Original Articles. PAX-2 Is a Helpful Marker for Diagnosing Metastatic Renal Cell Carcinoma

Original Articles PAX-2 Is a Helpful Marker for Diagnosing Metastatic Renal Cell Carcinoma Comparison With the Renal Cell Carcinoma Marker Antigen and Kidney-Specific Cadherin Ayhan Ozcan, MD; Qihui Zhai, MD; Rehana Javed, MD; Steven S. Shen, MD, PhD; Donna Coffey, MD; Bhuvaneswari Krishnan, MD; Luan D. Truong, MD N Context. The diagnosis of metastatic renal cell carcinoma (RCC) remains problematic. Objective. To evaluate the role of PAX-2, a renal tubular cell transcription factor, in the diagnosis of metastatic RCC. PAX-2 expression in metastatic RCC was compared with that of the renal cell carcinoma marker antigen (RCCM) and kidney-specific cadherin (KSC), which are 2 known markers for RCC. Design. Immunostaining for PAX-2, RCCM, and KSC was performed on consecutive tissue sections of 95 metastatic RCCs (77 clear cell, 8 papillary, 5 sarcomatoid, and 5 collecting duct) and 183 metastatic tumors other than RCC. Results. For PAX-2, positive immunoreactivity was detected in 77% clear cell, 75% papillary, 100% collecting duct, and 0% sarcomatoid metastatic RCCs. For RCCM, positive immunoreactivity was detected in 49% clear cell, 75% papillary, 0% collecting duct, and 0% sarcomatoid metastatic RCCs. For KSC, only 2 metastatic clear cell RCCs (3%) were positive. In combination, all markers were positive in 0% of cases; all markers were negative in 23% of cases (17 clear cell, 1 papillary, and for all 5 sarcomatoid); and at least 1 marker was positive in 76% of cases (PAX-2 only in 28% of cases [21 clear cell, 1 papillary, and 5 collecting duct] and RCCM only in 3% of cases [2 clear, 1 papillary]). Of 183 metastatic tumors other than RCC, 14 were positive for PAX-2 (nodal metastasis of carcinoma of colon [1], breast [1], endometrium [1], and ovary [1]; and omental metastasis of carcinoma of uterus or ovary [10]). Conclusions. PAX-2 is a sensitive and specific marker for metastatic RCC. The diagnostic yield would be marginally increased by adding RCCM, but not KSC, as an immunomarker. (Arch Pathol Lab Med. 2010;134:1121 1129) The diagnosis of metastatic renal cell carcinoma (RCC) remains problematic for several reasons. Renal cell carcinoma may present initially as a metastasis without clinical features referable to a primary renal mass. 1 Even when the diagnosis of metastatic RCC is suggested from biopsy examinations and the presence of a renal mass is subsequently confirmed by imaging studies, tissue from the renal mass may not be available for comparison. Renal cell carcinoma is well known as a tumor that can give rise to latent metastasis, even isolated metastases, decades after the initial diagnosis. 2 Metastatic RCC may be less Accepted for publication December 11, 2009. From the Department of Pathology, The Methodist Hospital, Houston, Texas (Drs Ozcan, Zhai, Javed, Shen, Coffey, and Truong); the Department of Pathology, Baylor College of Medicine, Houston, Texas (Drs Krishnan and Truong); the Department of Pathology, Weill Medical College of Cornell University, New York, New York (Drs Ozcan, Zhai, Shen, Coffey, and Truong); the Department of Pathology, Gülhane Military Medical Academy and School of Medicine, Ankara, Turkey (Dr Ozcan); and the Department of Pathology, The Methodist Research Institute, Houston, Texas (Drs Zhai, Shen, Coffey, and Truong). The authors have no relevant financial interest in the products or companies described in this article. Reprints: Luan D. Truong, MD, Department of Pathology, MS 205, The Methodist Hospital, 6565 Fannin St, Houston, TX 77030 (e-mail: ltruong@tmhs.org). differentiated than and thus looks different from its primary tumor, and the metastasis may not conform to the known morphologic spectrum of primary RCC. Metastatic RCC may develop in patients with a history of both RCC and a second primary tumor. Patients with RCC have a higher incidence (27.4%) than the general population for developing a second primary tumor. 3 Thus, when a new lesion develops in these patients, the diagnostic possibilities would include a new primary tumor, its metastasis, or metastatic RCC. This diagnostic problem is accentuated because, against these clinical contexts, needle core biopsy or fine-needle aspiration is often performed, resulting in limited available tissue. Several immunohistochemical markers are known to help in the diagnosis of renal neoplasms, among which are the renal cell carcinoma marker antigen (RCCM), kidneyspecific cadherin (KSC), and PAX-2. 4 12 Renal cell carcinoma marker antigen, a terminally differentiated molecule localized to the brush border of normal proximal tubular cells, is a rather specific marker for primary renal neoplasm, especially for neoplasms derived from proximal tubule, that is, clear cell or papillary RCC. 6,13 However, both the percentage of tumors that are stained (the staining frequency) and the percentage of stained tumor cells in each case (the staining extent) Arch Pathol Lab Med Vol 134, August 2010 PAX-2 and Metastatic Renal Cell Carcinoma Ozcan et al 1121

