Getting personal

Over the last few years, the concept of personalised medicine has started to move beyond its role in oncology. In 2012, the FDA approved Kalydeco (ivacaftor), developed by Vertex Pharmaceuticals, for use in cystic fibrosis (CF) patients with the specific G551D mutation. This mutated gene is responsible for about 4% of CF cases in the US. Kalydeco became the first CF drug to address the underlying cause rather than the symptoms of the disease. 

Pharma major Roche says the majority of its late-stage compounds for the treatment of rheumatoid arthritis, asthma and haematology, as well as for cancer, are now developed with a diagnostic test. Swiss pharma Janssen is developing diagnostics to guide treatment decisions in schizophrenia, while AstraZeneca has paired up with Montreal Heart Institute in Canada to find genes linked to cardiovascular diseases and diabetes, and with Abbott to develop companion diagnostic tests to identify asthma patients who would benefit from the investigational drug tralokinumab.

In January 2015, the US government announced a $215m investment in ‘a bold new research effort to revolutionise how we improve health and treat disease’. Called the Precision Medicine Initiative, this is expected to exploit advances in genomics and other technologies to find more personalised treatments, initially for cancer. It will also create a research cohort of one million Americans, which will extend the personalised medicine success in cancer to other diseases, including diabetes, heart disease, Alzheimer’s, obesity, mental illness and rare diseases. In the same month, the UK government announced a £13.7m investment in four new ‘stratified medicine’ collaborations to develop targeted treatments for cancer, heart disease, asthma and lupus.

Already research is advancing the field. For instance, Thomas explains, several studies have confirmed that blood thinner Plavix (clopidogrel) requires activation by an enzyme coded for by the CYP2C19 gene. Genetic variation within this gene, however, reduces the activation of the drug and about 25–30% of stent patients have a threefold risk of stent thrombosis when taking Plavix because of this variation. According to FDA, there are now a number of diagnostic tests that can reveal the patient’s risk and determine alternative treatments.

In other research, explains Nicholas Donoghoe, partner in the pharmaceuticals and medical products practice, McKinskey & Co, genetic studies of patients with high cholesterol have revealed mutations in the PCSK9 gene, which, under normal circumstances, codes for a protein that helps remove cholesterol from blood. This finding, he says, has seen the recent development of PCSK9 inhibitor drugs – Repatha (evolocumab) and Praluent (alirocumab) – to fight high cholesterol in patients. ‘As our knowledge of disease mechanisms grow, we expect the opportunity for personalised medicine will follow, which promises future breakthroughs,’ he says.

With all this interest in targeted therapies, Keeling says personalised medicine has become the dominant R&D model within the pharmaceutical industry. Indeed, the 2015 Tufts University study estimates that more than 40% of all drugs now in development are personalised medicines and forecasts a 69% increase in the number of tailored drugs in development over the next five years.

There are, however, major hurdles before personalised medicine can become the norm. For starters, a new approach to clinical trials will be necessary. According to Donoghoe: ‘To accommodate more targeted patient groups, clinical trials will likely get smaller and harder to enrol.’

Furthermore, he believes trials are likely to become less traditional overall, with variations seen in dose, administration and drug combinations as regimes are tailored for each patient in the trial. ‘These non-traditional trial set ups may prove to be a challenge for regulatory agencies, which will need to adapt to evolving ways of testing therapies,’ he says. 

Richard Erwin, general manager at Roche UK, says the use of targeted medicine will mean companies will have to reach out to more patients and carry out screening tests to identify relevant patients for specific trials. ‘This could increase the costs of clinical trials,’ he says; however, he adds: ‘The use of social media and technology will aid identifying specific patients until such a time as proactive screening is widespread.’

There have been concerns, too, that smaller trials will have trouble producing convincing results but Thomas says advances in diagnostics should provide greater information on how individual patients benefit. This should reduce failure rates and development costs. The industry is also likely to review old drugs – which had failed or were less successful than anticipated – in smaller subpopulations to find specific benefits, she adds. 

The shift from a one-size-fits-all approach to niche drugs also requires a change in pharma’s business model, says Thomas. The co-development of companion diagnostics is a potential challenge to companies entering the space.

As a 2013 FDA report on personalised medicine notes, this raises considerable technical, conceptual, organisational and procedural challenges. For instance, aligning the development of drug and diagnostic is an issue when both have different development cycles, and regulatory and reimbursement requirements.

However, steps in this direction are being taken. Roche, for example, has in place a companion diagnostic strategy alongside its drugs in development. And AstraZeneca has created an in-house genome centre and bespoke database of genome sequences to identify new drug targets and aid selection of patients for clinical trials. ‘Genomics will be fundamental to our laboratory research, our clinical trials and the launch of our medicines for patients,’ says Menelas Pangalos, VP of innovative medicines and early development at AstraZeneca.

Keeling notes that advances in diagnostics are out-pacing the ability to integrate these devices into the treatment pathway. ‘The concept of precision medicine is no longer regarded as niche as it was 10 years ago but we score a three out of 10 for addressing the infrastructure to support the testing needs of these great treatments. That’s not good for industry or patients.’

Keeling believes a personalised medicine model is ‘simply a better model’ that makes financial sense for pharma. But it will be challenging to make the business model fit for purpose. He says more collaborative and pre-competitive dialogue will be needed. David Delaney, head of Healthcare North America, agrees: ‘[To reach the full potential of PM] industry stakeholders must take definitive steps to invest in advanced technologies and workforce talent, adjust to new governance models, and accept significant cultural shifts around data sharing and standards that foster easy interoperability of information.’

There are also concerns around the potential cost of tailored drugs since those already approved have price tags in the hundreds of thousands of dollars. Erwin believes the costs will balance out. ‘It stands to reason that if you only use the medicine in those who benefit most and can exclude those who would not benefit you could potentially drive savings on the overall drug budget,’ he says.

But Thomas is more cautious: ‘We have a very steep learning curve ahead of us to work out how to balance all of these risks and benefits in terms of the implications for drug costs and the fact that healthcare systems globally are stretched. It is difficult to see how we will develop business models that make personalised medicine affordable.’

A final concern is that a lack of clinical knowledge on personalised medicines means the adoption of such drugs will be slow. Keeling says there is still a long way to go before all stakeholders are adequately educated to optimise the business model for PM. ‘Educating all stakeholders is one way we can help move the needle on these issues,’ adds Miller.