Lung adenocarcinoma with mixed bronchioloalveolar and invasive components: clinicopathological features, subclassification by extent of invasive foci, and immunohistochemical characterization.

A significant proportion of small lung adenocarcinomas consists of two components: bronchioloalveolar carcinoma (BAC) and invasive carcinoma. The purpose of this study was to compare their clinicopathologic features with those of BAC and those of invasive cancer without BAC, and to define "early invasive" lesions based on the extent of invasive foci. We reviewed 484 lesions of resected lung adenocarcinoma and classified them into three groups according to tumor growth pattern: group 1 (n = 102, BAC), group 2 (n = 216, adenocarcinoma consisting of BAC and invasive carcinoma), and group 3 (n = 166, invasive adenocarcinoma without BAC component). Group 2 was further subdivided according to the extent of the invasive area: group 2a (n = 54), BAC with invasive foci <or=5 mm; group 2b (n = 162), BAC with invasive foci >5 mm. These groups were compared with regard to their clinicopathologic features, expression of Ki-67 and p53, and expression of laminin-5, a putative marker for tumor invasion. The positivity rates of vascular, lymphatic, and pleural invasion in each group were as follows: 0%, 0%, and 0% in group 1; 5.5%, 14.8%, and 1.9% in group 2a; 45.7%, 41.4%, and 25.9% in group 2b; and 84.9%, 61.4%, and 60.8% in group 3. Notably, no lymph node metastasis occurred in either group 2a or group 1, but it was observed in 24.1% of group 2b and 47.0% of group 3. The mean Ki-67 labeling index, the frequency of p53 overexpression, and the frequency of laminin-5 overexpression increased from group 1 (11%, 4%, and 0%) to group 2a (16%, 20%, and 7%) to group 2b (24%, 41%, and 23%) to group 3 (35%, 38%, and 38%). In contrast, no clear differences were observed when lesions were subdivided according to size. Based on the distribution pattern of Ki-67-positive tumor cells, lesions were classified into two groups: marginal type (63%) and nonmarginal type (37%). The latter showed a significantly higher labeling index than the former. Moreover, the proportion of the marginal type clearly decreased from group 1 (85%) and group 2a (87%) to group 2b (55%) to group 3 (19%). Group 2 lesions showed characteristics intermediate between the BAC and invasive adenocarcinoma. According to the extent of the invasive area, we were able to define a subgroup of mixed-type adenocarcinomas (group 2a) that could be regarded as early invasive cancer because they showed low rates of vascular, lymphatic, and pleural invasion, and no nodal involvement.