Virchows Arch (2013) 463:819–825 DOI 10.1007/s00428-013-1497-y

ORIGINAL ARTICLE

Nephrogenic adenoma of the urinary tract: clinical, histological, and immunohistochemical characteristics José I. López & Marco Schiavo-Lena & Alexandra Corominas-Cishek & Adriana Yagüe & Kevin Bauleth & Rosa Guarch & Ondrej Hes & Regina Tardanico

Received: 19 August 2013 / Revised: 24 September 2013 / Accepted: 7 October 2013 / Published online: 19 October 2013 # Springer-Verlag Berlin Heidelberg 2013

Abstract Nephrogenic adenoma is a benign condition of the urinary tract resulting from the displacement and seeding of renal tubular cells from the renal pelvis to the urethra. A retrospective series of 134 cases collected from four hospitals in three different countries was analyzed in this study. Recorded clinical data included age and sex, topography, urological antecedents, coexistent lesions, and follow-up. Cytonuclear and architectural features were reviewed, and PAX-8, p63, PSMA, S100A1, CEA, EMA, CD117, cannabinoid receptor CB1, AMACR, E-cadherin, and CD10 antibodies were included in an immunohistochemical panel. Males predominated (105 M/29 F) with an average age of 66 years (range, 14–96). Urothelial carcinoma was the most frequent clinical antecedent (43.2 %) and also the most common coexisting lesion (14 %). Tubular architecture was the most frequent pattern detected (40 %) although most cases showed a mixed pattern (45.5 %). Deep infiltrative growth into the muscularis propria occurred in two cases. EMA and PAX-8 were expressed in 100 % of nephrogenic adenomas, while E-cadherin reactivity was observed in 66.6 % of cases, cannabinoid receptor CB1 in 25 %, CD10 in 13.6 %, CD117 in 4.1 %, and AMACR in J. I. López (*) : A. Corominas-Cishek Department of Anatomic Pathology, Hospital Universitario Cruces, University of the Basque Country (UPV/EHU), Plaza de Cruces s/n, 48903 Barakaldo, Bizkaia, Spain e-mail: [email protected] M. Schiavo-Lena : R. Tardanico Department of Pathology, Spedali Civili, University of Brescia, Brescia, Italy A. Yagüe : R. Guarch Department of Pathology, Hospital Virgen del Camino, Pamplona, Spain K. Bauleth : O. Hes Department of Pathology, Charles University Hospital, Plzen, Czech Republic

2.7 %. For the rest of the antigens, no reactivity was found. The average time lapse between the pathological antecedent and the discovery of a nephrogenic adenoma was 32 months. We conclude that nephrogenic adenoma displays a broad spectrum of histological features that may mimic malignancy. In our experience, CB1 immunostaining adds a further argument in favor of a renal origin of this lesion. The combination of PAX-8+, p63−, and EMA+distinguishes nephrogenic adenoma from urothelial and prostate carcinoma, its most frequent malignant look-alikes. Keywords Nephrogenic adenoma . Urinary tract . Renal tubule cell . Immunohistochemistry . Differential diagnosis

Introduction Nephrogenic adenoma (NA) is a well-recognized mimicker of malignancy along the urinary tract [1–6]. It was initially considered as a special response to injury as it was found in patients with chronic inflammation or lithiasis or those who had undergone urinary tract surgery, or with other irritating conditions in the ureter, urinary bladder, prostate and urethra in adults [7–9]. However, NA has also been described at pediatric age [10, 11]. Friedman and Kuhlenbeck [12] named this lesion “nephrogenic adenoma” in one of the first descriptions of this condition, owing to its resemblance to renal tubules. This idea was subsequently supported by ultrastructural studies [13]. Finally, Mazal et al. [14] used in situ hybridization with X and Y chromosome-specific probes in male to female renal transplant recipients to prove the renal origin of NA. Since then, several series have been published contributing to the immunohistochemical profile of this entity [4, 11, 15–21], but the results are far from uniform. In this report, we describe the clinical context, histology, and immunohistochemical markers of 134 cases of NA

820

Virchows Arch (2013) 463:819–825

collected from four hospitals from three different European countries. To the best of our knowledge, this is the largest series of NA published to date.

