Expression of Renal Cell Carcinoma Antigen (RCC) in Renal Epithelial and Nonrenal Tumors Diagnostic Implications
Nasir Bakshi, MD,* Lakshmi P. Kunju, MD,; Thomas Giordano, MD,t and Rajal B. Shah, MD+ ¿
Abstract: Antibody to renal cell carcinoma (RCC) antigen, a normal human proximal brush border antigen, has recently become commercially available and reported to be highly specific and a relatively sensitive marker for RCC. Of the nonrenal tumors occasional carcinomas have been reported to express RCC, notably breast carcinoma. Using tissue micro- arrays, we investigated the use of RCC on a large number of renal epithelial neoplasms (RENs) and nonrenal tumors, especially those potentially confused with REN. Three tissue microarrays containing 241 REN samples, 192 samples of a wide variety of neoplasms and 170 adrenal tumor samples, respectively, were stained with RCC monoclonal antibody. RCC expression was scored for staining intensity and percentage expression. Out of 241 REN, 173 were positive for RCC (sensitivity 72%): clear cell 72%, papillary 95%, chromophobe 91%, unclassified 85%, oncocytoma 75%, sarcomatoid 20%, and metastatic RCC 40%. The overall immunostaining intensity was consistently much higher in papillary and clear cell RCC than in other tumors. Seventy-six out of 362 nonrenal tumor samples demonstrated either focal or diffuse expression for RCC (specificity 79%). These included: adrenocortical neoplasms 37/170 (22%), colonic 11/29 (37.5%), breast 9/27 (33%), prostate 5/18 (27.7%), ovary 2/17 (11.7%), melanoma 3/18 (16.6%), lung 3/21 (14.2%), and parathyroid 3/3 (100%). RCC expression was seen equally among adrenal adenoma and carcinoma group. Eight out of 28 (28.5%) normal adrenal cores also stained for RCC. RCC is a relatively useful marker in the differential diagnosis of REN only if used in a panel with other positive and negative markers. RCC does not reliably differentiate REN, especially classic clear cell type, from adreno- cortical neoplasms, which are frequently confused due to close anatomic proximity and similar morphology. RCC also does not reliably differentiate subtypes of renal epithelial neoplasms.
Key Words: renal cell carcinoma antigen, renal epithelial neoplasm
From the *Department of Pathology, Oklahoma University Health Sciences Center, Oklahoma City, Oklahoma; Departments of #Pathology; and Urology, University of Michigan Medical Center, Ann Arbor, Michigan.
Reprints: Nasir Bakshi, MD, Department of Pathology, Oklahoma University Health Sciences Center, Oklahoma City, Oklahoma (e-mail: nasir-bakshi@ouhsc.edu).
Copyright @ 2007 by Lippincott Williams & Wilkins
(Appl Immunohistochem Mol Morphol 2007;15:310-315)
T he ability to distinguish renal cell carcinoma (RCC) from other malignancies by immunohistochemistry has been limited by the lack of a reliable positive marker for renal differentiation. EMA, CK7, and CD10 anti- bodies are traditionally used positive markers for RCC.1 The sensitivity and specificity of these markers is low as other types of carcinomas also stain with these markers. RCC antigen is a 200-kd glycoprotein expressed in normal human proximal brush border antigen.2 Antibody to RCC has recently become commercially available and has been reported to be a highly specific and relatively sensitive marker for RCC.3 Although its expression has also been reported in other body tissues like normal breast lobules and ducts, parenchymal cells of parathy- roid and thyroid follicles, only occasional nonrenal tumors have been reported to be positive for RCC.4 Furthermore to date very few studies have tried to assess RCC as an aid in morphologic subtyping of renal epithelial neoplasms (RENs).5 Using TMA, we investi- gated the use of RCC in different histologic subtypes of REN while at the same time comparing its expression in nonrenal tumors, especially those potentially confused with REN because of location or histologic resemblance.
MATERIALS AND METHODS
Patients
All the cases included in this study were obtained from the archives of the department of Pathology, University of Michigan Medical Center, Ann Arbor, Michigan. None of the patients had been treated with chemotherapy before surgery. Institutional review board approval was obtained to procure and analyze tissues used in the study.
