As for phase 1 studies, a phase 2 trial protocol must be developed that describes in detail the objectives of a study, how it will be carried out, and what data analyses are planned.

Protocols (and many other documents produced as part of a trial) should be controlled documents; version numbered and dated using a formalized convention.

Protocols should include as much relevant detail as possible, to ensure effective review by regulatory authorities. A range of organizations have produced guidance on the key elements of a protocol. The SPIRIT initiative has developed widely used guidance on what should be included in a clinical trial protocol, as well as tools to aid protocol development.

As for phase I trials, international guidelines such ICH E6 (R2) detail the sections that should be included in a phase 2clinical trial protocol. ICH E6 (R2) guidance provides a unified standard for the European Union, Japan and the USA to ensure mutual acceptance of clinical data by the regulatory authorities in these regions.

ICH E6 (R2) guidance includes a list of the essential documents required before, during and after completion of a trial. Essential documents are needed so that the conduct of a trial, and the quality of the data produced, can be evaluated by regulatory authorities and their advisers. They demonstrate the compliance of the investigator, sponsor and monitor with GCP standards and all applicable regulatory requirements.

In vivo activity: Evidence is also required of in vivo activity against target pathogens causing the indications for which patients will be enrolled in the clinical study.

Defining a regimen for a phase 2 trial

A critical aspect of a phase 2 trial is the appropriate dose and dosing interval. These can be defined through the use of a range of pharmacokinetic/pharmacodynamic (PK/PD) methods:

Dose fractionation studies: A set quantity of drug can be delivered via a range of different-sized doses, with differing dosing intervals, in in vitro and in vivo experimental systems, to determine impacts on microbiological outcomes.

These studies can be used to derive a range of pharmacodynamic indices that provide a measure of exposure – maximum concentration (Cmax), ‘area under the curve’ (AUC, a measure of cumulative exposure over time), and the proportion of time when free plasma concentrations of a drug are greater than a pathogen’s susceptibility threshold (minimum inhibitory concentration, MIC) (fT>MIC).

Drug concentrations are generally expressed in terms of free (f) concentrations, to take account of a drug adhering to bloodstream proteins and therefore not being able to act on pathogens. This requires detailed protein-binding studies, undertaken by a specialist laboratory.

As well as guiding choice of dose and dosing interval, these studies help to identify the best pharmacodynamic index to use to assess the likely efficacy of a drug.

Magnitude of the pharmacodynamic index: As antibiotic susceptibility is likely to vary across strains, experiments with multiple strains are required to estimate the magnitude of the relevant PK/PD index that results in microbiological outcomes of clinical relevance – i.e. what value of fT>MIC or fCmax is associated with bacterial killing.

The desired microbiological outcome varies by indication. For complicated urinary tract infections (cUTI), for example, halting of bacterial multiplication (stasis) in a murine thigh infection model is generally used. For pneumonia, the key outcome measure is the killing of bacteria (assessed by measuring changes in bacterial numbers on a logarithmic scale).

As well as impacts on bacterial growth and survival, it may also be important to identify drug exposures that suppress the emergence of resistance.

Target attainment analysis: Probability of target attainment (PTA) is a key variable, providing a measure of the likelihood that a chosen measure of exposure exceeds the value required for an efficacious response against an infection.

This analysis requires expertise in PK/PD modelling and integrates information from PK/PD, clinical pharmacology and clinical microbiology studies:

  • A population PK model is developed to simulate the distribution of drug exposures that will be expected when a population of patients receives a given regimen.
  • Across a range of MIC values, the proportion of the total number of simulated patients in which drug exposure reaches target values is calculated.
  • As the distribution of MICs is known from clinical microbiology studies, the likelihood of therapeutic success for a planned regimen can then be calculated. This is referred to as the fractional target attainment value.

High target attainment values (>80–90%) provide evidence that the selected dose and dose interval is likely to be efficacious because the number of patients that will suffer exposure dependent therapeutic failure is minimized.