Assuming positive results are obtained in a phase 1 trial, the next objective is to confirm safety in patients and generate data on efficacy. Conventionally, this involves progressively larger phase 2 and phase 3 trials. However, developers now have a range of options for the next stage of clinical evaluation:

  • A conventional phase 2 trial in patients (typically 150–300).
  • An adaptive phase 2/3 trial: a larger trial that can be adapted while in progress (in carefully defined, pre-set ways) in light of data obtained.
  • Immediate progress to a phase 3 trial, skipping phase 2: this approach has been followed for drugs addressing a high unmet need, but is high risk and rarely used.

Each approach has its advantages and disadvantages.

Approach Advantages Disadvantages
Phase 2 study ·     Provides an estimate of efficacy and safety in patients in a specific indication, decreasing risk
prior to investment in a large phase 3 trial.·     Provides an opportunity to evaluate the potential impact of factors that could affect efficacy
(known as covariates), such as physiological factors or other drugs taken at the same time; this
can be used to refine the inclusion/exclusion criteria for phase 3 trials.·     Any required longer-term toxicology studies can be conducted in parallel.

·     Phase 2 PK/PD data can inform dose selection for phase III study.

 

 ·     Protocol design must be prospectively defined. Minimal flexibility to modify the study without a
protocol amendment.·     May be associated with increased cost and longer duration of a development programme.
Phase 2/3 adaptive design ·     Pre-specified modifications to the study design or statistical analyses can be made while the
trial is ongoing without requiring a protocol amendment.·     Study can be stopped early for futility (lack of beneficial effect) or efficacy, which may
reduce the cost and size of a study.·     If sample size adjustment is planned, it is more likely that statistical power will be maintained,
increasing the likelihood of statistically significant results.

·     Seamless designs allow a more rapid transition from phase 2 to phase 3.

 

 ·     Significant planning is required to define potential modifications, decision rules, and
statistical methodology.·     Requires multiple interim analyses to evaluate the need for each possible adaptation. These must
be carefully planned to avoid the risk of unblinding as the study is modified.·     More complex statistical methodology is required: a statistician expert in adaptive design must be
involved in the planning, conduct and analysis of the trial.

·     Due to time and cost of ongoing data review and multiple interim analyses, may not result in a
shorter, smaller or less expensive study.
Direct to phase 3 (skip phase 2) ·     May shorten overall development timeline and lower costs. ·       Significantly increases risk of failure since no efficacy or safety data in patients are
available until the phase III trial is completed.·       Requires a large upfront investment to characterize a compound prior to phase III study,
including extensive phase I data to define potential factors affecting efficacy (e.g. age, renal or
hepatic impairment) and comprehensive PK/PD data to optimize the dose. 

Comparison of three post-phase 1 approaches and their advantages and disadvantages

Phase 2 Phase 2/3 adaptive design Direct to phase 3 (skip phase 2)
Study design Relatively small and homogenous population (~150–300 patients) Larger than a typical phase 2 study if adaptation is designed to combine phase 2 and 3 in one trial.
Sample size about the same as a phase 3 trial (but depends on adaptations planned)		 Has been used in a few instances for drugs to address high unmet need
Advantages Provides an estimate of efficacy and safety in patients in a specific indication, decreasing risk prior
to investment in a large phase 3 trial		 Pre-specified modifications to the study design or statistical analyses may be made while the trial is
ongoing without requiring a protocol amendment, providing greater flexibility		 May shorten overall development timeline and decrease cost
Can be used evaluate the potential impact of covariates and drug–drug interactions to refine
the inclusion/exclusion criteria for phase III trials		 Potential to stop the study early for futility or efficacy: may reduce the cost and size of
the study
Any required longer-term toxicology studies can be conducted in parallel If sample size adjustment is planned, increases assurance that statistical power will be
sufficient when the study ends increasing the likelihood of success
Phase 2 data can further refine PK/PD knowledge to strengthen the dose prediction for phase
3		 Seamless designs allow a more rapid transition from phase 2 to phase 3
Disadvantages Protocol design prospectively defined.  Minimal flexibility to modify the study without a
protocol amendment		 Significant advance planning is required to define potential modifications, decision rules
and statistical methodology.		 Significantly increases risk since no efficacy or safety data in patients are available
before the completion of phase 3 study
May be associated with increased cost and longer duration of a development programme Requires multiple interim analyses to evaluate the need for each possible adaptation. Must
be carefully planned to avoid the risk of unblinding related to study modifications		 Requires a large upfront investment to characterize a compound prior to phase 3, including:
comprehensive microbiology data; extensive phase 1 data on potential covariates (age, renal or hepatic
impairment); and comprehensive PK/PD data to ensure an optimal dose for phase III study
More complex statistical methodology required: a statistician with expertise in adaptive
design must be involved in the planning, conduct and analysis of the trial		 If phase 1 trials are not sufficiently comprehensive, there is significant risk of failure.
Phase 3 registration studies are to demonstrate efficacy/safety for approval
Due to time and cost for ongoing data review and multiple interim analyses, may not result
in a shorter, smaller, or less expensive study