Multiple aspects of a phase 2 trial need to be defined.

Indication

The specific indication (target pathogen(s), site of infection, health condition) being targeted should be defined. It should be consistent with the indication to be targeted in a phase 3 trial and in clinical practice, if approved. 

Blinding

A study could be double-blind (neither investigator nor patient knows what treatment they are receiving), single-blind (only the patient is unaware of treatment being given) or open label (both patient and researcher are aware of treatment being given. To avoid the risk of bias, the highest degree of blinding possible should be adopted.

Control group

A control group could be given a placebo or a comparator drug.

  • Placebo-controlled trial: Although there are some infections for which the FDA requires placebo-controlled trials, including sinusitis and bronchitis, in general a placebo control is not acceptable for any serious infection for which effective therapy is available. Patients in a control group must receive a therapy of demonstrated efficacy, or one that is likely to be efficacious if a licensed treatment is not available.
  • Comparator-controlled trial: Safety and efficacy of an investigational drug are compared with those of an existing drug or treatment recommended for the targeted indication. The dose, dosing regimen and infusion rates (for intravenous drugs) of the comparator should be as specified on its label.

Dosing regimen

With adequate pharmacokinetic/pharmacodynamic PK/PD data, it should be possible to define a single dose that has a high probability of target attainment (PTA), the exposure required for efficacy. However, there are situations where multiple doses may be evaluated (dose ranging), for example when the preclinical exposure/response relationship is not clear or PK/PD data are not available (e.g. for non-traditional therapies that are not directly antibacterial).

Sample size

A phase 2 study should be sufficiently large to provide an estimate of efficacy in a narrowly defined set of patients, and to determine common short-term side effects and potential risks. In general, phase 2 studies have a sample size between 150 and 300 patients. 

Clinical microbiology

Samples must be collected from all patients, pathogens identified, and in vitro susceptibility determined for all isolates. All isolates should be sent to a central microbiology laboratory, which is selected before the study starts. The identification and susceptibility data from the central lab are used for the microbiologic assessments for the clinical trial. Susceptibility testing at the trial site itself (disk testing) is desirable but not essential. Molecular tests may be used to identify isolates with specific resistance determinants of interest.

Pharmacokinetics (PK)

An appropriate sampling scheme should be defined in advance of a study. Data on study drug exposure in patients is important to refine the population PK model. Information from phase 1 studies can be used to identify maximally informative sampling times, so that PK characteristics can be assessed using the fewest samples. Generally, three to four samples are taken through the dosing period.

Selection of the treatment regimen

The nature of an indication will affect treatment strategy and choice of comparator. For example:

On the basis of this information, it can be determined whether an investigational drug is sufficient as monotherapy or if combination therapy is required. For example, if an antibiotic has activity against only Gram-negative pathogens, regimens would need to include metronidazole to treat cIAI and vancomycin or linezolid to treat HAP/VAP. Similar considerations would help to define an appropriate comparator.