A phase 3 trial protocol details the objectives of a study, how it will be carried out, and what data analyses are planned. Protocols should be controlled documents; version numbered and dated using a formalized convention.

Development of phase 3 trial protocol is conceptually similar to protocol development for phase 2 trials.

Defining a regimen for a phase 3 trial

Investigators, regulators and investors will all expect a high degree of certainty that the regimen that is chosen for a phase 3 trial, which will be the one used in clinical practice if the treatment is approved, will be the right one.

The evidence to support the regimen may come from a combination of pharmacokinetic/pharmacodynamic (PK/PD) data and the clinical results from phase 2 studies. Increasingly, however, phase 2 studies may be bypassed. In this circumstance, the PK/PD package must be extensive and provide all the evidence for the activity and efficacy of the new compound.

Dose and dose justification: The same approaches as described for phase 2 are used, but more work may be required to ensure the conclusions are robust and to assign and quantify any risk.  The following may need to be considered:

  • Further dose-fractionation studies may be required, potentially complemented by data from in vitro and in vivo model systems.
  • To provide confidence in the relevant pharmacodynamic target, multiple strains of pathogen may need to be studied. While there is no definitive number, it is important to include all relevant genera and species, and to take into account resistance mechanisms. Pathogens with a range of antibiotic susceptibilities (minimum inhibitory concentration (MIC) values) should be studied, including pathogens with MICs differing from wild-type.
  • Robust target attainment analyses will require reliable estimates of PK variability in the target population. These data may be obtained from PK studies embedded in phase 2 trials or from “run in” studies (carried out after recruitment has begun but before randomization has started).

Site of infection and proposed clinical indication:  PK/PD analyses should be aligned with the proposed clinical study and ultimate clinical indication. For example, a specific PK/PD programme that examines the utility of drug for pneumonia is required if the intended clinical indication is community-acquired pneumonia (CAP) or hospital-acquired pneumonia/ventilator-associated pneumonia (HAP/VAP).

This might require use of specific experimental model systems and clinical studies (e.g. to estimate the partitioning of a drug into the epithelial lining fluid of volunteers).

Preliminary breakpoint setting: Management of patients receiving an investigational drug requires estimation of a breakpoint, the concentration of an antibiotic likely to be effective in a particular treatment context. Extrapolation of PK/PD data can be used to provide a preliminary breakpoint value to guide clinical decision-making (“PK/PD bridging”).