RESUMO
AIMS: POLE exonuclease domain mutations identify a subset of endometrial cancer (EC) patients with an excellent prognosis. The use of this biomarker has been suggested to refine adjuvant treatment decisions, but the necessary sequencing is not widely performed and is relatively expensive. Therefore, we aimed to identify histopathological and immunohistochemical characteristics to aid in the detection of POLE-mutant ECs. METHODS AND RESULTS: Fifty-one POLE-mutant endometrioid, 67 POLE-wild-type endometrioid and 15 POLE-wild-type serous ECs were included (total N = 133). An expert gynaecopathologist, blinded to molecular features, evaluated each case (two or more slides) for 16 morphological characteristics. Immunohistochemistry was performed for p53, p16, MLH1, MSH2, MSH6, and PMS2. POLE-mutant ECs were characterised by a prominent immune infiltrate: 80% showed peritumoral lymphocytes and 59% showed tumour-infiltrating lymphocytes, as compared with 43% and 28% of POLE-wild-type endometrioid ECs, and 27% and 13% of their serous counterparts (P < 0.01, all comparisons). Of POLE-mutant ECs, 33% contained tumour giant cells; this proportion was significantly higher than that in POLE-wild-type endometrioid ECs (10%; P = 0.003), but not significantly different from that in serous ECs (53%). Serous-like features were as often (focally) present in POLE-mutant as in POLE-wild-type endometrioid ECs (6-24%, depending on the feature). The majority of POLE-mutant ECs showed wild-type p53 (86%), negative/focal p16 (82%) and normal mismatch repair protein expression (90%). CONCLUSIONS: A simple combination of morphological and immunohistochemical characteristics (tumour type, grade, peritumoral lymphocytes, MLH1, and p53 expression) can assist in prescreening for POLE exonuclease domain mutations in EC, increasing the probability of a mutation being detected from 7% to 33%. This facilitates the use of this important prognostic biomarker in routine pathology.