Table 2 Clinicopathological data in 134 nephrogenic adenomas of the urinary tract

Location

Materials and methods A total of 134 NA cases were retrospectively collected from four hospitals from Spain, Italy, and the Czech Republic, encompassing a 7-year period (2006–2012). Age, sex, topography of the lesion, and type of specimen (biopsy, transurethral resection, cystectomy, nephrectomy…) were retrieved from the clinical records. When available, clinical antecedents, coexisting lesions, and follow-up data were also obtained. Formalin-fixed and paraffin-embedded material was selected for H & E and periodic acid–Schiff (PAS) routine stains and for immunohistochemical analysis. A panel of commercially available antibodies (Table 1) was tested in three different laboratories under the same conditions on a randomly chosen number of cases following standard procedures. Selected slides were immunostained in an automated immunostainer (Envision FLEX, Dako Autostainer Plus), using Tris-EDTA for antigen retrieval. Tissue present in the paraffin blocks adjacent to NA was used as an internal negative control. Immunohistochemical results were assessed by visual impression.

Results Clinical data Males predominated in the series (105 M/29 F), the average age being 66 years (range, 14–96). The urinary bladder was the most common location of NA (92 cases, 68.6 %) followed by the urethra (18 cases, 13.5 %) (Table 2). Tissue for

Table 1 Immunohistochemistry panel in nephrogenic adenomas Antibodies

Source/clone

Dilution

PAX-8 AMACR p63 EMA CEA PSMA S100A1 CD117 Cannabinoid receptor CB1 E-cadherin CD10

Cell-Marque/polyclonal Biocare/P504S Biocare/BC4A4 Dako/E29 Dako/II-7 Dako/3E6 Dako/polyclonal Dako/polyclonal ABR/polyclonal Dako/NCH-38 Dako/56C6

Ready to use Ready to use 1:150 Ready to use Ready to use 1:100 Ready to use 1:700 1:1,000 Ready to use Ready to use

Histology

Coexisting lesion

Antecedent

Number

Percent

Renal pelvis Ureter Urinary bladder Prostate

11 11 92 2

8.2 8.2 68.6 1.5

Urethra Papillary Tubular cystic Flat Mixed Urothelial carcinoma Relevant inflammation Lithiasis Stenosis Others None Not recorded Urothelial carcinoma Prostate hyperplasia Prostate adenocarcinoma Others Not recorded

18 15 54 4 61 19 14 5 3 8 45 41 58 5 4 9 58

13.5 11.2 40.3 3 45.5 14 10 4 2 6 33.5 30.5 43.2 3.7 3 6.7 43.2

histological diagnosis was obtained mainly by transurethral resection (94 cases, 70 %) and biopsy (24 cases, 18 %). Cystectomy or nephroureterectomy had been performed in 16 cases (12 %). A specific urological disease had been recorded in the clinical histories of 76 of the cases (56.7 %) with an average time lapse between antecedent and NA of 32 months. As previous antecedents, urothelial carcinoma (58 cases, 43.2 %), prostatic adenomyomatous hyperplasia (5 cases, 3.7 %), or adenocarcinoma (4 cases, 3 %) was found. Other clinical antecedents included urethral stenosis (two cases), and interstitial cystitis, urothelial dysplasia, neurogenic bladder, bladder diverticulum, and renal cyst, one case each. A history of long-term bladder catheterization had also been recorded in one case. Morphologic findings NA showed varied morphology (Fig. 1), with papillary (15 cases, 11.2 %), tubular cystic (54 cases, 40.3 %), and flat (4 cases, 3 %) growth patterns. Cases composed of mixed growth patterns were predominant (61 cases, 45.5 %). NA tended to be located at the surface and in the underlying lamina propria and was well circumscribed, sometimes multifocal, when the surgical specimens were of sufficient extent. Tubules sometimes fused to form solid cords of cells, giving the appearance of malignancy. In addition, a pseudoinfiltrative

