Editor's Note: In May 2014, the Liddy Shriver Sarcoma Initiative awarded a $250K grant to the International Sarcoma Kindred Study.
Learn more about the study in this video:
The International Sarcoma Kindred Study (ISKS)1 was established to address a fundamental question faced by all afflicted by cancer: why me? There are many factors that align to explain the seemingly random impact of cancers. Environmental causes, like smoking, sunlight, diet and obesity, radiation and some chemicals, have long been known to modify cancer risk. What is remarkable, however, is how much variation exists between individuals exposed to the same cancer risk. Two smokers with the same number of pack years of tobacco exposure: one lives to 90; the other gets lung cancer in his fifties. Why? One answer is chance. Cancer is a complex disease, with thousands of genetic changes within the cancer cell, perhaps dozens of which are important to cancer development. Acquiring these changes (also known as mutations) takes time. Although exposure to environmental factors gives tumors a head start, multiple additional events are still required for a cancer to establish itself.
Another possible answer is that there appear to be enormous differences between individuals in cancer risk, which are due to inborn variation. This variation may be due to the genes we each inherit from our parents, or due to random changes in our genes that occur within us as we develop from an embryo. In either case, a substantial number of our normal cells are ‘primed’ by these changes to have a head start in becoming a cancer. In some cases, this risk is so extreme that their effects are visible in families that have far too much cancer to be the result of chance alone.
Heritable vs Genetic
The terms heritable and genetic are not interchangeable. A heritable event is one that can be passed from parent to child. Genetic simply means 'relating to genes.' Whilst there are heritable genetic events, such as the passing on of genes from parent to child, the term genetic does not always refer to something that is heritable. Individuals can have genetic changes occur within cancer cells during the development of a cancer that are not heritable. These are referred to as somatic changes, and they typically occur within cancer cells. Heritable genetic events affect what is sometimes referred to as the 'germline.'
Much of what we know about cancer biology has been learnt from such families—examples include TP53,2 RB1,3 and the breast cancer genes BRCA14 and BRCA2.5-7 The other major source of knowledge about cancer biology has come from animal models of heritable cancer predisposition,8,9 in which the role of certain key genes in cancer development has been. In these models, scientists engineer mice that carry mutated genes in their tissues, and show that such mice are then cancer prone.
The big change recently that puts people back in the frame as subjects for genetic research is Next Generation Sequencing (NGS). Previous techniques for identifying genetic changes (mutations) have been limited to a small number of genes at any one time. Given that each human genome carries over 20,000 genes, encoded by 3 billion bases, in only the most extreme familial cases of cancer could the genetic change responsible for the disease be identified. However, the past decade has seen astonishing advances in genetic technologies, so that we can now decode an entire human genome in a matter of weeks, even days, and at affordable costs.
What this means is that NGS techniques can now be applied to any person with cancer, to make sense of their personal cancer risk. It is important to note that carriage of a mutation that may increase one’s risk of cancer does not mean certainty that cancer will occur. The actual risk will vary according to the gene involved, and the effect of the mutation. Moreover, the risk is powerfully modified by our environment. This means that simple measures (like not smoking) may substantially mitigate the risk. There are many things that can potentially be done to modify cancer risk, but the task of turning the information into useful knowledge will require enormous effort. Mapping each mutation and change to its consequence is not a trivial task, let alone testing ways to protect people. This work will require large cohorts of patients in whom that mapping and testing is undertaken. These cohorts need to be meticulously annotated with respect to family history, environmental risk exposures, and personal cancer history. While such cohorts have been developed for breast cancer, bowel cancer, and ovarian cancer in particular, there is a lack of such a cohort for sarcomas.