Table 1. The Sites of Metastases From Renal Cell Carcinomas or From Other Tumors Renal Cell Carcinoma, No. of Cases Other Tumors, No. of Cases a Lung 33 13 Lymph node 9 84 Bone 9 6 Brain 8 8 Soft tissue 7 3 Pancreas 7 0 Adrenal gland 7 2 Liver 5 21 Bowel 4 1 Pleura 3 7 Skin 2 1 Spleen 1 0 Seminal vesicle 0 8 Omentum 0 21 Testis 0 1 Bladder 0 1 Kidney NA 1 Total 95 183 Positive cases 70 (74%) 14 (7.6%) b Abbreviation: NA, not applicable. a These include adenocarcinoma of breast, endometrium, lung, prostate, pancreas, gastrointestinal tract; squamous cell carcinoma of lung; neuroendocrine carcinoma including small cell carcinoma of lung; urothelial carcinoma; thyroid carcinoma; hepatocellular carcinoma; and malignant melanoma. b These positive cases include nodal metastasis of carcinoma of colon (1), breast (1), endometrium (1), and ovary (1); and omental metastasis of carcinoma of uterus or ovary (10). are markedly decreased for metastatic RCC, thus limiting the diagnostic utility of RCCM in the metastatic context. 13 Kidney-specific cadherin, found in the basolateral infolding membrane of normal distal tubular and collecting duct cells, is known to mark their neoplastic derivatives, that is, chromophobe RCC or oncocytoma, in most cases, but is usually not expressed by other types of RCCs. 7,8,14 Furthermore, KSC expression in metastatic RCC has not been studied. PAX-2 is a member of the paired box family of transcription factors, which is required for development and proliferation of the kidney, brain, and müllerian organs. 15 27 Several studies have established PAX-2 as a sensitive marker for RCC, regardless of histologic subtypes and differentiation levels. 9 12,25,28 However, the diagnostic utility of PAX-2 in the context of metastatic RCC has not been systematically evaluated. In this study, we evaluate the diagnostic sensitivity and specificity of PAX-2 in metastatic tumors from both RCC and other tumor types, and we compare the expression of PAX-2 in metastatic RCCs with that of RCCM or KSC. MATERIALS AND METHODS This retrospective study included 95 metastatic RCCs to various organs (Table 1). The diagnosis of metastatic RCC was confirmed by characteristic morphology, immunoprofiles, renal masses documented by imaging, or comparison with the primary RCCs, which were available in several but not all cases. Although the morphologic spectrum of metastatic RCC may not completely conform to that of primary RCC, the metastases were classified according to the 2004 World Health Organization classification scheme for primary RCC 29 (Tables 2 and 3). Also included were 183 metastases from tumors other than RCC to various organs (Table 1). The metastatic tumors included adenocarcinoma of breast, endometrium, lung, prostate, pancreas, gastrointestinal Table 2. Expression of PAX-2, RCCM, and KSC in Metastatic Renal Cell Carcinomas PAX-2, No. (%) RCCM, No. (%) KSC, No. (%) Clear cell (n 5 77) 59 (77) 38 (49) 2 (3) Papillary (n 5 8) 6 (75) 6 (75) 0 (0) Sarcomatoid (n 5 5) 0 (0) 0 (0) 0 (0) Collecting duct (n 5 5) 5 (100) 0 (0) 0 (0) Total (n = 95) 70 (74) 44 (46) 2 (2) Abbreviations: KSC, kidney-specific cadherin; RCCM, renal cell carcinoma marker antigen. tract; squamous cell carcinoma of lung; neuroendocrine carcinoma including small cell carcinoma of lung; thyroid carcinoma; hepatocellular carcinoma; and malignant melanoma. Since previous studies show that a percentage of müllerian-type tumors (ovary, fallopian tube, and endometrium) also express PAX-2, preference was given to their corresponding metastasis. To compare the expression of PAX-2, RCCM, and KSC in metastatic RCCs, consecutive tissue sections from each metastatic RCC were submitted for immunostaining with these 3 markers. To evaluate the diagnostic sensitivity and specificity of PAX-2 in the context of metastatic disease, sections from metastatic tumors other than RCC were submitted for immunostaining with PAX-2. These tissue sections were subjected to deparaffinization, hydration, and endogenous peroxidase blocking. Antigen retrieval was done in a similar fashion for each of these 3 target antigens: tissue sections were subjected to Dako Target Retrieval Solution (ph 6; Dako, Carpinteria, California) in a pressure cooker set at 95uC for 22 minutes, followed by gradual cooling for 20 minutes. The tissue sections were incubated for 30 minutes at room temperature with the RCCM antibody (1:10; Vector Laboratories, Burlingame, California), a monoclonal antibody against KSC (1:75; Invitrogen, Carlsbad, California), or an anti PAX-2 polyclonal antibody (1:75; Invitrogen). Detection of the staining reaction was achieved by an enzymeconjugated polymer complex adapted for automatic stainers from Ventana (Ventana UltraView; Ventana Benchmark XT; Ventana Medical