Virchows Arch (2013) 463:819–825

821

Fig. 1 Architectural patterns in nephrogenic adenoma: a papillary, b cystic, c tubule solid, d mixed (original magnifications: ×40, ×100, and×250)

A

B

C

D

pattern resembling malignancy was regularly seen. Two cases in this series presented invasion of the urinary bladder muscularis propria. Cells in NA also had a varied appearance (Fig. 2). As a general rule, papillae and/or tubules with variable cystic transformation were covered by a single layer of cuboidal cells with eosinophilic cytoplasm and dark, small, and round nuclei. Atypical nuclear features were not observed. However, flat elements resembling endothelial cells, columnar cells, signet ring-like cells, and clear cells were also occasionally seen. Edema and inflammation were seen in the stroma, but fibromyxoid features were not observed. Eosinophilic PAS-positive material resembling Tamm–Horsfall protein was frequently present in the cystic lumina. Microtubules sometimes contained a basophilic secretion product. One case in the series presented foci of oncocytic change in some tubules, resembling renal oncocytoma. NA was seen in association with other lesions in 46 cases (34.3 %), including urothelial carcinoma (19 cases, 41.3 %) or prominent chronic inflammation (14 cases, 30.4 %). Lithiasis (five cases), urethral stenosis (three cases), bladder diverticulum (two cases), squamous metaplasia (one case), postradiation cystitis (one case), and urothelial papilloma (one case) were other NA-associated findings. Immunohistochemical findings EMA (73/73) and PAX-8 (48/48) were positive in 100 % of the tested NA (Fig. 3). EMA had a predominantly

membranous staining pattern, and PAX-8 displayed nuclear staining (Fig. 3). PAX-8 was never expressed in urothelium and prostate epithelium. E-cadherin immunostained NA in 66.6 % of cases (32/48) but was positive in urothelium and prostate epithelium (Fig. 3). CB1 (12/48, 25 %; Fig. 4), CD10 (10/73, 13.6 %), AMACR (2/73, 2.7 %), and CD117 (2/48, 4.1 %; Fig. 4) were expressed in different subsets of NA. p63, CEA, S100A1, and PSMA were consistently negative. p63 was consistently expressed in urothelium (Fig. 3). Interestingly, in NA and in urothelium, the staining pattern of PAX-8 and p63 was exactly inverse (Fig. 3).

Discussion NA is the result of the seeding of renal tubular cells along the urinary tract. Although it is most frequently found in the urinary bladder, the entire urinary tract from the renal pelvis to the urethra may be involved. Solid evidence for a renal origin of NA was first provided by Mazal et al. [14] when they proved that these lesions in renal transplant recipients were donor kidney-derived. This finding ended the widespread belief of a “special inflammatory local response to injury” as the cause of this condition [7–9]. Interestingly, years before Mazal's paper [14], some authors reported ultrastructural similarities between NA cells and renal tubular cells [13], but in the absence of irrefutable evidence, the connection between normal renal tubular cells and NA was difficult to imagine and accept.

