Myoid differentiation in mesoblastic nephroma: Clinicopathologic and cytogenetic findings of a rare case

Myoid differentiation in mesoblastic nephroma: Clinicopathologic and cytogenetic findings of a rare case

Myoid Differentiation in Mesoblastic Nephroma: Clinicopathologic and Cytogenetic Findings of a Rare Case By Michael Guschmann, Holger To¨nnies, Christ...

195KB Sizes 0 Downloads 19 Views

Myoid Differentiation in Mesoblastic Nephroma: Clinicopathologic and Cytogenetic Findings of a Rare Case By Michael Guschmann, Holger To¨nnies, Christoph Bu¨hrer, Harald Mau, and Martin Vogel Berlin, Germany

The authors report the case of a benign renal mesenchymal tumor in a baby boy detected by ultrasound scanning during prenatal diagnosis. Histologically, the tumor was diagnosed as a congenital mesoblastic nephroma (CMN) with myoid differentiation. The tumor normally is characterized by a fascicular proliferation of bland, spindle-shaped cells. CMN is the most common renal tumor in the neonatal period and presumedly results from a neoplastic transformation affecting the pluripotent mesodermal nephric blastema. In embryonic life, tumorigenic influences acting on the nephric blastema might result in selective overgrowth of its mesoblastic derivates. CMN must be differentiated from other spindleshaped tumors, like Wilms’ tumor, rhabdoid tumor of the kidney, clear cell sarcoma, nephrogenic adenofibroma, fibroma and fibrosarcoma, leiomyoma, metanephric stromal tumor, and, in this case especially, from tumors with myoid

differentiation like infantile myofibromatosis. Numerical molecular abnormalities are observed frequently in renal mesenchymal tumors, especially in chromosome 11. Cytogenetic findings in our tumor after comparative genomic hybridization (CGH) showed full trisomies of chromosomes 20 and 22q, partial trisomies for the distal part of 11q and 1p, and an approximately full monosomy of chromosome 4 (4qter-4p15). The chromosomal imbalances of the tumor can be described as: rev ish enh(1p31pter,11q23qter,20,22), dim(4)(p15qter). J Pediatr Surg 37:E22. Copyright 2002, Elsevier Science (USA). All rights reserved.

T

evaluate the interrelationships of histology and chromosome aberrations. In this case report of an mesoblastic nephroma with myoid differentiation, we identified an unusual genetic finding, which, to our knowledge, had not been reported previously in mesoblastic nephroma. The newborn, a white male, was born at 38 weeks’ gestation, weighing 2,960 g and measured 51 cm in length. He was the first child of a 31-year-old primigravida. In the 34th week of pregnancy ultrasound examination diagnosed a renal tumor. A follow-up scan at 36th and 38th week were done, and a round hypoechoic tumor measuring 2.3 ⫻ 3.2 ⫻ 2.5 cm was seen (Fig 1). A computer tomographic scan of the tumor showed a hypodense lesion in the anterior part of the right kidney. No other anomalies were seen. A complete tumor resection was performed. The postoperative course was unremark-

HE MESOBLASTIC NEPHROMA is the most common renal neoplasm in the neonatal period. About 85% of this mesenchymal tumor occurs in neonates and infants.1,2 It was first described in 1967 by Bolande et al,3 as a tumor characterized histologically by a fascicular proliferation of spindle shaped cells. Until now, the microscopic appearence ranged from a bland, fibromatosis pattern to a much more cellular lesion sometimes with mitosis.4-6 We saw in our case a histologic mixed tumor variant with scattered to confluent nodules and bundles of eosinophilic spindle cells like those seen in infantile myofibromatosis or dermatofibromasarcoma protuberans. Because of the tremendous morphologic overlap that exists among other pediatric renal tumors such as metanephric stroma tumor, Wilms’ tumor or of infantile myofibromatosis, and the diagnostic dilemma thereby, we describe the differential diagnosis of mesoblastic nephroma and other renal spindle cell tumors. Efforts to understand the biology of mesoblastic nephroma have more recently expanded to include analysis of overall DNA content of renal tumors. Recent studies have found that several of the renal tumors have molecular alterations. Reports of karyotype findings, particularly gains like trisomy or tetrasomy of chromosome 11 have been published.4-8 We performed comparative genomic hybridization (CGH) with DNA extracted from paraffin-embedded tumor material to Journal of Pediatric Surgery, Vol 37, No 8 (August), 2002: E22