We believe that sarcomas have a disproportionately high genetic basis, unlike cancers that are clearly very largely due to environmental factors, like lung cancer and skin cancers. This is because they particularly affect the young, suggesting people with sarcomas have a head start. In fact, much of what was learnt about familial cancer was due to studies of families affected by sarcomas.10
Why is studying personal cancer risk important? For several reasons:
- It matters if you have an increased risk of cancer. If this is known, more careful attention can be paid to picking up cancers BEFORE they become incurable. This is already standard practice for people at higher risk for breast and bowel cancer. Since sarcomas are primarily cured by surgery, it makes sense to try to detect them as early as possible, before they spread. It also matters because in some cases, this information might be used to modify treatment or imaging techniques, to minimize needless cancer risk.
- It matters if your family shares that risk, for the same reason. Moreover, because sarcomas affect young people, this information could be used to help family planning, for examples including reproductive choices and adoption.
- Our understanding of what drives cancers has largely been learnt from genetic studies in human and animal experimental models. With the development of NGS, we think such studies can now be extended from rare and exceptional circumstances, or mice, to include all who may be at increased cancer risk. Improved knowledge of cancer biology will undoubtedly bring to light previously unknown cancer biology, and form the basis of better treatments.
It was in recognition of this opportunity and need that the ISKS was established in Australia in 2008, with the support of the Rainbows for Kate Foundation.11 Kate Boyson was diagnosed with Ewing’s sarcoma in May 2007, and after five months succumbed to the disease, leaving two baby daughters. In memory of Kate, her husband, Marcus Boyson, established the Rainbows for Kate Foundation. An alliance between Associate Prof David Thomas and the Rainbows for Kate Foundation brought the ISKS to reality. The foundation generously supported the establishment of the ISKS Australia and the global study database that links and serves all international sites.
The ISKS is a unique biological, epidemiological and clinical resource created to investigate the heritable aspects of adult onset sarcoma. Initiated in Australia, the ISKS is now active across four continents with more than 800 families participating. The ISKS resources are available to all researchers to maximize efforts to improve outcomes for families with sarcoma.
How the ISKS works
People who have sarcoma (probands) are recruited from major sarcoma treatment centres, regardless of their family history of cancer. With a focus on the adult rather than the pediatric population, probands must be greater than 14 years of age and have a histologically confirmed sarcoma. Further family members are invited to participate if the proband is under 45 years of age or there is an increased incidence of cancer in the family as ascertained by study-specific criteria. An age-matched (±5 years), non-blood relative is recruited for each proband.
Participants complete a study questionnaire containing demographic, medical, epidemiological and psychosocial information, including personal history of cancer and exposure to known risk factors for sarcoma. Participants donate a blood sample and consent to access of stored biospecimens, medical records and other relevant health documentation. Upon assembly of reported family information, the process of independent verification is undertaken. Cancer diagnoses and cancer related deaths are confirmed by reference to medical records, cancer registries, death certificates or the equivalent. For each proband, treatment histories are annotated in detail and slides and blocks of tumour material are collected.
We have begun to analyze these biospecimens for genetic changes, using both conventional and Next Generation sequencing techniques. In the event a genetic finding of clinical significance is made, the family is informed that a genetic change has been detected and referred for genetic counseling and confirmatory testing at a familial cancer service.
Progress to date
The ISKS began recruitment across Australia in 2009 and since then sites have opened at the Tata Memorial Hospital, India (2010); Centre Leon Berard, France (2011); Churchill Hospital, New Zealand (2011) and the Huntsman Cancer Institute, USA (2012). Sites will shortly open at the Royal Marsden and University College Hospitals in the UK, the Vancouver General Hospital in Canada, and the Mt Sinai Hospital in New York. The World Sarcoma Network has been a useful forum for collaboration and establishment of the ISKS internationally. All aspects of the study are governed by the ISKS Global Steering Committee, comprised of representatives from each recruitment site or local study centre. The global study centre is situated at the Peter MacCallum Cancer Centre, Australia and houses the global biospecimen collection and data repository.
