www.elsevier.com/locate/humpath

Human Pathology (2013) 44, 1412–1420

Original contribution

Renal cell carcinoma with areas mimicking renalangiomyoadenomatous tumor/clear cell papillaryrenal cell carcinoma☆,☆☆

Fredrik Petersson MD, PhDa,b, Petr Grossmann MSc, PhDb, Milan Hora MD, PhD c,Maris Sperga MDd, Delia Perez Montiel MDe, Petr Martinek MSc b,Maria Evelyn Cortes Gutierrez MD f, Stela Bulimbasic MD, PhDg, Michal Michal MDb,Jindrich Branzovsky MDb, Ondrej Hes MD, PhDb,⁎

aDepartment of Pathology, National University Health System, Singapore, SingaporebDepartment of Pathology, Charles University in Prague, Faculty of Medicine in Plzeň, Pilsen, Czech Republic 304 60cDepartment of Urology, Charles University in Prague, Faculty of Medicine in Plzeň, Pilsen, Czech Republic 305 99dDepartment of Pathology, Riga East University, Riga, Latvia 1038eDepartment of Pathology, Institute Nacional de Cancerologia, Mexico City 1400, MexicofDepartment of Pathology, Hospital Juarez de Mexico, Mexico City 1002, MexicogDepartment of Pathology, University Hospital Dubrava, Zagreb 10040, Croatia

Received 31 July 2012; revised 31 October 2012; accepted 2 November 2012

c

P

0h

Keywords:Clear cell papillary renalcell carcinoma;Clear renal cell carcinoma;Immunohistochemistry;Molecular genetics;Renal angiomyomatoustumor

Summary We present a cohort of 8 renal carcinomas that displayed a variable (5%-95% extent) lightmicroscopic appearance of renal angiomyoadenomatous tumor/clear cell papillary renal cell carcinoma(RAT/CCPRCC) without fulfilling the criteria for these tumors. All but 1 case predominantly (75%-95%extent) showed histopathologic features of conventional clear cell renal cell carcinoma. In 5 of 7 caseswith mostly conventional clear renal cell carcinoma (CRCC) morphology, a diagnosis of CRCC wassupported by the molecular genetic findings (presence of von Hippel-Lindau tumor suppressor [VHL]mutation and/or VHL promoter methylation and/or loss of heterozygosity [LOH] for 3p). Of the other 2cases with predominantly characteristic CRCC morphology, 1 tumor did not reveal any VHL mutation,VHL promoter methylation, or LOH for 3p, and both chromosomes 7 and 17 were disomic, whereas theother tumor displayed polysomy for chromosomes 7 and 17 and no VHL mutation, VHL promotermethylation, or LOH for 3p. One tumor was composed primarily (95%) of distinctly RAT/CCPRCC–like morphology, and this tumor harbored a VHL mutation and displayed polysomy for chromosomes 7and 17. Of the 5 cases with both histomorphologic features and molecular genetic findings of CRCC, wedetected significant immunoreactivity for α-methylacyl-CoA racemase in 2 cases and strong diffuse

☆ The authors declare that they have no conflict of interest.☆☆ The study was supported by the Czech Government grant agency (IGA NT12010-5) and by the project of the Ministry of Health, Czech Republic, for

onceptual development of research organization (00669806), Faculty Hospital in Pilsen, Czech Republic.⁎ Corresponding author. Department of Pathology, Charles University, Medical Faculty and Charles University Hospital Plzen, Alej Svobody 80, 304 60

ilsen, Czech Republic.E-mail address: hes@medima.cz (O. Hes).

046-8177/$ – see front matter © 2013 Elsevier Inc. All rights reserved.ttp://dx.doi.org/10.1016/j.humpath.2012.11.019

1413Renal cell carcinoma with areas mimicking RAT/CCPRCC

immunopositivity for cytokeratin 7 in 3 cases. Despite the combination of positivity for α-methylacyl-CoA racemase and cytokeratin 7 in 2 cases, there was nothing to suggest of the possibility of aconventional papillary renal cell carcinoma with a predominance of clear cells.© 2013 Elsevier Inc. All rights reserved.

1. Introduction

Renal angiomyoadenomatous tumor (RAT) and clear cellpapillary renal cell carcinoma (CCPRCC) are 2 relativelyrecently described low-grade neoplastic entities with over-lapping histologic features and with indolent clinicalbehavior so far [1-6]. Recently, RAT and CCPRCC havebeen thought of as 2 ends of a spectrum of 1 nosologic entity.However, there is still no consensus on this, and we willherein refer to them as RAT/CCPRCC, implying that we donot take a categorical stand on the issue of “lumping orsplitting” this group of tumors. Although the histopathologicfeatures of these tumors are characteristic, there existoverlapping features with other renal tumors predominantlycomposed of clear cells such as conventional clear cell renalcell carcinoma (CRCC) and papillary renal cell carcinoma(PRCC) with clear cells, as well as renal cell carcinoma(RCC) not otherwise specified (NOS). We have identified acohort of non-RAT/non-CCPRCC renal tumors displayingvariably extensive areas with a striking resemblance to RAT/CCPRCC. In this study, we characterize the clinicopatho-logic as well as immunohistochemical and molecular geneticfeatures of this group of tumors and discuss the implicationsfor the practicing surgical pathologist who is faced with thetask of diagnosing renal tumors.

