Identification of in-vivo generated human autoantibodies for the screening and early detection of primary colorectal cancer (CRC)

点击翻译按钮获取中文摘要

CRC is the second highest cause of cancer mortality in the western world: the most important determinant of survival is stage of disease when diagnosed, with over 90% survival for patients with early stage colorectal cancer. Randomised trials, involving faecal blood testing followed by colonoscopy (use of a surgical endoscope to examine the colon), have shown early detection with appropriate treatment results in mortality reduction of ~16%. Unfortunately the impact of such interventions is limited by a number of factors, including that fact that the majority of patients are not diagnosed until late stage disease, the low patient acceptability of home faecal sampling required for current tests (Uptake: UK 57%; European 34%) and the low sensitivity and specificity of existing faecal tests. Similarly, although an alternative colonic imaging method (sigmoidoscopy) appears to give higher mortality reductions (22%-31%), patient acceptability remains low (45%-58%) with significant social inequalities. Furthermore the examination is usually limited to only part of the colon (distal colon). Our solution is to provide a blood test that will improve public acceptability while also providing improved sensitivity & specificity compared to faecal screening. This will significantly improve clinical outcomes (improved survival rates) and cost-effectiveness.The immune system, which protects us from microbes, also mounts a response to molecules overproduced and released by cancer cells within a tumour. This response includes the generation of antibodies to these tumour-associated molecules. Because the molecules recognised by the antibodies originate from the patient, the antibodies are termed autoantibodies. We, and other groups, have good evidence that detecting these autoantibodies in the blood of patients can provide a route to improved methods for early detection of tumours. Our approach will require only a blood sample, avoiding faecal sampling or invasive imaging colon techniques. As such it is highly likely to dramatically increase patient acceptability, which in itself will improve detection rates and at the same time simplify the testing process and provide improved cost-effectiveness for the NHS.Our approach is to produce an assay enabling us to test for autoantibodies to multiple tumour molecules simultaneously. This in the form of a microarray (regularly spaced spots of selected tumour-associated molecules on a glass slide surface). Serum from patient blood is washed over the array, and if antibodies are present, they bind to one or more of the tumour-associated molecule spots. Bound antibodies are themselves detected with a fluorescent reporter and the signal intensity measured. Comparison to cutoffs based upon examination of known colorectal cancer patients and normal controls allows the identification of samples positive for tumour-molecule autoantibodies, indicating that further investigation is required. The small size of the microarray and spots allows for replication of spots for each tumour-associated molecule on the array, yet requires significantly reduced chemical volumes over tests which examine only one tumour-associated molecule per test.The potential of this microarray-based autoantibody screening in colorectal cancer is demonstrated in our preliminary work, using a limited panel of 32 tumour-associated molecules and a small sample set of 262 known cancer/normal individuals.This project will identify the optimal panel from the 32 tumour-associated molecules which are suitable for use as a colorectal cancer screening tool and validate that resultant panel using much larger, clinically well-defined sample sets than our preliminary study. Our prediction is that with a patient compliance rate of >90%, test specificity of 85% and sensitivity of 65% this would result in detection of 58.5% of the total number of CRCs -more than a doubling of the total number of CRCs currently detected.