822

Virchows Arch (2013) 463:819–825

A

B

C

D

E

F

Fig. 2 Cell morphology in nephrogenic adenoma: a cuboidal, b flat, c hobnail-like, d clear cell type, e oncocytic, f signet ring-like (original magnification, ×400)

Several authors [8, 9, 11, 22] have shown that earlier surgery, inflammatory conditions, coexisting neoplasia, lithiasis, and other injuries are frequently associated with NA. For instance, transurethral resection might eventually act as a triggering factor for exfoliated tubular cells to be attached to the local surgical bed, which are then later engulfed by the subsequent reparative process. This might explain why some rare cases of NA deep in the wall of the urinary tract display a pseudoinfiltrative pattern [6]. Urothelial carcinoma, either previous or synchronous, is most frequently associated with NA. We found a history of a resected urothelial carcinoma in 73 % of our 134 patients and coexistence of a urothelial carcinoma and NA in 13.5 %. Similarly, mucosal erosions provoked by lithiasis or a urinary catheter may predispose to the success of the seeding process. NA displays a wide variety of histological patterns, some of them simulating malignancy [3, 5, 9, 11, 23]. Tubular,

cystic, and papillary arrangements, often as a mixture, in a stroma with variable inflammation are typical. Eosinophilic (Tamm–Horsfall protein) and basophilic intraluminal secretion products are frequently detected. Some authors [24] have recently called attention to the underecognized flat pattern in NA and described pagetoid growth in the adjacent transitional mucosa, which might be related to incomplete resection and early recurrence because this mode of spreading can only be detected by microscopy. Particularly difficult to distinguish from malignancy are the (rare) infiltrative pattern of growth into the urinary tract wall [6, 23] and the presence of a spindle cell fibromyxoid stroma [25]. Cells usually have a columnar, cuboidal, flattened, signet ring, or hobnail-like appearance. However, some cases may show cells with clear cytoplasm, simulating clear cell adenocarcinoma, signet ring cell carcinoma, or prostatic adenocarcinoma [2, 3, 5, 9], notably in NA arising in the urethra. Oncocytic morphology, not previously

Fig. 3 Inverse immunoreactivity pattern of p63 and PAX-8 in urothelium and nephrogenic adenoma. While p63 is positive in urothelium and negative in nephrogenic adenoma (a), PAX-8 is positive in nephrogenic adenoma and negative in urothelium (b) (original magnification, ×250)

A

B

Virchows Arch (2013) 463:819–825

823

Fig. 4 CD117 expression in scattered cells of normal renal distal tubules (a) and in nephrogenic adenoma (b). Cannabinoid receptor CB1 is expressed in scattered cells of normal renal distal tubules (c) and in nephrogenic adenoma (d) (original magnification, ×400)

A

B

C

D

described, was detected in some tubular structures of one of our NA. Hyperchromatic large nuclei with prominent nucleoli were an occasionally observed atypical cell feature [1]. In recent years, several publications set out to define the immunohistochemical profile of NA [3, 4, 15–21, 24], but the results are contradictory. The most discriminative, consistent, and widespread marker expressed in NA is PAX-2 protein and, even better, PAX-8 protein [24]. PAX family genes are implicated in the developmental control of the kidney and other organs in vertebrates, and their mutations have been associated with various congenital diseases [5, 17, 26]. Ozcan et al. [27] reported that PAX-8 is positive in 100 % of NA, but also in malignant lymphomas and thyroid follicular-derived neoplasms, and in 90 % of Müllerian-derived and renal carcinomas. We also found PAX-8 to be positive in our NA cases. As we found PAX-8 not to be expressed in urothelium of prostatic epithelium, we consider this marker to be helpful to distinguish between NA and its most common diagnostic alternatives. PAX-2 is also a reliable marker for NA [24, 28], even in urinary cytology [29]. PAX-5 is negative [11]. p63, a well-known marker of normal and neoplastic urothelium [30], is constantly negative in NA [4, 21] which makes this antibody an ideal complement of PAX-8 to distinguish between NA and other urothelial-derived lesions. EMA, a good marker of distal renal tubular cells, is not expressed in prostatic adenocarcinoma and only in urothelial umbrella cells, but constantly positive in NA [4]. CEA is not expressed [21] but CK7 [5] is consistently positive in NA and constitutes reliable markers but does not rule out other possible diagnostic

