Symptomatic Patterns, Biologic Behavior, and Prognosis in Cancer of the Lung

Abstract

The present work demonstrates the necessity of conventional morphologic analyses with categorizations based on symptomatology. A previously-constructed system of classifying clinical symptoms, using principles of Boolean algebra and symbolic logic, has been applied to carcinoma of the lung. On the basis of the presenting symptoms, 442 patients with histologic evidence of lung cancer, treated during 1954-58, were divided into six clinical subgroups. The results demonstrate a method of quantifying the expectations of clinical judgment, and help confirm, in lung cancer, the biologic predeterminism hypothesis for solid neoplasms. According to this hypothesis, neoplasms are most likely to be curable if both anatomically excisable and slow-growing. As shown in the data, the best survival results occurred in patients with slow development of only primary symptoms, and in those discovered accidentally while asymptomatic. Since these distinctions could not have been detected from standard morphologic assessments, the data indicate that symptoms are important indices, particularly for rates of growth, of biologic behavior in human neoplasia.