Expression of uroplakin II and GATA-3 in bladder cancer mimickers: caveats in the use of a limited panel to determine cell of origin in bladder lesions.

Antibodies targeting uroplakin II (UPII) are highly specific for urothelial cells and are frequently used to determine if a primary bladder lesion or a metastatic lesion originates from the urothelium. However, to date, no studies have tested the expression of UPII in histological mimickers of bladder cancer that are nonurothelial in origin. Given the potential risk of misdiagnosis, immunohistochemical markers are often used to better characterize these lesions. In the present study, we analyzed the immunohistochemical expression of UPII in a set of urothelial carcinoma mimickers that included conventional nephrogenic adenoma (n = 8), papillary nephrogenic adenoma (n = 6), endometriosis/endosalpingiosis (n = 5), inflammatory myofibroblastic tumor (n = 4), ectopic prostate tissue (n = 2), and malakoplakia (n = 2). We also examined the expression of GATA-3, another commonly used immunohistochemical marker in bladder cancer diagnosis, in the same lesions. Weak immunoreactivity for UPII was identified in 6 of 27 mimickers (22%), and GATA-3 was expressed in 16 of 27 mimickers (59%). Strong immunoreactivity for UPII appeared to be a specific marker for urothelial cell of origin, although weak staining was seen in a significant proportion of mimickers. GATA-3 immunostaining was present in a greater number and broader spectrum of mimickers; however, only one case of papillary nephrogenic adenoma showed dual positivity for UPII and GATA-3. These findings support the immunohistochemical panel-based approach in the diagnosis of bladder lesions, especially if nonurothelial bladder cancer mimickers are in the differential diagnosis. Additional larger studies would be of value to expand on these findings.

Urothelial Differences in the Exstrophy-Epispadias Complex: Potential Implications for Management.

PURPOSE: The authors examined the urothelium of exstrophy-epispadias complex spectrum patients for histological differences and expression of terminal markers of urothelial differentiation. MATERIALS AND METHODS: Between 2012 and 2017 bladder biopsies were obtained from 69 pediatric exstrophy-epispadias complex patients. These specimens were compared to bladder specimens from normal controls. All bladder specimens underwent histological assessment followed by immunohistochemical staining for uroplakin-II and p63. Expression levels of uroplakin-II and p63 were then assessed by a blinded pathologist. RESULTS: Forty-three classic bladder exstrophy biopsies were obtained (10 newborn closures, 22 delayed closures, and 11 repeat closures). Additional biopsies from 18 cloacal exstrophy patients and 8 epispadias patients were also evaluated. These specimens were compared to 8 normal control bladder specimens. Overall, uroplakin-II expression was lower in exstrophy-epispadias complex patients compared to controls (p <0.0001). Among classic bladder exstrophy patients, there was reduced expression of uroplakin-II in the delayed and repeat closures in comparison to newborn closures (p=0.045). Expression of p63 was lower in patients with exstrophy-epispadias complex compared to controls (p <0.0001). Expression of p63 was similar among classic bladder exstrophy patients closed as newborns when compared to delayed or repeat closures. Classic bladder exstrophy patients had a higher rate of squamous metaplasia when compared to controls (p=0.044). Additionally, there was a higher rate of squamous metaplasia in the patients undergoing delayed closure in comparison to those closed in the newborn period (p <0.001). CONCLUSIONS: The urothelium in the exstrophy-epispadias complex bladder is strikingly different than that of healthy controls. Uroplakin-II expression is greatly reduced in exstrophy-epispadias complex bladders and is influenced by the timing of bladder closure. Reduced uroplakin-II expression and increased rates of squamous metaplasia in exstrophy-epispadias complex patients undergoing delayed closure suggests that exposure of the urothelium may induce these changes. These findings shed light on the molecular changes in exstrophy-epispadias complex bladders and may have implications on the appropriate timing of primary bladder closure, as those closed in the newborn period appear to have a greater potential for growth and differentiation.

Gynecologic Serous Carcinoma: An Immunohistochemical Analysis of Malignant Body Fluid Specimens.