alternatives. Similarly, E-cadherin is expressed in most NA in our series but also in urothelium. S100A1 has been recently referred to as a reliable marker of NA [18, 21], but this could not be confirmed in our cases, tested under conditions similar to those previously reported. Results with AMACR are controversial in NA [4, 11, 15, 16, 19–21]. Gupta et al. [15] obtained AMACR immunoreactivity in 58 % of 32 cases and stressed that NA is the first benign condition expressing this antigen. However, others [16, 21] have noticed that AMACR expression in NA may be related to the specific topographic location of the lesion since cases in the prostatic urethra were all positive whereas cases in the bladder were not, which makes this antibody unreliable for use in differential diagnosis. To make matters worse, Fromont et al. [19] criticized these results stating that AMACR expression in NA is an artefact, probably due to nonspecific background staining when using the avidin–biotin detection procedure. These findings have been offset by OrtizRey et al. [20] using biotin-free methods. CD10, a marker of proximal convoluted tubule cells in the kidney, has also been occasionally tested in NA [4, 11, 20]. CD10 expression has been recorded in roughly one third of cases [4, 20]. We have found focal cell surface expression of NA in 13.6 % of our cases. CD117 stains single cells in collecting ducts in the distal nephron [31]. This marker has been focally positive in two cases of our series (4.1 %) and in one case of a previous report [20]. Expression of cannabinoid receptors CB1 and CB2 has been recently tested in normal adult and fetal human kidneys

824

and in renal tumors. Results show that intercalated cells in the distal nephron and in related neoplasms, chromophobe renal cell carcinoma and renal oncocytoma, express CB1 but not CB2 [32, 33]. Interestingly, we found CB1 staining in 25 % of NA cases, adding further evidence to its renal tubule origin. Differential diagnosis of NA includes clear cell adenocarcinoma of the urinary bladder and urethra [34], urothelial carcinoma with deceptively bland features [35], and prostatic carcinoma [2]. Clear cell adenocarcinoma of the urinary tract is a rare tumor entity with a strong predilection for females and typically lacking the clinical features associated with NA. Clear cell adenocarcinomas are usually large tumors detectable on gross examination. Microscopically, they show a diffuse infiltrative growth pattern with clear cells with prominent pleomorphism, hyperchromatic enlarged nuclei, mitoses, a high rate of Ki67 expression, and extensive invasion of the muscularis, features not seen in NA [34]. Urothelial carcinomas with deceptively benign features include a nested pattern, small tubular/glandular features, a microcystic pattern, and inverted growth. Immunohistochemistry may be helpful since NA expresses PAX-8 but not p63 and urothelial carcinomas express just the opposite, as demonstrated in this study. Prostatic adenocarcinoma may mimic NA when the prostatic urethra is involved. Distinctive nephrogenic patterns, adjacent urothelium, cystic spaces with proteinaceous eosinophilic material, associated inflammation, and expression of CK7 are features useful in differential diagnosis.

Conclusions NA is a benign condition that may simulate malignancy, both cytologically and architecturally. The clinicopathological background in addition to a small panel of immunohistochemical markers, including PAX-8, p63, and EMA, will solve problematic cases and notably rule out urothelial and/or prostate adenocarcinoma. CB1 is expressed in a subset of NA. As to the etiopathogenesis of NA, given its renal tubular origin, we hypothesize that different cells from different levels of the nephron may give rise to this lesion, in view of the marker pattern detected which corresponds to various cells along the nephron.

Conflict of interest The authors declare that they have no conflict of interest.

