP conditions
 GLP Guidelines specify documentation details and processes to
provide a high level of QA and transparency of process
 Provides no reassurance of quality of study conduct, ie
competence, but does facilitate detection of fraud and poor study
design or execution.
 Critical elements of toxicity studies
11

 Dose:
 The amount of a substance administered per unit of weight/animal/surface area and
per unit of time at a given frequency for a specified duration by a specific route
 A dose is meaningless without specifying all of these parameters, eg
 5 mg/kg bw/day for 5 days orally
 1 mg/kg bw eight hourly for 1 month by ip injection
 Response:
 Effects in the test species from the administered substance; Response can be
pharmacological, toxicological or physiological/adaptive.
 Dose-response relationship
 establishes causality that a substance has in fact induced the observed effects
 establishes the lowest dose where an induced effect occurs and the highest dose where no
effect occurs- the threshold effect and no-effect dose
 Beware however that these are partially artefacts of study design
 Indicates the rate of increase of injury with increased dose - the slope for the dose response.
 Gives some indication of leeway in exposure – ie does a small increase in dose cause a large or
small increase in injury

BartCrofts Scientific Services Pty Ltd; [email protected]

 Animal versus human exposure (Dose)
12

 Not all tissues/effects will be best measured by the same

exposure parameter
 mg/kg bw may be a better comparator for
 GIT effects likely to be due to purely local concentrations
 Effects on the liver and kidney - may be related to the total flux of drug rather than the
AUC
 Bone toxicity - osteoporosis drugs generally concentrate in the bone and levels are not
directly proportional to blood AUC values
 Exposure ratio = AUC or Cmax in animals / AUC or Cmax in
humans
 This may be determined at the:
 NOEL (No Observable Effect Level), and/or
 NOAEL (No Observable Adverse Effect Level), and/or
 LOEL (Lowest Observable Effect Level) for a given toxicity, and/or
 HD (highest dose tested)
 Safety margin = ratio of exposure at no-effect dose in
animals cf the exposure at the maximal human dose
 Two models of dose response
13

 Threshold: when the data indicates a

threshold below which no adverse effects can be
measured. threshold

 Non-threshold: when the data cannot

identify a threshold for adverse effects. (In this
case, the dose-response curve may be assumed Dose
to pass through zero).
 In practice all effects are likely to have a threshold
but this may be at exceptionally small
doses/exposures for some effects such as No threshold
genotoxicity
 The non threshold approach is a regulatory policy
determined by government directives in the USA –
it is a matter of some contention and disagreement Dose
amongst academic toxicologists

BartCrofts Scientific Services Pty Ltd; [email protected]

 Dose Response Relationship
14

•Dose-response-relationship between magnitude of dose and incidence

or severity of adverse effect

BartCrofts Scientific Services Pty Ltd; [email protected]

 NOAEL and LOAEL
15

 No Observed Adverse Effect Level (NOAEL)

 The highest dose at which no adverse effects were observed
 Lowest Observed Adverse Effect Level (LOAEL)
 The lowest dose at which adverse effects were observed
 They are the actual data points from experimental
animal studies or human clinical trials.
 Partially artefacts of dose selection and statistical
power of the study
 NOAEL and LOAEL are pivotal in the conduct of risk
assessments
 Used as Points of Departure PoD for PDE estimation

BartCrofts Scientific Services Pty Ltd; [email protected]

 X X

LOAEL
Severity
of response
NOAEL
X

X
X X
Log Dose

Threshold - lowest dose which causes an effect, below this dose, no

toxicity is observed
NOAEL – highest dose employed in a study at which no adverse effect
was observed (derived experimentally)
LOAEL - lowest dose at which there was an observed adverse effect
(derived experimentally)
BartCrofts Scientific Services Pty Ltd; [email protected]
 Each toxicological or pharmacological end point
(effect) may have its own dose response curve
17
Response

NO(A)EL A NOAEL B NOAEL C

Dose
BartCrofts Scientific Services Pty Ltd; [email protected]
 Threshold Approach
18

 Used to derive PDE, (ADI, TDI etc) for

 Non-genotoxic carcinogens
 Non-cancer effects
 Where a threshold occurs due to biological mechanisms such
as;
 Excretion, metabolism,
 repair or adaptation

