Prognostic significance of grading in lung adenocarcinoma

Abstract

BACKGROUND: Although grading has prognostic significance for many tumor types, a prognostically significant grading system for lung adenocarcinoma has not yet been established. The aim of this study was to evaluate histologic characteristics included in tumor grading systems, establish optimal cutoff values that have the strongest association with overall survival, and develop a grading system incorporating the histopathologic characteristics that the authors found to have prognostic significance in patients with lung adenocarcinoma. METHODS: The authors studied lung adenocarcinomas from 85 consecutive patients, and evaluated the percentage of solid pattern (as a reflection of tumor architecture), the degree of cytologic atypia, and the mitotic count. RESULTS: In univariate analysis, overall survival was associated significantly with sex (P = .045), age (P = .0008), tumor status (P < .0001), lymph node status (P = .02), solid pattern (P = .046), and cytologic atypia (P = .01), but not with mitotic count (P = .26). On the basis of optimal cutoff values, the authors found that a solid pattern ≥90% and severe cytologic atypia were the best discriminators of worse outcome. A grading score, computed as the sum of the architecture score and cytologic atypia score (2 = well differentiated, 3 = moderately differentiated, 4 = poorly differentiated), was a significant predictor of overall survival in univariate analysis (median overall survival times, 72.4, 39.5, and 8.7 months for well, moderately, and poorly differentiated adenocarcinoma, respectively; P = .0001). Moreover, grading was an independent predictor of survival in multivariate analysis (P = .002). CONCLUSIONS: The authors describe a grading system that incorporates the percentage of solid pattern and degree of the cytologic atypia that is an independent predictor of survival in patients with lung adenocarcinoma. Cancer 2010. © 2009 American Cancer Society.