INDEX WORDS: Mesoblastic nephroma, morphology, congenital tumor, myoid differentiation, prenatal diagnosis, comparative genomic hybridization.

From Abteilung fu¨r Paidopathologie und Plazentologie, Institut fu¨r Humangenetik, Molekulare Zytogenetik, Klinik fu¨r Neonatologie, und Klinik fu¨r Kinderchirurgie, Charite´, Campus-Virchow Klinikum, Humboldt-Universita¨t zu Berlin; Berlin, Germany. Address reprint requests to Dr med. Michael Guschmann, Abteilung Paidopathologie und Placentologie, Institut fu¨r Pathologie, Charite´, Campus Virchow-Klinikum, Medizinische Fakulta¨t der Humboldt-Universita¨t zu Berlin, Augustenburger Platz 1, 13353 Berlin, Germany. Copyright 2002, Elsevier Science (USA). All rights reserved. 1531-5037/02/3708-0037$35.00/0 doi:10.1053/jpsu.2002.34498 1

2

GUSCHMANN ET AL

evaluated using an epifluorescence microscope (Axiophot; ZEISS, Germany) fitted with specific single band pass filter sets for DAPI, SpectrumGreen and SpectrumOrange. Image analysis and karyotyping (CGH) was performed using the ISIS analysis system (METASYSTEMS, Germany). Diagnostic thresholds used for the identification of chromosomal underrepresentations (deletions) and overrepresentations (duplications) were 0.75 and 1.25.

RESULTS

Pathologic Findings and Immunohistochemistry

Fig 1. Sonographic cross section of fetal kidney, 38th gestational week with a round hypoechoic tumor.

able. A cousin was diagnosed as having a Wilms’ tumor at birth. MATERIALS AND METHODS Macroscopically, the lesion was 3.3 ⫻ 2.2 ⫻ 1.3 cm. The cut surface was homogeneous, white/gray (Fig 2). The excised specimen was fixed in 5% buffered formalin, and representative blocks were processed and embedded in paraffin using standard methods. Sections measuring 4 ␮m were stained with H&E, periodic-acid-schiff (PAS), and masson trichrome stain. The sections were evaluated immunohistochemically using the APAAP technique. Primary antisera included mouse monoclonal antibodies against alpha smooth muscle-actin, desmin, vimentin, MIB-1, and CD 31 from DAKO (Copenhagen, Denmark).

DNA Extraction No fresh tumor tissue was available for cytogenetic analysis. Therefore, DNA from 6 4-␮m-thick, paraffin-embedded sections was isolated using a commercial kit (Invisorb Spin Tissue Kit, Invisorb, Germany) according to the manufacturer’s recommendations. The isolation technique is based on the binding of genomic DNA to a carrier matrix in the presence of chaotropic salts.9

