A 2-tier subdivision of papillary proliferations of the endometrium (PPE) only emphasizing the complexity of papillae precisely predicts the neoplastic risk and reflects the neoplasia-related molecular characteristics-a single-centered analysis of 207 cases.

Papillary proliferation of the endometrium (PPE) is subdivided based on the complexity of the papillae and the proliferation of lesions, and the complex group is considered to have an increased risk of concurrent/subsequent endometrial neoplasia. However, the current subdivision criteria fail to prove the equivalence of the quantity of simple papillae and structural complexity. In this study, we divided PPE of 207 cases from 2014 to 2022 into 3 groups according to structural complexity and proliferation degrees: Group 1 equaled to the simple PPE with a simple papillary structure and typical localized proliferation; group 2 had the simple structure similar to group 1 but occupy over 50% of the endometrial polyp or > 2 lesions in the surface of nonpolypoid endometrium; group 3 had the truly complex branching papillae despite of its proportion. Group 3 was implicated with significantly more concurrent endometrial neoplasia (EAH and carcinoma) compared with groups 1 and 2 (P < 0.01), while no difference was found between groups 1 and 2. In 128 cases with no concurrent endometrial abnormalities in the initial biopsy or curettage specimens, 4 cases presented endometrial neoplasia (3 carcinoma and 1 atypical hyperplasia) in the subsequent specimens, all of which presented PPE of group 3 but not group 1 or 2 in the prior tissues (P < 0.01). The immunochemistry of 83 cases showed similar expressions of ER, PTEN, ARID1A, PTEN, p16, β-catenin, and p53 between PPE and the surrounding normal endometrium. Nearly 100% of PPE cases lost expressions of PR. A total of 2/83 cases showing PAX2 expression were all in the group 3 and correlated with endometrial neoplasia (2/17, 11.76%, P < 0.05). 76/83 (91.57%) of PPE lesions had KRAS mutations, and the distributions of which were similar among 3 groups. The frequency of mucinous metaplasia was significantly higher in the PPE lesions with KRAS mutations (72/74, 97.30%, P < 0.01). Group 3 showed higher frequency of single KRAS mutations compared with the combination of groups 1 and 2 (P < 0.01). Finally, the concordance of KRAS mutation profiles between PPE and endometrial neoplasia was significantly higher in group 3 than either group 1 or 2 (P < 0.01), while no difference was found between group 1 and 2. Thus, a new 2-tier subdivision system only emphasizing the complexity of papillae is recommended, which might precisely predict the risk of endometrial neoplasia and neoplasia-related molecular characteristics.