 A very high likelihood that no effect will occur in humans

at or below threshold level (after adjusting for uncertainty)
 The Threshold approach doesn’t attempt to calculate a
numerical level of risk at low exposure (ie not quantitative)

BartCrofts Scientific Services Pty Ltd; [email protected]

 Threshold Approach
19
 Advantages
NOAEL relatively easy to determine

 Avoids the need to quantify the risk
 Default uncertainty factor of 100 is generally highly conservative
 Assumes all uncertain acts to increase risk in humans (unlikely)
 Long history of use to protect public health
 Crude, simple, easily explained, works reliably in almost all circumstances
 Disadvantages
 NOAEL depends on dose levels used
 Depends on biological effects monitored (if you do not look you may not
see)
 Limited by experimental protocol
 statistical & interpretive Power
 Relevance of the animal or other model to humans
 Limited use of dose response slope
 Choice of uncertainty factor largely arbitrary with poor substantiation
 Based on a Point estimate

BartCrofts Scientific Services Pty Ltd; [email protected]

 Dose Response
20

incidence
120
Mg/kg bw/d 0 2.3 4.6 9.3 19
100
No pregnant 20 22 21 23 21
animals 80
Foetal weight 3.41 3.32 3.33 3.31 3.16
60
Foetuses 122 126 118 133 131 incidence
% Foetuses with 0.8 1.0 16.1 71.4 97.0 40
cardiac dysplasia
20

0
0 5 10 15 20
-20

Dysplasia = abnormal development & probably abnormal function

BartCrofts Scientific Services Pty Ltd; [email protected]

 Selecting LOAELs
21

LOAELs

BartCrofts Scientific Services Pty Ltd; [email protected]

 Interpreting toxicity studies
22

BartCrofts Scientific Services Pty Ltd; [email protected]

 Interpreting toxicity studies
23

 Questioning the results

 Validity of older studies (pre 1970)
 Species-specific effects
 Statistical versus biological significance
 Weight of evidence-judgement
 adequacy, validity and appropriateness of data base.
 Is the effect a physiological or toxicological one?
 Physiology
 Variation within limits of normal function (e.g. variation in liver
enzymes following a meal).
 Toxicology
 Reversible/irreversible ?; Injurious ? and therefore adverse and
harmful

BartCrofts Scientific Services Pty Ltd; [email protected]

 Remember
24

 Hazards are the potential adverse effects a chemical

is capable of causing, risk is the likelihood of that
occurring at various exposures
 All substances are hazardous under certain
conditions of exposure (route & dose)
 Sometimes need only small amounts for it to be a
risk.
 Toxicity is a function of exposure and intrinsic hazard
- dose metrics are very important

BartCrofts Scientific Services Pty Ltd; [email protected]

 Safe Level of Exposure
(PDE)
25

Clinical and Preclinical

Toxicity study Data Set

No Observed Adverse Effect Level(s) (NOAEL)

Based on the most sensitive adverse effect(s)* in the
most sensitive animal study for that effect (mg/kg
bw/day)

Uncertainty factors:

Safe level of intake for humans for a lifetime.

PDE (mg/kg bw)
* Under some circumstances you may need different PDEs for different sub populations

BartCrofts Scientific Services Pty Ltd; [email protected]

 Setting a PDE
26

 Establishment of a PDE
 the amount of a substance in a medicine that can be administered
daily over a lifetime without appreciable health risk
 Highly conservative (safe) value
 Assumes the substance is administered every single day of a
persons life from infancy until death (VERY unlikely for an API as
contaminant)
 Assumes all uncertainty acts to increase the sensitivity of humans
compared to the test animals (also very unlikely)
 Other aspects of the Risk assessment such as
exposure assessment add additional layers of
conservatism
BartCrofts Scientific Services Pty Ltd; [email protected]
 Species Variation
27

 Dose vs. plasma concentration

 Sildenafil [Viagra]:
 Mouse: 1.0 mg/kg PO dose - plasma AUC 31 ng.h/ml
 Human: 1.0 mg/kg PO dose - plasma AUC 815 ng.h/mL
 Cause of the difference: Faster clearance in mice (91
ml/min/kg) than in humans (9.8 ml/min/kg)
 Consequence: significantly higher oral doses would need to be
administered to mice in the evaluation of the potential adverse
effects in humans.
 Preclinical safety testing: Oral doses up to 200 mg/kg bw/day
were studied in mice cf. clinical dose of 50 mg (1 mg/kg for a
50 kg patient).

