PURPOSE: The overexpression of cyclooxygenase-2 (COX-2) is associated with a worse prognosis and the development of distant metastases in cervical cancer. This matched-pair analysis examined whether COX-2 expression is associated with para-aortic lymph node (PALN) recurrence in uterine cervical cancer treated with radiotherapy (RT). METHODS AND MATERIALS: For this study, we matched 20 patients with PALN recurrence after definitive or postoperative RT by stage with 20 others who did not have PALN recurrence. Of the 20 patients with PALN recurrence, definitive or postoperative RT was performed in 11 and 9 patients, respectively. COX-2 expression was assessed immunohistochemically using a mouse monoclonal antibody on formalin-fixed paraffin-embedded tumor specimens taken before RT. A logistic regression model was used to predict for PALN recurrence. RESULTS: COX-2 was expressed in 28 (70%) of the 40 patients. The staining intensity was as follows: weak in 19 (47%), moderate in 6 (15%), and strong in 3 (8%) patients. The patients with PALN recurrence had much greater expression of COX-2 (17 patients, 85%) than did the control group (11 patients, 55%; p = 0.04). Strong staining intensity of COX-2 was seen only in the PALN recurrence group. The statistically significant factors associated with PALN recurrence were positive pelvic lymph nodes (odds ratio, 7.61; 95% confidence interval, 1.55-37.37; p = 0.01) and COX-2 expression (odds ratio, 1.47; 95% confidence interval, 1.04-2.09; p = 0.03). CONCLUSION: Our findings suggest that COX-2 overexpression in the initial tumor tissue might be associated with PALN recurrence after RT in cervical cancer patients.
PURPOSE: The overexpression of cyclooxygenase-2 (COX-2) is associated with a worse prognosis and the development of distant metastases in cervical cancer. This matched-pair analysis examined whether COX-2 expression is associated with para-aortic lymph node (PALN) recurrence in uterine cervical cancer treated with radiotherapy (RT). METHODS AND MATERIALS: For this study, we matched 20 patients with PALN recurrence after definitive or postoperative RT by stage with 20 others who did not have PALN recurrence. Of the 20 patients with PALN recurrence, definitive or postoperative RT was performed in 11 and 9 patients, respectively. COX-2 expression was assessed immunohistochemically using a mouse monoclonal antibody on formalin-fixed paraffin-embedded tumor specimens taken before RT. A logistic regression model was used to predict for PALN recurrence. RESULTS:COX-2 was expressed in 28 (70%) of the 40 patients. The staining intensity was as follows: weak in 19 (47%), moderate in 6 (15%), and strong in 3 (8%) patients. The patients with PALN recurrence had much greater expression of COX-2 (17 patients, 85%) than did the control group (11 patients, 55%; p = 0.04). Strong staining intensity of COX-2 was seen only in the PALN recurrence group. The statistically significant factors associated with PALN recurrence were positive pelvic lymph nodes (odds ratio, 7.61; 95% confidence interval, 1.55-37.37; p = 0.01) and COX-2 expression (odds ratio, 1.47; 95% confidence interval, 1.04-2.09; p = 0.03). CONCLUSION: Our findings suggest that COX-2 overexpression in the initial tumor tissue might be associated with PALN recurrence after RT in cervical cancerpatients.