Table 1 Probes for aneuploidy detection of chromosomes 7and 17

Chromosome Probe

7 CEP 7 (D7Z1)/7p11.1-q11.1Alpha Satellite DNA

17 CEP 17 (D17Z1)/17p11.1-q11.1Alpha Satellite DNA

2. Materials and methods

Of 290 cases deposited or catalogued in the Plzeň TumorRegistry under the diagnosis “unusual/hybrid/clear cell/RAT-like/renal carcinoma,” 8 cases of renal tumorspredominantly composed of clear cells with a variablepresence of areas reminiscent of RAT/CCPRCC wereretrieved. The cases were reviewed by the first and lastauthors of this study (F.P. and O.H.).

2.1. Immunohistochemistry

The immunohistochemical study was performed using aVentana Benchmark XT automated stainer (Ventana Med-ical System, Tucson, AZ). Antibodies against cytokeratin(CK) 7 (1:200, monoclonal, OV-TL 12/30; Dako, Glostrup,Denmark), pankeratin (polyclonal, AE1-AE3/PCK26, RTU;Ventana-Roche Diagnostics, Basel, Switzerland), high-molecular-weight CK K903 (1:300, monoclonal, 34βE12;Dako), CD10 (1:20, monoclonal, 56C6; Novocastra, Bur-lingame, CA), α-methylacyl-CoA racemase (AMACR;

1:200, monoclonal, 13H4; Dako), paired box protein Pax-2(1:100, polyclonal; Invitrogen, Camarillo, CA), transcriptionfactor E3 (1:100, polyclonal; Abcam, Cambridge, UK),vimentin (1:1000, monoclonal, D9; Neomarkers, Westing-house, CA), carbonic anhydrase 9 (CAIX) (1:100, mono-clonal, 303123; R&D Systems, Minneapolis, MN), Ki-67(MIB1, 1:1000, monoclonal; Dako), desmin (1:2000,monoclonal, D33; DakoCytomation, Carpinteria CA), andsmooth muscle actin (monoclonal, 1A4, RTU; Ventana-Roche) were applied.

2.2. Molecular genetic study

2.2.1. Fluorescence in situ hybridization methodsEach 4-μm-thick section was placed onto a positively

charged slide. Each hematoxylin and eosin–stained slide wasexamined for determination of areas for cell counting. Areason the sections featuring classical clear renal cell carcinoma(CRCC) and RAT/CCPRCCmorphology were annotated, andthe corresponding areas were manually cut from the blocks forseparate analyses. The unstained slide was routinely depar-affinized and incubated in the 1× Target Retrieval SolutionCitrate pH 6 (Dako) for 40 minutes at 95°C and subsequentlycooled for 20 minutes at room temperature in the samesolution. The slide was washed in deionized water for 5minutes, and tissue was covered with Proteinase K (20 mg/mL) (SERVA, Heidelberg, Germany) for 6 minutes at roomtemperature. The slide was then placed into deionized waterfor 5 minutes, dehydrated in a series of ethanol solutions(70%, 85%, and 96% for 2 minutes each), and air dried.

Probes for aneuploidy detection of chromosomes 7 and 17(Vysis/Abbott Molecular, Abbott Park, IL) were mixed withwater and LSI/WCP (locus-specific identifier/whole chro-mosome painting) hybridization buffer (Vysis) in a 1:2:7ratio (Table 1). An appropriate amount of probe mix wasapplied to each specimen, covered with a glass coverslip, andsealed with rubber cement. The slide was incubated in theThermoBrite instrument (StatSpin/Iris Sample Processing,Westwood, MA) with codenaturation parameters 85°C for

1414 F. Petersson et al.

8 minutes and hybridization parameters 37°C for 16 hours.The coverslip was then removed, and the slide was placed inposthybridization wash solution (2× SSC/0.3% NP-40) at72°C for 2 minutes. The slide was air dried in the dark,counterstained with DAPI I (Vysis), coverslipped, andimmediately examined.

2.2.2. Fluorescence in situ hybridization interpretationThe section was examined with an Olympus BX60

fluorescence microscope (Olympus Corp., Tokyo, Japan)using a 100× objective and filter sets Triple Band Pass(DAPI/Spectrum Green/Spectrum Orange) and Single BandPass (Spectrum Green/Spectrum Orange). Scoring ofaneuploidy was performed by counting the number offluorescent signals in 100 randomly selected nonoverlappingtumor cell nuclei. The slide was independently enumeratedby 2 observers (O.H. and P.G.). Monosomy and polysomyfor studied chromosomes were defined as the presence of 1signal per cell in more than 45% and 3 or more signals inmore than 10% (mean ± 3 SD in normal nonneoplasticcontrol tissues), respectively.

