Detection of occult metastatic breast cancer cells in blood by a multimolecular marker assay: correlation with clinical stage of disease.

ABSTRACT

Currently, molecular markers offer the unique opportunity to identify occult metastasis in early stage cancer patients not otherwise detected with conventional staging techniques. To date, well-characterized molecular tumor markers to detect occult breast cancer cells in blood are limited. Because breast tumors are heterogeneous in tumor marker expression, we developed a "multimarker" reverse transcription-PCR assay combined with the highly sensitive electrochemiluminescence automated detection system. Breast cancer cell lines (n = 7), primary breast tumors (n = 25), and blood from normal donors (n = 40) and breast cancer patients [n = 65; American Joint Committee on Cancer (AJCC) stages I-IV] were assessed for four mRNA tumor markers beta-human chorionic gonadotropin (beta-hCG), oncogene receptor (c-Met), beta 1-->4-N-acetylgalactosaminyl-transferase, and a tumor-associated antigen (MAGE-A3). None of the tumor markers were expressed in any normal donor bloods. Breast cancer cell lines and primary breast tumors expressed beta-hCG, c-Met, beta 1-->4-N-acetylgalactosaminyl-transferase, and MAGE-A3 mRNA. Of the 65 breast cancer patient blood samples assessed, 2, 3, 15, 49, and 31% expressed 4, 3, 2, 1, and 0 of the mRNA tumor markers, respectively. At least two markers were expressed in 20% of the blood specimens. The addition of a combination of markers enhanced detection of systemic metastasis by 32%. In patient blood samples, the MAGE-A3 marker correlated significantly with tumor size (P = 0.0004) and AJCC stage (P = 0.007). The combination of beta-hCG and MAGE-A3 mRNA markers correlated significantly with tumor size (P = 0.04), and the marker combination c-Met and MAGE-A3 showed a significant correlation with tumor size (P = 0.005) as well as AJCC stage (P = 0.018). A multimarker reverse transcription-PCR assay that correlates with known clinicopathological prognostic parameters may have potential clinical utility by monitoring tumor progression with a blood test.