Search

The expression of α-methylacyl-coenzyme-A racemase (AMACR) has previously been reported in 75% to 100% of urethral/bladder clear cell carcinomas, tumors that are known to display broad phenotypic overlap with their identically named mullerian counterparts. Herein, we assess the utility of AMACR in distinguishing endometrial clear cell carcinomas (CCCs) from endometrial serous carcinomas (ESCs) and endometrial endometrioid carcinomas (EECs). A total of 111 endometrial carcinomas in a tissue microarray, including 49 CCCs, 13 ESCs, and 49 EECs, were assessed for AMACR immunoreactivity, with results scored semiquantitatively (scores 0, 1+, 2+, 3+ for 0%, 1%-5%, 6%-50%, >50% immunoreactive cells, respectively). Fifty (45%) of the 111 carcinomas were AMACR positive, with the following score distribution: CCC: 0 (n = 12), 1+ (n = 12), 2+ (n = 3), 3+ (n = 22); EEC: 0 (n = 38), 1+ (n = 4), 2+ (n = 4), 3+ (n = 3); ESC: 0 (n = 11), 1+ (n = 1), 2+ (n = 0), 3+ (n = 1). AMACR expression was significantly more frequent in CCC (75%) than in ESC (15%) or EEC (22%); P < .0001. The sensitivity and specificity of AMACR expression in classifying a carcinoma as CCC were 0.75 (95% confidence interval [CI], 0.61-0.86) and 0.79 (95% CI, 0.66-0.88), respectively, with an odds ratio of 11.62 (95% CI, 5-28; P < .001) and an area under the curve of 0.79 (95% CI, 0.68-0.88). These findings indicate that AMACR expression is strongly associated with CCC and displays a relatively robust diagnostic test performance. However, its practical utility may be limited by the focal nature of its expression in 32% of the AMACR-positive CCC cases as well as its expression in 15% to 22% of the non-CCC histotypes.

Phenotypic differences between otherwise similar tumors arising from different gynecologic locations may be highly significant in understanding the underlying driver molecular events at each site and may potentially offer insights into differential responses to treatment. In this study, the authors sought to identify and quantify phenotypic differences between ovarian clear cell carcinoma (OCCC) and endometrial clear cell carcinoma (ECCC) using a proteomic approach. Tissue microarrays were constructed from tumor samples of 108 patients (54 ECCCs and 54 OCCCs). Formalin-fixed samples on microarray slides were analyzed by matrix-assisted laser desorption/ionization mass spectrometry, and 730 spectral peaks were generated from the combined data set. A linear mixed-effect model with random intercept was used to generate 93 (12.7%) peaks that were significantly different between OCCCs and ECCCs at the fold cutoffs of 1.5 and 0.667 and an adjusted P value cutoff of 1.0 × 10(-10). Liquid chromatography-tandem mass spectrometry was performed on selected cores from each group, and peptides identified therefrom were compared with lists of statistically significant peaks from the aforementioned linear mixed-effects model to find matches within 0.2 Da. A total of 53 candidate proteins were thus identified as being differentially expressed in OCCCs and ECCCs, 45 (85%) of which were expressed at higher levels in ECCCs than OCCCs. These proteins were functionally diverse and did not highlight a clearly dominant cellular theme or molecular pathway. Although ECCCs and OCCCs are very similar, some phenotypic differences are demonstrable. Additional studies of these differentially expressed proteins may ultimately clarify the significance of these differences.

Small cell neuroendocrine carcinoma of the prostate is a rare variant of prostatic cancer that shares morphologic similarity with prostatic adenocarcinoma of Gleason 5 pattern. It has also been considered morphologically and immunohistochemically indistinguishable from small cell neuroendocrine carcinomas of other origins. CD44 is a cell-surface molecule proposed to identify cancer stem/progenitor cells in prostate cancer. We performed immunohistochemical study for CD44 expression in 11 cases of prostatic small cell neuroendocrine carcinoma and compared its patterns of expression with 73 cases of prostatic adenocarcinoma and 47 cases of small cell neuroendocrine carcinomas of other organs. Strong and diffuse membrane staining for CD44 was observed in 100% of the prostatic small cell neuroendocrine carcinomas. In conventional adenocarcinomas of the prostate, positive staining was only seen in rare, scattered tumor cells; and CD44 staining was negative in most of the small cell neuroendocrine carcinomas of nonprostate origin. The difference in CD44 expression between small cell neuroendocrine carcinomas of the prostate and those of other organs are statistically significant (P < .001). Our study demonstrates the utility of immunohistochemical staining for CD44 in distinguishing prostatic small cell neuroendocrine carcinoma from its mimickers including prostatic adenocarcinoma of Gleason 5 pattern and small cell neuroendocrine carcinomas of other organs. CD44 is the first marker that shows a high degree of tissue/organ specificity for small cell neuroendocrine carcinomas. Because CD44 is a putative marker of prostate cancer stem cells, the strong and diffuse expression of CD44 and the lack of expression of prostate luminal differentiation markers androgen receptor and prostatic specific antigen in prostatic small cell neuroendocrine carcinomas suggest that the tumor cells may retain cancer stem cell features.