2.2.3. DNA extraction and bisulfite DNA conversionDNA for molecular genetic investigation was extracted

from formalin-fixed, paraffin-embedded tissue. Several 5-μm-thick sections were placed on the slides. Hematoxylinand eosin–stained slides were examined for identification ofneoplastic tissue. Subsequently, neoplastic tumor andnonneoplastic tissue from unstained slides were scraped,and DNA was isolated by the NucleoSpin Tissue Kit(Macherey-Nagel, Düren, Germany). Bisulfite conversionof DNA was performed using the EZ DNA Methylation-Gold Kit (DNA input 500 ng; Zymo Research, Orange,CA). All procedures were performed according to themanufacturers' protocols.

Table 2 PCR primers used in mutation analysis of the VHLgene and designed in Primer3 software [24]

Name Primer (sequence 5′ → 3′)

VHL/exon 1VHL e1-1 CGCGAAGACTACGGAGGTVHL e1-2 GTCTTCTTCAGGGCCGTAVHL e1-3 GAGGCAGGCGTCGAAGAGVHL e1-4 GCGATTGCAGAAGATGACCTVHL e1-5 GCCGAGGAGGAGATGGAGVHL e1-6 CCCGTACCTCGGTAGCTGTVHL e1-7 CCGTATGGCTCAACTTCGACVHL e1-8 GCTTCAGACCGTGCTATCGTVHL/exon 2VHL e2-1 ACCGGTGTGGCTCTTTAACAVHL e2-2 TCCTGTACTTACCACAACAACCTTVHL/exon 3VHL e3-1 GCAAAGCCTCTTGTTCGTTCVHL e3-2 ACATTTGGGTGGTCTTCCAGVHL e3-3 CAGGAGACTGGACATCGTCAVHL e3-4C CCATCAAAAGCTGAGATGAAAC

2.2.4. VHL gene analysisFoci of CRCC and RAT/CCPRCC morphology were

analyzed separately. Mutation analysis of exons 1, 2, and 3of the Hippel-Lindau tumor suppressor (VHL) gene wasperformed using polymerase chain reaction (PCR) and directsequencing. PCR was performed using the primers shown inTable 2. The reaction conditions were as follows: 12.5 μL ofHotStar Taq PCR Master Mix (QIAgen, Hilden, Germany),10 pmol of each primer, 100 ng of template DNA, anddistilled water up to 25 μL. The amplification programconsisted of denaturation at 95°C for 15 minutes and then 40cycles of denaturation at 95°C for 1 minute, annealing at55°C for 1 minute and extension at 72°C for 1.5 minutes forall amplicons. The program was finished by 72°C incubationfor 7 minutes.

The PCR products were checked via 2% agarose gelelectrophoresis. Successfully amplified PCR products werepurified with Agencourt AMPuremagnetic particles (Agen-court Bioscience Corporation, Beverly, MA), both sidessequenced using Big Dye Terminator Sequencing kit(Applied Biosystems, Foster City, CA) and purified withmagnetic particles Agencourt CleanSEQ (Agencourt Biosci-ence), all according to the manufacturer's protocol, andsubsequently run on an automated sequencer (ABI Prism3130xl; Applied Biosystems) at a constant voltage of 13.2kV for 20 minutes.

All samples were analyzed in duplicate. Analyses ofpositive samples were repeated.

2.3. Analysis of VHL promoter methylation

Foci of CRCC and RAT/CCPRCC morphology wereanalyzed separately. Detection of promoter methylation wasperformed via methylation-specific PCR, as described byHerman et al [7]. Briefly, 100 ng of DNA or 2 μL ofconverted DNA was added to the reaction consisting of 12.5μL of HotStar Taq PCR Master Mix (QIAgen), 10 pmol offorward and reverse primers (Table 3), and distilled water upto 25 μL. The amplification program comprised denaturationat 95°C for 14 minutes and then 40 cycles of denaturation at95°C for 1 minute, annealing at 60°C for 1 minute, andextension at 72°C for 1 minute. The program was finished byincubation at 72°C for 7 minutes. The PCR products werechecked via 2% agarose gel electrophoresis.

Table 3 PCR primers used in methylation status of the VHLgene

Name Primer (sequence 5′ → 3′)

Unmethylated VHLVHL-U-S GTTGGAGGATTTTTTTGTGTATGTVHL-U-A CCCAAACCAAACACCACAAAMethylated VHLVHL-M-S TGGAGGATTTTTTTGCGTACGCVHL-M-A GAACCGAACGCCGCGAA

Table 5 Basic clinicopathologic data

Case no. Age (y) Size (cm) Sex Site Follow-up

1 66 4.5 M L LF2 43 2 × 3 × 2 M R LF3 60 1.5 a F L LF4 51 5 × 4 × 4 F R LF5 74 1.8 a and 4.8 a F UK LF6 51 11.5 × 12 × 10.5 F L 3 y AW7 71 3.1 × 1.8 × 1.6 M R 1 y AW8 52 4 × 3.5 × 3 M R 5 mo AW

Abbreviations: M, male; F, female; R, right; L, left; LF, lost to follow-up; AW, alive and well; UK, unknown.

a Diameter.

1415Renal cell carcinoma with areas mimicking RAT/CCPRCC

A patient with known VHLmutation and fully methylatedHeLa cell DNA were used as a positive control for VHLmutation analysis and promoter methylation analysis,respectively. As a negative control, randomly selectedhealthy donor blood was used.

