OUP user menu

Different HER2 Protein Expression Profiles Aid in the Histologic Differential Diagnosis Between Urothelial Carcinoma In Situ (CIS) and Non-CIS Conditions (Dysplasia and Reactive Atypia) of the Urinary Bladder Mucosa

Sven Gunia MD, Stefan Koch MD, Oliver W. Hakenberg MD, Matthias May MD, Christoph Kakies MD, Andreas Erbersdobler MD
DOI: http://dx.doi.org/10.1309/AJCPKUZ69LXZGFEA 881-888 First published online: 1 December 2011


We evaluated HER2 expression profiles in 32 carcinoma in situ (CIS) and 31 non-CIS conditions (5 dysplasia and 26 reactive atypia) of the urinary bladder mucosa by applying breast cancer scoring rules. In situ hybridization was performed on tissue microarrays to assess HER2 gene amplification status. Our immunoprofiling data disclosed moderate to strong HER2 expression in CIS, including the basal layer of the urothelium, and absent to weak HER2 expression in non-CIS conditions. From the histologic differential diagnostic standpoint, immunostaining for HER2 protein represents a useful adjunct to aid in the delineation between CIS and non-CIS conditions of the bladder mucosa. Pathogenically, aberrant HER2 protein expression in CIS seems to be more commonly associated with polysomy than with gene amplification. From a therapeutic viewpoint, prospective clinical studies should investigate the potential benefit of HER2-targeted therapies in CIS, particularly in cases unresponsive to conventional therapeutic regimens.

Key Words:
  • Flat lesions of urinary bladder mucosa
  • Histologic differential diagnosis
  • Immunohistochemistry
  • HER2 expression profiles

Urothelial carcinoma in situ (CIS) represents a neoplastic flat lesion in which the surface epithelium contains cells showing high-grade cytologic atypia.1 In practice, the conventional histologic differentiation between CIS and non-CIS conditions of the bladder mucosa (dysplasia and reactive atypia) may be challenging owing to possible histomorphologic overlap.1,2 However, pathologic classification is crucial in terms of clinical management and prognosis.37 Although photodynamic diagnosis is recommended by the 2006 European Association of Urology guidelines for the diagnosis of CIS of the bladder, this method has not been implemented on a regular basis in clinical practice owing to its low specificity.8

In surgical pathology practice, only a few diagnostic markers are established to immunohistochemically aid in the differential diagnosis between CIS and non-CIS conditions, which not infrequently yield discrepant staining results when used in combination.2,912 Therefore, additional robust markers are needed to clarify the classification of conventionally and immunohistochemically equivocal bladder biopsy findings.

The human epidermal growth factor receptor-2 gene (HER2, also known as ERBB2 and HER2/neu) has been identified as an independent predictor for disease-related survival in muscle-invasive urothelial carcinomas (UCs).1316 However, its role in CIS has not been clarified. We looked at the so-far neglected diagnostic significance of HER2 protein expression profiles in flat lesions of the urinary bladder mucosa in terms of differentiating CIS from non-CIS conditions.

Materials and Methods

Patients and Samples

We used retrospective computerized database analysis to search the internal pathology database of the Institute of Pathology, HELIOS Clinic Bad Saarow, Charité-University Medicine Teaching Hospital, Bad Saarow, Germany, for cases of urothelial CIS, dysplasia, and reactive atypia. Archived conventionally H&E-stained sections were retrieved from the files, and tissue quality issues were carefully examined to exclude poorly oriented sections and specimens containing only minute foci of urothelium, cautery artifacts of the urothelium, metaplasia, or severe crushing artifacts. There were 63 adequately oriented sections obtained from 63 patients (52 men; mean age at diagnosis, 67.9 years; range, 43–87 years) remaining available for studies for which institutional review board approval was obtained by the Landesärztekammer Brandenburg. In all selected study cases, the entire resected tissue was subjected to histologic examination. The originally signed out database diagnoses made by at least 1 experienced senior clinical pathologist not involved in the performance of this study were as follows: CIS (31 cases), dysplasia (7 cases), and reactive atypia of the urothelium (25 cases).

