Personal treatment

In January 2009, an Icelandic company launched the first commercially available genetic tool to assess the personal lifetime risk of common cancers. Recent years have witnessed substantial investments in genetic tools in oncology, not only to calculate lifetime risk but also to predict treatment outcomes and risks of progression in individuals already diagnosed with cancer.

These prognostic tools use biomarkers based on specific DNA sequences or the presence or absence of drug receptors to calculate the probability that an individual will respond to, or progress on, a particular therapy. For many years, oncologists have puzzled over the fact that patients with seemingly identical cancers and associated symptoms can respond so differently to a given proven treatment. Such tools could help to solve this conundrum and identify those most likely to benefit.

The use of genetic testing in oncology is expected to increase significantly in the future, enabling drugmakers to gain a better understanding of how drugs perform according to patients' genetic makeup and enabling patients to receive personalised therapies with greater chance of success.

Cost and health benefits
Studies have predicted that the number of cancer patients will increase by 55 per cent by the year 2020, with cancer incidence in the developing world said to be soaring. Simultaneously – and now more than at any other time – the media has been dominated by concerns over the cost-efficacy of novel cancer therapies.

In such a climate, personalised treatment makes a lot of sense. Indeed, outgoing US Secretary of Health and Human Services, Michael Leavitt, recently advised that personalised healthcare should be an "explicit goal of healthcare reform" for the Obama administration. Almost in parallel, the UK's National Health Service has unveiled new guidance for the personalised care of patients with long-term illnesses.

Not only do tailor-made treatments offer the potential for greater efficacy and improved survival in cancer patients, but they may also offer cost benefits. By selecting patients for treatment according to genetic status or disease subcategory, healthcare providers can avoid funding treatments for patients who are likely to derive no benefit and, in so doing, save those patients from potentially harmful side effects. With novel cancer drugs commanding substantial prices, cost savings could be significant.

Developments in genetics
Tailored treatment is not a new phenomenon in oncology. It has long been known that the hormone receptor status of breast tumours dictates prognosis and response to hormonal therapies such as tamoxifen, a drug that has been commercially available since the 1970s. Similarly, Herceptin (trastuzumab; Roche/Genentech) has been licensed for over 10 years for the treatment of breast cancers that over-express the HER2 gene. Oncologists expect further refinement of the accuracy of HER2 testing and the addition of genotypic information to more drug labels in the coming years.

KRAS – a clue in colorectal therapy
Nowhere in oncology has personalised medicine undergone so much scrutiny in recent months as in the management of metastatic colorectal cancer. The monoclonal antibodies Erbitux (cetuximab; Eli Lilly/Bristol-Myers Squibb) and Vectibix (panitumumab; Amgen) target the epidermal growth factor receptor (EGFR) on the tumour cell's surface to slow disease progression and prolong survival; both drugs are licensed for the treatment of advanced colorectal cancer. However, studies have shown that these treatments are ineffective in the 40 per cent of metastatic colorectal cancer patients whose tumours exhibit a mutated KRAS gene, a gene that regulates cell division.

In an unprecedented move, the drugmakers approached the US Food and Drug Administration (FDA) in December 2008 to argue for label changes to support the use of their agents only in patients with normal KRAS genes who are likely to benefit from the drugs. The US National Comprehensive Cancer Network and the American Society of Clinical Oncology now strongly recommend genetic screening for the KRAS mutation in metastatic colorectal cancer patients before prescribing either treatment.

Use of the KRAS gene as a predictive biomarker could save substantial costs in the management of metastatic colorectal cancer, achieving an estimated annual net saving in the US of $604m for Erbitux alone. Genetic profiling could also be used to direct patients with the KRAS mutation to alternative treatments with greater probability of benefit.

Prognostic value of PTEN
The PTEN gene is the focus of considerable interest in the oncological community for its potential prognostic value in a number of cancers. PTEN is a tumour suppressor gene that is a major component of the EGFR cell-signalling pathway and is associated with programmed cell death in healthy tissues.

The particular appeal of PTEN is that loss of PTEN function occurs commonly in several cancers. For example, low PTEN levels have been associated with Herceptin resistance and poor prognosis in HER2 over-expressing breast tumours. In patients with advanced colorectal cancer, they have been correlated with lack of response to Erbitux; in metastatic prostate cancer, high PTEN levels have been linked to prolonged survival, with similar findings in melanoma and lung cancer.

Researchers now believe that PTEN may become a valuable tool to assess patients' suitability for specific anti-cancer agents. Indeed, a molecular diagnostic product has recently been developed in the US to measure PTEN expression, and there are high hopes that this product, and tools like it, will help to guide therapeutic decision-making in oncology in the future.

