Histologic and molecular staging of endometrial cancer

Jaime Prat (ES)

[Prat] Autonomous University of Barcelona

Context: The most important advance in the knowledge of endometrial carcinoma was made in 1983; i.e., separation of endometrioid (type I) from nonendometrioid (type II) which are two different diseases clinically, pathologically and regarding molecular genetics. Objective: The current FIGO staging system is almost 40 years behind. Indeed, the 2009 FIGO staging considers only one disease instead of “at least” two (endometrioid/low-grade and nonendometrioid/high-grade) which merit separate staging classification, particularly at stage I. Methods: The validity of this classification has been confirmed and expanded by 2013 The Cancer Genome Atlas (TCGA) (POLE/ultramutated, MI/hypermutated, low copy number and high-copy number; the latter group including serous carcinomas and serous-like G3 endometrioid carcinomas which behave like serous tumors; and POLE-mutated G3 endometrioid which behave very well even without chemotherapy). Main Outcome: Histotype should be reported; i.e., endometrioid or nonendometrioid (serous, clear cell, carcinosarcoma, mixed, undifferentiated, and dedifferentiated carcinomas); however, all nonendometrioid carcinomas are G3 and grading nonendometrioid tumors is useless. On the other hand, all G1-2 tumors are endometrioid carcinomas. Obviously, designing a specific staging system for each histotype would not be practical, but separating low-grade (G1 and G2) from high-grade (G3) for the purpose of staging -not for treatment but prognosis- is quite reasonable. Results: Distinguishing G1 from G2 endometrioid carcinoma may be difficult. There is not a clear cut division between G1 and G2 and overlapping is common. Regarding fertility sparing surgery in young patients, it wouldn’t matter whether the carcinoma is G1 or G2; it is G3 that really matters. G3 includes a gray zone pointed out by TCGA (serous and serous-like endometrioid) which can be reduced by immunohistochemistry (p53, MMR); also POLE mutation is worth investigating as positive cases may not need chemotherapy and have very good prognosis. Thus, p53, MMR (IHC) and POLE mutation analysis should be done in all G3 carcinomas. Noteworthy, G3 endometrioid carcinomas represent a small fraction of endometrial carcinomas compared with low-grade carcinomas, which are far more common. Conclusions: Low-grade (G1 and G2) and high-grade (G3) endometrial carcinomas represent two different diseases and a separate stage I classification is recommended.