The relationship between tumor markers and histopathologic and demographic features of patients with breast cancer and their distinctive role in metastatic process.
e11540 Background: Tumor marker monitoring is generally performed by the physicians, although in many guidelines, it has not been recommended for breast cancer. In this study, we aimed to research the role of CEA and CA15.3 levels in metastatic process. Methods: In between years 2000 January and 2011 August, the documents of 482 female patients followed by breast cancer diagnosis in Medical Oncology and Radiation Oncology clinics of Kartal Dr. Lutfu Kirdar Education and Research Hospital, were evaluated retrospectively. Results: To determine the role of CEA and CA15.3 levels in evaluation of metastatic process, ROC analysis was performed and its cut-off values were 1.3 ng/ml and 14 U/ml, respectively. Sensitivities of CA15.3 and CEA levels were detected as 83.33% and 87.50%; specificities were 40.80% and 41.14%, respectively. Evaluation of the relationship between CEA and CA15.3 levels and the status of hormone receptors by ROC analysis showed the cut-off values as 26 U/ml and 1,6 ng/ml, respectively. In determination of endocrine sensitivity, the sensitivities of tumor markers elevation were 88.17% and 54.12%, respectively. Their specificities were 26.92% and 73.08%. CEA level elevation in hormone sensitive patients was statistically significant (p: 0.007). When sensitivity and specificity of CA15.3 and CEA levels on detection of c-erbB2 positivity were evaluated by ROC analysis, cut-off values were determined as 17 U/ml and 1,7 ng/ml. Sensitivities of these distinctive cut-off values were found as 58.8% and 71.15% and specificities were 60% and 51.67% (p: 0.017). When the relation between CA15.3 and CEA levels and node involvement was evaluated by ROC analysis, cut-off values were respectively 16 U/ml and 0.9 ng/ml. Their sensitivities were 53.45% and 87.93% and specificities were 55.56% and 51.67%, respectively. Conclusions: The levels of tumor markers have been suggested to have a significant role in prognosis of the disease with respecttothe correlation between different histopathological and clinical features and tumor markers. Randomised controlled prospective studies planned and performed in this issue may clarify this uncertainity.