BartCrofts Scientific Services Pty Ltd;

[email protected]
 Statistical Versus Biological Significance
28

 Scientists often use a p value of 0.05 to justify a conclusion that

an effect is more likely to be real rather than a chance variation
 What this actually means however is that there is a 5% (or 1 in 20)
chance of the finding being due to simple random variation - or
put another way if the study was run multiple times the finding
would occur on average in 1 in 20 studies by random chance alone
 In a toxicology study there are hundreds or even thousands of
parameter combinations being measured so it is inevitable that a
number of parameters will vary from control simply by random
chance
 This is where the concept of biological significance comes in. A
finding might be concluded to be biologically significant if for
example:
 There is a dose relationship
 The effect is consistent with that on related parameters
 A similar effect is seen in other studies or in other species
 The magnitude of the effect is outside that observed in historical controls

BartCrofts Scientific Services Pty Ltd; [email protected]

 Statistical vs biological vs toxicological
significance
29
 A lack of statistical significance  no effect
 the study may lack statistical power
 if pre-treatment values are available compare these to the treated values for individual animals (unlikely to be
possible for most published studies)
 look for similar findings across studies/species if possible
 look for correlated findings eg spleen enlargement + anaemia
 Not all substances which cause cancer in animal studies are “carcinogens”
 altered hormonal balance, persistent irritation, persistent stimulation of a tissue can cause cancer by non (primary)
genotoxic mechanisms
 the mechanism involved may not be relevant to human exposures
 Statistical significance  Biological significance
 eg in a full chronic toxicity study many thousands of end points are measured so statistically significant
results will occur at random - look for trends over time and doses
 Biological significance  Toxicological significance
 a small change in a parameter at high doses which remains in the normal range may be treatment related
but not adverse &  not toxicologically significant
 A clear treatment related adverse effect may still not be toxicologically significant
 eg pregnant rabbit gavage dosed with a bad tasting irritant substance, stops eating for 2 days then
resumes normal eating
 pups have decreased body weights, decreased survival and increased anomalies (eg delayed
development)
 effect is 20 to maternal nutritional deficit
 What if overt terato-genesis is seen? Eg cleft palate
 May need specific studies on the effect of nutritional deficit at the same time on development
 Species Specificity
30

 In general the animal models in use ARE highly predictive of effects

in man but there are some important exceptions
 Most species have some natural susceptibility to diseases which may
differ quantitatively or qualitatively to man
 eg thyroid follicular cell tumours - rats more sensitive
 liver tumours - mice more sensitive
 We can use historical control data to provide some indication of
natural propensities but remember genetic drift.
 Animal characteristics
 species, strain, source, age, sex, pregnancy
 background incidences & natural disease susceptibility
 suitability as a model for humans
 route and nature of administration
 diet, drinking water, gavage, parenteral, topical
 relevance to human exposure route
 “Historical Controls”
31

 aggregated incidences of ‘spontaneous’ pathological values collated from

control groups over a period of time
 to be of value they must be
 same sex, age, species, strain, source, nutritional status, dosing vehicle
 should be as temporally close to the comparator study as possible (genetic drift)
 Provide a measure of ‘background’ incidences and species variance
 Particularly useful for
 small group sizes (but first consider comparison against pre treatment values where available)
 assessing low incidence, dose related, non significant effects
 CAUTION; The more studies in the HC database the greater the range between extremes
 therefore more likely for a treatment related effect to fall within the ever broadening range

 comparison against the mean + 2 SD may often be more appropriate

 Dosing vehicle (eg corn oil) may not be without physiological effect
 Genetic drift may quickly render HC data non comparable
 Some parameters change with age so need to age match the HC data
 Validity of Findings
32
 Questions to ask of positive results
 is the effect real?
 what is the possible/probable/proposed mechanism?
 is it likely to occur in humans at clinical exposures?
 is there a specific population at risk
 eg poor/extensive metabolisers?
 Questions to ask of negative results (often related to
study design)
 was sample size insufficient?
 was exposure inadequate (both degree & duration)?
 were appropriate parameters measured?
 is the animal model unsuitable?
 are there differences in animal cf human metabolism?
 is pre-existing pathology absent?
 does tolerance develop?
 Study Types
33
RANGE OF STUDIES NORMALLY CONDUCTED
PHARMACOKINETICS
ACUTE TOXICITY
REPEAT DOSE TOXICITY
CARCINOGENICITY
GENOTOXICITY
REPRODUCTION
DEVELOPMENTAL
CLINICAL

