Write a human PD or PK/PD study report

How should a human PD or PK/PD study report be written?

Define the response, exposure metric, timing, population and analysis that connect the measured drug effect to the study objective. Explain uncertainty and possible confounding. Distinguish a biomarker response or exploratory exposure association from established clinical benefit and from a justified dosing conclusion.

Before you begin

Studies primarily assessing pharmacodynamic effects in humans. Studies contributing an efficacy or safety demonstration may belong in 5.3.5 even when they use PD endpoints.

What you will prepare: A report that preserves endpoint meaning, measurement timing, analysis assumptions and limits of clinical interpretation.

Separate pharmacological effect from therapeutic benefit

Describe the measured response and its intended role: mechanism, biomarker, dose selection, short-term effect or another pharmacological question. M4E distinguishes healthy-subject PD/PK-PD studies at 5.3.4.1 and patient studies at 5.3.4.2. When the study is part of the efficacy or safety demonstration, examine 5.3.5 instead; endpoint terminology alone does not settle placement.

Collect the protocol, endpoint/assay documentation, dose and sample records, response assessment records, analysis plan and final outputs. A biomarker's relationship to clinical benefit needs evidence. If that relationship is unresolved, describe the biomarker change without renaming it a clinical improvement.

Make the exposure-response relationship auditable

Define the exposure metric, response scale, baseline, assessment windows and analysis population. Explain why the timing can capture the proposed relationship, including delayed or persistent effects where relevant. Distinguish prospective dose-response comparisons from exploratory associations based on observed concentrations.

For healthy subjects, explain the nontherapeutic objective and relevance of the model or challenge. For patients, describe disease course, background treatment and other factors that can affect response independently of exposure. Present uncertainty and sensitivity to analysis assumptions. If a model is used, supply its structure, development and evaluation rather than only a smooth fitted curve.

Challenge a plausible exposure-response story

Fictional exercise: patients who tolerate treatment remain on a higher dose, and a retrospective plot shows better outcomes at higher exposure. The reviewer should examine dose-selection and disease-related confounding before calling the relationship causal. Identify the analysis as exploratory and show what additional evidence supports the interpretation.

Adapt the CSR organization to communicate the actual PD task; E3 is not a rigid section-filling template. Reconcile biomarker units, timepoints and populations with the clinical pharmacology summary. If the report supports a proposed regimen, keep the evidence and limitations next to that proposal rather than moving them to an unrelated appendix.

Separate observation, association and clinical interpretation

Use a claim map before writing the discussion. Each step requires its own evidence; a smooth curve does not supply the missing links.

Writing exercise: the same response scale is measured at different postdose times in two cohorts. Before describing an exposure-response difference, establish whether timing or cohort characteristics could explain it. Obtain the analyst's comparison and preserve limitations rather than merging the points into one apparently uniform relationship.

Use the population PK guide when exposure estimates come from a model. Cross-reference the clinical efficacy summary only for the clinical claims the evidence supports; a pharmacodynamic signal should remain identifiable as such.

Frequently asked questions

Sources and revisions

Requirements, source recommendations and editorial preparation advice have different roles. Review the scope and revision of the source you use.

Related guides