Problems in Pharmacology: Clinical Trials and Molecular Markers

This week’s post follows on from something I touched on last week: the issues in the drug design process.

Drug design tends to be stem either from mimicry of molecules the chemist knows that the drug target already interacts with (such as a substrate that binds to an enzyme) or from knowledge of the potential’s drug target’s structure (where computer modelling, for example, can suggest which would be the best structure for a potential drug).Resized

These drugs are synthesized, and then undergo preclinical testing. This involves assessment of their stability and their interaction with an isolated drug target (called in vitro testing), and then testing inside animal models. After these stages, totalling up to 6 years, the drug will go to clinical trials in patients.

Clinical trials are in three stages, with the amount of patients in each stage increasing as confidence in the drug’s capabilities develops. The final stages are then approval by the authorities and marketing. This drug discovery process takes in total anything between 10 and 16 years and will cost the company up to a billion dollars. And yet, there are still issues with drug safety that slip through the net: think of thalidomide as the most popular example! Why is this?

There are a multitude of reasons of course. Once a drug has reached later trials, a company can become invested in it and wish to push it through. There may also be signs that the drug is unsafe missed, which may come from the design of the clinical trial itself.

In clinical pharmacology, a biomarker is a characteristic the patient has that can be measured during clinical trials, and is used as an indicator of their biological state. For example, a biomarker such as glucose or hormone levels can indicate the likelihood that a patient will get better or relapse.

A key issue with clinical trials is that they do not measure all relevant biomarkers. A different drug for the same medical condition will have a different action (mechanism) in the body and therefore will have a different effect on biomarkers. They may effect biomarkers that were unaffected by other drugs.

Therefore, in order to fully understand the effect of the drug in the body and truly assess its safety, trials should be randomised with controls (patients taking “placebo” pills with no drug in them to show how biomarkers might change with no drug present). The system currently unfortunately relies on the validation of biomarkers to display patient health, which is both a lengthy and difficult process.

New clinical trial designs are therefore being considered, such as the “I-SPY 2” trial which uses molecular tests to tailor treatment or “BATTLE” which uses biomarkers to tailor the treatment. These new designs are quite new, so unfortunately I am unable to provide more detail than this!

The future of drug discovery seems to be tending towards the more in silico side of research, which uses computational modelling of drug action to suggest how a drug will act in a given system. The future looks bright here, but again that is a discussion for a later post!

I based this post on a talk at the EACPT 2013 conference by Professor Max Parmar (UCL). This post is entirely my own opinion, based on my own experiences – feel free to disagree and share your thoughts in the comments! I’ll be continuing on the “Problems in Pharmacology” theme next week!


A note from the author: As my posts sometimes touch on emotive subjects, comments are disabled after 4 days. This is because, at this stage, I feel that ongoing discussions tend to stagnate.