No clinical trial has a 100% success rate and this is mainly due to the differing genetic makeup between individuals. Many experts have tried to improve the rates or offer explanations as to why one drug would fail in clinical testing while others work.

A year ago, in Canada, a pilot project called "Genomics for Precision Drug Therapy in the Community Pharmacy" set out to collect DNA samples from patients from 29 pharmacies around the Fairfield province. The aim was to analyse the DNA of the patients to determine how they respond to certain medications through a process called pharmacogenomic testing.

This spring, the project will be entering Phase Two where pharmacists will also be collecting samples from patients taking mental health medications and commonly-prescribed cardiovascular and pain medications for sequencing. Pharmacists and patients will then be informed about how their body is reacting to the medication and if they need a higher or lower dosage.

“I like to think it’s future-proofed because you have that information. It’s yours and because it’s coded in your DNA, it doesn’t change for your lifetime. You or I could potentially use that information for drugs that haven’t even been invented yet. I won’t have to provide your DNA again,” said Dr Corey Nislow, lead author of the project, adding the second phase is also expected to take 12 months.

Australia's government investing in pharmacogenomics

Nislow is not the only one attempting to address the issue. The Australian government's National Health and Medical Research Council has just launched a 20,000-person genomics study to determine which antidepressant worked the best for each individual.

Volunteers 18-years and older will be asked to complete a 15-minute online survey and based on their responses, will be asked to submit a saliva sample for genomic analysis.

The sample will be sequenced and analysed through 'genome-wide association scans". The data could inform new drug development and ensure that the right patients are recruited into the right trials. It could also explain the variable responses individuals have to different types and dosages of antidepressant medications.

“Since genetic factors account for up to 95 percent of drug-response variability, pharmacogenomics can help physicians identify which psychiatric drugs may work best for individual patients, before they even take them," said Paul Owens, CEO of Mayo Clinic startup OneOme.

Could natural knockouts provide clues as well?

Others such as heart disease geneticist Sekar Kathiresan of the Broach Institute in Massachusetts aim to explain why some drugs work for some individuals and not other.

His research team looked at individuals who are born as natural knockouts - missing one or more genes. These people are commonly found in Pakistan as they often marry their first cousins. This raises the odds that a mother and father will both pass identical copies of a mutation in a specific gene to their children.

Their study looked at the DNA of 10,503 Pakistanis participating in a long-term heart disease and diabetes study and found that 1317 genes were knocked out without causing obvious medical issues. The team then looked for abnormalities in about 200 clinical blood biomarkers such as cholesterol.

"The Human Genome Project gave us a parts list of 18,000 genes, and this project is trying to understand what missing a part means in terms of biological consequences," says heart disease geneticist Sekar Kathiresan of the Broad Institute in Cambridge, Massachusetts, who co-led the study.

Why some drugs work and others don't

They found seven genes disabled in at least two people which was matched up with a biomarker change, such as unusually low levels of insulin. Other genes missing include one that codes for an enzyme called Lp-PLA2, which is linked to arterial plaque. Drug companies have been spending billions of dollars to block Lp-PLA2 as it was believed that higher levels of Lp-PLA2 correlate with higher risks of heart attacks.

However the Pakistanis who lacked either one or both copies of the gene did not have a lower risk of heart disease, explaining why the drugs have failed in clinical trials. 

Now Kathiresan and others are calling for a Human Knockout Project that will collate all data from similar projects into a single database. This would make it possible for academic researchers and drug companies to trace an individual knockout from the database and carry out more clinical tests. MIMS

Read more:
Patient-led research may be the future of medical studies
Sharing clinical data: A way forward for scientific research
Immunotherapies might hasten tumour growth instead, suggests cancer researchers