Hepatocyte nuclear factor 1-beta (HNF1β) has recently emerged as a relatively sensitive and specific marker for ovarian clear cell
carcinoma. The purpose of this study is to assess the diagnostic utility of this marker for endometrial clear cell
carcinoma. Immunohistochemical analysis was performed on 75 endometrial tissues using a goat polyclonal antibody raised against a
peptide mapping at the C-terminus of human HNF1β
protein. The 75 cases included 15 clear cell
carcinomas, 20
endometrioid carcinomas, 15 endometrial serous
carcinomas/uterine papillary serous
carcinomas, 20 cases of normal endometrium, 2 cases of clear cell
metaplasia, and 3 cases of Arias Stella reaction. Staining interpretations were based on a semiquantitative scoring system, a 0 to 12+ continuous numerical scale that was derived by multiplying the extent of staining (0 to 4+ scale) by the intensity of staining (0 to 3+ scale) for each case. HNF1β expression was found to be present in a wide spectrum of tissues. Twenty-seven (54%) of the 50
carcinomas displayed at least focal nuclear HNF1β expression, including 11 (73%) of 15, 9 (60%) of 15, and 7 (35%) of 20 clear cell, serous, and
endometrioid carcinomas, respectively. The average nuclear staining scores for clear cell
carcinomas, endometrioid carcinomas, and serous
carcinomas were 5.2, 1.4, and 4.1, respectively. Clear cell
carcinomas and
endometrioid carcinomas displayed statistically significant differences regarding their nuclear staining scores (P = 0.0027), but clear cell
carcinomas and endometrial serous
carcinomas did not (P = 0.45). The calculated sensitivity of any nuclear HNF1β expression in classifying a
carcinoma as being of the clear cell histotype was 73%, whereas the specificity was 54%. Nineteen of 20 normal endometrium samples displayed at least focal nuclear expression of HNF1β, and this expression was often diffuse. The 5 cases of benign histologic mimics of clear cell
carcinomas (Arias Stella reaction and clear cell
metaplasia) displayed some degree of HNF1β immunoreactivity, with an average nuclear staining score of 7.3. We conclude that although HNF1β is frequently expressed in clear cell
carcinomas, it should be used with caution as a diagnostic marker because of its lack of specificity. It neither distinguishes endometrial serous
carcinomas from clear cell
carcinomas nor clear cell
carcinomas from its benign mimics. The greatest diagnostic utility of HNF1β expression may be in a supportive evidentiary role favoring clear cell
carcinoma when the principal differential diagnostic consideration is
endometrioid carcinoma.