Research opportunities in personalized cancer medicine

Advances in gene sequencing and targeted drug development are helping make personalized healthcare a reality. What does this mean for research careers in oncology?

The thing that's going to distinguish people who get jobs and those who don't is having unique skill sets. Hearn Jay Cho, New York University (NYU) Cancer Institute

ONE OF the tantalizing possibilities of the Human Genome Project is the prospect of healthcare tailored to an individual's genetic profile — and the most tangible opportunity for personalized medicine is in the field of oncology. “A decade after the original publication of the draft human genome sequence, there are now incredibly exciting prospects for cancer patients,” says Paul Workman, director of the Cancer Research UK Cancer Therapeutics Unit at the Institute of Cancer Research (ICR) in the United Kingdom. “Personalized medicine is already making a difference.”

In personalized medical treatment, diagnostics can pinpoint which molecular abnormalities underlie a disease. Therapies designed to target those specific aberrations can then, in principle, provide maximum benefit for the patient with minimal effects on healthy cells. It's an approach that could revolutionize healthcare. “This is really the future of medicine,” says Garret Hampton, senior director of development oncology diagnostics at California-based biotechnology company Genentech.

Credit: GENENTECH

A natural target

Oncology is an ideal starting point for personalized medicine because cancer is essentially a blanket term for a multitude of disorders whose features are largely unique to each patient. For instance, when scientists at Washington University School of Medicine in St Louis, Missouri, and their co-researchers sequenced the genomes of breast tumours from 50 patients, they discovered the tumour's genetic mutations were highly diverse — of the 1,700 mutations discovered in total, most were unique to each patient's tumour, and only three mutations occurred in 10 percent or more.

“Cancer is a genetic disease, but one that can reflect changes in hundreds of genes,” says Andrew Simpson, scientific director at the Ludwig Institute for Cancer Research (LICR), a global non-profit research organization with bases in Europe, the United States, Brazil and Australia. “The permutations in any one particular tumour are rarely, if ever, reproduced between individuals.”

However, some tumour characteristics are more common than others — for example, around one in five cases of breast cancer involve the over-expression of the HER2 protein, a characteristic that is targeted by the drug trastuzumab (Herceptin; Genentech). In April 2010 researchers from Stanford University's School of Medicine and their collaborators published a paper in Nature that showed around one-third of breast cancer patients have tumours that express high levels of an RNA transcript called HOTAIR.

The discovery of specific biomarkers and the development of targeted treatments and diagnostics requires a wide variety of researchers, says Wilbert Zwart, junior group leader at the department of molecular pathology at the Netherlands Cancer Institute (NKI). These include molecular biologists, cell biologists, geneticists, biochemists, pharmacologists and molecular imagers. The spectrum of disciplines provides the in-depth knowledge needed to evaluate a potential new treatment. “The success of a molecule in clinical development really depends on the extent to which one understands the mechanism of action a drug has against its target,” says Genentech's Hampton, adding that a diagnostic test to correctly target the drug is another essential component.

A deluge of information is being generated by high-throughput sequencing of cancer genomes. This data flow will increase as the sequencing of patients' genomes and those of their tumours becomes a more affordable and integral aspect of standard care. National and regional screening programmes will contribute to the information flow. In the United Kingdom, for example, Cancer Research UK is leading a new Stratified Medicine Programme, with support from AstraZeneca and Pfizer, which will store DNA samples from the tumours of 9,000 cancer patients from the National Health Service (NHS) in a two-year pilot project. Mike Burgess, head of the oncology discovery and translational area of Roche's Pharma Research and Early Development organization, describes the breadth of opportunities that will arise as a result. He says people with skills in bioinformatics will obviously be vital to help make sense of the data, but expertise in mathematics and statistics will be invaluable.

Multidisciplinary medicine

Much of the data generation occurs in an academic or clinical setting and involves a wide range of disciplines. “We're following more than 80,000 patients right now and we're enrolling another 300 to 400 a week, so that's a lot of data,” says William Dalton, chief executive officer and centre director at Moffitt Cancer Center in Tampa, Florida, which is storing the data for use in future studies. “The goal is to develop associations between different aspects of a patient's history and molecular data. We have major efforts in bioinformatics and information technology to provide quality data in a manageable way to clinicians, scientists, administrators and also patients — the effort will require a huge workforce,” Dalton explains. They've also hired mathematicians “who've never touched a test tube” to work with colleagues in the lab to understand and model systems for better predictions.

The main hubs of Roche's global R&D network are located in Basel, Switzerland; Nutley, New Jersey, United States (pictured); and San Francisco, United States. Credit: ROCHE

Although researchers who are expert in their specialties are needed, so too are those “who have an overview and working knowledge of the whole spectrum of skills from gene to clinical trial,” says ICR's Workman. “None of these activities are happening in isolation — each person has to work with colleagues in other areas, to help fit together the jigsaw puzzle of personalized medicine.”

