Nowadays, thousands and thousands of biomarkers are on the market to measure in human blood plasma (in addition to other markers in (for example) urine, hair, breath, et cetera). Biomarkers become more and more valuable, because they can at an early stage be used to predict (at least to some extent) what the clinical effect may/might be of the drug. 

An important aspect to realize is that drugs may work, but not help. What is meant is, that a drug may activate specific pharmacological pathways, but that the clinical outcomes (feeling, survival rate, etc.) are not affected. The effect is measurable on a pharmacological level, but the changes for the patient remain unnoticible.

(Note: You also have drugs that help, but do not work, e.g. placebo drugs).

The same is true for biomarkers. Many biomarkers demonstrate changes in the human body, but which markers are (from a safety of efficacy perspective) relevant for a clinical trials remains to be discussed. Choosing the correct biomarkers for answering the correct scientific questions is a skill in itself. One should thereby realize the differences between biomarkers, clinical markers (or clinical endpoints), and surrogate markers/endpoints:

• Biomarker (Biological marker): Indicator for (ab)normal biological process/response that can be objectively measured and evaluated
• Clinical endpoint: Characteristic that reflects the patients feeling/functioning/survival
• Surrogate endpoint: Biomarker that substitutes clinical endpoint

Note that not always proper clinical endpoints can be measured in a study. For example in a study that investigates the effect of long-term effect of daily intake of a drug to prevent hospitalization (e.g. use of statins to prevent cardiovascular disease). This could take years (or even tens-of-years). A clinical trial of such a long period is of course not possible, therefore proper surrogate endpoints need to be included in the clinical trial: a certain indicator that predicts on a short term basis the long term clinical outcome (e.g. cholesterol level in the case of cardiovascular disease). 

Validation of biomarkers that can be useful as surrogate endpoints is very important nowadays for drug development.