Principles and Practice of Pharmaceutical Medicine

structure–activity similarities to known carcino-
gens, evidence of preneoplastic lesions in
repeated-dose nonclinical studies, or long-term
tissue sequestration of the drug. Another guideline
(Federal Register, 1 March 1995) addresses the
complex issue of the selection of doses for these
studies; this responds to much criticism of the prior
recommendation to use the maximum tolerated
dose (which had been suggested by the National
Toxicology Program; Haseman and Lockhart,
1994). The current ICH guideline recommends a
high dose causing up to 25-fold greater plasma
AUC in rodents compared to the AUC in humans
at steady state. A subsequent amendment to this
guideline (Federal Register, 4 December 1997)
adds a further proviso that the highest dose in a
carcinogenicity study need not exceed 1500 mg
kg
^1 per day when (a) there is no evidence of
genotoxicity and (b) the maximum recommended
human dose is no more than 500 mg per day. The
basis for species selection, circumstances needing
mechanistic studies and exploitation of pharmaco-
kinetic information in carcinogenicity testing is
described in yet another guideline (Federal Regis-
ter, 21 August 1996).
Modern protocols for carcinogenicity studies
have changed little since they were first established
in the early 1970s. In recent years, the use of mice
(historically the second of the two required species
in addition to rats) has come under scrutiny
because they may be inappropriate models, with
unusual sensitivity to certain classes of chlorinated
hydrocarbons. The most recent ICH guideline
(Federal Register, 21 August 1996) allows for the
option of using transgenic mice and study designs
of somewhat shorter duration.
Of growing importance is the interaction of fac-
tors that are critical to a successful toxicology
program. For example, if a transgenic mouse
model is selected, then the choice of strain is
important and may depend on whether the drug is
non-genotoxic (TG.AC model) or genotoxic (p53
model). Metabolic and pharmacokinetic data are
important to ensure that the selected models handle
and metabolize the drug in a fashion at least reason-
ably similar to humans and may vary for the
same drug according to the toxic effect of interest.
Perhaps the most important factor is the relevance

of the doses selected to those in humans. Although

this has been a subject of controversy for years, a

recent ICH guideline allows for the use of toxico-

kinetic measurements, and states that doses that

produce an AUC in the carcinogenicity model that

are 25 times that seen in humans at steady state may

no longer have to be used under some circum-

stances. A recent review of the status of carcino-

genicity testing (Reno, 1997) addresses the many

factors that should be considered in a carcinogeni-

city program.

Special studies

It is not uncommon in drug development programs

for specific toxicities to be uncovered. In most

cases, additional studies are then carried out that

will attempt to elucidate additional information

with regard to the mechanism of the effect. For

example, the identification of a non-specific beha-

vioral effect (e.g. tremors and/or convulsions) may

trigger the performance of a neurotoxicity study,

which includes an exhaustive evaluation of the

potential effects on the central and the peripheral

nervous tissues. The identification of an effect on

reproduction may warrant the performance of

detailed studies to identify the specific mechanism

or phase of the reproductive cycle that is affected.

In-depth metabolic studies may prove that the

effect is related to a metabolite in animals that

has no relevance to man, and prevent the abandon-

ment of an otherwise promising drug. It is rare that

a drug development program does not involve

some type of special study.

6.8 Product licence/new drug

application requirements

Format and content of the application

Although differing in format for each application,

an integrated summary that interrelates the phar-

macology, pharmacokinetic and toxicology study

information, and what significance the data has to

human safety, is paramount (Pecket al., 1992). A

72 CH6 NONCLINICAL TOXICOLOGY