Molecular profile of grade 3 endometrioid endometrial carcinoma: is it a type I or type II endometrial carcinoma?

Two types of endometrial carcinoma (EC) have been delineated on the basis of clinicopathologic studies. Low-grade endometrioid carcinoma (EEC) is the prototype of type I EC and is characterized by microsatellite instability and PTEN, K-ras, and/or β-catenin gene mutations, whereas type II EC is typically represented by serous and clear cell carcinomas (SCs/CCCs), the former frequently showing p53 mutations and c-erb-2 overexpression; however, the molecular profile of grade 3 EEC has not yet been well characterized. The goal of this study was to define the immunohistochemical and molecular profile of grade 3 EEC. We studied 25 patients with grade 3 EEC ranging in age from 35 to 87 (mean 61) years. At the time of initial diagnosis, 16 patients had stage I tumors, whereas 3, 5, and 1 had stages II, III, and IV tumors, respectively. Only 1 patient with stage IV tumor had disease in the peritoneum because of direct extend of tumor through the uterine wall. Two tissue microarrays were constructed from paraffin-embedded blocks and stained for MLH-1, MSH-2, p16, cyclin D1, C-erb-B2, WT-1, and p53. Loss of MLH-1 and MSH-2 was seen in 3 of 25 and 1 of 24 tumors, respectively; none showed loss of both. Diffuse p16 nuclear expression was found in 7 of 23 cases; diffuse and strong nuclear immunostaining for p53, cyclin D1, and Her-2 was seen in 9 of 24 neoplasms, 9 of 25, and 3 of 25 carcinomas, respectively. WT-1 was negative in all 25 tumors. One of the 3 grade 3 EECs with Her-2 overexpression showed gene amplification by fluorescence in situ hybridization analysis. No gene amplification for cyclin D1 was found. Follow-up information was available for all patients. Sixteen had stage I tumors. Of these patients, 11 were alive and well (AW), 3 died of disease (DOD), and 2 died of unrelated causes (DUC), with a mean follow-up time of 56 months (range, 24 to 96 mo); 2 of 3 patients with stage II tumors DOD, and 1 was AW with a mean follow-up time of 81 months (range, 6 to 66 mo); of the 5 patients with stage III tumors, 2 DOD, 1 was AW, 1 was alive with lung metastases, and 1 DUC [mean follow-up of 29 months (range, 12 to 74 mo)]; the only patient who had a stage IV tumor DOD 12 months later. Interestingly, patients with grade 3 EECs showing loss of MLH-1/MSH-2 had stage I tumors, and all were AW (60 to 84 mo). Seventy-seven percent (7 of 9) of patients with tumors showing cyclin D1 overexpression were stage I, and none died of disease, whereas 85% (6 of 7) of patients with p16-positive tumors were high stage (2 stage II, 3 stage III, and 1 stage IV), and 5 DOD. All but one of these patients had tumors that also had p53 overexpression. All 3 patients with Her-2 overexpression DOD (stages I, III, and IV). In conclusion, this study shows that grade 3 EEC shares with low-grade EEC the overexpression but not amplification of cyclin D1 and low frequency of Her-2 overexpression and amplification. Grade 3 EEC shares with SC the relatively common p53 and p16 overexpression and low frequency of loss of mismatch repair genes. However, in contrast to SC ECs, which often show WT-1, cyclin D1 amplification, and Her-2 overexpression and/or amplification, grade 3 EECs rarely overexpressed any of these markers. Moreover, in this study, patients with tumors showing loss of MLH-1/MSH-2 or cyclin D1 overexpression were more likely to have low-stage tumors (stage I), whereas patients with tumors that overexpressed p53, p16, or Her-2 were frequently associated with high-stage tumors.