Microbiologic surrogate end points in clinical trials of infectious diseases: example of acute otitis media trials

Abstract

Clinical outcomes that measure how patients feel, function, or survive are the most important and relevant outcomes of therapy in clinical trials and in clinical practice. Surrogate end points, which do not directly measure clinical benefit to the patient, may function as substitutes for clinical end points in clinical trials. Such surrogates are attractive as they may allow measurement of outcomes earlier in time or with a smaller sample size than with clinical outcomes. Microbiologic biomarkers, such as culture results at a specific time after start of therapy, or pharmacodynamic analyses of the effect of drugs on organisms often are proposed as surrogate end points in clinical trials of therapies for infectious diseases. However, evaluation of biomarkers as surrogate end points poses distinct challenges, and only a few biomarkers have been useful replacements for clinical end points. Evaluation of biomarkers as potential surrogate end points first requires an understanding of the differences among measurements of the cause of a disease, risk factors for outcome, and measurements of treatment effects. We will discuss the definitions of clinical and surrogate end points and the reasons why surrogate end points may not predict the true clinical benefit of therapies. We will use the example of the biomarker of microbiologic outcomes from tympanocenteses performed during therapy as the sole measure of clinical effectiveness in clinical trials of acute otitis media to illustrate the challenges in evaluating biomarkers as surrogate end points.