The respiratory epithelium. IX. Validity and reproducibility of revised cytologic criteria for human and hamster respiratory tract tumors.

Cytologic criteria, based upon features of differentiation at the cellular level, have been established previously for human and hamster respiratory tract tumors. In order to validate these histogenetic revised criteria and to study their reproducibility, imprint smears of human and hamster tumors, hamster tracheal washing specimens and preoperative human specimens of sputa, bronchial washings and bronchial brushings were coded randomly, followed by blind cytologic study of the tumor cells. The hamster tumors were induced by intratracheal instillation of benzo(a)pyrene-ferric oxide or sequential administration of benzo(a)pyrene-ferric oxide and N-methyl-N-nitrosourea. Greater accuracy and reliability were obtained for the histogenetic cytodiagnosis versus the histogenetic histologic evaluation than were achieved for the conventional cytodiagnosis versus the World Health Organization histologic evaluation of 1967. The histogenetic classification utilizes diagnostic criteria that may be universally applicable by pathologists and cytopathologists. Many poorly to well-differentiated epidermoid carcinomas, poorly to well-differentiated adenocarcinomas, large-cell carcinomas and giant-cell carcinomas, diagnosed as such by conventional cytologic criteria, possessed histogenetic criteria of combined epidermoid and adenocarcinomas. The new data clarify conventional cytodiagnoses of poorly differentiated epidermoid carcinomas and poorly differentiated adenocarcinomas and eliminate vague categories, such as large-cell undifferentiated, giant-cell and small-cell anaplastic carcinomas.