A National Platform for Molecular Diagnostics: Results from Phase I of the Cancer Research UK Stratified Medicine Programme

Colin R Lindsay1,Emily Shaw1,James D Brenton2,Rachel Butler3,Jim Davies4,David Gonzalez de Castro5,Mike Griffiths6,Andrew Hanby7,Steve Harris4,David J Harrison8,Stephen R D Johnston9,Malcolm Mason10,Dion Morton11,Andrew G Nicholson12,Karin Oien13,Doris Rassl14,Jane Rogan15,Ian Walker1,Peter W M Johnson1

1Cancer Research UK, London, UK,2Cancer Research UK Cambridge Institute, Cambridge, UK,3Institute of Medical Genetics, Cardiff, UK,4University of Oxford, Oxford, UK,5The Centre for Molecular Pathology, Surrey, UK,6West Midlands Regional Genetics Laboratory, Birmingham, UK,7Leeds Institute of Cancer and Pathology, Leeds, UK,8University of St. Andrews, St. Andrews, UK,9The Royal Marsden NHS Foundation Trust, London, UK,10Cardiff University, Cardiff, UK,11The Queen Elizabeth Hospital, Birmingham, UK,12Royal Brompton and Harefield Hospitals NHS Foundation Trust, London, UK,13NHS Greater Glasgow and Clyde, Glasgow, UK,14Papworth Hospital NHS Foundation Trust, Cambridge, UK,15Manchester Cancer Research Centre, Manchester, UK

Presenting date: Tuesday 3 November
Presenting time: 16.10-16.25


There is an increasing demand for early analysis of multiple prognostic and predictive genetic markers in patient tumour samples. Phase I of the Cancer Research UK Stratified Medicine Programme (SMPI) was designed to develop a platform to incorporate molecular diagnostics into the normal pathway of patient care and demonstrate the applicability of a nationwide testing network, with results linked to healthcare records and held in a central database.


From August 2011 to June 2013, patients with breast, colorectal, prostate, lung or ovarian cancer, or metastatic melanoma, were approached at 26 hospitals for consent to molecular testing of tumour samples in three centres. Formalin-fixed paraffin-embedded sections and peripheral blood samples were collected for analysis of a small panel of molecular aberrations in driver genes. Results were transmitted directly to participating clinical centres for inclusion in medical records. An agreed clinical dataset was obtained for all patients and stored using the English national cancer registration system. Sequence mutations were assessed by Sanger sequencing, pyrosequencing or screening methods (BRAF, DDR2, EGFR, KIT, KRAS, NRAS, PIK3CA, PTEN, TP53). Gene rearrangements were assessed by fluorescent in situ hybridisation (ALK and TMPRSS2-ERG).


10754 patients (98% of those approached) consented, with 9010 samples sent for analysis. Confirming a level of quality in the clinical data and corresponding genetic analysis, the results across all six cancers are consistent with what has been reported previously: examples include near-mutual exclusivity of EGFR mutation (8.3% of samples) and ALK rearrangement (1.9%) in NSCLC (1889 samples in total), and KRAS (39%), NRAS (4%) and BRAF mutation (11%) in colorectal cancer (1605 patients). We also observed significantly high levels of TP53 mutation in late stage breast (75% incidence in stage IV, 1873 patients) and colorectal cancer (64% incidence in stage IV), and high levels of BRAF mutation (39%, 52 patients) in mucinous colorectal cancer. With large patient numbers facilitating an examination of gene-gene and genetic-clinical correlations on a largely unselected population-wide level, SMP1 also offers a number of novel insights which will require future validation: a high incidence of KRAS mutation (37%) in UK lung adenocarcinoma, a significant elevation of TP53 mutation in ductal carcinoma of the breast (27%; 1425 patients) compared to lobular carcinoma of the breast (3%, p=<0.0001; 179 patients), and similar rates of genetic aberrations between superficial spreading and nodular melanomas. Correlations between survival and genetic modification will be presented with a minimum of 2 years follow up.


Here we report the first UK-wide study assessing the molecular drivers of cancer within nationalised healthcare systems. We have confirmed the feasibility of incorporating molecular diagnostics in the normal pathway of UK cancer care. SMP1 features large patient numbers and offers a unique dataset, unrepresented in other genomic studies: one that is not subject to planned selection bias and features large patient numbers.