[The solitary pulmonary nodule: the preliminary results in differential diagnosis by high-resolution computed tomography with a contrast medium].

High-resolution computed tomography (HRCT) with iodinated contrast material has been used by many authors to study solitary pulmonary nodules (SPNs). The degree of enhancement was correlated with the nodule malignancy.Forty adult patients were examined, before and after contrast agent administration, with incremental dynamic CT. We selected 22 patients with SPNs (3-30 mm phi, except one with 40 mm phi). The CT numbers of the inner nodule were calculated before and 1, 2 and 3 minutes after the i.v. administration of a weight-related dose (1.5 mL/kg/min) of nonionic iodinated contrast agent. A dose of 100 mL contrast agent was used in the first 6 patients. The difference in CT numbers between unenhanced images and the images with maximum enhancement (max. attenuation) was also calculated.Histologic diagnoses included 4 tuberculomas, 3 hamartomas and 15 malignant tumors (9 adenocarcinomas, 5 squamous cell carcinomas and 1 non-Hodgkin lymphoma). The CT numbers (in Hounsfield units, HU) of malignant nodules ranged 12-31 HU (mean: 21.5 HU) before contrast agent administration; the "long-standing" tuberculomas ranged 11-22 HU (mean: 16.5 HU) and the hamartomas had a mean density of 10.5 HU. We excluded for the study two "fresh" tuberculomas, one of which was surrounded by a low-attenuation infiltrate (the halo sign). We selected a threshold value of 20 HU on enhanced CT images to distinguish malignant (> or = 20 HU) from benign (< or = 20 HU) nodules. All lung cancers had complete enhancement (mean density: 35.5 HU). With 20 HU as the threshold value for a positive test, sensitivity was 100\%, specificity 85.7\%, positive predictive value 93.8\% and negative predictive value 100\%; test bias was 1.067.Positron emission tomography (PET) with 2-[fluorine-18] fluoro-2-deoxy-D glucose is reported to be as accurate as enhanced HRCT, but it does not provide accurate morphological information, is not widely available and it is quite expensive: therefore, in our opinion, CT should be preferred. After examining over 100 patients, we may use our results in the decision analysis comparing surgical risk with cancer risk.