2.4. 3p Loss of heterozygosity analysis

Foci of CRCC and RAT/CCPRCC morphology wereanalyzed separately. For loss of heterozygosity (LOH)analysis of neoplastic tissue DNA, 10 STR (short tandemrepeats) markers—D3S666, D3S1270, D3S1300, D3S1581,D3S1597, D3S1600, D3S1603, D3S1768, D3S2338, andD3S3630—located on the short arm of chromosome 3 (3p)were chosen from the database (GenBank UniSTS). Theprimers are listed in Table 4. PCR conditions were the sameas mentioned earlier. Successfully amplified PCR productswere mixed with Gene Scan-500LIZ Size Standard (AppliedBiosystems) and run on an automated genetic analyzer ABIPrism 3130xl (Applied Biosystems) at a constant voltage of15 kV for 20 minutes. A sample was considered LOHpositive if the ratio of nontumor DNA to tumor DNA wasgreater than 1.5 or less than 0.66. All samples were analyzedin duplicate.

3. Results

Basic clinical data are presented in Table 5. Briefly, therewere 4 male and 4 female patients with an age range of 43 to74 years (median, 56 years). The size of the tumors ranged

Table 4 PCR primers used in LOH analysis of chromosome 3p

Marker Name

D3S666 D3S666-SK#15D3S666-SK#16

D3S1270 D3S1270-FD3S1270-R

D3S1300 D3S1300SFD3S1300A

D3S1581 D3S1581-FD3S1581-R

D3S1597 D3S1597-FD3S1597-R

D3S1600 D3S1600-FD3S1600-R

D3S1603 D3S1603-FD3S1603-R

D3S1768 D3S1768SFD3S1768A

D3S2338 D3S2338-FD3S2338-R

D3S3630 D3S3630-FD3S3630-R

from 1.5 to 11.5 cm (median, 3.5 cm). Follow-up wasavailable for 3 patients (range, from 5 to 39 months; median,12 months).

3.1. Histology

All but 1 (case 1) of the tumors showed predominanthistopathologic features of CRCC with a variable (5%-25%)presence of areas that displayed morphologic featurescharacteristically seen in RAT/CCPRCC (Figs. 1-3). Case1 was predominantly (95%) composed of RAT/CCPRCCareas including smooth muscle (“leiomyomatous”) stroma,elongated variably open and collapsed tubules lined by asingle layer of columnar-shaped clear cells with basallylocated low-grade nuclei (Fig. 4), and a prominentperitubular capillary network. In addition, areas morereminiscent of conventional clear cell (alveolated pattern

Primer (sequence 5′ → 3′)

CAAGGCATTAAAGTGGCCACGCGTTTGAACCAGTTTCCTACTGAGTGGAACTGTATCAAAGGCTCTTGCATTAGNATTCTCCAGAAGCTCACATTCTAGTCAGCCTGCCAATTCCCCAGATGCAGAACTGCCAAACCAGGGTAACAGGAGCGAGAGTACAAATACACACAAATGTCTCGCAAATCGTTCATTGCTATCACCATCATCTGCCTGTCTGCTTGCCTTGGGATTTACCCTAACTCCACTTGAAAGCTCAGCGAACAGCAACAAATGGTTGCTGCCAAAGATTAGACACTGTGATTTGCTGTTGGAGAAGCCAGCAGTTTCTCCTGTATTGTTTTCCAGGATAAGAAGGGATAAGCTGCAAATCAACCAAATACAATTCATGAGACCTGA

Fig. 1 Low-power view of tumor (case 2) predominantlycomposed of clear cells and with focal areas showing histologicfeatures strongly reminiscent of RAT/CCPRCC (hematoxylin-eosin, ×40).

Fig. 3 CRCC with long branching tubules surrounded by finecapillary network typical for RAT/CCPRCC (case 5). Inset: similarstructures shown in case of typical RAT/CCPRCC (hematoxylin-eosin, ×100).

1416 F. Petersson et al.

with cysts) were also present. The clinicopathologic andgenetic features of case 1 have been published previously [8].In 2 cases (cases 2 and 5), small foci of osseous metaplasiawithin the stromal component were found. Case 3 containedabundant smooth muscle stroma encircling variably sizedtumor cell aggregates with clear cell morphology. Also, incases 1, 7, and 8, a variable presence of smooth musclestroma was identified within the tumors (Fig. 5A-C). Mostsmooth muscle stroma seemed to emanate from moderatelylarge veins in the periphery of the tumors.

3.2. Immunohistochemistry

The immunohistochemical findings are summarized inTable 6. The results of the immunohistochemical study were

Fig. 2 Areas with RAT/RAT-like morphology (case 6): predom-inantly CRCC with “shark smiles.” Inset: RAT/CCPRCC withtypical shark smile structures lined by columnar cells with blister-like luminal parts (hematoxylin-eosin, ×100).

mostly consistent with CRCC in cases 2 to 8 (correspondingto the predominant light microscopic features). Despite thepredominant morphology of CRCC, strong immunoreactiv-ity for CK7 was identified in cases 2, 3, 5, and 8 (Fig. 6). Thepositive staining was seen in 40% of the tumor cells in case 2and was diffuse in cases 3, 5, and 8. Extensive immunore-activity for AMACR with variable intensity (weak to strong)was seen in cases 2 and 4 (Fig. 7). Moderate staining forAMACR was present in case 8. The tumor cells in case 4 didnot express vimentin.