Systems; Tucson, Arizona). For each case, the staining intensity was graded from 0 to 3+ (0 5 no stain; 1+ 5 unequivocal but weak; 2+ 5 moderate; and 3+ 5 strong), and the estimated percentage of stained tumor cells for each of the 3 markers was also evaluated. The staining pattern (nuclear, cytoplasmic, or membranous) was also noted. The expression of PAX-2, RCCM, and KSC was correlated with histologic types of metastatic RCC. RESULTS Light Microscopic Features of Metastatic RCC The sites of metastasis are listed in Table 1. The histologic types included clear cell (77 cases), papillary (8), sarcomatoid (5), and collecting duct RCCs (5). Table 3. PAX-2, RCCM, and KSC Profiles of Metastatic Renal Cell Carcinoma According to Histologic Subtypes a PAX-2 + 2 + + + 2 2 2 RCCM + 2 2 + 2 + 2 + KSC + 2 2 2 + + + 2 Clear cell (n 5 77) 0 17 21 35 2 0 0 2 Papillary (n 5 8) 0 1 1 5 0 0 0 1 Sarcomatoid (n 5 5) 0 5 0 0 0 0 0 0 Collecting duct (n 5 5) 0 0 5 0 0 0 0 0 Abbreviations: KSC, kidney-specific cadherin; RCCM, renal cell carcinoma marker antigen; +, positive; 2, negative. a The number indicates the number of cases with a specific immunoprofile. 1122 Arch Pathol Lab Med Vol 134, August 2010 PAX-2 and Metastatic Renal Cell Carcinoma Ozcan et al

Hematoxylin-eosin tissue sections or pathology reports of the primary RCCs were available for review in 34 cases. In these cases, although the histologic types of the primary and metastatic tumors were similar, the nuclear atypia seemed to be more pronounced in the metastases. Within the clear cell RCCs, poorly differentiated features including predominance of granular cells, prominent nucleoli, rhabdoid changes, myxoid stroma, focal and faint spindle cell morphology, or marked necrosis were noted in 21 cases (27%). The sarcomatoid, collecting duct, and papillary RCC metastases showed a similar morphologic spectrum as that described for their primary counterparts. PAX-2 in Metastatic RCC PAX-2 was successfully detected against a background of appropriate positive and negative controls. The staining was nuclear, without cytoplasmic staining, or with faint focal cytoplasmic staining in few cases. PAX-2 nuclear staining was detected in 70 of 95 tumors (74%), with a high frequency for clear cell (77%), papillary (75%), and collecting duct (100%) RCCs, but was negative in all 5 sarcomatoid RCCs (Table 2). Staining was noted in a few to virtually all tumor cells (mean, 60.8%; with the staining of more than 50% of tumor cells in 42 tumors), with an intensity score of 1+, 2+, and 3+ in 18, 25, and 27 cases, respectively. Within the clear cell RCC group, PAX-2 staining was equal in both intensity and extent for the welldifferentiated tumors as well as the poorly differentiated tumors (Figure 1, A, B, D, E, G, H, J, and K; and Figure 2, A, B, D, E, G, H, J, and K). For the papillary tumors, 35% to 95% of tumor cells were positive (mean, 64%), with at least 2+ intensity (Figure 3, A and B). For collecting duct RCC, 10% to 100% of tumor cells were stained (mean, 45%), with an intensity score of 1+, 2+, and 3+ in 2, 2, and 1 case, respectively (Figure 3, D and E). There was no staining for sarcomatoid RCC (Figure 3, G and H). RCCM in Metastatic RCC Renal cell carcinoma marker antigen was successfully detected against a background of appropriate positive and negative controls. The staining was noted predominantly in cell membrane but cytoplasmic staining was also seen in some cases. There was no nuclear staining. Positive staining for RCCM was detected in 44 of 95 tumors (46%). It was seen in 49% of clear cell and 75% of papillary RCCs, but was not present in sarcomatoid or collecting duct RCCs. Staining was noted in a few to virtually all tumor cells (mean, 46.5%; with the staining of more than 50% of tumor cells in 16 tumors), with an intensity score of 1+, 2+, and 3+ in 2, 8, and 34 cases, respectively. Within the clear cell RCC group, in contrast to PAX-2, RCCM staining was more frequent in the well-differentiated tumors than in the poorly differentiated tumors (Figure 1, A, C, D, F, G, I, J, and L; and Figure 2, A, C, D, F, G, I, J, and L). Staining with RCCM was shown in papillary carcinoma (Figure 3, A and C), but not in collecting duct RCC (Figure 3, D and F) or sarcomatoid RCC (Figure 3, G and I). KSC in Metastatic RCC Kidney-specific cadherin was successfully detected against a background of appropriate positive and negative controls. The staining was noted predominantly in cell membrane but cytoplasmic staining was also seen. There was no nuclear staining. Kidney-specific cadherin was noted in 2 of 95 tumors, both of which were clear cell RCCs. Comparison of PAX-2, RCCM, and KSC Staining in Metastatic RCC The frequencies of tumor types, stratified against various combined staining patterns, are summarized in Table 3: all 3 markers positive (0%; 0 of 95); all 3 markers negative ([24%; 23 of 95]; 17 clear cell, 1 papillary, and 5 sarcomatoid); at least 