Evidence for the clinical use of tumour markers

Michael J Duffy
DOI: https://doi.org/10.1258/0004563041731529
2004-09-01
Abstract:Testing for tumour markers should only be performed if it results in a better patient outcome, increased quality of life or reduced overall cost of care. Ideally, the clinical value of a tumour marker should be validated in a large prospective study or a meta-analysis of small-scale retrospective/prospective studies (i.e. a level 1 evidence study) prior to routine use. Markers that have been validated in such a level 1 evidence study include carcinoembryonic antigen in the surveillance of patients with diagnosed colorectal cancer, alphafetoprotein, human chorionic gonadotrophin and lactate dehydrogenase for evaluating prognosis in patients non-seminomatous germ cell tumours, CA 125 for monitoring therapy in patients with ovarian cancer, oestrogen receptors for predicting response to hormone therapy in breast cancer, HER-2 for predicting response to trastuzumab in patients with advanced breast cancer and urokinase plasminogen activator/plasminogen activator inhibitor type 1 for determining prognosis in breast cancer. Although currently in widespread use, the value of prostate-specific antigen in screening for prostate cancer has yet to be validated in a large prospective randomized trial.
What problem does this paper attempt to address?