3.3. Molecular genetic study

The results of the molecular genetic study are summarizedin Table 7. A mutation in the VHL gene was detected in cases1, 2, 7, and 8. LOH for 3p was identified in cases 2, 3, and 8,

Fig. 4 Low-power view of case 1 resembling RAT/CCPRCC(hematoxylin-eosin, ×40).

Fig. 5 Leiomyomatous stroma is a characteristic feature of RAT.A, A bona fide RAT/CCPRCC. B, CRCC (case 4) with RAT/CCPRCC features. C, Leiomyomatous stroma may also beextensive in non-RAT/non-CCPRCC tumors, as shown in case 3.Hematoxylin-eosin: A and C: ×40, B: ×200.

1417Renal cell carcinoma with areas mimicking RAT/CCPRCC

whereas methylation of the VHL promoter was detected incase 6. Polysomy of both chromosomes 7 and 17 was presentin cases 1 and 5 in RAT/CCPRCC-like areas. Polysomyof chromosome 7 was identified in cases 7 and 8 in

RAT/CCPRCC-like areas. In all other cases (both CRCC andRAT/CCPRCC-like areas), chromosomes 7 and 17were disomic.

4. Discussion

Herein we present a cohort of renal tumors predominantlycomposed of clear cells and with focal areas showinghistologic features strongly reminiscent of RAT/CCPRCC,yet without fulfilling the diagnostic criteria for thesetumors. All cases except case 1 showed unquestionable lightmicroscopic features of CRCC. The distinction betweenCRCC and RAT/CCPRCC is critical because no case ofRAT/CCPRCC has, to date, behaved aggressively, that is, nolocal recurrences or metastatic behavior has been documented.Morphologically, RAT/CCPRCC is typified by tubules, smallcompact nests and abortive papillae lined by neoplasticcolumnar epithelial cells with optically clear cytoplasm,frequently with an apical “blister” quality and low-grade,basally located nuclei; a well-developed peritubular capillarynetwork; and slightly open, variably angulated tubules thathave been likened to the “smile of a shark” and variablepresence of a leiomyomatous/myofibroblastic stroma [1-5].Immunohistochemically, the neoplastic epithelial cells ofRAT/CCPRCC express CK7 and Pax-2, whereas expressionof CAIX and vimentin is variable and only few tumor cells insome tumors have expressed CD10 or AMACR.

On a molecular genetic level, RAT/CCPRCC lacks VHLgene mutations and shows absence of LOH for 3p [1,4,9-11].Although trisomy and polysomy of chromosome 7 have beendescribed [9,12], numerical aberrations of chromosome 17appear to be highly unusual, with only 1 published case thatshowed trisomy 17 [1]. The rarity of this finding is supportedby Rohan et al [10], who could not demonstrate trisomies ofchromosome 7 and 17 in any of their cases. Loss ofchromosome Y has not been documented in RAT/CCPRCC.Of note is that Kuroda et al [13] have shown that CRCC maynot infrequently show gain of chromosome 7.

In all but 1 case (case 1) in this series, the overwhelminglydominant light microscopic pattern was unquestionably thatof CRCC, and in 5 of 7 cases, this was supported by themolecular genetic findings (VHL mutation or VHL promotermethylation or LOH for 3p). However, although conven-tional CRCC shows VHL mutation in only approximately34% to 56% of cases, the finding of a VHLmutation in a casewith clear cell morphology strongly argues for a diagnosis ofCRCC [14].

Although case 1 also displayed a VHL mutation, theoverwhelmingly large part displayed distinct RAT/CCRCC-like morphology. This case is highly unusual in anotheraspect as well, namely, that both chromosomes 7 and 17were polysomic. Despite showing clear-cut CRCC morphol-ogy, case 4 did not reveal any molecular genetic abnormal-ities, that is, no VHL mutation, VHL promoter methylation,or LOH for 3p, and both chromosomes 7 and 17 were

Table 6 Immunohistochemical results

Caseno.

CK7 K903 AE1-AE3/PCK26

CD10 CANH AMACR TFE3 Actin Desmin PAX2 Ki-67 Vim

1 +++ Pos +to +++

+++ Mostly neg++ 5%

Variable, negto + to ++

Neg Neg Weak + Weak + + b5% +++

2 Strong+++ 40%neg 60%

Neg Strong+++ 30%neg 70%

+++ Variable, +to ++/+++;some neg areas

Pos +to +++

Neg Neg Neg Neg b5% +++

3 +++ ++/+++ +++ +++ Foc + to +++ Neg Neg + + Weak focal +++4 Neg Neg Neg +++ Neg Foc +

to ++Foc +to +++

Neg Neg Neg b5% Neg

5 +++ + to +++ +++ Large areasneg; someareas ++to +++

Neg to + to++ to +++

Foc + Fewcells ++

+ + Neg 50%weak +50%

b5% +++

6 Neg Neg +++ +++ +++ mem Foc + Foc + Foc + Neg Weakfocal +

20% +++

7 Foc+(minimal)

Neg Neg +++ ++/+++;some areas neg

Neg Neg +++ Neg Neg ++/+++

8 +++ Foc + +++ ++/+++ +++ mem ++ Neg +++ Neg +++ b5% +++

Abbreviations: AE1-AE3/PCK26, pankeratin; TFE3, transcription factor E3; Vim, vimentin; pos, positive; neg, negative; foc, focally; +, weak positivity; ++,moderate positivity; +++, strong positivity; mem, membranous positivity; foc, focal positivity.

