Tumor markers in metastatic breast cancer: High tumor specificity within the reference range

Abstract

10554 In breast cancer as well as in other solid tumors all “tumor markers” actually available are not tumor specific but can be measured in different extent also in all healthy individuals. By consequence any fixed cut off value used must lead to false positive and negative test results for the individual patient. In a retrospective analysis we investigated the diagnostic capacity of CEA and CA 15–3 for the early detection of metastatic breast cancer comparing the use of fixed cut off values with kinetics. After complete removal of the tumor (R0 resection) breast cancer patients reach 4 weeks after the end of adjuvant radio/chemotherapy their individual “normal” values: these median values for CEA (Abbott, AxSYM) and CA 15–3 (Roche, Elecsys) are comparable to the medians of healthy individuals (CEA: 1 ng/ml; CA 15–3: 13 U/ml). A reproducible increase of 100% of at least one of these markers was postulated as indication for metastatic disease. A total of 899 breast cancer patients with a non metastatic primary tumor and available individual baseline values after primary treatment was included. Of these 717 remained disease-free and served as control group, whereas 182 patients developed metastatic disease during follow-up. On this basis we reached a specificity of more than 98% for both biomarkers and a sensitivity of 16% for CEA alone, 24% for CA 15–3 alone and another 25% for both markers resulting in 66% true positives or 34% false negatives. Using fixed cut off values (CEA: 4 ng/ml and CA 15–3: 30 U/ml) we ended up with 19% false positives (specificity 81%) and a sensitivity of 9% for CEA, 30% for CA 15–3 and another 27% for both resulting also in 34% false negatives. Using higher cut off values where the probability of metastatic breast cancer is increased (CEA: 6 ng/ml and CA 15–3: 60 U/ml) we reached a specificity of 96% and a sensitivity of 10% for CEA and 19% for CA 15–3, another 14% for both, resulting in 57% false negatives. We conclude that the interpretation of these markers in follow-up care using individual baseline values and kinetics leads to a significant superior profile of specificity and sensitivity. No significant financial relationships to disclose.