IMMUNOS-p53-Learn from history, avoid mistakes

Immunohistochemistry for diagnosis in pathology started as routine techniques in the 80's. Immediately was adopted in numerous departments and classical surgical pathology changed forever. I think nobody realized at that time that 30 years later, when reviewing difficult cases, residents will be more interested in learning what immunos to request than learning the basic features of H&E. I remember that a malignant tumor positive for keratin was a carcinoma, and cases positive for vimentin were sarcomas. Today we know that some lymphomas and sarcomas can be positive for keratin, and some carcinomas can be negative for keratin. We need to learn from this to avoid committing the same mistake.

Papillary serous carcinoma, of the ovary or uterus, was described based on H&E, and it is a high grade carcinoma, different from endometrioid carcinoma. In most cases it is not difficult to recognize a papillary serous carcinoma and an endometrioid carcinoma. Then, p53 came to help us. Papillary serous carcinoma is positive for p53, and endometrioid carcinoma is negative. But now we have some problems with the interpretation of p53:

1- What percentage of cells should be positive to designate a tumor as positive?

According to IJGynPath 23: 129, 2004= 50%

According to Cancer Medicine 4:75,2015= 10%

According to Pathology Outline = 5%

2- A case with 1%, 2%, or 3% of positive cells is wild pattern or null pattern?

3- 10% to 15% of high grade serous carcinomas do not have a p53 mutation. Abstracts 1254, and 1104. USCAP 2016

4-Cases with TP53 mutations have better survival. Abstract 1104, USCAP 2016

5- 3% of grade 1, and 11% of grade 2 endometrioid adenocarcinomas have p53 mutations,not associated with outcome, or are + for p53

IJGynPath 35:289,2016, and abstract 1179, USCAP 2016

6- According to abstract 1132, USCAP 2016, p53 can be used as a surrogate for copy number status. This new genomic classification of endometrial carcinoma is trying to replace a morphologic "irreproducible" classification.

7- According to IJGyPath 35:289,2016, p53 cannot be used to classify cases in the new genomic classification because 10% of the copy # high cases are wild type, and 30% of the POLE cases, and 8% of the MSI cases have Tp53 mutations.

Suggestion to young pathologists.

New techniques are great and welcome, but they should be used to confirm the diagnosis. An endometrioid carcinoma, positive for p53 is still an endometrioid carcinoma. Diagnosis should be based on what the cells are doing ( morphology) , not what they are thinking ( immuno). In cases where the H&E does not agree with the immunos, think about the patient. You might feel great rendering a brave diagnosis which will show that you are very well informed, but do you have prove that the patient needs an aggressive treatment?