TMAs
The TMAs were constructed from archival forma- lin-fixed and paraffin-embedded tissue blocks using a tissue arrayer (Beecher Instruments, Silver Spring, MD) according to a previously described protocol. Tissue cores, 0.6-mm in size taken from a representative area of each tumor were used to assemble the arrays. Three TMAs were constructed: renal array, containing 241
REN samples, multitumor array containing 192 samples of a wide variety of neoplasms and an adrenal array comprising of 170 adrenal tumor and 28 normal adrenal samples. Based on several previous studies indicating that 2 or more 0.6-mm cores on TMAs give a better representation than single cores, 3 cores per case were used.6,7 Quality of histologic sections was assessed on hematoxylin and eosin stained tissue array section before immunohistochemical stains. Cores containing normal renal cortex were included to serve as positive internal control in all 3 TMA.
Immunohistochemistry
Four-micrometer sections were cut from the TMAs and transferred to silanized glass slides. Immunostaining was performed by standard avidin-biotin peroxidase complex technique. Sections were stained with a mouse monoclonal antibody (clone 66.4.C2, dilution of 1:50 with 30 min incubation) specific for RCC (Vector Labora- tories, Burlingame, CA). Immunohistochemistry was performed using an automated immunostaining system (Dako Corporation, Carpentaria, CA). Any staining in tumor cells either cytoplasmic or cell membrane or both was regarded as positive.
Scoring and Statistical Analysis
TMA with satisfactory staining for RCC were scanned by automated microscope (Bacus Laboratories, Lombard, IL) and transferred into a web-based platform (Profiler) for evaluation by 2 of the 3 pathologists in the study (N.B. and R.B.S.). Sections were scored using a semiquantitative scale for each individual tumor tissue cylinder on the array slide. RCC expression was scored both for staining intensity and percentage expression as follows:
Staining Intensity
0 = negative, 1+ = weak, 2+ = moderate,
3+ = strong.
Percentage Expression
Negative: no tumor cells showing immunoreactivity, Focal: only < 25% tumor cells showing immuno- reactivity,
Diffuse: ≥25% tumor cells showing immunoreac- tivity.
Scores were entered into a Microsoft Excel spread- sheet. Noninterpretable results due to excessive back- ground staining, loss of representative tissue or lack of enough tumor tissue were not used for statistical calculations. Data extracted form the database with the appropriate information was matched with the staining results. The final data was recorded in a Microsoft Excel datasheet and used for statistical analysis.
| Histologic Subtype | Positive | Percentage |
|---|---|---|
| Classic clear cell | 46/64 | 72 |
| Papillary renal cell | 40/42 | 95 |
| Unclassified | 33/38 | 85 |
| Sarcomatoid | 6/30 | 20 |
| Oncocytic | 18/24 | 75 |
| Chromophobe | 20/22 | 91 |
| Metatastatic | 10/23 | 40 |
| Total (all types) | 173/241 | 72 |
RESULTS
All cases were analyzed for results and each tumor type had at least one interpretable core.
Immunohistochemical Results in Normal Renal Parenchyma
Strong (3+) membranous staining with RCC was seen in the epithelia of all proximal renal tubules. Glomeruli, endothelial cells, and extraglomerular capil- laries showed no staining.
Immunohistochemical Results in Renal Neoplasms
Regardless of subtype, 173 of a total of 241 REN were RCC positive (sensitivity 72%). The break-up of various subtypes was: classic clear cell 72%, papillary 95%, chromophobe 91%, oncocytoma 75%, sarcomatoid 20%, unclassified 85%, and metastatic RCC 40%. These results are presented in Table 1. The overall immuno- staining intensity was consistently much higher in papillary and classic clear cell RCC demonstrating moderate-to-strong (2+ to 3+) diffuse membranous and cytoplasmic staining than in other tumors with oncocytic histology and chromophobe RCCs, which showed weaker staining as shown in Graph 1. RCC staining patterns in various subtypes of RCC are depicted in Figure 1. Percentage expression in various subtypes of RCC is illustrated in Graph 2.
90
80
70
60
Papillary renal cell
Classic Clear cell
50
Chromophobe
40
Unclassified
Oncocytic
30
Sarcomatoid
20
10
0
0
1+
2+
3+
A
B
C
D
E
F
G
H
I
J
K
L
Immunohistochemical Results in Adrenal Neoplasms
The adrenal array included 28 normal adrenal cores of which 8 (28.5%) stained for RCC. Regardless of the type, 37/170 (22%) adrenocortical neoplasms stained for RCC. Immunohistochemical expression for RCC was seen almost equally among adrenal adenoma (21%) and carcinoma (22.5%) groups (Table 2).
