Interpretation of P53 Immunohistochemistry in Endometrial Carcinomas: Toward Increased Reproducibility

Abstract

P53 immunohistochemistry has evolved into an accurate surrogate reflecting the underlying TP53 mutation status of a tumor, and has utility in the diagnostic workup of endometrial carcinomas. Recent work predominantly carried out in tubo-ovarian high-grade serous carcinoma has revealed 4 main patterns of p53 staining (normal/wild-type, complete absence, overexpression, and cytoplasmic); the latter 3 patterns are variably termed abnormal/aberrant/mutation-type and are strongly predictive of an underlying TP53 mutation. The aim of this review is to provide practical advice to pathologists regarding various aspects of p53 immunohistochemical staining. These include laboratory methods to optimize staining, a description of the different patterns of staining, advice regarding the interpretation, and reporting of p53 staining and practical uses of p53 staining in endometrial carcinoma diagnosis. Illustrations are provided to aid in the interpretational problems.

FIG. 1

Different patterns of p53 expression. (A) Endometrial endometrioid carcinoma showing normal wild-type pattern of p53 expression with variable proportion of tumor cell nuclei staining with variable intensity. Note, this wild-type pattern should not be reported as “positive,” because this is ambiguous reporting language. (B) Endometrial endometrioid carcinoma, grade 3, with overexpression, showing strong staining in virtually all tumor cell nuclei, much stronger compared with the internal control of fibroblasts in the center. Note, there is some cytoplasmic background indicating that this staining is quite strong but this should not be interpreted as abnormal cytoplasmic pattern. (C) Endometrial serous carcinoma showing complete absence of p53 expression with internal control showing moderate to strong but variable staining. Note, wild-type pattern in normal atrophic glands at 12 and 6 o’clock. (D) Endometrial endometrioid carcinoma showing cytoplasmic p53 expression with internal control (stroma and normal endometrial glands) showing nuclear wild-type pattern. The cytoplasmic pattern is accompanied by nuclear staining of similar intensity.

FIG. 2

(A, B) Two low-grade endometrial endometrioid carcinomas with p53 staining in the majority of tumor cell nuclei but with variable intensity (some staining strong, some moderate, some weak, few negative). The intensity is within the range of the internal control but less intense compared with Figure 1B (same protocol). There is some cytoplasmic blush suggesting that the staining is bordering on too strong. (C, D) Two endometrial serous carcinomas with areas of “mosaic” staining bordering wild-type pattern (more in C than D) that show otherwise abnormal overexpression pattern in the remainder of the sample (see inset). The distinction of spurious “mosaic” staining due to poor fixation from true heterogenous staining (compare with Fig. 4) can be challenging.

FIG. 3

(A, B) Endometrial serous carcinoma with complete absence pattern of abnormal p53 expression stained on 2 different platforms. (A) Nonspecific nuclear staining interpreted as wild-type pattern; (B) shows complete absence of nuclear staining but a weak cytoplasmic blush indicating staining bordering on too strong. (C) Endometrial endometrioid carcinoma with wild-type staining with slight cytoplasmic blush on the left and true abnormal cytoplasmic staining on the right (compare with low-power view in Fig. 4C). The true abnormal cytoplasmic staining is accompanied by a variable nuclear staining of similar intensity but not strong diffuse. (D) Endometrial endometrioid carcinoma with wild-type pattern showing weak cytoplasmic staining probably due to too strong staining. This should not be interpreted as abnormal cytoplasmic staining.

FIG. 4

(A) Endometrial endometrioid carcinoma showing a combination of wild-type pattern on the left and overexpression on the right (normal+abnormal=heterogenous). (B) Endometrial endometrioid carcinoma with heterogenous staining (normal wild-type pattern+abnormal overexpression and complete absence). (C) Endometrial endometrioid carcinoma with heterogenous staining (normal wild-type pattern and abnormal overexpression and cytoplasmic staining). (D) Endometrial endometrioid carcinoma with heterogenous staining (normal wild-type pattern and abnormal overexpression and complete absence). (E) High power from (C) showing transition from overexpression to wild-type pattern. (F) Endometrial endometrioid carcinoma with variable wild-type pattern (not heterogenous), “high” wild-type on left versus “low” wild-type on the right.

FIG. 5

Staining results from 4 immunohistochemistry platforms: protocol #1 used EnVision Flex target retrieval solution (TRS), high pH, ready to use (RTU) primary antibody clone DO7 for 30 min with 10 min linker. Protocol #2 used the same as #1 except the RTU primary antibody clone DO7 was diluted (1/5). Protocol #3 used EnVision Flex TRS, high pH, RTU primary antibody clone DO7 for 20 min without linker (Dako vendor protocol). Protocol #4 used the same as #3 except the RTU primary antibody clone DO7 was diluted (1/10). Six specimens representing the 4 p53 staining patterns in columns from left to right: wild-type pattern in a germinal center of tonsil (“low expressor” positive control), overexpression in an endometrial serous carcinoma (“high expressor” positive control), 2 wild-type endometrial endometrioid carcinomas showing the range of wild-type expression, cytoplasmic and complete absence of staining (note presence and absence of internal control).