OBJECTIVE: To evaluate the diagnostic accuracy of the percutaneous fine needle aspiration cytology (FNAC) for mediastinal lesions by using histology or follow-up clinical diagnosis as gold standard. PATIENTS AND METHODS: CT-guided percutaneous FNAC was performed on 131 patients with mediastinal lesions. Helical CT was used with 3-10 mm image thickness range and low radiation dose (40 mAs, 120 kV). Samples were immediately examined by a cytologist to determine if they were representative. Histological samples were obtained by means of biopsy or resection specimens in 73 patients and clinical follow-up in 50. RESULTS: The material was satisfactory for diagnosis in 126 patients (95.2 %), in whom 103 lesions (78.6%) were considered malignant (62 primary tumours and 41 metastases) and 23 (17.6%) benign. In the 123 patients with clinical monitoring or pathological diagnosis, using FNAC led to the identification of malignancy with a sensitivity of 95.2 % (95%CI: 89.2-97.9%), specificity 84.2% (95%CI: 62.4-94.5%), positive predictive value 97.1% (95%-CI: 91.7-99.0%), negative predictive value 76.2% (95%CI: 54.9-89.4%), likelihood-ratio positive 6.03 (95%CI: 2.13-17.05) and accuracy 93.5% (95%CI: 87.7-96.7%). Pneumothorax was the most frequent complication (3 cases). There was good agreement between the cytological findings and the histological findings, not only for malignant lesions (kappa coefficient: 0.641) but also for benign (kappa 0.607). CONCLUSIONS: CT-guided percutaneous FNAC is a safe and effective technique for the diagnosis of the mediastinal masses, with a high diagnostic yield for malignancy depicting. Copyright 2009 SEPAR. Published by Elsevier Espana. All rights reserved.
OBJECTIVE: To evaluate the diagnostic accuracy of the percutaneous fine needle aspiration cytology (FNAC) for mediastinal lesions by using histology or follow-up clinical diagnosis as gold standard. PATIENTS AND METHODS: CT-guided percutaneous FNAC was performed on 131 patients with mediastinal lesions. Helical CT was used with 3-10 mm image thickness range and low radiation dose (40 mAs, 120 kV). Samples were immediately examined by a cytologist to determine if they were representative. Histological samples were obtained by means of biopsy or resection specimens in 73 patients and clinical follow-up in 50. RESULTS: The material was satisfactory for diagnosis in 126 patients (95.2 %), in whom 103 lesions (78.6%) were considered malignant (62 primary tumours and 41 metastases) and 23 (17.6%) benign. In the 123 patients with clinical monitoring or pathological diagnosis, using FNAC led to the identification of malignancy with a sensitivity of 95.2 % (95%CI: 89.2-97.9%), specificity 84.2% (95%CI: 62.4-94.5%), positive predictive value 97.1% (95%-CI: 91.7-99.0%), negative predictive value 76.2% (95%CI: 54.9-89.4%), likelihood-ratio positive 6.03 (95%CI: 2.13-17.05) and accuracy 93.5% (95%CI: 87.7-96.7%). Pneumothorax was the most frequent complication (3 cases). There was good agreement between the cytological findings and the histological findings, not only for malignant lesions (kappa coefficient: 0.641) but also for benign (kappa 0.607). CONCLUSIONS: CT-guided percutaneous FNAC is a safe and effective technique for the diagnosis of the mediastinal masses, with a high diagnostic yield for malignancy depicting. Copyright 2009 SEPAR. Published by Elsevier Espana. All rights reserved.