Immunohistochemical Studies

Following deparaffinization of wax-embedded tissue specimens cut at 4 μm, antigen demasking was accomplished by heat retrieval (100°C) in 0.01 mol/L citrate buffer for 30 minutes, automatically performed by the bond system (Vision BioSystems, Melbourne, Australia). Then, sections were stained for cytokeratin (CK)20 (monoclonal, Progen Biotechnik, Heidelberg, Germany), p53 (monoclonal, Ventana, Tucson, AZ), and HER2 protein using the HercepTest (DAKO Denmark A/S, Glostrup, Denmark) according to the manufacturers’ protocols. Primary antibodies were incubated at room temperature for 1 hour. Subsequently, sections were washed with phosphate-buffered saline and incubated with rabbit antimouse IgG, dilution 1:50, and following that, mouse peroxidase-antiperoxidase conjugate, dilution 1:200.

Positive controls for p53 (breast cancer with adjacent normal breast tissue) and CK20 (colonic adenocarcinoma with adjacent normal mucosa) were consistently performed when immunohistochemical analysis was used to aid in the classification of conventionally challenging study cases. Moreover, HER2+ standard controls were run in all study cases according to the manufacturer’s instructions because internal quality assurance is pivotal for HER2 protein expression scoring.17,18 Negative controls were also consistently performed and had primary antibodies replaced by buffer.

Tissue Microarray Construction and Automated Silver-Enhanced In Situ Hybridization

The tissue microarray technique was used to investigate HER2 gene amplification. Suitable areas for tissue retrieval were marked on H&E-stained sections and punched out of the paraffin block (1.5-mm punch diameter), and 1 punch from each study case was inserted into a recipient block. Then, the tissue array was cut into 4-μm-thick sections. Evaluation of HER2 gene amplification status was performed on 48 study cases because 15 samples (24%) were lost during tissue arraying.

Briefly, deparaffinized 4-μm-thick tissue microarray sections were subjected to automated silver-enhanced in situ hybridization (SISH) performed according to the manufacturer’s instructions for the INFORM HER2 DNA and chromosome 17 probes (Ventana). Both probes were labeled with dinitrophenol and optimally formulated for use with the ultraVIEW SISH Detection Kit (Ventana) and accessory reagents on the Ventana BenchMark series of automated slide stainers. Following pretreatment with CC2 (citrate buffer pH6, 3 times at 90°C for 12 minutes), digestion with protease 3 (enzyme activity 0.02 U/mL) was performed (16 minutes at 37°C). The HER2 DNA probe was denatured at 95°C for 12 minutes, and hybridization was performed at 52°C for 3 hours. Subsequently, stringency washes (3 times for 8 minutes at 72°C with 2× saline sodium citrate solution) were done. The chromosome 17 probe was denatured at 95°C for 12 minutes, and hybridization was carried out at 44°C for 2 hours. Following that, stringency washes (3 times for 8 minutes at 59°C) were repeated. Both probes were visualized by using the rabbit antidinitrophenol linker antibody. For detection of the HER2 gene, the ultraVIEW SISH Detection Kit was used with horseradish peroxidase–labeled secondary antibody (SISH DET HRP) incubated for 16 minutes at 37°C. For chromosome 17 detection, the ultraVIEW alkaline phosphatase (AP) Red ISH Detection Kit (Ventana) was used. This kit contains the AP-labeled secondary antibody DET AP multimer, which was used for 12 minutes at 37°C. The slides were then counterstained with Ventana Hematoxylin II.

Evaluation of Study Cases

Based on conventionally H&E-stained sections, both raters (S.G. and A.E.) independently classified all study cases as reactive atypia of the urothelium, dysplasia, or CIS according to the diagnostic criteria devised by the recent World Health Organization Classification of Tumours.19 Because the histologic criteria for distinguishing severe dysplasia from CIS are unreliable, they were combined into a single category, namely, CIS.20

After a few days had elapsed, both raters assessed the immunostained slides in 3 subsequent rounds of histologic examination, again separated by a period comprising a few days to avoid possible subconscious influence caused by the previously read tissue sections. Moreover, the order in which the study cases were arranged in a box was changed coincidentally before every new round of histologic examination, and both raters were blinded with respect to patient identification and clinical information, the originally signed out diagnoses, and the diagnostic category previously allocated by themselves and by the other pathologist.

