Positron emission tomography (PET) is a new imaging technique which, by measuring focal metabolic activities, can make a qualitative statement (benign or malignant) about a tumour. PET has been described in many studies to provide a high diagnostic accuracy for the evaluation of pulmonary coin lesions. However, these studies were not always supported by histological confirmation of the results. In a controlled prospective study, it was investigated whether the diagnostic accuracy of PET is sufficiently high to allow omission of diagnostic thoracotomy or thoracoscopy in the case of a negative finding.
A PET scan was carried out before operation using [18F]fluorodeoxyglucose (FDG) in 50 patients with pulmonary coin lesions (diameter 30 mm or less). All of these lesions were completely removed thoracoscopically or by a formal thoracotomy and were examined histologically. Using the histology results, the diagnostic accuracy of the PET procedure with regard to a benign or malignant diagnosis was evaluated and compared with that of computed tomography (CT).
From a total of 54 coin lesions (four of the 50 patients had two lesions) there were 31 malignant (19 primary bronchial carcinomas, 12 metastases) and 23 benign diagnoses. With the PET procedure 28 of 31 malignant and 19 of 23 benign lesions were classified correctly (sensitivity 90 per cent, specificity 83 per cent). False negatives included two bronchial carcinomas and one metastasis. CT had a sensitivity of 100 per cent and specificity of 52 per cent.
FDG PET cannot generally be considered as a replacement for diagnostic thoracoscopy or thoracotomy at the present time. However, by combining FDG PET with radiological follow-up, clinical applications may evolve in patients at low risk for a malignant tumour or at high risk for surgical complications. © 1998 British Journal of Surgery Society Ltd