Quantification of MUC1 in breast cancer patients - A method comparison study

Objective: To compare the performance of four serum assays for the quantifi cation of MUC1 in breast cancer patients. Study design: A total of 282 seru m samples were evaluated with two automated (Boehringer Mannheim Enzymun-Te st(R) CA 15-3 and Chiron ACS(TM) BR) and two manual assays (Centocor CA 15- 3(R) radioimmunoassay [RIA] and Biomira Truquant(R) BR RIA(TM)). Sera were obtained from healthy controls (n=50), patients with benign (n=25) and mali gnant breast disease (n=77) and patients with other malignancies (n=69). In addition, sera from pregnant women (n=56) and patients with liver cirrhosi s (n=5) were included. Results: Intraassay coefficients of variation (C.V.s ) were highest for the manual Centocor CA 15-3 assay (7.4% for values below 50 kU/1 and 8.1% for values above 180 kU/1). Interassay C.V.s were highest for the manual Truquant BR assay (11.7% for the lower concentration values and 18.6% for the higher concentration values). False positive rates range d between 0% for the Centocor CA 15-3 RIA and 14% for the ACS BR assay (cut -off: 30 kU/1). In monitoring breast cancer patients all four assays show s imilar patterns, although absolute MUC1 values found may differ up to 50%. Conclusion: For monitoring purposes all assays perform equally well, howeve r, automated assays show lower inter- and intraassay variability, especiall y in the higher value range. Therefore we recommend the use of the same, au tomated, assay for quantification of MUC1 during the follow-up of breast ca ncer patients. (C) 1999 Elsevier Science Ireland Ltd. All rights reserved.