One hundred forty-three patients with bronchogenic carcinoma were studied prospectively with computed tomography (CT) to determine the accuracy of CT in the evaluation of mediastinal nodal metastases. Mediastinal lymph nodes were localized according to the lymph node mapping scheme of the American Thoracic Society and were considered abnormal if they exceeded 1 cm in short-axis diameter. All patients underwent surgical staging, which consisted of either mediastinoscopy alone or mediastinoscopy and thoracotomy. At the time of surgical staging, all accessible nodes were either removed or sampled. The sensitivity of CT for mediastinal nodes on a per-patient basis was 64%, with a specificity of 62%. The sensitivity of CT for individual nodal stations involved with tumor was only 44%. The presence of obstructive pneumonitis did not appreciably alter the sensitivity of CT, but the specificity was lower (43%). The likelihood of metastases increased with lymph node size; however, seven of 19 (37%) lymph nodes that measured 2-4 cm in short-axis diameter were hyperplastic and did not contain metastases. The relative insensitivity of CT makes formal nodal sampling at the time of mediastinoscopy or thoracotomy essential to detect lymph node metastases.

Localized Castleman disease manifests as either a solitary, well-circumscribed mediastinal mass or an infiltrative mass with associated lymphadenopathy on CT or MR images. Disseminated Castleman disease manifests with diffuse mediastinal lymphadenopathy.

No abstract

No abstract

No abstract

No abstract

No abstract

No abstract