ABSTRACT
BACKGROUND: We reviewed a series of 188 resected pulmonary mucinous adenocarcinomas (MAs) to clarify the prognostic significance of lepidic and non-lepidic patterns. METHODS: Non-lepidic patterns were divided into bland, non-distorted acini with uncertain invasiveness (pattern 1), unequivocal invasion into stroma (pattern 2), or invasion into alveolar spaces (pattern 3). RESULTS: The mean proportion of invasive patterns (patterns 2 and 3) was lowest in small (≤ 3 cm) tumors, and gradually increased in intermediate (> 3 cm and ≤ 7 cm) and large (> 7 cm) tumors (8.4%, 34.3%, and 50.1%, respectively). Adjusted T (aT) stage, as determined by the size of invasive patterns, was positively correlated with adverse histologic and clinical features including older age, male sex, and ever smokers. aTis tumors, which were exclusively composed of lepidic pattern (n = 9), or a mixture of lepidic and pattern 1 (n = 40) without any invasive patterns, showed 100% disease- free survival (DFS). The aT1mi tumors, with minimal (≤ 5 mm) invasive patterns (n = 63), showed a 95.2% 5-year DFS, with recurrences (n = 2) limited to tumors greater than 3 cm in total size (n = 23). Both T and aT stage were significantly associated with DFS; however, survival within the separate T-stage subgroups was stratified according to the aT stage, most notably in the intermediatestage subgroups. In multivariate analysis, the size of invasive patterns (p = .020), pleural invasion (p < .001), and vascular invasion (p = .048) were independent predictors of recurrence, whereas total size failed to achieve statistical significance (p = .121). CONCLUSIONS: This study provides a rationale for histologic risk stratification in pulmonary MA based on the extent of invasive growth patterns with refined criteria for invasion.