Immunohistochemical Results in Nonrenal Neoplasms
Regardless of type, out of 192 primary and metastatic tumors of nonrenal origin 39 were positive for RCC. These include: parathyroid 3/3 (100%), colonic 11/29 (37.5%), breast 9/27 (33%), prostate 5/18 (27.7%), cervix 3/122 (25%), melanoma 3/18 (17%), lung 3/21
(14%), and ovary 2/17 (11.7%). Tumors that did not stain include: urothelial carcinoma, pancreatic carcinoma, testicular tumors, hepatocellular carcinoma, lymphoid neoplasms, and endometrial carcinoma. These results are presented in Table 3. RCC staining patterns in various nonrenal primary epithelial neoplasms including adrenal neoplasms are depicted in Figure 2.
DISCUSSION
RENs comprise several different histopathologic types of tumors. In most cases it is possible to distinguish different types of renal neoplasms on the basis of conventionally stained tissue sections, alone. However, overlapping morphologic characteristics can sometimes pose difficulties in discriminating tumors like some chromophobe RCCs, oncocytomas, and clear cell RCC
Sarcomatoid
Oncocytic
Unclassified
☐ Negative Focal (<25%) ☒
☐
Diffuse (>25%)
Chromophobe
Classic Clear cell
Papillary renal cell
0%
20%
40%
60%
80%
100%
with oncocytic change. Because different types of renal neoplasms have differing prognosis, many immunohisto- chemical markers have been tested to support a correct morphologic decision.8-10 RCC, an antibody directed against the brush border of proximal tubules along with the recently studied markers like carbonic anhydrase IX11 are being tried as possible useful markers in this setting. We, in our institution have been using RCC for period of time. Due the sparse number of well-documented studies comparing RCC in various tumors we embarked on the present study.3,5 Using the versatile technology of TMA we investigated the use of RCC in different histologic subtypes of RENs and its expression in a variety of nonrenal neoplasms, normal adrenal gland, and adrenal neoplasms.
Our results support RCC as a sensitive marker of renal differentiation. Seventy-two percent of REN showed staining at any level irrespective of the tumor type, whether primary or metastatic. When analyzing primary REN alone the sensitivity of RCC was slightly higher at 75%. We also observed that poorly differen- tiated REN are more likely to be negative for RCC than are better-differentiated cases. Eighty percent of sarco- matoid REN were either completely negative or showed only focal staining. In contrast 85% of unclassified REN were positive for RCC, nearly half of which showed less
| Normal/Tumor Type | Positive | Percentage |
|---|---|---|
| Adrenocortical carcinoma | 16/72 | 22.5 |
| Adrenocorical adenoma | 12/57 | 21 |
| Adrenal hyperplasia | 7/27 | 26 |
| Pheochromocytoma | 3/14 | 14 |
| Total (all types) | 37/170 | 22 |
| Tumor Type | PositivePercentage | |
|---|---|---|
| Colonic adenocarcinoma | 11/29 | 37.5 |
| Breast-ductal and lobular carcinoma | 9/27 | 33 |
| Prostatic adenocarcinoma | 5/18 | 28 |
| Cervix-squamous and endocervical adenocarcinoma | 3/12 | 25 |
| Malignant melanoma | 3/18 | 17 |
| Lung-squamous and adenocarcinoma | 3/21 | 14 |
| Ovary-serous, mucinous, clear cell | 2/17 | 12 |
| Parathyroid adenoma | 3/3 | 100 |
| Lymph node-malignant B and T cell lymphomas | 0/31 | 0 |
| Urinary bladder-urothelial carcinoma | 0/6 | 0 |
| Testes-seminoma and embryonal carcinoma | 0/3 | 0 |
| Pancreatic adenocarcinoma | 0/3 | 0 |
| Liver-hepatocellular carcinoma | 0/3 | 0 |
| Endometrial carcinoma | 0/3 | 0 |
| Total (all types) | 39/192 | 20 |
than 10% staining (Graph 2). These findings lend some support to the notion that poorly differentiated REN tend to lose reactivity for most of the antigenic markers and RCC is no exception. Our experience also indicates that heterogeneous and sometimes focal nature of the staining in some clear and unclassified REN may lead to false- negative results, particularly if attempted on small biopsy samples. We therefore do not recommend using RCC in discriminating the histologic subtypes of REN when the sample size is small.
The immunohistochemical results in nonrenal neo- plasms show that 39 out of 192 (20%) tumor samples were positive for RCC. In contrast to our study, McGregor et al5 described RCC staining in 15/146 (11%) total nonrenal neoplasms performed on routine histologic sections. The results of our study emphasize the relatively frequent RCC staining of conventional breast carcinomas and colonic cancer. This is a significant limitation as metastatic breast carcinoma is frequently in the differential diagnosis of a bony lesion subject to biopsy. We also occasionally found RCC to stain several other tumors for which staining have not been previously reported. These include carcinomas of lung, colon, endocervix, prostate, and melanoma. Certain of these may be considered significant, even though they may not pose significant diagnostic problems because of their overall higher frequency. Due to histologic similarity of nonrenal papillary neoplasms originating from other sites such as lung, RCC could potentially be used as a helpful and reliable marker because a majority (95%) of papillary RCCs consistently show a strong expression for RCC.