1 marker positive (76%; 72 of 95); PAX-2 positive only ([28%; 27 of 95]; 21 clear cell, 1 papillary, and 5 collecting duct); RCCM positive only ([3%; 3 of 95]; 2 clear and 1 papillary); and KSC positive only (0%; 0 of 95). Staining on consecutive tissue section of the same tumor allowed for direct, case-by-case comparison of the staining of the 3 markers. Among the 77 clear cell RCCs, 21 were positive for only PAX-2, 2 were positive for only RCCM, and none were positive for only KSC. Most of the PAX-2 positive tumors in this histologic group displayed poorly differentiated features described above (Figure 1, D, E, J, and K; and Figure 2, D, E, J, and K). Among the 8 papillary RCCs, 1 each showed staining for only PAX-2 or RCCM, and none showed staining for KSC. PAX-2, but not RCC or KSC, was expressed by each of the 5 collecting duct tumors (Figure 3, D through F). None of the markers were noted in 5 sarcomatoid RCCs (Figure 3, G through I). PAX-2 Staining of Metastases From Tumors Other Than RCC The metastatic sites and the types of primary tumors are listed in Table 1. The tumor types included adenocarcinoma of breast, endometrium, lung, prostate, pancreas, gastrointestinal tract; squamous cell carcinoma of lung; neuroendocrine carcinoma including small cell carcinoma of lung; urothelial carcinoma; thyroid carcinoma; hepatocellular carcinoma; and malignant melanoma. Fourteen of 183 metastatic tumors expressed PAX-2, all of which were in lymph nodes (4 of 84 cases) or omentum (10 of 21 cases). The 4 PAX-2 positive nodal metastases included adenocarcinoma of colon (1) and breast (1), with the primary tumors being PAX-2 negative (Figure 4, A and B); and carcinoma of endometrium (1) and ovary (1), with strong PAX-2 expression in the primary tumors. The 21 omental metastases included 16 tumors from uterus or ovary, 10 of which expressed PAX-2 (Figure 4, C and D), whereas the 5 metastases from other organs were negative for this marker. COMMENT Data from our study demonstrate that PAX-2 is a sensitive and specific marker for metastatic RCCs. The diagnostic yield would be marginally increased by adding RCCM, but not KSC, as an immunomarker. PAX-2 is a member of the paired box gene family, which consists of 9 members, PAX-1 through PAX-9, each encoding a transcription factor. These transcription factors are expressed in an orderly fashion during fetal development and are implicated in proper organogenesis. PAX-2 is known to control the development of the central nervous system, the kidney, 18,22,23,26 and the müllerian organs. 30 32 During nephrogenesis, PAX-2 appears very Arch Pathol Lab Med Vol 134, August 2010 PAX-2 and Metastatic Renal Cell Carcinoma Ozcan et al 1123

Figure 1. Metastatic clear cell/granular renal cell carcinoma. A, B, and C, Lung, clear cell type, Fuhrman nuclear grade 1, diffuse staining for both PAX2 and renal cell carcinoma marker antigen (RCCM) (original magnifications 3100). D, E, and F, Adrenal gland, clear cell type, Fuhrman nuclear grade 3, no PAX-2 staining and strong RCCM staining (original magnifications 3200). G, H, and I, Lung, adjacent high and low nuclear-grade areas, strong PAX-2 staining in high-grade area and weak or no staining in low-grade area, negative for RCCM in both areas (original magnifications 3200). J, K, and L, Lung, granular cell type, nuclear grade 3, strong PAX-2 staining, no RCCM staining (original magnifications 3100). Abbreviation: H&E, hematoxylin-eosin. early in the renal blastema and promotes mesenchymal cell proliferation and apoptosis, and mesenchymal-epithelial transformation, with formation of immature renal tubules and glomeruli. However, maturation of these renal tubules depends partly on disappearance of PAX-2; thus, in normal adult kidney, PAX-2 is seen only focally in parietal 1124 Arch Pathol Lab Med Vol 134, August 2010 epithelial cells and collecting ducts. In both humans and experimental animals, PAX-2 transgene is associated with glomerulosclerosis and renal cystic changes, whereas PAX2 deletion induces renal tubular atrophy.10,15 27 Several previous studies including ours have shown that PAX-2 is a sensitive marker for primary renal neoplasm, PAX-2 and Metastatic Renal Cell Carcinoma Ozcan et al