1418 F. Petersson et al.

disomic. Also, on an immunohistochemical level, case 4showed unusual features. Although CD10 was positive,vimentin, CAIX, and Pax-2 were completely negative. Inaddition, significant immunoreactivity for AMACR waspresent. Case 5, which also displayed histologic featurespredominantly of CRCC and an immunohistochemicalprofile that largely was the one expected for CRCC (exceptfor strong and diffuse positivity for CK7 but negative forAMACR), displayed polysomy for chromosomes 7 and 17selectively—in the part of the tumor that showed RAT/CCRCC-like morphology—whereas no VHL mutation,VHL promoter methylation, or LOH for 3p was present inany (CRCC or RAT/CCRCC-like) part of the tumor.

Fig. 6 Immunoreactivity for CK7 is considered an integral part ofthe RAT/CCPRCC immunohistochemical profile, here shown in aCCRCC (case 2) with RAT/CCPRCC features (×40).

Of the 5 cases with both histomorphologic features andmolecular genetic findings compatible with CRCC (presenceof VHL mutation and/or VHL promoter methylation and/orLOH 3p), we detected some unexpected immunohistochem-ical features. Significant immunoreactivity was seen forAMACR in 2 cases (nos. 2 and 8) and strong and frequentlydiffuse immunopositivity for CK7 in 3 cases (nos. 2, 3, and8). As early as 1995, expression of CK7 in CRCC (albeit notstrong and diffuse) has been documented by Gatalica et al[15], who described immunoreactivity for CK7 in 24% ofclassic CRCC.

Despite the combination of positivity for AMACR andCK7 in cases 2 and 8, these tumors did not show any lightmicroscopic features to suggest the possibility of a

Fig. 7 CCRCC with RAT/CCPRCC features (case 8) showingstrong AMACR positivity (×40).

Table 7 Summary of morphologic, unusual immunohistochemical, and molecular genetic findings

Case %CRCC

% RAT/CCPRCC

Unusual IHC Molecular genetics

CRCC RAT/CCPRCC

1 5 95 CK7+ NP VHL mut+ a; promoter VHL methylation−;LOH 3p−; 7,17 polysomic

2 95 5 CK7+,AMACR+

VHL mut+ b; promoter VHL methylation−;LOH 3p+; 7,17 disomic

VHL mut+ c; promoter VHL methylation−;LOH 3p+; 7,17 disomic

3 90 5 CK7+ VHL mut−; promoter VHL methylation−;LOH 3p+; 7,17 disomic

VHL mut−; promoter VHL methylation−;LOH 3p+; 7,17 disomic

4 90 10 AMACR+, vim−,CAH IX−,PAX2−

VHL mut−; promoter VHL methylation−;LOH 3p−; 7,17 disomic

VHL mut−; promoter VHL methylation−;LOH 3p−; 7,17 disomic

5 85 15 VHL mut−; promoter VHL methylation−;LOH 3p−; 7,17 disomic

VHL mut−; promoter VHL methylation−;LOH 3p−; 7,17 polysomic

6 75 15 VHL mut−; promoter VHL methylation+;LOH 3p+; 7,17 disomic

VHL mut−; promoter VHL methylation+;LOH 3p+; 7,17 disomic

7 90 10 VHL mutNA; promoter VHL methylation−;LOH 3pNA; 7,17 disomic

VHL mut+ d; promoter VHL methylation−;LOH 3p+; 7 polysomic, 17 disomic

8 95 5 CK7+, AMACR+ VHL mut+ e; promoter VHL methylation−;LOH 3p+; 7,17 disomic

VHL mut+ e; promoter VHL methylation−;LOH 3p+; 7 polysomic, 17 disomic

Abbreviations: % of CRCC, % of area compatible with CRCC; % of RAT/CCPRCC, area compatible with RAT/CCPRCC; IHC, immunohistochemistry;vim, vimentin; CAH9, carboanhydrase IX; +, positive; −, negative; VHL, VHL gene; NP, not performed; NA, not analyzable.

a c.374dupA/p.H125Qfs*7.b c.234T N G/p.N78K.c c.234dupA/p.N78fs*54.d c.213delC/p.S72fs*87.e c.529A N T/p.R177*.

