| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Oncology, Markers, Clinical Correlates |
Divisions of Human Genetics and Molecular Biology [J. A. B., B. F., J-Y. Z.] and Oncology [A. J. J.], and Department of Pathology [L. B. R.], The Children’s Hospital of Philadelphia and Department of Pediatrics, University of Pennsylvania School of Medicine [J. A. B., A. J. J.], Philadelphia, Pennsylvania 19104; Departments of Experimental Pathology [C. D. J.] and Neurosurgery [C. R.] Mayo Clinic, Rochester, Minnesota 55905; Department of Neurology, Beth Israel Medical Center, North Division, New York, New York 10128 [J. C. A.]; and Department of Pathology, Kaplan Cancer Center, New York, New York 10016 [D. Z.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
AT/RTs of the brain are clinically aggressive malignancies that have overlapping clinical, histological, and radiographic features with MB/PNET (6) . AT/RT of the brain is almost exclusively a tumor of infants, with the majority of patients being diagnosed in the first 2 years of life. In the past, the majority of AT/RTs have been misclassified as MB/PNETs because two-thirds of AT/RTs contain fields of primitive neuroepithelial cells characteristic of MB/PNET (6 , 7) . Cytogenetic and molecular studies are a useful adjunct in the differential diagnosis of children with malignant brain tumors. An isochromosome 17q is present in approximately 30–40% of MBs and has not, to date, been observed in AT/RT (8, 9, 10) . Conversely, although monosomy or deletion of chromosome 22 is seen in the majority of rhabdoid tumors in all anatomical sites, the nonrandom association of monosomy 22 with MB/PNET is still a subject of debate (8) .
Identification of hSNF5/INI1 as a candidate tumor suppressor gene on chromosome 22 for malignant rhabdoid tumors was recently reported (11) . The hSNF5/INI1 (12 , 13) gene is a component of the mammalian SWI/SNF complex, which functions in an ATP-dependent manner to remodel chromatin, thus allowing transcription factor binding to DNA (14) . We demonstrated germ-line and somatic mutations of INI1 in CNS AT/RTs as well as in rhabdoid tumors in the kidney and soft tissue (15) . We, therefore, hypothesized that the deletion or mutation of INI1, which is associated with an aggressive and usually fatal clinical course in children with AT/RT, could also account for the poor prognosis observed in some infants with classified MB/PNET. At the same time, the specificity of INI1 mutations for rhabdoid tumors in the CNS could be determined.
In the present study, we evaluated MB/PNET biopsy samples from 52 children by FISH and molecular genetic techniques to detect deletions and to determine loss of heterozygosity in chromosome 22. Tumors were then analyzed for intragenic deletions and mutations in the INI1 gene by a combination of heteroduplex and direct sequence analysis.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Karyotypes were prepared from 10 primary tumors and have previously been published (9) . Cytogenetic pellets or touch preparations from frozen tumor tissue were analyzed by interphase FISH (10 cases) using cosmid probes for INI1 in 22q11.2 and EWS in 22q12 (15) . DNA isolated from fresh or frozen tissue was analyzed by PCR-based microsatellite analysis (49 of 52 cases) for loss of heterozygosity using probes that map to 22q according to previously reported methods (15 , 16) .
The 28 tumors analyzed at The Children’s Hospital of Philadelphia were screened by PCR and heteroduplex analysis of each of the 9 coding exons of the INI1 gene, as described previously (15) . Exons with bandshifts suggestive of a mutation were then sequenced to identify the base alteration. For the 24 Mayo Clinic cases, the 9 exons and corresponding exon/intron boundaries were directly sequenced if there was evidence of loss of heterozygosity for markers on chromosome 22 (5 tumors) or if loss of heterozygosity could not be determined (1 tumor; Ref. 11 , 16 ).
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The patients ranged in age between 4 days and 20 years, with a median
age at diagnosis of 4 years. Twenty-three (44%) of the 52 children
were 
3 years of age at diagnosis. The Children’s Hospital of
Philadelphia patients were selected from a larger cohort based on age
and available DNA samples, and, therefore, the high percentage of young
children in this study does not reflect the natural age-specific
incidence of MB/PNET.
