Choroid Plexus Carcinomas andRhabdoid Tumors: Phenotypic andGenotypic Overlap

JOSEPHINE WYATT-ASHMEAD,1* BETTE KLEINSCHMIDT-DEMASTERS,2

GARY W. MIERAU,1 DAVID MALKIN,3 EDMUND ORSINI,1 LORIS MCGAVRAN,4 ANDNICHOLAS K. FOREMAN5

1Pathology Department, The Children’s Hospital, 1056 East 19th Avenue B120, Denver, CO 80218, USA2Neuropathology Department, University of Colorado Health Sciences Center, 4200 East 9th Avenue, Denver, CO80262, USA3Department of Pediatrics, Division of Oncology, The Hospital for Sick Children and University of Toronto, 555University Avenue, Toronto, Ontario, Canada M5G 1X84Colorado Genetics Laboratory, University of Colorado Health Sciences Center, 4200 East 9th Avenue, Denver,CO 80262, USA5Neuro-Oncology Department, The Children’s Hospital, 1056 East 19th Avenue B120, Denver, CO 80218, USA

Received December 5, 2000; accepted May 20, 2001.

ABSTRACTFive of six poorly differentiated choroid plexus carcinomasidentified at our institution contained cells displaying arhabdoid phenotype. Immunoperoxidase stains showed fo-cal positivity for cytokeratin, epithelial membrane antigen,glial fibrillary acidic protein, S100, and vimentin. TheMIB-1 proliferative index ranged from 7.0% to 27.1%. Allsix tumors were p53 positive. Only the one child withLi-Fraumeni syndrome had a p53 germline mutation. Elec-tron microscopy verified choroid plexus differentiation andthe co-existence of rhabdoid cells. Of the five studied, fourhad deletions of chromosome 22 [three with monosomy 22and one with del(22)(q12)]. Thus, there was a phenotypicand genotypic overlap between choroid plexus carcinomasand rhabdoid tumors.

Key words: choroid plexus carcinoma, rhabdoid tumor,deletion of chromosome 22, monosomy 22, p53

INTRODUCTIONChoroid plexus carcinomas, aggressive tumors of-ten affecting young children, are rare [1–6]. Cyto-

genetic studies on choroid plexus carcinomas areeven rarer [2–4]. A review of our choroid plexuscarcinomas revealed a phenotypic and genotypicoverlap between choroid plexus carcinomas andrhabdoid tumors. Ultrastructural studies verifiedthat these tumors were, indeed, choroid plexuscarcinomas and not rhabdoid tumors secondarilyinvolving choroid plexus.

METHODSOver a 13-year period (1987–2000), six choroidplexus carcinomas were accessioned at our insti-tution (6 million population base).

For light microscopic examination, tissuewas fixed in neutral buffered formalin, cut at 4 �

microns, and stained with Harris’ hematoxylin andeosin. For immunoperoxidase staining, sialinatedsections were stained by the avidin-biotin tech-nique using a variety of antibodies (Table 1) alongwith appropriate positive and negative controls.*Corresponding author

Pediatric and Developmental Pathology 4, 545–549, 2001

DOI: 10.1007/s10024-001-0085-3

Pediatric and Developmental Pathology

© 2001 Society for Pediatric Pathology

Four of the six children were studied forgermline mutations [7,8].

For ultrastructural examination, tissue sam-ples from all six tumors were fixed in glutaralde-hyde and osmium tetroxide, dehydrated in gradedalcohols, and embedded in epoxy resin. Sectionswere cut at 60 nm, stained with uranyl acetate andlead citrate, and viewed with an electron micro-scope (Zeiss EM-10).

Samples from five of the six tumors (fourprimary and one secondary) were sent for cytoge-netic studies. Independent primary in situ cultureswere established and harvested in 4–6 days [9].

RESULTSThe age at diagnosis of the six children rangedfrom 0.83 to 15 years with a mean age of 3.84years (four of the six being less than 2 years ofage). The male/female ratio was 2:4. The lateral/third/fourth ventricle ratio was 2:0:4. Except forthe child with Li-Fraumeni syndrome, none hadfamilial syndromes or other tumors. The lengthof survival after diagnosis ranged from 1 monthto 50� months with 50% surviving less than 1year. In four of the six children, the tumors hadtime to recur.

By light microscopy, all of the choroid plexuscarcinomas exhibited sheets of epitheliod cellswith little choroid plexus differentiation (papillary

formations and perivascular rosettes in which the

bases of the tumor cells were flush with the base-

ment membranes of the vessels) (Fig. 1). Rhabdoid

cells (eosinophilic “bellies” of cytoplasm and ec-

centric nuclei with large, eosinophilic nucleoli)

were seen in five of the six choroid plexus carcino-

mas, including the choroid plexus carcinoma from

the child with Li-Fraumeni syndrome (Fig. 2). The

rhabdoid cells were seen in clusters and as single

cells scattered within the sheets of the poorly dif-

ferentiated cells.

