Transthoracic fine-needle aspiration cytology. An analysis of 500 cases

Existing cytological and radiological services enabled the introduction of transthoracic fine-needle aspiration (FNA) in Johannesburg in 1983. The results achieved in the first 500 patients subjected to this procedure are presented, together with histological and clinical correlation. Of a total of 330 malignant tumours in this series, 306 were successfully sampled by FNA. Sensitivity of the technique in malignant disease was 92.7%, with 98.8% specificity. Positive and negative predictive values were 99.4% and 85% respectively. FNA was less accurate in the assessment of 24 benign thoracic tumours, only 75% of which were adequately sampled. Similarly, definitive cytological diagnosis of specific non-neoplastic pulmonary lesions, including granulomatous disease, was possible in only 78.6% of cases. The problem of aspirates in which no cellular abnormality (benign or malignant) was observed is discussed in relation to accuracy of needle placement. The special advantages of this technique included non-surgical verification of malignancy in radiologically or clinically inoperable cases; documentation of metastases; and detection of those tumour types (small-cell undifferentiated carcinoma and lymphoma) more appropriately treated by chemotherapy than by surgery. Pneumothorax requiring intercostal drain insertion occurred in 24 patients (4.8%). The advantages of rapid, accurate, inexpensive, non-operative diagnosis of thoracic masses was felt to outweight this single major complication.