Figure 2. Metastatic clear cell/granular renal cell carcinoma. A, B, and C, Lung, clear cell type, isolated tumor cells in alveolar spaces, not readily distinguishable from reactive cells such as macrophages. Some of these cells are strongly positive for PAX-2, but are negative for renal cell carcinoma marker antigen (RCCM) (original magnifications 3200). D, E, and F, Soft tissue, spindled cell with clear cytoplasm and low nuclear grade, showing unequivocal staining for PAX-2, but not RCCM (original magnifications 3200). G, H, and I, Bone, tumor cells with clear/granular cytoplasm and high nuclear grade against an abundant myxoid background. They strongly express PAX-2, but not RCCM (original magnifications 3200). J, K, and L, Lung, tumor cells with granular cytoplasm, high nuclear grade, necrosis, and rhabdoid features, which are diffusely positive for PAX-2 but not RCCM (original magnifications 3100). Abbreviation: H&E, hematoxylin-eosin. with an overall detection rate of 85%. This rate is however variable among histologic subtypes (92% for clear cell, 87% for papillary, 83% for chromophobe, 0% for sarcomatoid, 100% for collecting duct, and 88% for oncocytoma).10 12,28 Little is known on PAX-2 expression by metastatic RCCs. Arch Pathol Lab Med Vol 134, August 2010 PAX-2 expression was noted in 23 of 27 metastatic RCCs (85%), as reported by Gokden et al,33 and in 23 of 29 metastatic RCCs (79%) obtained by fine-needle aspiration in a study by Wasco and Pu.34 Both studies included a relatively small number of cases and did not address the PAX-2 and Metastatic Renal Cell Carcinoma Ozcan et al 1125

Figure 3. Metastatic papillary, collecting duct, and sarcomatoid renal cell carcinoma. A, B, and C, Lymph node, papillary, low nuclear grade. Diffuse staining for both PAX-2 and renal cell carcinoma marker antigen (RCCM) (original magnifications 3100). D, E, and F, Liver, collecting duct type characterized by tubular formation, high nuclear grade, and desmoplastic stroma. The tumor cells are diffusely positive for PAX-2, but negative for RCCM (original magnifications 3100). G, H, and I, Lung, sarcomatoid. The spindle tumor cells admixed with inflammatory cells are negative for both PAX-2 and RCCM (original magnifications 3200). Abbreviation: H&E, hematoxylin-eosin. histologic types of the metastatic tumors or the degree of tumor differentiation in relation to the staining patterns. The current study expands the results previously presented in abstract format,35 by using a large number of metastatic RCCs obtained either by surgical excision or fine-needle aspiration, and confirms PAX-2 as a sensitive marker for metastatic RCC, with an overall detection rate of 74%. However, since PAX-2 is a transcription factor, only nuclear staining is considered specific. The current study indeed shows that in most positive cases only nuclear staining is observed. However, in a few positive and negative cases, weak and focal cytoplasm staining was also noted. This observation has been previously reported.28 The reason for this aberrant PAX-2 expression is not clear. It probably is not related to a technical aspect, such as endogenous avidin-binding activity, since an enzyme-conjugated polymer complex technique was used, which bypasses any biotin-avidin related step. Regardless of its etiology, cytoplasmic staining of PAX-2 1126 Arch Pathol Lab Med Vol 134, August 2010 should be considered negative for diagnostic purposes. Generally, an immunomarker is less often expressed and, even in positive cases, shows fewer stained cells in metastases than in the corresponding primary tumors, thus limiting its diagnostic utility in the former.33,34 This limitation is perhaps not a problem for PAX-2, since the discrepancy in PAX-2 expression frequency in primary tumor versus metastatic RCC (85% versus 74%) under identical staining conditions was relatively minor. In PAX-2 positive metastatic RCCs, usually a large percentage (mean, 60.8%) of tumor cells were positive, a result similar to that for primary tumors (50%).13,28 These observations add significant diagnostic advantages to PAX-2 as a marker for metastatic RCC because, often, only limited tissue is available for diagnostic purposes in this context and because marked attenuation in staining, associated with metastasis in terms of both frequency and extent, is well known for RCCM,13 the only other popular kidney-specific marker. Within the category of PAX-2 and Metastatic Renal Cell Carcinoma Ozcan et al

Figure 4. A, Breast ductal carcinoma metastatic to a lymph node showing strong PAX-2 staining. B, The primary tumor is negative for PAX-2. C, Undifferentiated ovarian carcinoma metastatic to omentum. D, Tumor cells with strong PAX-2 staining (PAX-2, original magnifications 3200 [A and B] and 3100 [D]; hematoxylin-eosin, original magnification 3200 [C]). clear cell/granular metastatic RCC, we have noted that although morphology typical for RCC is observed in many cases, in up to 28% of these cases atypical features are noted, probably reflecting metastasis-associated dedifferentiation, including predominance of granular cells, prominent nucleoli, rhabdoid changes, myxoid stroma, focal and faint spindle cell morphology, and marked necrosis. It is within this group, for which the histologic diagnosis of metastatic RCC is not obvious, that PAX-2 is useful. Many such tumors are unequivocally positive for PAX-2, but negative for RCCM. Another diagnostic advantage of PAX-2 is that it stains all metastatic collecting duct carcinoma in the current study, as well as their primary tumors, which confirms a finding that was observed previously.28 Collecting duct