BartCrofts Scientific Services Pty Ltd;

[email protected]
 Studies Providing Toxicology Data
34

 Preclinical
 In Silico - ie computer modelling
 Unlikely to be useful for PDE determination
 In vitro
 most important are genotoxicity assays
 In vivo
 Animal studies will often yield key data for PDE determination but
can be hard to obtain the data
 Clinical
 Always important but may not cover some of the critical
endpoints (reproduction and development)

BartCrofts Scientific Services Pty Ltd; [email protected]

STANDARD PRECLINCAL STUDIES
35

 Primary pharmacodynamics
 Mechanism, comparators, implications
 Secondary pharmacodynamics
 Mechanism, comparators, implications
 Cardiovascular and respiratory effects
 Autonomic and Central Nervous system
 Gastrointestinal
 Renal
 Endocrinological
 Immunological
 Pharmacokinetics and metabolism
 Species comparisons
 Effect of repeat dosing
 Effect of route of administration, vehicle, dose response
PRECLINCAL STUDIES - 2
36

 General Toxicity
 Acute
 Local
 Subacute and chronic
 Carcinogenicity
 Genotoxicity
 Gene mutations
 Chromosomal damage
 (DNA damage)
 Reproductive toxicity
 Pharmacokinetics
 Fertility
 Oraganogenesis
 Peri and postnatal development
 Additional studies
 In vitro Studies
37
 Generally cheaper than animal studies and more consistent (reproducible)
 May require animal tissues but generally reduces animal use
 Often adaptable for high throughput screening
 All have some, generally considerable, limitations in predictivity (also somewhat
true for animal studies)
 Irritation
 Eye
 Chemistry (strong acids or alkalies irritant)
 Excised living bovine cornea
 Cell culture
 Chicken egg chorioallantroic membrane
 Enucleated chicken eye
 Skin
 Simple chemistry
 eg strong alkalies are known skin irritants so no need to test further
 Keratinocytes in culture
 Skin organ culture
 Reconstituted human skin models
 With the exception of genotoxicity studies, in vitro studies will generally not be
important in establishing a PDE

BartCrofts Scientific Services Pty Ltd; [email protected]

 ANIMAL TOXICITY STUDIES
38

 Majority of toxicity data is obtained from animals, primarily rodents and dogs
 Monkeys and other species generally used sparingly
 Acute Toxicity Studies
 Used to establish starting doses for longer term studies
 An indication of the likely effects of acute overdose in humans,
 Usually by the oral, dermal and inhalational routes (for OH&S) and the pharmaceutical route of
administration
 Species:
 3 or 4 mammalian species of known strain (e.g. mice, rats & giunea pigs, rabbits , - dogs, & monkeys may also be used less
commonly)
 Key results/outcomes
 Minimum lethal dose and maximum non-lethal dose.
 Clinical signs of toxicity in relation to the onset of deaths,.
 Effects on body weight, food intake, water consumption.
 Identification of target organs for toxicity (possibly).
 Identification of cause of death (where possible).
 Will generally not be important for establishing a PDE but can give an indication of the likely hazard band
a drug will fall into

BartCrofts Scientific Services Pty Ltd; [email protected]

If LD50 at lower concentration for IV than for oral this implies that
metabolism is important for detoxification of this chemical
May also be a Cmax effect or bioavailability effect
If LD50 higher (less toxic) for IV than oral what does this imply ?
 Study Types
40

PHARMACOKINETICS

BartCrofts Scientific Services Pty Ltd;

[email protected]
 Pharmacokinetics
41

Ideal key data requirements

 Absorption and bioavailability
 For test species yielding pivotal NOAELS
 For humans
 For all relevant routes of exposure in each species
 Interspecies/strain differences
 Any relevance for cross species extrapolation

