The promise and challenges of rare cancer research.

136 www.thelancet.com/oncology Vol 17 February 2016 It is commonly stated that 20–25% of all oncology patients have a diagnosis of rare cancer. The US National Cancer Institute defi nes rare cancers as those with an incidence of less than 15 cases per 100 000 per year. On the basis of this defi nition, Greenlee and colleagues analysed data for more than 9 million adults in whom cancer was diagnosed between 1995 and 2004, and reported that 60 of 71 cancer types, primarily defi ned by anatomical location, were classed as rare. Those cancers accounted for 25% of all adult cancer diagnoses. In a subsequent study, the Surveillance of Rare Cancers in Europe (RARECARE) consortium analysed European population-based cancer registry data for patients with cancer diagnosed between 1988 and 2002. In that study, rare cancers were defi ned as those with an incidence of less than six per 100 000 per year, and a newly generated list of cancer types was used; 22% of all new cancer diagnoses were classed as rare cancers. With evolving understanding of cancer biology and the associated changes in the taxonomy of malignant disease, these numbers now underestimate the incidence of rare cancers. Large-scale eff orts, including the International Cancer Genome Consortium, the Cancer Genome Atlas, and the Cancer Genome Project, have systematically catalogued genomic alterations in diff erent cancer types and have provided the impetus for changing the classifi cation of cancers from histologically based to molecularly based. As part of a Series of three papers on rare tumours in The Lancet Oncology, Niki Boyd and colleagues describe the eff ects of this change. As a result of the reclassifi cation of common cancers based on molecular and genomic markers, rare cancers are becoming increasingly common. For example, melanoma, which has an overall estimated incidence of 73 870 (23 cases per 100 000) in the USA in 2015, is not a rare cancer as defined by the US National Cancer Institute. However, on the basis of work by The Cancer Genome Atlas and others, cutaneous melanoma can now be divided into distinct molecular subtypes, including those harbouring BRAF mutations, NRAS mutations, NF1 mutations, or no mutations in these genes. The annual incidences of these molecularly defined melanoma subtypes are 12·0 per 100 000, 6·5 per 100 000, 3·2 per 100 000, and 1·4 per 100 000, respectively. Under this new taxonomy, melanoma is no longer classed as one common cancer: rather, it is a collection of rare cancers, each with a unique biology with distinct implications for therapy—an increasingly common pattern in oncology. Outcomes in patients with rare cancers are worse than those in patients with more common tumour types. For individuals whose cancers were diagnosed between 1995 and 1999, 5-year relative survival was 47% for those with rare cancers and 65% for those with common cancers. Some of the key issues contributing to poor outcomes in rare cancers include diffi culties or delays in diagnoses, limited access to centres with clinical expertise, less eff ective standard treatments, and inadequate funding for preclinical and clinical research programmes. In the second paper in this Series, Jean-Yves Blay and colleagues outline the importance of collaboration between investigators, research consortia, industry, regulatory bodies, and patient advocacy groups to overcome these challenges and enable the advancement of drug development and improve outcomes for patients with rare cancers. The rapid accrual of patients in the AstraZenecasponsored phase 3 trial of standard chemotherapy with or without selumetinib for advanced uveal melanoma (SUMIT) shows the potential of eff ective collaboration between industry, academia, and patient advocacy groups. In the National Cancer Institute Cancer Therapy Evaluation Program’s randomised phase 2 trial of selumetinib alone versus chemotherapy, 101 patients with advanced uveal melanoma were randomly assigned at 15 centres within 3 years. By contrast, in SUMIT, 152 patients were screened and 129 randomly assigned from 29 centres within 9 months. This increase in accrual in a shorter amount of time was achieved by doing the trial in 11 countries, identifying and targeting key referral centres, and partnering with doctors committed to advancing the treatment of uveal melanoma and patient advocacy groups such as CureOM (an ocular melanoma advocacy initiative of the Melanoma Research Foundation). Despite this success, international trials in rare tumours are still associated with challenges that are The promise and challenges of rare cancer research