By immunoperoxidase staining (Table 1), the

tumor cells showed focal positivity for cytokeratin,

epithelial membrane antigen (EMA), glial fibrillary

acidic protein (GFAP), S100, and vimentin. The

MIB-1 proliferative index ranged from 7.0% to

27.1% with five being �13.0%. All of the choroid

plexus carcinomas were positive for p53, although

only the child with Li-Fraumini syndrome had a

p53 germline mutation.

Ultrastructural studies verified that all the tu-

mors were, indeed, choroid plexus carcinomas and

not rhabdoid tumors secondarily involving choroid

plexus. Choroid plexus differentiation was con-

firmed through the demonstration of bulbous mi-

crovilli, cilia, junctional complexes, cytoplasmic

glycogen, and basal lamina (Fig. 3A). Electron mi-

croscopy also confirmed the presence of rhabdoid

Table 1. Immunoperoxidase stains—antibodies and findings

Antibody Sourcea Clonality Dilutions Findings

CK Shandon/Lipshaw M 1:100 � (focal)

DES Immunotech/Coulter M 1:100 �

EMA Dako Corporation M 1:100 � (focal)

GFAP Dako Corporation P 1:300 � (focal)

MIB-1 Immunotech/Coulter M 1:50 7.0 to 27%

MSA ENZO Diagnostics M 1:8000 �

NSE Dako Corporation P 1:1000 �

p53 Biogenex M 1:200 � to ���

S100 Dako Corporation P 1:3000 � (focal)

SYN Biogenex M 1:50 �

VIM Immunotech/Coulter M 1:50 � (focal)

CK, cytokeratin (AE1/AE3); DES, desmin; EMA, epithelial membrane antigen; GFAP, glial fibrillary acidic protein; M, monoclonal; MSA, muscle-specific actin; NSE, neuron-specific enolase; P, polyclonal; SYN, synaptophysin; and VIM, vimentin.aSuppliers: Biogenex, San Ramon, CA; Dako Corporation, Carpinteria, CA; ENZO Diagnostics, Farmingdale, NY; Immunotech/Coulter, Westbrook, ME;Shandon/Lipshaw, Pittsburgh, PA.

546 J. WYATT-ASHMEAD ET AL.

cells with their characteristic whorls of intermedi-ate filaments (Fig. 3B).

Of the five choroid plexus carcinomas studied,four had deletions of chromosome 22 (Table 2).Three had monosomy 22 and one had a del(22)(q12).

DISCUSSIONThis is the first series showing a phenotypic andgenotypic overlap between choroid plexus carcino-mas and rhabdoid tumors. We were initially skep-tical that our six tumors were actually choroid

plexus carcinomas. We suspected that these tu-mors, especially those in the posterior fossa, hadbeen miscategorized and were actually atypicalteratoid/rhabdoid tumors secondarily involvingthe choroid plexus. Light microscopy, includingimmunohistochemical staining, was not particu-larly helpful. Electron microscopy often providedthe only reliable means of verifying choroid plexusdifferentiation and the coexistence of rhabdoid tu-mor cells. In addition to a phenotypic overlap be-tween choroid plexus carcinomas and rhabdoid

Figure 1. The choroid plexus carcinomas were oftenformed by sheets of poorly differentiated cells (A). Fo-cally, the tumor cells lined up around vessels (B). Rarely,papillary formations were found (C). H&E (A–C); �100(A), �400 (B), �200 (C).

Figure 2. There were foci of rhabdoid cells in five ofthe six choroid plexus carcinomas (A–C). The rhabdoidcells had cytoplasm stuffed with whorls of filaments andeccentric nuclei with prominent nucleoli (C). H&E (A–C);�100 (A), �400 (B), �1000 (C).

CHOROID PLEXUS CARCINOMAS/RHABDOID TUMORS 547

tumors, we found a genotypic overlap. Of the fivechoroid plexus carcinomas studied, four had dele-

tions of chromosome 22 with three having mono-somy 22 and one having del(22)(q12).

Concurrently, in a broad survey of 229 humantumors for hSNF5/INI1 somatic mutations, Sev-enet and colleagues detected an apparent geno-typic overlap between choroid plexus carcinomasand rhabdoid tumors [10]. Four of six of theirtumors carrying the file diagnosis of choroidplexus carcinoma had hSNF5/INI1 somatic muta-tions. However, since their study did not include alight microscopic or electron microscopic review,they did not document a phenotypic overlap be-tween their choroid plexus carcinomas and theirrhabdoid tumors.