CONTEXT.­: Evaluation of fluid specimens involved by serous carcinoma might potentially include PAX8, GATA3, Uroplakin II, SOX2, and SALL4 antibodies. Those markers are commonly employed for diagnosing carcinomas of various types, including urothelial malignancies and germ cell tumors. There have been no comprehensive immunohistochemical studies, to our knowledge, for those markers on fluid specimens involved by serous carcinoma. OBJECTIVE.­: To evaluate immunohistochemical markers PAX8, GATA3, SOX2, uroplakin II, and SALL4 in the diagnosis of high-grade serous carcinoma in fluid specimens. DESIGN.­: We examined 113 fluids (96 ascites specimens and 17 pleural fluid specimens) that were positive for carcinoma. Most (94 cases; 83.2%) consisted of high-grade serous carcinoma of Müllerian origin. Nineteen cases of non-high-grade serous carcinoma (including one case of low-grade serous carcinoma) of gynecologic origin were also included as anecdotal data. RESULTS.­: In 113 fluid specimens with positive results for carcinoma, including nonserous types, 99 (87.6%) had positive results for PAX8, 19 (16.8%) for GATA3; 19 (16.8%) for SOX2, 23 (20.4%) for uroplakin II, and 8 (7.1%) for SALL4. Of 94 fluids (83.2%) involved with high-grade serous carcinoma, 84 (89.4%) had positive results for PAX8, 18 (19.1%) for GATA3, 17 (18.1%) for SOX2, 22 (23.4%) for uroplakin II, and 8 (8.5%) for SALL4. Some of these specimens showed reactivity for more than one immunohistochemical marker. CONCLUSIONS.­: Most fluids involving high-grade serous carcinoma showed positive results for PAX8, and some cases expressed GATA3, SOX2, uroplakin II, and SALL4. Serous carcinoma in fluids may be positive for immunohistochemical markers not thought of traditionally as associated with gynecologic malignancy, an important consideration in avoiding misdiagnosis.

Immunohistochemical Differentiation of Plasmacytoid Urothelial Carcinoma From Secondary Carcinoma Involvement of the Bladder.

Plasmacytoid urothelial carcinoma (UC) is a rare variant of UC that can histologically mimic metastatic cancer involving the urinary bladder. A total of 45 cases of plasmacytoid UC were collected and reviewed histologically. The following immunohistochemical markers were performed: CDX2; polyclonal carcinoembryonic antigen (p-CEA); gross cystic disease fluid protein 15 (GCDFP-15); mammaglobin; estrogen receptor (ER); progesterone receptor (PR); GATA 3 and uroplakin II. In all cases, the plasmacytoid variant of UC lacked expression of ER and mammaglobin. In contrast, GCPDFP-15, PR, CDX2 and p-CEA showed positive staining in 11 (24.4%), 6 (13.3%), 8 (17.7%), and 22 (48.8%) cases, respectively. GCPDFP-15 was expressed in 4/8 female cases with 1 concurrently focally (+2) expressing PR. GATA 3 and uroplakin II was positive in 37/45 cases (82.2%) and 15/45 (33.3%) cases, respectively. A tissue microarray with 40 cases of infiltrating lobular carcinoma of the breast was stained for uroplakin II, and was negative in all cases. Tissue microarrays with 46 cases of gastric signet ring cell adenocarcinomas were all negative for GCDFP-15, ER, PR, GATA3, uroplakin II, and mammaglobin. A panel of stains including mammaglobin, ER, and uroplakin II is recommended to exclude metastatic lobular breast carcinoma to the bladder in cases where a conventional UC component is not present. Immunohistochemistry for CDX2 and p-CEA cannot be utilized to differentiate signet ring cell adenocarcinoma of the gastrointestinal tract from plasmacytoid UC; GATA3 or uroplakin II immunoreactivity can rule out a gastric primary given their negativity in signet ring cell adenocarcinoma of the stomach.

GATA3 Positivity in Benign Radiated Prostate Glands: A Potential Diagnostic Pitfall.