 Gender differences
 Linearity
 If absorption is saturated is this above or below the NOAEL
 Accumulation
 Does the drug build up, bioaccumulate, eg bisphosphonates in bone, and this potentially adverse
 Half life
 If a once daily medication (long half life) contaminates a thrice daily medication (short half life)
could this result in build up in the contaminant blood levels

BartCrofts Scientific Services Pty Ltd; [email protected]

 Animal versus human exposure
42

 Not all tissues/effects will be best measured by the same

exposure parameter
 mg/kg bw may be a better comparator for
 GIT effects likely to be due to purely local concentrations
 Effects on the liver and kidney - may be related to the total flux of drug rather than the
AUC
 Bone toxicity - osteoporosis drugs generally concentrate in the bone and levels are not
directly proportional to blood AUC values
 Exposure ratio = AUC or Cmax in animals / AUC or Cmax in
humans
 This may be determined at the:
 NOEL (No Observable Effect Level), and/or
 NOAEL (No Observable Adverse Effect Level), and/or
 LOEL (Lowest Observable Effect Level) for a given toxicity, and/or
 HD (highest dose tested)
 Safety margin = ratio of exposure at no-effect dose in
animals cf the exposure at the maximal human dose
 Study Types
43

REPEAT DOSE
TOXICITY STUDIES

BartCrofts Scientific Services Pty Ltd; [email protected]

 Repeat Dose Toxicity
44

 Identify most sensitive target organ(s) of toxicity

 NOAEL and LOAEL for these effects in the test
species and AUC at these doses (to estimate Human
Equivalent Dose HED)
 Likely to be the most difficult data to locate for older
drugs
 For newer drugs may be covered in and EPAR or FDA review
 Is there any evidence for this toxicity in human
clinical trials or post market review

BartCrofts Scientific Services Pty Ltd; [email protected]

 Group sizes
45

 The greater the number of animals/subjects per group the

greater the statistical power
 For small group sizes, (low statistical power) ideal is to have
pre-treatment (baseline) investigations
 Each animal then serves as its own control
 Number of groups also affects power (trend analysis)
 For some serious but rare pathologies no practicable group
size will ever have sufficient power to detect even substantial
increases in incidence (eg gliomas in rats)
 Dose selection can compensate for (or increase) statistical
power in some circumstances
 Increased magnitude of effect reduces the number of subjects to reliable
detect that effect
 REPEAT-DOSE TOXICITY STUDIES
46
 Species:
 Two mammalian species, one of which must be non-rodent.
 Rats are usually preferred for oral and inhalation, rabbits and mini pigs for dermal, dogs for oral and intravenous,
monkeys for biologicals...
 Pharmacokinetics, mode of action, species-specific factors etc. should be taken into account when choosing an
animal model for toxicity testing.
 Age
 Generally young healthy adult animals
 May use juvenile animals for products intended for children
 Or where toxicity specific to a life stage is of concern
 Duration
 Short term usually 2- 4 weeks
 Long term up to a year
 Chronic, lifetime
 Dose levels
 Usually three Sometimes (4) dose levels plus a control group.
• The highest dose should cause target organ or non specific toxicity
• The lowest dose should cause no toxicity (NOAEL), but for a drug is ideally sufficient to produce a
pharmacological/therapeutic effect and is associated with an AUC (ie systemic exposure) comparable to the maximum
anticipated human value.
• Doses usually escalate at approx multiples of √10 (3.2X)
 Group size
 10/sex/dose for rodents, 4/sex/dose for non-rodents in short-term studies.
 20/sex/dose for rodents, 4 – 6 /sex/dose for non-rodents in long-term studies (usually including recovery
subgroups).
 Satellite groups are also used for toxicokinetic testing.
 If the product is intended to be used clinically in only one gender (e.g. oral contraceptive), then use of the relevant
gender alone may be acceptable.
 Duration of repeat-dose toxicity studies to support
clinical trials
47

Duration of repeat-dose toxicity studies to support Phase I & II clinical trials in EU

Duration of clinical Minimum duration of repeat-dose toxicity studies

trials

Single dose 2-4 weeks 2 weeks

Up to 2 weeks 2-4 weeks 2 weeks
Up to 1 month 1 month 1 month
Up to 3 months 3 months 3 months
Up to 6 months 6 months 6 months
>6 months 6 months Chronic