carcinoma is a type of renal RCC that can closely simulate other types of highgrade carcinoma, such as urothelial carcinoma, highgrade papillary RCC, or metastatic carcinoma. No specific immunohistochemical marker is available currently to differentiate them.36,37 PAX-2, however, fails to detect metastatic sarcomatoid RCC, as well as its primary tumor.10 12,28 Although PAX-2 is expressed in up to 83% of primary chromophobe RCCs,28 its expression in metastatic chromophobe RCC remains unknown. Our Arch Pathol Lab Med Vol 134, August 2010 study, despite its relatively large number of cases, did not include metastatic chromophobe RCC. This type of RCC accounts for 5% to 7% of all RCCs, is often indolent clinically, and rarely metastasizes (4.2%).38 Indeed, it accounts for only 1.9% of 910 metastatic RCCs reported by Hoffmann et al.39 Renal cell carcinoma marker antigen is an excellent marker for primary RCC, with an overall detection rate of 80% to 85%,6,13,28 which is variable among histologic types (clear cell, 84% 92%; papillary, 93% 96%; chromophobe, 0% 45%; sarcomatoid, 25%) but negative for oncocytoma. Expression of RCCM is, however, much attenuated in metastatic RCC, in terms of both frequency and extent, with a detection rate between 40% and 70%, in studies by McGregor et al,13 Gokden et al,33 and Gokden et al.40 These observations are corroborated by the current study, in which only 46% of metastatic RCCs express RCCM, with fewer stained cells (mean, 46.5%; with more than 50% tumor cells stained in 17% of tumors) than in the primary tumors (with more than 50% tumor cells stained in 72% of tumors). This attenuation is probably due to the fact that RCCM is a terminally differentiated molecule, abundantly expressed in brush border of normal tubular epithelial cells, and that its expression may be inversely proporpax-2 and Metastatic Renal Cell Carcinoma Ozcan et al 1127

tional to tumor differentiation. 6,13 This is in keeping with the observation, at least in the category of clear cell RCC, that the metastatic tumors are usually poorly differentiated and that these tumors do not express RCCM. Thus, RCCM was not seen in any metastatic collecting duct carcinoma or sarcomatoid RCCs, and it was not seen in their primary tumors either. 6,13 These considerations detract from the diagnostic utility of RCCM in the metastatic context. A previous report 13 showed expression of RCCM by about 10% of nonrenal primary tumors, including those from breast and parathyroid, with rare expression in metastatic breast carcinoma. Kidney-specific cadherin is a sensitive marker for oncocytoma (38% to 95%) and chromophobe RCC (58% to 100%), 14,28 but is rarely expressed by other RCCs (0% to 14%), in keeping with the fact that KSC is normally limited to the distal convoluted tubules and collecting ducts, and oncocytoma and chromophobe RCC are thought to derive from these tubular segments. Kidney-specific cadherin expression by metastatic RCCs has not been previously studied. We found that KSC was expressed by a small number of tumor cells in only 2 of 95 metastatic RCCs (2%), both of which were of clear cell type. These observations, plus the rarity of metastatic chromophobe RCC, discourage the use of this marker in the context of metastatic RCC. To the best of our knowledge, a comprehensive study of KSC expression in metastatic RCC, or its expression in primary or metastatic tumors other than those from the kidney, is not available. Simultaneous staining for PAX-2, RCCM, and KSC affords insight into the possible complementary diagnostic utilities of these markers in diagnosing metastatic RCC. Our study shows that among these markers, PAX-2 is the most sensitive (with an overall detection rate of 74%, 46%, and 2% for PAX-2, RCCM, and KSC, respectively), and it helps detect a significant number of tumors in instances where the 2 other markers fail. Indeed, up to 28% of metastatic RCCs in this series are detected only by PAX-2 (21 clear cell, 1 papillary, and 5 collecting duct tumors). These observations indicate that PAX-2 not only helps recognize additional cases of metastatic clear cell and papillary RCCs that elude their traditional markers, but also, more importantly, detects all metastatic collecting duct carcinomas, which are uniformly negative for both RCCM and KSC. Although RCCM was expressed by 44 of 95 metastatic RCCs (46%), 40 also expressed PAX-2, such that 4 tumors expressed RCCM only. These observations, plus the fact that RCCM was never the only positive marker for the rest of the cases in this study, imply that there is very little additional diagnostic gain in using RCCM over that provided by PAX-2. Since only 2 tumors were positive for KSC, and each was also strongly positive for PAX-2, KSC may not be relevant in diagnosing metastatic RCC. In a previous study, 28 we have noted that up to 90% of primary RCCs are positive for at least 1 marker among PAX-2, RCCM, and KSC. However, only 