Expression of CK20 and p53 was evaluated in a dichotomized manner (CK20 expression in umbrella cells only vs throughout the entire urothelium; p53 expression absent or with only low nuclear staining intensity predominantly in the lower half of the urothelium vs at least moderate staining intensity throughout the entire mucosa). HER2 scoring was performed by separately using the breast cancer scoring rules to the upper (luminal) and lower (basal) half of the urothelium considering only linear membranous staining. By adopting the 10% area cutoff, each study case was scored as 0, 1+ (unequivocal intercellular membranous staining confirmed only at ×40 objective), 2+ (membranous staining disclosed at medium magnification, ×10 objective), or 3+ (unequivocal membranous staining already visible at low magnification, ×2.5 objective). Only distinct, complete, membranous intercellular linear staining (chicken wire–type at cell-cell contact sites) was scored. Cytoplasmic and/or nuclear staining and incomplete (eg, basolateral or only lateral) staining, granular (not linear) pseudomembranous staining, and ring-shaped staining of a single isolated cell surrounded by an artificial shrinkage rim were not scored.

To qualify as CIS, both raters had to classify the case as CIS conventionally, or the lesion had to be classified as CIS by 1 pathologist based on conventionally stained sections, and the staining patterns for CK20 and p53 were required to confirm this classification. The dysplastic category was assigned when both raters independently classified the case as dysplasia based on H&E-stained sections. The remaining cases were classified as reactive atypia of the bladder urothelium.

Evaluation of in situ hybridization was performed according to the American Society of Clinical Oncology/College of American Pathologists guidelines.21 Briefly, at least 50 nuclei were selected for scoring in each study case. Only nonoverlapping intact nuclei of urothelial cells were scored, and nuclei showing only 1 spot and those without any spots were excluded from analysis. Umbrella cells of the urothelium were not considered because they can be tetraploid in the absence of malignancy.22,23 HER2 gene amplification status was classified as negative (HER2/chromosome 17 ratio <1.8), equivocal for HER2 gene amplification (HER2/chromosome 17 ratio between 1.8 and 2.2), or positive (HER2/chromosome 17 ratio >2.2).21

Statistical Analysis

For statistical analysis, the statistical software package SPSS 18.0 was used (SPSS, Chicago, IL). Briefly, κ values were calculated to assess interobserver variability in evaluating immunostaining profiles. Spearman analyses were performed to assess the correlation between the HER2 scores assigned by both raters. Receiver operator characteristic curves for diagnostic testing and sensitivity, specificity, accuracy, and positive and negative predictive values were calculated for each marker evaluated. For all calculations performed, the significance level was set at .05.


According to consensus diagnoses, 32 CIS and 31 non-CIS conditions (5 dysplasia and 26 reactive atypia of the urothelium) were assessed in the present study.

The κ statistics showed low interobserver agreement in HER2 protein expression in the upper vs lower half of the mucosa (κ = 0.03, P = .54 vs κ = 0.18, P = .003) and excellent interobserver agreement in terms of CK20 and p53 staining evaluation (κ = 0.67, P < .001 vs κ = 0.93, P < .001). Although 1 rater had a tendency to assign higher HER2 scores to the study cases, the scores independently assigned by both raters showed a significant positive correlation for the upper (Rho = 0.734; P < .001) and lower (Rho = 0.853; P < .001) halves of the urothelium Table 1.

Concerning the lower (basal) half of the urothelium, non-CIS conditions showed absent or only weak HER2 expression with sparing of the basal cell layer in all non-CIS study cases evaluated (Table 1) Image 1 and Image 2, whereas the majority of CIS cases displayed moderate to strong HER2 expression, including the basal cell layer of the neoplastic urothelium (Table 1) Image 3. This difference was statistically significant (rater A, mean score, 1.59 vs 0, P < .001; rater B, mean score, 2.09 vs 0.77, P < .001).