You may be able to get some information about the short term
repeat dose toxicity studies from published reports on phase I
trials
 REPEAT-DOSE TOXICITY STUDIES
Observations

 Pre-treatment (baseline) values (especially important in wild-

caught large animals eg primates)
 Mortality and clinical signs of toxicity (appearance & behaviour)
 Body weights, food and water consumption
 Laboratory tests (Clinical Chemistry, Haematology etc) , ECG,
ophthalmoscopy and body temperature
 Post-mortem examinations (organ weight, gross and
histopathology)
 Reversibility of drug-related lesions (in recovery animals)
 Toxicokinetics (in satellite animals)
 Typical Parameters in 1- 3 Month Studies
Investigation Performed Parameters
General Observations
Clinical Observations Daily 49
Food & Water Consumption Weekly
Body Weights Weekly
Ophthalmoscopic Examinations Once pretest &
termination

Clinical Pathology
Urine Chemistry Pretest and 4 Weekly
Urinalysis Pretest and 4 Weekly Color, appearance, pH, protein, glucose, ketones, bilirubin, blood, urobilinogen, microscopic examination of
sediment
Sodium, potassium, chloride, calcium, phosphorus, creatinine clearance, volume
osmolality
Serum Chemistry Pretest and 4 Weekly Glucose, urea nitrogen, creatinine, alanine aminotransferase, aspartate aminotransferase, alkaline
phosphatase, gamma-glutamyl transferase, bilirubin, total protein, albumin, globulin, albumin/globulin
ratio, cholesterol, triglycerides, sodium, potassium, chloride, calcium (total)
Phosphorus, lactate dehydrogenase, testosterone (males only)

Hematology Pretest and 4 Weekly RBC, Hb, HCT, MCV, MCH, MCHC, Ret, Plat., total leukocyte count, differential leukocyte count,
Coagulation (PT, APTT)

Post Mortem
Investigations
Necropsy Gross pathology of all organs

Organ Weights & Adrenal Glands, Muscle - Biceps Femoris, Aorta, Thoracic Nerve, Bone, Ovaries, Pancreas, Bone Marrow,
Histopathology Parathyroid , Pituitary Gland, Brain, Prostate Gland, Cecum, Salivary Glands, Colon, Seminal Vesicles,
Duodenum, Skin, Epididymides, Spinal Cord Esophagus, Spleen, Eyes with Optic Nerve, Stomach,
Harderian Glands, Testes, Heart, Thymus, Ileum, Thyroid Gland, Jejunum, Tongue, Kidneys, Trachea,
Liver, Urinary Bladder, Lungs (plus Bronchi), Uterus (plus Cervix), Lymph Nodes ,Vagina, Lymph Nodes,
Mammary Gland

Ultrastructural Pathology Selected pathology

 Study Types
50

CARCINOGENICITY
STUDIES

BartCrofts Scientific Services Pty Ltd;

[email protected]
 Carcinogenicity
51

 Is the drug carcinogenic in animal studies

 Both rats and mice or just one
 Is there any indication of carcinogenicity in humans
 Clinical trials (very unlikely)
 Post market surveillance -
 If data is confined to animals, is the tumour type,
location & mechanism relevant to humans ?
 Prescriber information will usually provide adequate
data for this endpoint and indicate the conclusions of the
respective regulators as to the relevance of the findings if
any.
 Just use the regulatory conclusions
 Usually provides an indication of NOAEL or MOE calculation at
specific rat dose

BartCrofts Scientific Services Pty Ltd; [email protected]

 Carcinogenicity - long term study
52
 Internationally harmonised guidelines
 Species selection
 Rat, mouse
 Factors for species selection:
 Background tumor rates
 pharmacology
 repeat dose toxicity
 metabolism and toxicokinetics

 Duration – lifetime (24 months for rats,18-24 months for mice)

 Dose levels: generally 3, plus a control group
 Group size: 50 (desirable survival at termination 25/group)
 Complete necropsy and histopathology
 Carcinogenicity - long term study
53