Histologic changes following radiation therapy to the prostate include multilayering of glands, atrophy, squamous metaplasia, and often marked random nuclear atypia. We have seen multiple consultation cases where the differential diagnosis of these radiated prostate glands included urothelial carcinoma, with multilayered to solid-appearing proliferations that were positive by immunohistochemistry for GATA3. To formally investigate this issue, 30 cases of benign prostate tissue with radiation atypia, from 1990 to 2015, were obtained from our institution. Cases were evaluated by immunohistochemistry for the prostate-specific markers prostate-specific antigen (PSA), P501S (Prostein), and NKX3.1 and urothelial markers GATA3 and uroplakin 2. GATA3 was positive in 100% of cases, with 70% showing moderately strong to strong staining in a mostly patchy manner within a gland. PSA was positive in 93.3% of cases, with 89.2% showing moderately strong to strong staining in a mostly diffuse manner. P501S was positive in 96.7% of cases, with 93.1% showing moderately strong to strong staining in a mostly patchy manner. NKX3.1 was positive in 82.8% of cases, with 33.3% showing moderately strong to strong staining in a mostly patchy manner. Uroplakin 2 was negative in 100% of cases. Our findings highlight that GATA3 is often positive in benign prostate glands with radiation atypia, which along with the morphologic features present a pitfall for the misdiagnosis of urothelial carcinoma. A combination of PSA and P501S is the best prostate-specific panel for use in radiated prostate, with the caveat that they are often patchy and do not stain all radiated glands.

A combination of p40, GATA-3 and uroplakin II shows utility in the diagnosis and prognosis of muscle-invasive urothelial carcinoma.

Recently developed antibodies against uroplakin II and ΔNp63 (p40) show utility in diagnosing primary bladder urothelial carcinoma, although application in metastatic bladder cancer and patient outcomes has been less well defined. We evaluated these antibodies by immunostain intensity and H-score, in conjunction with GATA-3 immunoreactivity, in 89 patients with muscle-invasive urothelial carcinoma and 35 paired metastasis. The maintenance of immunoreactivity in metastatic lesions and the association to disease recurrence and survival was assessed. Bladder urothelial carcinoma showed immunoreactivity for GATA-3 in 99% (88/89), p40 in 87% (77/89) and uroplakin II in 80% (71/89) of cases, with a positive correlation between GATA-3 and uroplakin II H-score (0.44; p < 0.0001). All lesions were positive for at least one marker, reinforcing the use of these markers as a panel. In 35 patients with paired lymph node metastasis, uroplakin II and GATA-3 were retained in 90% and 75% of patients, respectively, suggesting these markers may have relatively high sensitivity in diagnosing metastatic urothelial carcinoma. High intensity p40 immunostain (3+), however, was significantly associated with reduced patient survival (p = 0.03). These results suggest that whereas GATA-3 and uroplakin II may be most useful in the diagnosis of urothelial carcinoma metastasis, p40 may be additionally suited as a prognostic marker.

Utility of uroplakin II expression as a marker of urothelial carcinoma.

Uroplakins are markers of terminally differentiated urothelium. Uroplakin II (UPII) is a newly described sensitive marker for urothelial carcinoma (UC). The expression profile of UPII in different types of UC and its utility in the diagnostic setting are needed. We evaluated UPII expression in bladder tissue microarrays, including urothelial neoplasm of low malignant potential (n = 8), low-grade papillary UC (n = 72), noninvasive high-grade papillary UC (n = 77), UC in situ (n = 27), and invasive high-grade UC (INVUC) (n = 122). UPII expression in 52 breast carcinomas and 38 high-grade prostate adenocarcinomas was also assessed. UPII expression was compared with GATA binding protein 3 (GATA3) and estrogen receptor for its role in facilitating the differential diagnosis of the above 3 types of malignancy. UPII labeling was seen in 83.0% of UC overall, including 95.7% of noninvasive UC and 65.6% of INVUC. UPII labeling was not found in any breast and prostate carcinomas. In comparison, GATA3 labeling was seen in 91.6% of all UCs, including 96.4% of noninvasive UCs and 85.1% of INVUC, with stronger intensity and extent compared with UPII (P < .005). GATA3 labeled 2 (5%) of 38 high-grade prostate adenocarcinoma. Estrogen receptor nuclear labeling was seen in 13.0% of UCs and 12.5% of prostate carcinomas. UPII was highly specific (100%) but only moderately sensitive for UC and can therefore be a potentially useful marker to identify urothelial lineage and help distinguish UC from prostate cancer or, in conjunction with GATA3, from metastatic breast cancer.

Uroplakin II is a more sensitive immunohistochemical marker than uroplakin III in urothelial carcinoma and its variants.