Although previously McGregor et al5 reported that none of the 21 hepatocellular carcinoma stained for RCC and only 2 out of 16 testicular tumors expressed the antigen, ideally additional larger sample size than analyzed in this study would be necessary to assess the RCC expression in these and other tumor types like urothelial, pancreatic and endometrial tumors
We also evaluated RCC staining in adrenal neo- plasms, both carcinoma and adenoma. To our knowledge
A
B
C
D
E
F
G
H
1
J
K
L
M
N
0
P
our study is the only of its kind evaluating a large number of adrenal tumor samples for a renal-specific marker (RCC). Our results indicate that in general the number of adrenal neoplasms staining for RCC is much higher than expected for a marker that is relatively renal specific (Table 2). This fact should be borne in mind when using this antibody in differentiating morphologically difficult cases of adrenal carcinoma. A panel of antibodies including RCC, cytokeratin 7, CD10, EMA, and inhibin might be more appropriate than using any single one of them.12
In summary, this study supports RCC as a useful marker for RCC. It is not, however, entirely specific, showing relative frequent staining in colonic carcinomas, breast carcinoma, and malignant melanoma. Occasion- ally, several other tumors also stain for RCC, including ovarian carcinoma and adenocarcinoma of lung. The use of RCC in conjunction with other positive and negative markers of REN like epithelial membrane antigen, CD10, cytokeratin, and vimentin is recom- mended in the differential diagnosis of REN. RCC does
not reliably differentiate REN, especially classic clear type from adrenocortical neoplasms with which they could be frequently confused due to close anatomic proximity.
REFERENCES
1. Avery AK, Beckstead J, Renshaw AA, et al. Use of antibodies to RCC and CD10 in the differential diagnosis of renal neoplasms. Am J Surg Pathol. 2000;24:203-210.
2. Pan CC, Chen PC, Ho DM. The diagnostic utility of MOC31, BerEP4, RCC marker and CD10 in the classification of renal cell carcinoma and renal oncocytoma: an immunohistochemical analysis of 328 cases. Histopathology. 2004;45:452-459.
3. Gokden N, Mukunyadzi P, James JD, et al. Diagnostic utility of renal cell carcinoma marker in cytopathology. Appl Immunohisto- chem Mol Morphol. 2003;11:116-119.
4. Koga F, Kawano K, Honda M, et al. Sarcomatoid renal cell carcinoma with scant carcinomatous components. Int J Urol. 2000;7:58-60; discussion 61.
5. McGregor DK, Khurana KK, Cao C, et al. Diagnosing primary and metastatic renal cell carcinoma: the use of the monoclonal antibody ‘Renal Cell Carcinoma Marker’. Am J Surg Pathol. 2001;25:1485-1492.
6. Hoos A, Urist MJ, Stojadinovic A, et al. Validation of tissue microarrays for immunohistochemical profiling of cancer specimens using the example of human fibroblastic tumors. Am J Pathol. 2001;158:1245-1251.
7. Rimm DL, Camp RL, Charette LA, et al. Tissue microarray: a new technology for amplification of tissue resources. Cancer J. 2001;7:24-31.
8. Javidan J, Stricker HJ, Tamboli P, et al. Prognostic significance of the 1997 TNM classification of renal cell carcinoma. J Urol. 1999;162:1277-1281.
9. de Peralta-Venturina M, Moch H, Amin M, et al. Sarcomatoid differentiation in renal cell carcinoma: a study of 101 cases. Am J Surg Pathol. 2001;25:275-284.
10. Amin MB, Tamboli P, Javidan J, et al. Prognostic impact of histologic subtyping of adult renal epithelial neoplasms: an experience of 405 cases. Am J Surg Pathol. 2002;26:281-291.
11. Bui MH, Seligson D, Han KR, et al. Carbonic anhydrase IX is an independent predictor of survival in advanced renal clear cell carcinoma: implications for prognosis and therapy. Clin Cancer Res. 2003;9:802-811.
12. Fetsch PA, Powers CN, Zakowski MF, et al. Anti-alpha-inhibin: marker of choice for the consistent distinction between adrenocor- tical carcinoma and renal cell carcinoma in fine-needle aspiration. Cancer. 1999;87:168-172.