Concordance analysis of paired cancer antigen (CA) 15-3 and 27.29 testing.

Abstract

Cancer antigens (CA) 15-3 and 27.29 are used in the clinical management of many breast cancer patients. Given that immunoassays for CA 15-3 and CA 27.29 target epitopes on the same glycoprotein-Mucin 1 (MUC1)-the present analysis was conducted to evaluate the potential concordance of tumor marker results when both tests were ordered by providers on the same specimens. A retrospective limited dataset of paired CA 15-3 (Roche Diagnostics) and CA 27.29 (Siemens Diagnostics) test results was obtained from a national clinical reference laboratory. Concordance according to reference interval (RI) status and percent (%) change between consecutive test results was analyzed. 37,652 paired results from 12,470 distinct patients were obtained. The correlation between CA 15-3 and CA 27.29 results was high (correlation coefficient: Pearson, 0.967), although across the dataset a significant difference between CA 15-3 and CA 27.29 results was observed (P<0.05). RI concordance between CA 15-3 and CA 27.29 results was observed in 93.7% of pairs (35,280 of 37,652). Correlation was also observed in the % change of CA 15-3 and CA 27.29 results between consecutive specimens for individual patients. Using doubling or halving thresholds (i.e., 100% increase or 50% decrease), concordance in % change was observed between CA 15-3 and CA 27.29 in approximately 90% of cases. Individual patient results trended similarly across both markers over time. While generally concordant, CA 15-3 and CA 27.29 results should not be used interchangeably. The present report provides no evidence for added value in performing both tests routinely for individual patients.