PAS-D: If mucin is diffusely present, than endocervical adenocarcinoma. Focal mucin positivity can be seen in endometrial adenocarcinoma.

Best is to do IHC.

IHC: ER, PR, CEA, p16, and vimentin

ER and PR strongly positive in endometrial adeno and negative/weak positive in endocervical.

CEA and p16 positive in endocervical and negative in endometrial adeno.

Vimentin negative in endocervical and positive in endometrial adeno.

Presence of HPV by in-situ hybridisation in endocervical and not in endometrial adeno

I agree with Mayank Gupta. In addition to immunohistochemistry, the presence of adenocarcinoma in situ would also favor an endocervical origin, while the presence of atypical endometrial hyperplasia would favor and endometrial origin.

Also, beware with p16 since it is often expressed in serous adenocarcinoma of the endometrium as well (due to other molecular mechanisms not linked to high-risk HPV infection).

Gael Phillips Pathology Queensland / University of Queensland

Vimentin is a very useful antibody I also use to distinguish endocervical from endometrial adenocarcinoma but this may not help in cases where an endocervical adenocarcinoma has endometrioid differentiation. It is most useful in cases where there is the more usual type of endocervical adenocarcinoma, in which case the epithelial cells will be negative. The clinical appearances also need to be taken into account , such as the apparent location, either the cervical canal or endometrial cavity. I agree that p16 can be helpful but is also expressed in some endometrial cancers. I also agree that an in situ component in the cervix is helpful in identifying a cervical primary site and mucin stains should al

ways be performed.

Sometimes in fractional curettage, we get endocervical dissociated glands along with separate fragments of adenocarcinoma with borderline feature. In these cases it is difficult to decide. I do agree with all the suggestions made above. I would like to emphasize when everything is 50-50, we can bet on endocervical adenocarcinoma because it is more common than endometrial adenocarcinoma.

A panel of immunohistochemical stains ER,PR,CEA is working well for us in making a differential diagnosis of endometrial and endocervical adenocarcinomas.A strong ER,PR staining with a weak or negative staining for CEA favour endometrial while strong CEA expression favour endocervical adenocarcinoma.