Difference between biomarkers and clinical endpoints

Last month I met some people Sanwa Biotech, who told me about their medical device ALiA and its possible applications in the research with customisable biomarkers. I did not know what are biomarkers, so I googled and read three research papers and wikipedia, trying to figure out what is that. Here is my recall.

So essentially there are three important components to understand: biomarkers, surrogate endpoints and clinical endpoints. Those are mostly used in clinical trials. 

Biomarkers are characteristics that measure how patients feel, function and survive. Endpoints are used in clinical trials to measure an outcome that represent benefit to the patient. Clinical endpoints are outcomes of patients condition that give direct clinical benefit - survival, decrease of pain or absence of disease. And surrogate endpoints substitute clinical endpoints and can predict clinical benefit. For example, reduction of tumor in size is commonly used surrogate endpoint to predict longer survival of a patient with cancer, while testing new treatment drug.  

For example, we assume that high level of cholesterol leads to heart attacks. Clinical end point would be a heart attack, surrogate end point would be a level of cholesterol in blood. The question is how significant is a correlation between clinical end point and its surrogate that preludes it. If it is the only one or major one - then it can be used as surrogate for the purpose of the trials (using clinical surrogates shorten time and cost for drug development and approval). If connection is not validated - then level of cholesterol would be one of biomarkers.