One of the biggest cost factors for clinical trials in medical research is ensuring you have the right patients. An analysis of clinical-trial data from January 2000 up to April 2019 estimated that only around 12 per cent of drug-development programmes ended in success. Next-generation sequencing can act as a matchmaker between people with particular illnesses (or a genetic predisposition to them) and researchers in related fields. Selection is typically based on data such as age, gender, medical history and current stage of a disease. But that doesn’t always tell the whole story. Vetting participants via NGS could find more patients likely to respond to specific treatments, which would result in fewer failed drugs.
The ability to edit genes could transform our ability to create and deliver therapies. By removing a gene and analysing which functions are affected, researchers can perform ‘knockout screening’ to identify drug targets. They can also take specific genes out of cells, before administering drugs to see if those cells become more sensitive to treatment. In some early clinical trials with CRISPR, scientists are removing cells, editing the DNA and then re-injecting them in an effort to find more effective treatments for cancer and blood disorders.
AI underpins these scientific breakthroughs and is already starting to provide valuable support to clinical trial participation. Natural language processing enables computers to analyse text and speech far quicker than humans. When applied to medical R&D, algorithms can scour doctor’s notes and reports to identify suitable participants for clinical trials. In one recent US pilot study, IBM’s Watson for Clinical Trial Matching system increased the average monthly enrolment for breast-cancer trials by 80 per cent. AI can also search important sources of information like comparable studies, clinical data and regulatory information to support researchers in the trial design phase.