Pharmacogenomics in breast cancer
Despite significant milestones in breast cancer treatment with the development of the blockbuster agents Herceptin and, more recently, Avastin (bevacizumab; Roche/Genentech), breast cancer remains the leading cause of cancer death in European women. Pharmacogenomic studies are aiming to address this challenge and determine why some patients respond to particular therapies and others do not.

A Swiss research team has used sophisticated computational methods to investigate whether specific gene activity patterns are associated with treatment responses in newly-diagnosed patients receiving chemotherapy prior to surgery. Somewhat surprisingly, they discovered that the activity of the tumour's micro-environment, rather than the tumour itself, predicted response to chemotherapy.

Chromosomal damage
Another genomic study has revealed high levels of chromosomal damage in a breast cancer cell line. These genomic changes may enable cells to grow uncontrollably and survive outside their normal environment, which are essential attributes of tumour cells. In addition to improving understanding of the genetics of breast cancer, such findings may support the development of future prognostic markers and tailored therapies in breast cancer.

Even in patients receiving tamoxifen, pharmacogenomic developments could help to guide treatment decision-making. A specific metabolic enzyme, CYP2D6, can act as a rate-limiting step in the conversion of tamoxifen into its metabolites. Pre-treatment screening to measure activity of this enzyme could help to determine which patients are most likely to respond to tamoxifen and how long they should remain on treatment in order to achieve optimal benefit. Such screening is already being requested by some patients although, as yet, no consensus is available to guide clinicians on the use of screening in this setting.

Measuring risk of recurrence
Until recently, the reasons why some cancer survivors remain cancer-free and others develop recurrent disease have largely eluded researchers. However, developments in a number of tumour types are beginning to shed light on this perplexing situation and identify potential prognostic and therapeutic strategies.

Data have shown that high expression levels of the newly-identified gene, MACC1, are associated with substantially elevated risks of developing secondary tumours in patients with colorectal cancer. MACC1 is thought to activate the HGF/Met signalling pathway, which is important for tumour growth and metastatic formation. Once this pathway has been activated, tumour cells proliferate and migrate faster, to settle in sites far from the primary tumour.

In a similar development, a gene called metadherin has been linked to treatment resistance and metastatic spread in breast cancer. Previously, it was thought that breast cancer spread initially to nearby organs and lymph nodes before reaching more distant parts of the body. It is now believed that metadherin, which enables tumour cells to 'stick' to blood vessels, allows breast cancer cells to break away from the primary tumour at a much earlier stage in tumour development. In the study, cancers with high metadherin levels were more likely to be treatment-resistant and spread to the bones, lungs and other vital organs. The development of prognostic markers and tailored treatments targeted to metadherin could inhibit secondary tumour formation in the 30–40 per cent of breast cancer patients in whom this gene is over-expressed, representing a significant benefit to a large patient population.

Studies are also helping to elucidate the mechanisms behind metastatic progression in lung cancer. In the future, greater understanding of the genes and proteins involved in metastatic pathways may enable the development of predictive tools to identify patients at risk of disease recurrence so that they can be offered more intensive treatment and monitoring.

Future treatment selection
With the availability of increasingly sophisticated research methods and greater understanding of the complex pathways involved in tumour growth and recurrence, oncologists are beginning to elucidate the mechanisms through which particular therapies succeed or fail in patients with common cancers. Even in traditionally treatment-resistant malignancies, such as pancreatic and ovarian cancer, researchers are identifying potential therapeutic targets and prognostic markers that may help to guide treatment selection in the future.

These significant advances are likely to have an impact on more than just treatment outcomes. The cost-efficacy of novel targeted therapies, particularly for patients with metastatic cancers, is already the subject of considerable media attention. Personalised medicine offers the opportunity to select for treatment only those patients most likely to benefit, so that cost-efficacy can be maximised.

Even the arena of drug trial design may be open to change, following the FDA's recommendation that biomarker development should form an integral part of clinical drug development programmes.

"An ideal scenario is one in which the relationship of the biomarker to potential action of the drug is recognised very early," the FDA reported in 2008. Certainly, collaborations between drugmakers and the developers of molecular assays are likely to be witnessed with increasing frequency in the future.

Personalised medicine is already becoming a reality in the management of some common malignancies. With increasing research focus on genetic profiling, predictive biomarkers and tailored treatments, it is likely to become the norm in the management of a wide range of cancers in the future.

The Author
Anna Hopkins is a medical writer
To comment on this article, email pm@pmlive.com