 Dose level selection

 Toxicity MTD (maximum tolerated dose/Minimally Toxic Dose)
 Pharmacokinetic end points
 Pharmacodynamic end points
 Maximum feasible dose (5% in diet, local tolerance, practicality) - for low toxicity
substances
 Limit dose (1500 mg/kg/day for non-genotoxic drugs with human dose of < 500
mg/day)
 Other (case-by-case basis)
 Minimally Toxic dose (MTD) (More than mild toxicity invalidates the
study)
 produce at least minimum toxic effect over the course of the carcinogenicity
study
 Some toxicity necessary to establish “LOD”
 Must not cause alterations in physiological function which alter the animal's
normal life span or interfere with interpretation of the study
 not more than 10% decrease in body weight gain relative to controls
 Must not cause marked target organ toxicity
 No substantial alterations in clinical pathology parameters
 Carcinogenicity studies
54

 Evaluation of carcinogenicity studies

 tumour incidence and latency (time to onset)
 ie how many of each type and when did they appear
 statistical analysis
 dose-related trend
 historical control data
 pharmacokinetics and exposure
 mechanism of action: genotoxic or non-genotoxic
 If non-genotoxic,
 mechanism and threshold?
 relevance to humans?
 Assessing a non-genotoxic carcinogen
55

 To discount a non-genotoxic carcinogen finding in

animals consider:
 There should be a plausible non-genotoxic mechanism
explaining the animal carcinogenicity, supported by data or
published information.
 A threshold dose exists in animal studies, and the threshold
exposure level in animals is higher than the likely human
exposure.
 Tumours only occur in one species or one sex (sex specific
metabolism or mechanism?). The mechanism does not operate
or is weak in humans.
 There is extensive human data to indicate lack of
carcinogenicity in humans by that mechanisms.
 Non-genotoxic carcinogenicity
56

 Lymphoma
 Immunosuppression (cyclosporin, azathioprine)

 Relevance to humans: high incidence of lymphoma was found

in organ transplant patients,
 BUT strong risk/benefit

 SO real risk but acceptable to the patient group

 Non-genotoxic carcinogenicity
57

 Kidney tumours - binding to 2µ-globulin (specifically

synthesised by male rats)
 male rats only
 Not applicable to female rats or other rodents
 mechanism:
 binding to 2µ-globulin (present in lysosomes)
 globulin accumulates in the lysosomes of the proximal tubules
 Causes lysosome dysfunction
 Results in release of digestive enzymes
 Causes renal epithelial necrosis
 Leads to cell proliferation
 Results in tumours
 relevance to humans:
 species specific (male rats only)
 unlikely to occur in humans so disregard in risk assessment
 Study Types
58

GENOTOXICITY

BartCrofts Scientific Services Pty Ltd;

[email protected]
 Common Genotoxicity Tests
59

 Bacterial Reverse Mutation (Ames test)

 Gene mutation in mammalian cells (V79, CHO etc)
 Unscheduled DNA synthesis (DNA damage) – in vitro and/0r in vivo
 Sister chromatid exchange assay (SCE)
 Saccharomyces cerevisiae (yeast) (gene mutation or mitotic
recombination)
 Dominant lethal test (heritable genetic alteration assay)
 Mouse spot test
 Heritable translocation assay
 Drosophila (fruit fly) sex-linked recessive lethal mutation
 In vivo micronucleus assay in mice
 Genotoxicity
60

 The Genotoxicity of a compound is determined on a

weight of evidence basis from the results of a battery
of in vivo and in vitro tests
 ie no one test is usually sufficient in isolation to classify a
compound as genotoxic or not.
 Generally little or no value in assessing the
underlying studies for API PDE purposes
 The Prescriber Information (authorised label) will
usually provide adequate data for this endpoint and
indicate the conclusions of the respective regulators
as to the relevance of the findings if any

BartCrofts Scientific Services Pty Ltd; [email protected]

 Study Types
61

REPRODUCTIVE AND
DEVELOPMENTAL TOXCITY

BartCrofts Scientific Services Pty Ltd;

[email protected]
 Reproduction and Development
62

 Does the drug affect;

 male or female fertility
 Foetal survival
 Foetal development
 Neonatal development
 The presence or absence of relevant effects will be noted
in the authorised prescriber information but there may
not be any information on the threshold doses.
 Will often need to look at post market studies to gauge
the relevance of animal data in clinical practice
 Clinical trials will generally be of no value (exclude
pregnant women)

