US Food and Drug Administration Recommendations on the Use of Surrogate Measures as End Points in New Anti-infective Drug Approvals.

Importance: Regulatory and scientific guidelines stipulate that indirect, surrogate measures of patient benefit, such as a change in microbial culture status, should be used as primary end points only in pivotal trials of chronic conditions that are serious or life threatening and when the experimental therapy is expected to offer substantial benefit compared with available therapy. However, many recent US Food and Drug Administration (FDA) anti-infective drug approvals for acute and/or non-life-threatening diseases have been based on pivotal trials using surrogate measures as primary end points rather than clinical outcomes, such as symptom resolution or survival. Objectives: To review FDA recommendations for primary end points in pivotal trials of new anti-infective drugs and assess the concordance of those recommendations with the regulatory and scientific conditions for the appropriate use of surrogate measures as primary trial outcomes. Evidence Review: All guidance documents for antimicrobial drug development hosted on the FDA website were searched in November 2017; the search was updated in June 2018. For each document, 2 reviewers independently extracted data on the recommended primary end points for a pivotal or phase 3 trial. Findings: Twenty-two FDA guidance documents met the inclusion criteria, which included recommendations for primary end points in pivotal clinical trials in 27 infectious disease indications. Twenty-one of 27 indications recommended surrogate outcomes as either the sole primary end point or as components of composite end points. None of the recommendations for the use of surrogate measures matched the regulatory and scientific conditions favoring indirect outcomes in place of clinical outcomes. Conclusions and Relevance: The FDA guidance documents for developing new anti-infective agents frequently recommend indirect measures of patient benefit, rather than direct measures of patient benefit, as sole primary end points or components of primary end points. Existing guidance documents should be updated and revised to recommend appropriate clinical outcomes consistent with general scientific and regulatory parameters.