Analysis of cell type and radiographic presentation as predictors of the clinical course of patients with bronchioloalveolar cell carcinoma

Study objectives: Bronchioloalveolar carcinoma is a primary lung neoplasm of variable histopathologic, radiologic, and clinical expression. There are three cell types described in bronchioloalveolar carcinoma: Clara cells, mucin-producing cells, and alveolar type II epithelial cells. It is unclear whether these three tumor cell types are associated with a specific radiologic presentation and clinical course. In this study, we investigated whether tumor cell type, identified by transmission electron microscopy, correlated with a specific radiologic pattern, and whether tumor cell type or radiologic presentation correlated with the patient's clinical course and outcome. Design: Transmission electron microscopy was used to restudy tissue blocks from the original surgical histopathologic specimens in 54 patients with primary bronchioloalveolar carcinoma diagnosed over a 10-year period (1980 to 1990). The pretreatment radiographs were reviewed in each case, and the first chest radiograph obtained at the time of the discovery of the tumor in each patient was compared with the results of the ultrastructural study. The medical records of each patient were examined to obtain pertinent radiologic, clinical, and patient outcome information. Measurement and results: There were 32 Clara cell tumors, 10 mucin-producing cell tumors, and 1 alveolar type II epithelial cell tumor in this series. Eleven additional tumors had mixtures of two or more cell types. No statistically significant relationship was detected between tumor cell type and radiologic presentation or patient mortality pattern. There was increased mortality among patients who presented radiologically with segmental, lobar, multifocal, or diffuse disease compared with those patients exhibiting a solitary pulmonary nodule at presentation. Conclusion: Radiologic presentation, rather than tumor cell type, provides prognostic information that aids in predicting patient outcome.