Comparative Genomic Hybridization Extracted test DNA from the tumor specimen was labelled by DOP-PCR incorporating SpectrumGreen-dUTP (VYSIS) for green fluorescence. High-molecular-weight normal male reference DNA was labelled by nick translation with SpectrumOrange-dUTP (VYSIS). The DNA fragment size of both test and reference DNA after labelling was compared on an agarose gel (1,5%). Both DNAs showed the same size range of 100 to 1,000 base pair. Metaphase spreads were prepared from PHA-stimulated peripheral blood lymphocytes from a karyotypically normal male and pretreated by incubation with RNase (stock solution: 20 mg/ml RNase A, 10 mmol/L Tris-HCI pH 7.5, 15 mmol/L NaCl) before hybridization. Ethanol-stored chromosome preparations were equilibrated in 2 ⫻ SSC at room temperature and incubated with RNase solution (100 ␮g/mL) at 37°C for 1 hour. For each hybridization, 200 ng of labelled test DNA, 200 ng reference DNA, and 12.5 ␮g Cot-1 DNA were coprecipitated, resuspended in 14 ␮L hybridization mix containing 50% formamide, 2 ⫻ SSC and 10% dextran sulphate, denatured at 70°C for 5 minutes, and hybridized to denatured metaphase spreads. Slides were incubated at 37°C for 2 days. After standard posthybridization washes, images of the hybridized metaphases were

In general, the tumor is composed of interlacing plump connective tissue cells. The tumor cells were elongated and spindle shaped. The nuclei were focal cigar shaped. Areas of storiform and herringbone patterns were observed. The cytoplasm of these cells was fibrillar, eosinophilic, and partially slightly PAS-positive. In addition, some areas showed an admixture of cells ranging from small ovoid cells to large spindle cells. Here, nuclear pleomorphism and solitary mitoses were observed. A special feature in our case was the presence of scattered bundles and small nodules (Fig 3) composed of bland spindle with well-defined cytoplasmatic margins and vesicular tapering nuclei with inconspicuous nucleoli, closely resembling smooth muscle cells or myofibroblasts. These areas reprensented about 20% of the tumor tissue. Numerous vascular slits and spaces of variable size were seen (Fig 4). Immunohistochemical analysis showed diffuse positivity for vimentin und smooth muscle-actin and vimentin and negativity for desmin. Comparative Genomic Hybridization CGH was applied to DNA of paraffin sections to detect chromosomal imbalances in the tumor. The average ratio profile of 10 metaphases showed different chromosomal imbalances (Fig 5). Ratio profile deviations after CGH showed full trisomies of chromosomes 20 and 22q, partial trisomies for the distal part of 11q and 1p, and an approximately full monosomy of chromosome 4 (4qter-4p15).

Fig 2. Cut surface of a congenital mesoblastic nephroma showed homogen white/gray to yellow.

MYOID DIFFERENTIATION IN MESOBLASTIC NEPHROMA

Fig 3. Mesoblastic nephroma, histologically with a cellular fascicular proliferation of bland, spindle-shaped cells and with myoid nodules and vascular spaces of variable size. (H&E, original magnification ⴛ100.)

There was no evidence of further deletions or amplifications. Because of known methodic problems, centromeric and heterochromatic regions, p-arms of acrocentric chromosomes, and telomeric regions were excluded from evaluation.10 Summarizing the results of CGH analysis the chromosome copy number determination of the tumor can be described as: rev ish enh(1p31pter,11q23qter,20,22), dim(4)(p15qter). DISCUSSION

Congenital mesoblastic nephroma (CMN) is the most common renal neoplasm in neonates and represents 80% of solid renal masses in the neonatal period.6,11 The tumor is more common than Wilms’ tumor in newborn infants less than 3 months of age.8,12 The majority of patients are subsequently free of recurrence or metastasis.12-14 Recognized as a distinct entity in 1967, it was described by Bolande3 as a mesenchymal tumor characterized histologically by a fascicular proliferation of bland, spindle-shaped cells. After the first histologic description it was recognized that the microscopic appearance ranged from the bland fibromatosis pattern initially described to a much more cellular, less-differentiated lesion.3,6,15 The mixed histologic character of the tumor indicates a histogenetic relationship to Wilms’ tumor, although the malignant nephroblastic or epithelial constituency typical of Wilms’ tumor is inconspicuous or totally lacking. Currently, the pathogenetic relationship between CMN and Wilms’ tumor is unclear, although both tumors originate from embryonal blastema.3,16 CMN may present in utero with nonimmune fetal hydrops and polyhydramnios. Hypertension has been reported as a sign, especially in children with cellular mesoblastic nephromas. Renin synthesis appears to be