References 1. Cheng L, Cheville JC, Sebo TJ, Eble JN, Bostwick DG (2000) Atypical nephrogenic metaplasia of the urinary tract: a precursor lesion? Cancer 88:853–861

Virchows Arch (2013) 463:819–825 2. Allan CH, Epstein JI (2001) Nephrogenic adenoma of the prostatic urethra. A mimicker of prostate adenocarcinoma. Am J Surg Pathol 25:802–808 3. Rahemtullah A, Oliva E (2006) Nephrogenic adenoma: an update on an innocuous but troublesome entity. Adv Anat Pathol 13:247–55 4. Xiao GQ, Burstein DE, Miller LK, Unger PD (2006) Nephrogenic adenoma. Immunohistochemical evaluation for its etiology and differentiation from prostatic adenocarcinoma. Arch Pathol Lab Med 130:805–810 5. Alexiev BA, LeVea CM (2012) Nephrogenic adenoma of the urinary tract: a review. Int J Surg Pathol 20:123–131 6. Diolombi M, Ross HM, Mercalli F, Sharma R, Epstein JI (2013) Nephrogenic adenoma: a report of 3 unusual cases infiltrating into perinephric adipose tissue. Am J Surg Pathol 37:532–538 7. O'Shea PA, Callaghan JF, Lawlor JB, Reddy VC (1981) “Nephrogenic adenoma”: an unusual metaplastic change of urothelium. J Urol 125: 249–252 8. Ford TF, Watson GM, Cameron KM (1985) Adenomatous metaplasia (nephrogenic adenoma) of urothelium. An analysis of 70 cases. Br J Urol 57:427–433 9. Young RH, Scully RE (1986) Nephrogenic adenoma. A report of 15 cases, review of the literature, and comparison with clear cell adenocarcinoma of the urinary tract. Am J Surg Pathol 10:268–275 10. Voss K, Peppas D (2013) Recurrent nephrogenic adenoma: a case report of resolution after treatment with antibiotics and nonsteroidal anti-inflammatory medication. Urology doi:. doi:10.1016/j.urology. 2013.04.024 11. Kao CS, Kum JB, Fan R, Grignon DJ, Eble JN, Idrees MT (2013) Nephrogenic adenomas in pediatric patients: a morphologic and immunohistochemical study of 21 cases. Pediatr Dev Pathol 16: 80–85 12. Friedman NB, Kuhlenbeck H (1950) Adenomatoid tumors of the bladder reproducing renal structures (nephrogenic adenoma). J Urol 64:657–670 13. Pierre-Louis ML, Kovi J, Jackson A, Ucci A, Pinn-Wiggins VW (1985) Nephrogenic adenoma a light and electron microscopic and immunohistochemical study. J Natl Med Assoc 77:201–205 14. Mazal PR, Schaufler R, Altenhuber-Muller R, Haitel A, Watschinger B, Kratzik C, Krupitza G, Regele H, Meisl FT, Zechner O, Kerjaschki D, Susani M (2002) Derivation of nephrogenic adenomas from renal tubular cells in kidney transplant recipients. N Engl J Med 347:653–659 15. Gupta A, Wang HL, Policarpio-Nicolas ML, Tretiakova MS, Papavero V, Pins MR, Jiang Z, Humphrey PA, Cheng L, Yang XJ (2004) Expression of alpha-methylacyl-coenzyme A racemase in nephrogenic adenoma. Am J Surg Pathol 28:1224–1229 16. Skinnider BF, Oliva E, Young RH, Amin MB (2004) Expression of alpha-methylacyl-CoA racemase (P504S) in nephrogenic adenoma: a significant immunohistochemical pitfall compounding the differential diagnosis with prostatic adenocarcinoma. Am J Surg Pathol 28: 701–705 17. Tong GX, Weeden EM, Hamele-Bena D, Huan Y, Unger P, Memeo L, O'Toole K (2008) Expression of PAX8 in nephrogenic adenoma and clear cell adenocarcinoma of the lower urinary tract: evidence of related histogenesis? Am J Surg Pathol 32:1380–1387 18. Cossu-Rocca P, Contini M, Brunelli M, Festa A, Pili F, Gobbo S, Eccher A, Mura A, Massarelli G, Martignoni G (2009) S-100A1 is a reliable marker in distinguishing nephrogenic adenoma from prostatic adenocarcinoma. Am J Surg Pathol 33:1031–1036 19. Fromont G, Barcat L, Gaudin J, Irani J (2009) Revisiting the immunophenotype of nephrogenic adenoma. Am J Surg Pathol 33: 1654–1658 20. Ortiz-Rey JA, Antón-Badiola I, Pérez-Pedrosa A, Peteiro-Cancelo A, González-Carreró J (2012) Nephrogenic adenoma: an immunohistochemical analysis using biotin-free methods. Appl Immunohistochem Mol Morphol 20:386–391