The difference in HER2 expression encountered in the upper (luminal) half of the urothelium also differed significantly between CIS and non-CIS conditions (rater A, mean score, 1.72 vs 0.77, P < .001; rater B, mean score 2.37 vs 1.87, P = .003).

The 5 dysplastic study cases showed HER2 expression profiles closely related to the staining profiles observed in reactive atypia of the urothelium (Table 1). In particular, all dysplastic study cases assessed failed to exhibit HER2 expression in the basal cell layer of the urothelium (Image 2).

Sensitivity, specificity, positive and negative predictive values, and accuracy for the 3 markers evaluated are summarized in Table 2.

The receiver operating characteristic analyses Table 3, Figure 1, Figure 2, and Figure 3 calculated an area under the curve equal to 0.75 (rater A, 95% confidence interval, 0.62–0.87; P = .001) and 0.86 (rater B, 95% confidence interval, 0.76–0.96; P < .001), indicating that HER2 expression in the lower half of the urothelium has a high accuracy for delineating CIS from non-CIS conditions.

View this table:
Table 1

SISH performed on 26 CIS and 22 non-CIS conditions confirmed absent HER2 gene amplification in all non-CIS conditions and in 25 CIS cases, whereas 1 CIS case was equivocal for HER2 gene amplification.


The diagnosis of CIS has important prognostic and therapeutic implications for patients. The present study assessed the so-far neglected diagnostic role of HER2 protein expression profiles in flat lesions of the urinary bladder mucosa, placing the focus on differentiating CIS from non-CIS conditions (dysplasia and reactive atypia of the urothelium). To this aim, HER2 expression was determined by separately applying the breast cancer immunohistochemical scoring rules to the upper and lower halves of the bladder urothelium to account for heterogeneous HER2 expression within the mucosa in non-CIS conditions (Images 1 and 2).

Image 1

Reactive atypia of the urinary bladder mucosa shows weak HER2 protein expression predominantly in the upper (luminal) half of the mucosa (anti-HER2/neu, ×20 objective).

Image 2

HER2 protein expression encountered in urothelial dysplasia (anti-HER2/neu, ×20 objective).

View this table:
Table 2
View this table:
Table 3

No internal HER2+ standard controls were provided to the raters, which clearly explains the interindividual differences in immunohistochemical scoring in our study (Table 1). This approach was chosen for 2 reasons. First, we aimed to avoid interlaboratory staining variation not infrequently encountered in daily practice. Second, in practical terms, concerning the question of how evaluation of HER2 status could be applied in assisting to resolve the ongoing challenge of discriminating between CIS and non-CIS conditions, the presence or absence of HER2 expression in the lower half of the urothelium and, particularly, in the urothelial basal cell layer should be considered according to our data, irrespective of the “absolute score” of marker expression. The significant positive correlation between the scores independently assigned by both raters supports this notion.

Image 3

Urothelial carcinoma in situ displays HER2 protein expression in the upper (luminal) and lower (basal) half of the urothelium, including the basal cell layer of the neoplastic urothelium (anti-HER2/neu, ×20 objective).

In terms of comparing the diagnostic performance of the 3 markers used, staining evaluation of CK20 and p53 showed superior interobserver agreement, whereas HER2 expression in the lower half of the urothelium turned out to be superior to CK20 and p53 in specificity and positive predictive value for differentiating CIS from non-CIS conditions (Table 2). Therefore, according to our data, HER2 immunohistochemical analysis might contribute to clarification of the classification of conventionally equivocal bladder biopsy samples exhibiting discrepant staining profiles with respect to established diagnostic markers.2,912

Except for 1 CIS case equivocal for HER2 gene amplification, the latter was absent in all CIS cases investigated in our study. The slightly different results recently published by Schwarz et al24 might be attributable to the different methods used for in situ hybridization (FISH vs SISH). In agreement with the latter study, our data suggest that aberrant HER2 expression in CIS seems to be attributable more commonly to polysomy than to gene amplification.24 These findings mirror the situation encountered in invasive UC in which HER2 overexpression is accompanied by HER2 gene amplification in only a subset of cases.25 On the other hand, at variance with the high rate of HER2 expression in CIS observed in our study, other investigators reported HER2 overexpression to be present in only 8.2% vs 9.2% of superficial vs muscle-invasive UC.13,26 These data suggest a different pathogenic role of HER2 overexpression that might be stage-dependent during the multistep carcinogenesis of invasive bladder cancer.

