The first full day of the NCRI conference started with two more contrasting plenary lectures. Jeff Settleman from Genentech in California spoke about the use of cell lines derived from tumour cells to model drug sensitivity and resistance. This addresses two of the most intractable problems in cancer drug development: matching patients to appropriate therapies, and dealing with the almost inevitable development of drug resistance. Both these problems are caused by the heterogeneity of tumours, the first by molecular differences between patients with similar tumour histology and the second by differences between cells within a single tumour. Both can be modelled in vitro using tumour-derived cell lines.

Settleman then presented a number of different examples. In studying heterogeneity between patients his group has used large panels of cell lines (one example including 130 different lines from non small cell lung cancer tumours) and subdivided them into those with genetic changes that predict sensitivity or resistance to a particular drug treatment. In just one example, the gene NRG1 is expressed much more strongly in tumours that are sensitive to the drug Lapatinib than in those that are resistant to it.

Multiple molecular mechanisms are known to be involved in the development of drug resistance. A key driver, however, is simple Darwinian evolution; a drug that kills the majority of cells in a tumour will leave behind a population of "fit" cells known as drug tolerant persisters (DTPs) that will multiply, re-forming the tumour. These cells share some properties with cancer stem cells. Finding ways to target these cells preferentially may delay or even prevent the development of a more stable drug-resistant phenotype.

Sir Mike Richards, Clinical Director for Cancer and End of Life Care in England, spoke eloquently about the need to improve cancer survival in England, drawing on evidence for where the problems occur and offering practical solutions. It is fairly well known that England (and the other countries of the United Kingdom) lag seriously behind the best-performing developed countries in cancer survival rates; it is estimated that improving performance to match those countries would save about 10,000 lives each year. Interestingly, however, the prognosis for patients who survive for at least one year in England is very close to that in similar countries. The English problem is very much one of early detection and diagnosis.

Richards went on to discuss interventions that would help improve this early part of the English "cancer pathway", from patient awareness and self-referral through the GP system to diagnostic procedures. One important aim is to reduce the high proportion of patients first presenting as emergencies, as these have particularly poor prognosis. Campaigns to improve public awareness of key cancer symptoms are now being rolled out across England, starting with bowel cancer. Improving the availability and take-up of screening programmes is another important target.

The topic of molecular heterogeneity between tumours of similar histology was picked up later in the morning in a well-attended symposium on stratified medicine. Two of the three speakers, Alberto Bardelli of the University of Torino, Italy and Jean Charles Soria of Institute Gustave Roussy, Paris, France, discussed molecular markers of drug response in colorectal and non-small cell lung cancer respectively; the third, David Gonzales de Castro from the Institute of Cancer Research in London presented some practical implications for implementing molecular diagnostics in standard clinical practice.The British Association for Cancer Research, the largest membership organisation for cancer research in the UK, offers an annual Translational Research Award to a researcher under the age of 35 who has made a significant contribution to research bridging the laboratory and the clinic. The winning applicant is always invited to give a presentation at the NCRI conference, This year, the prize was awarded to James Flanagan from Imperial College, London. His talk, intriguingly entitled "Epigenetic Epidemiology", was given on Monday afternoon, in a wide-ranging session on the overall theme of Diagnosis and Therapy.

Flanagan explained that epigenetic features of the genome – that is, aspects of gene expression that do not arise from changes to the base sequence of DNA – can contain markers for cancer risk and prognosis. One of the most important epigenetic changes is the methylation of DNA bases, which can act as an "on-off switch" for gene expression. Flanagan compared patterns of DNA methylation in candidate genes in blood cells taken from participants in a prospective study of breast cancer, and found one marker in the promoter region of the gene ATM to be a prognostic marker for breast cancer risk, particularly for younger women.

This was one of three parallel sessions presenting submitted abstracts deemed to be among the best. Other noteworthy papers were given by Rosalind Eeles of the Institute of Cancer Research, Sutton, UK (who also chaired the session); Xueying Mao, from Queen Mary Hospital, London, UK, and Diana Eccles from the University of Southampton, UK. Eeles presented results from the international PRACTICAL prostate cancer consortium that identified and validated seven further genetic markers of susceptibility to this condition. Mao described an intriguing study showing that there were clear differences between the genetic changes associated with prostate cancer in Chinese patients compared to Western ones. And Eccles described findings from the POSH study of breast cancer patients diagnosed at or below the age of 41 showing that patients carrying a BRCA1 mutation had improved distant relapse free survival. This might indicate that tumours bearing BRCA1 deletions are more sensitive than others to anthracycline based chemotherapy.

Another highlight of the afternoon was the regular Clinical Trials Showcase, which included four presentations of recent trials that all, in the words of session chair Matt Seymour of the National Cancer Research Network in Leeds, had produced "really important and [clinical] practice-changing results".

Bone metastases from prostate cancer have sometimes been treated with radio-isotopes emitting beta particles. This treatment was fairly effective but highly toxic, the toxicity resulting from the long range of the radioactive particles. Alpha particle emitters are potentially both more effective and less toxic, as the alpha particle has a much shorter range and can be targeted more precisely. Radium-223 chloride is an alpha-emitting radio-isotope that is first-in-class for this patient population. Christopher Parker of the Royal Marsden Hospital, London, UK presented some very promising results from the ALSYMPCA Phase III trial that compared radium-223 and best standard care to best standard care alone in men with progressive, castration-resistant prostate cancer and bone metastases, Median overall survival was 11.2 months in the placebo arm and 14.0 months in the radium arm, a highly significant result (p=0.00185) that led to the trial being stopped early. The side effect profile was highly favourable, and Parker concluded that this therapy may become part of a new standard of care for these patients.

James Larkin, also from the Royal Marsden Hospital, presented results from the BRIM3 Phase III trial comparing a novel BRAF inhibitor, vemurafenib, with dacarbazine in patients with advanced malignant melanoma bearing the common BRAF mutation V600E. These patients are very poorly served by conventional treatments, with typical survival times of just 6-8 months that have been almost unchanged for decades. Hazard ratios for overall survival (0.44) and progression free survival (0.26) very strongly favoured vemurafenib; the most dramatic results were seen for progression free survival, with median times of 5.3 months in the vemurafenib arm compared to 1.6 months for dacarbazine. Although vemurafenib has disadvantages including a significant side effect profile and high cost, it still represents an important advance in treatment options for these difficult to treat patients.