[Interventional computed tomography in thoracic oncology. Apropos of 170 cases]

The authors report their experience with follow-up cytohistological examination of subclinical thoracic lesions in oncology by transparietal puncture with CAT-scanning localization in 170 patients. They detail a few special points concerning the material used and their techniques, in view of reducing the most common incidents and of managing the most severe complications. They analyze their results in two groups of patients, one investigated in a context of known malignancy and the other, with no history of malignancy in whom bronchoscopy, bronchial lavage and brushing were negative. They found an overall sensitivity and specificity of 86% and 100%, respectively, in the diagnosis of malignant disease. The method showed an efficacy of 83.5%. For the two populations, sensitivity was 84% and 87% respectively, and efficacy was 82 and 85% respectively. Thus, the results were better in the patients who had never received anticancer therapy. They also observed significant differences depending on whether the targets were mediastinal, pleuropulmonary or parietal. While the rapid diagnostic information appears invaluable in both groups of patients, in contrast, the small volume of the specimens, which may be hemorrhagic, necrotic, with more or less extensive fibrosis of the stroma, and the immaturity of certain tumors, sometimes make it impossible for the pathologist to give a precise diagnosis of the tumor other than the benign or malignant structural character of the lesion. This represents a significant obstacle in hematology for affections which, although non surgical, nevertheless require precise identification of the lesion for effective therapy. The elaboration of more effective sampling material, repeated punctures and immunohistoenzymatic analysis techniques would probably further improve the results of the method.