In an ailing industry desperately casting about for a way to save itself, the rapidly emerging field of pharmacogenomics could provide a life-saving tonic.
There can be no denying the fact that the pharmaceutical world is in flux, if not downright turmoil. The old blockbuster model has fallen by the wayside, and even the big players are seeking a cure, with some even gobbling up their rivals in an attempt to boost flagging pipelines. But there is another option that could be a lifesaver for the industry - pharmacogenomics, also personalised medicine.
Personalised medicine in healthcare is defined as the need to consider an individual patient's unique lifestyle and medical history, and how this impacts on his or her response to treatment. In pharmaceuticals, however, personalised medicine aims to develop drug therapies that have efficacy within narrowly targeted groups of patients, based on each person's genetically programmed reaction to the drugs.
The foundation of this type of personalised medicine is the fact that two people can take the same medication and have completely different responses: one may have severe side effects, and the other none at all. Or the treatment may lead to remission in one person, and seem to have no effect in the other.
One reason for these different outcomes is the variation in our inherited genetic make-up. Genetic variations can determine how we respond to treatment, similarly to the way they cause differences in eye or hair colour. For example, if you have a genetic variation that causes a drug to stay in your body longer than normal, this may cause unwanted side effects.
Researchers are working to identify these genetic variations, and match them with responses to medications, so that physicians can take this into account when prescribing drugs. In addition to the usual information such as weight, age, medical history and how any relatives may have reacted to the same medication, doctors in the future may be able to take into account your own personal genetic make-up.
The field of personalised therapies - and the diagnostics used to develop them - is growing so rapidly that at the end of last year PricewaterhouseCoopers released Diagnostics 2009, the first in a series of annual reviews of significant events for personalised medicine.
According to Tony Pillari, PwC's Director, Healthcare Advisory Services, the report was developed for three key reasons. First, the level of deal activity in the in vitro diagnostics sector, which PwC feels is significant and likely to increase in the coming years, creating a multitude of business opportunities.
Second, the growing importance of diagnostics in the practice of medicine and in the emergence of personalised medicine. And third, the many exciting advances that have been made in recent months in the field of personalised medicine, including those related to diagnostics.
In PwC's definition, a personalised medicine diagnostic is any tool that allows for the generation of data that are then used to tailor prevention and care strategies the needs of the individual. Such tools would include companion diagnostics, early diagnostics and prognostics.
Tailoring the treatment to the patient is not an entirely new idea. What's different about this new definition, says Pillari, is the specificity and accuracy that advances in genomics and proteomics and a whole host of related technologies now make possible.
"We now have a much deeper understanding of many diseases, including complex diseases like cancer, at the molecular level, and our knowledge continues to grow," he says. "That kind of understanding simply wasn't possible before. As a result, the way we treat disease, including how we design treatment options for patients, has changed to a point where trial and error will hopefully be replaced by trial and success."
Pharmaceutical companies looking to maximise success rates will be able to determine which subsets of patients would be more likely to respond positively to their drugs. This happens to some extent already, with companies often targeting subsets of patients if initial trials with larger groups don't yield the expected results. AstraZeneca, for example, did this with its anti-cancer drug Iressa. In 2004, a large randomised study failed to demonstrate a survival advantage for the drug in the treatment of non-small cell lung cancer, but it has since been shown to be effective in patients with mutations relating to epidermal growth factor receptor.
Jan Lundberg, formerly AstraZeneca's EVP for Discovery Research and now head of R&D at Lilly says that even in the last couple of years there have been major developments in personalised healthcare and combined diagnostics. He highlights the CNS area of Alzheimer's disease, where the pathophysiology is increasingly known, based to a large extent on genetic analysis of patients with hereditary Alzheimer's. "Here we have a PET ligand as a new diagnostic tool," he says, "and you can see if patients have accumulated amyloid in the brain, which is a strong signal that the risk of developing Alzheimer's is high. You can follow the accumulation of amyloid and potentially prevent amyloid deposits or reduce them if they are already there, which is likely to influence the disease.
"There are some new agents coming in, which are tested to prevent accumulation of or to deplete amyloid, which offer a way of combining diagnosis and potentially following the effects of anti-amyloid treatments and correlating that with improvement in cognition, again taking a personalised healthcare treatment approach."
Lundberg believes treatments developed for specific patient populations will become more common within the pharmaceutical industry, as traditional business models change and as progress in technology and science make this possible.
This applies particularly in oncology, as he explains: "In oncology we have the ability to analyse tumour biopsies or blood samples for their respective tumours. We can let science drive the benefits for patients both in relation to maximising the potential for clinical effect and also reducing risk. It's still an evolution, but the trend is very clear."
Lundberg does point out, however, that ideally R&D organisations should work out beforehand which groups of patients are likely to benefit, rather than waiting until the drug has been developed and then looking for people it can help.
"A key aspect will be to identify these opportunities early enough. With Iressa, we only established which patients would respond when the drug was already on the market. It would be better if we could do this at the time of, or even before, nominating compounds to the clinic in a new area. We could analyse this tumour population to see if there are specific pathways that are particularly prone to be more responsive, and then design the compounds against these pathways and select patients in early clinical trials that could be maximally responsive, based for instance on genetic analysis of tumour biopsies."
PwC's Pillari believes specialised therapies can help pharmaceutical companies improve their bottom line. "The blockbuster model has been the predominant model in the pharmaceutical industry for some time and until recently, it has been very successful. However, the poor return on R&D investment realised by that model over the past few years, in terms of new drugs developed and introduced into the marketplace, has been well documented, as has the number of drugs coming off patent in the next few years.
