Frequent expression of napsin A in clear cell carcinoma of the endometrium: potential diagnostic utility.

The histotyping of high-grade endometrial carcinomas with clear cells may be subject to significant interobserver variability, which suggests that a biomarker that can distinguish endometrial clear cell carcinoma (CCC) from its mimics would be of diagnostic utility. This study assessed the usefulness of napsin A immunohistochemistry in the diagnosis of CCC, on the basis of an analysis of 77 cases diagnosed as such at 9 institutions. After being independently reviewed by a subset of 3 pathologists, cases for which there was diagnostic consensus among all 3 reviewers in agreement with the primary contributor (n=60) were used to establish a "consensus group" that served as a gold standard relative to which napsin A performance was assessed. Duplicate, 1.0-mm-core tissue microarrays were constructed from the 54 cases in the consensus group for which requisite materials were available, as well as from 49 endometrial endometrioid carcinomas (all grades) and 17 endometrial serous carcinomas. Napsin A immunohistochemical analysis was performed on the microarrays and on the 17 cases for which there was no diagnostic consensus, with scoring based on the proportion of immunoreactive cells (0, 1+, 2+, and 3+ indicative of 0, 1% to 25%, 26% to 49%, and ≥50% immunoreactive cells, respectively). The distribution of scores for the 49 CCC cases with evaluable cores was as follows: 0, n=6; 1+, n=6; 2+, n=8; 3+, n=29. Among the evaluable cases, the frequency of ≥1+ napsin A immunoreactivity was significantly higher in CCCs (43/49, 88%) than in endometrial serous carcinomas (1/13, 7.7%; P<0.0001) and endometrial endometrioid carcinomas (0/49, 0%; P<0.0001). The sensitivity, specificity, negative predictive value, and positive predictive value of ≥1+ napsin A expression in predicting the consensus clear cell histotype were 0.88 (95% confidence interval [CI], 0.75-0.95), 0.98 (95% CI, 0.9-1), 0.91 (95% CI, 0.86-0.96), and 0.98 (95% CI, 0.86-1), respectively. Napsin A expression was not associated with survival or clinicopathologic factors. In the group of cases without diagnostic consensus for CCC, 50% showed ≥1+ napsin A expression; all napsin A-negative cases had previously been classified as non-CCC by ≥2 reviewers, whereas only 37.5% of the napsin A-positive cases had been classified as CCC by 2 of the 3 reviewers. In conclusion, napsin A is a sensitive and specific biomarker of the clear cell histotype in endometrial carcinomas and accordingly may have diagnostic utility in their histotyping.