Preoperative evaluation of macrophage colony-stimulating factor levels in patients with endometrial cancer

OBJECTIVE: Our purpose was to examine the relationship between preoperative serum levels of macrophage colony-stimulating factor, alone and in combination with CA 125, and the presence of prognostic clinicopathologic factors and subclinical metastases in women with endometrial cancer. STUDY DESIGN: Ninety-two women who underwent primary exploration for endometrial adenocarcinoma had preoperative serum samples evaluated for macrophage colony-stimulating factor and CA 125 levels. Multivariate analysis was used to determine the associations of surgicopathologic findings with macrophage colony-stimulating factor and CA 125 levels. Logistic regression analysis was used to identify factors associated with the risk of extrauterine disease. The association of macrophage colony-stimulating factor and CA 125 levels with stage, grade, and depth of myometrial invasion and histologic characteristics were analyzed with Fisher's two-tailed exact test. RESULTS: Elevated levels of macrophage colony-stimulating factor were not associated with depth of myometrial invasion, histologic grade, or histologic cell type; however, advanced stage (p = 0.02) and the presence of lymph node metastases (p = 0.04) were associated with elevated levels. Sensitivity and specificity of macrophage colony-stimulating factor for predicting extrauterine disease were 42% and 89%, respectively. If either an elevated macrophage colony- stimulating factor or an elevated CA 125 level was used to predict extrauterine disease, the sensitivity was increased to 67% but the specificity was decreased to 78%. Macrophage colony-stimulating factor elevations predicted lymph node metastases with a sensitivity of 50% and a specificity of 86%. A multivariate regression model showed CA 125 to be the most significant predictor of extrauterine disease; macrophage colony- stimulating factor also contributed prognostic information (p = 0.02). The sensitivity and specificity of the multivariate model for predicting extrauterine disease were 75% and 73%, respectively. CONCLUSION: Macrophage colony-stimulating factor and CA 125 are neither sensitive nor specific enough to be used as predictors of the presence or absence of extrauterine disease in patients with endometrial cancer.