Precision medicine is fundamentally changing how we treat patients, but is it too early to get excited about this burgeoning area?
'Precision medicine' is not a very precise term, and pundits have used it to mean anything from avoiding one-size-fits-all drugs to individualised treatments for every patient. Everyone agrees, though, that it is key to the future of medicine – even if many also feel that there is still a long way to go to perfect it.
"For us, precision medicine is about understanding and using the molecular basis of a patient's cancer in order to guide their treatment," says Dr Rowena Sharpe, head of precision medicine at Cancer Research UK and leader of the charity's Stratified Medicine Programme, which is working to prepare the NHS for this new era of medicine. "It's being discussed at the global level."
Cancer is perhaps the best showcase for the paradigm-shifting potential of precision medicine, with targeted drugs available for more than 25 different cancers and many more in clinical trials. What's more, some of the best-known and most successful targeted medicines are in cancer, including drugs that specifically target breast cancer patients with the BRCA mutation or lung cancer patients expressing PD-L1.
A recent meta-analysis by the University of California San Diego confirmed the importance of targeted treatments in cancer. It looked at 346 phase I clinical cancer trials and found that tumour shrinkage rates were 30.6 percent in treatment arms employing precision medicine compared to 4.9 percent in other arms.
Although precision medicine can use any biomarker, most decisions about targeted cancer treatment today involve testing for genetic changes that drive tumour growth.
However, genetics are not the be-all-and-end-all of precision medicine, says Steve Gardner, CEO of healthcare data firm RowAnalytics. "The one-to-one link between precision medicine and genetics is a little bit blinkered," he told the audience at a Life Sciences Hub Wales' seminar on personalised medicine in June. "Looking only at genetics is a black and white photo; we want multi-dimensional analysis because it enable tests to become better."
He cited the prostate-specific antigen (PSA) test as an example. "PSA is only about 90 percent accurate, but if you combine certain single nucleotide polymorphisms (SNPs) with the PSA antigen value, you can improve the accuracy rate to 99.9 percent for a pretty large cohort of the population. Protein-level plus genetics has a potentially huge impact."
Cancer Research UK's Sharpe agrees: "Although precision medicine to date has focused on genomically driven targeted therapies, we also want start moving towards a multi-omic signature approach, combining different types of 'omics' to get an accurate prediction.
"You have no guarantee that a fault in the DNA will lead to a fault in the production or expression of a protein, so we need to look more at expression analysis, proteomics and metabolomics. Those areas are in their infancy at the moment but we're working with the scientific community to identify key areas that show promise so we can validate the technology and take it into a trial setting," she says.
There is also a need to treat patients with a combination approach, says Sharpe. "If you know that a patient has several drivers, you want to target those mutated and activated pathways with several drugs at the same time. The issue with combination therapies is that toxicity needs to be taken into account; when you move into the dual or triple combination era you will get an ever-increasing risk of safety implications for the patients, so we need to tread carefully."
She says a multifaceted approach could become particularly important with immuno-oncology drugs, as current targeted treatments for immunotherapies still have room for improvement. "Immuno-oncology is a really exciting avenue that has, to date, had durable response in patients with melanoma and lung cancer. What we really need is a predictor for which patients will and won't respond so we can guide treatment accordingly.
"At the moment, we've got PD-L1 expression, which isn't the best of biomarkers, but we need to make sure we standardise the limits of those tests. For example, how much expression gives a positive result? We also need to bring in other elements to make it truly predictive, as some people with PD-L1 expression don't respond to the antibody and some who don't have expression do respond."
Such an approach will, of course, require more precision drugs to be developed, so pharma has an important role to play going forward, says Sharpe. "The pharma industry is crucial to this area and we are keen to further our collaboration. The more we can use multiple drug pipelines the more we can bring these combination therapies into the clinic and derive benefit from them.
"Our aim with the Stratified Medicine Programme is to set up a consortium so that pharma partners can open up discussions with each other – this model is working really well at the moment and we hope to expand it. We need to make sure that we bring funding and drugs to the UK and open up those pipelines so we can bring our academic brains to the mix and start to design interesting studies that will hopefully benefit everyone."