To this end, researchers with multidisciplinary expertise who can help translate research to the clinic are especially welcome in the field. For example physician-scientists — scientifically trained clinicians who work in both patient care and research — are sought after. “There's definitely a market for MD-PhDs who can treat patients but also understand the genomics,” Workman says. “Molecular pathologists are also really important — individuals who understand disease pathology in relation to different tumour types but also have a deep molecular understanding of the disease.”

In the same vein, a collaborative spirit among researchers is crucial. “We don't need people working in their own ivory towers — we need the kind of personality that can work and communicate with others, understand what contributions they bring to the table, and seeks diverse talent,” says Dalton of Moffitt Cancer Center. “We need team players.”

Collaborations between institutions will also be essential for the development of personalized medicine. Because the targeted therapies are specific to relatively small groups, researchers may need to form consortia to find enough patients to recruit for studies. Dalton highlights an example of this at Moffitt Cancer Center: “We had a new drug that was being tested in metastatic colorectal cancer, and we found 37 patients in three and a half months using our database, which pools together information from 18 partners in 10 different states.” He and his colleagues soon hope to include international partners in the database.

Developing targeted drugs

As well as research collaborations, personalized medicine for cancer will necessarily involve combinations of drugs — specialized therapies might only be useful to combat one aspect of what are often very complicated cancers. “Many drugs will be available that may not be really effective on their own, so there will be a lot of work available in analyzing what combinations are effective,” says LICR's Simpson.

And, while a single drug might prove effective against a cancer for a set period, resistance to that drug may later develop. The solution to this may also be a combination of therapies. “It's like chess — we have to anticipate whatever move a cancer might make and counteract it before it adapts,” Dalton says. This approach will provide opportunities for those working in the relatively new field of systems biology. “You want systems biology expertise to help understand how perturbations in one system might cause changes in another,” Dalton explains.

The move toward treatment combinations will lead to new strategies for clinical trials. “You'll see multiple small trials [running] together to accomplish more,” says Anna Pavlick, director of the melanoma programme at the New York University (NYU) Cancer Institute. “For instance, after you prove safety with a single therapy in a phase I trial, as it moves into phase II you'll combine that therapy with others in another phase I trial.”

Novel therapies are not the only focus of personalized medicine. Any treatment requires complementary diagnostics to pinpoint which specific patients it will suit best. As well as sensitivity and precision, speed is an important factor in designing diagnostics. “You want rapid turnaround time — sometimes patients may not have three weeks [for you] to run a test,” Pavlick explains. “There's a desire for researchers to develop assays that work quickly.” To explain Pavlick points to a diagnostic test for vemurafenib (Zelboraf; Roche/Plexxikon), a drug that received approval from the US Food and Drug Administration (FDA) in August this year for use in patients with inoperable or metastatic melanoma that carries the BRAF V600E mutation. The test, called the cobas BRAF Mutation Test, is PCR-based and will indicate whether a patient is suitable for treatment with vemurafenib within 72 hours.

Around 1,000 staff and postgraduate students work for the UK's Institute of Cancer Research (ICR). Credit: ICR

Given the move toward niche drugs that target smaller populations of patients, LICR's Simpson sees plenty of opportunities for very small companies that each develop a single reagent. “You'll see these ‘micro-biotechs’ designed as rather temporary things, forming and closing as they either bring drugs through trials and license them to a larger entity or see failure of that particular reagent,” he says. “There'll be real demand for people with a mixture of drug-making skills and business skills to run those companies.”

An international job market

Although there are opportunities for a wide range of disciplines in personalized medicine, the sector is subject to the tight job market affecting all others. Despite that, there are ways to stand out from the crowd. “The thing that's going to distinguish people who get jobs and those who don't is having unique skill sets, such as information technology, computer science and systems engineering in addition to understanding biology,” says Hearn Jay Cho, assistant professor of medicine and pathology at the NYU Cancer Institute. “That will really make people competitive.”

Employment possibilities in personalized cancer healthcare research are not restricted to the United States and Europe. “China is a big growth area for this type of expertise [and it] is going to expand exponentially — they have a large population and they have the infrastructure necessary to analyze these types of data sets,” Roche's Burgess says.

James Christensen, director of translational pharmacology at Pfizer, concurs. “Pfizer is doing more and more work in Asia, particularly China — we're working with academic centres and contract research organizations there on drug discovery,” he says. “There's also interest there in understanding the molecular basis of common malignancies in Asia, such as hepatocellular carcinoma and gastric cancer.”

Personalized cancer medicine is also advancing rapidly in Australia, says Pavlick of the NYU Cancer Institute, also pointing to development in the Middle East. “There's a huge push right now to build science in Dubai,” she says. NKI's Zwart mentions Singapore as another country advancing in this field and attracting very skilled scientists. Simpson from LICR says that the potential for micro-biotechs in the coming decade will make it easier for contributions from developing countries where drug development is not currently well-established — such as Latin American countries.

As the foundations of personalized medicine are laid across the world, the likelihood increases that the Human Genome Project will benefit all of humanity in the fight against cancer. “This field is really becoming global,” Burgess concludes.

Nature editorial staff have no responsibility for content