3

localized in the entrapped renal parenchyma within the tumor.17 The tumor usually is discovered as a palpable abdominal mass or, rarely, present with anemia or hematuria.11,12 There have been only 2 cases reports involving hemorrhage and shock.18,19 Local invasion into the perinephric tissues may be present in all types of CMN, but nephrectomy is curative.1 CMN is unilateral and often large, 5 to 10 cm in diameter. On a cut surface the tumor is glistening, whorled, or finely trabeculated and gray-white to yellow in color. Normally, the border between the tumor and the uninvolved kidney is without a capsule, and in histologic sections tumor tissue is mixed with kidney parenchyma. CMN is characterized by fingerlike projections extending into the adjacent kidney tissue.1 Microscopically, the bulk of the CMN is composed of interlacing sheets of connecting tissue with fibroblasts myofibroblasts. Smooth muscle elements may be demonstrated rarely. A small number of atrophic kidney tubules and glomeruli are embedded within the tumor. Cellular variants may contain foci of cystic, dysplastic, or immature renal tubules within its mesoblastic substance. It is this mixed character that suggests an origin in the embryonal nephric blastema much like Wilms’ tumor.2,20-22 CMN are classified as either classic, cellular, or mixed histologic variants. Classic CMNs are characterized by relatively hypocellular interlacing bundles of bland spindle cells. Neither pleomorphism or mitosis are observed. Cellular CMNs, in addition to being more cellular, frequently have mitosis and necrosis.23,24 Nuclear CMNs range from round to small and spindle shaped, occasionally giving the lesion a sarcomatous appearance. When both classic and cellular patterns are present, as in our case, the tumor is considered mixed.6 An alternative histogenetic viewpoint is that the tumor is fundamentally a fibromatous process.3 The CMN is

Fig 4. Mesoblastic nephroma with fascicular proliferation of bland, spindle-shaped cells and numerous vascular slits and spaces. Left side shows a myoid nodule. (H&E, original magnification ⴛ200.)

4

GUSCHMANN ET AL

Fig 5. Averaged (A) and single (B) ratio profiles of the chromosomal imbalances detected in the tumor genome by comparative genomic hybridization after DOP-PCR. Full trisomies for chromosome 20 and 22q, partial trisomies for the distal part of 11q and 1p, and an approximately full monosomy of chromosome 4 (4qter-4p15) are indicated by ratio profile deviations reaching the diagnostic thresholds of 0.75 and 1.25, respectively.

readily distinguishable from the medullary or capsular fibromas and leiomyomas occuring in adults. The latter tumors are typically small and discrete, whereas their histologic composition is uniformly fibromatous or myomatous.3 Sometimes the cellular variant of CMN is histologically the renal counterpart of the congenital fibrosarcoma, which tends to occur in the soft tissue in the extremities and axial region. This tumor occurs almost exclusively in children younger than 2 years and is described mainly in newborns.7,25 The nephrogenic adenofibroma exhibits a predominant and benign-appearing mesenchymal proliferation surrounding multifocal nodules of blastema differentiating toward glandular or papillary structures. This tumor is confused easily with CMN, but presents at later age (mean 13.3 years), and may have multiple nephrogenic rests (nephroblastomatosis).26 The classical variant of CMN is morphologically identical to infantile fibromatosis and affects the same age group. This suggests that CMN represents infantile fibromatosis of the kidney as a distinct entity.4,5 The metanephric stromal tumor is a primary pediatric renal neoplasm with a fibrous lesion centered in the renal medulla. In contrast to CMN the metanephric stromal tumor includes typically onion-skin cuffing around entrapped renal tubules and angiodysplasia.4,5 The clear cell sarcoma and the malignant rhabdoid tumor of the kidney occur chiefly in children younger than 2 years. In contrast to CMN, clear cell sarcoma contains a high content of vessels and consists of cells with vacuolized cytoplasm. Malignant rhabdoid tumor possesses cells with large roundish nuclei, pale chromatin, and very prominent nucleoli.1