Figure 1

Receiver operating characteristic graphs for HER2 protein expression in the upper and lower half of the urothelium scored by 2 independent raters. Top, upper (luminal) half of the urothelium; Bo, lower (basal) half of the urothelium; R1, rater A; R2, rater B.

Figure 2

Receiver operating characteristic graphs for cytokeratin (CK)20 expression in the urothelium assessed by 2 independent pathologists. R1, rater A; R2, rater B.

Figure 3

Receiver operating characteristic graphs for nuclear p53 expression in the urothelium evaluated by 2 independent raters. R1, rater A; R2, rater B.

Patients with invasive bladder UC have recently been suggested as potential candidates for targeted therapy, and chromogenic in situ hybridization provides an accurate and practical alternative to FISH for the detection of HER2 gene amplification in UC.13,27,28 However, the role of HER2 expression in CIS with regard to potential targeted therapy has not been clarified. We are aware that possible therapeutic implications of HER2 overexpression in CIS are speculative at present. However, our data hint at the need for advanced prospective clinical studies aiming to clarify whether HER2-targeted therapies are beneficial in CIS, particularly in cases unresponsive to currently established therapeutic regimens.

Limitations of our study that merit further discussion include the low number of dysplastic study cases evaluated and the fact that discohesive CIS and non-CIS conditions were not assessed. Although our data suggest distinctive HER2 expression patterns in urothelial dysplasia and CIS (Images 2 and 3) that might be more helpful to separate both lesions compared with p53 and CK20, the number of dysplastic cases examined is not sufficient to draw definitive conclusions in this regard. However, the histologic differentiation of urothelial dysplasia (which, by definition, is a low-grade lesion) from reactive atypia is of little clinical importance at present because no uniform guidelines have been established for the management of patients with dysplastic bladder urothelium.

Concerning the second limitation of our study, mucosal denudation is not infrequently encountered in bladder biopsies performed by cold-cup forceps or wire loop electrocautery.29 In this scenario, sometimes only a few urothelial cells situated (“clinging”) on top of the basement membrane are retained for histologic examination. Notably, according to our findings, HER2 immunostaining is of particular diagnostic value to resolve these challenging cases because it is especially the HER2 expression pattern encountered in the basal urothelial cells that is helpful in the differentiation between CIS and non-CIS conditions (moderate to strong in CIS vs absent to weak in non-CIS conditions). In this regard, HER2 offers a strong advantage over other diagnostically useful markers that require intact full-thickness urothelial mucosa for reliable interpretation.2,912

Our study hints at a so-far neglected diagnostic role of HER2 expression in flat lesions of the urinary bladder mucosa. From a diagnostic standpoint, our data suggest that different HER2 protein expression profiles encountered in the lower half of the urothelium aid in the differentiation between CIS and non-CIS conditions (dysplasia and reactive atypia of the urothelium). Pathogenically, aberrant HER2 protein expression in CIS seems to be more commonly associated with polysomy than with gene amplification. From a therapeutic viewpoint, our findings invite advanced prospective clinical studies to clarify the so-far unresolved future role of HER2-targeted therapies in CIS of the urinary bladder.


Upon completion of this activity you will be able to:

  • define the immunohistochemical HER2 expression profiles in various flat urothelial lesions of the urinary bladder.

  • employ HER2 expression profiles to differentiate various types of flat urothelial lesions.

  • describe gaps in current knowledge regarding clinical implications of HER2 expression in urothelial carcinoma in situ.

The ASCP is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The ASCP designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 Credit ™ per article. Physicians should claim only the credit commensurate with the extent of their participation in the activity. This activity qualifies as an American Board of Pathology Maintenance of Certification Part II Self-Assessment Module.

The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose.

Questions appear on p 994. Exam is located at www.ascp.org/ajcpcme.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
View Abstract