OBJECTIVES: Uroplakin (UP) II and UPIII are highly specific immunohistochemical markers for urothelial differentiation. Here we studied the sensitivity of UPII and UPIII in conventional and variant urothelial carcinomas (UCs). METHODS: Immunohistochemical staining for UPII and UPIII was performed on tissue microarray slides, including 105 conventional bladder UCs (BUCs), 90 upper urinary tract UCs (UUTUCs), and 47 micropapillary, 16 plasmacytoid, 22 small cell carcinoma, and 41 sarcomatoid UC variants. RESULTS: UPII expression was significantly higher than UPIII expression in conventional BUC (44% vs 17%, P < .001) and UUTUC (67% vs 46%, P = .045). UPIII expression was significantly higher in UUTUC than in BUC (P < .001). In UC variants, UPII expression was significantly higher than UPIII expression in micropapillary (91% vs 25%, P < .001), plasmacytoid (63% vs 6%, P < .001), and sarcomatoid (29% vs 5%, P = .032) variants. Only rare cases of the small cell carcinoma variant had focal UPII and UPIII expression. Compared with conventional UC, the sarcomatoid variant had significantly lower UPII expression, whereas the micropapillary variant had significantly higher UPII expression (P < .001). CONCLUSIONS: UPII demonstrates a significantly higher sensitivity than UPIII in conventional and variant UCs. Thus, UPII is a more valuable marker than UPIII in immunohistochemical analyses for confirming the urothelial origin of carcinomas.

Uroplakin II outperforms uroplakin III in diagnostically challenging settings.

AIMS: We performed a head-to-head comparison of an antibody against uroplakin III (UP3) and a new uroplakin II (UP2) antibody that remains untested in diagnostically challenging settings. METHODS AND RESULTS: We immunostained high-grade bladder neck carcinomas (n = 35), high-grade upper tract urothelial carcinomas (UC) and renal carcinomas (n = 85), metastases of UC (n = 30) and a multicancer tissue microarray (n = 88) for UP3 and UP2, and scored staining intensity and proportion. UP3 showed membranous plaque-like expression, while UP2 staining showed both membranous and cytoplasmic positivity. Significantly greater intensity (P = 0.003) and proportion (P = 0.03) of staining was noted for UP2 among bladder neck lesions, with UP2 staining showing greater sensitivity (63% versus 19%) and similar specificity (95% versus 100%) for UC over prostate carcinoma (P = 0.02). Among upper tract lesions, UP2 staining showed greater intensity and proportion than UP3 (both P < 0.001), including improved sensitivity (68% versus 23%) and equal specificity (both 100%) for UC (P = 0.006). Among UC metastases, UP2 staining showed greater intensity and proportion (both P < 0.001) with higher sensitivity (73% versus 37%, respectively, P = 0.001). Of 88 additional cases tested, no non-urothelial cases stained for either UP. CONCLUSIONS: The UP2 antibody outperforms the UP3 antibody, including in diagnostically challenging settings, and is a useful addition to the armamentarium of biomarkers for UC.

Expression and localization of four uroplakins in urothelial preneoplastic lesions.

In superficial umbrella cells of normal urothelium, uroplakins (UPs) are assembled into urothelial plaques, which form fusiform vesicles (FVs) and microridges of the apical cell surface. Altered urothelial differentiation causes changes in the cell surface structure. Here, we investigated ultrastructural localization of UPIa, UPIb, UPII and UPIIIa in normal and cyclophosphamide-induced preneoplastic mouse urothelium. In normal urothelium, terminally differentiated umbrella cells expressed all four UPs, which were localized to the large urothelial plaques covering mature FVs and the apical plasma membrane. The preneoplastic urothelium contained two types of superficial cells with altered differentiation: (1) poorly differentiated cells with microvilli and small, round vesicles that were uroplakin-negative; no urothelial plaques were observed in these cells; (2) partially differentiated cells with ropy ridges contained uroplakin-positive immature fusiform vesicles and the apical plasma membrane. Freeze-fracturing showed small urothelial plaques in these cells. We concluded that in normal urothelium, all four UPs colocalize in urothelial plaques. However, in preneoplastic urothelium, the growth of the uroplakin plaques was hindered in the partially differentiated cells, leading to the formation of immature FVs and ropy ridges instead of mature FVs and microridges. Our study demonstrates that despite a lower level of expression, UPIa, UPIb, UPII and UPIIIa maintain their plaque association in urothelial preneoplastic lesions.