"As a result, and quite understandably, pharmaceutical companies are very concerned about pipelines and future revenues. In this context, the move to specialised therapies is relevant to the pharmaceutical industry for two reasons. First is the potential to reduce drug development time and costs and increase success rates through enriched clinical trials.
"These are trials that only include those people who, based on the analysis of their make-up at the molecular level, are most likely to respond to a given drug. Second is the opportunity to move to a 'niche buster' model, that is, one that generates revenue because of the value a specialised therapy delivers to a specific subset of patients."
There are some within the industry who see a potential downside to the advent of personalised medicine. David Lathbury, AstraZeneca's Director of Process Chemistry, believes it will only help big pharma improve its business models if it translates to a much more reduced cost of development. If it doesn't do this, he says, the simple net present value calculation means that companies won't make money. "If you can very quickly identify the target groups, if you can get through phase III with, for example, 200 patients, then absolutely it will be a benefit. But if you've got to go through the same process before you identify that population, it could be very difficult."
Lathbury says that where personalised medicine could make a difference is in dosing regimens, which could help drive through a better result. "We tend to give everyone an average dose, that's the way most clinical programs are designed. If better titration of dose will reduce adverse events, increase compliance and get a better return on the drug, those steps I think we can take."
Then there's the view we could go much further. Dorman Followwill. Partner and VP, Healthcare EIA for Frost & Sullivan, describes his idea of the ultimate version of personalised medicine. "If you're talking about the holy grail of personalised medicine, it would be walking into a physician's office, pulling out a smart card that has your genotype on it, and having them plug it into their system and then diagnose you or, dispense medication for your genotype.
"Estimates of when we might achieve that vary from 10 to 15 years, to as long as 100 years. However long it takes, we will see increased personalszation over time because that is a mega trend in the world today. In my role, I get to look at some of the 360-degree views of many industries produced by our team, and the topic of personalisation comes up big time in healthcare, in automotive and transportation, and in environment and building technologies."
The version of personalised medicine as it exists today has proven popular with regulators. In some cases, they now insist on biomarker testing to guide drug prescribing. Tony Pillari believes their main concern is efficacy. He points the testing for infection by a specific HIV subtype that is required prior to the use of Pfizer's Selzentry treatment, which he says is based on the fact that Selzentry blocks the specific receptor that HIV subtype uses to attach to and infect white blood cells.
"Similarly," he says, "testing for epidermal growth factor receptor (EGFR) expression is required prior to the use of Erbitux because Erbitux binds to EGFR and blocks certain growth factors that ultimately promote the expansion and spread of tumours. We believe this practice will become more common, as everyone in the healthcare value chain, from regulators to pharma to providers to patients to payers, appreciates the benefits of improving drug efficacy as well as reducing the frequency of adverse events.
Yet despite the fact that the clinical case for developing companion diagnostics is strong, the PwC report points out that there has yet to be significant deal-flow between the pharmaceutical and diagnostics industries, because while the existing pathway for drug approval and reimbursement is clear and well-established, this is less so for diagnostics and even less so for drugs and diagnostics developed in tandem.
Pillari does point out that there are a number of encouraging initiatives underway, including by the FDA and the Critical Path Institute, to better define the pathway for drug and companion diagnostics development. As this pathway becomes clearer, and the risks associated with such strategy are mitigated, he expects deal-flow between the two to increase.
However you look at it, and even with its potential downsides, there is no backing away from pharmacogenomics, with its potential to bring once-dead drugs back to life by allowing them to be repurposed for new populations. In a beleaguered industry suffering from the twin ills of dry pipelines and no blockbusters, personalised medicine could provide a life-saving shot in the arm.
Reaping the benefits
The potential benefits of pharmacogenomics could include:
The following genetic tests are currently in use. They help guide dosing and prevent toxic levels of medication from building up in patients who lack certain enzymes.
Cytochrome P450 test. This group of enzymes are responsible for metabolising more than 30 types of medications. The test can determine dosing and effect of some antidepressants, anticoagulants, proton pump inhibitors and a number of others.
Thiopurine methyltransferase test. This enzyme breaks down the chemotherapy drug thiopurine, which is used to treat leukaemia and autoimmune disorders.
UGT1A1 TA repeat genotype test. This test detects a variation in a gene that affects the UGT1A1 enyme, which determines how the body breaks down irinotecan, a drug used to treat colorectal cancer.
Dihydropyrimidine dehydrogenase test. This enzyme breaks down the drug 5-fluorouracil is a commonly used chemotherapy medication.
What are biomarkers?
Biomarkers can be a sign of a normal/abnormal process, or of a condition or disease. For example, blood pressure is widely accepted as a biomarker because large epidemiologic databases exist demonstrating a correlation between elevated blood pressures and adverse cardiovascular outcomes. This has been supported by the numerous placebo-controlled studies showing an effect on stroke and coronary heart disease outcomes from lowering blood pressure.
US FDA Pharmacogenomics Guidance further defines three categories of biomarkers: exploratory, probable and known valid. Markers are included in these categories based upon available scientific information.
Biomarkers can be further divided into the categories of predictive or prognostic. A prognostic biomarker is associated with the likelihood of an outcome (eg. survival, response, recurrence) in a population that is untreated or on 'standard' (non-targeted) treatment.
A predictive biomarker can predict differential effect of treatment on outcome. A predictive biomarker is a biomarker that is present prior to an event occurring and which predicts that outcome. For example, the KRAS oncogene can be considered a negative predictive biomarker for response to treatment with the EGFr (epidermal growth factor receptor) class of drugs since it can identify which patients are unlikely to respond to treatment with an EGFr inhibitor.
This article was first published in Next Generation Pharmaceuticals magazine: www.ngpharma.eu.com/article/Miracle-cure