1419Renal cell carcinoma with areas mimicking RAT/CCPRCC

conventional PRCC with a predominance of clear cells. Incontrast, and as mentioned earlier, 1 case (no. 2) revealedLOH for 3p and (different!) mutations in the VHL gene inthe CRCC and RAT/CCRCC-like areas, whereas chromo-somes 7 and 17 were disomic (again in both parts of thetumor). Also, in case 8, the same VHL mutations and LOH3p were detected in the CRCC and RAT/CCRCC-likeareas; chromosomes 7 and 17 were disomic in the CRCCarea, whereas in the RAT/CCRCC-like area, 17 wasdisomic and 7 was polysomic. Similarly, in case 7, bothVHL mutation and LOH 3p were present in the RAT/CCRCC-like area, which was polysomic for chromosome 7(and disomic for 17). Regrettably, the part of the tumorwith CRCC morphology was not analyzable for VHLmutation and LOH 3p.

Case 5 had an immunohistochemistry profile that wasconsistent with CRCC. No VHL gene mutation, LOH 3p, orVHL promoter methylation was detected. However, poly-somy of both chromosomes 7 and 17 was present in theRAT/CCRCC-like part of the tumor, whereas the CRCC areawas disomic for both chromosomes. This is a highly unusualcombination for CRCC, and from a chromosomal aberrationpoint of view, the cytogenetic profile of the RAT/CCRCC-like area in this case is more characteristic of conventionalPRCC or tubulocystic RCC [16,17].

Two highly interesting phenomena were identified in 3 ofour cases. (1) As stated earlier, 2 different VHL mutationswere documented in case 2, corresponding to the 2 different

morphologies (CRCC and RAT/CCPRCC-like). (2) Differ-ences in numerical chromosomal aberrations (between theparts of the tumors with CRCC and RAT/CCRCC-like) weredocumented in cases 7 and 8. These findings are inaccordance with findings of Gerlinger et al [18], whoshowed striking genetic heterogeneity in RCCs.

As mentioned previously, given the (so far) indolentbehavior of RAT/CCPRCC, the distinction from other clearcell neoplasms with RAT/CCPRCC-like features is ofutmost clinical relevance. Moreover, given the emergingdifferentiated targeted oncologic treatment of renal carcino-mas (CRCC versus translocation RCC versus PRCC),correct tumor classification is becoming an even morecrucial task for the diagnostic surgical pathologist [19,20].

One important point of our study is that relying too muchon immunohistochemistry when classifying clear cell tumorsis fraught with danger. We have documented strongimmunoreactivity for CK7 and AMACR, either in combi-nation or in isolation, in tumors that, given the lightmicroscopic features and major genetic findings, cannot beclassified as anything other than CRCC. We feel that thisdiagnostic aspect of clear cell renal neoplasms has not beenadequately highlighted in the literature. As clearly shown inour series, “aberrant” expression of CK7 and AMACR inCRCC should not lead the diagnostician astray from adiagnosis of CRCC.

The leiomyomatous stroma that originally was thought tobe a highly specific feature of RATmaymerit reconsideration.

1420 F. Petersson et al.

We and others [21-23] have clearly demonstrated that bonafide CRCC may harbor a smooth muscle–rich stroma. Thismay at times be extensive, as demonstrated by Canzonieri et al[21] and in our case 3. One case (no. 1) in this series representsa significant differential diagnostic problem as far as tumorclassification is concerned. For the time being, we considerthis case to belong to the very heterogeneous category ofRCC-NOS. Also, case 5 represents another difficult case froma nosologic point of view. The light microscopic features wereundoubtedly those of CRCC. However, from a genetic pointof view, this case shows more similarities to conventionalPRCC or tubulocystic RCC. On an immunohistochemicallevel, strong immunoreactivity was detected for CK7.Otherwise, the findings were those expected for CRCC.

In conclusion, renal carcinomas predominantly composedof clear cells may show areas indistinguishable on a lightmicroscopic level from RAT/CCPRCC. This is important tobear in mind when faced with limited material (core biopsy orinadequate sampling) from a low-grade renal neoplasmcomposed of clear cells. The immunohistochemical detectionof significant expression of CK7 and/or AMACR in bona fideCRCC should not create diagnostic confusion. In selectedcases, performing molecular genetic studies pertaining to VHLmutation, LOH of 3p, and numerical analysis of chromosomes7 and 17 would contribute to reaching the correct diagnosis. Inthis context it is important to bear in mind that polysomy ofchromosome 7 is a nondistinguishing feature between RAT/PRCC and CRCC. Moreover, despite the use of these adjuncttechniques, there exists a small subset of tumors showingvariably extensive RAT/CCPRCC-like morphology thatcurrently defies classification. This group of tumors is, forthe time being, best classified as RCC-NOS. The biologicalbehavior of these tumors requires further study.

Our data also highlight the importance of adequatesampling for morphologic, immunohistochemical, andmolecular genetic studies within the currently applieddiagnostic framework. On a more theoretical level, ourresults also corroborate emerging data on intratumoralgenetic heterogeneity in renal neoplasms, which likely willaffect the classification of renal tumors in the future.

References

[1] Aydin H, Chen L, Cheng L, et al. Clear cell tubulopapillary renal cellcarcinoma: a study of 36 distinctive low-grade epithelial tumors of thekidney. Am J Surg Pathol 2010;34:1608-21.