ISKS site investigators
A/Prof David Thomas, PI
Dr Gillian Mitchell
Ms Heather Thorne
Prof Loane Skene
Dr Kathy Tucker
Dr Craig Lewis
Prof Martin Tattersall
A/Prof Sandro Porceddu
Dr Michael Gattas
A/Prof Susan Neuhaus
Dr Richard Carey-Smith
Ms Mary-Anne Young
Dr Gillian Dite
Prof Graeme Suthers
Dr Paul James
Dr Paul Clarkson
Dr Isabelle Ray-Coquard
Prof Ajay Puri
Prof Rajiv Sarin
Dr Iain Ward
Prof Ian Judson
Dr Beatrice Seddon
Dr Charlotte Benson
Prof Bass Hassan
A/Prof Joshua Schiffman
Prof Lor Randall
Dr Bob Maki
Dr Angela Cioffi
There are currently 831 families (January 2013) enrolled in the ISKS consisting of 831 sarcoma probands (54% male), 811 family members and 403 controls. The ISKS has pedigree information such as birth date, death date and cause of death on approximately 24 000 individuals within these families. Self reported ethnicity was 80% Caucasian/white, 7% Indian,4% South East Asian including Chinese with the remainder from diverse ethnic backgrounds. The relationship status of probands was 64% partnered, 36% single. Education levels were reported as 54% of probands completing university or vocational training. There have been 697 proband blood samples collected and 716 study questionnaires completed. Family members and control subjects have contributed a further 584 blood samples and 894 questionnaires.
Clinical classification schemes for LFS/LFL families
The Li Fraumeni Syndrome (LFS), first documented in 196912 and later expanded10 described families with an autosomal dominant pattern of inheritance displaying a range of young-onset cancers. The LFS spectrum of cancers include soft tissue sarcoma, osteosarcoma, breast cancer, brain tumour, adrenocortical carcinoma, leukemia and lung cancer. It was later established that hereditable mutations in the TP53 gene were often associated with LFS.2 People with a heritable TP53 mutation have a greatly increased risk of cancer. Female mutation carriers have a 50% chance of cancer by 30yrs and a more than 90% chance by 50yrs compared to a 20% and 60% chance in males at the same ages.13-16 In order to estimate individual cancer risk and assist in the decision making for predictive genetic testing, families have traditionally been classified according to clinical criteria. The classic LFS criteria is a proband with a sarcoma diagnosed <45yrs & a first degree relative with any cancer <45yrs & a first or second degree relative with any cancer <45yrs or a sarcoma at any age. The TP53 mutation rate in families meeting this criteria is 60-70%.10 The clinically utilized Chompet criteria (mutation rate 29-35%)17,18 is a proband who has an LFS tumour <46yrs & at least one first or second degree relative with an LFS tumour (except breast cancer if the proband has a breast cancer) <56yrs or with multiple tumours, or a proband with multiple tumours (except multiple breast tumours), two of which are LFS tumours & the first of which occurred <46yrs, or a proband with adrenocortical carcinoma or choroid plexus tumour, irrespective of family history. Less stringent criteria describing Li Fraumeni like (LFL) families include Birch8,19 and Eeles8,20 with TP53 mutation rates being 22% and 8%, respectively.
Sixty-eight percent of the proband sarcomas were of soft tissue subtypes; 20% liposarcoma, 19% sarcoma NOS, 17% leiomyosarcoma, 11% fibromyxosarcoma, 10% synovial sarcoma, 4% angiosarcoma, 3% GIST and 3% MPNST with the remaining bone sarcomas comprised of 37% osteosarcoma, 29% Ewing/PNET, 27% chondrosarcoma and 7% rhabdomyosarcoma. The mean age at sarcoma diagnosis for soft tissue and bone subtypes was 51.5 years and 36.9 years, respectively. In addition to the proband sarcomas, another 2463 cancers have been reported in the ISKS families with a mean age at diagnosis of 58.0 years, significantly younger than the general population (65.6 years , Australian Institute of Health and Welfare).
