Since the first clinical use of tumour markers, quality assurance has been
considered only in a restrictive manner, that is, as a surveillance of the
analytical process. In other words, quality assessment was roughly viewed a
s a synonymous with quality control. This is not surprising, since tumour m
arkers are almost exclusively assayed by radioimmunoassays, whose analytica
l performance were suboptimal in the 1970s. Furthermore, tumour marker conc
entrations in biological fluids were very low (in the ng range); in additio
n, primary standards were not available and dose-response curves were set u
p with conventional calibrators. Therefore, quality control programmes have
become mandatory to restrict intra-and inter-laboratory variability. (C) 2
000 Harcourt Publishers Ltd.