Prognostic significance of flow cytometry in lung cancer. A 5‐year study

Flow cytometrically determined cellular DNA content has been measured on specimens from 101 patients affected by lung cancer (40 epidermoid cell carcinoma, 22 adenocarcinoma, 21 large cell carcinoma, 11 small cell carcinoma, and seven undifferentiated carcinoma), and one by mesothelioma. Ninty-eight of 102 (96%) patients with neoplastic disease evidenced the occurrence of at least one cytometrically aneuploid cell subpopulation. Fifty-five of 102 (54%) cases evidenced the occurrence of multiclonality, that is, the presence of more than one aneuploid stem cell line. However, the incidence of multiclonality in lung carcinoma was statistically different in surgical cases (where multiple site sampling from the primary and lymph nodes was possible) in comparison to the nonsurgical ones (e.g., bronchial washing): 48/77 (62%) and six of 24 (25%). respectively. Therefore, only the 77 surgical patients were used for further analysis. The cases were classified according to the DNA index (DI) in the following way: Group A (tumors with one or more stem lines with D1 ranging from 1 to 2) and Group B (tumors with at least one stem line with DI 2). A significant correlation has been found betyeen the cytometric ploidy condition so defined (Groups A and R) and the tumor mass doubling time (DT), Group B being associated with fast growing tumors (DT lower than 90 days). A qtatistically better 12-month survival rate (5-year maximum follow-up) was observed in Group A (88%) in respect to Group B (47%) and is evident in the patient survival time course. A better prognostic indication can be achieved by stratifying the patients according to both the cytometric ploidy condition and the tumor DT. Flow cytometric data can usefully contribute to the prognostic assessment of lung carcinoma provided that representative cellular material is collected by multiple site sampling. Cancer 60:844-851, 1987. HE STUDY OF the growth characteristics and cellular T heterogeneity of a highly malignant tumor, such as the bronchopulmonary carcinoma, could be useful to reach a better understanding of tumor natural history. Furthermore, this kind of information could be used to design of new therapeutic approaches, because no advances have been achieved with respect to long-term survival despite the recent introduction of combined therapeutic modalities.’** Until now, the definition of this type of cancer and its prognosis have been assessed on the basis of histologic and clinical parameters, that is, on essentially descriptive and qualitative parameters. The possibility that some quantitatively determined biologic feature^,^ such as growth kinetics and cellular DNA