Two children (4 days and 1 year of age) received no further therapy
after surgery and died within a year of diagnosis. Two patients (6 and
7 years of age) were treated with surgery and radiation therapy and are
alive 9 and 12 years from diagnosis. The other 48 patients were treated
with combined chemotherapy according to one of six different protocols.
Forty of the 48 patients also received radiation treatment. Ten (43%)
of 23 patients 3 years of age have died, and 10 (34%) of 29
patients >3 years of age at diagnosis have died. A multivariate
statistical analysis to assess correlations between age at diagnosis
and outcome with respect to mutation or deletion status could not be
performed because of the variability in treatment protocols.
Deletion and INI1 Mutation Analysis.
As shown in Table 1
, a total of 16 (30%)
of 52 cases demonstrated monosomy 22, deletion of 22q11.2, or loss of
heterozygosity for markers that map to 22q11-q12. None of the tumors
were homozygously deleted for the INI1 region in chromosome
22q11.2. Eight (35%) of the 23 patients who were 3 years of age
at diagnosis had loss of chromosome 22, whereas 8 (28%) of 29 patients
>3 years of age demonstrated loss of chromosome 22 in their tumors.
|
, mutations of
INI1 were identified in 4 of the 16 tumors with loss of
chromosome 22. All 4 of the mutations were found in children less than
3 years of age. Eighteen tumors without loss of chromosome 22 (11 cases
<3 years of age; 7 cases >3 years of age) were also screened for
alterations in the INI1 gene, but no additional mutations
were identified.
|
The second case was a pineoblastoma (PNET of the pineal gland)
diagnosed in a 7-month-old infant. Karyotypes demonstrated monosomy 22
as the only abnormality. A missense mutation in exon 2 in
INI1 was identified in the remaining allele. A C
T bp
change that introduces a novel stop codon (codon 40) in place of an
arginine is predicted to result in premature truncation of the protein.
Normal tissue was not available, and thus a germ-line mutation could
not be ruled out. The histology and immunophenotyping was reviewed;
however, based on the limited material available, there was no evidence
to support a change in diagnosis to AT/RT. This child also died because
of disease progression.
An identical single bp deletion of one of four cytosines in codons 382
or 383 (exon 9) was detected in cases 3 and 4. The sequence profile
from one of the tumors is shown in Fig. 1
. This mutation is predicted to result
in the addition of an extra 100 amino acids to the end of the INI1
protein (Fig. 2)
. Normal tissue was not
available in either case to determine whether the mutation was
constitutional.
|
|
The other patient with the exon 9 mutation was diagnosed at 8 months of age with a brain stem tumor. The original diagnosis was PNET, although small areas with rhabdoid cells were noted in the histological sections from the surgical sample. Microsatellite analysis of DNA isolated from the formalin-fixed and paraffin-embedded tumor tissue demonstrated loss of alleles at three markers from 22q11.2, consistent with loss of heterozygosity. The loss of chromosome 22 and subsequent identification of the exon 9 mutation in a tumor with rhabdoid cells is consistent with a diagnosis of AT/RT. The child had been treated on a Children’s Cancer Group therapeutic study (CCG-9921) with combined induction chemotherapy consisting of ifosphamide, vincristine, etoposide, and carboplatin, followed by maintenance chemotherapy and delayed radiation therapy at 36 months of age. At last follow-up, she was 6 years of age and showed no evidence of disease.
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cytogenetic and/or molecular genetic deletions of chromosome 22 were identified in 16 of 52 tumors. Coding sequence mutations in INI1 were identified in 4 of these 16 tumors. Eighteen additional tumors were screened by heteroduplex analysis and/or direct sequencing, but there were no apparent mutations. The remaining 17 tumors without loss of heterozygosity for markers on chromosome 22 were not screened for an INI1 mutation. Bi-allelic INI1 mutations could have been missed in these cases. This possibility is reduced by the fact that, to date, we have only identified one rhabdoid tumor with mutations in both alleles and no loss of chromosome 22 (Ref. 15 and unpublished data).4
In two of the four patients (cases 3 and 4 in Table 2
), the finding of
an INI1 mutation provided evidence for a change in diagnosis
from MB to AT/RT. Because of insufficient tissue for the first two
cases, additional studies that may have been used to confirm a
diagnosis of AT/RT could not be performed. Because at least two-thirds
of AT/RTs contain areas of tumor that resemble PNET (6)
,
it is not possible to exclude a diagnosis of AT/RT based on the
histological identification of such tissue in the biopsy specimen.
