David S. Feigin

Sensitivity of chest radiography, computed tomography, and gallium scanning to metastasis of lung carcinoma

To determine the efficacy of radiologic techniques in preoperative staging of the mediastinum for lung carcinoma, the authors studied 45 patients with chest films supplemented with oblique views, esophagrams, gallium scans, and computed tomograms (CT). They interpreted the studies and correlated surgical findings using a modified classification of lymph node regions. The mediastinum was positive on chest films in 14 of the 21 cases with pathologically proved mediastinal metastases (33% false‐negative). Gallium scans in cases with a positive primary were positive in 12 of 15 cases with mediastinal or hilar metastases (20% false‐negative). Computed tomography showed nodes over 1 or 1.5 cm in size in or adjacent to the biopsy‐positive node region in 18 of 19 patients (5% false‐negative), extranodal mediastinal involvement, and 9 of 10 proven hilar metastases. Computed tomography is a sensitive screening technique in patients who would otherwise require an invasive staging procedure, but is not highly specific (false‐positive rate 38%).

Postpartum Pleural Effusion: A Common Radiologic Finding

We did a retrospective study of 112 patients and prospective study of 30 patients and 30 controls to ascertain the incidence of postpartum pleural effusion. An effusion was shown on chest radiographs in 51 patients (46%) in the retrospective study, and 20 patients (67%) in the prospective study. The incidence of pleural effusion in obstetric patients in the prospective study was significantly greater than that of controls (chi-squared test, 20.4, p less than 0.001). There was no difference between patients with effusion and those without in age, race, parity, antepartum complications, method of delivery, fetal outcome, or postpartum complications. We conclude that pleural effusion occurs frequently in the first 24 hours after delivery and that, in the absence of symptoms or signs of cardiopulmonary disorder, no intervention is necessary.

Misconceptions regarding the pathogenicity of silicas and silicates

Several inhaled substances, from occupational or other environmental exposure, produce significant pulmonary disease and abnormalities demonstrated by pulmonary imaging. Areas of controversy and misconception relate principally to the extent and nature of both the clinical disease and the imaging abnormalities specific to each substance. The size and shape of the inhaled particles is an important determinant of the nature and severity of the disease produced, with fibrous shapes usually being the most pathogenetic. Fibrogenicity is another important pathogenetic characteristic of talc and kaolin, as well as asbestos. Talc produces four distinct forms of pulmonary disease, depending not only on the other substances with which it is inhaled, but also whether it is inhaled or injected intravenously. When inhaled alone, talc does not appear to produce significant pulmonary fibrosis or malignancy. Kaolin, mica, fullers earth, zeolite, and fiberglass all vary in disease production according to their shape and fibrogenicity. Silica, diatomaceous earth, and other forms of silica are all highly fibrogenic and thus produce clinically obvious disease with sufficient inhalation. The largest particles usually produce nodular patterns in the upper pulmonary fields, as is typical of silicosis. The fibrous particles are more likely to manifest themselves as interstitial patterns in the lower pulmonary fields.

Vasculitis in the lung

Pulmonary vasculitis occurs most commonly in the collagen vascular diseases and in granulomatous pulmonary disease. In the collagen vascular group, vasculitis causes diffuse interstitial inflammation and subsequent fibrosis, resulting in interstitial radiographic patterns, especially in the lower lung fields. Vasculitis accompanied by granulomatosis typically produces focal inflammation and is, therefore, manifested as nodules and masses. The more typical collagen vascular diseases include rheumatoid arthritis, systemic lupus erythematosus, systemic sclerosis, and dermatomyositis. The most common vasculitis granulomatosis is Wegeners, with similar radiographic abnormalities occurring in lymphomatoid granulomatosis. Atypical examples of vasculitis pulmonary disease include ankylosing spondylitis, in which upper-lung field fibrobullous changes are seen. Periarteritis nodosa and Behcets syndrome include abnormalities of large vessels and thromboembolic phenomena. Bronchocentric granulomatosis and allergic granulomatosis involve airway abnormalities as well as vasculitis and granuloma formation.

Improving specificity of computed tomography in diagnosis of malignant mediastinal lymph nodes

Four parameters of lymph node appearance on computed tomography were studied to improve specificity for malignancy in bronchogenic carcinoma: 1) node location, 2) homogeneity, 3) border definition, and 4) delineation by fat. Of 54 carcinoma patients, nodes were pathologically malignant in 21. Computed tomography showed enlarged nodes (over 1 cm) in 20 of these (true-positive rate, 96%), but also in 13 of the 33 patients with pathologically benign lymph nodes (false-positive rate, 39%). A combination of all four computed tomography parameters reduced the false-positive rate from 39 to 21% and decreased the true-positive rate from 96 to 86%. The criterion of border definition alone reduced the false-positive rate from 39 to 6%, but decreased the true-positive rate from 96 to 61%. The most useful computed tomography parameter was delineation, which reduced the false-positive rate to 15% and decreased the true-positive rate only to 86%. The computed tomography appearance of enlarged lymph nodes may be used to improve selection of patients with enlarged nodes for preoperative biopsy.

The effect of vasospasm on perfusion radionuclide imaging in experimental pulmonary embolism

Experiments were conducted to further evaluate previously observed discrepancies between postembolic perfusion scans and angiograms, and the relationship of these discrepancies to pulmonary vasospasm. Nonresorbable emboli were formed from Ivalon, muscle, or heated autologous clot. Scans and angiograms were compared at 24-hour intervals for two to eight days. During the first two days after embolism, angiographic-scan discrepancies occurred in most dogs, characterized by decreased size of scintigraphically demonstrated perfusion defects at the same times that the angiogram showed stable or even progressive obstruction. This discordant behavior of scan and angiogram is thought to reflect subsidence of postembolic vasospasm. In the 17 experimental dogs, 13 showed evidence of vasospasm, and four did not. In those that developed spasm, direct angiographic evidence of such spasm was present in all but one, manifest as diffuse constriction of small arteries and diminished parenchymal stain in vascular beds that were not directly obstructed mechanically. Although spasm diminished rapidly during the first 6 hours, there was continued and appreciable further resolution for over 48 hours in some dogs. This study reconfirms the frequent existence of vasospasm in three different embolic models, and indicates subsidence of spasm over a much longer time frame than was formerly supposed. Regression of spasm over an interval of several days suggests that perfusion scans should be performed clinically as soon as possible after a suspected embolic episode, prior to subsidence of spasm, in order to exploit the added sensitivity provided by spasm.