Pharmacodynamic measurements in healthy volunteers or patients may be used for establishing equivalence between two pharmaceutical products. This may be necessary if the drug and/or its metabolite(s) in plasma or urine cannot be determined quantitatively with sufficient accuracy and sensitivity. Furthermore, pharmacodynamic studies in humans are required if measurements of drug concentrations cannot be used as surrogate end-points for the demonstration of the efficacy and safety of the particular pharmaceutical product; this applies, for example, to topical products where it is not intended that the drug should be absorbed into the systemic circulation.

If pharmacodynamic studies are used, the conditions under which they are performed must be as rigorously controlled as those of bioequivalence studies, and the requirements of the guidelines for good clinical practice (GCP) for trials on pharmaceutical products (5) must be satisfied.

The following requirements must be taken into account in planning, conducting and assessing the results of a study intended to demonstrate equivalence by means of measurements of pharmacodynamic drug responses:

- the response measured should be a pharmacological or therapeutic effect relevant to the claims of efficacy and/or safety;

- the methodology must be validated for precision, accuracy, reproducibility and specificity;

- neither the test nor the reference product should produce a maximum response in the course of the study, since it may be impossible to distinguish differences between formulations given in doses that produce maximum or near-maximum effects; investigation of dose-response relationships may be a necessary part of the design;

- the response should be measured quantitatively under double-blind conditions and be recordable by means of a suitable instrument on a repetitive basis to provide a record of the pharmacodynamic events which are substitutes for plasma concentrations; where such measurements are not possible, recordings on visual analogue scales may be used, and where the data are limited to qualitative (categorized) measurements, appropriate special statistical analysis will be required;

- non-responders should be excluded from the study by prior screening, and the criteria whereby responders and non-responders are identified must be stated in the protocol;

- where an important placebo effect can occur, allowance for this effect should be made in the study design by including placebo treatment as a third phase in that design;

- the underlying pathology and natural history of the condition should be considered in the study design, and information on the reproducibility of baseline conditions should be available;

- where a cross-over design is not appropriate, a parallel group study design should be chosen.

In studies in which continuous variables can be recorded, the time course of the intensity of the drug action can be described in the same way as in a study in which plasma concentrations are measured, and parameters can be derived which describe the area under the effect - time curve, the maximum response and the time when that response occurred.

The statistical methods for the assessment of the outcome of the study are, in principle, the same as those outlined for bioequivalence studies. However, a correction should be made for the potential non-linearity of the relationship between the dose and the area under the effect - time curve, based on the outcome of a dose-response study. However, it should be noted that the conventional acceptance range as applied for bioequivalence assessment is usually too large and therefore not appropriate; for this reason, it should be defined on a case-by-case basis and described in the protocol.