In our case, we saw myofibroblastic differentiation with scattered to confluent nodules and bundles of spindle cells associated with well defined cytoplasmic margins. The myoid areas were sm-actin positive and desmin negative. Our tumor, like infantile myofibromatosis, dermatofibrosarcoma protuberans, and other myofibroblastic lesions had a myofibroblastic differentiation, which gives support to the theory of a fibroblastic/myofibroblastic line of differentiation for this type of mesonephroblastic tumors.27,28 CMN has mixture of mesenchymal tissue containing foci of cystic, dysplastic renal tubules within its mesoblastic substance. It is this mixed character that suggests an origin in the embryonal nephric blastema, much like Wilms’ tumor. The bulk of the tumor is composed of fibroblasts or myoblasts. Although fibroblasts may be predominant, as in our case, smooth muscle elements may be present. The pluripotentiality of this mesoblastic tissue is shown clearly by such different tissue.3 An alternative histogenetic viewpoint is that the tumor is fundamentally a fibromatous process. But the immature nature of many of the tubules plus their presence in sites of extrarenal invasion indicate that they are an intrinsic part of the neoplastic process. Because these features indicate an origin of the tumor in embryonal mesoderm, specifically the nephric blastema, the designation congenital mesoblastic nephroma is preferred. CMN, like Wilms’ tumor results from a neoplastic transformation affecting the pluripotent mesodermal nephric blastema. This mesodermal tissue is capable of differentiating into glomeruli and tubules, as well as a variety of connective tissues. In embryonic life, influences acting on the nephric blastema might result in the selective

MYOID DIFFERENTIATION IN MESOBLASTIC NEPHROMA

5

overgrowth of its mesoblastic or mesenchymal derivates, whereas epithelial derivates might be affected only minimally. The mesoblastic component may be unaffected or stimulated to greater proliferation in this process.3,8,29 It has been proposed that there is a close relationship between fibroblasts and myofibroblasts and that former might undergo isoformic transitions into the latter in relation to functional demand. This is illustrated in the reparative response in wounds, but also can occur in neoplasms as a reactive stromal response or in neoplastic cells per se, as happens in infantile myofibromatosis, fibromatoses, fibrosarcomas, or dermatofibrosarcomas.30 So tumors that show fibroblastic differentiation also may be capable of differentiating toward myofibroblasts, although myoid differentiation in CMN is a very infrequent phenomenon.28 Our finding of areas of myoid differentiation in an otherwise ordinary congenital mesoblastic nephroma gives further support to the fibroblastic/myofibroblastic line of differentiation for the pluripotent mesodermal nephric blastema. Recognition of this described morphologic variant of CMN is not only important with regard to histogenesis but also because it can be confused with myofibroblastic lesions, especially when small biopsy sections are evaluated. The focal biphasic appearance with myoid nodules and fascicles, combined with more cellular areas typical of CMN, might be confused on low power with an myofibromatosis lesion.27,28 The discoveries of chromosomal aberrations associated with pediatric renal tumors represent major advances toward understanding the molecular pathogenesis of these lesions.4,5 Recent studies have found that several renal stromal tumors like CMNs or Wilms’ tumors have molecular alterations identical to those of other extrarenal tumors and, hence, are not kidney specific. The cellular variant of CMN has been shown to bear the translocation t(12; 15)(p13;q25) characteristic for infantile fibrosarcoma.4,5 Bourgoeis et al7 and Argani et al4,5 recently have identified a novel ETV6-NTKR3 fusion gene resulting from this rearrangement. The classical variant of CMN did not

