RESUMO
INTRODUCTION: We investigated whether a group of pathologists could reproducibly apply the International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society (IASLC/ATS/ERS) classification for lung adenocarcinoma to a cohort of stage 1 tumors and whether this architectural classification and/or other parameters could demonstrate survival advantage. METHODS: A total of 145 cases of 7 edition of tumor, node, metastasis stage 1 adenocarcinoma were retrospectively reviewed for predominant architectural pattern, including cribriform pattern, nuclear grade, mitotic index, and necrosis. The parameters were assessed for reproducibility and survival and using multivariate analysis, compared with stage, age, and sex. RESULTS: The majority of tumors had a mixed architecture with the acinar pattern being the most common predominant architecture. Micropapillary and cribriform architecture were the least frequent patterns. This study demonstrated that a group of five pathologists could reproducibly apply the IASLC/ATS/ERS classification. Although there were insufficient cribriform-predominant adenocarcinomas for assessment, when the percentage of all cribriform was combined with other architectures, it was associated with a worse prognosis. The majority of the parameters assessed demonstrated significance with univariate analysis but only mitotic index, as assessed by the highest count/10 high-power fields remained significant with multivariate analysis. CONCLUSION: In this study of resected stage 1 primary lung adenocarcinoma, we found mitotic index to be the only independent prognostic marker. It was more closely associated with outcome than either pathologic T stage or IASLC/ATS/ERS architecture-based classification. Further validation of concordance and reproducibility in reporting mitotic index, as well as validation of prognostic significance, needs to be undertaken in independent data sets.