BartCrofts Scientific Services Pty Ltd; [email protected]

 Reproductive & Developmental toxicity

 ICH and OECD protocols differ

 ICH approach recognises 6 distinct phases and allows greater flexibility in
study designs but essentially covers the same ground as OECD
 OECD uses two stages – reproduction and development
 Developmental Toxicity
 Treatment of pregnant females only, from time of implantation until just before birth
 Generally Rat and Rabbit, about 20/sex
 Pups are removed by caesarean section and examined
 Reproductive toxicity
 Treatment of 2 generations from before mating through to birth, lactation, weaning,
maturation of pups, mating ….etc
 Treatment of males for 2-4 weeks prior to mating, during cohabitation, until termination
when the female is pregnant
 Treatment of females for 2 weeks prior to mating, during cohabitation, through birth,
lactation and weaning
 Then treatment of 4 pups from each litter at each dose right through until they give birth
and wean their pups
• ie 2 whole generations
 Developmental toxicity endpoints
64

• Pre-implantation loss
• Early resorption (post implantation death)
• Late death
• Malformations/variations/anomalies
• Growth retardation, size & weight
• Sex ratio

BartCrofts Scientific Services Pty Ltd; [email protected]

 Study Types
65

CLINICAL DATA

BartCrofts Scientific Services Pty Ltd;

[email protected]
 First Dose in Man - MABEL
66

 Minimum Anticipated Biological Effect Dose

 Combines a pharmacological with a toxicological estimate of first dose
in man and uses the lesser of the two
 Pharmacology
 Estimate human “Minimal Anticipated Biological Effect Level”
(MABEL)
 justify based on pharmacology
 adjust for anticipated exposure in man
 include anticipated duration of effect
 adjust for inter-species differences in affinity / potency
 Toxicology
 Determine “No Observable Adverse Effect Level” (NOAEL)
 Convert NOAEL to a “Human Equivalent Dose” (HED)
 adjust for anticipated exposure in man
 adjust for inter-species differences in affinity / potency
 Apply >10-fold safety factor
 Phase I Clinical Trials

 Open Label and small numbers 10-100 healthy participants (often men
between 18-40)
 Low single dose initially based on the NOAEL from the most sensitive animal
species (or MABEL see later)
 Convert NOAEL to Human Equivalent dose (HED) using body weight scaling
(W0.67) which compensates for surface area and comparative metabolic rate
between smaller and larger species
 Consider
 Differences in ADME
 Experience with the chemical class
 Cross species expression of relevant receptors

 Divide by uncertainty factors to account for interspecies variability, steep

dose response curve etc (usually at least 10 to 100)
 Progressively increase if no evidence of toxicity
 Primarily looks at basic pharmacokinetics, safety, tolerability
 Lasts months to 1 year
 May provide an excellent estimate of a no effect dose in humans
 Phases II & III & IV

 Larger numbers of patients as subjects over periods

of years
 Randomised, controlled, blinded
 Efficacy and safety
 Phase IV is post market surveillance
 e.g. Vioxx (rofecoxib) was used for RA for 5 years before a post
approval clinical trial revealed increased risks of heart attack
and stroke
 Withdrawn from the market
 Key Points
69

 Prescriber information
 Carcinogenicity and genotoxicity are usually well covered
 Reproductive effects often not covered in sufficient detail to yield
POD
 Repeat dose toxicological effects generally not well covered
 Side effects, exaggerated pharmacology effects usually well covered
 Published papers
 Phase I trial reports may provide information on MABEL or human
pharmcokinetic endpoints (EC50, IC50) that are a useful source of
potential POD values
 Phase II & III and other studies may yield data on off target effects,
reproduction
 EPAR, US FDA reviews
 May be required for Pharmacokinetic data and PODs for repeat dose
effects in animals

BartCrofts Scientific Services Pty Ltd; [email protected]

 Not What You Know
70

 Toxicology covers every aspect of biological science

so no one can be truly an expert in the traditional
sense
 The key is;
 Asking the right questions
 Knowing where to get the answers
 Accurately interpreting the available data
 Balancing caution with pragmatism

BartCrofts Scientific Services Pty Ltd; [email protected]

 Got all that ?
Time for a break !
71