Expression of epithelial membrane antigen, often used to make a
diagnosis of AT/RT in this setting, could not be tested in the first
case and was negative in the second case. However, this might also be
expected if only a small sample of the tumor were available. Given that
both children were less than 1 year of age at the time of diagnosis,
the possibility of AT/RT must be considered.
The mutations identified in the present report are consistent with what others and we have previously published for hSNF5/INI1 (11 , 15 , 17 , 18) . In fact, an identical exon 9 deletion was noted in two other sporadic CNS AT/RTs in our previous series of tumors (15) . Recently, Sevenet et al. (17 , 18) screened a series of brain tumors for deletions or mutations of hSNF5/INI, including 53 MB/PNETs. Two of 17 PNETs and 5 of 36 MBs had mutations in INI1. Interestingly, mutations were also detected in four of six choroid plexus carcinomas. These authors suggest that a variety of CNS tumors, including MB, "central" PNET, and choroid plexus carcinoma share a common molecular genetic pathway in tumor development based on the finding of INI1 deletions and mutations in each of these tumor types. Historically, most AT/RTs have been misclassified as MBs, ependymomas, germ cell tumors, choroid plexus carcinomas, or supratentorial PNETs (6 , 7) . We recently reported a patient with a constitutional mutation of exon 7 who developed a CNS tumor and, 6 months later, presented with a renal rhabdoid tumor (19) . Although the original diagnosis of the brain tumor was choroid plexus carcinoma, it was reclassified as an AT/RT by histology and immunophenotyping before the identification of an INI1 mutation. Sevenet et al. (17) did not indicate whether the histology of their cases with INI1 mutations was reviewed. This might rule out the possibility that tumors with INI1 mutations identified as MB, PNET, or choroid plexus carcinoma were actually AT/RT. Regardless of the histology, however, the finding of a homozygous deletion or mutation in INI1 suggests that these tumors be considered the same disease.
Sevenet et al. (17) also noted that the tumors with INI1 mutations seemed to be clustered among the younger patients. Mutations were identified in 4 of 9 MB/PNET patients who were less than 36 months of age, compared with only 3 of 44 patients who were older than 36 months at diagnosis. In the present study, we selected additional tumors from children less than 3 years of age from our larger cohort of brain tumor patients to increase the likelihood of detecting mutations. Overall, the frequency of INI1 mutations in this series of MB/PNET patients was quite low. Not surprisingly, all of the mutations we found were in tumors from young children. In fact, three of the four children were less than 1 year of age at diagnosis. Statistical analysis of the survival data were not feasible in this study because the patients were collected over a ten-year time period, and they were treated with one of six different protocols. However, it is notable that one of the children with a mutation in exon 9 mutation was still alive six years from the date of diagnosis. Clinical correlative studies designed to compare INI1 deletion and mutation status and outcome will clearly be of interest.
Our findings also suggest that the poor prognosis associated with young
age at diagnosis for children with MB/PNET is not likely to be
associated with a deletion of chromosome 22q11 or mutation of the
INI1 gene. As shown in Table 1
, there was a higher frequency
of chromosome 22 deletions in the younger age group (35% in children
less than 3 years versus 28% in children greater than 3
years). However, if we exclude the two patients with INI1
mutations for whom the diagnosis was changed to AT/RT, then 6 (29%) of
21 children 3 years at diagnosis had loss of chromosome 22,
which is similar to what was seen in the older children. The finding
that approximately one-fourth (14 of 52, or 27%) of MB/PNETs may have
loss of heterozygosity for markers that map to 22q suggests the
presence of another tumor suppressor gene that maps to this chromosome.