74% of metastatic RCCs reach this level, highlighting the need for more sensitive RCC immunomarkers. Another potentially useful technique in the differential diagnosis of metastatic neoplasms, when a renal primary tumor is being considered, is electron microscopy. Furthermore, the combined use of immunostaining and electron microscopy may provide a significantly increased diagnostic sensitivity. Unfortunately, tissue was not submitted for electron microscopic evaluation in this retrospective study. Since the diagnosis of metastatic RCC is often made in the context of possible additional primary tumors, it needs to be determined whether primary or metastatic tumors other than RCCs can express PAX-2 (that is, the specificity of PAX-2 for metastatic RCCs). Aside from primary and metastatic RCCs, PAX-2 has been reported in all 39 nephrogenic adenomas tested 41 and most (67%) ovarian serous carcinomas. 32 In a study addressing the diagnostic specificity of PAX-2, in which primary tumors that could potentially be confused with RCCs were selected, Gokden et al 33 reported PAX-2 expression in 1 of 3 parathyroid carcinomas, 3 of 7 clear cell carcinomas of ovary, and 1 of 1 papillary cystadenoma of the epididymis, but no expression in 45 other tumors. In a systematic but preliminary study, 35 we have noted PAX-2 expression in 1 of 16 parathyroid hyperplastic lesions, 9 of 17 endometrioid carcinomas of uterus or ovary, and 3 of 9 serous carcinomas of uterus or ovary, but no expression in another 415 tumors. It is known that aside from parathyroid lesions, all tumor types other than RCCs that are positive for PAX-2 are related to the target organs of PAX-2 in term of transcription factor pathway. Since nephrogenic adenoma is potentially pertinent to renal tubular cell differentiation, 41 and serous or endometriod carcinomas are evidently related to müllerian organs, PAX-2 expression in some of these tumors is not unexpected. True aberrant PAX-2 expression is indeed exceptional. The current study showed the same spectrum of PAX-2 expression for metastatic tumors of nonrenal origin. Although 14 of 183 metastases (7.6%) in this study expressed PAX-2, only 2 (adenocarcinomas of colon and breast) represent aberrant PAX-2 expression, since their primary tumors were negative for PAX-2. PAX-2 expression in the other 12 metastases (2 nodal and 10 omental) is, however, expected, since their primary tumors are all of müllerian origin and these primary tumors also expressed PAX-2. In conclusion, among the known markers for RCC, PAX-2 is the most sensitive and specific, especially in the metastatic context. Furthermore, PAX-2 may also be helpful in diagnosing metastatic tumors of müllerian derivation. References 1. Azam F, Abubakerr M, Gollins S. Tongue metastasis as an initial presentation of renal cell carcinoma: a case report and literature review. J Med Case Reports. 2008;2:249. 2. Lordan JT, Fawcett WJ, Karanjia ND. Solitary liver metastasis of chromophobe renal cell carcinoma 20 years after nephrectomy treated by hepatic resection. Urology. 2008;72(1):230.e5 e6. 3. Rabbani F, Reuter VE, Katz J, Russo P. Second primary malignancies associated with renal cell carcinoma: influence of histologic type. Urology. 2000; 56(3):399 403. 4. MacLennan GT, Farrow G, Bostwick DG. Immunohistochemistry in the evaluation of renal cell carcinoma: a critical appraisal. J Urol Pathol. 1997;6: 195 203. 5. Cochand-Priollet B, Molinie V, Bougaran J, et al. Renal chromophobe cell carcinoma and oncocytoma: a comparative morphologic, histochemical, and immunohistochemical study of 124 cases. Arch Pathol Lab Med. 1997;121(10): 1081 1086. 6. 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(2):203 210. 7. Mazal PR, Exner M, Haitel A, et al. Expression of kidney specific cadherin distinguishes chromophobe renal cell carcinoma from renal oncocytoma. Hum Pathol. 2005;36(1):22 28. 8. Adley BP, Gupta A, Lin F, Luan C, Teh BT, Yang, XJ. Expression of kidneyspecific cadherin in chromophobe renal cell carcinoma and renal oncocytoma. Am J Clin Pathol. 2006;126(1):79 85. 1128 Arch Pathol Lab Med Vol 134, August 2010 PAX-2 and Metastatic Renal Cell Carcinoma Ozcan et al

9. Gnarra JR, Dressler GR. Expression of Pax-2 in human renal cell carcinoma and growth inhibition by antisense oligonucleotides. Cancer Res. 1995;55(18): 4092 4098. 10. Daniel L, Lechevallier E, Giorgi R, et al. Pax-2 expression in adult renal tumors. Hum Pathol. 2001;32(3):282 287. 11. Mazal PR, Stichenwirth M, Koller A, Blach S, Haitel A, Susani M. Expression of aquaporins and PAX-2 compared to CD10 and cytokeratin 7 in renal neoplasms: a tissue microarray study. Mod Pathol. 2005;18(4):535 540. 12. Memeo L, Jhang J, Asaad AM, et al. Immunohistochemical analysis for cytokeratin 7, KIT, and PAX2: value in the differential diagnosis of chromophobe cell carcinoma. Am J Clin Pathol. 2007;127(2):225 229. 13. 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(12):1485 1492. 