A long road ahead?
However, Gardner of RowAnalytics cautions that widespread adoption of targeted drugs is a long way away. "Trying to develop a whole raft of new drugs – pushing them through the approvals process, getting them on the market and on to the formularies then getting them prescribed – is a generational change. It will happen in particular areas because certain disease areas require them and certain patient populations demand them, but I don't think it will happen everywhere for at least 20 years. It isn't going to be in routine practice through primary, secondary and tertiary care in the NHS anytime soon."
The same is true for the diagnostics required to make these drugs work, says Tim Pitfield, sales director EMEA and China at Janssen Diagnostics. "You need to deliver not just in R&D but in front-end development, the regulatory environment, in marketing and sales, and in manufacturing. There is a huge amount of work getting from a new biomarker that might be a great concept to developing it in routine practice. It is a big step to take a test from a research centre and put it in a hospital and get the same result."
Targeted treatments can't exist without these diagnostics, but there are many other hurdles to overcome in this area, he says. "The business model doesn't stack up very well for diagnostics because you have to run lots of patients in a large study. It can cost millions to generate the evidence required to show its effectiveness, and there is also an expectation around what you can charge for a diagnostic.
"Evidence generation has been a problem for many diagnostics companies so they are launching without specific intended use or any clinical outcome connection, and the evidence to support the diagnostic is generated in routine use in the laboratory post-launch," says Pitfield.
Luckily, there are several supporting mechanisms in place in the UK to help generate this evidence, he says. "We have the NIHR's diagnostic evidence co-operatives in London, Leeds, Newcastle and Oxford, which have been established as research groups to help SMEs with a new biomarker or potential diagnostic put together studies. There's also the Companion Diagnostics Catapult, which is another large amount of funding to support biotechnology companies as they develop new innovative diagnostics."
Breaking into the NHS
The NHS itself is another hurdle to clear, as it needs to be equipped to deal with the new process precision medicine requires, says Gardner. "The NHS isn't remotely ready to do much of this. There's limited resource, the technologies are changing all the time and the amount of validation required for technologies and clinical practices is quite a bit higher than what you use in the research to find the biomarkers in the first place."
The health service needs to look at incentives throughout the entire process, he adds. "Not to put too fine a point on it, doctors are paid to prescribe medicines; that's where their incentives come from. It's not about curing people, it's about treating people. The clinical population has got to know about precision medicine, it has to become educated about it, it needs to be confident in it, then you need the incentives to work in your favour so they prescribe a more accurate test."
Pathology in particular is an area that needs more resource, says Sharpe. "The demands we're putting on the health service are ever increasing, and the types of requests we're making with genomic analysis is a shift in practice for pathology systems. It's not just asking for another section and performing another stain, it's adapting the whole fixation process to preserve that nucleic acid and make sure it's of a good enough quality to yield a useful result.
"As results increase in complexity, we need to make sure we inform and educate the medical community so that they understand the basic biology and feel comfortable using these results to inform treatment and discuss them with patients," she says.
Old meets new
Precision medicine may excite many, not least patients, but even when it becomes widespread it is unlikely to completely replace traditional treatments, says Sharpe. "Even treatments like chemotherapy have beautiful response rates in early-stage testicular cancer, for example. If it's a treatment that works and it gives a high response rate then we should stick with it.
"Ultimately, you will get resistance kicking in to precision medicine, so what we've tended to do is focus the targeted therapy on late-stage cancer patients who won't benefit from radiotherapy or surgery, or they've had both and they've relapsed."
Resistance is an issue but there are plenty of possible solutions, she says. "We could use blood-based technology to monitor response and recurrence or resistance when using targeted therapy and switch at the appropriate moment. Or we could use combination therapies up front to inhibit the multiple pathways early on and make sure patients get a longer response."
In the end, though, most challenges with precision medicine seem like small quibbles compared to the huge promise offered by the field. The future of medicine may yet be slow to arrive but it is certainly on its way.