Quality control for estrogen and progesterone receptor assay in human breast cancer: the influence of computation methods on intra and interlaboratory variability.

: The importance of evaluating receptors for estrogen and progestin in human breast cancer has been pointed out by many authors. In the absence of a reference standard, receptor assays must be controlled by intra and interlaboratory quality control programs. Much interlaboratory variability exists due to non-uniform analytical protocols, non-uniform ligands, intrinsic errors and also errors in computation methods. The goals of our Italian Quality Control Program on Multicenter Trials are to standardize the analytical procedures and computation methods. Twenty Italian laboratories participated in the Quality Control Program. Each specimen was assayed for steroid receptor content according to the standardized dextran-coated-charcoal method. Data were subjected to computerized analyses by 5 different methods of calculation (Scatchard plot, direct plot, Lineweaver-Burk method, Brunauer-Emmet-Teller analysis, single-point approach). The results were than evaluated to identify intra- and inter-assay variation coefficients and to define other statistical parameters. The authors suggest different calculation methods depending on the specific experimental and/or physiopathological conditions.