OBJECTIVES: A panel composed of three immunomarkers, cytokeratin 19, HBME-1 and galectin-3, is recommended for improving the accuracy of diagnosing papillary thyroid carcinoma (PTC), but the results are sometimes inconsistent and difficult to interpret. We used immunocytochemistry to assess the utility of chemokine CXC ligand 12 (CXCL12) as a novel diagnostic marker for PTC employing paraffin-embedded cell blocks. METHODS: We analyzed the expression of CXCL12 using immunocytochemical staining in 82 cases of thyroid lesions (47 PTCs and 35 thyroid lesions other than PTC). To determine the optimal cut-off value for the assessment of CXCL12 positivity, we used receiver operating characteristics (ROC). RESULTS: ROC curves showed that the optimum diagnostic cut-off was 10% (area under the curve 0.950, 95% confidence interval 0.891-1.008), with exclusive CXCL12 expression in PTC compared to other thyroid lesions (p < 0.001). In total, more than 90% of the PTCs were associated with CXCL12 immunohistochemical staining, while only up to 11.4% of thyroid lesions other than PTC were positive for CXCL12. The follicular variant PTC showed 90% CXCL12 expression compared to 10.5% positivity in follicular neoplasm. CONCLUSION: Our findings indicate that CXCL12 may be an effective supplementary diagnostic marker for PTC in preoperative fine-needle aspiration cytology using the cell block method.
OBJECTIVES: A panel composed of three immunomarkers, cytokeratin 19, HBME-1 and galectin-3, is recommended for improving the accuracy of diagnosing papillary thyroid carcinoma (PTC), but the results are sometimes inconsistent and difficult to interpret. We used immunocytochemistry to assess the utility of chemokine CXC ligand 12 (CXCL12) as a novel diagnostic marker for PTC employing paraffin-embedded cell blocks. METHODS: We analyzed the expression of CXCL12 using immunocytochemical staining in 82 cases of thyroid lesions (47 PTCs and 35 thyroid lesions other than PTC). To determine the optimal cut-off value for the assessment of CXCL12 positivity, we used receiver operating characteristics (ROC). RESULTS: ROC curves showed that the optimum diagnostic cut-off was 10% (area under the curve 0.950, 95% confidence interval 0.891-1.008), with exclusive CXCL12 expression in PTC compared to other thyroid lesions (p < 0.001). In total, more than 90% of the PTCs were associated with CXCL12 immunohistochemical staining, while only up to 11.4% of thyroid lesions other than PTC were positive for CXCL12. The follicular variant PTC showed 90% CXCL12 expression compared to 10.5% positivity in follicular neoplasm. CONCLUSION: Our findings indicate that CXCL12 may be an effective supplementary diagnostic marker for PTC in preoperative fine-needle aspiration cytology using the cell block method.