From our study we were able to conclude thatthere is a phenotypic and genotypic overlap be-tween choroid plexus carcinomas and rhabdoidtumors. The finding of rhabdoid tumor cells anddeletions of chromosome 22, especially the longarm, may well equate with even more aggressivetumor behavior.

Further molecular studies of the hSNF5/INI1gene in these and other choroid plexus carcinomasverified by electron microscopy are needed to fur-ther characterize the overlap between these twohighly aggressive tumors—choroid plexus carcino-mas and rhabdoid tumors—that may share a com-mon genetic pathway.

Table 2. Karyotypes

48,X,�X,der(1)(1pter��1q32::1p32��1pter),t(3;11)(q29;q23),�8,�del(9)(q11q22),add(13)(p12),del(16)(q11.1),�18,add(20)(q13.3),del(22)(q12)[cp4]/93–99,idem[cp4]/46,XX[2]a

46,XX[2]

45,XX,�22[13]/46,XX[2]

33,XY,�1,�2,�7,�10,�14,�19,�20,�21,�mar[17]/66,XXY,�Y,�1,�2,�7,�10,�14,�19,�20,�21,�3,�4,�5,�6,�8,�9,�11,�12,�13,�15,�16,�17,�18,�22,�marx2[6]/46,XY[1]b

Not done

45,XY,�22[8]/46,XY[12]

aSecondary tumor.bLi-Fraumeni sydrome.From ISCN (1995): An International System for Human CytogeneticNomenclature. Mitelman F, ed; Basel: S. Karger, 1995.

Figure 3. Electron microscopy, in all cases, revealed fea-tures of choroid plexus differentiation, such as bulbousmicrovilli (curved arrow) and cilia (straight arrow) (A). Afilamentous “rhabdoid” inclusion (asterisk) is shown in asubjacent cell (B). �7300 (A); �3700 (B).

548 J. WYATT-ASHMEAD ET AL.

A C K N O W L E D G M E N T S

This study was presented at the Society for PediatricPathology meeting in New Orleans in Spring of 2000and published in abstract form in Modern Pathology[11]. The authors thank Dr. Robert Garcea for all hishelp and support. The authors also gratefully ac-knowledge the manuscript preparation by Ms. Gin-ger Woodward, the technical assistance of Ms. BillieCarstens in karyotyping these tumors, and the sup-port of the Children’s Hospital histology staff.

R E F E R E N C E S1. Berger C, Thiesse P, Lellouch-Tubiana A, Kalifa C, Pierre-

Kahn A, Bouffet E. Choroid plexus carcinomas in child-hood: clinical features and prognostic factors. Neurosur-gery 1998;42:470–475.

2. Dobin SM, Donner LR. Pigmented choroid plexus carci-noma: a cytogenetic and ultrastructural study. CancerGenet Cytogenet 1997;96:37–41.

3. Li YS, Fan YS, Armstrong RF. Endoreduplication andtelomeric association in a choroid plexus carcinoma. Can-cer Genet Cytogenet 1996;87:7–10.

4. Mertens F, Heim S, Mandahl N, et al. Recurrent chromo-

somal imbalances in choroid plexus tumors. Cancer GenetCytogenet 1995;80:83–84.

5. Newbould MJ, Kelsey AM, Arango JC, Ironside JW, BirchJ. The choroid plexus carcinomas of childhood: histopa-thology, immunocytochemistry and clinicopathologicalcorrelations. Histopathology 1995;26:137–143.

6. Shinoda J, Kawaguchi M, Matsuhisa T, Deguchi K, SakaiN. Choroid plexus carcinoma in infants: report of twocases and review of the literature. Acta Neurochir (Wien)1998;140:557–563.

7. Malkin D. Germline p53 mutations and heritable cancer.Annu Rev Genet 1994;28:443–465.

8. Quesnel S, Verselis S, Portwine C, et al. p53 compoundheterozygosity in a severely affected child with Li-Frau-meni syndrome. Oncogene 1999;18:3970–3978.

9. Mierau GW, Tyson RW, McGavran L, Parker NB, Parting-ton MD. Astroblastoma: ultrastructural observations on acase of high-grade type. Ultrastruct Pathol 1999;23:325–332.

10. Sevenet N, Lellouch-Tubiana A, Schofield D, et al. Spec-trum of hSNF5/INI1 somatic mutations in human cancerand genotype-phenotype correlations. Hum Mol Genet1999;8:2359–2368.

11. Wyatt-Ashmead J, Mierau GW, Foreman NK, Orsini E,McGavran L, Kleinschmidt-DeMasters BK. Choroidplexus carcinomas: electron microscopic, cytogenetic, andp53 analyses. Mod Pathol 2000;13:6P (#34).

CHOROID PLEXUS CARCINOMAS/RHABDOID TUMORS 549