tolerability may be dose-dependent, and this can
only be assessed when studied in a systematic
fashion. Lastly, most drugs are just one of a series
of compounds which share closely related proper-
ties in preclinical testing. It is impossible to know
which of these is the most promising, when only
one has been tested.
Assuming that reasonable tolerability, reason-
able understanding of pharmacokinetics and (pre-
ferably) a relevant pharmacodynamic effect has
been observed in normal volunteers (see Chapter
8), then the first task is to reassess all of these
in a relevant disease state. This is slower and uses
more patients than the professor’s uncontrolled
observations. But at the end of a small number of
such small studies, there ought to be good informa-
tion about the feasibility of a pivotal clinical
trials program and, if not, then the feasible
course corrections (e.g. alternative indications).
Note that one such course correction may be ceas-
ing to develop the drug, and switching to another
member in the series. Arguably, the appropriate
‘killing’ of drugs is the most valuable thing that a
phase II program can accomplish before too much
time and money has been wasted.
When choosing a clinical trial design
(Table 9.2), economic factors include numbers of
patients, time that will elapse, drug supply and total
cost. Although these economies are important and
relevant in all design choices, they should also
be factored against the end points that may or
will be measured. The relevance of an end point
and its sensitivity to detect a drug-related effect
may be primarily dependent upon the duration of
patient exposure. For example, a short period of
observation is unlikely to detect a difference in
time to next seizure in a study of antiepileptic
drug with an add-on design in patients who are
only moderately disabled by epilepsy. On the
other hand, the identification of a PK interaction
between a new and an established therapy in the
same populationmay only requirevery short obser-
vation periods.
There are several common classes of study
design. These classes apply to almost all phases
of drug development. No list of trial designs can beTable 9.2 Basic trial designs and the factors that are suited and unsuited to each
Trial type Factors suited Factors unsuited
Parallel-group, single Episodic disease Rare disease
treatment Imperfect placebo matching
Blinding difficult (e.g. surgical
procedures, psychtropic drugs)
Parallel-group, chronic Stable disease state Unethical to use active comparator
treatment or placebo
Crossover with washout Stable disease state Untreated washout not ethical
Ethical to use placebo after active
Sequential Rare disease Complicated tolerability profile
Homogeneous disease state Many concomitant disease factors
Urgent need to save life
‘Nof 1’ Stable disease state Few or no feasible alternative
therapies
‘Large simple’ Very common disease Tolerability issues not closely related
Easily measured end-points to efficacy variable
Well-understood drug
Open label Tolerability issues only Spontaneous adverse event frequency
high
Within-patient Stable disease state Drug tolerance
dose ranging Intolerable high initial dose
Combination therapy A priorireason to expect favorable Unethical to use single therapy
drug interaction9.9 COMMON PHASE II/III STUDY DESIGNS 107