Primary non-Hodgkin's lymphoma and pseudolymphoma of lung: a study of 161 patients.

The authors studied 161 cases of primary non-Hodgkin's lymphomas and pseudolymphomas of lung. Small lymphocytic proliferations, which they believe to be lymphomas, constituted 31.6 per cent of cases; plasmacytoid lymphocytic and small cleaved follicular center cell lymphomas (Lukes-Collins system), 22.4 and 11.8 per cent of cases, respectively; and the remaining follicular center cell lymphomas and B-immunoblastic sarcomas, 5.6 per cent of cases. Pseudolymphomas constituted 14 per cent of cases. Most patients were elderly and asymptomatic; in most cases a solitary nodule or infiltrate was observed on a chest radiograph. Radiographic evidence of effusion was found in both lymphomas and pseudolymphomas, but hilar adenopathy was restricted to lymphomas. A few peribronchial reactive germinal centers and intralesional giant cells/granulomas were seen frequently in unequivocal lymphomas, so their presence cannot be used to exclude neoplasia. A generally monomorphic cell population and invasion of bronchial cartilage or visceral pleura are suggestive of malignancy, whereas primitive cytologic appearance and invasion of lymph nodes or parietal pleura are pathognomonic of malignancy. Diffusely admixed mature lymphocytes and plasma cells with numerous reactive follicles suggest pseudolymphomas. Immunologic determination of clonality may be diagnostically definitive. Most localized lesions in lung were treated by surgical resection, whereas in cases of extensive pulmonary disease, biopsies were performed and patients were treated by chemotherapy or irradiation. Both lymphomas and pseudolymphomas recurred, most often within three years. Pseudolymphoma recurred only in lung. When distant spread of lymphoma occurred, it commonly involved extranodal sites. Only 18 of 101 patients with lymphoma died with or of tumor, and no patient with pseudolymphoma died of disease. Neither histologic subtype among the "small cell" lymphoid lymphomas nor the presence of regional node involvement was prognostically significant, but pleural effusion on the initial chest radiograph was a significant predictor of both recurrence and mortality.