contain the fusion transcript. Argani4,5 believe that the mixed type of CMN represents more than one genetic entity, because some mixed CMNs that had both classic and cellular areas show the fusion transcript, whereas some did not. There are reports on DNA aneuploidy and gains of chromosome 11 and intratumor heterogenity in mesoblastic nephroma.6 Schofield described trisomy or tetrasomy for D11Z1 (centromere of chromosome 11) only in CMNs with mixed histology.6 In contrast, chromosome 11 aberrations were not found in classic histology CMNs. Speleman et al8 described trisomy 11 in mixed and cellular CMN. These reports suggest an oncogenetic basis for histologic variability in CMNs Trisomie 11.6 Trisomy 11 has been reported in mesenchymal breast tumors, infantile fibrosarcomas, and benign chondroid tumors. These observations suggest that trisomy 11 might convey a proliferative advantage to several mesenchymal cell types.6 So abnormal proliferation in some or all of these tumors might result from increased copy number of one or more genes on chromosome 11.6 Additional aneuploidies in cellular or mixed histology CMNs include extra copies of chromosoms 8 and 17.6 The genetic analysis in our case included a trisomy of chromosome 20 and 22q, a trisomy for the distal part of 11q and 1p, and an approximately full monosomy of chromosome 4 (4qter-4p15). These findings suggest that different chromosomal imbalances could be the genetic background of the tumor appearance. We have shown a new variant of CMN with myoid differentiation, which gives further support to the fibroblastic/myofibroblastic line of differentiation for the pluripotent mesodermal nephric blastema. This focal biphasic appearance with myoid nodules and fascicles, combined with more different cellular areas of CMN, might be confused on differential diagnosis with other renal stromal tumors and especially with myofibromatosis lesion like infantile myofibromatosis. We identified an unusual genetic finding of the tumor that had not been reported previously in mesoblastic nephroma.

REFERENCES 1. Harms D, Schmidt D, Leuschner I: Neue Aspekte der Nephroblastome (Wilms-Tumoren) und anderer metanephrogener Neoplasmen. Verh Dtsch Ges Path 73:350-371, 1989 2. Kleist B, Lorenz G, Dietrich H, et al: Mesoblatisches Nephrom des Erwachsenenalters. Pathologe 19:226-229, 1998 3. Bolande RP, Brough AJ, Izant RJ: Congenital mesoblastic nephroma of infancy: A report of eight cases and the relationship to Wilms’ tumor. Pediatrics 40:272-278, 1967 4. Argani P, Fritsch M, Kaskol SS, et al: Detection of the ETV6-NTRK3 chimeric RNA of infantile fibrosarcoma/cellular congenital mesoblastic nephroma in paraffin-embedded tissue: application to challenging pediatric renal stromal tumors. Mod Pathol 13:29-36, 2000

5. Argani P, Beckwith B: Metanephric Stromal Tumor. Am J Surg Pathol 24: 917-926, 2000 6. Schofield DE, Yunis EJ, Fletcher JA: Chromosome aberrations in mesoblastic nephroma. Am J Pathol 143:714-724, 1993 7. Bourgeois JM, Knezevich SR, Mathers JA, et al: Molecular detection of the ETV6-NTRK3 gene fusion differentiates congenital fibrosarcoma from other childhood spindle cell tumors. Am J Surg Pathol 24:937-946, 2000 8. Speleman F, van den Berg E, Dhooge C, et al: Cytogenetic and molekular analysis of cellular atypical mesoblastic nephroma. Genes Chrom Cancer 21:265-269, 1998 9. Vogelstein B, Gillespie BE: Preparative and analytical purification of DNA from agarose. Proc Natl Acad Sci U S A 76:615-619, 1979