A familial cancer syndrome is characterized by a recognized aggregation of certain cancer types within a family and causes family members to be at increased risk of developing cancer.21 Several familial cancer syndromes have been identified and they are often associated with mutations in a specific gene. The Li-Fraumeni syndrome (LFS; Box 2) is most often associated with sarcomas and is most often caused by mutations in the TP53 gene. Pedigree analyses of the first 630 ISKS pedigrees has shown 378 (60%) families have a history of cancer. Nine of these families meet the classical LFS criteria and 60 meet the clinically utilized Chompret LFS classification (Box 3). Of the remaining 309 families with a history of cancer, 301 are Li Fraumeni like (LFL) and there are 2 Gorlins’s syndrome, 1 familial schwannomatosis, 2 neurofibromatosis and 3 families with breast or colon cancer syndrome characteristics. There was no history of cancer in 222 ISKS families with an additional 30 families being uninformative.
Initial testing, using a combination of high resolution melt analysis and conventional Sanger sequencing, of the first 559 ISKS probands showed a heritable TP53 gene mutation in 17 families (3% of cases compared to approximately 0.0005% in the general population). Interestingly, whilst 10 of these mutations are seen in LFS or LFL families the remaining 7 families have no family history of cancer. These findings are challenging current clinical practice where family history dictates predictive genetic testing. Within the 17 families with germ line TP53 mutations there are 104 relatives (parents, siblings, children, grandchildren) that are potentially affected. To date 31 individuals have been tested with 13 found to be positive. We have begun to move from single gene studies into Next Generation Sequencing studies, to more fully explore the frequency and types of mutations that may be present—which we anticipate to exceed 1 in 6 individuals.
ISKS resources are available
The ISKS resources are available to all researchers with a peer reviewed project and human research ethics approval, subject to the ISKS Global Steering Committee approval. The ISKS holds detailed family pedigree, clinical, epidemiological, histological and molecular information. In addition, biological material includes peripheral blood DNA and FFPE tumor tissue. Several groups worldwide are currently working on projects utilizing the ISKS material (Table 1). The first publication arising from the ISKS was by Downing et al,22 and described an increased incidence of Hodgkin’s lymphoma in patients with sarcoma.
|Blood test for ALT-positive cancers||Dr Jeremy Henson||Children’s Medical Research Institute, Sydney, Australia|
|Osteosarcoma Genome-wide Association Scan||Dr Sharon Savage||National Cancer Institute, Rockville, MD, USA|
|Mapping of genetic modifier alleles in the p53 pathway in patients with sarcoma||Dr Gareth Bond||Ludwig Institute for Cancer Research, University of Oxford, UK|
|Pilot Surveillance Study in Multi-Organ Cancer prone syndromes – SMOC||Dr Gillian Mitchell||Peter MacCallum Cancer Centre, Melbourne, Australia|
|Telomere length in Familial Sarcoma||A/Prof Joanne Dickinson||Menzies Research Institute Tasmania, University of Tasmania, Australia|
|Research participants’ and healthcare professionals’ views of the feedback of genetic test information following participation in the ISKS||Ms Mary-Anne Young||Peter MacCallum Cancer Centre, Melbourne, Australia|
|Knowledge and attitudes of people with sarcoma, their family and health professionals towards genomics and genetic research||Prof Jane Halliday||Murdoch Children’s Research Institute, Melbourne, Australia|
|Whole exome sequencing of high risk sarcoma kindreds||A/Prof David Thomas||Peter MacCallum Cancer Centre, Melbourne, Australia|
|Personalized risk assessment for families with Li Fraumeni Syndrome||A/Prof Wenyi Wang||MD Anderson Cancer Centre, Houston, Texas, USA|
The ISKS is now five years old. The value of the cohort is increasing as recruitment extends to new sites globally, and as scientists are utilizing the resource to answer important questions for the lives of patients with sarcomas, and their families. The impact of philanthropic support in establishing this initiative cannot be overstated. Global collaborative research is not readily funded within any one country, and sarcomas are rarely prominent in the priorities of funding agencies. Ironically, such collaborations are particularly important for rare diseases. The ISKS is one of several examples of international collaborations fostering research into these devastating diseases.