Future studies will be designed to explore this possibility as novel
candidate genes from chromosome 22 are identified. At present, because
of the clinically aggressive nature of AT/RT, we suggest that newly
diagnosed infants with malignant CNS tumors be screened for loss of
chromosome 22 and mutation of the INI1 gene. This will
become critical as tumor-specific treatment protocols for AT/RT and
MB/PNET are developed.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 This work was supported by NIH Grant CA 46274
and Grant NS34514 (to J. A. B.) and was facilitated in part by
the Children’s Cancer Group, which is supported by National Cancer
Institute Grant 5U10 CA13539. 
2 To whom requests for reprints should be
addressed, at The Children’s Hospital of Philadelphia, Room 1002,
Abramson Research Building, 3516 Civic Center Boulevard, Philadelphia,
PA 19104. Phone: (215) 590-3856; Fax: (215) 590-3764; E-mail: biegel{at}mail.med.upenn.edu
3 The abbreviations used are: MB,
medulloblastoma; PNET, primitive neuroectodermal tumor; AT/RT, atypical
teratoid/rhabdoid tumor; CNS, central nervous system; FISH,
fluorescence in situ hybridization.
4 J. A. Biegel, unpublished data.
Received 2/ 4/00; revised 4/ 3/00; accepted 4/10/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. Fujita, J. Igarashi, E. R. Okawa, T. Gotoh, J. Manne, V. Kolla, J. Kim, H. Zhao, B. R. Pawel, W. B. London, et al. CHD5, a Tumor Suppressor Gene Deleted From 1p36.31 in Neuroblastomas J Natl Cancer Inst, July 2, 2008; 100(13): 940 - 949. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.P. Meyers, Z.P. Khademian, J.A. Biegel, S.H. Chuang, D.N. Korones, and R.A. Zimmerman Primary intracranial atypical teratoid/rhabdoid tumors of infancy and childhood: MRI features and patient outcomes. AJNR Am. J. Neuroradiol., May 1, 2006; 27(5): 962 - 971. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. M. Tekautz, C. E. Fuller, S. Blaney, M. Fouladi, A. Broniscer, T. E. Merchant, M. Krasin, J. Dalton, G. Hale, L. E. Kun, et al. Atypical Teratoid/Rhabdoid Tumors (ATRT): Improved Survival in Children 3 Years of Age and Older With Radiation Therapy and High-Dose Alkylator-Based Chemotherapy J. Clin. Oncol., March 1, 2005; 23(7): 1491 - 1499. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Ray, M. Ho, J. Ma, R. K. Parkes, T. G. Mainprize, S. Ueda, J. McLaughlin, E. Bouffet, J. T. Rutka, and C. E. Hawkins A Clinicobiological Model Predicting Survival in Medulloblastoma Clin. Cancer Res., November 15, 2004; 10(22): 7613 - 7620. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Hilden, S. Meerbaum, P. Burger, J. Finlay, A. Janss, B. W. Scheithauer, A. W. Walter, L. B. Rorke, and J. A. Biegel Central Nervous System Atypical Teratoid/Rhabdoid Tumor: Results of Therapy in Children Enrolled in a Registry J. Clin. Oncol., July 15, 2004; 22(14): 2877 - 2884. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. J. Guidi, T. M. Veal, S. N. Jones, and A. N. Imbalzano Transcriptional Compensation for Loss of an Allele of the Ini1 Tumor Suppressor J. Biol. Chem., February 6, 2004; 279(6): 4180 - 4185. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Roy, I. L. de la Serna, and A. N. Imbalzano The Myogenic Basic Helix-Loop-Helix Family of Transcription Factors Shows Similar Requirements for SWI/SNF Chromatin Remodeling Enzymes during Muscle Differentiation in Culture J. Biol. Chem., September 6, 2002; 277(37): 33818 - 33824. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Wild, P. Langer, A. Ramaswamy, B. Chaloupka, and D. K. Bartsch A Novel Insulinoma Tumor Suppressor Gene Locus on Chromosome 22q with Potential Prognostic Implications J. Clin. Endocrinol. Metab., December 1, 2001; 86(12): 5782 - 5787. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. W. M. Roberts, S. A. Galusha, M. E. McMenamin, C. D. M. Fletcher, and S. H. Orkin Haploinsufficiency of Snf5 (integrase interactor 1) predisposes to malignant rhabdoid tumors in mice PNAS, December 5, 2000; 97(25): 13796 - 13800. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