6

10. Kallioniemi OP, Kallioniemi A, Piper J, et al: Optimizing comparative genomic hybridization for analysis of DNA sequence copy number changes in solid tumors. Genes Chrom Cancer 10: 231-243, 1994 11. Willert JR, Feusner J, Beckwith JB: Congenital mesoblastic nephroma: a rare cause of perinatal anemia. J Pediatr 134:248, 1999 12. Howell CG, Othersen HB, Kiviat P: Therapy and outcome in 51 children with mesoblastic nephroma: A report of the National Wilms’ Tumor Study. 17:826-831, 1982 13. Hrabovsky E, Othersen HB, deLorimier A, et al: Wilms’ tumor in the neonate: A report from the national Wilms’ Tumor Study. J Pediatr Surg 21:385-387, 1986 14. Trillo AA: Adult variant of congenital mesoblastic nephroma. Arch Pathol Lab Med 114:533-535, 1980 15. Kumar S, Marsden HB, Carr T, et al: Mesoblastic nephroma contains fibronectin but lacks laminin. J Clin Pathol 38:507-511, 1985 16. Stambolis C: Benigne und potentiell maligne metanephrogene Neoplasmen: Morphologie, Diagnose und klinische Bedeutung in Doerr W, Leonhardt H (Hrsg): Normale und Pathologische Anatomie, (1984); Bd 50. Thieme, Stuttgart New York, Seite 6-42 17. Malone PS, Duffy PG, Ransley RG, et al: Congenital mesoblastic nephroma, renin production, and hypertension. J Pediatr Surg 24:599, 1989 18. Goldberg J, Liu P, Smith C: Congenital mesoblastic nephroma presenting with hemoperitoneum and shock. Pediatr Radiol 24:54-55, 1994 19. Zach TL, Cifluente RF, Strom RL: Congenital mesoblastic nephroma, hemorrhagic shock and disseminated intravascular coagulation in a newborn infant. Am J Perinatol 8:203-205, 1991 20. Durham JR, Bostwick DG, Farrow GM, et al: Mesoblastic nephroma of adulthood: Report of three cases. Am J Surg Pathol 17:1029-1038, 1993

GUSCHMANN ET AL

21. Ganick DJ, Gilbert EF, Beckwith JB, et al: Congenital cystic mesoblastic nephroblastoma. Human Pathol 12:1039-1043, 1981 22. Schmidt D: Nephroblastome (Wilms-Tumoren) und Nephroblastom Sondervarianten. Pathologie, Klassifikation, Differentialdiagnose, In Dhom G, Eder M, Fischer R, et al (Hrsg): Vero¨ffentlichungen aus der Pathologie, Bd 133. Fischer, Stuttgart New York, (1989); Seite 116-121 23. Pettinato G, Manivel JC, Wick MR, et al: Classical and cellular (atypical) congenital mesoblastic nephroma: A clinicopathologic, ultrastructural, immunohistochemical, and flow cytometric study. Human Pathol 20:682-690, 1989 24. Joshi VV, Kasznica J, Walters TR: Atypical mesoblastic nephroma: Pathologic characterization of a potentially agressive variant of conventional congenital mesoblastic nephroma. Arch Pathol Lab Med 110:100-106, 1986 25. Reddemann H, Lorenz G, Rehberg U: Konnatales mesoblastis¨ bersicht zur Rezidivha¨ufigkeit und ches Nephrom im Sa¨uglingsalter—U zu Therapiergebnissen von 74 Fa¨llen. Klin Pa¨diatr 192:509-516, 1980 26. Hennigar RA, Beckwith JB: Nephrogenic Adenofibroma. Am J Surg Pathol 16:325-334, 1992 27. Beckwith JB: Wilms tumor and other renal tumors of childhood: An update. J Urol 136:320-324, 1986 28. Calonje E, Fletcher CDM: Myoid dfferentiation in dermatofibrosarcoma protuberans and ist fibrosarcomatous variant: clinicopathologic analysis of 5 cases. J Cutan Pathol 23:30-36, 1996 29. Snyder HM, Lack EE, Chetty-Baktavizian A, et al: Congenital mesoblastic nephroma: Relationship to other renal tumors in infancy. J Urol 126:513-516, 1981 30. Katenkamp D, Stiller D in Weichgewebstumoren. Johann Ambrosius Barth Verlag, Leipzig (1990); Seite 43-46