Virchows Arch (2013) 463:819–825 21. Quinones W, Ziober A, Yao Y, Bing Z (2013) Immunohistochemical markers for the differential diagnosis of nephrogenic adenomas. Ann Diagn Pathol 17:41–44 22. Gokaslan ST, Krueger JE, Albores-Saavedra J (2002) Symptomatic nephrogenic metaplasia of ureter: a morphologic and immunohistochemical study of four cases. Mod Pathol 15:765–70 23. Oliva E, Young RH (1995) Nephrogenic adenoma of the urinary tract: a review of the microscopic appearance of 80 cases with emphasis on unusual features. Mod Pathol 8:722–730 24. Piña-Oviedo S, Shen SS, Truong LD, Ayala AG, Ro JY (2013) Flat pattern of nephrogenic adenoma: previously unrecognized pattern unveiled using PAX2 and PAX8 immunohistochemistry. Mod Pathol 26:792–798 25. Hansel DE, Nadasdy T, Epstein JI (2007) Fibromyxoid nephrogenic adenoma: a newly recognized variant mimicking mucinous adenocarcinoma. Am J Surg Pathol 31:1231–1237 26. Dahl E, Koseki H, Balling R (1997) Pax genes and organogenesis. Bioessays 19:755–65 27. Ozcan A, Shen SS, Hamilton C, Anjana K, Coffey D, Krishnan B, Truong LD (2011) PAX 8 expression in non-neoplastic tissues, primary tumors, and metastatic tumors: a comprehensive immunohistochemical study. Mod Pathol 24:751–64 28. Amin W, Parwani AV (2010) Nephrogenic adenoma. Pathol Res Pract 206:659–662

825 29. Herlitz LC, Tong GX, Hamele-Bena D, Greenebaum E (2008) Nephrogenic adenoma identified on urine cytology using PAX-2 immunostaining. Diagn Cytophathol 36:47–49 30. Ho PL, Kurtova A, Chan KS (2012) Normal and neoplastic urothelial stem cells: getting to the root of the problem. Nat Rev Urol 9: 583–594 31. Kato N, Honma K, Hojo H, Sasou S, Matsuzaki O, Motoyama T (2005) KIT expression in normal and neoplastic renal tissues: immunohistochemical and molecular genetic analysis. Pathol Int 55:479–483 32. Larrinaga G, Varona A, Pérez I, Sainz B, Ugalde A, Cándenas ML, Pinto FM, Gil J, López JI (2010) Expression of cannabinoid receptors in human kidney. Histol Histopathol 25: 1133–1138 33. Larrinaga G, Sanz B, Blanco L, Cándenas ML, Pinto FM, Irazusta A, Gil J, López JI (2013) Cannabinoid CB1 receptor is expressed in chromophobe renal cell carcinoma and renal oncocytoma. Clin Biochem 46:638–641 34. Herawi M, Drew PA, Pan CC, Epstein JI (2010) Clear cell adenocarcinoma of the bladder and urethra: cases diffusely mimicking nephrogenic adenoma. Hum Pathol 41:594–601 35. Amin MB (2009) Histological variants of urothelial carcinoma: diagnostic, therapeutic and prognostic implications. Mod Pathol 22(suppl 2):S96–S118