14. Shen SS, Krishna B, Chirala R, Amato R, Truong LD. Kidney-specific cadherin, a specific marker for the distal portion of the nephron and related renal neoplasms. Mod Pathol. 2005;18(7):933 940. 15. Dressler GR, Douglass EC. Pax-2 is a DNA-binding protein expressed in embryonic kidney and Wilms tumor. Proc Natl Acad Sci U S A. 1992;89(4): 1179 1183. 16. Dressler GR, Wilkinson JE, Rothenpieler UW, Patterson LT, Williams- Simons L, Westphal H. Deregulation of Pax-2 expression in transgenic mice generates severe kidney abnormalities. Nature. 1993;362(6415):65 67. 17. Maulbecker CC, Gruss, P. The oncogenic potential of Pax genes. EMBO J. 1993;12(6):2361 2367. 18. Torres M, Gomex-Pardo E, Dressler GR, Gruss P. Pax-2 controls multiple steps of urogenital development. Development. 1995;121(12):4057 4065. 19. Mansouri A, Hallonet M, Gruss P. Pax genes and their roles in cell differentiation and development. Curr Opin Cell Biol. 1996;8(6):851 857. 20. Stuart ET, Gruss P. PAX: Developmental control genes in cell growth and differentiation. Cell Growth Differ. 1996;7(3):405 412. 21. Dahl E, Koseki H, Balling R. Pax genes and organogenesis. Bioassays. 1997;19(9):755 765. 22. Eccles MR. The role of Pax-2 in normal and abnormal development of the urinary tract. Pediatr Nephrol. 1998;12(9):712 720. 23. Dressler GR, Woolf AS. Pax2 in development and renal disease. Int J Dev Biol. 1999;43(5):463 468. 24. Kuure S, Vuolteenaho R, Vainio S. Kidney morphogenesis: cellular and molecular regulation. Mech Dev. 2000;92(1):31 45. 25. Igarashi T, Ueda T, Suzuki H, et al. Aberrant expression of Pax-2 mrna in renal cell carcinoma tissue and parenchyma of the affected kidney. Int J Urol. 2001;8(2):60 64. 26. Eccles MR, He S, Legge M, et al. PAX genes in development and disease: the role of PAX2 in urogenital tract development. Int J Dev Biol. 2002;46(4): 535 544. 27. Bouchard M, Souabni A, Mandler M, Neubüser A, Busslinger M. Nephric lineage specification by Pax2 and Pax8. Genes Dev. 2002;16(22):2958 2970. 28. Ozcan A, Zhai J, Candice H, et al. PAX-2 in the diagnosis of primary renal tumors: immunohistochemical comparison with the renal cell carcinoma marker antigen and kidney-specific cadherin. Am J Clin Pathol. 2009;131(3):393 404. 29. Eble JN, Sauter G, Epstein JI, Sesterhenn IA, eds. Pathology and Genetics of Tumours of the Urinary System and Male Genital Organs. Lyon, France: IARC Press; 2004. World Health Organization Classification of Tumours; vol 7. 30. Kuschert S, Rowitch DH, Haenig B, et al. Characterization of Pax-2 regulatory sequences that direct transgene expression in the Wolffian duct and its derivatives. Dev Biol. 2001;229(1):128 140. 31. Fickenscher HR, Chalepakis G, Gruss P. Murine Pax-2 protein is a sequence-specific trans-activator with expression in the genital system. DNA Cell Biol. 1993;12(5):381 391. 32. Tong GX, Chiriboga L, Hamele-Bena D, Borczuk AC. Expression of PAX2 in papillary serous carcinoma of the ovary: immunohistochemical evidence of fallopian tube or secondary Müllerian system origin? Mod Pathol. 2007;20(8): 856 863. 33. Gokden N, Gokden M, Phan DC, McKenney JK. The utility of PAX-2 in distinguishing metastatic clear cell renal cell carcinoma from its morphologic mimics: an immunohistochemical study with comparison to renal cell carcinoma marker. Am J Surg Pathol. 2008;32(10):1462 1467. 34. Wasco MJ, Pu RT. Comparison of PAX-2, RCC antigen, and antiphosphorylated H2AX antibody (gamma-h2ax) in diagnosing metastatic renal cell carcinoma by fine-needle aspiration. Diagn Cytopathol. 2008;36(8):568 573. 35. Zhai QJ, Ozcan A, Coffey D, et al. PAX-2 as a renal and Müllerian tumor marker: a comprehensive study. Mod Pathol. 2009;22(suppl 1s):391A. 36. Pan CC, Chen PC, Tsay SH, Ho DM. Differential immunoprofiles of hepatocellular carcinoma, renal cell carcinoma, and adrenocortical carcinoma: a systemic immunohistochemical survey using tissue array technique. Appl Immunohistochem Mol Morphol. 2005;13(4):347 352. 37. McCluggage WG, Wilkinson N. Metastatic neoplasms involving the ovary: a review with an emphasis on morphological and immunohistochemical features. Histopathology. 2005;47(3):231 247. 38. Amin MB, Amin MB, Tamboli P, et al. Prognostic impact of histologic subtyping of adult renal epithelial neoplasms: an experience of 405 cases. Am J Surg Pathol. 2002;26(3):281 291. 39. Hoffmann NE, Gillett MD, Cheville JC, Lohse CM, Leibovich BC, Blute ML. Differences in organ system of distant metastasis by renal cell carcinoma subtype. J Urol. 2008;179(2):474 477. 40. Gokden N, Mukunyadzi P, James JD, Gokden M. Diagnostic utility of renal cell carcinoma marker in cytopathology. Appl Immunohistochem Mol Morphol. 2003;11(2):116 119. 41. Tong GX, Melamed J, Mansukhani M, et al. PAX2: a reliable marker for nephrogenic adenoma. Mod Pathol. 2006;19(3):356 363. Arch Pathol Lab Med Vol 134, August 2010 PAX-2 and Metastatic Renal Cell Carcinoma Ozcan et al 1129