[Contrast-enhanced study of solitary pulmonary nodules with thin-section computed tomography]

Since many benign and malignant pathologic conditions can appear as solitary pulmonary nodules, to establish nodule nature is always necessary for correct patient management. Recently, some authors have demonstrated the effectiveness of incremental dynamic CT in distinguishing cancerous from noncancerous lesions. The purpose of this work is to report our personal experience in this field.

We reviewed the incremental dynamic CT scans of 21 patients with a solitary pulmonary nodule < 3.5 cm phi without any calcifications, cavities and fat--namely, 15 carcinomas, 3 granulomas, 2 hamartomas, 1 abscess. Lesion density was evaluated before and 30 s, 1, 2, 3 and 5 min after contrast agent administration; we used a circular region of interest consisting of the central portion of the nodule in all cases and of 60-70% of its area in most cases. We subdivided the nodules into two groups, according to their enhancement: the nodules with > 20 HU and those with < 20 HU. All the lesions were submitted to surgery and histologic studies.

Fourteen of 15 carcinomas and one hamartoma had contrast enhancement > 20 HU; an abscess exhibited marked ring-shaped contrast enhancement (positive predictive value: 87%). One carcinoma, three granulomas and one hamartoma had no contrast enhancement, or else it was < 20 HU (negative predictive value: 80%).

Recently, some authors have demonstrated that malignant nodules, studied with incremental dynamic CT, have higher contrast enhancement than benign nodules. A value > 20 HU is a good predictor of malignancy (positive predictive value: 90%) and, conversely, a value < 20 HU is an unquestionable sign of benignity (negative predictive value: 100%). Our findings confirm the positive predictive value of enhancement > 20 HU, but not its negative predictive value because we found a malignant nodule without contrast enhancement.

Incremental dynamic CT is an effective indicator of solitary pulmonary nodule nature, but its predictive value is not absolute and therefore this technique should be integrated with biopsy in the cases which are clinically or radiologically suspicious.