OBJECTIVE: To assess the additive value of pleural fluid carcinoembryonic antigen (CEA.PF) level in the diagnosis of malignant pleural effusion. METHODS: Thoracentesis and closed pleural biopsy were performed in consecutive patients with pleural effusions. CEA.PF, cell analysis, and biochemical, cytopathologic and microbiologic studies were carried out. Further diagnostic interventions were undertaken if initial tests were inconclusive. RESULTS: A total of 176 patients were evaluated. The effusions proved malignant in 78 patients (44%). Benign etiologies were diagnosed in 89 cases, comprising 51 tuberculous pleurisies, 12 empyemas, 26 others. The cause was unknown in 9 patients. Median (range) in ng/ml of CEA.PF were 233 (1-12,500) in malignant vs. 2.5 (0.3-9) in tuberculosis, 1.4 (0.1-2) in transudates, 19.4 (0.6-312) in empyemas, p < 0.001. Receiver operating characteristic curve identified 10 ng/ml as the best cut-off for CEA.PF, yielding a sensitivity of 0.77, a specificity of 0.94, a positive and negative predictive value of 0.92 and 0.82, respectively. Among the 78 patients with malignant effusions, CEA.PF was elevated but initial cytopathologic study was nondiagnostic in 14 patients (18%). Prompted by the raised CEA.PF, further diagnostic interventions were undertaken and secured the diagnosis of malignancy in all of these 14 patients. CONCLUSIONS: CEA.PF level adds limited value on cytopathologic study in the diagnosis of malignant pleural effusions. It potentially identifies 18% of patients with malignant effusions who require further investigations despite negative initial cytopathologic study. Copyright 2000 S. Karger AG, Basel
OBJECTIVE: To assess the additive value of pleural fluid carcinoembryonic antigen (CEA.PF) level in the diagnosis of malignant pleural effusion. METHODS: Thoracentesis and closed pleural biopsy were performed in consecutive patients with pleural effusions. CEA.PF, cell analysis, and biochemical, cytopathologic and microbiologic studies were carried out. Further diagnostic interventions were undertaken if initial tests were inconclusive. RESULTS: A total of 176 patients were evaluated. The effusions proved malignant in 78 patients (44%). Benign etiologies were diagnosed in 89 cases, comprising 51 tuberculous pleurisies, 12 empyemas, 26 others. The cause was unknown in 9 patients. Median (range) in ng/ml of CEA.PF were 233 (1-12,500) in malignant vs. 2.5 (0.3-9) in tuberculosis, 1.4 (0.1-2) in transudates, 19.4 (0.6-312) in empyemas, p < 0.001. Receiver operating characteristic curve identified 10 ng/ml as the best cut-off for CEA.PF, yielding a sensitivity of 0.77, a specificity of 0.94, a positive and negative predictive value of 0.92 and 0.82, respectively. Among the 78 patients with malignant effusions, CEA.PF was elevated but initial cytopathologic study was nondiagnostic in 14 patients (18%). Prompted by the raised CEA.PF, further diagnostic interventions were undertaken and secured the diagnosis of malignancy in all of these 14 patients. CONCLUSIONS:CEA.PF level adds limited value on cytopathologic study in the diagnosis of malignant pleural effusions. It potentially identifies 18% of patients with malignant effusions who require further investigations despite negative initial cytopathologic study. Copyright 2000 S. Karger AG, Basel