[2] Gobbo S, Eble JN, Grignon DJ, et al. Clear cell papillary renal cellcarcinoma: a distinct histopathologic and molecular genetic entity. AmJ Surg Pathol 2008;32:1239-45.

[3] Michal M, Hes O, Havlicek F. Benign renal angiomyoadenomatoustumor: a previously unreported renal tumor. Ann Diagn Pathol 2000;4:311-5.

[4] Michal M, Hes O, Nemcova J, et al. Renal angiomyoadenomatoustumor: morphologic, immunohistochemical, and molecular geneticstudy of a distinct entity. Virchows Arch 2009;454:89-90.

[5] Tickoo SK, dePeralta-Venturina MN, Harik LR, et al. Spectrum ofepithelial neoplasms in end-stage renal disease: an experience from 66tumor-bearing kidneys with emphasis on histologic patterns distinctfrom those in sporadic adult renal neoplasia. Am J Surg Pathol2006;30:141-53.

[6] Ross H, Martignoni G, Argani P. Renal cell carcinoma with clear celland papillary features. Arch Pathol Lab Med 2012;136:391-9.

[7] Herman JG, Graff JR, Myohanen S, et al. Methylation-specific PCR: anovel PCR assay for methylation status of CpG islands. Proc NatlAcad Sci USA 1996;93:9821-6.

[8] Petersson F, Yan B, Huang J, et al. Low-grade renal carcinoma withhistologic features overlapping with renal angiomyoadenomatoustumor and featuring polysomy 7 and 17 and a mutation in the vonHippel-Lindau gene: report of a hybrid tumor and a few comments onrenal angiomyoadenomatous tumor and papillary renal tumors withclear cells. Ann Diagn Pathol 2011;15:213-6.

[9] Wolfe A, Dobin SM, Grossmann P, et al. Clonal trisomies 7, 10 and12, normal 3p and absence of VHL gene mutation in a clear celltubulopapillary carcinoma of the kidney. Virchows Arch 2011;459:457-63.

[10] Rohan SM, Xiao Y, Liang Y, et al. Clear-cell papillary renal cellcarcinoma: molecular and immunohistochemical analysis with em-phasis on the von Hippel-Lindau gene and hypoxia-inducible factorpathway-related proteins. Mod Pathol 2011;24:1207-20.

[11] Kuroda N, Michal M, Hes O, et al. Renal angiomyoadenomatoustumor: fluorescence in situ hybridization. Pathol Int 2009;59:689-91.

[12] Kuroda N, Shiotsu T, Kawada C, et al. Clear cell papillary renal cellcarcinoma and clear cell renal cell carcinoma arising in acquired cysticdisease of the kidney: an immunohistochemical and genetic study. AnnDiagn Pathol 2011;15:282-5.

[13] Kuroda N, Tamura M, Shiotsu T, et al. Chromosomal abnormalities ofclear cell renal cell carcinoma: frequent gain of chromosome 7. PatholInt 2010;60:9-13.

[14] Eble JN, Sauter G, Epstein JI, Sesterhenn IA, editors. Pathology andgenetics of tumours of the Urinary System and Male Genital Organs.World Health Organization classification of tumours. Lyon: IARC;2004.

[15] Gatalica Z, Kovatich A, Miettinen M. Consistent expression ofcytokeratin 7 in papillary renal cell carcinoma. J Urol Pathol 1995;3:205-11.

[16] Zhou M, Yang XJ, Lopez JI, et al. Renal tubulocystic carcinoma isclosely related to papillary renal cell carcinoma: implications forpathologic classification. Am J Surg Pathol 2009;33:1840-9.

[17] Yang XJ, Zhou M, Hes O, et al. Tubulocystic carcinoma of the kidney:clinicopathologic and molecular characterization. Am J Surg Pathol2008;32:177-87.

[18] Gerlinger M, Rowan AJ, Horswell S, et al. Intratumor heterogeneityand branched evolution revealed by multiregion sequencing. N Engl JMed 366:883-92

[19] Linehan WM, Bratslavsky G, Pinto PA, et al. Molecular diagnosis andtherapy of kidney cancer. Annu Rev Med 2010;61:329-43.

[20] Pfaffenroth EC, Linehan WM. Genetic basis for kidney cancer:opportunity for disease-specific approaches to therapy. Expert OpinBiol Ther 2008;8:779-90.

[21] Canzonieri V, Volpe R, Gloghini A, et al. Mixed renal tumor withcarcinomatous and fibroleiomyomatous components, associated withangiomyolipoma in the same kidney. Pathol Res Pract 1993;189:951-6[discussion 957–9].

[22] Kuhn E, De Anda J, Manoni S, et al. Renal cell carcinoma associatedwith prominent angioleiomyoma-like proliferation: report of 5 casesand review of the literature. Am J Surg Pathol 2006;30:1372-81.

[23] Shannon BA, Cohen RJ, Segal A, et al. Clear cell renal cell carcinomawith smooth muscle stroma. HUM PATHOL 2009;40:425-9.

[24] Rozen S, Skaletsky H. Primer3 on the WWW for general users and forbiologist programmers. Methods Mol Biol 2000;132:365-86.