Cancer: what's on the cards?

How likely we are to develop cancer is determined by our genes and our lifestyle. Likening this to a hand of cards, the risk of cancer depends on whether we inherit good cards or bad cards and also how we play them (our lifestyle). Some genes carry a very high risk, which is why some individuals have a particularly strong family history of cancer. Most of the time, however, cancer risk is determined by a combination of genes conferring a more moderate risk; nevertheless, the overall risk can be high (a bad hand) if there are enough of these genes.

At Strangeways Research Laboratory (SRL) in Cambridge, the research groups of Professor Doug Easton, Dr Paul Pharoah and Dr Alison Dunning, supported with over £9 million of funding from Cancer Research UK, are working out the role of normal human genetic variation in cancer risk – essentially, which hands of cards are worse than others.

The SEARCH begins

Several high-risk gene defects, such as mutations in BRCA1 and BRCA2 in breast cancer, have been identified through family studies. Work in the 1990s at SRL by Professor Sir Bruce Ponder, who is now Director of the Cancer Research UK Cambridge Research Institute (CRI), and Professor Easton helped track down these two genes. However, among the general population, these defects are usually rare, and most cancers are the result of inheriting several more-common gene mutations. For breast cancer, these more-common gene defects account for as much as 80% of inherited risk, according to findings at SRL.

‘The trouble is, the individual effects of common genetic variants are small and to get reliable evidence about specific variants you need to sample large numbers of people,’ explained Professor Easton, Director of the Cancer Research UK Genetic Epidemiology Group. Key to finding these mutations has been the assembly at SRL of one of the largest population studies of cancer ever conducted, along with the unique expertise that the team has gathered together in cancer epidemiology, biostatistics, large-scale genetic analysis and public health medicine.

This work was started in 1996 by Professor Ponder and Professor Nicholas Day, who recruited Dr Paul Pharoah as a Clinical Fellow to enrol patients with breast cancer. The enrolment was later extended to include patients with ovarian, colorectal or uterine cancer, as well as participants with no history of cancer. The project, now called SEARCH, was further extended five years ago to include bladder, brain, kidney, oesophageal and pancreatic cancers, as well as melanoma and non-Hodgkin’s lymphoma. ‘Our early investment in well-curated and very large study sets, with blood samples, pathology review and clinical data, has been absolutely crucial as it has provided the statistical power for reliable conclusions about common genetic variants,’ explained Professor Ponder. Today, SEARCH numbers nearly 27,000 cases and normal controls from the East Anglia region, providing a remarkable and growing resource: by 2013, the hope is that this will have expanded to 35,000.

Going global

‘SEARCH has shown that the size of the dataset is really important for assessing the impact of common genetic variants accurately,’ explained Dr Pharoah, who leads a research group at SRL from the Department of Oncology. ‘The logical next step was to combine data from SEARCH with other studies that had been happening worldwide.’

Cambridge now coordinates five international consortia of study groups: two consortia studying breast cancer, one each studying ovarian and prostate cancers, and a newly formed consortium to examine genetic differences underlying adverse side-effects from cancer radiotherapy (see below). From SRL, the consortia pull in not just SEARCH but also other studies such as ProtecT, a prostate cancer study led by Cambridge’s Professor of Surgical Oncology, David Neal, together with Professor Freddie Hamdy in Oxford and Professor Jenny Donovan in Bristol; and the familial breast cancer study EMBRACE, led by Professor Easton. The scale of the endeavour is unprecedented in population studies, and the European Union has recently awarded €12 million to coordinate these large-scale genetic studies in breast, ovarian and prostate cancers.

‘Apart from increasing the reliability of the data,’ said Professor Easton, ‘the international consortia afford the opportunity to study populations from different parts of the world where different genetic and lifestyle factors are operating.’

Gene hunting

Improvements in technology that would have been hard to imagine when SEARCH began are now being used to analyse the data, demonstrating the enormous foresight in setting up such a resource a decade ago.

‘We know that there are about 10 million variants in the genome, but choosing the right ones to test for association with cancer has in the past owed a great deal to chance, with the result that very few positive associations were identified,’ said Dr Dunning, who leads the high-throughput laboratory team within the SRL. ‘Now, though, thanks both to the ability to carry out genome-wide scans and the samples collected through the international consortia, we can pinpoint the variants that are definitively linked to the risk of cancer.’

A full genome scan for breast cancer, the first of its kind, was completed in 2007 by the researchers and published in Nature. Full genome scans of prostate and ovarian cancers have since followed in Nature Genetics. In the latest scan, published in October 2009, the genomes of 38,000 men with and without prostate cancer were analysed for over 43,000 single differences in DNA (called single nucleotide polymorphisms or SNPs), revealing multiple new cancer gene regions.

To date, 13 predisposing gene regions have been identified for breast cancer, five for ovarian cancer and 27 for prostate cancer, findings that have significant implications for targeted screening and prevention in the future. Since most of these newly discovered regions contain genes that had not previously been considered in cancer, they will also provide new insights into the biology of the disease. Going forward, Professor Ponder’s group at the CRI is developing phenotypic assays as a read-out of cancer risk, studying how risk genes exert their function and searching for molecular markers for future studies of early diagnosis and prevention. Professor Fiona Watt at the Wellcome Trust Centre for Stem Cell Research is also studying the biological effects caused by these gene defects to understand what goes wrong in cancer.

Professor Neal, in the Department of Oncology and CRI, has been investigating whether a protein made by one of the newly discovered prostate loci can be used as a screening and diagnostic marker in prostate cancer since it can be measured in serum and urine. Early results suggest not only loss of the protein in prostate cancer, but also a decrease in men who possess the high- risk form of the gene but who have not yet developed prostate cancer.

Translational tools

But what does this all mean to understanding our own risk? For breast and ovarian cancer, Dr Antonis Antoniou and Professor Easton have developed a computer model named BOADICEA that can predict an individual’s risk of these cancers. The tool is already being used by genetic counsellors to identify high-risk individuals, referring them for counselling and regular screening if appropriate, and providing advice about ways to lower their risk. As new data come to light from the genome scans, BOADICEA will continually be improved, providing increasingly accurate information to individuals wishing to know the hand of genetic cards that they’ve been dealt.

BCAC: Breast Cancer Association Consortium

BOADICEA: Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm

CIMBA: Consortium of Investigators of Modifiers of BRCA1/2

EMBRACE: Epidemiological Study of Familial Breast Cancer

OCAC: Ovarian Cancer Association Consortium

PRACTICAL: Prostate Cancer Association Group to Investigate Cancer Associated Alterations in the Genome

ProtecT: Prostate Testing for Cancer and Treatment

SEARCH: Studies of Epidemiology and Risk Factors in Cancer Heredity

For more information, please contact Professor Doug Easton (doug.easton@srl.cam.ac.uk), Dr Paul Pharoah (paul.pharoah@srl.cam.ac.uk) and Dr Alison Dunning (alisond@srl.cam.ac.uk) at the Strangeways Research Laboratory or visit the SEARCH website.