A comparison of bone-related biomarkers and CA27.29 to assess response to treatment of osseous metastatic breast cancer.

BACKGROUND: The assessment of bone metastases by clinical examination or imaging techniques is still considered unreliable. We compared a specific marker of bone resorption, urinary deoxypyridinoline (DPD)-crosslinks, with serum calcium (Ca), alkaline phosphatase (AP) and CA27.29, to evaluate the status of bone metastases in patients with breast cancer.

MATERIALS AND METHODS: Second morning voided urine was collected from 2 groups of patient (pts), those without evidence of disease (n = 118), and those with bone metastases (n = 85) under specific therapy plus pamidronate. DPD and CA27.29 were measured on the automated ACS180 system (Bayer Diagnostics, Tarrytown, NY, USA). Receiver operating characteristics (ROC) curves were established for each of the 4 biomarkers to determine whether they could distinguish the 2 subsets of pts with clinically sufficient validity, and to establish the corresponding cut-off values.

RESULTS: Neither Ca nor AP was useful in discriminating the 2 subgroups. At a DPD cut-off of 13 nmol/mmol, we found a specificity of 69% and a sensitivity of 53% for diagnosing bone metastases. Best results, however, were seen for CA27.29. A cut-off value of 30 U/ml resulted in a specificity of 62% and a sensitivity of 81%.

CONCLUSIONS: CA27.29 was the best parameter for the discrimination of stage IV breast cancer with bone metastases. The primary advantage of DPD lies in the monitoring of bone metastases under specific therapy.